Question 1

Standard grinding equipment lead time is 1-2 months.

Accepted Answer

For standard specification products, our typical lead time is within 1–2 months from order confirmation. These items are manufactured based on established designs and tested processes, allowing us to ensure consistent quality and timely delivery.However, for customized products—such as those requiring integration with specific rail systems or tailored performance parameters—the lead time varies significantly. Custom solutions often involve detailed technical consultations, design iterations, compatibility assessments, and multiple rounds of repetitive testing to validate performance under real-world operating conditions. This rigorous development cycle ensures that the final product meets both safety standards and the unique operational requirements of each client’s infrastructure (e.g., track gauge, signaling systems, load capacity, environmental conditions, etc.). As a result, delivery timelines for bespoke equipment are determined on a case-by-case basis and typically require close collaboration between our engineering team and the customer throughout the project lifecycle.We always strive to provide an accurate and realistic schedule during the initial quotation phase, and we keep clients informed of progress at every stage to ensure transparency and alignment.

Question 2

We fully support standard Chinese rail profiles (50/60/75 kg/m), covering speeds from 120-350 km/h.

Accepted Answer

Our rail grinding solutions are engineered to serve a comprehensive range of railway systems, including high-speed rail networks, heavy-haul freight lines, urban mass transit (such as subways and light rail), commuter rail systems, and industrial or mining railways. Each application presents unique challenges—high-speed lines demand ultra-smooth profiles for passenger comfort and safety; heavy-haul routes require aggressive defect removal due to intense wheel-rail contact; while metro systems prioritize noise reduction and precise profile control in confined tunnels. Our equipment portfolio includes both on-track grinding trains and smaller, modular units that can access tight curves or depot tracks. Through adaptable grinding patterns, variable power settings, and interchangeable grinding modules, we ensure optimal performance across diverse track gauges (standard, narrow, or broad), rail profiles (e.g., UIC, AREMA, GB standards), and operational environments—from arid deserts to humid coastal zones. We continuously collaborate with global operators to refine our technology for local conditions, ensuring maximum efficiency, longevity, and compliance with regional rail safety regulations.

Question 3

We provide full-cycle service from training to long-term support.

Accepted Answer

Yes, we offer comprehensive post-purchase support that begins before delivery and extends throughout the equipment’s operational life. Our operator training program includes both theoretical instruction and hands-on practice, covering machine operation, safety protocols, routine maintenance, troubleshooting, and data interpretation from onboard monitoring systems. Training can be conducted at our factory, at the customer’s facility, or virtually—tailored to the client’s schedule and language requirements. Beyond initial training, we provide ongoing technical support through a dedicated global service team available 24/7 for remote diagnostics and emergency assistance. We also offer scheduled maintenance planning, software updates, spare parts logistics, and periodic on-site performance reviews. For long-term partnerships, we can establish localized service centers or train in-house technical teams to enhance self-sufficiency. Our goal is not just to deliver equipment, but to ensure our customers achieve maximum uptime, operational efficiency, and return on investment over the entire asset lifecycle.

Question 4

We employ a data-driven, customized approach instead of fixed intervals.

Accepted Answer

Rail grinding cycles are not fixed but are instead determined by a dynamic set of interrelated factors, making a customized, condition-based approach essential for optimal track maintenance. Key determinants include traffic volume and axle load (higher intensity accelerates rail wear), train speed (affecting contact stress and fatigue development), environmental conditions (moisture, temperature extremes, and contaminants like sand or leaves influence corrosion and adhesion), rail material properties, track geometry (sharp curves experience asymmetric wear), and historical defect patterns (such as corrugation, shelling, or rolling contact fatigue). Modern rail asset management systems integrate data from track inspection vehicles, ultrasonic testing, and grinding history to generate predictive maintenance plans. Rather than adhering to a rigid calendar-based schedule, we recommend a performance-driven strategy: grinding is performed when predefined thresholds for surface roughness, profile deviation, or crack depth are exceeded. This adaptive method maximizes rail life, reduces lifecycle costs, and enhances safety. We work closely with clients to develop tailored grinding programs aligned with their network priorities, budget constraints, and regulatory requirements—ensuring every pass delivers measurable value.

Question 5

Rank by time to the threshold, weighted by traffic and gated by consequence. Severity is what has accumulated; priority is what will accrue before the next window can open.

Accepted Answer

How do you prioritise which track sections get ground first?Rank by time to the threshold, not by size. For each candidate section, take the distance the defect still has to travel before it has to be ground, and divide it by the rate at which that defect is measurably developing at the points that will fail first; the smallest quotient goes first. Then let consequence override the arithmetic in the few places where being late is not repaid by an extra pass — turnout areas and the tightest curves, where the fallback is a speed restriction or a rail change rather than another visit. Kilometres per section, and "worst condition first", both produce a queue that is easy to defend in a meeting and wrong on the track.A queue is a statement about what will happen next, not about what has already happenedSeverity is a stock, priority is a flowDepth and amplitude describe what has accumulated. The rate describes what will accumulate before the next window can be opened. Those are different quantities, and only the second one belongs in a queue. Two sections can read identically on the same afternoon and still be different jobs: one has been stable at that reading for several cycles, the other reached it since the last visit. Sorting by condition ranks the stock, and the window being allocated is about the flow. This is the single most common reason a well-run programme still arrives late at a section it had already measured and already ranked.The rate also has to be measured rather than assumed. Measuring it means the same fixed points compared across two visits — the technique itself is a separate subject, covered in how rail wear is measured and when that decides grinding or replacement, and this answer does not restate it. Traffic is the driver of the rate, not evidence of it. It enters the ranking as a weight on sections whose rate you have, never as a substitute for sections whose rate you have not.DimensionWhat it has to describeWhere the number comes fromThe proxy usually substituted insteadWhat the proxy does to the orderDevelopment rateHow fast the defect is growing at the points that will reach the limit firstThe same fixed points, compared across two visits, in the unit your interval is written in"How bad it looks now"Puts the deepest but stable section at the front, and leaves the shallow fast-developing one for the visit after the one it will fail beforeTrafficThe loading that drives the rate on that sectionTonnage where duty is heavy; wheel passes and axle load where it is notNetwork-wide annual tonnage applied to every sectionEvery section inherits one average, so a lightly used siding and a loaded curve move up and down the list togetherConsequenceWhat being late actually costs at that locationThe fallback for that defect there — another pass, a speed restriction, a rail changeLength of the section, or defect depthThe longest section outranks the turnout carrying the same traffic, and the cheap fallbacks crowd out the expensive oneConsequence is deliberately not a score to be added to the other two. It is a gate. Arithmetic decides the order of ordinary sections; consequence decides which sections may not be overtaken by arithmetic at all. Mixing the two into one weighted total is what produces a ranking that nobody can explain to the engineer who has to work it.Some sections are ranked by how fast they change, others by what happens if they are missedWhere the queue normally settles, and whyTight curves and turnout areas come out near the front of almost every honest ranking, and they get there for two different reasons — one on rate, one on consequence. On the unballasted side of the network, curves below about 1,000 m radius are where side wear and corrugation concentrate, because the gauge corner is doing work the rest of the head is not. In a single curve of 800 m radius ground in six passes, the steel taken at the maximum was 1.8 mm on one side of the head against 0.3 mm on the other. That asymmetry has a direct consequence for how the list is built: ranking at the level of the section hides the fact that one side of that section is the whole problem. The unit of ranking is section, rail and defect type — not the section alone.Turnouts sort to the front on the second dimension instead. The contact is concentrated onto a fixed geometry with very little room to move, the tolerance on where the contact band should sit is tight — the turnout grinding reference puts it at 25–30 mm for high-speed work and 35–40 mm for conventional and heavy-haul work — and the possession is measured in minutes rather than hours. A defect there is not usually cosmetic, it reads as dynamic response, and the intervention window is short enough that being a cycle late can remove the possibility of doing it by grinding at all.Section classWhat drives the rate thereWhat being late costsWhere it usually sortsCurves below roughly 1,000 m radiusSide wear and corrugation concentrate on the gauge corner; the rate is geometry-driven and largely independent of how the section looks from a distanceThe fix migrates from profile work to profile restoration, and eventually to replacement of one rail in the curveFront, on rateTurnout areasContact is forced onto a fixed geometry with a narrow acceptable contact bandDynamic response rather than a visible defect, on a possession measured in minutesFront, on consequenceHeavy-haul plain lineHigh tonnage against a straight section, so the rate is moderate but relentless; squats and lip build rather than side wearClosure cost on a line that cannot easily be spared, and a defect that travels with the wheel pathMiddle, and this is the block that gets cut first when money is shortYards, sidings and low-speed terminal trackLow traffic, but corrosion and lip accumulate where wheels stand and slipRide quality and noise at the platform end, rarely a safety consequenceLast, and correctly so, provided the fallback is genuinely cheapThe ordering is stable in shape from network to network and completely unstable in detail, which is why a static list is a liability. The rates that produced this year's order were measured before the pass that changed them: grinding alters the very profile that was generating the rate, traffic mix shifts, lubrication practice changes, and a section that was front-ranked can become ordinary within a cycle. Re-ranking is not a review of last year's decision, it is the decision.The order has to be rebuilt from the measurement that the last pass made possibleWhen the money does not reach the end of the listWhat to doWhy it is the better use of a short budgetThe alternative that looks cheaper and is notClose whole sections at the front of the queue, and leave the tail untouched until the next roundA section ground to its planned depth comes inside the envelope and its clock starts again. A section given a fraction of its planned depth does neither — it stays above the threshold and stays on the listShaving a pass off every section on the list. The budget is spent in full and not one section leaves the queue, which is the most expensive possible outcomeDecide finishability before entering the section, and only start what the window can completeDepth per pass multiplied by the passes the window allows is either enough to reach the plan or it is not. A pass abandoned part-way leaves a transition in the middle of the section that the next campaign inheritsEntering every front-ranked section for a first pass and returning for the rest — the return trip is a new mobilisation for a section that had already been paid for onceRe-rank before the next window rather than at the next budgetThe order is a forecast, and every pass updates the inputs to it. A mid-year re-rank costs a meeting; a mid-year re-rank done late costs a sectionCarrying the agreed list through the year because it was agreed, and discovering the error at acceptanceThe budget that constrains all of this is built on quantity rather than on price — the reasoning behind what drives the cost per kilometre is where the pass-kilometre basis comes from, and the queue is only as good as the acceptance test that closes each section: the criteria that decide whether a job passed is what converts a finished section into a closed one. Two further references belong with this question rather than in it: the curve-specific case for managing side wear on curved track, and the service side of turnout work in turnout grinding. The interval that sets how long the queue may wait is set from tonnage and measured condition, and the measurements themselves only count if they are kept — the records a programme has to keep is what makes next cycle's ranking possible. Grinding prioritisation is a scheduling discipline, and a Molaton wheel cannot rank anything for you. What it can do is keep a section's removal per pass predictable enough that a queue built on pass counts stays valid when the campaign reaches the last section on it — which is the only part of the list that ever gets tested.Why trust this answer — RailwayCare (product brand Molaton) has made rail grinding wheels since 2004, drafted the industry standard JB/T 11431 for them, and holds ISO 9001 / ISO 45001 / CRCC certification. Prioritisation is decided long before a wheel is ordered, but it is visible from where we stand: on operator programmes the sections that get ground last are usually the ones whose rate was never measured between visits, and the wheel that arrives for them is being asked to remove an interval's worth of damage in a single campaign. The ranking logic below is written from that observation.Related questions you may also askRail wear measurement: how to measure it and when to grind or replaceRail grinding on curves: managing side wearWhat does rail grinding cost per kilometre, and what drives it?What acceptance criteria decide whether a grinding job passed?How is a rail grinding interval set by tonnage?What records should a rail grinding programme keep?Rail turnout grinding solutionsMore answers on rail grinding strategy and equipmentRanking a network you cannot afford to finish?Send RailwayCare the sections on your candidate list, the readings you hold at fixed points from the previous visit, and the tonnage or wheel passes for each. We will set out what the development rates imply for order, which sections are finishable inside the possessions you can buy, and which Molaton wheel keeps removal per pass stable enough for the ranking to hold to the end of the list.Rank My Sections

Question 6

Three layers in order: a pass-kilometre requirement, a price per pass-kilometre, and a contingency sized from your own rework history rather than a round percentage.

Accepted Answer

How do you set an annual grinding budget for a network?Build it in three layers, in this order: a pass-kilometre requirement, a price per pass-kilometre, and a contingency sized from your own last cycle. The figure that gets circulated first — network kilometres multiplied by a price — is not a budget. It is a quote for track nobody has surveyed this year, and it fails in both directions at once. It over-funds long straight sections that need few passes, it under-funds the curves that need many, and it contains nothing at all for the two things that reliably move the outturn: re-grinding a pass that failed acceptance, and inserting a section that crossed its threshold between planned visits. A budget is a purchasing plan for a quantity you can define, not a price applied to a quantity you happen to know.The number has to describe work that can actually be bought and measuredLayer one: the quantity is pass-kilometres, and passes are an outputThe unit that describes grinding work is the pass-kilometre: grinding length multiplied by the number of passes. Passes are not a policy choice, they are a result — the profile deviation at the control points divided by what one pass of that machine removes at that speed, rounded up. That has a direct consequence for budgeting: this year's requirement cannot be typed in from an intention, it has to be projected from what last cycle actually consumed, held separately for each class of track.Two numbers from a single operator campaign show why the classes cannot be averaged. On one programme, straight sections were ground in four passes and curve sections in six — the same kilometre of track consuming half again as much work in one place as in the other. And inside one curve of 800 m radius ground in six passes, the steel removed at the maximum reached 1.8 mm on one side of the head against 0.3 mm on the other. The section is budgeted as a whole, but it cannot be described by an average, because the average is neither side.Line in the budgetUnit it has to be built inWhat it is made ofWhat goes wrong when it is built in track-kilometresQuantityPass-kilometres: length × passesThe pass-kilometres measured in the previous cycle, held separately for straight track, curve bands and turnout areasLength becomes the only variable, so the longest sections absorb the programme and the curves that need six passes are funded as though they needed fourPrice basisCost per pass-kilometre, with campaign overheads as their own lineThe grinding itself, the consumable it burns, and the mobilisation and possession cost that is charged per campaign rather than per passEvery bidder has to fold the passes into one rate, and that rate is then wrong for every section on the list — high on the straight, low on the curveContingencyPass-kilometres, not currencyA re-work allowance for passes rejected at acceptance, plus an insertion allowance for sections crossing a trigger out of sequenceThere is nowhere to charge a second pass, so the first section that fails acceptance is paid for by cancelling the next one on the listDelivery capacityPossessions: windows and hoursThe requirement divided by the pass-kilometre output measured per window on this networkThe money is approved and there are no windows to spend it in, which is the most common way a funded programme ends the year under-spent and still behindThat fourth line is the one that connects money to the timetable, and it is a division rather than an assumption. One machine put down 18.72 pass-kilometres inside a single two-hour possession on the Hengliu line in November 2022. Divide the requirement by the output measured on your own track and the answer is the number of windows that have to be bought; skip the division and the budget authorises work for which no possession exists.Curves and straight track do not consume the same quantity per kilometreSizing the contingency from records rather than from habitCampaign shape is a budget decision in its own right, because overheads do not scale with pass-kilometres. The same annual requirement delivered in fewer, longer campaigns carries less mobilisation and fewer possessions; delivered in more, shorter ones it fits a fragmented possession pattern. That trade has a limit on the other side, since a longer interval lets more damage accumulate between visits, and the depth required to clear it grows with the interval — so the cheap campaign shape is not always the cheap programme.A contingency that exists as a percentage survives review and then behaves like spare money. A contingency with named components behaves like a forecast. There are two that carry almost all of the variance. The first is re-work: a pass that met the plan but not the clause is a second pass paid for out of the same envelope. This is usually a criterion mismatch rather than a bad crew — one intercity operator's published grinding quality index sets a depth floor of 0.1 mm for preventive work against 0.2 mm for pre-grinding and for corrective work, so a compliant preventive pass can be rejected by a corrective clause and ground again.The second is the out-of-sequence insertion: corrugation depth or lip width past 0.3 mm, a crack reading at the window, a ride complaint. These are not failures of the programme, they are what the programme exists to catch, and a budget with no line for them borrows from the next section on the list every time one fires.ComponentWhat puts it in playHow to size it from what you already holdRe-grind after a failed acceptanceA pass that met the plan but not the clause: a depth floor written for a different strategy, surface roughness, a continuous blue band, coverage short of the target profileThe ratio of re-ground to accepted pass-kilometres in the previous cycle. If that ratio is not held, the honest position is that the line cannot be sized this year, and that recording the inputs is the condition for next year's number being a forecastOut-of-sequence insertionCorrugation depth or lip width past 0.3 mm, a crack reading taken at the window, a ride complaint that arrives before the next planned visitThe count of unplanned interventions last cycle multiplied by their average pass-kilometres. The count predicts better than the cause does, because the causes change and the count does notShort-section and transition workTurnouts, platforms, level crossings and anything under about a kilometre, where the machine has to ramp in and out and the overhead sits on very little lengthThe pass-kilometres booked on sections under a kilometre. This is the line that quietly doubles a campaign's overhead when it is folded into a network-wide average instead of held apartNone of the three can be sized from a percentage, and all three are cheap to hold once the records separate them. That is the argument for treating the budget and the record-keeping as one decision rather than two.Money approved for pass-kilometres still needs windows to be spent inWhat the budget meeting will ask, and the number that answers itQuestion from the meetingWhat actually answers itThe number you must already holdWhat a wrong answer looks likeWho owns the answerWhy not grind everything deeply once and skip a year?The envelope. Depth is a lifetime account on that rail, and one campaign spends a slice of it rather than resetting itSteel already spent against the profile envelope, and the per-job acceptance ceiling that caps any single visit"We will go deeper and not come back" — which spends two years of account in one possession and still does not remove the next visitPermanent way engineerWhy is our rate per kilometre higher than the neighbouring network's?The section mix, expressed as passes per kilometre, not the price of a passPass-kilometres per track-kilometre for this network and for theirsA comparison of two averages over two different mixes, which settles nothing and usually becomes an argument about suppliersWhoever owns the budget modelCan half of it move into next financial year?The interval against the damage rate on the sections deferred — deferred as classes, never as a slice off the top of the listMeasured development rate per section class, in the unit your intervals are already written inDeferring by value rather than by class, so the fastest-developing sections leave the programme alongside the cheapest onesPermanent way engineerWhy is there a contingency line at all?Last cycle's actual-to-plan ratio, split into re-ground and inserted pass-kilometresRe-work and insertion pass-kilometres as a share of planned, taken from records rather than from memoryA round percentage that passes review and is spent in the first quarter, leaving the last quarter exposedFinance, with engineering signing the ratioWhat does the extra money buy?Coverage and frequency. It does not buy depth, and a budget that promises depth is promising what the envelope will refuseSections still outside their interval at the funded pass-kilometre level"Deeper passes" — the one thing an increase cannot purchase, because the ceiling is geometricAsset managerTwo of these answers depend on documents the budget process does not produce, which is why budgets get defended on assertion instead of evidence. The rate per class comes from setting the interval from tonnage and measured condition; the actual-to-plan ratio comes from the records a programme has to keep, and if those do not separate re-ground from inserted pass-kilometres the contingency has nothing to be built from. The price side is set out separately in what operators pay per kilometre and why and in the drivers behind a cost per kilometre; the commercial side of the same problem is the KPI set a contract has to carry; and the ceiling that keeps depth out of the purchasing conversation is the grinding envelope. None of them yields a price without a network, and neither does this answer. What a Molaton wheel contributes is narrower and more useful: a consumable whose cost is already expressed per pass-kilometre, and whose removal per pass is consistent enough that the projection from last cycle to this one is a projection rather than a hope. A budget resting on a pass count that changes when the wheel changes is not a budget, it is a range.Why trust this answer — RailwayCare (product brand Molaton) has supplied rail grinding wheels since 2004, drafted the industry standard JB/T 11431 for them, and holds ISO 9001 / ISO 45001 / CRCC certification. Because our wheels are quoted against pass-kilometres on operator campaigns, we sit inside the arithmetic this answer describes: we see which budgets survive the year with their contingency intact and which spend it in the first quarter on re-ground passes. The structure below is written from that position, and contains no unit prices, because a price without a network is a number nobody can check.Related questions you may also askRail grinding cost per kilometre: what operators payWhat does rail grinding cost per kilometre, and what drives it?What KPIs should a rail grinding contract include?What records should a rail grinding programme keep?How is a rail grinding interval set by tonnage?What is the grinding envelope, and who sets its limits?How to get a rail grinding quote: tips for buyersMore answers on rail grinding strategy and equipmentBuilding a grinding budget that survives its own year?Send RailwayCare the last cycle's pass-kilometres by section class, the rate you pay per pass-kilometre and the re-work share you actually incurred. We will come back with the pass-kilometre requirement the mix implies, the campaign shape it fits, and the Molaton wheel specification whose removal per pass makes that projection hold across a season.Build My Budget Basis

Question 7

Three steps in a fixed order: baseline the sections, set a clock the track responds to, then rewrite acceptance and payment so a preventive depth is allowed to pass.

Accepted Answer

How does a network move from corrective to preventive grinding?Corrective grinding decides depth from a defect you can see. Preventive grinding decides it from a clock and a tolerance you agreed to — and that swap is the real work. The move between them is not a change of depth setting, it is a change in what has to be proven at the end. A corrective job is accepted when the defect is gone: one measurement, at one moment, settles it. A preventive job cannot be accepted that way, because a shallow pass over rail that is already near its target profile produces rail that looks finished whether or not the amount removed was right. What makes it provable is a comparison — damage accumulated since the previous pass against steel removed in this one — and a comparison needs a starting point. So the transition runs in three steps in a fixed order: baseline the sections that will go on the clock, set an interval from a clock the track actually responds to, then rewrite the acceptance and payment wording that decides whether a preventive depth is allowed to pass. Each step is what makes the next one checkable. Networks that reverse the order end up with a preventive campaign ground to a corrective depth, or one that cannot be closed out at all.Changing who decides the depth changes what has to be proved afterwardsThree steps, in this orderStepWhat it puts in placeWhat breaks when it is skippedBaseline the sections that will go on the clockA measured starting condition at fixed points: profile against the target, a defect census by type, and the section still left in the headThere is nothing to subtract. The interval cannot be defended and the first pass cannot be judged, because nobody recorded what was there before itSet the interval from a clock that section responds toAn interval tied to the mechanism actually damaging that track — accumulated tonnage where duty is heavy, wheel passes and measured condition where it is notA round number is adopted instead. It is then either too long, so damage outruns the depth, or too short, so the section is spent for nothingRewrite the acceptance and payment linesA criterion a preventive depth can satisfy, a unit of payment that does not reward depth for its own sake, and a proof defined across the interval rather than at a single momentThe strategy changes and the paperwork does not. The crew grinds to the old depth to be paid, which is a corrective cut charged to a preventive clockIn practice the third step is where transitions fail, and the reason is organisational rather than technical. The first two can be done by the engineering side alone. The third needs procurement, the contractor and the acceptance authority to agree on a number that is smaller than the one they used last year — and nobody is thanked for proposing that. It is still the step that decides whether the programme survives its first cycle, because a criterion written for defect removal will reject a compliant preventive pass every single time.The strategy is only as durable as the clause that pays for itWhere the wording has to change with the strategyLine in the documentsWhat it says under corrective grindingWhat it has to say under preventive grindingWhat leaving it alone costsWho owns the changeMinimum depth at acceptanceA removal deep enough to take the defect out, which for most visible defects means a value in the upper tenths of a millimetreA value matched to a planned preventive depth, which is smaller by design. One Chinese intercity operator's published grinding quality index, for example, sets 0.2 mm for pre-grinding and for corrective work but 0.1 mm for preventive workA compliant preventive pass is rejected as shallow, or the crew doubles the depth to pass and the section is consumed at twice the planned rateThe acceptance authority, not the contractorInterval clauseSilent, or written as "as defects appear"A stated range on a stated clock, plus the condition that triggers an out-of-sequence interventionThe programme reverts to reactive within a year, and the transition is remembered as a cost with no visible benefitThe permanent way engineerUnit of paymentPer millimetre removed, or per defect treatedPer kilometre brought inside the envelope, with the depth band fixed and coverage measuredThe commercial incentive points straight at depth, so every pass drifts to the deepest value the criterion will allowProcurementProof required at acceptanceThe defect is gone; a single inspection closes the jobAccumulated damage is smaller than steel removed; two measurements across the interval close the jobCorrective proof is applied to preventive work, and it always fails — there is no defect left to point atThe acceptance authority and whoever holds the recordsWhat has to be recorded per cycleThe defect, the depth used, the resultProfile against target at fixed points, depth per pass, the interval actually achieved, and which condition trigger fired if one didThe interval can never be revised on evidence, so it slowly becomes whatever the first guess was, defended on seniorityWhoever holds the recordsTwo of these lines are worth arguing about in advance rather than at the first acceptance meeting. The depth line sets the physical ambition of the programme, and the documentation for how deep is safe before rail is over-ground is where the upper end of that band belongs. The proof line sets what the contractor has to be able to show, and the criteria themselves are set out in what acceptance criteria decide whether a job passed. A Molaton wheel does not decide any of this; what it does is make the shallow end of the depth band repeatable, because a programme whose whole economics rest on removing a small, consistent amount cannot afford a pass that skates and has to be repeated.Two measurements across the interval are what turn a shallow pass into evidenceReading whether the transition is holdingWhat you see at the end of a cycleWhat it meansWhat to do nextThe reading that would mislead youMeasured damage over the interval is smaller than the steel removed, with margin leftThe clock is at least as short as the damage rate requiresHold the interval and start widening the list of sections on itThat the rail looks good. It looks good after a corrective pass too, which is exactly the confusion being designed outDepth achieved per cycle drifts up while the interval stays unchangedCoverage is being bought with depth — something inside the pass is being missedCheck the pass pattern and the coverage before touching the intervalThat acceptance readings still pass. They are passing on depth, not on shapeUnplanned corrective interventions fall but do not disappearThe programme is working on the defects it was designed for and not on the othersSplit the census by defect type before widening the scopeA lower total count, which can hide a rise in the one defect the programme does not addressRejections cluster on a single criterionOne line of the wording was never changed, or was changed and never explained to the crewRewrite that line and brief the crew before the next possession, not after itThat the crew is at fault. The criterion is doing exactly what it saysThe economics are the reason the transition is worth its friction at all. Reported defect probabilities differ sharply by strategy — one published comparison puts them at roughly 15% with no grinding, 8% with corrective grinding and 4% with preventive grinding — and the gap comes from removing damage while it is still shallow rather than after it has become a defect with a depth of its own. That is the same mechanism that lets a grinding programme extend rail service life, and it is why the interval has to be defended on measured accumulation rather than on a nationally quoted figure — the reasoning behind setting an interval by tonnage is conditional for the same reason.Why trust this answer — RailwayCare (product brand Molaton) has made rail grinding wheels since 2004, drafted the industry standard JB/T 11431 for them, and holds ISO 9001 / ISO 45001 / CRCC certification. Our wheels have been run in side-by-side trials against imported wheels on mainline preventive programmes where the outcome was judged by the operator's own acceptance records — which is where the difference between a preventive and a corrective criterion shows up first. The transition logic below is written from that position: we see which programmes hold their interval and which quietly revert.Related questions you may also askWhat is rail grinding: a complete guide for track operatorsPreventive vs corrective rail grinding: what is the difference?How is a rail grinding interval set by tonnage?What grinding depth is safe before you over-grind a rail?What KPIs should a rail grinding contract include?What records should a rail grinding programme keep?More answers on rail grinding strategy and equipmentMoving a network onto a preventive clock?Send RailwayCare the sections you intend to baseline, the defect census you already hold and the acceptance wording currently in use. We will set out what the baseline has to contain, what interval the measured accumulation supports, and which Molaton wheel holds a shallow preventive depth consistently enough to keep the contract wording honest.Plan My Transition

Question 8

Inside a tunnel the limit is the air, not the rail: extraction capacity caps the dust rate, visibility fails before depth does, and the plan has to slow down.

Accepted Answer

What operating constraints apply to grinding inside tunnels?Inside a tunnel the constraint is not the rail, it is the air. A pass that removes the same fraction of a millimetre underground as it does in the open removes exactly the same amount of steel, and turns it into exactly the same amount of dust — but in the open that dust is carried away, and in a tunnel it is released into a volume that is barely being replaced. Three things follow, and none of them is "grind shallower". First, the extraction system becomes the ceiling: it is rated as a volumetric flow, so the mass you can make per hour is capped, which turns the night from a depth question into a rate question. Second, visibility goes before the depth target does, and the operator loses the one cue that told him the full band was covered. Third, sparks and heat that are harmless outdoors become a fire and an exposure question, so the mitigations — suppression, water mist, forced ventilation, respiratory protection — add steps the possession has to pay for. The practical result is a slower night with fewer passes, not a shallower plan.What the pass removes is fixed by the plan; what it leaves in the air is notWhat the tunnel takes away, and what it costs to get it backConstraintOn open trackInside a tunnelWhich planning variable it takes awayWhat the workaround costsDust dispersalFine particulate is carried off by ordinary air movement; the concentration never reaches anything the crew noticesThere is no dispersal. Concentration rises with the mass removed per hour and stays where the machine is workingThe rate — and through the rate, the number of passes a possession can carryTime. The cut is either slowed or interrupted while the air is cleared, and neither is free inside a fixed windowVisibilityThe bright worked band on the rail head stays readable for the whole pass, so coverage can be watched as it happensThe cloud builds ahead of the machine; the band fades at precisely the point where a missed band would show upThe certainty of coverage, and with it the confidence to keep cuttingA second way of confirming coverage — marking, measurement between passes, or shorter work blocksSparks and heatSparks land, cool and are gone; radiant heat is shed upwards to open skyA dust cloud plus sparks in a confined space is an ignition path, and the heat has nowhere to goContinuity. Every stop for fire watch or cooling is productive time leaving the nightSuppression and water mist, and then drying time before anything can be measured honestlyAir changeNot a variable at all — the track is outsideForced ventilation has a finite throughput, and a single-bore tunnel with one access is worst at the far endHow far into the bore a single night can reasonably reachSequencing. Work has to start at the end that keeps the cloud behind the machine, not at the nearest point to the accessExposure of peopleDispersed long before anyone breathes it at concentrationThe dust is not only rail steel: high contact temperatures add fine particulate from the tool itself, and the exposure limit applies where the crew is actually standingWhether the work can proceed at all on a given nightRespiratory protection, rotation, and a stop-work threshold written down in advance rather than judged on the nightRead the five rows together and one pattern is hard to miss: every one of them converts into time, and not one of them converts into depth. That is why a tunnel programme copied from an open-track programme does not produce a shallow grind — it produces an unfinished one, with the last stretch of the bore left half covered because the night ran out at the point where the dust was thickest. The rate itself is set by a combination of depth, working speed and how many cutting heads are engaged — and speed alone is never a complete instruction, as set out in whether grinding speed changes the result.Every constraint that appears underground is paid for in time, not in millimetresRewriting the plan rows for a tunnelRow of the planHow it is written for open trackHow it has to be written for a tunnelWhy the open-track form failsWorking speedChosen from the removal rate needed to reach the planned depth in the passChosen from what the extraction can clear at that depth, and then the depth is checked back against itSpeed is being set by the wrong ceiling, so the first symptom is not a slow cut — it is a dusty onePasses per possessionLength divided by speed, with a standard allowance for set-upDerived after deducting clearing time from the possession and allowing for passes lost to the stop for airThe arithmetic is correct and the possession still runs out, every night, by the same amountWhere the night startsAnywhere. Order does not change the resultAt the end that leaves the cloud behind the machine, finishing back towards the accessThe machine works into its own dust and loses the far end first, which is the stretch nobody can inspect cheaply laterAcceptance readingTaken as the cut finishes, on a surface the crew can seeTaken after ventilation, on a dry surface, with the clearing time recorded beside the numberThe reading is taken through a film of dust and water, and the argument about it arrives a week later with no way to settle itThe wet-surface row is the one that catches experienced crews. Water mist is a reasonable answer to the spark problem, and it leaves the rail head damp; a profile measurement taken on a damp head is not automatically wrong, but it is taken under conditions that nobody wrote down, and a reading without its conditions cannot be defended. The same discipline that governs preventing burn in a light-restricted working environment applies here: the mitigation and the measurement have to be planned as one step, not two.A tunnel plan is judged by whether the far end got finished, not by whether the depth looks rightDecisions that have to be made before the machine entersDecisionWhy it cannot wait until the crew is insideWhat the night looks like when it was not madeThe depth plan and the speed it impliesIn a tunnel these are one decision, not two. A depth chosen without an extraction rate behind it is a target nobody can reach within the airThe crew reaches the depth by cutting faster than the air can clear, and the final third of the bore is measured blindWhere the air goes, and how fast it is replacedThe ventilation path fixes the clearing time, and the clearing time is part of the possession rather than an afterthought at the end of itClearing is discovered at the end of the night, and the acceptance check is taken out of the following night's productionWhich point will be measured, and at what hourA measurement point has to be reachable and dry at a known time; that is a plan input, not a convenienceThe reading is taken late, at the most accessible spot, and the section that actually mattered has no number attached to itWho has authority to stop on air qualityA threshold that needs a discussion on the night is not a thresholdThe crew keeps cutting because stopping is somebody else's call, and the exposure is written up as normal working conditionsNone of this makes tunnels ungrindable; it makes them a different planning problem, which is why a network with a large underground share cannot be planned with a mainline template — the reasoning is in planning grinding for a metro or light rail network. What a wheel contributes here is narrower than suppliers like to suggest: a Molaton wheel does not change the mass a given depth removes, and it cannot make the air clear faster. What it can do is keep cutting at the reduced speed a tunnel imposes, so the night is spent on the planned passes instead of on a second, dustier attempt at the same metre. The records that make that visible afterwards are the same set a network needs anyway, described in the records a grinding programme should keep.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 and drafted the industry standard JB/T 11431 for them, working from the friction-materials laboratory background of Wuhan University of Technology. Wheels we made have been run in enclosed and confined working conditions where the acceptance was scored on a written protocol, not on a supplier claim. The constraints below are written from the operational side of grinding — what the dust, the air change and the possession actually do to a plan — not from a machine brochure.Related questions you may also askHow to grind subway rails: urban transit grinding programmesHow do you plan grinding for a metro or light rail network?Does grinding speed change the result?How do you prevent rail burn during grinding?What acceptance criteria decide whether a grinding job passed?What records should a rail grinding programme keep?More answers on rail grinding strategy and equipmentGrinding a tunnel section this season?Send RailwayCare the bore length, the ventilation arrangement, the possession you get and the defect you are chasing. We will set out how the dust rate caps the passes you can plan, which depth is reachable at tunnel speeds, and which Molaton wheel suits a shallow cut where the air, not the rail, sets the pace.Plan My Tunnel Programme

Question 9

A metro network is planned in metres per night, not kilometres per shift. Split it by work face, and stop using tonnage as the clock underground where it does not fit.

Accepted Answer

How do you plan grinding for a metro or light rail network?A metro network is not a short mainline. It is a different unit of planning. On a mainline the plan unit is pass-kilometres per shift and the binding constraint is the machine's rate — double the rate and you halve the nights. Underground almost none of that holds. The non-productive part of a night is close to fixed: traction power has to be isolated, the unit has to come out of a depot and go back into one, and the dust and water you can carry is a capacity, not an assumption. None of that shrinks when you plan fewer metres. So the variable you actually control is length, not rate, and a transit plan is written as metres per night multiplied by nights. Everything else becomes a question of which metres. The second thing that changes is the clock. Accumulated tonnage is the mainline clock because damage tracks axle load; underground the axle loads are light but the number of wheel passes through the same curve is large and concentrated in the same few hundred metres, and that repetition — not tonnage — is what drives corrugation and side wear. An interval lifted from a mainline tonnage figure will under-call the curves and over-call the tangent. The third is that one network contains several work faces whose constraints do not resemble each other, so a plan organised by line cannot be executed; a plan organised by work face can.The unit on the plan sheet decides what the plan can expressWhat a transit plan counts, and what it cannot borrowDimension of the planHow a mainline writes itHow transit has to write itWhat goes wrong if the mainline form is copiedUnit of measureKilometres, or pass-kilometres, per shiftMetres per night, against a named list of sectionsA nightly figure in kilometres reads as a target; the crew either abandons it or rushes the set-ups and leaves sections half groundTime budgetMost of the possession is productive timeA fixed share of the possession goes before the first cut and after the last — and that share does not scale with lengthEfficiency is calculated on productive minutes that were never available, and the programme slips every night by the same amountBinding constraintThe machine's removal rateThe window length, and how much dust and water the unit can carry through itDepth is planned as if capacity were free; the extraction system fills up before the metres are done and the unit has to leave and re-enterClock for the intervalAccumulated tonnage through the sectionWheel passes through the same curve, held against measured condition — with tonnage only where the duty is genuinely heavyCurves that corrugate fastest read as low-tonnage and get deferred; tangent that is barely working gets called early. The clock answers a question the track is not askingOnly the last row is a strategy error rather than an arithmetic one, and it is the one that survives longest, because the tonnage figure is real and defensible on the section it came from. On a network where the duty is light but the passes are dense, the honest move is to treat a tonnage-based interval as one input among several and let measured condition trigger the rest. There is no universal transit interval, and any supplier who offers one is describing a different railway.One network, several work faces, and a different constraint on each of themSplit the network by work face before splitting it by lineWork faceWhat dominates thereWhat constrains the planHow the metres get allottedWhat class of unit fitsOpen section between stationsRolling contact fatigue and short-pitch corrugation, much as on a mainlineAccess and the window; the track itself behaves conventionallyBy measured deviation against the target profile, not by length of trackThe same class you would use on a mainline; the difference is logistics, not capabilitySmall-radius curve clustersGauge-corner side wear and corrugation — on transit the curve radius commonly sits under about 1000 m, and that is where the damage concentratesThe curve is where the night runs out: the same metres take longer, and the set-up cannot be shortenedDisproportionately. These curves should take the largest share of metres per track-kilometre in the whole planWhatever holds the crown through the curve; hand-guided units where the curve is tighter than the machine can follow — see grinding curves differently from tangentTurnouts and depot roadsGauge-corner wear, squats and local deformationAccess, and the number of separate set-ups rather than the lengthIn set-up counts, not metres — allot a night's worth of set-ups and let the length followUnits that can be carried in and set up by hand; the division of labour is in large-machine and small-machine grindingEmbedded or street-running trackCrown corrugation, and the noise complaints that follow itThe gauge face is not accessible at all, so only the crown is in scopeBy the length where the crown is genuinely reachable, surveyed on foot before the windowCompact units that can be moved in without disturbing the surrounding surfaceRead as a whole, the table explains why a transit programme that reports success in kilometres is usually under-reporting two of its four faces. Turnouts are not measured in metres, and embedded track is not fully grindable, so both disappear from a kilometre target — which is exactly where the residual manual work sits, and where side wear on the tightest curves is left to develop until it becomes a speed restriction.The plan is only as good as the faces it names, and the metres it gives each of themBuilding the plan in four stepsStepWhat it establishesWhat skipping it costsBaseline each work face separately — condition, defect type and remaining section, recorded against fixed pointsA starting state that a later pass can be subtracted from, per faceEverything downstream is judged against the section that happens to be easiest to measure, and the hard faces never get a baseline at allFix the metres for a night from the fixed time first, then derive the rate you needA length budget that survives the isolation, the travel and the extraction capacityThe rate is chosen for the machine and the length is whatever is left — which is how a programme ends up with passes it cannot finish and sections it re-enters twiceGive each work face its own interval clock — passes and condition on curves, tonnage where the duty is heavyAn interval that tracks the mechanism actually wearing that faceOne clock is applied to four different mechanisms, and the faces that develop fastest are the ones deferred longestWrite acceptance per night into the contract, in the same units the plan usesA night that can be closed out on the spot rather than reconciled at the end of the yearQuality is discovered late and in aggregate, when the only remaining remedy is another possession — the criteria themselves are in what acceptance criteria decideTwo disciplines make the difference between a plan and a wish list. The first is refusing to state a nightly figure you have not measured a rate for — a rate derived from one section, on one surface, at one setting, is not a network rate, which is why the pass count has to come from a measured removal figure rather than from a target — see how many passes a rail needs. The second is treating the wheel as part of the plan rather than a purchase order, because on a short window the cost of a wheel that loads up or skates is paid in metres, not in wheel life. That behaviour, across the shallow preventive cut a transit programme mostly runs and the deeper cut a curve cluster needs, is set out in the Molaton wheel range. A Molaton wheel cannot make a window longer; it can stop a wheel problem from spending the window you have.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431 for rail grinding wheels, are certified to ISO 9001 / ISO 45001 / CRCC, and have supplied wheels into urban transit trials where the acceptance was scored against a written test protocol rather than a supplier claim. The planning logic below comes from those trials and from possession-limited work on transit and mainline networks.Related questions you may also askHow to grind subway rails: urban transit grinding programmesHow is a rail grinding interval set by tonnage?Should curved track be ground differently from tangent track?What is the difference between large-machine and small-machine grinding?What acceptance criteria decide whether a grinding job passed?More answers on rail grinding strategy and equipmentPlanning a transit network’s grinding year?Send RailwayCare the network layout, the possession window you get per night and the defect types you are seeing. We will set out how to split the network by work face, what each face can realistically contribute in metres per night, and which Molaton wheel suits a shallow preventive cut and a curve cluster in the same programme.Plan My Network Programme

Question 10

Surface condition does not change the depth you plan. It changes whether the first pass can cut at all - four states, the assumption each one breaks, and what stays the same.

Accepted Answer

How does rail surface condition change the grinding plan?Surface condition does not change how much steel you plan to remove. It changes whether the first pass can cut at all. A grinding plan is a subtraction: target profile minus measured profile, divided by the per-pass removal rate. Every term in that subtraction rests on two assumptions — that the surface being measured is steel, and that the wheel will bite it. Surface condition attacks those assumptions, not the arithmetic. Four states do it, and each one breaks a different term. Rust and mill scale put a loose oxide film on the head: the profile you measure includes it and the first pass removes it, so both the before-reading and the removal figure are partly oxide. Oil and grease cut the friction at the contact: removal collapses while heat does not, and the reflex response — more pressure, less speed — produces a glazed wheel and a burn in the same shift. White etching layer is not parent metal at all: measured at roughly 870 HV0.3 against about 340 HV0.3 below it, 45-320 µm thick depending on parameters, with cracks reaching about 0.2 mm. A pass sized for parent metal skates across it, or presses it in. Indentations mean the head is not a continuous plane: a plan set from the average crown leaves the pit edge standing and over-cuts the metal around it. All four call for the same shape of answer and a different setting — make the first pass a pass that reads the surface, then re-measure before the campaign commits its depth.The first pass answers a question the plan sheet never askedFour states, four different assumptions brokenWhat is on the railWhat you see before the passWhich assumption it breaksWhat the first pass has to changeWhat does not changeRust and mill scaleA dull brown or grey film, often patchy, sitting differently on the running band than in the four-footThat the measured profile is steel. The gauge reads the oxide, and the oxide comes off easily, so the pass reports removal that is not metal removal — see what grinding actually removesOne light conditioning pass, then re-measure at the same points and set the removal depth from the new reading. This pass is a measurement, not the first removal passThe target profile. It describes where the head should end up, not what it looks like nowOil and greaseA dark wet sheen, characteristically near a lubricator, a crossing, a fuelling point or a depot throatThat the wheel will bite. Contact friction falls, removal drops away, and the surface stays grey and unmarked while the temperature climbsClean first, then grind. If the possession forces grinding anyway, cut light and watch colour rather than depth — the wheel is the thing being tested, as set out in a wheel that glazes instead of cuttingNothing in the depth plan is valid until friction is back. Depth computed from a skating pass is a number without a sourceWhite etching layerBright, almost polished bands that do not respond to a normal cut; martensitic rather than the surrounding structureThat the material is uniform. Hardness and crack depth vary through the layer, so one depth cannot serve it and the metal under itA shallow pass read against the layer thickness, not against the profile deviation — the grading logic is in controlling white etching layer by grindingThe measurement discipline. Layer thickness and crack depth are readings you take, never values you assumeIndentationsA regular outline with sharp edges, repeated at even spacing along the same rail; it catches a fingernailThat the head is a continuous plane. A pit is local metal below the plane, so a uniform depth is wrong in two directions at onceAn exploratory, light pass, with the depth set from the deepest pit rather than the average crown; the blending length then follows that depth, as described in the indentation grinding protocolThe acceptance thresholds. Roughness at or under 10 µm and no continuous bluing apply whatever the surface looked like going inThe pattern across those four rows is worth naming, because it is what makes this a planning question rather than a housekeeping one: not one of them is answered by grinding more. Rust answers to measurement, oil answers to cleaning, the hard layer answers to depth control, and a pit answers to where the depth is referenced from. Every one of them turns the first pass into an instrument, and an instrument that is read as a working pass corrupts the depth budget of the whole campaign from its first metre.Four surface states, four different settings, one shared discipline of reading before committingReading the first pass when the surface was not cleanWhat the first pass showsWhat it actually meansWhat to do nextWhat not to doRemoval well below plan, surface clean and grey, no marksThe wheel is sliding. Lubrication on the rail or a loaded wheel face; the cut never engagedClean the rail, dress or change the wheel face, and run one pass again before reading anything into the numberDo not add pressure or slow the machine down. That converts a friction fault into a heat fault — the mechanism in preventing rail burnRemoval below plan, surface bright and polishedThe pass cut oxide and stopped at metal it had not yet reached. The reading describes the film, not the headRe-measure now and re-plan the depth from this pass. Record it as the conditioning passDo not count it as the first removal pass, and do not average it with the passes that followRemoval on plan, but colour rises in one band and fades either sideThe wheel found a harder band. The single depth number is an average across two materialsReduce depth per pass in that band and re-order the angles so the same band is not worked continuously — how the angle sequence is setDo not accept the average as proof of removal. Half the head was never cut to that figureRemoval on plan and even, but a bright strip survives at the pit or along one shoulderThe plan was referenced from the average crown, so the lowest point never came into the cutRe-set the depth from the deepest point and read the first pass again at that pointDo not add a pass to the same pattern. It deepens what was already cut and leaves the strip where it wasThat last row is the expensive one, because it is invisible in an average. Removal across a rail head is never even — a single machine in a single working can take several times more metal on one side of the head than the other — so a first pass read at one position can certify a surface it never touched. Reading the crown and both shoulders is the minimum that makes the reading mean anything, and it is also the only way to tell an oil film from a missed angle: one leaves the whole band grey, the other leaves a strip.What the surface looked like is a plan input, and it belongs on the sheet in writingFour lines to add to the plan sheetLine on the planWhat it recordsWho fills itWhat it costs to leave outSurface condition by chainageWhich of the four states, over which length, and how thick or how deep the reading said it wasThe person who walked the section before the possession — not the operator on the nightThe first pass is set for the wrong surface, and the discovery happens after the depth has already been spent on itFirst-pass type and capWhether this pass is a conditioning pass or a removal pass, and the depth ceiling for whichever it is — against the 0.5 mm limit on parent metal in the safe depth before over-grindingThe planner, from the measured conditionTwo shifts read the same pass two ways and neither number transfers to the next sectionPre-treatment and who pays for itClean, scrape or dry; who does it, and whether it sits inside the possession or before itThe planner together with whoever owns the maintenance of wayCleaning turns up as an unplanned stoppage in the middle of the window, and the metres planned for the night do not happenRe-measure points and the stop triggerThe fixed points where the before-reading is taken again after conditioning, and the reading that stops the job rather than changing itThe supervisor, with whoever owns the envelope for anything outside tolerance — see who sets the envelope limitsThe original depth survives the new reading, and it is paid for out of the section — the deficit described in rail that has already been over-groundNone of these lines is a machine setting. They are records of what the rail was, and they exist so that the trial pass has something valid to test against instead of testing the plan against itself. Which wheel holds a consistent cut through oxide, through a lubrication film and through a hard band is a consumable question rather than a planning one, and it is answered in the Molaton wheel range. A Molaton wheel behaves predictably; it still cannot tell you what it is cutting through unless the surface is written down before the window opens.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431 for rail grinding wheels and are certified to ISO 9001 / ISO 45001 / CRCC. The readings and thresholds below come from acceptance clauses, bench work and our own on-track trials, and they are the same four questions our engineers ask before a campaign sets its first pass.Related questions you may also askWhat does rail grinding actually remove from the rail head?How do you control white etching layer by grinding?What is the role of a trial pass before full grinding?How do you prevent rail burn during grinding?What grinding depth is safe before you over-grind a rail?More answers on rail grinding strategy and equipmentGrinding a section you have not seen clean yet?Send RailwayCare the rail section, what the surface looks like along the length, and the possession you have. We will set out which surface states change the first pass, what the pass has to read before it becomes a removal pass, and which Molaton wheel holds a consistent cut through the condition you actually have.Plan My First Pass

Question 11

One measured pass proves three things: removal reached the plan, no burn, the angles cover the band. If any fails, change the setting - more passes repeat the fault.

Accepted Answer

What is the role of a trial pass before full grinding?It is the only pass in the campaign you can afford to get wrong. Everything after it has already spent the possession, so the trial exists to convert a plan into a measurement while the cost of being wrong is still one pass wide. One pass, measured at the same fixed points before and after, has to settle exactly three things. Did the removal reach the planned depth? Pre-grinding and remedial work plan at least 0.2 mm at the crown, preventive work at least 0.1 mm, and no single job goes past the 0.5 mm cap on parent metal. Did the cut leave the surface out of the burn range? No continuous bluing, roughness at or under 10 µm, no periodic wheel marks. Do the angles cover the band the profile asks for? One continuous fresh cut across the whole zone, no bright strips left between cut bands. If any of the three fails, the answer is a setting change — depth per pass, travel speed, power, or the angle pattern. It is not another pass. Adding passes re-runs the same fault on metal that has not yet been touched, and on a rail it is metal you will want in the next cycle.The trial is a measurement taken while the decision is still reversibleThree readings, and nothing else, decide whether the window opensWhat the pass has to proveHow it is readWhat passing looks likeWhat failing looks likeWhere the threshold comes fromRemoval reached the planned depthSame points, same method, measured immediately before and immediately after a single passA cut equal to the mode's plan: at least 0.2 mm at the crown for pre-grinding and remedial work, at least 0.1 mm for preventive work, with 0.5 mm as the absolute ceiling on parent metalA clean, grey, almost polished surface with a cut below the plan. The metal looks good — that is exactly what makes it easy to missThe grinding mode plan and the parent-metal cap; contract clauses are typically written as at least 0.2 mm at the crown at 7 km/h with roughness at or under 8 µm and no bluingNo burnColour along the whole trial length, plus roughness, taken on the same pass the depth was measured onRail colour, roughness at or under 10 µm, no continuous blue band, no evenly spaced wheel marksPale yellow turning blue, and roughness climbing with each successive readingTemperature against colour: normal colour below about 471 °C, pale yellow burn around 471-600 °C, blue burn around 600-735 °CThe angle set covers the bandWhich parts of the head carry a fresh cut and which are still untouched, read across the width, not along the lengthOne continuous cut through the zone the target profile asks for, edges includedBright untouched strips between cut bands, or the cut sitting to one side of the crownThe pass pattern against the target profile — see how a pass pattern is setThe section is still inside the envelopeTrial length measured against the envelope, not against the previous contractor's reportDeviation inside region 1 at +0.2 / −0.4 mm and regions 2-4 at ±0.2 mm, with depth still available for the rest of the jobThe trial length already at or over the band on one pass — the plan, not the machine, is wrongThe envelope and its owner — who sets the limitsOne detail matters when reading the first row: removal is not the same at every position across the head. In a six-pass campaign on an 800 m-radius curve, the same high rail gave up about 0.3 mm at the outer shoulder against about 1.8 mm on the inner side — a sixfold spread on one rail, from the same machine in the same working. A trial that reads the crown only will therefore report a number that is true and useless. Read at least the crown and both shoulders, which is also the only way the third reading — coverage — can be answered at all.Depth, heat and coverage are three separate questions asked of the same single passReading a failure: which parameter to move, and why a pass will not doWhat the trial showsWhat it actually meansWhat to change before the next passWhat adding a pass would do insteadDepth short, surface cleanPower or travel speed is wrong for this rail — a faster pass shortens the contact arc and takes less per passSlow down or raise power until a single pass reaches the planned depth, and record the pair — a speed without its power is not a setting — see what speed actually changesRepeats the same shallow cut. You reach the depth eventually, but every repeat is another heat cycle on the same metre, and the burn reading gets worse while the depth reading gets betterDepth fine, colour blue at one zoneHeat per metre is too high there: too much depth in one pass at too low a speed, or the head dwelling over one zoneCut the depth per pass, raise speed, shorten the dwell, and re-order the angles so no single zone is worked continuously — the sequence is in preventing rail burnBuys nothing on depth, adds heat, and thickens the white layer that a bench programme measured running from roughly 45 µm to roughly 320 µm as the parameter roseDepth fine, no burn, bright strips untouchedAn angle is missing from the set, or the head is being worked across a narrower band than the profile requiresChange the pattern, not the depth — a missing angle is a coverage fault, and coverage does not improve with repetitionRe-runs the same angles over the same strips: the untouched strip stays untouched and the worked strip simply goes deeperDepth, burn and coverage all fine, but the crown is already at the bandThe trial found the limit the hard way: the whole job cannot run without leaving the envelopeStop and re-baseline the plan with whoever owns the envelope. This is a scope question, not a parameter questionMakes it worse on every pass, and creates the deficit described in what a programme should recordThe pattern in that table is worth naming, because it is the most common mistake in trial work: depth can be bought with passes, but nothing else can. A campaign that answers every failed reading by grinding again converts a parameter problem into a section problem. The three readings are deliberately separate for that reason — each one points at a different setting, and only one of them points at the pass count.The per-pass rate the trial measures is the number the rest of the plan is built onWhere the trial sits in the windowStepWhat it producesWhat skipping it costsPick the trial length — representative of the worst condition in the job, same rail section and same class of deviation, not the smoothest stretch availableA result that transfers to the kilometres you have not seen yetA trial that proves the easy part of the job and says nothing about the length that will actually decide the outcomePre-measure at fixed, marked points with the same method that will be used for acceptanceA before-picture that the after-reading can be subtracted fromRemoval becomes an impression rather than a number, and the acceptance report has no baseline to stand on — the method is set out in verifying a job without an inspection carRun one pass only, then read depth, burn and coverageA per-pass removal rate — the single input the whole pass count is computed fromYou learn the rate after the depth has already been spent, and the plan cannot be corrected without paying for it twice — see how many passes a rail needsFix the setting from the readings, or exit the jobA parameter set the rest of the window can be run on without re-decidingEvery later pass becomes another uncontrolled experiment on rail that cannot be re-cutThat is why the trial is best understood as the gate rather than the warm-up: it is where the pass count, the pattern and the speed all get their numbers from one measured reality instead of three assumptions. Which wheel holds that depth at that speed without adding heat is a consumable question, answered in the Molaton wheel range; what the finished surface must read before the job is signed off is set out in the acceptance criteria and, at length, in how acceptance standards are written. A Molaton wheel is only as good as the pass that proves what it does on your rail.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431 for rail grinding wheels and are certified to ISO 9001 / ISO 45001 / CRCC. The depth, roughness, colour and coverage readings below come from acceptance clauses and our own on-track trials, and they are the same three readings our engineers ask for before a campaign commits its window.Related questions you may also askHow many grinding passes does a rail need?How is a grinding pass pattern set, and why does the angle sequence matter?How do you prevent rail burn during grinding?What acceptance criteria decide whether a grinding job passed?How do you verify a grinding job without an inspection car?More answers on rail grinding strategy and equipmentPlanning your first pass on a new section?Send RailwayCare the rail section, the measured deviation and the possession you have. We will set out what one trial pass has to show on your rail, which angle pattern covers the band, and which Molaton wheel holds the depth without putting heat into the surface.Plan My Trial Pass

Question 12

You cannot put the crown back. Read the band, then either bring the shoulders down to the crown you cut, or stop and plan replacement. Check the wheels either way.

Accepted Answer

How do you treat a rail that has already been over-ground?Grinding cannot undo grinding. There is no pass that puts the crown back, so the first thing to abandon is the instinct to run the machine again over the low spot. What is still open is a choice between three moves, and the choice is settled by measurement rather than by another pass. Read the running band first. If the polished band has stopped running through the middle of the head and now appears on the two shoulders with an untouched strip between them, the crown was cut below its neighbours — the profile is concave across its width, and the wheels are riding two shoulders instead of one crown. When the shoulders still carry metal above the wear limit, the profile can be made convex again by bringing the shoulders down to the crown you cut; that works, and it is the only grinding-based repair there is, but it spends more section, so the whole profile then sits permanently lower than the plan said. When the shoulders are already at their limit, further grinding is not a repair and the conversation moves to replacement planning. Either way, inspect the wheels, because a band that has narrowed or split loads them first — and the component that has already failed is often not the rail.An over-cut crown is a subtraction already spent; the repair question is what the remaining section can still affordWhat an over-ground head looks like, and why the score can still be greenWhat you see on the headWhat the section has actually doneWhat the profile score can sayWhy it is a loss you cannot recoverCrown sits at or below the bottom of its zone band — region 1 is allowed +0.2 / −0.4 mm against the target profileMetal has left the part of the head the design expected to be thereStill passes while every other zone stays inside its own allowanceEvery later cycle starts from a lower crown, and nothing in the process adds steel back — the arithmetic is set out in whether to grind or replaceThe polished band no longer crosses the middle; it shows on both shoulders with an untouched strip between themThe crown was cut below its neighbours, so the profile is concave across its widthOften still passes, because the shoulders read at zero deviation and the crown reads at the edge of region 1The wheels stop touching the middle of the rail altogether, and the area that carries the load shrinks to two stripsCrown flattened into a wide polished plane instead of a crowned arcThe transverse radius has been taken out of the headCan improve — a flat face can sit inside the band across more measuring points than a curved oneThe score measures fit inside a band, never how much metal is left. A comfortable number over a thin section is the trap described in the grinding envelopeColour on the surface, or a bright etched layer under it, with the geometry otherwise compliantHeat went into the metal as well as the shape — usually the signature of chasing depth with extra passes at low travel speedNothing: surface condition is a separate acceptance item from profile fitA hard, brittle white layer measured at about 870 HV0.3 against about 340 HV0.3 for parent metal. In bench work its thickness ran from roughly 45 µm at 1,000 rpm to roughly 320 µm at 2,000-3,000 rpm — a layer that later breaks up and seeds cracksNote what the second and third rows have in common: the measurement is not lying, it is answering a narrower question. A profile score asks how closely the head matches a shape and how far that shape is from the design line, zone by zone. It does not ask how much of the head is left, and it does not ask where the wheel actually touches. Both of those are separate readings, and an over-ground rail can be perfectly compliant on the first while being deficient on the other two.The band tells you where the wheel went; the section tells you whether there is still money in the accountThree readings decide it — and each one has a different ownerWhat to measureMeasured against whatWhere there is still a route backWhere there is noneWho decidesRemaining section at the crown and at the gauge cornerThe wear limit for that rail section — for 43 kg/m rail 9 mm vertical and 15 mm lateral, for 50 kg/m rail 10 mm and 17 mmStill above the limit with the next cycle's planned depth in hand: accept it, record it, and bring the next interval forwardAt or below the limit: this is rail replacement planning, no matter how good the profile looksThe infrastructure manager, on the section record — see what a programme should recordProfile deviation across the four head zonesThe envelope: region 1 at +0.2 / −0.4 mm, regions 2, 3 and 4 at ±0.2 mm, inside a 0.5 mm cap on parent metal removal in one jobDeviation that sits on the side where cutting is still allowed — the shoulders can be lowered to the crownDeviation on the side of too little metal at the crown. Cutting operations only move one wayThe envelope owner. A crew cannot re-open the band on site to make a shallower crown legalRunning band position and widthWhere the target profile puts contact — on the crown, over the zone the wheels' nominal rolling circle meets. As a reference for guidance values, turnout practice quotes 25-30 mm on high-speed lines and 35-40 mm on conventional and heavy-haulA band that is narrow or split but still on the crown, with metal on the shoulders: restore convexity by taking the shoulders down, then re-measureA band that has moved onto the gauge corner, or that has no continuous cut through the middle because the crown is below its neighbours and the shoulders are exhaustedJointly: the track engineer and the rolling-stock engineer. Rail geometry alone cannot settle itSurface right after the cutNo continuous bluing, roughness at or under 10 µm, no periodic wheel marksLight colouring that a finishing pass inside the remaining budget will take offA white layer already formed, with the section already spent producing itThe acceptance authority — and the reading is recorded, not negotiatedThe second row is where most arguments start, and it is worth stating plainly: you cannot lift a crown, you can only lower the shoulders to it. That is a real repair, it works, and it is routinely the right call — but it is a repair paid for out of the same account that was already overdrawn. The profile afterwards is correct in shape and lower in level, which means the next cycle arrives sooner for the whole width, not just for the part that was over-cut. On heavy-haul and mining lines, where tonnage consumes the section budget quickly, that acceleration compounds: the section economics of a long-haul programme are set out in track care for heavy-haul and mining railways.A narrowed or split band raises contact stress before it raises any alarm on the railWhy the wheels usually fail before the rail doesPartWhy it goes firstWhat to look at on siteWhat grinding can and cannot doWheel treadA narrower band loads a smaller area. A tread that has hollowed meets the rail on two shoulders rather than one crown, so the two profiles stop agreeing even if each is individually acceptableThe shape of the tread across its width, and whether the polished mark on the tread is one band or two. If it is two, the rail is not the only thing that needs restoringGrinding can move where contact happens. It cannot change the wheel's own shape — that is wheel-set reprofiling, a different job in a different placeWheel flange and gauge faceOnce the band migrates towards the gauge corner, the lateral force carried at the flange rises, and both faces wear fasterFlange wear against the previous inspection, and how high the contact mark has climbed up the gauge cornerGrinding can restore the gauge-corner relief the target profile specifies — see how grinding addresses side wear. It cannot restore a worn flangeGauge-corner fatigue on the railThe corner was never designed to carry the load that a narrowed band puts on itCracking and spalling in the corner zone, not only at the crownThis is where grinding still has genuine work to do — but it is now remedial work created by the earlier over-cut, and it is paid for in more sectionThe crown itselfIt is already below the level the plan assumedDepth at the same fixed measuring points each cycle, compared with the envelope rather than with the last contractor's reportNothing. This row is arithmetic, and it is where the honest answer to the question is “this is the depth that was not safe” rather than “grind it again”The direction of that last mechanism is not a matter of opinion. A rail defect that departs from the intended profile degrades the wheel-rail contact condition, and that degradation raises noise and vibration and accelerates the deterioration of both the vehicle components and the rail itself. Over-grinding is exactly such a departure — manufactured, not worn. So the sequence on site is: read the band, read the section, inspect the wheels, and only then decide whether the machine comes back. If the crown is below the plan and the band has left it, the next programme should carry the deficit forward as a recorded input and set its interval from remaining section rather than from the tonnage schedule, which is what a programme that extends life actually does. Which wheel can hold a light finishing pass without adding heat is a consumable question and is answered in the Molaton wheel range. And where the crown has gone below what the target profile allows at all, the question stops being a grinding question and becomes the one in who defines the target profile.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431 for rail grinding wheels and are certified to ISO 9001 / ISO 45001 / CRCC. The envelope figures, section limits and white-layer measurements below come from railway specifications and our own on-track records, not from a brochure, and we say plainly where grinding stops being a repair.Related questions you may also askWhat grinding depth is safe before you over-grind a rail?What is the grinding envelope, and who sets its limits?When should you grind a rail instead of replacing it?How does grinding address rail side wear?What records should a rail grinding programme keep?More answers on rail grinding strategy and equipmentHas a rail been cut below its band?Send RailwayCare the measured profile, the running band and the section record for the length in question. We will tell you whether the shoulders still hold enough metal to bring the profile back to convex, what that costs in remaining section, and whether a Molaton wheel can take the finishing pass off without adding heat.Assess My Over-Ground Section

Question 13

The envelope is the tolerance band around the target profile plus the depth floor and ceiling. The infrastructure manager sets it and publishes it; it is never a field adjustment.

Accepted Answer

What is the grinding envelope, and who sets its limits?A target profile says where the rail head should be. The envelope says how much rail you are allowed to spend getting there. It is drawn as a band around the target — a tolerance on each zone of the head — together with a depth floor and a depth ceiling: deep enough that the defect is genuinely removed, no deeper than the parent-metal cap, and never deeper than the section can still afford over its remaining life. The infrastructure manager sets it, per rail section and per remaining section, and publishes it alongside the target profile as part of the maintenance specification. It is not a live setting. An operator who works outside it has not improved a surface; he has spent a future the asset owner had not agreed to spend.The envelope is written before the machine arrives and is settled afterwards from measurements, not from opinionThree things the paperwork names separately, and crews keep mergingWhat it is calledWhat it actually constrainsWhat goes wrong when it is treated as one of the other twoTarget profileWhere the running band should sit and what shape the head should present to the wheel. On Chinese plain line it is published as a profile family — 60N or 75N, per Q/CR 583—2017 — not as a rail weight name like 60 kg/mIf "grind to UIC60" or "grind to 60 kg/m" is accepted as an instruction, the crew has a rail section and no shape, and will invent a shape from whatever the head already looks like. The full argument is in who defines the target rail profileTolerance band around that profileHow close each zone has to sit to the target. Chinese profile-quality practice allows plus 0.2 to minus 0.4 mm at the crown zone and plus or minus 0.2 mm across the other three zones, scored as a grinding quality indexTreated as a target, the band becomes a place to aim for rather than a limit to stay inside; the crew grinds until the gauge says "green" and stops, whatever depth that tookDepth envelopeHow much metal this cycle may remove and how much the section can still give over its whole life, with a hard ceiling on parent-metal removalTreated as a target, it turns into a depth to reach for; treated as absent, it is simply whatever the passes happened to take — and no record exists to subtract from next timeThe three are related but none of them substitutes for another: the profile is a shape, the band is an accuracy, the envelope is a budget. A job can land perfectly inside the band and still have overspent the budget, because the budget is measured against the rail's remaining section rather than against the drawing.Two questions with two different owners: what shape, and how much metal may that shape costWhat an envelope is actually built fromInputWhich end of the envelope it setsWhat breaks if it is left outWhere the figure comes fromHow often it changesRail section and profile versionThe shape the band is drawn aroundThe band has no centre, so every zone tolerance is applied to whatever shape was inheritedSection standard for the rail, profile family from Q/CR 583—2017 for plain line, plus the crossing standard for turnoutsOn relaying, on major repair, or when the manager revises the profileZone tolerances on that profileThe ceiling on how close, and how far, each zone may sitAcceptance becomes a single number for the whole head, and local over-grinding hides inside an averageThe manager's own profile-quality scheme, with its zone definitions and scoring bandsRarely — but it is versioned, so the version belongs on the sheetPlanned depth for the grinding modeThe floor — the minimum a pass must achieve to countPasses polish the surface and the defect survives; the next cycle has to be remedialMode rules: pre-grinding and remedial work from 0.2 mm, preventive work from 0.1 mm, measured near the crown at the running bandWith the mode, not with the mood — it changes when the campaign type changesSection already spent in earlier cyclesThe remaining headroom under the ceilingThe plan starts from the wrong remaining section, which is the commonest way a programme runs out of metal earlyYour own records: measured profile before and after each cycle — see what records a programme should keepEvery cycleRemaining section against wear limitsThe absolute ceiling, and the point where grinding stops being the toolGrinding continues on a rail that is already at its maintenance limitWear limits by rail type — for example 9 mm vertical and 15 mm side for 43 kg/m rail, 10 mm and 17 mm for 50 kg/m railEvery measurement, and it only ever moves one wayParent-metal removal capThe per-job hard stop, independent of how the budget looksA generous envelope plus a deep defect invites a single cut that no acceptance sheet would passAcceptance rules, where parent-metal removal is capped — the figures themselves are set out in what grinding depth is safeWith the standard, not per jobNothing in that table is a preference the crew can re-rank on site. Three of the six rows are published figures, one is a mode rule, one is an arithmetic subtraction from your own records, and one is a physical limit that does not respond to scheduling pressure. That is why the envelope is signed at specification level: it is the one document that lets a later argument be settled from paper.Every cycle spends metal that does not come back, so the budget has to survive being read after the factFive ways a crew leaves the envelope while every reading looks legalWhat happens on siteWhy it looks compliant afterwardsWhich line it actually crossedWho pays for itGrinding until the index score is high, then stoppingEvery reading sits inside the band, often comfortablyThe per-cycle depth budget, because the metal was spent to improve a score rather than to remove a defectThe next cycle, which has less to spend — and eventually the relaying dateChasing a corrugation peak with more depth on the same spotThe reported depth is an average, and averages stay modestThe parent-metal cap, and the local section in one zoneThe asset, permanently — and the reason is set out in why grinding can make corrugation worseSpot re-grinding between campaigns, logged as maintenanceEach individual visit is inside the envelopeThe cumulative spend, which is the only figure the envelope really tracksWhoever plans the next interval on a remaining section that no longer existsTreating the band edge as the finish lineThe surface is inside tolerance on every zoneThe depth floor, which is a separate requirementThe defect, which comes back at its own rate regardless of how tidy the profile looksApplying the main-line envelope inside a turnout or a restricted zoneThe numbers on the sheet did not changeThe local geometry budget, where target and budget are both section-specificThe restricted zone, which is either over-cut or left proud depending on which way the assumption fellThe governance part is short. The envelope is issued by the infrastructure manager with the profile, quoted in the specification, and accepted against measured before-and-after data. The contractor's part is to stay inside it and to hand back the evidence; the authority's part is to confirm the evidence and to authorise the line. On that reading, the envelope is not the grinder's business at all — the grinder is only the instrument that has to be capable of respecting it. What makes a Molaton wheel relevant here is narrow and mechanical: it has to hold a stable cut at the depth the envelope allows, pass after pass, so the depth that is budgeted is the depth that actually arrives. Verification without a recording car is its own procedure, covered in verifying a grinding job without an inspection car, and how much time that budget buys between cycles is interval set by tonnage.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431 for rail grinding wheels, are certified to ISO 9001 / ISO 45001 / CRCC, and we sit on both sides of this boundary: supplying the consumable to campaigns, and being asked to accept a result against an envelope the asset owner wrote.Related questions you may also askWhat is the target rail profile and who defines it?What grinding depth is safe before you over-grind a rail?When should you grind a rail instead of replacing it?What acceptance criteria decide whether a grinding job passed?Rail profile grinding techniques, standards and best practicesMore answers on rail grinding strategy and equipmentDo you have the envelope in writing, or just a profile drawing?Send RailwayCare your target profile, the zone tolerances you must meet and the measured section you are starting from. We will set out what depth per cycle the envelope really allows, what a Molaton wheel holds at that depth, and where your current specification leaves the budget open to interpretation.Review My Grinding Envelope

Question 14

Speed moves removal and heat in opposite directions, so the answer is a window: fast enough to fit the possession, slow enough for a pass to reach the defect without bluing.

Accepted Answer

Does grinding speed change the result?Yes, and not in one direction. Travel speed is the one setting on a grinding train that moves two things at once, and they move opposite ways. Raise it and each abrasive grain crosses a shorter contact arc, so a pass takes less metal; raise it far enough and the cut stops reaching the bottom of the defect while the surface still looks clean. Lower it and removal per pass climbs, but the heat delivered per metre of rail climbs with it, because the same contact patch sits over the same metal for longer. So the question is not fast or slow. It is which speed makes one pass at least as deep as the defect in front of it while keeping the finished surface out of the bluing range. That is a window, not a preference — and a speed figure means nothing until it is quoted with the power and the depth it was measured at.Speed is chosen from the measured defect, not from the habit of the previous campaignWhat a change of speed actually changes on the railWhat movesSetting speed higherSetting speed lowerWhere the limit comes fromContact arc between wheel and railShortens, so each grain cuts a shorter chipLengthens, so each grain cuts a longer chipGeometry of wheel and head; no amount of pressure makes a fast pass cut like a slow oneMetal removed in one passFalls with the shorter chipRisesThe floor is the defect's own depth — see what grinding depth is safeHeat delivered per metreFalls per metre of railRises per metre, and the surface starts to colourBluing threshold and the roughness limit — see preventing rail burnPressure fluctuation at the headFalls; the head rides more steadily over irregularityRises; the head works harder to hold a constant loadSimulated grinding pressure on real track irregularity, where excursion grows with irregularity amplitude and falls as travel speed risesPossession consumedLess per pass, so a fixed window covers more kilometresMore per pass, so the window covers fewer kilometresThe possession, not the grinder — and the trade is priced in cost per kilometreThe mechanism is geometric rather than a matter of operator feel: as the train runs faster, the contact arc between wheel and rail shortens, and a shorter arc is a shorter cut per grain. A bench programme that varied wheel speed from 1,000 to 3,000 rpm makes the same point from the other end — and it is worth reading as a bench result, not as a travel-speed curve. Metal removal did not climb in step with the setting; it fell and then rose. Temperature rose continuously. Roughness went 6.6, 7.2, 8.4 µm. The white layer thickened from roughly 45 to roughly 320 µm, and at the top of the range the removal mechanism stopped being cutting and became cutting plus spalling. Pushing the parameter up buys damage faster than it buys depth.Two campaigns at the same speed can remove different depths, which is why speed is only half of a parameter pairA speed number on its own proves nothingRecordTravel speedPower or wheel speed at the timeWhat the same record statesContract acceptance clause7 km/hNot stated — speed was never meant to stand aloneMetal removal at the rail crown of at least 0.2 mm per pass, no burn, no bluing, Ra at or under 8 µmTrial on a Loram-format grinding train, Jiangmen, 20267 km/h80% of available powerProfile-measured cut of 0.142-0.241 mm at the inner rail zone, 0.165-0.326 mm at the outer zone, 0.193-0.222 mm at the crown; no bluingHalf-train comparison trial, Urumqi, 200911 km/h18 kWReached 0.2 mm of depth in a single pass with no bluingProduction campaign, Yiyang line12 km/h15.6 kW145 pass-km at an average consumable loss of 25.45 mm — a wheel-life figure, not a rail figureProduction campaign, Shuohuang heavy-haul line15 km/h13.8 kW75 kg/m rail on one track, 60 kg/m on the other, in the same workingPreventive campaign, Changfu line15 km/h68% of power, about 20.4 kW, at 3,600 rpm48 wheels fitted on one side of the train for a controlled comparisonPreventive campaign, Hewu high-speed line, 202016 km/h68% of power, about 20.4 kW, at 3,600 rpmNo continuous bluing; the trial reported 4.28 against 3.27 pass-km per mm of wheelProcurement acceptance for high-speed passive wheels25-55 km/h, set by line speedNot applicable — the stones are free-rollingOne row of stones to grind 10-15 km continuously, Ra at or under 10 µm, no bluingRead the first three rows together and the trap shows: 7 km/h, 11 km/h and 16 km/h have all produced 0.2 mm-class removal, because the power behind them was not the same. A tender that asks for a speed without asking for the depth that speed must deliver has specified nothing. The last row is the honest exception: in passive grinding the stone rolls with the train, removal per pass is deliberately tiny, and the strategy is many light cycles rather than one deep cut — which is set out in the operating parameters of high-speed passive rail grinding. Its speed cannot be borrowed for an active campaign, and an active campaign's depth cannot be demanded of it.The window closes from both sides: too fast leaves the defect root, too slow leaves heat in the same metreChoosing the speed from the defect, not from habitJob in front of the machineDepth one pass must reachWhich side of the window to workWhy that sideWhat the wrong choice looks likeRolling contact fatigue, crack depth within the 0.2 mm windowThe crack's own depth, so typically a 0.2 mm-class cutThe slower side, to keep the contact arc longDepth below the crack root leaves a bright surface over a live defectA polished rail that reports the same cracks at the next inspectionCorrugation, peaks of 0.05-0.3 mm, remedial trigger at 0.3 mmEnough at the peak to cut it down, not an average over the faceModerate speed with the peak addressed deliberatelySpeed alone does not differentiate peak from trough; the pattern doesPeaks knocked back, troughs untouched, amplitude unchanged — the failure mode in why grinding can make corrugation worsePreventive grinding, planned depth from 0.1 mmAt least 0.1 mmThe faster side is available, provided the floor still holdsLight and frequent is the design intent of preventive workPasses that polish rather than cut, so the next cycle has to be remedialRemedial or pre-grinding, planned depth from 0.2 mmAt least 0.2 mmThe slower side, paired with enough power to reach itDepth and speed are one decision; power is the third member of the setWindow hours saved on the first pass and spent again on a second oneWeld zoneAbout 0.2 mm per pass inside a capped parent-metal budgetSlow, over a short strokeThe zone already carries a geometry deviation and a heat historyA step at the stroke end, or a hollow beside a clean-looking weld — see planning grinding around weldsTurnout and restricted zoneSet by the local deviation, not by the main-line planSlow entry, short work blocksThe machine stops and starts far more often than on plain trackMinutes lost in setting and hand-back, not in grinding — covered in grinding a turnout without closing the lineTwo limits do the clamping, and both are already written down: the depth floor from the grinding mode, and the surface ceiling from burn and roughness. Anything between them is a legitimate speed for that job; anything outside is a trade someone else pays for. One boundary stays on the consumable side — which wheel can hold a 0.2 mm cut at the speed you need, and still take the full Molaton range, is a wheel question, and it is answered there. What the campaign should spend in total belongs to interval set by tonnage.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431 for rail grinding wheels, are certified to ISO 9001 / ISO 45001 / CRCC, and the speed, power and depth pairings quoted above come from our own on-track campaigns and acceptance clauses, not from a brochure.Related questions you may also askHow many grinding passes does a rail need?How is a grinding pass pattern set, and why does the angle sequence matter?What grinding depth is safe before you over-grind a rail?How do you prevent rail burn during grinding?Operating parameters of high-speed passive rail grindingMore answers on rail grinding strategy and equipmentWhich speed fits your defect, and at what depth?Tell RailwayCare the defect type and measured depth on your section, the power available at the head and the possession you work in. We will set out the speed window that satisfies both limits, what a Molaton wheel holds at that speed, and where your current setting is spending metal it does not get back.Set My Grinding Speed Window

Question 15

Four record sets carry a grinding programme: condition before the work, the parameters actually run, acceptance data after it, and wheel batches with consumption.

Accepted Answer

What records should a rail grinding programme keep?The rail will be measured again next cycle. The numbers behind this campaign will not exist unless someone wrote them down. Four record sets separate a grinding programme from a series of grinding jobs: the condition found before the machine arrived, the parameters the job actually ran at, the acceptance data taken after it left, and the wheel batches with what each consumed. Their real purpose is not compliance and not the audit file. It is to hand the next plan a measured baseline instead of last year's assumption — because the passes a section needs are derived from the deviation the last campaign did not remove, and that deviation only exists as a number if somebody recorded it.A possession produces two outputs, a ground rail and a data set, and only one of them can be re-measured from the track afterwardsThe four record sets, and the decision each one feedsRecord setWhat goes into itWhich decision it feedsWhat breaks when it is missingCondition before the workProfile readings by section against the target, the defect map with measured depth at each point, rail section and tonnage carried since the last campaignHow many passes and which pattern the section needs, and where the depth budget is already spentPasses get carried over from the previous job or from the standard pattern, and depth is discovered to be unavailable mid-windowParameters actually runPass count, working speed, the angle sequence used, depth achieved per pass, section actually completed, and every stoppage with its reasonWhether the outcome came from the plan or from improvisation — and what a repeatable job looks likeThe result cannot be reproduced, so a good campaign cannot be copied and a bad one cannot be diagnosedAcceptance data after the workProfile against the target, 0.3 mm under a 1 m straightedge, Ra at or under 10 µm, no continuous bluing, parent-metal removal within 0.5 mm, ramp steeper than ‰1Handover, payment, and the starting position for the next intervalDisputes are settled by opinion, and the next plan starts from a shape nobody verifiedWheel batches and consumptionBatch identity, wheels fitted and removed per shift, pass-kilometres per millimetre achieved, and any abnormal wear, glazing or breakage with the section it happened onRe-order quantities, and whether a consumption change is a wheel problem or a depth problemConsumption anomalies only appear as an unexplained purchasing increase months laterNone of the four is a formality. Each one closes a question that arrives later, usually from someone who was not on site. The method used to take the readings is deliberately left open here — a recording car, a hand instrument or a marked reference section all work, provided the same method is used each cycle; the options are compared in verifying a grinding job without an inspection car.What the machine did is not recoverable from the rail afterwards, which is why the parameters have to be logged while the window is openThe questions that arrive later, and the entry that answers themWhat someone asks, months laterThe minimum entry that answers itWas the section inside the profile envelope when you handed it over?The post-work profile readings per section, filed against the target they were compared with — not against a nominal drawingWhy did this campaign take four passes when the last one took three?Pre-work deviation at the same sections and the depth each pass actually achievedDid that bluing come from the machine or was the rail already like that?Speed and dwell per pass, plus a surface note at the section taken before work startedWhy are wheels being ordered ahead of schedule?Pass-kilometres per millimetre per batch, per section — the figure reads as consumable life, never as rail lifeWas the interval too long, or was the last job too shallow?Tonnage carried since the previous campaign, next to the depth that campaign left behindWho accepted this work, and against what?The signed acceptance entry naming the criteria, the measured values and the person who signedTwo numbers are routinely misfiled, and both make a programme look better than it is. Pass-kilometres per millimetre describes how long a wheel lasted — on one high-speed campaign two sets ran at 4.28 and 3.27 pass-kilometres per millimetre, and neither figure says anything about the rail. Pass-kilometres per window hour describes throughput — the measured 18.72 pass-kilometres in a two-hour possession on a Chinese high-speed line is a productivity figure, and filing it as evidence of rail preservation is a category error.A record that cannot be compared with the same measurement from the previous campaign is a file, not a baselineHow long each record keeps earning its placeHorizonWhat is retrieved from the records at that pointWhy it cannot be rebuilt laterEnd of the shiftMetres covered against metres planned, passes run, and every stop with a reasonThe window is gone; a stop that was not written down becomes an unexplained shortfall in the next reportEnd of the campaignAcceptance values, the depth consumed, and the deviation that remainsThe rail has already been changed by the work, so the pre-work state exists only as the recordNext intervalThe tonnage and the residual deviation that together set the next pass countWithout it the interval becomes a calendar habit rather than a measured one — the arithmetic is in setting the interval by tonnageLife of the assetHow much of the design wear allowance has been spent across all campaigns, and how often each section has been workedThis is the only place the cumulative cost of grinding — in steel rather than money — is visible; see the safe grinding depth per railThe record set also settles commercial questions without argument, which is why it belongs in the contract rather than in good intentions: a per-kilometre rate can only be checked against the pass basis and window actually run, and the KPI structure that makes those numbers contractual is set out in what KPIs a grinding contract should include. A Molaton wheel batch is traceable by identity for exactly this reason — if consumption moves, the batch record tells you whether the wheel changed or the job did.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431, hold ISO 9001 / ISO 45001 / CRCC certification, and run batch traceability on every wheel that leaves the plant — the same discipline a grinding programme needs in its own records.Related questions you may also askGrinding acceptance standards and profile toleranceHow do you verify a grinding job without an inspection car?What acceptance criteria decide whether a grinding job passed?What KPIs should a rail grinding contract include?How long does a rail grinding wheel last per grinding cycle?How do rail grinding programs extend rail service life?More answers on rail grinding strategy and equipmentCan you reconstruct what your last campaign actually did?Send RailwayCare one shift of your existing paperwork, a profile survey and the wheel usage for that shift. We will show you which of the four record sets is missing, what the next interval can and cannot be set from, and how a Molaton wheel batch record closes the consumption question.Review My Grinding Records

Question 16

Welds need their own short stroke, lighter cut and separate finishing pass, with the stroke length taken from the weld's own excess height, not from the plain-track plan.

Accepted Answer

How should grinding be planned around rail welds?A weld is not a short length of plain track, and it must not be planned as one. It arrives with its own deviation from the target shape — excess height where it was not trimmed flush, a slight dip where it was — and with a heat-affected zone on either side. So the plan has to give it three things plain track never asks for: a short stroke, a lighter cut, and a finishing pass of its own. The parameter that decides whether that works is the grinding length, the distance the stroke runs out beyond the weld, and it is the number most plans leave blank. Too short and the cut ends in a step that the next wheel strikes; too long and the pass spends parent metal the interval cannot afford. The length is not a preference — it is derived from the weld's measured excess height and from the ramp the acceptance sheet requires.A weld is planned in centimetres while the rest of the section is planned in kilometres, and the two plans have to meet at the same passWhat a weld actually asks of the planZone in the planWhat the plan must give itWhy plain-track settings fail thereWho owns the decisionWeld crown and the first tens of centimetres either sideA stroke measured in centimetres, run at a lighter pressure than the main patternThe same head pressure applied over a few centimetres of harder, hotter metal removes more than intended and drives heat where the metallurgy is already alteredThe planner, from the measured excess heightThe heat-affected zone on both sidesShorter dwell and lower pressure until the finished surface stays clear of the bluing thresholdPlain-track speed and pressure were chosen for sound metal, not for a zone that has already been through a thermal cycleThe machine crew, under a rule written into the work orderThe run-out at each end of the strokeA ramp long enough to be a ramp — acceptance on Chinese networks calls for more than ‰1Ending the stroke where the wheel happens to reach depth leaves a step, and a step is a geometric defect, not a surface-finish item that can be polished awayWhoever signs acceptance, which is why the length appears on the sheetA weld inside a curve or a crossingBlending to the local rail shape at that point, not to the tangent-track targetA target copied from tangent track leaves the weld high on one leg or hollow on the otherThe planner together with whoever owns the target profile — see who defines the target rail profileA cluster: closure panels, plug rails, a crossingOne grouped zone with a single setup rather than a series of isolated pointsGrinding each weld as a standalone point multiplies setups inside the same possession and eats the window before the plain track is reachedThe possession planner, together with the grinding planThe reason length deserves its own line in the plan is arithmetic. On Chinese networks the weld zone is worked to roughly 0.2 mm per pass, while parent metal removal is capped at 0.5 mm across the whole job. A weld delivered with excess height above the cut a single pass can take at acceptable heat is therefore a multi-pass job by definition — and the answer is more passes over a longer stroke, never more pressure over a short one. Depth rules on their own are set out in what grinding depth is safe. How the length itself is measured, ramped and verified at the seam is the subject of its own field analysis: why grinding length is the unseen critical parameter in weld seam finishing.Length is the one planning input that no acceptance sheet records directly, yet it decides whether the finished weld is smooth or steppedThe sequence a weld should force on the work orderStepDecided before the window opensWhat it changes downstreamCount and locate the welds in the sectionPosition, welding method as delivered, and whether each sits on tangent, curve or crossingConverts the section into a list of zones with known geometry, which is what makes a per-weld stroke possible at allMeasure the excess height at each weldA reading taken before the machine arrives, not guessed from the previous campaignSets both the pass count and the stroke length, because the ramp has to bleed the height out over a finite distanceFix the stroke length and the rampLength in centimetres either side, with an end ramp steeper than ‰1Removes the two standard failures at once: the step left by a short stroke, and the hollow left by a stroke long enough to cut sound railSplit the pass allocationWhich passes the whole section shares, and which passes only the weld zone receivesStops the weld from absorbing passes budgeted for corrugation, and stops plain-track passes from re-cutting a finished weldWrite the finishing ruleLighter pressure, own stroke, own final pass — stated as a rule, not left to feelMakes the result repeatable across crews and shifts, which is what a programme needs and a one-off job does notRecord the outcome against the numberThe acceptance values for the weld itself: 0.3 mm under a 1 m straightedge, Ra at or under 10 µm, no continuous bluingTurns the next campaign's stroke length into a measured starting point instead of an inherited habitA weld is ground in a different rhythm from the rail around it, which is why the crew and the plan have to agree on where the zone begins. The same principle governs grinding a turnout without closing the line, where restricted zones inside a possession break the pattern in a similar way.The stroke ends where the plan says it ends, not where the operator judges the metal looks cleanWhat the finished weld tells you about the plan behind itSymptom at the weldWhat it usually meansWhat the plan got wrongWhat to change next cycleA visible step at the end of the strokeThe cut stopped at depth instead of ramping out of itStroke length was copied from plain-track practice or left to the operatorSet the ramp from the measured excess height and put a length on the work orderBlue or straw colour at the toe of the weldHeat went in where the zone was already alteredSame pressure and dwell as sound rail, or depth chased with pressure instead of passesLighter cut and shorter dwell in the zone; see preventing rail burn during grindingWeld still proud after a full campaignDepth was limited by the pass count rather than by the metal availablePasses were budgeted for the section and never allocated to the zoneGive the weld its own pass allocation instead of borrowing from the plain trackHollow either side of a weld that looks cleanThe stroke overshot and took parent metal on the rampsLength was set generously to be safe, without a depth budgetReconcile stroke length with the 0.5 mm parent-metal cap; a longer stroke needs lighter passes, not fewerA crack reappears at the same weld within one intervalEither the seam was not fully cleaned out, or the geometry still concentrates loadThe plan treated the weld as a finish problem instead of a profile problemRe-read the deviation, not just the surface; a repeat defect is a planning input, and the failure logic is set out in why grinding can make a defect worseTwo boundaries keep this a grinding question. Which consumable should be fitted to the weld zone is a separate decision, answered on that side: which wheel is used for weld seam grinding; a Molaton weld-zone wheel simply gives the plan the removal consistency the stroke length assumes. How the weld was made — alignment, preheat, cooling — belongs to rail replacement and welding, and grinding starts from whatever geometry that process delivered. And where the whole campaign sits in time is a separate arithmetic: interval set by tonnage.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431 for rail grinding wheels, are certified to ISO 9001 / ISO 45001 / CRCC, and every claim above is backed by on-track tests on high-speed, heavy-haul and metro networks, where weld zones are ground alongside plain track.Related questions you may also askWhy grinding length is the unseen critical parameter in weld seam finishingHow many grinding passes does a rail need?How is a grinding pass pattern set, and why does the angle sequence matter?What grinding depth is safe before you over-grind a rail?What acceptance criteria decide whether a grinding job passed?More answers on rail grinding strategy and equipmentDo your welds have a length on the work order?Send RailwayCare the weld count in your section, the welding method used and the excess height you measure at the seam. We will set out how the stroke and the pass allocation should be split, what a Molaton weld-zone wheel would hold at that depth, and where your current plan is most likely to leave a step.Plan My Weld Zone Grinding

Question 17

You do not need a recording car to accept a job: fix the sample plan first - a chord and template, profile readings at marked sections, and before and after at the same point.

Accepted Answer

How do you verify a grinding job without an inspection car?You do not need a recording car to accept one grinding job — you need a measurement plan that was fixed before the machine arrived. Three substitutes cover what a car would have told you about that job: a chord and template for longitudinal shape, profile readings on marked sections for the cross-section, and before/after readings at the same chainage for what the pass actually removed. The rule that makes any of them worth recording is the same: the sample must be the same physical point before and after, marked on the rail, so the two readings describe one place instead of two neighbouring places that differ by more than the tolerance being judged.Verification is decided by where the readings are taken, not by how many of them existWhat each substitute can prove, and what it cannotSubstitute for the carThe number it givesWhich acceptance point it provesWhat it still cannot tell youA chord and a template laid across the headGap under a 1 m straightedge, and the gap between the head and the target shape at that pointLongitudinal irregularity against the 0.3 mm in 1 m limit, and whether the local shape matches the target drawingNothing beyond the length of the straightedge, and nothing about the next kilometreProfile readings at marked sectionsCross-section of the head, zone by zone, compared with the target: +0.2 / −0.4 mm at the crown centre, ±0.2 mm in the other zonesWhether the geometry that was ordered is the geometry that was delivered, at the points measuredWhat happened between the sections; a single skipped band can sit between two passing readingsBefore/after readings at the same chainageRemoval achieved per position: crown, gauge corner, field sideThat the plan was followed and the parent metal was not cut past the 0.5 mm limitWhy a position was missed — it is a result, not a diagnosisSurface and heat inspection of the finished headHeat tint and burn on the running surface, roughness where it is measured, the position and width of the polished bandThat no burn was left behind, that roughness met the 10 µm limit, and that the band sits where the target wanted it (25–30 mm across a high-speed turnout, 35–40 mm on conventional track)Any depth figure at all; it is a screen that decides where to measure properlyThe sampling position is not a formality. On the same high rail, after the same six passes, removal measured 0.3 mm at one position and 1.8 mm at another. The spread across one head is far larger than the ±0.2 mm zone tolerance, so a reading taken a hand's width up or downstream is not a slightly different measurement — it is a different answer. That is the whole reason the car is convenient: it always knows where it is.Repeating a reading at the wrong place looks like measurement and behaves like noiseFix the sample points before the machine arrivesStepWhat is doneThe rule that keeps it validWhat goes wrong if it is skipped1. Split the job into blocksDivide the campaign length into work blocks and place at least one section in each, plus every section with a known defectSections are chosen from the plan and the defect list — never from how the finished surface looksSections cluster where the result was already good, and the report describes the easy part of the job2. Mark them on the railChainage, rail side and a painted mark across the head, with the orientation of the gauge notedTwo marks per point so the instrument goes back in the same place and the same orientationReadings drift by centimetres between visits and the removal figure becomes meaningless3. Read before the machineStarting shape, band position, defects and roughness at every marked pointSame day, same instrument, same person, whole set togetherDepth has to be estimated afterwards, and estimation always flatters the result4. Read after, in the same orderRepeat every reading at the marked point, before any re-work discussionThe pair at one mark is the unit of evidence; nothing is compared across marks"Improvement" appears where the second reading simply landed on a different part of the head5. Decide block by blockPass, fail or unverified, per block, with the numbers attachedOne failed section in a block makes the block unverified until the section is re-readA clean average hides a missed band that will be the first thing the next inspection findsWhere a previous campaign failed, put sections there at double density. A car gives continuous data along the whole length; a sample plan gives honest data at chosen points, and its weakness is precisely the length between them. Density is how you buy back that confidence, and it is set before the possession, not after.The record is what makes a sample plan defensible a year later, when the surface itself has already changedWhat the handover record has to containRecord itemWhat it must showThe decision it supportsSection scheduleChainages, both rails, sections per work block, and how each was chosenWhether the sample was neutral enough to accept the block at allStarting profile per sectionDeviation from the target profile, zone by zone, taken before the first passWhich depths the plan was entitled to takeAchieved removal per positionCrown, gauge corner and field side, next to the planned figureWhether the envelope held, in particular the 0.5 mm ceiling on parent metal — see safe grinding depthSurface and heat resultHeat tint, roughness readings, band position and width across the headWhether the job can be signed at all, or holds for re-work — see acceptance criteria for a grinding jobPass pattern actually runOrder of the bands cut, and every departure from the planned patternWhy the removal figures differ by position — see how a pass pattern is setNon-conformance and dispositionSection, defect, re-work carried out or deviation accepted, with its justificationWhether the block closes, reopens, or moves to the next interval as a known defectThe sheet is signed by both parties at handover, while the machine is still on site. Two boundaries keep this honest. First, a sample plan cannot certify track it did not touch: if the report has no reading for a block, that block is unverified, not passed. Second, verifying a shape is not the same as defining it — the numbers above are judged against the target profile and its tolerances, which come from the operating authority, as set out in what the target rail profile is and who defines it. Where a campaign covers turnouts under traffic, the sections have to follow the sequence in grinding a turnout without closing the line. Which instrument a network standardises on is an equipment decision rather than a verification-plan decision; the wider acceptance standards are reviewed in grinding acceptance standards. Removing the same depth on every shift is the other half of comparability: a Molaton wheel keeps its cut from the first block to the last, so two readings taken a week apart describe the same plan rather than two different ones.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431 for rail grinding wheels, are certified to ISO 9001 / ISO 45001 / CRCC, and every claim above is backed by on-track tests on high-speed, heavy-haul and metro networks.Related questions you may also askWhat acceptance criteria decide whether a grinding job passed?What is the target rail profile and who defines it?What grinding depth is safe before you over-grind a rail?How is a grinding pass pattern set, and why does the angle sequence matter?How many grinding passes does a rail need?More answers on rail grinding strategy and equipmentWant the sample plan before the possession?Send RailwayCare the campaign length, the defect list and the target profile you are working to. We will return a section schedule with chainages and rails marked, the readings each point has to produce, and the record sheet a Molaton wheel schedule has to be signed against — so the job can be accepted without waiting for a recording car.Get a Verification Plan

Question 18

Cost per kilometre is a division, not a rate card: machine time and consumables divided by kilometres actually finished. Passes, speed, possession use and wheel life move it.

Accepted Answer

What does rail grinding cost per kilometre, and what drives it?Cost per kilometre is a division, not a rate card. The numerator is what the machine, crew and consumables cost for the time the track is handed over; the denominator is the kilometres you actually finish inside that time. Four things move either side of that line — passes, working speed, possession utilisation and wheel consumption — and the cheapest quoted rate usually wins only by assuming the best possible value for all four at once. No unit prices are given here, because they belong to a specific network, machine class and contract; the structure is what makes two quotes comparable.The rate and the result come out of the same possession, which is why the denominator deserves as much scrutiny as the priceThe four drivers, written as one divisionDriverWhere it sits in the divisionWhat moves itWho controls itPassesMultiplying the denominator: work is counted in pass-kilometres, so a section ground three times is three times the distance billed for the same track lengthHow far the measured head sits from the target shape, and whether the plan is built on that measurement or on last year's number — see how many grinding passes a rail needsThe planner, from the profile survey — this is the only driver the owner can cut without touching qualityWorking speedSets the denominator directly: at an unchanged depth per pass, a higher working speed buys more kilometres from the same windowDepth per pass and the heat balance that limits it, so speed cannot be raised on its own — the trade is set out in preventing rail burnThe machine and the pass plan, jointly; not a commercial leverPossession utilisationDecides how much of the window is productive: two hours of occupation is never 120 grinding minutesTravel to and from the section, waiting for isolation, pattern resets and unplanned stopsOperating and possession planning, before a contractor is pricedWheel consumptionThe only driver normally quoted as consumable life, in pass-kilometres per millimetre, which is how possession time converts into consumable spendRemoval achieved per pass, and whether depth is reached by the plan or chased with pressureWheel performance plus depth discipline — the interface where a Molaton wheel specification and the grinding plan have to agreeAn anchor makes the arithmetic visible without quoting a price. On one Chinese high-speed line a GMC-96X produced 18.72 pass-kilometres inside a two-hour possession — 9.36 pass-kilometres per hour of window, not per hour of grinding. A 20 km section ground three times is 60 pass-kilometres, roughly 6.4 window hours at that measured rate; dropping the third pass takes 20 pass-kilometres off, about a third of the time, the wheels and the occupation together. That is where the largest savings sit — on the denominator, not on the rate.Every extra pass is paid for twice: once in possession time and again in the consumable that had to do the cuttingWhy the lowest rate per kilometre tends to cost the mostHow the rate is writtenHow it looks in a tenderWhat actually comes back laterWhose budget it lands onA rate tied to a fixed number of passesThe most comparable number in the roomThe bid built on the fewest passes is the cheapest; when the measured deviation needs one more, the extra pass returns as a variationThe owner's maintenance budget, in-yearA rate that excludes wheel consumptionClean, low, easy to defendConsumption is the input that rises first when depth is chased with pressure instead of planThe owner, as re-ordered wheels or extra shiftsA rate built on a full two-hour windowIdentical for every bidderWhen windows shrink, a fixed rate per kilometre is recovered through fewer metres per shift or a re-mobilisationThe owner, as a section left unfinished into the next cycleA rate that stops at the last pass, not at the profileInvoice matches the work orderAcceptance failure is outside the price, so the re-work is a second campaign inside the same intervalThe owner, and again in the next intervalA rate with mobilisation priced separatelyThe machine day looks efficientOn networks of many short or remote sections, mobilisation repeats while kilometres stay modestThe owner, per visit, for the life of the contractNone of this argues for a dearer contractor. It argues for reading a per-kilometre figure against the pass basis and the window it assumes. A Molaton wheel can hold removal so the planned depth is reached inside the planned passes; it cannot make a rate built on two passes pay for three.Comparability comes from the stated basis of the rate, not from the size of the number in the comparison columnWhat a per-kilometre rate must state to be comparableLine in the rateWhat it must stateWhy it changes the numberWho carries the risk if it is silentPass basisPlanned passes and the depth each pass is expected to takeA rate for two passes is not a rate for three, and the difference is a third of the jobThe owner, as a variationMeasurement basisWhich sections were surveyed, how the starting deviation from the target profile was takenDepth is derived from that deviation, so the survey sets the passesThe contractor first, then the owner as a claimWindowWindow length per shift and the productive share assumedThe denominator is window hours, not machine hoursThe owner, as unfinished metresWheel basisWhether consumption is inside the rate, and the pass-kilometres per millimetre it assumesConsumable life varies by section: 4.28 against 3.27 pass-kilometres per millimetre on two sets in the same high-speed campaign, and that figure describes the wheel, never the railThe owner, as re-ordersAcceptanceWhich criteria, measured by whom, and what happens on a failA failed job is a second campaign, priced or not — criteria are set out in acceptance for a grinding jobWhoever the contract leaves holding itRe-mobilisationHow short, remote or split sections are pricedOn a network of many short sections this line can outweigh the rate itselfThe owner, every visitRecordsWhat is handed over per shift, and in what formNothing above can be checked after the fact without itThe owner, permanentlyTwo adjacent questions are deliberately excluded. What the wheel itself costs per kilometre of track is a consumable calculation, answered on that side: wheel cost per kilometre of track. What a grinding machine costs to buy is an equipment question: the cost of a rail grinding machine. This page is the cost of the job, which is the rate multiplied by passes, then divided by whatever the window actually delivers. Set the interval first — interval by tonnage — because a cheap campaign run too often costs more per year than a properly scoped one.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431 for rail grinding wheels, are certified to ISO 9001 / ISO 45001 / CRCC, and every claim above is backed by on-track tests on high-speed, heavy-haul and metro networks.Related questions you may also askHow to calculate rail grinding wheel cost per kilometer of trackHow many grinding passes does a rail need?What KPIs should a rail grinding contract include?Should you buy a grinder or contract the grinding out?How do rail grinding programs extend rail service life?More answers on rail grinding strategy and equipmentWhat is your cost per kilometre actually made of?Send RailwayCare the section length, the planned passes, the window you can offer and the profile survey you are planning from. We will set out which of the four drivers is inflating your figure, what a Molaton wheel schedule would consume at that depth, and how the same work could be scoped in fewer passes.Break Down My Cost per Kilometre

Question 19

Grinding adds no steel. It moves the running band off the gauge face onto the crown, and stops being the right answer once the remaining head cannot hold the target shape.

Accepted Answer

How does grinding address rail side wear?Only in one way: it moves the running band back onto the crown. Side wear is not a stain that can be cut away — it is the result of the wheel contacting lower and lower on the gauge face, and every millimetre of steel that made that happen is already gone. Grinding therefore works on the shape, not on the worn area: the gauge corner is relieved and the crown is brought back so the wheel is invited to run higher. Adding passes to the same shape does not slow side wear down, it just narrows a head that has already lost section.The correction is a change of contact position, so the geometry is the deliverable rather than the finished surfaceWhy more passes do not slow side wearWhat operators tryWhat it looks like afterwardsWhat actually happened in the steelWhat to do insteadGrind the shiny gauge face harderClean, bright, uniform metal on the faceThe wheel still rides at the same low point; the head is now narrower and contact stress is higher than beforeGrind to the target profile, and treat brightness on the gauge face as a symptom rather than an instructionRepeat the plain-line profile for another cycleAll zone tolerances pass on paperNothing in the geometry invites the band upward, so wear returns at roughly the previous rateUse the profile that carries gauge-corner relief for that curve class, and check where the band landsTake a deeper cut on the high rail onlyThe high rail reaches tolerance firstDepth per pass may pass the envelope while the other rail is left outside its own toleranceBudget removal per rail and per position inside the same envelope, not per campaignChase a uniform band with extra passesThe surface looks even across the headYou have polished the band that was already wrong, and spent section that could have held the correctionStop and measure the band position against the crown before adding anythingThis is the mechanism that separates side wear from corrugation, which is an amplitude problem solved by cutting peaks, and from a gauge-corner crack field, which is a depth problem solved by taking the cracked layer while it is still shallow. Side wear is a position problem: the only lever long enough to reach it is the target profile.Once the band sits on the crown, the surface condition stops being the deciding evidenceThe sequence that actually shifts contactStepThe number you needWhere it comes fromWhat goes wrong if you skip it1. Measure the worn shape, not just the wearHead width, gauge-face angle, vertical loss at the crown centre, and the present band positionA profile gauge across the head, taken at fixed chainages on both railsYou plan depth against a single wear figure and never learn where the contact actually sits2. Fix the target the work is judged againstThe target profile drawing for that curve class, with its tolerance bandThe operating authority's maintenance specification — see what the target rail profile is and who defines itEach shift grinds to whatever the last one left, and side wear comes back without anyone changing a parameter3. Split the depth by positionTwo depths per rail: gauge corner and crown, each computed from its own deviationMeasured removal per pass at the working speed on that curve, capped by the envelope: 0.1 mm upward as a planned preventive depth, 0.2 mm upward as a corrective one, total removal on the parent metal no more than 0.5 mm (safe grinding depth)One average depth either leaves the gauge corner proud or over-cuts the crown that was already correct4. Run the angle sequence so relief comes before finishingThe order in which the bands are cut, and which band each head angle coversThe pass pattern rules in how a pass pattern and angle sequence are setFinishing the crown first removes the material the gauge-corner relief still needed5. Prove the band movedBand position relative to the crown after the pass, plus zone tolerances: +0.2 / −0.4 mm at the crown centre and ±0.2 mm in the other zonesThe post-grind profile report, per rail, checked against acceptance criteria for a grinding jobA job can pass on roughness and still leave the contact point where it was, so the wear clock restarts immediatelyThe consumable side of this is narrower than it looks. A Molaton wheel does not change the target profile; what it has to do is hold removal on the gauge side — where the contact arc is longest and the cooling worst — so the planned relief is reached without adding passes or pressure. A wheel that fades halfway through forces exactly the behaviour this page is warning against. Matching a wheel format to a machine is a separate question, answered on the consumable side: rail grinder wheel selection.There is a point where the remaining head can no longer carry the shape the curve needs, and grinding cannot move itWhen grinding stops being the answerCriterionHow it is measuredWhat it means at that levelWho decides the next stepWear approaching the limit set for that rail sectionVertical and side wear against the limit for the section in use — the maintenance rules set them per rail type, for example 9 mm vertical and 15 mm side on 43 kg/m rail, 10 mm and 17 mm on 50 kg/mGrinding restores shape, never section: past this point each campaign spends the rail-head allowance that was reserved for the next oneThe infrastructure owner, against its own wear table and replacement planThe target profile can no longer be reached inside the envelopeRequired gauge-corner relief compared with the removal still allowed on that headGrinding to tolerance would only flatten the surface; the band cannot be brought back up, so wear continues regardless of qualityOwner plus the grinding planner, before the campaign is booked — see grinding versus replacing a railCracking or crushing alongside the wearCrack depth against the roughly 0.2 mm window in which a pass still removes the root as well as the surfaceDeeper than that, the surface comes out clean and the root stays, which is a fatigue question rather than a shape question (how grinding prevents rolling contact fatigue)Owner's engineering function, with the rail scheduled for attentionRepeat campaigns with no fall in wear rateWear rate per 100 MGT on the same sections across consecutive cyclesEither the profile is not being reached or the traffic has changed; neither is fixed by another passOwner's maintenance planning, together with the grinding contractorOne habit keeps this decision honest: a speed restriction is an operating measure, not a grinding parameter, and it should never be used as a way to postpone a replacement that the section already justifies. While a rail waits for replacement, the sensible grinding scope is to hold shape and keep the band off the gauge face — not to chase the original profile with section that no longer exists. How side wear behaves across a whole curve, including the outer rail's longer contact arc, is set out in managing side wear on curved track.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431 for rail grinding wheels, are certified to ISO 9001 / ISO 45001 / CRCC, and every claim above is backed by on-track tests on high-speed, heavy-haul and metro networks.Related questions you may also askWhen should you grind a rail instead of replacing it?What is the target rail profile and who defines it?How is a grinding pass pattern set, and why does the angle sequence matter?What grinding depth is safe before you over-grind a rail?What acceptance criteria decide whether a grinding job passed?More answers on rail grinding strategy and equipmentIs the profile still able to move the contact band?Send RailwayCare one measured profile from the worn high rail and the target drawing for that curve. We will tell you how much gauge-corner relief is still available, whether the band can be brought back to the crown, and what a Molaton wheel schedule would need to remove to get there without extra passes.Check My Side Wear Case

Question 20

A curve is not a tangent job with more passes. Two rails, two contact problems, one depth budget: the plan changes in shape, in where metal comes off, and in how soon it returns.

Accepted Answer

Should curved track be ground differently from tangent track?Yes, but not because a curve needs more passes. On tangent track one target profile and one depth figure fit both rails. On a curve the two rails carry two different contact problems and give up metal in different places, so the plan has to change in three things only: which shape each rail is ground to, how the depth budget is split across the head, and how soon the section comes back. The arithmetic underneath does not change — depth still comes from measured deviation, and total removal still stays inside the same envelope.On a curve the same pass plan meets two different contact conditions a few metres apartWhat changes from a tangent plan, and what does notItemTangent trackCurved trackWhat happens if you reuse the tangent planTarget shapeOne profile applied to both rails; the contact band sits near the crown by itselfTwo targets: the high rail needs the gauge corner relieved so the band climbs back toward the crown, the low rail needs a crown that stays wide and centralThe high rail is ground smooth on the gauge face while the wheel keeps running on it, so side wear returns at close to the previous rateWhere the metal comes offClose to symmetrical across the headDeliberately unequal: on the high rail of an R=800 m curve, six passes at 12 km/h took up to 0.3 mm on the field side but up to 1.8 mm on the gauge side — a six-fold spread inside a single pass planOne average depth per rail over-grinds the side that was already close and leaves the gauge corner still proudHow often it returnsThe long clock: about 100 Mt on plain line and large radius above 1,200 mThe short clock: 30-50 Mt on small-radius curves; in UK practice curves are planned near 15 Mt against about 45 Mt on straight trackThe curve arrives at the next inspection already outside the band, and the correction then costs section instead of preventing itDefect mix to read firstCorrugation, squatting, rail-head fatigue spread across the headHigh rail: side wear and gauge-corner fatigue. Low rail: corrugation and surface crackingYou plan for the defect you can see from the cab instead of the one that decides the profileThe tonnage numbers above are conditional ranges, not a schedule: the transferable part of the UK comparison is the ratio of roughly three to one between curve and straight, which is what a plan should carry over to another network — not the absolute tonnage. How the interval itself is calculated from a removal rate is set out in how a grinding interval is set by tonnage.Two rails, two jobs: the shape on each side of a curve is chosen separately, not averagedHigh rail and low rail are two different jobsRailWhat it losesThe grinding objectiveWhat to check afterwardsHow a wrong plan looksHigh (outer) railMetal from the gauge face and gauge corner, plus rolling-contact fatigue at the cornerProfile recovery: move the contact band up and off the gauge face, and take the cracked layer with itHead width, gauge-face angle, and where the running band actually sits relative to the crown — not just how bright the surface isThe gauge face looks clean and even polished, yet the band is still low and side wear continuesLow (inner) railCorrugation amplitude, surface cracking, light plastic flow on the crownAmplitude removal: cut the peaks below the valleys instead of lowering the whole surface by one depthThe residual amplitude measured on the same chord at the same position, before and afterThe surface looks freshly ground and the corrugation is back within a season, because the peak-to-peak distance was never reduced (see why grinding can make corrugation worse)This is also where curve work splits between machine types. A long machine works along the length and applies one angle set to every head; on the short transitions for a small radius, or where the high rail's gauge corner needs a shape the main head cannot reach, part of the job belongs to a smaller unit. That division is covered in large-machine and small-machine grinding, and the geometry of which angle covers which band of the head is in how a pass pattern and angle sequence are set.Unequal removal across a curve is the expected result of a correct plan, not evidence that the job went wrongUneven removal is normal — check these insteadWhat the report showsNormal or a faultHow to verify itWhat to changeGauge side lost several times the metal of the field sideNormal on any curve carrying side wear — the wear was there before the machine arrivedPer-position measurement on that rail, not a fleet or section averageNothing, provided the total stays inside the envelope and the contact band has moved upOne rail inside tolerance, the other still outsideNormal; the two rails started from different shapesTwo separate tolerance reports, one per railPlan and accept the two rails separately from the startWhole curve still short of target after the planned passesA faultCompare the removal actually achieved per pass with the figure the pass count was computed fromPasses were calculated from a tangent-line removal figure; recompute from measured removal at that speed on that curveGauge corner shiny again within weeksA fault, and not a grinding-quality faultMeasure where the running band sits relative to the crownThe plan changed the surface but not the contact position — go back to the target profile drawing, covered in what the target rail profile is and who defines itTwo numbers stay shared with tangent work and are the reason a curve job is not a licence to cut deeper. Planned depth runs from 0.1 mm upward for preventive work and 0.2 mm upward for corrective work, with total removal on the parent metal capped at 0.5 mm, and the working window for surface cracks is about 0.2 mm deep — the same figures as on plain line, detailed in safe grinding depth. What a Molaton wheel has to deliver for a curve plan is narrower than it sounds: the same removal per pass on the gauge side, where the contact arc is longest, from the first metre to the last. A wheel that fades there forces the operator to add passes or pressure, and both of those spend section the curve does not have.Where a curve also carries corrugation, the set of defects becomes a combination rather than one item, and the treatment order changes; that case, including the specification side of it, is handled in managing side wear on curved track and in wheel selection for small-radius curve corrugation, which belongs to the consumable question rather than the strategy one.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431 for rail grinding wheels, are certified to ISO 9001 / ISO 45001 / CRCC, and every claim above is backed by on-track tests on high-speed, heavy-haul and metro networks.Related questions you may also askWhat is the target rail profile and who defines it?How is a grinding pass pattern set, and why does the angle sequence matter?How is a rail grinding interval set by tonnage?Why does grinding sometimes make corrugation worse?What grinding depth is safe before you over-grind a rail?More answers on rail grinding strategy and equipmentDoes your curve plan treat both rails as one job?Send RailwayCare the curve radius, the two rails' measured profiles and the last pass plan. We will tell you which rail is being averaged into the other, where the depth budget is being spent, and what a Molaton wheel schedule that holds removal on the gauge side would look like for that section.Review My Curve Plan

Question 21

Burn is heat that could not leave the rail before the next head arrived. Depth, speed, head pressure and cutting time decide it. Blue on the surface means stop, not slow down.

Accepted Answer

How do you prevent rail burn during grinding?Burn is a heat balance, not a fault in the wheel. Every pass puts heat into the rail, and that heat has to leave through the steel before the next head arrives at the same point. Four things decide whether it does: depth per pass, working speed, head pressure and how long one section keeps being cut. The field rule needs no instrument: once the surface turns blue the steel has already passed the temperature where its surface layer changes, so you stop and re-plan the pass instead of finishing it more slowly.Whether a pass leaves heat damage is decided before the machine moves, by four settings and one clockThe four variables you actually controlVariableWhat it changesWhich direction raises the heatThe limit or check you already haveDepth per passThe volume of steel taken per unit length, and therefore the energy any one head has to put into the surfaceDeeper cut at the same speedPlanned depth runs from 0.1 mm and up for preventive work to 0.2 mm and up for corrective work, with single passes typically 0.2-0.3 mm; total removal on the parent metal is capped at 0.5 mmWorking speedThe time the contact arc spends on any one point, and at a fixed depth the power needed to hold the cutFaster at the same depth, because removal rate per hour goes up with itSpeed should come from removal measured at that speed on that line, never from the schedule aloneHead pressureHow hard the wheel is pressed into the steel, and how much of the work turns into rubbing instead of cuttingPressure above what the cut needs, which is a friction source rather than a removal sourceThe standard keeps contact pressure below 10 kN; on a qualification run at 30% power we measured 3,120 N, so the cap is a ceiling, not an operating pointContinuous cutting timeHow many times a section is re-heated before it gets a chance to coolLong runs with no stop and no change of patternSet a distance per run and an inspection stop after it, and record the stop on the shift sheetThe four are not equal levers. Depth multiplies the heat directly and is the one number the plan already fixes; pressure and speed mostly decide how much of that energy becomes cutting rather than friction. That is why the correction for a hot job is almost never a single knob: take depth down first, then check whether the pattern is stacking cuts on one band, and only then look at speed.The surface after the pass is the record of what the heat balance did, and it is readable without an instrumentRead the surface instead of the gaugeSurface after the passRail temperature it indicatesWhat has happened in the steelWhat to do nowNormal steel, no discolourationBelow about 471 °CNothing structural; the pass stayed within the removal the plan asked forCarry on, and keep the acceptance record on roughness and profileStraw to light yellowAbout 471-600 °CThe first stage of change at the surface, and roughness is already climbingReduce depth on the next pass and re-check roughness before continuing; do not add passes to make up for itBlueAbout 600-735 °CThe band in which the surface layer transforms into the hard, brittle layer that spalls laterStop. Re-plan depth, speed and the number of cuts landing on that point — a blue streak is not a cosmetic mark to be ground out at the same settingNo colour, roughness drifting upBelow the visible threshold, so the colour gives no warningInvisible heat damage: on trials, taking spindle speed from 1,000 to 3,000 rpm moved the white layer from about 45 to about 320 µm and roughness from 6.6 to 8.4 µm, with spalling appearing at the top of that rangeTreat a roughness trend as the warning the colour did not give, and pull speed back before the next runThe fourth row is the reason a colour check alone is not an acceptance method. Oxidation confirms what the eye already sees: oxygen content on the surface rises as the damage gets worse, and it rises further on the underside of the removed chips than on the rail, which is what you would expect if the heat is generated in the cut and left in the steel. The acceptance standard therefore does not ask whether the rail looks acceptable. It asks for no continuous blue, a workpiece temperature below 150 °C, and no cracking, breakage or abnormal vibration during the work — measured, not judged by eye.When the colour changes, the order in which you change the settings is what decides how much section the correction costsThe order to change things in when heat shows upWhat you seeFirst thing to changeWhy that one firstWhat not to doBlue along the crown for the whole lengthDepth per passIt is the variable that scales the heat directly and the one the plan already fixes, so the correction is arithmetic rather than guessworkDo not slow the machine to hide it: slower at the same depth keeps the heat on the point for longer, not shorterBlue in one zone only, typically the gauge cornerThe pattern, not the depthHeat appearing in one band means too many cuts landed on that band within one run, while the rest of the head was barely touchedDo not take depth off the whole head; the zones that were already on target will fall shortLight yellow that appears late in the runContinuous cutting time, then pressureA late onset is the signature of heat accumulating in a section that keeps being re-cut rather than a single overloaded passDo not run on to finish the shift and inspect afterwardsNo colour, but roughness above 10 µm or above the 8 µm contract targetSpeed, then the state of the cutting faceRoughness is the earliest visible evidence that energy is going into rubbing rather than cuttingDo not sign the job off on colour aloneSpalling or a crack found after a hot jobStop and re-open the sectionOnce the layer has transformed, the damage shows up after the wheel has gone; the surface you are looking at is the consequence, not the defectDo not grind it again at the same setting to smooth it overNotice what this table does not contain: any change to the wheel. Whether a consumable is suited to your rail is a real question, but it is a different one, and it is answered on the diagnosis side in why a grinding wheel burns the rail surface and in our article on burning prevention and wheel choice. What a Molaton wheel has to do for this page to be useful is narrower: hold the same removal per pass from the first kilometre to the last, because a wheel that stops cutting mid-job is exactly what tempts an operator to raise pressure and turn a depth problem into a heat problem.Two of the numbers on this page are also the two a contract should ask for. Depth is checked against the safe grinding depth, and the heat result is checked with the rest of the job in the acceptance criteria for a grinding job. If the blue has already appeared on a section that was ground last season, the sequence that produced it is usually a pass plan copied from another line, which is covered in how many grinding passes a rail needs.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431 for rail grinding wheels, are certified to ISO 9001 / ISO 45001 / CRCC, and every claim below is backed by on-track tests on high-speed, heavy-haul and metro networks.Related questions you may also askWhy does my rail grinding wheel burn the rail surface?What acceptance criteria decide whether a grinding job passed?What grinding depth is safe before you over-grind a rail?How do you control white etching layer by grinding?Preventive vs corrective rail grinding: what is the difference?More answers on rail grinding strategy and equipmentIs your plan asking for more heat than the rail can shed?Send RailwayCare your depth per pass, working speed, pressure setting and one photo of the finished surface. We will tell you which of the four variables is out of budget, what to change first, and which depth your line can still take — with a Molaton wheel schedule built to hold removal without pushing you into pressure.Review My Heat Budget

Question 22

A head tilted at one angle touches only a band of the rail head. A pattern is a set of overlapping bands, and the sequence decides whether peaks go first or the shape does.

Accepted Answer

How is a grinding pass pattern set, and why does the angle sequence matter?A grinding head never touches the rail head; it touches a band of it. A pattern is therefore a set of bands laid side by side from the gauge corner round to the field side, and it only works if adjacent bands overlap — a strip that falls between two angles is never cut by either. The sequence matters because the order fixes what the first passes are allowed to do: take the peaks off while the head is still rough, or bring the head into the target shape first and trim the peaks afterwards. Choose the order for the state the rail is actually in, and verify the result by position, not by pass count.What a pattern can and cannot do is decided by where each band lands on the head, not by how many passes were runWhy one angle cuts one band, not the headBand on the headWhat the head has to doTolerance it has to meetSignature when the strip is missedRunning surface, crownPresented roughly square to the running surface, so the band straddles the centre of the headCrown centre inside +0.2 to -0.4 mm of the target profileThe running band measures narrower than the 25-30 mm expected on high-speed track or 35-40 mm on conventional and heavy haul, and the wheel rides the edge of itGauge cornerTilted towards the gauge corner, so the band wraps the shoulder the wheel loads on curvesZones 2 to 4 within ±0.2 mm, side profile includedA flow lip and side wear keep growing, and the crack source the pass was meant to remove stays below the depth a later pass can reachField side shoulderTilted the other way, covering the outer shoulder used when the wheelset shifts outboardZones 2 to 4 within ±0.2 mmContact migrates inboard, the shoulder carries no load, and the head settles into a shape narrower than the drawingBetween those threeIntermediate angles whose only job is to make the bands above meetNo periodic wheel marks on the finished surface, roughness Ra ≤10 µm, or ≤8 µm where the contract asks for itA raised stripe that no angle reached, visible as a line after the job and as a periodic mark that comes back with the next wheelThe fourth row is where most patterns quietly fail, because the three main bands can all be correct while the joins between them are not. Overlap is not a refinement, it is the condition that turns a set of angles into a surface. Band width per angle is fixed by the wheel diameter and the tilt, so with a given number of heads the spacing either overlaps or it does not — there is no operating skill that closes a gap the geometry left open.Overlap between adjacent bands is what turns a list of angles into one continuous surfaceWhy the order is not interchangeableOrder chosenWhat the first passes are allowed to doWhich rail condition it suitsWhat it costs on a rail in the wrong conditionPeaks first, shape secondCut the high spots and flow lips while the head is still far from the drawingHeavy corrugation, deep lips, a head that has drifted a long way out of bandThe peaks are gone but the shape is still wrong, so the finishing angles have to take metal from the whole head to reach the bandShape first, peaks secondBring the head inside the target band, then trim what still stands above the crownA head close to the drawing with only surface irregularity leftA peak that stands above the band is outside the reach of the finishing angles, so the pattern is spent and the peak survivesCrack depth first, shape secondRemove the layer the cracks start from before anything is formedRolling contact fatigue, or a white etching layer at the surfaceA perfect profile cut over a crack root that is still below reach simply moves the same defect into the next intervalAmplitude first, profile laterSet the crest target from the measured amplitude itselfWavelength confirmed and amplitude measured on the same chord and positionA uniform depth over crest and trough removes metal from both and leaves the amplitude exactly where it wasRead the rows as answers to one question: what is standing between this head and the drawing. If it is metal above the band, peaks go first. If it is a surface layer the cracks live in, the layer goes first even though it is not the biggest dimensional problem. If it is a wave, the target for the crest is the amplitude, because only a differential cut changes a wave — the arithmetic behind that is set out in our answer on why grinding sometimes makes corrugation worse.The order is written in the plan and checked against the head, not left to the operator on the nightSetting the bands for your own lineStepInput you needWhat it decidesCheck before you leave1. Measure the headA transverse profile taken on the section before the job, on the same chord and position afterwardsWhich zones are already out of band, and therefore which bands carry the removalThe before and after traces are from the same section, not from two convenient ones2. Count the sectionsThe zones the acceptance standard already names on the headHow many bands the pattern needs before overlap is even possibleEvery named zone has at least one angle assigned to it3. Set the spacingThe width each angle actually cuts on this machine at this wheel diameterWhether adjacent bands overlap or leave a stripNo gap between bands anywhere from gauge corner to field side4. Order the sequenceDefect type, wave amplitude, and how far the head sits from the drawingWhether peaks go first or the shape doesThe purpose of the first pass is written in the plan, not chosen on the night5. Verify by positionRemoval measured per zone rather than per passWhether the pattern did what this table said it wouldPer-zone removal matches the plan and no periodic mark appearsStep five is the one that turns a pattern from a setting into a strategy. On one curve at R=800 m, a single six-pass plan at 12 km/h took up to 0.3 mm from the high-rail field side and up to 1.8 mm from the low-rail side — one pattern, six times the removal in one place. Our own cutting trials show the same thing in a different form: on the same machine and the same setting, measured removal ran 0.142-0.241 mm on the inner rail, 0.165-0.326 mm on the outer rail and 0.193-0.222 mm at the crown. A pattern is therefore set per position, and a pass count copied from another line tells you nothing about what any one zone received. The bands only make sense once the target they are aiming at is fixed, which is covered in our answer on the target rail profile, and the depth ceiling they may not cross in our answer on safe grinding depth.What a Molaton wheel has to deliver for this to hold is narrow and testable: the removal per pass at your working speed, and the same removal held for the whole job instead of fading as the wheel wears. A wheel that loses its cut halfway through leaves the last kilometres of the plan uncut, and the unground strip appears in the zones nobody thought to check. If you want the pass count question on its own, it is answered in how many grinding passes a rail needs; how a restored contact surface is judged afterwards is in our article on restoring the contact surface.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431 for rail grinding wheels, are certified to ISO 9001 / ISO 45001 / CRCC, and every claim below is backed by on-track tests on high-speed, heavy-haul and metro networks.Related questions you may also askHow many grinding passes does a rail need?What is the target rail profile and who defines it?What acceptance criteria decide whether a grinding job passed?Why does grinding sometimes make corrugation worse?Rail profile grinding: techniques, standards and practiceMore answers on rail grinding strategy and equipmentDo you know which zone your pattern never reached?Send RailwayCare the before-and-after profile from one section, the pass list you ran and your working speed. We will show you where the bands overlap, where a strip was left uncut, and what order your head actually needed — with a Molaton wheel schedule sized to hold that removal across the whole job.Review My Pass Pattern

Question 23

Four parts, one programme: the wheels that touch the rail, the machines that reach what a long one cannot, the vehicles that supply throughput, and the measurement that proves it.

Accepted Answer

Which rail grinding and maintenance solutions does RailwayCare supply?Four parts that only make sense together: the grinding wheels that actually touch the rail, the grinders that reach where a long machine cannot, the grinding vehicles and trains that give throughput, and the inspection that turns a job into a signed record. RailwayCare supplies any one of them on its own, but a programme needs all four in the right order. This page is an index of what each part decides; the specifications of each part live in their own sections and are deliberately not repeated here.Throughput, reach, consumable and evidence are four different purchases, and a programme needs all of themThe four parts, and the decision each one settlesPart of the programmeWhat it is, in one sentenceThe decision it settles in your planWhere the detail livesGrinding wheelsThe only part of the chain that actually touches the rail; everything else positions it.How much metal one pass can take at your working speed, and therefore how many passes a campaign needs.Wheel range by machine and rail typeGrindersPortable and on-track machines that reach what a long machine cannot: turnouts, crossings, weld and weld-adjacent zones, isolated defects.Who does the last few tens of metres, and inside which minutes of the possession.Grinder rangeGrinding vehicles and trainsThroughput: how much line one possession can take down to a set depth.Whether the interval your tonnage demands is physically deliverable inside your windows.Grinding vehicle rangeInspection and acceptanceTurns a grinding job into a record: geometry, surface, depth, before and after.Whether the campaign can be evidenced, and therefore whether it counts as life bought.Inspection tools and the acceptance criteriaNotice that none of the four rows is a product feature. Each is a decision that someone has to take before a machine is booked, and each one belongs to a different person in a different department in most organisations. The value of buying them from one place is not a discount; it is that the four decisions are written to fit each other.The chain only holds if each part is specified against the same target shape and the same depth budgetSix lines that turn four products into one programmeLine in the specificationWhich part it constrainsWhat happens when it is missingTarget profile drawing, with zone tolerancesAll fourThe machine defines the shape by default, and no part of the chain can be measured against anythingRemoval per cycle, and the envelope balanceWheel and vehicleDepth is set by the defect every single time, and the head's design wear allowance is spent without a planDuty: depth per pass at the planned working speedWheel and machine togetherPass counts rise to compensate, the surface heats instead of being shaped, and blue appears where geometry was wantedAccess and reach limits, stated by location typeGrinders, and the vehicle at the boundaryWork stops at the first restricted zone, or worse, is attempted inside itMeasurement plan: section, chord, transverse positionInspectionTwo people measure two different things, and both of them are rightEvidence pack: measurement files, machine settings, consumable batchInspection and consumableThe result exists on the day it was produced and cannot be defended a year laterEvery row here is a sentence you can put in a tender document or a contract schedule, and every row costs nothing to write. The rows that are usually missing are the third and the sixth, and they are the two that decide whether a future argument about a failed campaign can be settled with paper instead of opinion. The commercial version of the same discipline is set out in our answer on grinding contract KPIs.A programme is a set of written hand-offs, not a list of equipmentHow much of the chain to takeYour starting pointWhat we supply in that caseWhat stays on your sideYou already run machines, and campaigns slip for one reason: the consumable does not deliver the removal the plan assumedA consumable schedule matched to your machines, your rail grade and your working speed, with batch traceability on every setThe plan, the crew and the acceptanceYou have vehicles, but the ends of every job keep going undonePortable machines for turnouts, crossings and weld zones, plus the consumable schedule for themThe possession plan and the accessYou are building capacity for a defined tonnage programmeVehicle or train capability sized to the removal per cycle, with the consumable programme behind itThe operating crew and the network accessYou need the work evidenced, not only deliveredA measurement and acceptance protocol: baseline, thresholds, records packNothing, because the record is the deliverableThe third row is where most enquiries start and where most disappointment comes from: a machine is bought against a tonnage figure, and the depth per cycle that the tonnage actually requires is never converted into a consumable and pass plan. RailwayCare is a manufacturer first: since 2004 we have made the wheels that do the cutting, we drafted the industry standard JB/T 11431 for them, and the machines and vehicles we supply are specified around the same removal-per-pass arithmetic. That is why a Molaton schedule is quoted per working speed and per rail grade rather than per catalogue number.Two boundaries are worth stating plainly, because they keep this page honest. First, wheel specifications, abrasives and bonds are documented in the grinding wheel section, machine specifications and comparison between machine types in the grinding machine section, and instrument accuracy in the inspection section; repeating them here would only produce a third version that disagrees with both. Second, the strategy layer, which is what this group covers, is the part that decides whether any of the four is worth buying: who owns the target shape, which depth a cycle may take, and when the next cycle is due. If that layer is missing, the other four are equipment looking for a purpose. The full solution scope, including training and on-site support, is on the rail grinding solution page, and the build-it-in-house alternative is weighed in buy or contract.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431 for rail grinding wheels, are certified to ISO 9001 / ISO 45001 / CRCC, and every claim below is backed by on-track tests on high-speed, heavy-haul and metro networks.Related questions you may also askWhat is the target rail profile, and who defines it?What grinding depth is safe before you over-grind a rail?What is the difference between large-machine and small-machine grinding?What acceptance criteria decide whether a grinding job passed?What rail grinding is, and how a programme is builtMore answers on rail grinding strategy and equipmentBuilding a programme, or fixing one that keeps slipping?Tell RailwayCare the line, the rail grade, the tonnage you need to serve, your window length and what the last campaign actually removed. We will set out which of the four parts you already have, which one is the constraint, and what a Molaton consumable schedule plus the matching machine scope would look like against that target shape.Scope My Programme

Question 24

A programme buys life through three links: remove the crack source, restore the contact shape that sets stress, and slow how fast the defect returns. Depth and interval decide.

Accepted Answer

How do rail grinding programs extend rail service life?A programme extends life through a chain of three links, and it only counts if all three hold: remove the source the cracks start from, restore the contact shape that sets the stress, and slow the rate the same defect comes back. Depth per cycle governs the first two. Interval governs the third. A campaign that takes metal without moving all three has spent section and bought nothing, which is why the honest measure of a programme is not kilometres ground but the removal per cycle and the interval both staying inside budget.What separates a programme that buys life from one that only spends metal is written in three numbers, not in the tonnage passedThree links between one pass and a later renewalLinkWhat the pass actually changesThe number that proves it happenedWhat the renewal date does if the link is skipped1. Crack sourceRemoves the surface layer in which rolling contact fatigue initiates, including the hard, brittle layer sitting above itCrack depth at the surface after the job, against the 0.2 mm window; the layer itself measures about 870 HV0.3 against about 340 HV0.3 for parent metal on our samplesThe face looks clean while the crack root stays below reach, so the same defect returns inside the next interval and the renewal moves forward rather than back2. Contact shapePuts the head back inside the target profile band, so the wheel rolls where the design intendedCrown and zone deviation against the band, plus running-band width: 25-30 mm on high-speed, 35-40 mm on conventional and heavy haulContact migrates onto a narrow strip, contact stress rises, and side wear and plastic flow accelerate; the rail loses section without being worn evenly3. Rate of returnTakes the peaks and flow lips off before they grow into a shape that keeps regenerating the same defectMetal removed per position, against the section the line still has left to spendThe same defect class returns faster every cycle, so the budget is spent without buying time4. Dynamic loadLowers the impact a wheel delivers where geometry is discontinuous, which is where crack growth is driven hardestLateral acceleration measured at turnouts fell from about 0.20 to 0.04-0.08 m/s² after grinding on our projectsThe rail keeps absorbing loads the shape was supposed to remove, and fatigue accumulates at the rate of the unground surfaceRead the four rows as a sequence, because each one is a precondition for the next. There is no point restoring the shape over a crack root that will surface again, and no point removing the crack if the geometry that created it is left alone. Life is what you get when the sequence is completed in order, cycle after cycle.Removal at the crest, in the right zone, at the depth the crack window allows: the sequence is the strategyWhat our own track records showLine and yearWhat was measuredResultWhat it means for the renewal dateLiuzhou, 2022Output of a production campaign inside a single possession18.72 pass-km in one two-hour possessionInterval is a throughput question before it is a metallurgical one: if one night cannot deliver the removal the plan needs, the interval slips for reasons that have nothing to do with steelHewu high-speed line, 2020Passes delivered per millimetre of consumable consumed, on one machine and one line4.28 pass-km per millimetre of consumable, 214.22 pass-km on a single set, no continuous blueA campaign can only be worked to the planned depth if the consumable lasts the campaign; when it does not, depth is traded for changes and the second cycle is where the defect returnsShuohuang heavy haul, 2024Same measure, much higher tonnage per pass4.98 pass-km per millimetre, 199 pass-km per setOn heavy haul the tonnage pressure lands mostly on the interval, which is why depth discipline matters more there, not lessShuohuang, single-machine trial, 2023Same measure, with a changed machine set-up rather than a changed plan5.96 against 4.99 pass-km per millimetre, best single set 332.3 pass-kmThe plan is not the only lever: how the machine applies the cut changes what share of the plan is deliverable inside the windowYiyang, R=800 m curve, 2024Metal removed per position inside one six-pass plan at 12 km/hUp to 0.3 mm on the high-rail field side, up to 1.8 mm on the low-rail sideOne plan, six times the removal in one place: the deep side is where the remaining section is genuinely being spent, so any life claim must name the position it refers toThese are our own records, and they are deliberately reported as production measures rather than as a service-life figure. They answer the question a programme manager actually has to settle: can the work be delivered, at the depth the plan requires, with a consumable that survives long enough to finish it. Every row above can be re-measured on your line in one possession.A programme is judged on what it delivered inside the possession, not on what it intendedFive ways a programme spends section without buying lifeWhat goes wrongHow it shows in numbersWhat it costsThe correctionDepth set by the defect every cycleCumulative removal runs against the crown band of +0.2 to -0.4 mm and against the head's design wear allowanceSection meant to last years is consumed in a few campaignsBudget removal per cycle against the envelope still available, and keep corrective work as the exception it was meant to bePasses copied from the last jobSame pass count, but removal per position differs across the headOne zone is left above the crack window while another is over-cutSet passes from measured removal per zone on this machine and this lineInterval read off the calendarThe job starts after the 0.3 mm trigger instead of before itThe crack root sits below the depth a normal pass can reachSet the interval as a division and cap the result with the defect clockSurface restored, shape ignoredCrown inside tolerance, running band outside 25-30 or 35-40 mmThe same defect is regenerated by the contact geometry itselfHand back against the profile drawing, not only against roughnessNo baseline, so nothing is provableThe acceptance sheet is signed without a before readingYou cannot tell a campaign that bought life from one that only spent metalMeasure the same section, the same chord and the same transverse position twiceThe fifth row is the one that quietly costs the most, because it removes the ability to learn. Two programmes with identical removal can have opposite effects on life, and only a baseline taken on the same section distinguishes them. That is also why the plan, the depth and the acceptance thresholds belong in the same document: who owns the target shape is answered in our answer on the target rail profile, and the depth limits themselves in our answer on safe grinding depth.We do not publish a life extension in years or in per cent, and a supplier who has not measured your line should not either. What can be stated without invention is the arithmetic: the rate at which the defect returns and the amount of section each cycle costs are both measurable, and the programme that keeps both inside budget is the one that defers renewal. A Molaton consumable is qualified on exactly the two numbers this arithmetic depends on: removal per pass at the planned working speed, and passes delivered per millimetre consumed. Where the defect has already passed the point of grinding, the decision turns into the one covered in grind or replace.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431 for rail grinding wheels, are certified to ISO 9001 / ISO 45001 / CRCC, and every claim below is backed by on-track tests on high-speed, heavy-haul and metro networks.Related questions you may also askHow is a rail grinding interval set by tonnage?What grinding depth is safe before you over-grind a rail?How many grinding passes does a rail need?When should you grind a rail instead of replacing it?What rail grinding is, and how a programme is builtMore answers on rail grinding strategy and equipmentIs your programme actually buying years, or only spending metal?Send RailwayCare the interval you are running, the depth you take per cycle, the rail grade and one before-and-after measurement from the same section. We will show you which of the four links is missing, and what depth and interval your remaining section can still afford — with a Molaton wheel schedule built to deliver that removal at your working speed.Review My Programme

Question 25

Corrugation is a wavelength, not a thickness. A pass that removes no more metal at the crest than in the trough leaves the wave in place, and three other settings reprint it.

Accepted Answer

Why does grinding sometimes make corrugation worse?Because corrugation is a wavelength, and a grinding pass either breaks that wavelength or reprints it. Amplitude is measured from crest to trough, so a cut that removes the same depth everywhere moves the whole surface down and leaves the wave exactly as it was. Three further causes add to it: the wheel rubbing instead of cutting, the machine's own repetition landing near the wavelength, and a hand-back with no before-and-after measurement. All four are visible in numbers long before a machine arrives.Corrugation is removed only where the pass takes more metal from the crest than from the troughFour ways a campaign leaves the wavelength alive or makes it deeperCauseWhat you see on the lineWhat is really happeningWhat to changeThe cut is uniform instead of differentialThe trace looks cleaner but the same period is still readable, and the depth reading has barely movedEvery point of the head moves down by the same amount, so crest-to-trough distance is unchanged. Flattening a wave requires removing more metal at the crest than in the troughCompare the corrective depth at the crest with the measured wave amplitude, not with a nominal depth per passToo few passes for that amplitudePeak amplitude falls just under the trigger line, so the job is signed off, then the wave is back within part of the intervalThe crest was shaved, not brought down to trough level. What remains is a shallower wave of the same wavelength, and the same contact geometry regrows itPasses are an output of removal per pass on this machine and this line, not a number copied from the last jobThe wheel rubs instead of cuttingFace of the wheel looks loaded and shiny, the spark pattern changes, and metal seems to be smeared rather than cutWithout self-sharpening the grain rubs: heat rises, peaks are pushed sideways into the trough, and roughness climbs even though the vehicle kept movingStop the campaign and check the wheel duty. Adding passes to a rubbing wheel only adds heatMachine repetition near the wavelengthThe surface carries a regular pattern with a spacing that matches the head layout rather than the original waveThe heads strike the same phase repeatedly and the machine writes its own periodicity into the surface, which then grows under traffic as if it were real corrugationChange the angle sequence and the working speed so successive heads do not repeat the same phaseThe arithmetic behind the first two rows is worth stating plainly, because it explains most disappointing campaigns. Corrugation depth is a peak-to-trough distance. Removing a uniform layer from a wavy surface does not change that distance at all; it only lowers the average. Only removal that is deeper at the crest reduces the amplitude, and the depth required at the crest is the amplitude itself, not the depth that happens to be comfortable for the machine.A wave is a shape, so the plan has to be written against the shape and verified against itProving the wavelength is gone, not just the peakWhat to measureA pass looks likeA false pass looks likeAmplitude, same section, same chord, same transverse positionRe-measured on the exact section that triggered the job and below the corrective trigger of 0.3 mm depth on a conventional lineA figure quoted from a neighbouring section, a different chord length, or a different position across the headWavelength content of the traceThe dominant period is no longer present, or is reduced across the band rather than at one pointThe peak is lower but the same period is still readable, which means the job traded amplitude for timeWhole contact band, every zoneCrown within +0.2 to -0.4 mm, zones 2 to 4 within plus or minus 0.2 mm, and GQI at 85 or better where that index is usedOnly the crown measured; the gauge corner left as found, so the wave migrates rather than disappearsSurface condition at the same timeRoughness at or under 10 µm, no blue, and a 1 m straight edge showing less than 0.3 mmA cured wave replaced by a burnt, torn or glazed surface, which is a new defect recorded as a repairBehaviour after the first block of tonnageAmplitude stable over the traffic that follows, so the result survives contactMeasured only at hand-back, with regrowth already visible within the intervalThis is where a verification habit beats a specification. The five rows above are the same measurement repeated, not five different instruments: fix the section, fix the chord, fix the position, and take the reading twice. Without a baseline the acceptance sheet says nothing — and a corrugation job with no baseline is indistinguishable from a job that never happened.Verification is a repeat of the same measurement, taken on the same section before and afterThe order of work that avoids reprinting the waveStepDecision to takeInput you must already have1. Characterise the waveAmplitude, wavelength band, transverse position and extent along the lineA recording over the affected section, taken before any machine is booked2. Choose the modePreventive work that keeps the shape, or corrective work where the defect governs the depthDefect depth against the envelope the section still has left3. Set the passes from measured removalNumber of passes at the planned speed, from removal per pass on this machine and this lineOne measured pass on the same rail grade, not a figure from another network4. Set the depth per zoneDeepest where the wave is deepest, not one setting across the headPer-zone deviation. On an R=800 m curve we measured up to 0.3 mm on the high-rail field side and up to 1.8 mm on the low-rail side in the same six-pass plan5. Verify and hand backAmplitude, wavelength, profile and surface condition, all on the same sectionThe baseline from step 1 and the acceptance thresholds agreed before the jobTwo things sit outside this list and still decide whether the wave returns. The first is the target profile: corrugation follows the contact geometry that produces it, so a campaign that restores the surface without restoring the intended shape is grinding a symptom. Who owns that shape, and how it is specified, is answered in our answer on the target rail profile. The second is the wheel: a Molaton wheel is qualified on removal per pass at the planned speed, which is the number the pass arithmetic depends on — if a wheel cuts less than the plan assumed, the wave survives even though the campaign ran exactly as scheduled.The honest summary is that grinding does not make corrugation worse by accident. It makes it worse when the plan is written against tonnage and pass counts instead of against the measured wave, and when nobody measures the same section twice. Our corrugation grinding and verification study follows the same argument from the mechanism side; this page is the work instruction that comes out of it, and Molaton wheels are what the plan asks to deliver that removal, pass after pass, at the depth the section was promised.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431 for rail grinding wheels, are certified to ISO 9001 / ISO 45001 / CRCC, and every claim below is backed by on-track tests on high-speed, heavy-haul and metro networks.Related questions you may also askHow to remove rail corrugation and verify the resultHow many grinding passes does a rail need?What acceptance criteria decide whether a grinding job passed?What grinding depth is safe before you over-grind a rail?More answers on rail grinding strategy and equipmentCorrugation coming back after every campaign?Send RailwayCare the corrugation trace, the section and chord it was measured on, the pass plan that was used and the rail grade. We will work out whether the wave is being reprinted, and what removal per pass at your working speed it actually takes to flatten it — with a Molaton wheel schedule that delivers that depth.Review the Wave Plan

Question 26

WEL is a different material on the rail head, not surface dirt: its hardness and depth decide whether a graded pass removes it or the wheel presses it into the steel.

Accepted Answer

How do you control white etching layer by grinding?You control it by grading the work to the layer, not by taking a bigger cut. A white etching layer is not a stain on the rail head: it is a different material sitting on top of the parent steel, and in our measurements it reaches about 870 HV0.3 against 340 HV0.3 for the metal below. Its thickness is also an output of the grinding parameters rather than a property of the track. So two numbers decide the plan before a machine moves — how deep the layer is, and how deep the cracks are underneath it.Grinding is the routine operation that can reach a metallurgical layer without removing the railThree measured facts, and what each one forces into the planMeasured propertyWhat the number isWhat it forces into the grinding planHardness of the layerAbout 870 HV0.3 at the surface against about 340 HV0.3 in the parent metal, a rise of roughly 155%, with a sharp boundary visible in a sectionRemoval per pass is not constant through the layer. The wheel meets a much harder material at the top than at the bottom, so the first pass and the last pass do different amounts of workThickness of the layerIn controlled tests about 45 µm at 1000 rpm, rising to about 320 µm at 3000 rpm, with roughness moving from 6.6 to 8.4 µm over the same rangeThickness is produced by the grinding setting. A setting chosen to save time can generate the very layer the job was sent to remove, so speed and feed are part of the metallurgy, not just of productivityDepth of what sits under itContact fatigue cracks; grinding still removes them reliably while they are about 0.2 mm deep or lessThe layer is the symptom and the crack depth is the decision variable. Past 0.2 mm the surface can be cleaned while the crack root stays in the railThis is why a white etching layer cannot be handled with one setting for the whole network. The same instruction — take off 0.2 mm — is generous on a thin surface film and insufficient on a 300 µm band with cracks below it. Grade the work from the section, and the instruction becomes defensible.A layer that is harder than the metal beneath it turns every pass into a decision about how the grain meets the surfaceWhy one uniform pass presses the layer in instead of taking it offFailure modeHow it shows upWhat actually happensWhat the plan must changeOne deep pass over a thin filmSpindle load climbs, the surface goes bright and then tints, and roughness after the job is worse than before itA hard, brittle film fractures under the grain; the fragments are not carried away, they are rolled into the softer parent metal by the following wheelsRemove the film from the top down in graded passes, so each pass cuts metal rather than fragmentsSpeed raised to shorten the jobThe surface looks patchy and spalled rather than cut, and the measured layer is thicker after the campaign than before itHeat input and grain load rise together; the pass leaves about 320 µm of white layer instead of about 45 µmHold the spindle and feed setting, then buy time with passes rather than with speedRepeating a shallow cutThe band looks clean by eye but the measured profile still shows a hollow or a lipThe wheel follows the existing shape, so a shallow pass polishes the layer it failed to removeA depth decision is also a profile decision: verify per zone, not at one pointWorking above the temperature limitDiscolouration beyond light straw, and a surface that is soft where it should be hardThe rail head is being tempered or re-hardened by grinding heat rather than cut; the standard working limit for the workpiece is 150 °CTreat temperature as an acceptance item, reduce depth per pass and let the surface cool between passesThe reason this matters beyond the day of work is the service behaviour of the layer itself. A hard, brittle white layer breaks up early in traffic, and the fragments are pressed into the matrix beneath; from there they act as crack initiation sites. A campaign that leaves the layer in place does not postpone the problem, it puts a supply of hard fragments directly into the crack path.One boundary worth stating plainly: which wheel to buy for this duty — abrasive, bond, grade — is a purchasing question answered in our rail grinding wheel answers. What belongs here is the strategy: how deep, in how many passes, and what has to be true before the job is called finished.The depth budget is finite, so a graded plan spends it in slices rather than in one biteA graded plan you can write into a specificationGrade from the surveyEntry testGrinding settingWhat acceptance must showSurface film, no crackingLayer visible on the running band, no crack found under it by the non-destructive check in usePreventive pass at 0.1 mm or more, inside the profile envelopeProfile inside the tolerance band for every zone, roughness at or under 10 µm, no blue, no abnormal vibrationBanded layer with cracks up to 0.2 mmCrack depth read from a section sample at the affected location, at or under 0.2 mmGraded: two to three passes, total depth inside the 0.5 mm acceptance ceiling, cooling between passesThe last pass no longer opens cracks, and the contact band measures 25-30 mm on high-speed or 35-40 mm on conventional and heavy-haulLayer with cracks beyond 0.2 mm, or a squat already spalledCrack depth past 0.2 mm, or metal already missing from the running surfaceGrinding becomes a holding action, not a repair; the section is referred for replacement or milling assessmentDo not write a grinding depth that promises removal. Record the referral and the interval to the next inspectionTwo disciplines hold the plan together. The first is the envelope: 0.1 mm preventive, 0.2 mm or more corrective, 0.5 mm as the acceptance ceiling — three numbers describing one finite budget of section, which a graded plan spends slowly and a single deep pass spends at once. The second is position. On a 68 km heavy-haul line we ground an R=800 m curve with six passes at 12 km/h and measured up to 0.3 mm on the high-rail field side against up to 1.8 mm on the low-rail side. The layer sits where the contact sits, so any depth claim that is not positional is a claim about an average nobody can inspect.That positional detail is also what the wheel has to survive: a Molaton wheel asked to hold a graded cut through a 300 µm hard band on one side of the head and a softer surface on the other is being asked for consistent removal per pass, which is the only way the pass arithmetic holds. Our squat and white layer study sets out the formation mechanism and the graded approach in detail; the plan above is how it turns into a work instruction, and the Molaton range is what has to deliver the depth the plan assumes.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431 for rail grinding wheels, are certified to ISO 9001 / ISO 45001 / CRCC, and every claim below is backed by on-track tests on high-speed, heavy-haul and metro networks.Related questions you may also askRail squat formation, white etching layer and graded grindingWhat grinding depth is safe before you over-grind a rail?How does rail grinding prevent rolling contact fatigue?How many grinding passes does a rail need?More answers on rail grinding strategy and equipmentFound white etching layer or squat on your line?Send RailwayCare the rail grade, annual tonnage, the affected locations and any section or crack-depth readings you already have. We will turn them into a graded pass plan and a Molaton wheel schedule that matches the layer you actually have.Plan the Graded Work

Question 27

Tonnage does not change the rules of grinding, it changes the clock: wear, deformation and contact fatigue speed up at once, so interval, depth and profile move together.

Accepted Answer

How does heavy-haul tonnage change grinding strategy?Tonnage does not change the rules of grinding; it changes the clock. Every tonne puts the same wheel over the same steel, so wear, plastic deformation and rolling contact fatigue all accumulate at once, and the plan that was maintenance at 20 MGT a year is a debt at 100 MGT. What moves is not the method but four settings: how often, how deep, which profile, and how the track is divided.Tonnage behaves like a clock, and every mechanism on the rail head reads the same clockThree mechanisms running on one clockMechanismWhat it does to a heavy-haul rail headWhat it forces into the planAbrasive wearMetal leaves the gauge face and the crown steadily; the head narrows and flattensThe interval is written in tonnes, not in monthsPlastic deformationHigh contact stress from heavy axle loads pushes metal sideways, so the crown cups and a lip builds at the gauge corner or field sidePreventive depth has to clear the deformed layer, not polish its surfaceRolling contact fatigueCrack layers form where the wheel keeps loading the same bandThe interval also has a ceiling: grinding must arrive before cracks pass the depth it can still reachThe defect mix says the same thing more bluntly. On Chinese heavy-haul track the dominant family is deformation — shelling and flange build-up — whereas high-speed lines are dominated by contact fatigue and metro curves by side wear and corrugation. A heavy-haul grinding plan is therefore mostly a deformation control plan, and its success is measured by how little deformation reaches the next campaign.The steel is built for that duty. 75N-U78CrV raises carbon to 0.75-0.82%, silicon to 0.70-0.80%, manganese to 0.80-1.10% and adds 0.30-0.50% chromium, while holding sulphur and phosphorus at or below 0.02%: a metallurgy chosen for wear resistance and load capacity. It is also why heavy-haul rail is the hardest grinding duty there is — the same pass removes less metal from this head than from a softer one.When deformation leads the defect mix, the plan is written around removing it before it becomes a lipFour settings that move with the tonnageSettingLight-traffic defaultHeavy-haul settingWhere the figure comes fromIntervalAbout 100 Mt per pass on straight track and large-radius curves30-50 Mt, and shorter again where curves and gradients concentrate the damageChinese permanent way requirements; what transfers between networks is the curve-to-tangent ratio, not the absolute numberDepth per passPreventive work at 0.1 mm or morePreventive depth set to remove the deformed layer, corrective work at 0.2 mm or more, always inside the 0.5 mm acceptance ceilingMode of work plus a finite profile envelope; the ceiling is a ceiling, never a targetProfile target60N on 60 kg/m main line75N where 75 kg/m rail is laid; 60N stays on the 60 kg/m sectionsQ/CR 583-2017 profile guidance; the Shuohuang corridor carries both, 75N uphill and 60 downhillAcceptance bandContact band 25-30 mm at high-speed turnouts35-40 mm on conventional and heavy-haulTurnout grinding practice, split by traffic rather than by preferenceConverting millimetres into tonnes is where plans quietly break. Our own grinding project records include a heavy-haul line where the rail head was consuming roughly 0.2 mm of section per 25 MGT. That single in-service figure does most of the work, because it answers two questions at once: how many campaigns the rail has left inside its envelope, and how much of that envelope this year's plan is about to spend. It is a measurement, not a standard — a line with different grades, gradients and curve content will produce a different number, and adopting ours without measuring yours is how intervals end up set by optimism.The same discipline applies to position rather than to the average. On a 68 km heavy-haul line in Shandong carrying about 40 million tonnes a year, six passes at 12 km/h through an R=800 m curve removed up to 0.3 mm on the high-rail field side and up to 1.8 mm on the low-rail side. One pass plan, six times the metal in one place. Any heavy-haul depth figure quoted as a single number is therefore a planning average, and the acceptance measurement has to be positional.Depth is a positional claim on a finite envelope — which is why an average depth flatters a campaign that under-ground the curveWhen the tonnage outruns the planConstraint you hitHow the numbers expose itWhat you must not do about itWindows cannot carry the workload the tonnage createsRequired pass-kilometres divided by output per window exceeds the nights available; one machine produced 18.72 pass-kilometres inside a single 2-hour night window on the Hengliu lineDo not cut depth to fit the window — a shallow campaign spends envelope without removing the deformationCorrective work becomes the majority of campaignsEnvelope consumed per year climbs while the profile keeps degradingDo not answer with deeper corrective passes; the answer is shorter intervals or a different split of the trackThe same defects return inside one intervalThe deformed layer was thicker than the depth actually appliedDo not add passes without adding depth — repeating a shallow cut polishes the layer it failed to removeCurve and tangent are planned as one blockGauge-face wear stops being the governing defect, so the tangent interval no longer fitsDo not apply a tangent interval to a curve; separate the two plans and separate their acceptance pointsThe economics follow the same arithmetic, and it is worth stating as structure rather than as price. Wheel consumption scales with pass-kilometres, occupation time is a fixed cost per night whatever that night produces, and rail replacement is the terminal cost both of them exist to postpone. A heavy-haul strategy that shortens intervals deliberately spends more of the first two in order to protect the third — that trade is the whole argument, and it is worth making in numbers rather than in adjectives.There is a product consequence too, and it is the reason RailwayCare publishes heavy-haul consumption separately from high-speed results: a wheel that holds a 35-40 mm contact band on 75N rail while removing a deformation layer pass after pass is being asked for something different from a wheel polishing a high-speed head. Molaton heavy-haul wheels are qualified on exactly that duty, and our heavy-haul line case record reports the consumption and profile results from it. The strategy sets the interval; the Molaton wheel is what has to survive inside it.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431 for rail grinding wheels, are certified to ISO 9001 / ISO 45001 / CRCC, and every claim below is backed by on-track tests on high-speed, heavy-haul and metro networks.Related questions you may also askRail grinding for heavy-haul and mining railwaysHow is a rail grinding interval set by tonnage?How many grinding passes does a rail need?How does rail steel hardness affect the wheel choice?More answers on rail grinding strategy and equipmentRunning tonnage your plan cannot keep up with?Send RailwayCare your annual gross tonnage, rail grades, curve list and window length. We will turn them into campaign intervals, positional depths and a Molaton wheel schedule that fits the nights you actually get.Plan the Campaign

Question 28

The target profile is the track owner's specification, not a machine setting: a named rail-head template with tolerance zones, chosen from rail section, traffic and wheel contact.

Accepted Answer

What is the target rail profile and who defines it?The target profile is not a machine setting. It is a maintenance specification owned by the track owner: a named template of rail-head coordinates, with tolerance zones drawn around it, chosen from the rail section, the traffic that runs on it and where the wheel actually contacts the steel. A grinding machine can follow a template perfectly and still leave the wrong shape, because the choice of template is made before the machine is switched on.A grinding campaign starts with a document, not with a machine: the template decides what "correct" will mean at acceptanceThree layers: who decides, what is written down, what gets checkedLayerWho owns itWhat happens if the grinder decides insteadTemplate, the target head shapeTrack owner, working to its national or company profile standardEach pass copies whatever shape is already on the rail, so the error is inherited rather than correctedTolerance zones around that templateTrack owner, or the acceptance standard it works to; the inspector applies themMetal is spent polishing zones nobody asked about while the zone that governs ride quality is missedContact band position and widthTrack owner, from wheel-rail contact studiesThe band drifts to the field side or onto the non-working edge, the ride deteriorates, and the profile still "passes"Depth and life budgetPermanent way engineers who hold the design wear allowanceSection is spent in the wrong place, and the rail is scrapped years before its planned lifeMeasurement referenceAgreed before work starts: same instrument class, same named positionsWith no shared baseline, no result can be either proven or rejectedThe machine's part is execution. Profile control systems follow a stored template by copying the head closed-loop, and they follow it well — the accuracy of that following is a separate question, covered in how intelligent copying improves profile accuracy. But following is the whole of what it does. Nothing inside a grinding control system knows whether the loaded template suits a 60 kg/m main line carrying heavy axles rather than a lightly used siding; that judgement lives in the specification and in the engineer who wrote it.A stored template turns the machine into a precise follower — precise about a shape somebody else choseUIC60, 60 kg/m, 60N: the rail name is not the target shapeNameWhat it actually designatesWhere you meet itWhat the grinder has to be toldUIC60A European rail section designation in the 60 kg/m classEuropean infrastructure managers and UIC-based track standardsThe profile drawing and revision the owner works to — the name alone is not a shape60 kg/mThe Chinese standard rail section (TB/T 2344.1; switch and crossing rails under TB/T 2344.2)Chinese main line, almost universallyWhether the target is the as-rolled head or the adapted profile below60NAn adapted, post-grinding rail-head profile rather than a rail section; Q/CR 583-2017 directs main lines to 60N or 75NChinese main line profile grinding specificationsThis is the grinding target, and this is what gets loaded as the template75NThe adapted profile of the 75 kg/m familyHeavy-haul main line; the Shuohuang corridor runs 75N-U78CrV uphill and 60-U75V downhillWhich of the two templates applies to which track, direction by directionThis is where enquiries go wrong. "Grind to UIC60" or "grind to 60 kg/m" states which rail is in the track; it does not state which shape is to be left behind. On the Chinese network the shape has an official designation and a clause behind it (60N, 75N); on UIC60 track the shape is issued by the infrastructure manager and has to arrive with the enquiry. If a package contains only the rail name, the correct response is a request for the profile drawing and its zone table — not an assumed template.When the shapes do not match, the consequence shows up as a defect rather than as a paperwork error. Chinese permanent way guidance records the mechanism plainly: because wheel treads have adapted to the 60 kg/m and 60N profile, running on 50 kg/m rail readily produces abnormal rail-head wear and flange build-up. Nothing was ground badly in that example. The target shape and the wheel simply disagreed, and the rail paid for it. It is the same mechanism behind most complaints that "grinding does not last" on mixed-traffic networks.Two networks can grind to the same rail name and still hold two different shapes — the difference only appears where wheel and steel meetHow to write the target into a grinding specificationItem in the specificationWhy it must be explicitHow it is verified afterwardsTemplate identity: designation, drawing number, revisionTwo drawings carrying the same name can differ between revisionsLoaded template record compared against the contract drawingZone definition and per-zone tolerancesThe crown and the side zones do not share limitsZone-by-zone comparison against the pre-work baselineContact band expectationBand width and position are the outcome the wheel actually feelsBand checked at the same named positions, not "somewhere on the head"Depth budget per campaignGrinding spends a finite design allowanceMetal removed recorded and subtracted from the remaining allowanceReference methodA result is comparable only if it was measured the same way twiceSame instrument class, same positions, pre and post, signed by both partiesChinese acceptance practice shows how tight the box is. At the crown the tolerance band runs +0.2 / −0.4 mm, and ±0.2 mm across zones 2, 3 and 4, with a profile quality index of 85 or better rated excellent and 70 or better acceptable. Contact band targets split by traffic: 25-30 mm on high-speed turnouts, 35-40 mm on conventional and heavy-haul. Depth is banded too — at least 0.2 mm per pass for pre-grinding and corrective work, at least 0.1 mm in preventive mode, and 0.5 mm on the parent metal as an acceptance ceiling that is never a target.A wheel cannot choose a profile, but it can lose one. Shape-keeping depends on grade, balance and the way the wheel wears across its face, which is the manufacturing half of this same problem — the reason Molaton wheels are checked dimensionally and for balance batch by batch, and the reason RailwayCare treats the template and the consumable as two parts of one specification rather than two separate purchases. The Molaton range is listed by machine family in which wheel suits UIC60 and 60 kg/m profile work, which is the question that follows this one.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431 for rail grinding wheels, are certified to ISO 9001 / ISO 45001 / CRCC, and every claim below is backed by on-track tests on high-speed, heavy-haul and metro networks.Related questions you may also askRail profile grinding: techniques, standards and best practiceWhat acceptance criteria decide whether a grinding job passed?What grinding depth is safe before you over-grind a rail?Target profiles and restricted zones in turnout grindingMore answers on rail grinding strategy and equipmentNot sure which shape you are buying?Send RailwayCare the rail section, the traffic mix and any profile drawing you hold. We will tell you which template applies, where the tolerance zones fall on your head, and which Molaton wheel family holds that shape on the machines you run.Check the Template

Question 29

Take one road or one end at a time, not the whole layout. Short occupations, the point of switch last, and a handback the operator authorises - the grinding itself is unchanged.

Accepted Answer

How do you grind a turnout without closing the line?You do not need the whole turnout. You need one road, or one end, at a time. The job is planned as a series of short occupations — each one ground and given back before the next is asked for — with the point of switch taken last and shortest, because it is the only place where the machine itself, not just the rail, has to be checked before handback. The grinding is identical to open track. What changes is the size of the unit you ask the operator for, and the fact that the operator, not the grinding crew, puts the road back into service.An occupation is a purchased block of minutes: the plan is built to fit inside it, not the other way roundWhat actually sets the size of the occupationConstraintWhat it fixesHow it changes the planHow many roads the operator will withdrawWhether a one-road-at-a-time sequence exists at allIf only one road can go, the layout is worked in two fronts with a re-index between themDistance from point of switch to crossingHow far one machine setting reachesA long layout buys an extra stop, not a faster passPasses the defect needsThe only quantity that scales with the damagePasses come from profile deviation divided by per-pass removal at the working speed, rounded upRe-index and re-clamp timeThe number of stops, not the metres groundFewer stops beat faster grinding, every timeHandback check at the point of switchThe last block before the road returnsCannot be compressed by working harder, so it must be planned, not squeezedThe start signal is worth stating, because it decides how much of that list you need. Corrective turnout grinding is triggered on conventional lines when corrugation depth or flange build-up width passes 0.3 mm; on high-speed lines the triggers are ride quality, lateral acceleration alarms, a poor contact band, or fish-scale depth of 0.5 mm at turnout rails. Below those thresholds you are maintaining, not repairing — and a preventive occupation is a smaller unit to ask for than a corrective one.Fewer, better-planned occupations beat more frequent short ones: every occupancy carries the same fixed setup costWhere the minutes in a turnout occupation goBlockWhat it consumesWhat genuinely shortens itProtection, access, isolation of the roadA fixed cost charged to every occupation, whatever its sizeFewer and larger occupations — not working fasterBaseline measurement and target settingMeasurement time, no metal removedRecord named positions once, then reuse exactly the same ones at acceptanceGrinding inside the roadThe only block that changes the railDepth per pass and fewer re-index stopsMachine handling between zonesHandling inside a restricted spaceSequence the zones so the machine travels one way through the layoutPoint-of-switch check and handbackThe operator's own verificationHave the measurement record ready before the machine leaves the siteTo see the scale: one machine on the Hengliu line produced 18.72 pass-kilometres inside a 2-hour night window. That is what a window is worth when the machine never stops moving. A turnout is worked in metres with the machine repeatedly stopping, so its governing arithmetic is setup and handback time — which is exactly why the job is split across several small occupations instead of one long one that the operator will not grant.What must not happen is trading depth for time. If the passes the deviation requires will not fit the occupation you have been given, the answer is another occupation, not a shallower job. Per-pass depth is fixed by the mode of work and the acceptance ceiling is a ceiling, not a target — the depth rules do not bend because the window is short.The point of switch is the shortest block and the last one: nothing is handed back until the machine, not only the rail, has been checkedThe last block: point-of-switch check and handbackItemWhat is checkedWho authorises itSwitch rail against stock railSeating and throw exactly as found before the workPermanent way and signalling, per their rulebookPoint machine and lockingUntouched, still detecting and locking as designedSignalling disciplineInsulated joints and track circuitsNo bridging, no disturbance, no debris across the gapSignalling disciplineFastenings, baseplates, check railsSound, complete, correctly seated after grindingPermanent wayGrinding debris and foreign objectsCleared from switch, crossing and check-rail gapWork party, then reported to the operatorProfile and contact bandContact band 25-30 mm on high-speed turnouts, 35-40 mm on conventional and heavy-haul; zone tolerances held within ±0.2 mm on zones 2, 3 and 4 and +0.2 / −0.4 mm at the crownTrack owner or acceptance inspectorTwo limits here are absolute and are worth writing into the method statement. Grinding never adjusts switch throw, locking or track-circuit settings — those belong to the signalling discipline, not to a grinding plan. And no member of the grinding party declares a turnout fit for traffic: the crew presents its record, the operator authorises handback under its own rules, and if any item on the list cannot be confirmed the occupation ends with the road still withdrawn. A grinding crew that stays inside those limits stays invited back; one that does not creates a safety argument no quality result can win.Get the organisation right and the result shows up where the passenger feels it: lateral acceleration through ground turnouts drops from about 0.20 m/s² to 0.04-0.08 m/s². That improvement is the reason to spend several short occupations on one layout instead of waiting for a blockade the timetable will not give you. The Molaton turnout wheel range exists for the same reason — a restricted zone needs a wheel that reaches into it, not a bigger machine. RailwayCare has run this work on switch and crossing maintenance in a working freight yard, where the roads could not be closed either.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431 for rail grinding wheels, are certified to ISO 9001 / ISO 45001 / CRCC, and every claim below is backed by on-track tests on high-speed, heavy-haul and metro networks.Related questions you may also askPrecision turnout grinding: safety, smoothness and longevityWhy do rail grinder trains work at night?Removing flange build-up on switch railsHow is a rail grinding interval set by tonnage?More answers on rail grinding strategy and equipmentTurnouts you cannot close?Send us the layout, the roads you can withdraw and the length of window you can grant. RailwayCare will break the job into occupations that fit, name the zones each front covers, and match Molaton wheels to the machines that can actually reach them.Plan the Possession

Question 30

Not two sizes of the same job: the train grinds kilometres, the small machine grinds metres of restricted trackwork. Reach decides the split; handover is on profile, not passes.

Accepted Answer

What is the difference between large-machine and small-machine grinding?They are not two sizes of the same job. A mainline grinding train removes metal along a length of track: it works in kilometres, presents one angle set across a long chassis, and its output is measured in pass-kilometres per possession. A small machine removes metal at a place: it works in metres, at a turnout, a crossing or a weld, where the target geometry can change within a few hundred millimetres. Reach decides the split — and whatever the train cannot physically enter must be handed to a small machine inside the same possession, or it simply does not get ground.Length and place are the two units of account: pass-kilometres on the running line, millimetres of profile tolerance at the special trackworkWhere the boundary actually falls inside one turnoutZoneWho works itWhy the other cannotHandover pointRunning line between turnoutsGrinding trainA small machine produces no kilometres; the window would be spent moving, not grindingLast profile record before the turnoutTurnout approach on open trackTrain, if it is scheduled thereNothing yet restricts the head array; the profile is still repetitiveWhere the head array starts to foul the switchStock rail beside the switch railSmall machineThe running surface is interrupted and lies at a diverging angle a running-line head set cannot followFree edge of the stock railSwitch rail, point of switch to heelSmall machineTarget geometry changes over a short distance, while a long chassis presents one angle to every headHeel blockCrossing nose and wing railsSmall machineDeep, narrow, discontinuous geometry at the noseWing rail endCheck railsSmall machineThe check-rail gap is a restricted zone no train-mounted head can enterEnd postInsulated jointsSmall machineMetal removal has to stop at the joint without loading the insulated assemblyJoint barThat list is the practical content of large-machine / small-machine synergy: the small machine is not a backup for a train that did not turn up, and the train is not a bigger version of the small machine. One buys passes over distance; the other buys millimetres at a place. Confirm the real reach of your own machine against the manufacturer's restriction drawing before you write the sequence — the zone list above is generic, the foul points are machine-specific. The restricted-zone geometry behind those foul points is what makes the train stop where it stops.Two working fronts in one possession: kilometres on the running line, metres inside the special trackworkThe handover: five decisions, in this orderOrderDecisionWho owns itWhat breaks if it is skipped1Baseline: both machines record the same measurement positions before any metal is removedSurvey / acceptance partyNo baseline, no acceptance — the job cannot be judged afterwards2Target profile: the operator's template, not whatever the machine last storedTrack ownerThe two machines aim at different shapes and leave a visible step at the boundary3Sequence: train first along the running line, small machine second inside the restricted zonesWork plannerTurnout ground first is re-cut by the train running through it afterwards4Depth budget: one shared envelope, spent by bothBoth parties, in writingEach machine spends the allowance and the total runs past the acceptance cap5Acceptance: same positions, before against afterTrack owner / inspectorOne job gets two verdicts that cannot be comparedThe depth side is where handovers quietly fail. Per-pass depth is set by the mode: pre-grinding and corrective work take ≥ 0.2 mm per pass, preventive work ≥ 0.1 mm, and the acceptance ceiling on the rail head is 0.5 mm. Those are three different numbers, and they belong to one envelope. If the train spends 0.3 mm on the approach and the small machine then works the switch to its own 0.3 mm target, the two of them together have taken 0.6 mm of a section that only had 0.5 mm to give.The handover is settled by residual profile deviation at named positions, never by the number of passes each machine has madeWhere the division goes wrong in practiceFailureWhat it looks like on the railWhat to do insteadHanding over by pass countOne side of a curve ground far deeper than the other, boundary unblendedHand over by residual profile deviation at named positionsLeaving the restricted zone for next timeA transit trough through the switch that the next train pass cannot reach eitherSame possession, second front, small machinePushing the train into the restricted zoneSwitch rail and crossing untouched, angle sequence dragged off the running lineStop the train at the foul point; accept that these metres need another machineSending a small machine onto the open linePlenty of wheel wear, very few kilometresAccept that output per possession is the train's jobThe pass-count failure is easy to demonstrate. On the Yiyang R=800 m curve, six passes at 12 km/h removed up to 0.3 mm on the high side and up to 1.8 mm on the low side of the same alignment — a six-fold spread inside one pass plan. Two machines can be given the same pass count and still remove very different amounts, which is why the handover has to be a number on a profile record, not a number in a work order.Why bother sequencing at all? Because possession time is the scarce resource and its unit price is measurable. One machine on the Hengliu line produced 18.72 pass-kilometres inside a 2-hour night window. Every minute the train spends crawling through a switch it cannot grind is a minute bought at that rate and thrown away. On the other side of the same ledger, turnout grinding shows up in ride quality: lateral acceleration through turnouts falls from about 0.20 m/s² to 0.04-0.08 m/s² once the profile and the contact band are right.The wheels are where the two fronts physically differ, because a train head and a portable head do not share a wheel. Molaton builds both families from one plant founded in 2004 and drafted the industry standard JB/T 11431, so a single possession is not split across two supply chains — the wheel families and the machine range are on the Molaton range page and the rail grinding machines page.Why trust this answer — RailwayCare (product brand Molaton) has manufactured rail grinding wheels since 2004 — the first dedicated producer in China, born from the friction-materials laboratory of Wuhan University of Technology. We drafted the industry standard JB/T 11431 for rail grinding wheels, are certified to ISO 9001 / ISO 45001 / CRCC, and every claim below is backed by on-track tests on high-speed, heavy-haul and metro networks.Related questions you may also askHigh-speed turnout grinding: large-machine and small-machine synergyWhat acceptance criteria decide whether a grinding job passed?Should you buy a grinder or contract the grinding out?Turnout grinding solutionsMore answers on rail grinding strategy and equipmentSplitting one possession between two machines?Send us the turnout layout, your possession limit and the target profile you are working to. RailwayCare will set out which metres the train can take, where it must stop, what the small machine has to cover, and the Molaton wheels for both fronts.Plan the Grinding Split

Professional Rail Track Maintenance Solutions Provider

Safe Tracks Proven by China

FAQs

Are you looking for a reliable manufacturer of railway grinding and track maintenance equipment?

Safe Tracks Proven by China

Solution

Products

Support

Company

Professional Rail Track Maintenance Solutions Provider

Safe Tracks Proven by China

SOLUTION

PRODUCTS

MARKET

COOPERATION

SUPPORT

ABOUT

FAQs

Are you looking for a reliable manufacturer of railway grinding and track maintenance equipment?

Safe Tracks Proven by China

Solution

Products

Support

Company