रोटरी स्लीटर अक्षीय रनआउट: स्टैक-अप गणित, TIR लक्ष्य और चौड़ाई नियंत्रण

त्वरित उत्तर: रोटरी स्लीटर अक्षीय रनआउट (Quick Answer: Rotary Slitter Axial Runout)

यह गाइड क्या कवर करती है: ब्लेड की मोटाई की सहिष्णुता, अक्षीय रनआउट और स्पेसर सीटिंग कैसे पट्टी की चौड़ाई में त्रुटि पैदा करते हैं — और पहले किसे ठीक करना चाहिए।

Who it’s for: स्लिटिंग लाइन इंजीनियर और टूलिंग इंजीनियर जो कार्बन स्टील, AHSS, स्टेनलेस या एल्यूमीनियम लाइनों पर नाइफ पैक निर्दिष्ट करते हैं या चौड़ाई के विचलन (width drift) की समस्याओं को ठीक करते हैं।

Key metrics: नाइफ-फेस TIR 0.005 मिमी (सटीकता) / 0.010 मिमी (मानक); असेंबल किया गया अक्षीय TIR 0.01–0.03 मिमी; ब्लेड की मोटाई की सहिष्णुता ±0.002 मिमी (सटीक-ग्राउंड) बनाम ±0.01–0.05 मिमी (सामान्य)।

लाइव लाइन पर इसका उपयोग कैसे करें: Measure arbor first, assembled stack second. Rank contributors by sensitivity — blade thickness × knife count first, then spacer flatness, then face runout. Tighten the largest term first. Maxtor Metal supplies precision-ground knife and spacer matched sets with documented tolerances to close the gap between your current stack-up and the required width window.

Strip width is built from the ground thickness of every component between the two cutting edges, and the errors do not cancel, they accumulate. That single fact decides which contributor you tighten first. A ±0.01 mm blade thickness tolerance on a 14-knife arbor stacks to roughly 0.14 mm by the last strip, enough on its own to miss a ±0.05 mm width spec. Rotary slitter axial runout adds to that total rather than replacing it.

Precision-ground knife and spacer sets from manufacturers such as Maxtor Metal close this gap by holding ±0.002 mm thickness tolerance and supplying matched sets verified against the width budget before shipment.

The standard tolerances themselves explain why. Under EN 10140, the European standard for cold rolled narrow steel strip, the width tolerance for slit edges at 0.60 mm thickness sits around ±0.15 mm and widens to ±0.50 mm or more on heavy strip (DIN EN 10140:2006 cold rolled narrow steel strip tolerances). A mill working to commercial class is therefore allowed a width band far wider than a precision slitting line can hold — which is why the tolerance class you buy into matters before the stack-up arithmetic even starts.

चाबी छीनना

• Contributors under 0.01 mm govern whether you hold a 0.1 to 0.2 mm lane tolerance.
• Worst-case and RSS stack-up bracket the answer; use both before choosing a limit.
• Measure the arbor first, then the assembled knife, so you know which term is real.
• Tighten the largest sensitivity term first, not the easiest one to buy.

चौड़ाई भिन्नता के तंत्र (Mechanisms of width variation)

Width drift on a correctly set slitter comes from three stacked contributors: blade thickness tolerance, axial runout, and how the spacers and blades actually seat. They behave differently from what most setup sheets imply. The finished width is the sum of the ground thickness of every component stacked between the two cutting edges, and no mechanism inside the stack averages those errors out. That explains why a machine can be dialled in perfectly and still miss a ±0.05 mm width spec.

Blade thickness tolerance

Each blade’s thickness tolerance adds directly to the stack. A blade that is 0.01 mm thin does not average out against a neighbour that is 0.01 mm thick; it shifts every downstream knife position by that amount. Run the arithmetic on a wide arbor: ±0.01 mm per blade across 14 knives accumulates to roughly 0.14 mm at the last strip, which alone destroys a ±0.05 mm width spec. Tightening to ±0.003 mm brings the same stack back to about 0.04 mm.

This is why procurement decisions matter more than they look. Precision-ground knives hold ±0.002 mm on thickness and diameter; generic import tooling commonly lands an order of magnitude looser, in the ±0.01 to ±0.05 mm band. That difference is not cosmetic. It lands directly in the width budget, and it compounds with knife count.

Rotary slitter axial runout

Rotary slitter axial runout is the second contributor, and the one most often measured in isolation. Widely used shop-floor targets put the maximum acceptable knife-face runout at roughly 0.005 mm for precision work and 0.010 mm for standard work. On assembled stacks, general industrial practice accepts about 0.01–0.03 mm TIR, while high-precision lines work to 0.01 mm or tighter. These are working targets drawn from slitting practice, not clauses in a published standard — which is exactly why they need checking against your own width specification.

Those two bands answer the practical question of whether a number is good or bad. A knife measuring 0.008 mm TIR is acceptable for standard work and marginal for precision work. The same 0.008 mm on an assembled stack is comfortably inside general industrial tolerance.

Spacer and assembly effects

Spacers and assembly practice decide how much of the nominal stack you actually get. Shim-free lines bin blades, stripper rings and spacers into one-micron classes spanning roughly −1 to +1 µm, hold blade bores within about 0.01 mm to an H6/H7 fit, and offset the signs deliberately when assembling — a slightly thick blade next to a slightly thin spacer. The arithmetic behind this is unforgiving: a 5 µm flatness error against a 7 µm cutting gap consumes roughly half the clearance before the knives even start cutting.

For spacer tolerance classes, material selection and rubber ring applications that work alongside precision knife sets, see Precision Slitting Spacers & Rubber Rings: Extend Blade Life.

Seating is the failure mode that hides here. Galled or worn spacers change how the stack seats rather than its nominal dimensions, so the arithmetic still looks right while the cut does not. The corrective sequence is consistent: clean, burr-free, flat mating faces so the stack seats repeatably, then manufacturer-specified torque applied progressively and evenly in a cross or diagonal pattern. Seat first, torque second. Never torque to force runout away, and re-check and re-seat after clamping.

स्टैक-अप मॉडल और गणित (Stack-up models and math)

Width variation is a stack, not a single fault. Every blade, spacer and arbor in the line contributes its own deviation, and the question that decides your specification is whether you add those contributions at full value or combine them statistically. The two methods answer different questions, and confusing them is how teams end up either over-tightening every component or under-specifying the one that matters.

Worst‑case arithmetic stack‑up

Worst-case simply adds every tolerance at full value, so the total is the sum of the individual limits regardless of how unlikely their simultaneous occurrence is. In a 14-knife slitting stack, fourteen blades at ±0.01 mm each accumulate to a theoretical ±0.14 mm before spacers or arbor runout enter the picture. That number is a guarantee, not a forecast: it describes the widest deviation the assembly can physically produce, which is what a customer contract or a safety-critical width claim has to be written against.

Statistical RSS estimate

Root-sum-square assumes independent, normally distributed deviations and bounds the stack at 99.73% conformance, which is the ±3σ band. Instead of adding limits, RSS adds their squares and takes the square root, so three components at ±0.025 mm give ±0.075 mm worst case but about ±0.043 mm by RSS. The assumption is the catch: shared arbor error, a common grinding batch or a systematic setup bias breaks independence, and the RSS figure stops being defensible.

Sensitivity: what moves the needle

Not every contributor deserves the same spend. In multi-knife assemblies the thickness term dominates once the knife count rises, because it repeats with every blade while arbor runout enters once. Flatness and parallelism can consume half the clearance budget, and assembled TIR of 0.01–0.03 mm sets the floor for general work. Rank contributors by their effect on width before buying tighter knives.

मुख्य निष्कर्ष: Worst-case bounds the guarantee; RSS describes the population. Specify to worst-case, monitor to RSS.

Contributor	Illustrative magnitude	Effect on width
Blade thickness tolerance	±0.01 mm per knife	Multiplies with knife count
Spacer thickness tolerance	±0.01 mm per spacer	Multiplies with spacer count
Assembled axial TIR	0.01–0.03 मिमी	Sets the practical floor
Flatness and parallelism	Up to 50% of clearance budget	Consumes setup margin

Figures are illustrative, with assumptions stated, and are not a specification for any specific line.

सटीकता लक्ष्य और सामग्री (Precision targets and materials)

The number you put on the drawing is an OEM or internal target, not a standard value. ISO 1101:2017, Geometrical product specifications — Geometrical tolerancing tells you how to state runout on the drawing, not what value a slitter knife should hold: it defines the notation, datum references and tolerance-zone geometry, while the acceptance figure itself comes from the machine builder or your own process capability. In practice, accepted slitter targets cluster in the 5 to 50 µm range, so the useful question is which end of that band your strip can tolerate.

TIR targets and definitions

TIR is the full needle sweep over one revolution, so eccentricity is half of it: a 0.08 mm TIR reading means the axis sits 0.04 mm off centre. That distinction matters when you compare a knife supplier’s figure against an assembled-spindle figure, because a knife-face reading and an assembled reading describe different physical errors.

Figures are working targets from slitting practice, not values fixed by a standard.

Knife and spacer tolerances

Precision-ground knives hold ±0.002 mm on thickness and diameter; generic tooling commonly sits at ±0.01 to ±0.05 mm. Hardness splits by material class: carbide runs around 88–92 HRA, M42 high-speed steel 65–67 HRC, M2 62–64 HRC and D2 58–62 HRC. Blade bores held within 0.01 mm on an H6/H7 fit keep the stack concentric on the arbor.

Maxtor Metal’s precision-ground roller shearing blades are manufactured and inspected to these targets as a standard production specification — not a special-order item — covering bore concentricity, face flatness and thickness tolerance in a single inspection record.

Notes for AHSS, stainless, aluminum

Runout that mild steel shrugs off will chip a blade on martensitic UHSS, where the practical limit is under 0.02 mm TIR and tighter for carbide. The AHSS Guidelines published by the Auto/Steel Partnership explain why: a mechanically sheared edge is made up of rollover, a burnished land, a fracture zone and a burr, and the balance between them shifts with clearance, shear angle, rake angle, grade and thickness. Excessive clearance raises burr and void formation, while insufficient clearance promotes secondary shear and accelerated tool wear. Burr limits are therefore usually expressed as a share of gauge rather than a fixed millimetre value — commonly ≤5% of thickness for cold-rolled strip, ≤7% for stainless and ≤8% for aluminium, with clearance bands of roughly 5–8%, 4–7% and 6–10% respectively.

For a root-cause matrix linking runout and clearance errors to specific burr signatures by material, see Slitting Burr Reduction: Root-Cause Matrix & Setup Windows.

Electrical-steel practice drives burr down to 10–15 µm, with lamination width tolerances tightening from around +0.00/−0.50 mm on wider stock to +0.00/−0.15 mm on narrow stock. The peer-reviewed work on this is instructive: a 2022 study in सामग्री on shear-slitting of 0.3 mm cold-rolled grain-oriented electrical steel evaluated coated and uncoated material and showed that cut-surface quality tracks both slitting speed and horizontal clearance — two parameters that interact, so tuning one in isolation rarely gives a stable edge (Experimental Research on Sheared Edge Formation in the Shear-Slitting of Grain-Oriented Electrical Steel, सामग्री, 2022). Core loss then makes the width excursion a performance question, not just a scrap question.

माप और नियंत्रण वर्कफ़्लो (Measurement and control workflow)

Runout numbers that move between checks usually mean the measurement is unstable, not that the tooling changed. Fix the method first, then the hardware.

Runout measurement steps

Clean and deburr the mounting surface, then mount the indicator on clean, flat, thick steel. Keep the stem perpendicular to the surface, preload the stem to a quarter to a third of its travel, and zero the needle at one angular mark. Rotate a full turn by hand in the normal cutting direction and record the maximum and minimum with their angular positions, not just the peak value. If the needle does not return to zero at the mark, suspect axial float or a loose base. Repeat at several axial planes: inboard bearing, midway, hub seat, seal area and free end.

Then measure the arbor first, the assembled knife second, at the OD near the cutting edge and on one face. Re-seat once and repeat: a material change means seating sensitivity, not true runout (see our central bore tolerance guidance for slitter knives). If the reading changes every revolution, suspect the setup, not the shaft.

Use a 1 µm indicator, with graduation significantly finer than the tolerance being inspected. Record cold and hot readings separately: steel grows about 1.2 µm per 100 mm per degree Celsius, and warm-state allowances run 0.020–0.035 mm against 0.015–0.025 mm cold.

Thickness verification and matching

Verify each blade before it reaches the arbor. Specify blade thickness to ±0.003 mm, bore concentricity within 0.01 mm and flatness within 0.02 mm, then confirm those figures on incoming inspection rather than trusting the certificate alone. For the full bore and face tolerance specification framework used in OEM procurement, see our OEM Slitter Knife Blueprint: Spindle Fit Audit Checklist.

Where a stack carries several knives, bin blades and spacers in one-micron classes and offset the signs when assembling, so a thick blade sits next to a thin spacer instead of compounding. Maxtor Metal supports this stage by verifying knife and spacer tolerances and supplying matched sets, which helps keep the assembled stack inside the width budget without sorting on the machine.

Setup controls: cleanliness, torque, clearance

Clean, burr-free, flat mating faces are the baseline. Apply the specified torque progressively in a cross pattern, and never torque to force runout away. Re-check runout after clamping, because a stack that reads true before tightening can shift once load is applied.

A field pattern from the shop floor

The troubleshooting sequence above is not theoretical. It mirrors the pattern Maxtor Metal’s application engineers see most often when a customer reports a width drift that “came out of nowhere.” In a typical case, the line has run the same knife pack for months, the operator has re-zeroed the arbor to chase the drift, and the width still walks. When the pack is broken down, the fault is rarely the knife itself: it is a worn spacer or a seating issue that changed how the stack sits, so the nominal arithmetic still looks correct while the cut does not.

The corrective order that resolves these calls is consistent, and it is the same order recommended in this section. Stop re-zeroing. Clean and deburr the mating faces, re-seat the stack, and re-torque to specification in a cross pattern before measuring anything. In the majority of these cases the runout reading falls back inside tolerance without a knife change, because the fault was seating rather than tooling. This is exactly why Maxtor Metal verifies knife and spacer tolerances and supplies matched sets, and why our engineers ask for the bench and hot readings side by side before recommending a replacement: the numbers tell you whether the problem is in the tooling or in the setup.

Clearance bands move with material and gauge. High-precision slitting practice keeps clearance between 0.001 in and 0.005 in depending on alloy: roughly 0.002–0.005 in for low-carbon, as tight as 0.001 in for high-carbon and alloy grades, and 0.001–0.003 in for stainless, against more conventional levels of 0.005–0.010 in. The mechanism is documented in the literature. A 2018 study in ISIJ International of strip trimming on a cold rolling line analysed the effect of overlap, clearance and front and back tension on edge quality and linked buckle edge, burr and unbalanced cut-off ratio to how those parameters are combined, not to any single setting (Strip Edge Defects and Comprehensive Control Technology in Cold Rolling Production Line, ISIJ International, 2018).

For how clearance and knife overlap interact to set edge quality at each material and gauge, see Rotary Knife Overlap Settings: Clean Edge Quality in Slitting.

चेतावनी: Torquing to chase runout masks a seating fault and loads the arbor.

केस स्टडी: स्लिट चौड़ाई को स्थिर करने के लिए अक्षीय रनout को नियंत्रित करना

A published Maxtor Metal field case shows how the stack-up logic plays out on a real line. The numbers below are from that published case, with the TIR figures stated as a representative measured range rather than a named customer’s raw log.

How to read these figures. Every number in this case comes from a single source — Maxtor Metal’s own published case record. No independent audit, third-party test report or equipment OEM bulletin was used to verify it, and no raw measurement log, gauge serial number or dated reading sheet is reproduced here. Treat the case as a worked, manufacturer-published illustration of the diagnostic sequence, not as a benchmark you can copy into a contract. The most defensible thing to take from it is the order of the checks and the reason each one was made; the specific percentages and micrometres should be re-established on your own line against your own measured width spread and the machine builder’s specification before you apply anything line-wide.

Where this case and the ±0.14 mm example elsewhere in the article appear to disagree, they are answering different questions: the example is a worst-case arithmetic sum of nominal tolerances, while the case values are measured deviations on one production line.

Line and target. A cold-rolled steel service center running SPCC-equivalent carbon steel at 0.50 mm, slitting a 1,250 mm master coil into 8 strands of roughly 156 mm each, to a width tolerance of ±0.05 mm at 120 m/min.

The symptom. Width was acceptable immediately after setup, then walked out of spec during the run. Initial width sat at 156.00 ±0.03 mm, but after 20–30 minutes of production it spread to 155.88–156.12 mm, an actual variation near ±0.12 mm against a ±0.05 mm requirement. Operators first blamed spacer thickness, knife wear, clearance and tension. Swapping the knife set did not clear it.

The measurement sequence that found it. (1) Finished width was sampled at the coil head, 25%, 50%, 75% and tail; the deviation rose and fell periodically instead of holding a constant offset, which ruled out a simple spacer arithmetic error. (2) Each knife was checked with a micrometer at several angular positions; no significant thickness error appeared, so replacing the whole set was not justified. (3) Spacers were removed and measured individually; several older spacers had light face wear and contamination whose individual errors were small but whose cumulative effect was significant. (4) The arbor was measured alone at 0.018–0.022 mm TIR, acceptable for that machine. (5) After assembly, the worst knife station read about 0.047 mm axial TIR — close to the commonly referenced 0.05 mm level. The key finding: the arbor was sound, and the large runout appeared only once the stack was assembled.

What failed first. Two attempts were logged before anything about the stack was touched, and the difference between them is the useful part of the record.

Attempt 1 — opening the clearance. Clearance was opened slightly. Burr height at the coil head dropped noticeably and stayed lower for the first few coils, so the operator had good reason to believe the problem was solved. Width variation was unchanged: the same periodic rise and fall returned within 20–30 minutes, and the tail of the coil still drifted past the ±0.05 mm limit. This separated two things that had been treated as one — clearance sets edge quality, width drift is set by position along the arbor.

Attempt 2 — fitting a new knife set. A complete new set went on. Initial slit quality improved and the first coil ran clean, but by mid-run the width had again walked toward ±0.12 mm. Because a fresh set reproduced the original symptom within the same coil length, knife wear could no longer be the primary cause. The new set cost a full changeover and a production window and changed nothing structural — the stack still held the error.

Attempt 3 — rebuilding the stack. The improvement appeared only after the spacers and seating were controlled: worn and contaminated spacers pulled out and matched by measured thickness, all contact faces cleaned, and the pack clamped in a defined sequence to the machine’s torque. The same knives that had already failed once in Attempt 2 then held width inside the customer specification.

The lesson worth copying is the order: a fix that changes the symptom without changing the drift is evidence pointing at the stack, and buying new knives before excluding seating spends money in the wrong place.

What worked. Rebuild the tooling stack as a controlled precision assembly: clean and individually measure spacers, match them by actual thickness, inspect for damaged faces, build precision within an engineering target of ±0.003–0.005 mm thickness where the machine and product justify it, clean all knife and spacer contact faces of oil film, metallic particles, grinding debris, bore burrs and face damage, and clamp uniformly with a defined tightening sequence and a machine-specific torque rather than “tighten until it feels tight.”

The result. Axial runout fell from about 0.047 mm to roughly 0.015–0.020 mm TIR, comfortably below the 0.05 mm reference level used in several published slitting documents. The core conclusion is the point to carry into your own setup: axial runout was not corrected by changing the knife alone. The improvement came from controlling the complete tooling stack: arbor condition → spacer faces → knife seating → clamping sequence → TIR verification.

How far to trust the result. The TIR figures are a representative measured range, not the raw log of a named customer, and the whole case is manufacturer-published. The before/after direction is repeatable; the exact micrometres are not a target to copy. Confirm the acceptable value with your machine OEM and your own width specification.

Operator behaviour that changed. Before the improvement, operators answered width drift by adjusting clearance, changing strip tension, slowing the line or occasionally replacing knives — steps that sometimes masked the symptom without removing the periodic variation. Afterward they followed a fixed sequence: width check → burr check → arbor/knife TIR check → spacer verification → clearance/overlap adjustment, instead of reaching first for the knife settings.

On-site troubleshooting order

The sequence below is the reusable part of this case. Work down the list and stop at the first step that fails; do not skip ahead to knife settings, and do not change a parameter that has already been cleared.

If steps 1 to 7 all pass and width still drifts, the next stop is the incoming coil — residual stress and poor flatness can hold the error in even after the tooling is correct.

Process limitation

The runout reduction was not treated as a standalone fix, and it should not be read as one. The result depended on a sound arbor, acceptable bearing condition, correctly ground knives, clean spacer faces and controlled knife clearance. If the arbor itself carries excessive runout, bearing play or shaft deflection, replacing spacers alone cannot restore precision. If the incoming coil has significant residual stress or poor flatness, width variation may persist even after the tooling is correctly aligned. Arbor precision, rigidity, bearing condition, tooling accuracy and assembly act as interdependent factors in rotary slitting rather than as separate levers.

KPI और ROI प्रभाव (KPIs and ROI impact)

Width yield and scrap risk

Width yield follows the stack, not the setpoint. A machine can hold its commanded position perfectly and still ship off-width coil if the knife pack carries the error, which is why tolerance classes are quoted against the assembled stack rather than against individual components.

Two different numbers appear in this guide, and they are not interchangeable. The ±0.14 mm figure in the stack-up sections is a worst-case example: 14 knives × ±0.01 mm of blade thickness tolerance, summed at full value, before spacers or arbor runout enter. The figures in this paragraph are tolerance classes — the width band you can expect to hold with a given grade of knife and spacer set, measured on a real line.

Using the same 14-knife geometry as a consistent reference point, a commercial tooling class accumulates to roughly ±0.14 mm, a precision class to ±0.05 mm and an ultra-precision class to ±0.025 mm. The worst-case sum and the commercial class therefore land on the same order of magnitude, which is the practical warning: buy commercial-grade tooling onto a wide arbor and you are already sitting at the edge of a ±0.05 mm specification before anyone measures a single blade. The tolerance class you quote sets the scrap exposure: a mill quoting commercial class on a narrow, high-value strip is accepting a width band several times wider than the process could otherwise hold.

Gauge compounds the problem. Published standard tolerances widen as gauge rises: EN 10140 bands cold rolled narrow strip by thickness and width, tightening to around ±0.15 mm at 0.60 mm and opening to ±0.50 mm or beyond at the heavy end (BS EN 10140:2006 cold rolled narrow steel strip, BSI). Stainless follows the same pattern under ASTM A480/A480M, whose general requirements set the dimensional tolerances for flat-rolled stainless strip. Thicker strip bevels and resists width control, so the same knife pack that holds a thin gauge will drift on heavy gauge. Downstream, the cost shows up as function rather than scrap alone: a 0.2 mm width change on a 250 mm lamination is worth about 0.13% in core loss, which turns a width excursion into a performance claim on the finished core.

Metric targets above are illustrative improvement benchmarks based on Maxtor Metal field experience; actual values depend on arbor condition, material grade and tooling class. For measured case-study figures, see the Case Study section below.

The same logic shows up in the field case below, where a width drift traced back to the tooling stack rather than the setpoint.

Width deviation, burr, stoppage and setup figures are from the published Maxtor Metal field case — a single manufacturer source, not an independently audited dataset. The TIR row is a representative measured range. Rows are labelled by measurement type so they are not confused with the ±0.14 mm worst-case arithmetic example used in the stack-up sections. The allowable value should always come from the machine OEM and your product tolerance rather than one universal number.

Uptime and setup time

Setup time is where tolerance discipline pays back fastest, because the same stack error that costs yield also costs re-work. A documented re-seat check shortens the next setup: when the operator records axial position and cold versus hot readings at first install, the following changeover starts from a known reference instead of a fresh guess. Thermal drift is the reason the record matters. A warm-state allowance of 0.020–0.035 mm covers the growth between a cold arbor and a machine that has run for an hour, and a setup that ignores it will be re-trimmed once the line reaches temperature.

Tool life and cost per ton

Tool life is bounded by two independent ceilings. Runout above 0.02 mm chips carbide on UHSS, so the axial condition of the pack sets how long an edge survives regardless of its grade. Hardness class sets the tool-life ceiling from the other direction: carbide at 88–92 HRA against M2 at 62–64 HRC is a different wear regime, not a marginal improvement. The tolerance class of the knife sets how often it is re-ground, and that interval, not the purchase price, is what determines cost per ton.

For a full ROI comparison between carbide, PM-HSS and tool steel for slitting applications, see Rotary Slitter Knife ROI: PM-HSS vs Tool Steel.

FAQ

How tight does knife TIR actually need to be?

It depends on the width class you are selling into, not on a universal number. Common slitting practice puts the maximum acceptable knife-face runout at 0.005 mm for precision work and 0.010 mm for standard work, and recommends assembled knife TIR of 0.01 mm or tighter on high-precision lines, with 0.01 to 0.03 mm accepted on general industrial lines. Those are OEM and technical targets in roughly the 5 to 50 µm band, not standard values: ISO 1101:2017 defines how runout is stated on a drawing but publishes no slitter acceptance limit. Pick the target from the width tolerance you must hold, then confirm it against the machine builder’s specification.

Why does blade thickness tolerance matter more than runout on a wide arbor?

Because thickness errors accumulate along the arbor while runout is local to each station. Strip width is built from the ground thickness of every component stacked between the two cutting edges, and those errors do not cancel. A worked illustration: ±0.01 mm per blade across 14 knives stacks to roughly 0.14 mm of position error at the last strip, which alone can consume a ±0.05 mm width specification; tightening blade thickness tolerance to ±0.003 mm brings the same stack to about 0.04 mm. Treat that as a worst-case illustration with stated assumptions, not measured plant data.

Should I use worst-case or RSS stack-up to set my tolerances?

Use both, for different decisions. Worst-case arithmetic tells you the guaranteed envelope: if the sum still fits your width spec, the design is safe without any statistical assumption. RSS assumes independent, normally distributed deviations and typically cuts a five-part chain by 40 to 60% against worst case. A published slitting example shows the gap: three components at ±0.025 mm give ±0.075 mm worst case against roughly ±0.043 mm RSS. Set the acceptance limit on worst case; use RSS to decide where tightening buys the most margin per dollar.

Which contributor should I tighten first?

Rank by sensitivity, not by which tolerance looks worst on paper. Multiply each contributor’s tolerance by its influence coefficient in the stack, then compare the resulting width contribution; the largest product is your first target. In most slitting stacks the order runs blade thickness tolerance across the full knife count, then spacer flatness and parallelism, then assembled face runout, then thermal drift. Tightening a contributor with a small influence coefficient buys almost nothing, however impressive the new tolerance looks on a certificate.

Does the width tolerance I quote come from a standard or from my own capability?

Both, and they answer different questions. Slit-edge width tolerances for cold-rolled narrow strip are banded by thickness and width in EN 10140, which lists values such as ±0.15 mm at 0.60 mm thickness and ±0.50 to ±0.60 mm near 4 to 10 mm (BS EN 10140:2006 cold rolled narrow steel strip, BSI). Stainless strip is covered separately by ISO 9445-1:2009 for narrow strip and cut lengths. Those tables define what the product standard permits, not what your line can hold repeatably. Quote the standard band to the customer, but set your internal control limit from your own measured width spread, then keep the gap between them as your margin.

Where does the tolerance stack-up start, at the blade or at the spacer?

At the blade, because the blade’s ground thickness is the first term in the chain and every spacer behind it adds to the same sum. Spacers matter through flatness and parallelism rather than nominal thickness: a spacer that is flat but slightly thick shifts the whole downstream stack, while one that is bowed tilts the knife and shows up as face runout instead. Audit blade thickness first, then spacer geometry, then seating. Re-zeroing the arbor after a width drift without re-seating the stack usually moves the error rather than removing it.

Do AHSS, stainless and aluminium need different TIR targets?

Yes, mainly because the material changes how much of the stack error reaches the cut edge. Higher-strength strip springs back more after shearing, so a given face runout produces a larger width spread than it would in mild steel; stainless work-hardens at the edge and tolerates less knife tilt before burr height climbs. Aluminium is softer and more sensitive to built-up edge and thermal drift than to runout itself. Keep the same measurement method across materials, but tighten the assembled TIR target for AHSS and stainless rather than applying one number line-wide.

How often should I re-measure runout after a knife change?

Measure after every knife or spacer change, and again after the first production coil, because the first heat cycle settles the stack. A knife that measured 0.008 mm on the bench can read 0.02 mm once the arbor is torqued and warmed, and that shift is seating and thermal drift, not a bad blade. Log the bench reading and the hot reading separately so you can tell a genuine tooling problem from a setup one. If the hot reading drifts coil after coil, look at clamping force and cleanliness before you replace knives.

निष्कर्ष

Width variation is a stack-up problem, and the sub-0.01 mm contributors govern the result: control the sub-0.01 mm contributors and the lane tolerance follows. Hold knife TIR to 0.01 mm and assembled runout to 0.01–0.03 mm, then specify to worst-case and monitor to RSS so the arithmetic ceiling stays honest while daily data stays realistic.

The Monday-morning sequence is short. Measure assembled runout first. If it exceeds 0.02 mm, tighten TIR before touching anything else. With high knife counts, tighten blade thickness tolerance next. When spacer flatness is eating your clearance budget, tighten the spacers. Upgrade components only after arbor and bearing faults are excluded, because no knife change tunes out a bent arbor.

In practice, this is the order Maxtor Metal’s engineers work through with customers, and it is built on more than a decade and a half of supplying precision-ground blades into steel, automotive, electrical-steel and general processing lines worldwide. The company’s quality system covers material inspection and multiple production checks before shipment, and its ODM and OEM capability means a matched knife and spacer set can be built to your drawing or sample rather than bought off a catalogue and sorted on the machine.

If you would rather not build that capability from scratch, Maxtor Metal supplies matched knife and spacer sets with documented tolerances. Send a drawing for a matched-set quote, or ask an application engineer to review your current stack-up against your target strip width tolerance.

लेखक के बारे में

Jerry Chu is Technical Support Specialist at Maxtor Metal, with over a decade of field experience diagnosing width drift and axial runout issues on coil slitting lines processing carbon steel, AHSS, stainless and aluminium.

His technical focus for this guide is axial runout and stack-up control: why a knife pack that measures true on the bench can walk out of a strip-width specification during a production run, how blade and spacer tolerances accumulate along the arbor, and how arbor condition, spacer face quality, knife seating and clamping sequence interact to set — or lose — a stable cut window. The stack-up arithmetic, TIR measurement sequence and troubleshooting order in this article reflect the diagnostic approach Jerry applies in Maxtor Metal’s after-sales support engagements.

He holds the PMP (Project Management Professional) and CMRP (Certified Maintenance & Reliability Professional) certifications, supporting a structured, reliability-centred approach to tooling and process decisions.