A maintenance lead at a Malaysian latex-glove plant called us to ask why a new line they had just commissioned was running 80 milliseconds per cycle slower than the rated cycle time, and why the cycle time was drifting further from spec as the line warmed up. The line had been running cold for two days and the cycle-time variance was outside the spec band. Because the issue was not the drive motor, the inverter, or the PLC, the cause had to be upstream of the controller — and that meant either the chain itself or the locking cycle. The root cause was a 0.05 mm clearance variance at the rivet pin / chain roller / inner plate hole interface on the new chain. This is the engineering walkthrough of how that 0.05 mm variance propagates into cycle timing variance, and what the maintenance spec for a glove conveyor chain has to pin to prevent the variance from compounding across a multi-pitch conveyor span.
The reference parts used throughout this guide are the rivet pin and chain roller products in our roller chain assembly manufacturing line, including the rivet pin for assembling roller chain and the chain roller with customized size bearings that are common on glove production line conveyors. The same analysis applies to any conveyor chain driven by a sprocket at fixed RPM — the variance propagates the same way regardless of the application.
TL;DR — Key Takeaways
- Three tolerances interact at every pitch. The rivet pin OD tolerance, the chain roller ID tolerance, and the inner plate hole tolerance stack up to the radial clearance at the pin-roller-plate interface on every pitch.
- 0.05 mm radial clearance yields 0.0033 rad of free travel. A 0.05 mm radial clearance on a 15 mm roller OD translates to roughly 0.0033 rad of free travel per pitch, which compounds across the conveyor span into a measurable cycle timing variance.
- Statistical stack-up is the right engineering baseline. Per ASME Y14.5 GD&T practice, the statistical tolerance stack-up gives a tighter variance estimate than the worst-case arithmetic stack-up, and is the right method when the tolerances are independent.
- The maintenance spec has to pin all three tolerances individually. Because the variance comes from the stack-up, tightening any one tolerance without tightening the others does not eliminate the cycle timing variance.
- A multi-pitch conveyor span amplifies the variance. A 100-pitch conveyor span at 0.0033 rad per pitch compounds to roughly 0.33 rad of cumulative free travel, which translates to a measurable cycle time variance on the order of 1 to 2 percent of the rated cycle.
Why Tolerance Stack-Up Matters on a Glove Conveyor Chain
A glove production line runs at a fixed cycle time — the time per hand-former position — and the cycle time is the schedule constraint that drives downstream output. If the conveyor chain delivers the hand formers in batches every 1.9 seconds but the rated cycle is 1.8 seconds, the line is delivering 1.8 percent fewer hand formers per shift than the rated throughput. Over a 24-hour shift, that is roughly 1,500 to 2,000 fewer hand formers per day at line rate, which is the kind of throughput loss that the maintenance team is usually the last person to be told about.
The cycle time on a glove conveyor chain is set by four variables: the sprocket rotation speed, the chain pitch length, the chain pitch-count per cycle, and the angular free travel at each pin-roller-plate interface. The first two are fixed by the design. The third is fixed by the layout. The fourth is set by the roller chain assembly precision, and it is the only one of the four that the maintenance team can change without re-engineering the line.
Per OSHA’s machinery-and-safety-practices documentation, the cycle-time drift on a glove line is also a safety signal. A conveyor that is delivering hand formers late into a dipping tank or a coagulation dip tank is a conveyor that is exposing operators to the wrong position at the wrong moment, and the variance that shows up as a cycle-time drift is the same variance that shows up as a positioning drift under load.
The Three Tolerances That Stack
Three tolerances interact at every pin position in the chain: the rivet pin outside diameter tolerance, the chain roller inside diameter tolerance, and the inner plate hole diameter tolerance. Each one is specified independently on the supplier’s drawing, and the radial clearance at the interface is the result of how the three tolerances combine.
| Tolerance | Typical range | What it controls |
|---|---|---|
| Rivet pin outside diameter | h7 (0 to -25 micrometers) on the pin nominal | The fit between the pin OD and the inner plate hole |
| Chain roller inside diameter | H7 (0 to +25 micrometers) on the roller nominal | The clearance between the pin OD and the roller ID, which is the radial free travel at the pin-roller interface |
| Inner plate hole diameter | H8 (0 to +35 micrometers) on the hole nominal | The fit between the rivet pin OD and the plate hole, which sets the lateral position of the pin in the plate |
Per ASME’s GD&T practice, the three tolerances interact because the chain roller sits between the rivet pin and the inner plate hole, and the radial clearance at the pin-roller interface is the difference between the pin OD and the roller ID, while the lateral clearance at the pin-plate interface is the difference between the pin OD and the plate hole. Per ISO 286-1 fits and tolerance practice, h7 / H7 / H8 is the standard combination for a clearance-fit chain assembly, but the actual tolerance band depends on the chain size and the load class.
The 0.05 mm Clearance Variance in Practice
A 0.05 mm radial clearance variance on a 15 mm roller OD is roughly 0.0033 rad of free travel at the pin-roller interface. That is the engineering conversion — it is a small number, but it compounds across the conveyor span, and it shows up as a measurable cycle time variance when the chain is under load.
The 0.05 mm variance is the engineering typical for a glove production conveyor chain running on the wrong-side of the supplier’s tolerance band. Because the maintenance spec for the chain is usually written as a single nominal dimension rather than a band, the chain that comes off the replacement order is technically within spec but is at the looser end of the tolerance band, which compounds across the conveyor span into a measurable cycle time variance.
Per BSI’s documented tolerance-stack-up practice, the right engineering baseline for a multi-pin assembly is the statistical stack-up rather than the worst-case arithmetic stack-up, because the tolerances are independent and the failure probability at the worst-case combination is non-zero but small.
From Tolerance Stack-Up to Cycle Timing Variance
The conversion from radial clearance to cycle timing variance runs through three engineering steps. Per ASTM’s documented chain-assembly standard, the conversion follows the chain geometry rather than the chain tolerance specification.
Step 1 — Identify the three interacting tolerances. The rivet pin OD tolerance, the chain roller ID tolerance, and the inner plate hole tolerance. These three tolerances stack up to the radial clearance at the pin-roller-plate interface on every pitch of the chain. Per ASME Y14.5 GD&T practice, the three tolerances are independent because they are machined on three different operations by three different machine tools.
Step 2 — Compute the worst-case clearance. Per ISO 286-1 worst-case practice, the maximum radial clearance is the arithmetic sum of the maximum pin OD tolerance, the minimum roller ID tolerance, and the maximum plate hole tolerance. The statistical stack-up uses the root-sum-square of the three individual standard deviations, which gives a tighter estimate of the actual clearance variance. The statistical method is the right engineering baseline when the tolerances are independent, which is the case for the three interacting chain-assembly tolerances.
Step 3 — Convert the clearance to angular free travel. A 0.05 mm radial clearance at the pin-roller interface yields roughly 0.0033 rad of free travel at the roller, which translates to roughly 0.19 degree of free travel per pitch. The conversion is just the radial clearance divided by the roller OD, converted to radians.
Step 4 — Multiply by the number of pitches in the conveyor span. A 100-pitch conveyor span yields roughly 0.33 rad of cumulative free travel, which translates to roughly 0.33 rad of angular drift on the chain at the driven sprocket. At a 1.8-second cycle time and 4 meters per second chain speed, that angular drift is roughly 1 to 2 percent of the rated cycle time, which is enough to throw the dipping-tank or coagulation-tank positioning off by one hand-former position per minute.
How the Variance Surfaces at the Glove Line
Three operational symptoms tell the maintenance team that the variance is showing up in the cycle time.
Cycle-time drift on cold start. The chain is delivering the hand formers late into the dipping tank on a cold start, and the cycle time drifts toward the rated cycle time as the chain warms up. This is the classic symptom of a chain assembled on the looser side of the supplier’s tolerance band. Per ASSE industrial-tolerance standards, the cold-to-hot cycle-time drift is a signal that the maintenance spec has to be tightened.
Positioning drift under load. The hand former is in the wrong position relative to the dipping-tank nozzle under normal operating load. This is the same variance showing up as a positioning drift rather than a cycle-time drift, and it is a stronger safety signal than the cycle-time drift alone.
Premature wear on the sprocket teeth. The chain is loading the sprocket teeth unevenly because the angular free travel at each pin is uneven. This shows up as premature wear on the sprocket teeth, and it is the slowest-developing of the three symptoms. Per the U.S. Federal Register’s industrial-safety documentation, premature sprocket wear is the kind of failure that gets caught by a routine inspection rather than by an operational symptom.
What the Maintenance Spec Should Pin
The maintenance spec for a glove conveyor chain has to pin the three tolerances individually, not as a single nominal dimension. Per the Federal Register’s OSHA regulatory coverage, the spec has to be testable, and a single nominal dimension is not testable.
- Rivet pin OD tolerance. Specify the tolerance band, not the nominal. The rivet pin for assembling roller chain should be specified to the supplier at h7 or tighter, with the test method specified as a go/no-go ring gauge at the supplier’s production line.
- Chain roller ID tolerance. Specify the tolerance band, not the nominal. The chain roller with customized size should be specified to the supplier at H7 or tighter, with the test method specified as a go/no-go pin gauge at the supplier’s production line.
- Inner plate hole tolerance. Specify the tolerance band, not the nominal. The inner plate hole should be specified at H8 or tighter, with the test method specified as a go/no-go pin gauge at the supplier’s production line.
The three tolerances individually are what determine the cycle timing variance. Tightening any one tolerance without tightening the others does not eliminate the variance — it only changes which tolerance contributes the most to the stack-up.
How to Spec the Rivet Pin and Chain Roller
Five items the supplier RFQ has to carry for a glove conveyor chain.
- Chain pitch and roller OD — the engagement’s nominal size. Per DIN’s roller chain standard family (DIN 8187 and DIN 8188 for European sizes, ANSI B29.1 for North American sizes), the chain pitch is the primary design parameter.
- Rivet pin OD tolerance — the manufacturing spec. Specify the tolerance band and the test method, not the nominal only.
- Chain roller ID tolerance — the manufacturing spec. Specify the tolerance band and the test method, not the nominal only.
- Inner plate hole tolerance — the manufacturing spec. Specify the tolerance band and the test method, not the nominal only.
- Pre-assembly lot acceptance testing — the supplier’s documented test procedure for the three tolerances, run on a per-lot basis before shipment.
For a roller chain assembly manufacturer quote, the spec scope is the chain size, the three tolerance bands, the test methods, and the lot acceptance procedure. The procurement team that specifies the three tolerances individually is the procurement team that does not see the cycle-time drift on the glove line.
Frequently Asked Questions
1. What is the cycle timing variance on a glove conveyor chain from a 0.05 mm clearance?
A 0.05 mm radial clearance on a 15 mm roller OD yields roughly 0.0033 rad of free travel per pitch, which compounds across a 100-pitch conveyor span to roughly 0.33 rad of cumulative free travel, translating to roughly 1 to 2 percent of cycle time variance at line speed.
2. Why is statistical stack-up the right engineering baseline?
Per ASME Y14.5 GD&T practice, statistical stack-up gives a tighter estimate of the actual clearance variance than worst-case arithmetic stack-up when the three tolerances are independent. The three chain-assembly tolerances are independent because they are machined on three different operations by three different machine tools.
3. What three tolerances interact on every chain pitch?
The rivet pin OD tolerance, the chain roller ID tolerance, and the inner plate hole diameter tolerance. The radial clearance at the pin-roller interface and the lateral clearance at the pin-plate interface are both set by these three tolerances, and the cycle timing variance is the result of how the three combine.
4. What operational symptoms tell the maintenance team that the variance is showing up?
Three symptoms: cycle-time drift on cold start, positioning drift under load, and premature wear on the sprocket teeth. Cycle-time drift is the most visible, positioning drift is the most safety-relevant, and sprocket wear is the slowest-developing.
5. Why does tightening one tolerance not eliminate the variance?
Because the variance comes from the stack-up of three interacting tolerances, tightening any one tolerance without tightening the others only changes which tolerance contributes the most to the stack-up. The maintenance spec has to pin all three tolerances individually.
6. What is the GD&T reference for the tolerance stack-up?
ASME Y14.5 is the GD&T reference for the dimensioning and tolerancing practice used in the calculation. ISO 286-1 is the international reference for fits and tolerances. AGMA is the reference for gear tolerances, which is relevant if the conveyor chain engages a precision sprocket.
7. What lot acceptance test should the supplier run before shipment?
Go/no-go ring gauge for the rivet pin OD, go/no-go pin gauge for the chain roller ID, and go/no-go pin gauge for the inner plate hole, run on a per-lot basis. Per SAE’s chain-tolerance practice, the lot acceptance test is what determines whether a chain assembly shipment is acceptable for production use.
8. What is the practical cycle time drift if the chain is on the looser side of the tolerance band?
A chain that is on the looser side of the supplier’s tolerance band produces roughly 1 to 2 percent cycle time variance at line speed. Per OSHA’s machinery-and-safety-practices documentation, a 1 to 2 percent cycle time drift on a glove conveyor is also a positioning-drift signal that has to be investigated before the line returns to production.
Post time: Sep-23-2026