Date | 2022-11-23 15:51:16
A single molded part either meets the drawing or it does not. A production run is a different question: what matters is how far the parts drift from the nominal as the tool heats up, as the material lot changes and as the press operator changes shift. Buyers who assess a supplier on one sample are measuring the wrong thing, and suppliers who quote only a nominal tolerance are answering the wrong question.
For a thermoset like BMC the drift has a specific character. The part does not shrink on cooling the way a thermoplastic does; most of the dimensional change happens in the first hours after demolding, as the cure finishes and residual stress relaxes. That means the number that matters is not the dimension at the press but the dimension after the part has settled — and the second number is the one that has to be controlled.
Five variables account for most of the spread in a well-run BMC job. Material lot is the first: fibre length, filler loading and catalyst level shift slightly between batches, and each shift moves the flow and the shrinkage. Charge weight is the second, because a heavier charge raises cavity pressure and packs the part differently. Mold temperature and cure time together are the third, since an under-cured part continues to move after it leaves the press.
The fourth is the tool itself. A mold that is thermally unbalanced — hotter on one side, or with a cold insert — produces parts that differ across the cavity, so a single measurement tells you nothing about the part at the other end. The fifth is handling: parts stacked warm, or dropped into a bin while still soft, take a set that no amount of inspection afterwards will remove.
The table maps each source of variation to the dimension it moves and to the control that actually reduces it. Note that only one of the five is a machine setting; the rest are incoming material, tooling and handling decisions.
| Source | What it moves | Control |
|---|---|---|
| Material lot | Shrinkage and flow length between batches | Incoming inspection on each lot; first-off approval before the lot runs |
| Charge weight | Cavity pressure, then wall thickness and flash | Weighed charge against a fixed cavity volume |
| Mold temperature and cure time | Post-molding shrinkage and residual stress | Cavity-surface temperature logging; cure verified by the qualified cycle |
| Thermal balance of the tool | Cavity-to-cavity and side-to-side differences | Map the mold surface temperature, not the platen; balance the circuits |
| Handling of warm parts | Flatness and straightness | Fixtured cooling or flat trays; no stacking before the part is cool |
Inspecting a part straight off the press is useful for setting up and misleading for control. The dimension that ships is the dimension after settling, so the inspection point should be defined in the drawing or the control plan: measured at 24 hours, at a stated temperature, on a stated fixture. Without that definition two inspectors can disagree honestly and neither is wrong.
For volume production the useful record is a short control chart on the two or three dimensions that the assembly actually depends on, taken at a fixed interval, with the trend reviewed rather than the individual points. A dimension that drifts steadily over a shift is telling you about thermal balance or tool wear; a dimension that jumps between lots is telling you about material. Reading the pattern is more valuable than reacting to the number.
Not on its own. A tolerance is a limit, not a capability. What improves consistency is removing the sources of variation listed above and then measuring the result over a run long enough to include a material change and a shift change. A tolerance set tighter than the process can hold simply moves the problem to the inspection bench.
It depends on wall thickness and on how close the cure is to completion when the part is demolded, so it is established per part during qualification rather than assumed. The practical approach is to measure the same parts at intervals until two consecutive readings agree; that interval becomes the inspection rule for that part.
Usually larger when measured warm, because thermal expansion at demolding temperature outweighs the shrinkage still to come, but the balance differs between grades and between thick and thin sections. The direction has to be measured for the specific part rather than reasoned from first principles, which is exactly why the settling study comes before the control plan.
For technical data sheets, grade recommendations, or a discussion about your specific part and production volume:
📧 wendy.qiu@smcbmc.com 📞 +86-13868305300