Printed Wax Pattern Shrinkage Compensation: What to Scale

Printed wax pattern shrinkage compensation is set by scaling the pattern inversely before it is printed: a casting that shrinks 1.5–2.5% needs a pattern that is 1.5–2.5% larger, and the first article is then measured to correct the factor. The number is not a single constant; it moves with alloy, section and the way the mould constrains the part.

Getting this wrong is expensive in a quiet way. A pattern printed at nominal size produces a casting that is uniformly small, and the error is only visible after the first pour, when the flange is already below its machining line and the pattern has to be reprinted.

What sets printed wax pattern shrinkage compensation?

The first input is the alloy's solidification shrinkage. Stainless casting steel sits in the 1.5–2.5% band used by most foundries, and the exact value depends on grade and section thickness rather than on a handbook constant that applies to everything.

The second input is the drawing convention. If the CAD is the finished metal part, the pattern has to carry the whole compensation. If the CAD is already compensated pattern geometry, applying the factor again double-compensates and produces a casting that is too large. The RFQ should state which geometry is supplied.

The third input is section thickness. A thick flange solidifies slowly and pulls the thin wall with it, so the effective shrinkage across a manifold is not uniform; a single factor is a starting point, not a finish line.

The practical output is a pattern scale that the foundry and the pattern supplier agree on before printing. Writing "1.5–2.5%" into the drawing without saying whether it is applied to the pattern or to the casting is how the same number produces two different parts.

How is the first article used to correct it?

The first article is measured on a CMM against the pattern model, and the difference is converted back into a shrinkage correction. If the measured flange is 0.4% small, the next pattern is printed 0.4% larger in the affected directions rather than by a blanket percentage.

One correction round is normal for a thin-wall, long-runner manifold. The first pour confirms the fill and the gating; the second confirms the compensation. Projects that skip the measurement and print a batch of patterns usually pay for the batch twice.

The correction has a cost: a new pattern, a new shell and another furnace slot. That is still cheaper than machining a casting that has no stock on the flange, which is why the measurement belongs in the first article and not in the third.

The measurement should cover the flange faces and the runner ends, not only overall length, because shrinkage is not isotropic: the long axis and the wall-thickness direction move by different amounts, and one overall dimension can hide a flange that is out of position.

What happens when the wall is thin?

Thin walls move for a second reason. A 1.0 mm wall cools faster than the flange beside it, and the constraint of the flange turns that differential into a bow rather than a uniform contraction. The measured movement at the flange face is typically 0.3–1 mm on a thin-wall manifold, and it does not disappear when the shrinkage factor is corrected.

The design answer is allowance rather than a better factor: 0.5–1.5 mm of straightening stock on the flange, cut after stress relief and re-datuming. The machining plan assumes that stock exists, so the pattern has to leave it.

Finally, the compensation record belongs with the casting. If a later revision changes the alloy or the wall, the factor has to be re-established, because the movement that was measured belongs to the old combination and not to the new drawing.

Send the CAD with its compensation status to Neway; we will confirm whether the scale applies to the pattern or to the casting before the first pattern is printed.