PA66+GF30 warpage is controlled by three levers in this order: balanced cooling, CRIMS-validated Moldflow, and staged packing. In a case where a part measured 1.8 mm (0.07 in) diagonal warpage at T0, these three steps brought it to 0.3 mm — an 83% reduction.
Compliance: Jinsu PA66+GF grades meet REACH and RoHS requirements. Third-party test reports are available on request.
Cooling rate determines whether PA66 keeps shrinking after demolding
PA66 reaches maximum crystallinity of 40–55%, and that value depends on how fast it cools through the 180–210°C window. Cooling at 5°C/min yields 48–52% crystallinity and 1.5–1.8% total mold shrinkage, with only 0.1–0.2% additional shrinkage in the following 24 hours. Cooling at 100°C/min drops crystallinity to 28–35% and total shrinkage to 0.8–1.1%, but adds 0.6–1.0% of post-crystallization shrinkage after demolding.
| Cooling rate (°C/min) | Final crystallinity (%) | Total mold shrinkage (%) | 24-hour shrinkage increment (%) |
|---|---|---|---|
| 5 | 48–52 | 1.5–1.8 | 0.1–0.2 |
| 20 | 42–47 | 1.3–1.6 | 0.3–0.5 |
| 50 | 35–40 | 1.0–1.3 | 0.5–0.8 |
| 100 | 28–35 | 0.8–1.1 | 0.6–1.0 |
A part that measures in tolerance at ejection can therefore move out of tolerance by the next morning. Measure PA66 dimensions after 48 hours of conditioning under ISO 291 standard atmosphere, not at the press.
Glass fiber turns shrinkage into a three-direction differential
Glass fiber shrinks essentially nothing (CTE about 5×10⁻⁶/°C) while the PA66 matrix shrinks far more (about 80×10⁻⁶/°C). The result is directional shrinkage that no single shrinkage value can describe: 0.2–0.4% parallel to fiber orientation, 0.6–1.2% perpendicular to it, and 1.0–1.5% through the thickness.
That ratio — roughly three to one between perpendicular and parallel — is the arithmetic reason GF nylon warps. It also explains why a shrinkage figure entered into a mold design as one number will always be wrong in one direction or the other.
Grade selection reduces the ratio before the mold is cut
Low-warpage formulations narrow the perpendicular-to-parallel gap at the material level. Measured anisotropic shrinkage below is per ISO 294-4.
| Grade | Reinforcement | Shrinkage parallel / perpendicular (%) | Typical fit |
|---|---|---|---|
| PA66-GF30-LW | GF30 + low-shrink additive | 0.25 / 0.45 | Precision housings and covers |
| PA66-GF30-MH | GF20 + 15% mineral hybrid | 0.30 / 0.50 | Large thin-wall panels |
| PA66-GF50-HD | GF50 + optimized nucleation | 0.20 / 0.40 | Structural brackets |
| PA66-IMF | Impact-modified low-warpage package | 0.35 / 0.55 | Snap-fit assemblies |
Default Moldflow settings under-predict warpage by four times
On the same engine cover, three analysis rounds produced three very different numbers against a measured 1.8 mm. Default settings predicted 0.4 mm — under by more than 4×. Adding the RSC fiber-orientation correction and measured PVT data raised the prediction to 1.5 mm, still 17% low. Adding CRIMS data and transient cooling analysis produced 1.75 mm, a 2.8% deviation from measurement.
The parameters that carry the most weight are 3D tetrahedral mesh (a midplane mesh cannot resolve through-thickness shrinkage) and the CRIMS residual-stress model. Without supplier CRIMS data, Moldflow defaults to generic values and the prediction reverts toward the optimistic case.
Conformal cooling removes the temperature difference that drives warping
Drilled channels run straight, so channel-to-cavity distance varies across a complex part. Varying distance means varying cooling rate, which means varying crystallinity, which means varying shrinkage. Conformal channels printed by SLM/DMLS follow the cavity contour and hold cooling pitch at 1.5–2.5× wall thickness.
Across the projects Jinsu has supported, conformal cooling cut GF30 warpage by 40–60% and cycle time by 15–25%. The trade-off is tooling cost: DMLS printing plus post-processing typically exceeds the cost of drilled channels, so the case rests on volume and tolerance.
Comparing warpage-control routes across PA66+GF30 suppliers
The comparison below is based on the material data and analysis support each route typically provides. Verify current values against the supplier's own TDS before specifying.
| Route | Anisotropic shrinkage, parallel / perpendicular (%) | CRIMS data supplied | Moldflow support included | Typical lead time |
|---|---|---|---|---|
| Jinsu PA66-GF30-LW | 0.25 / 0.45 | Yes, on request with sample kit | Yes, warpage analysis review | Sample 5–10 kg, trial order from 25 kg |
| Jinsu PA66-GF30-MH | 0.30 / 0.50 | Yes, on request with sample kit | Yes, warpage analysis review | Sample 5–10 kg, trial order from 25 kg |
| Standard GF30 grade, general market | 0.3–0.5 / 0.8–1.2 | Not typically released | Not included | Varies by distributor stock |
| Branded global GF30 grade | 0.3–0.4 / 0.7–1.0 | Available to qualified projects | Charged separately | Often 6–12 weeks |
Staged packing redistributes stress instead of only adding material
Packing does two jobs in a GF30 part: it compensates volumetric shrinkage and it shifts where residual stress sits. A single-stage profile pushes stress toward the gate end; a multi-stage profile that drops pressure as the gate freezes distributes it more evenly. Because stress relaxation is a creep process measured in hours, an asymmetric distribution keeps deforming the part long after ejection — which is the mechanism behind the 24-hour change.
Where the 1.8 mm part ended up
The engine cover reached 0.3 mm after gate relocation from one side of the part to the center of the long axis, combined with an optimized cooling layout. That is a reduction from 1.8 mm to 0.3 mm, or 83%, on the same tool and the same material.
Working on a PA66 warpage issue now? Send your 3D part file to sales@maxnmtech.com for a warpage review, or request a low-warpage PA66 sample kit including LW and mineral-hybrid grades. Application engineers reply within 4 hours, no obligation.
Values are typical and not intended for specification purposes. Shrinkage per ISO 294-4 / ASTM D955. Actual performance depends on part geometry, processing conditions, material batch, and mold design. Verify current TDS and certified test reports with Jinsu before production.
Frequently asked questions
Why does a PA66+GF30 part warp only after 24 hours?
Crystallization is incomplete at ejection. Amorphous chains keep crystallizing at room temperature, adding 0.3-0.8% shrinkage that is uneven across wall thickness.
What mold temperature keeps PA66 dimensions stable?
80-100°C mold surface. This holds the part in the 180-210°C window long enough to reach 45%+ crystallinity, cutting post-shrinkage below 0.2%.
Can PA66+GF30 warpage be fixed without changing the mold?
Partly. Staged packing plus 80-100°C annealing for 2-4 hours speeds stress relaxation. Warpage above 1 mm usually still needs gate or cooling changes.
How accurate is Moldflow at predicting glass-filled nylon warpage?
With CRIMS plus measured PVT and transient cooling, deviation from measurement can fall within 3%. With default generic data, under-prediction of 4x is common.