Warpage in glass-filled PA66 is a fibre-orientation and cooling-balance problem, not a press-setting problem. On this bracket, re-gating to two balanced drops, adding conformal core cooling and correcting the steel for measured shrinkage cut twist from 1.2 mm to 0.35 mm over a 400 mm span.
Key takeaways
- Glass fibre makes warpage worse, not better. Fibres shrink about 0.3% along the flow direction and up to 1.2% across it, so orientation sets the twist.
- Hold pressure and cooling time do not repair a gating or cooling-layout problem. They only move the distortion somewhere else.
- Measure warpage in a gauged, datum-constrained state. A free-state CMM reading overstates twist and drives the wrong steel correction.
- Core-to-cavity temperature difference is the biggest lever. Holding it under 5 °C removed roughly two thirds of the twist.
- A post-mold cooling fixture is cheap insurance: 90 seconds in a gauge while the part is warm locks in the geometry.
- Two rounds of steel correction, driven by measured shrinkage, brought hole-to-hole position inside ±0.15 mm across a 400 mm span.
The project at a glance
A European industrial equipment manufacturer needed a structural mounting bracket for a machine assembly: an L-section channel roughly 400 mm long, with two mounting bosses at one end and a hinge lug at the other. It carries load, sits flat against a machined face, and is molded in 30% glass-fibre reinforced PA66 — a resin chosen for stiffness and heat resistance, and one of the harder materials to hold straight.
Part: structural mounting bracket, L-section, 400 mm span
Material: PA66 with 30% glass-fibre reinforcement
Nominal wall: 2.5 mm, with a 4 mm boss at the mounting end
Critical requirement: hole-to-hole position within ±0.15 mm, flatness within 0.3 mm in the gauged state
Tooling: one 1-out-of-2 production mold, hot runner, hardened 1.2343 (H11) inserts
Delivered from: RCH Plastic, Dongguan, China
The first tool, built by a previous supplier, produced parts twisted by up to 1.2 mm across the span. Operators had to force the bracket into position before starting a bolt — a two-person job that cracked the mounting boss on roughly one part in six.
The challenge: a 400 mm span that would not stay flat
Warpage is never a single cause, and here three mechanisms ran at once. Separating them was the difference between a real fix and more guesswork on the press.
Fibre orientation
Short glass fibres align with the flow direction as the melt fills the cavity. Because the fibres barely shrink while the polymer between them shrinks a great deal, the part contracts roughly 0.3% along the fibre direction and up to 1.2% across it. A single gate at one end of a 400 mm channel creates a long, highly oriented flow path along the span and a much less oriented region around the boss. Those regions want to shrink differently, and the part bows.
Cooling imbalance
The bracket is an L-section, so the core side of the tool carries most of the heat while the cavity side is easy to cool. When the core runs hotter, one face stays molten longer and shrinks more after ejection — the classic mechanism behind a part that bows toward the hot side. On the original tool the measured core-to-cavity difference was 14 °C.
Measurement error
The original inspection measured the part free-standing on a surface plate. That number is real, but it is not what the assembly cares about: once bolted to a machined face the bracket is partly constrained, and the error at the bolt holes is smaller than the free-state twist. Chasing a free-state number would have over-corrected the steel.
Diagnosis: orientation warp versus cooling warp
Before changing anything we ran a mold flow analysis on the original design, then a physical experiment on the existing tool. The simulation predicted flow-front pattern and fibre orientation; the experiment isolated cooling as a variable.
The experiment is worth repeating on any warping part: shoot a short series at production settings and measure each part twice — immediately after ejection, then after 24 hours of conditioning. Here, twist measured straight after ejection was 0.8 mm and grew to 1.2 mm over 24 hours, so most of the distortion was already locked in by the time the part left the tool. That points at gating and cooling, not post-mold shrinkage.
Sectioning a part showed the mounting boss was 4 mm against a 2.5 mm wall — 160% of nominal, well above the 50–60% rule for ribs and bosses. That thick mass cools last and pulls the surrounding surface with it.
The fix: gating, cooling layout and steel correction
Three changes, in this order of impact.
1. Re-gate to two balanced drops
The single end gate became two gates feeding the part symmetrically, shortening the maximum flow length and balancing the pressure gradient across the span. Balanced fill means balanced packing, and balanced packing means both halves shrink by the same amount. The weld line the second gate creates was placed in a low-stress region away from the hinge lug, verified in simulation before any steel was cut.
2. Balance the cooling
We rebuilt the core side with baffled and bubbler circuits following the part contour, and split core and cavity onto separate temperature control circuits. The core ran slightly cooler than the cavity to offset the natural imbalance of an L-section. Core-to-cavity difference dropped from 14 °C to under 5 °C, with mould temperature held at 90 °C — which PA66 needs for crystallisation and stable dimensions.
3. Correct the steel against measured shrinkage
With fill and cooling balanced, the remaining error was a consistent, repeatable bias — exactly what you correct in the tool. We measured a 30-piece sample across both cavities, calculated effective shrinkage in each principal direction, and machined the core to compensate. Two rounds took the bias out. This is only safe once the process is stable: correcting steel against a drifting process bakes the drift in.
We also cut the mounting boss from 4 mm to 2.8 mm with a cored pocket, and added a post-mold cooling fixture — a gauge that holds the part in its correct geometry for 90 seconds while it cools. The fixture costs a fraction of a tool modification and removes most of the remaining spring-back.
Trials and the measured results
Trials ran at our partner molding plant, since we build tools in-house and run trials and production through vetted partners. Each round was delivered as a documented package, not a verbal report.
Trial 1 verified fill balance across both cavities with short shots and confirmed the weld-line position against simulation. Trial 2 ran the production resin with a gate seal study to set hold time, and measured a 30-piece sample on a CMM using the part datum system in a gauged state. Trial 3 confirmed the process window at the extremes of the recommended settings.
Twist: 1.2 mm to 0.35 mm over the 400 mm span, measured in the gauged state
Hole-to-hole position: within ±0.15 mm across both cavities
Process capability: Cpk 1.42 on the two critical hole positions over a 500-piece run
Cycle time: 32 seconds at 90 °C mould temperature
Scrap: 18% on the original tool, 2.4% after correction
Every number above is backed by a document issued against the mold number: the DFM report, the CMM dimensional report on trial samples, the full process parameter sheet, and the trial photos and video.
Our answer: how we keep warpage out of a tool
RCH Plastic has been building injection molds in Dongguan since 2015, with more than 1,500 molds delivered to customers in North America and Europe, at 20–30 molds a month. Our answer to warpage starts before steel: we model your part in UG and run it through Moldflow, then return a written DFM report within 3 working days at no cost, so gating and cooling are settled on screen rather than on the press. Five mold designers averaging 8 years work alongside three project engineers and twelve fitters averaging 6 years — 49 staff in total, 37 in production and 12 in the office.
The machines that hold the geometry are 13 CNC machining centres with travels up to 1,600 mm, 6 EDM machines including Sodick mirror-finish and twin-head units, 3 wire EDMs, and CMM inspection on trial samples. Our machining tolerance is ±0.005 mm on critical mold dimensions; the molded part tolerance that tool can then hold is ±0.01 mm on critical dimensions, subject to resin and geometry — and glass-filled PA66 sits at the demanding end of that range.
Tools go up to 1,600 × 1,000 × 500 mm and 15 t, with up to 48 cavities, in steels including 1.2344 (H13), 1.2343 (H11), 1.2738HH and S136, on LKM, DME or HASCO bases, with hot runners from HRS, Yudo, Husky, Incoe or Mold-Masters. Lead time to T1 is 35 days on a simple tool, 42 medium, 50 complex, and 30 days when the drawing is finalised and components are in stock. Tool life is up to 1,000,000 shots depending on steel grade and maintenance, with a 12-month warranty. We make molds in-house and run trials and production through vetted partner factories — which keeps our engineering focused on the tool, where cost and lead time are actually decided.
What this means for your project
Nothing here is specific to brackets. Any part in a fibre-reinforced or semi-crystalline resin — glass-filled PA66, PBT, POM, talc-filled PP — warps for the same three reasons: orientation, cooling imbalance, and a process that was not stabilised before the steel was corrected. The sequence is always the same. Model the flow. Balance the cooling. Stabilise the process. Only then correct the steel, and measure in the state the assembly actually sees.
If you are fighting warpage on an existing tool, the fastest path is a fresh DFM and flow review against the current design. We do that before quoting a replacement tool, because a fair share of the warping parts we are sent do not need a new tool at all.
Related guides in this series
- Injection molding defects and quality guide — the full defect and quality library
- Common injection molding defects — root causes and the fixes that work
- Mold flow simulation guide — reading fill, pressure and warpage predictions
- Robot gripper components case study — holding tolerances on an automation part
Warping on a part you already have?
Send the 3D model, the resin grade and a photo of the distorted part, and tell us how the dimension is measured today. We will come back with the likely mechanism, the first three things to check, and whether the fix is a process change, a tool modification or a new tool. Ask our engineers for a free DFM and flow review.