PFMEA 不是填表交差。拿注塑成型这道典型工序，把常见的四种失效模式拆开看——后果是什么、根因在哪、现行控制怎么写——你就知道分析该往哪使劲。工序背景：汽车内饰件，材料 PP+TD20，一模两穴，壁厚 2.5mm，分析聚焦"熔体从射入型腔到冷却定型"全过程的潜在失效。
▌ 失效模式一：缺胶（短射）
后果：产品填充不满、尺寸缺肉，装配间隙超差，严重时整件报废。
原因：注射压力不足、熔体温度偏低、模具排气不良、流道偏小或冷料阻塞。根因往往不是单一参数，而是温度和压力的叠加效应——料温低 10 度，流动距离缩短近三成，单调一个参数没用。
现行控制：首件全检外观、过程巡检每 2 小时抽 5 件、模温机在线监控。
▌ 失效模式二：飞边（披锋）
后果：产品边缘多出薄片毛刺，外观判废；后续人工修剪增加工时和划伤风险，严重时合模面损伤模具。
原因：锁模力不足、注射压力过高、模具分型面磨损或贴合不严、塑料黏度偏低。多型腔模具中某一穴反复出飞边，多半是该穴分型面已经被压塌——先查分型面，再调参数。
现行控制：合模力每班开机确认、飞边目视检查 100%、模具保养每 500 模次检修分型面。
▌ 失效模式三：缩水印
后果：表面出现凹陷或银纹，外观件直接判废；结构件局部壁厚不足，强度下降。
原因：保压压力不够、保压时间太短、壁厚差异大、料温过高导致收缩率增大。缩水印常出现在壁厚突变处——加强筋根部、柱位顶部，这些位置最后冷却，补偿跟不上。
现行控制：保压参数纳入工艺卡、关键尺寸 SPC 监控、壁厚用超声测厚仪抽检。
▌ 失效模式四：翘曲变形
后果：平面度超差，装配后产生应力，长期使用异响或开裂；尺寸超差导致返工或报废。
原因：冷却不均匀、分子取向差异、脱模受力不均、环境温度骤变。翘曲是注塑件最复杂的失效——应力在冷却过程中冻结、脱模后释放，有时 24 小时后才变形到位。
现行控制：冷却时间固化在工艺参数中、出炉后 2 小时尺寸检测、关键件做时效后测量。
▌ 实操要点：原因往工艺上挖，别甩给人
- 失效原因别写成"操作工失误"——往工艺参数和设备条件上挖，才有可执行的纠正措施；
- 每个失效模式至少对应一条现行控制，写清"怎么测、多久测一次、超差怎么办"；
- PFMEA 的价值不在表格本身，在于把车间经验沉淀下来，让新员工少踩坑、同类问题不复发。
记一句：注塑 PFMEA 写得好不好，看失效原因是停在"人"，还是挖到了"参数"。
想把注塑车间的失效经验沉淀成可复用的失效库？续创的 FMEA 失效分析与失效库系统正好干这件事。欢迎联系续创信息技术咨询。
A PFMEA is not a form-filling exercise. Take the classic injection moulding process and break open its four common failure modes — what are the effects, where are the root causes, how should current controls be written — and you will know where to aim your analysis. The process context: an automotive interior trim part, material PP+TD20, one mould with two cavities, 2.5 mm wall thickness, and the analysis focuses on potential failures across the whole journey of the melt from injection into the cavity to cooling and solidification.
▌ Failure Mode 1: Short Shot
Effect: the part is under-filled and dimensionally short, assembly gaps fall out of spec, and severe cases are scrapped outright.
Cause: insufficient injection pressure, low melt temperature, poor mould venting, undersized runners or cold-slug blockage. The root cause is rarely a single parameter — it is the combined effect of temperature and pressure: a melt 10 degrees cooler loses nearly 30% of its flow length, so tuning one parameter alone gets you nowhere.
Current controls: full visual check on first articles, patrol inspection of 5 pieces every 2 hours, online monitoring by the mould temperature controller.
▌ Failure Mode 2: Flash
Effect: thin burrs along the part edge fail the appearance check; manual trimming adds labour and scratch risk, and severe cases damage the mould at the parting line.
Cause: insufficient clamping force, excessive injection pressure, a worn or poorly seated parting surface, or low melt viscosity. In multi-cavity moulds, when one cavity flashes repeatedly, its parting surface has most likely collapsed — check the parting surface first, then touch the parameters.
Current controls: clamping force verified at every shift start-up, 100% visual inspection for flash, parting-surface maintenance every 500 shots.
▌ Failure Mode 3: Sink Marks
Effect: surface dents or silver streaks — appearance parts are scrapped on the spot; structural parts suffer locally thin walls and reduced strength.
Cause: insufficient holding pressure, holding time too short, large wall-thickness variation, or excessive melt temperature raising the shrinkage rate. Sink marks typically appear at abrupt wall-thickness transitions — rib roots, the tops of bosses — which cool last, when compensation can no longer keep up.
Current controls: holding parameters written into the process card, SPC monitoring on critical dimensions, wall thickness sampled by ultrasonic gauge.
▌ Failure Mode 4: Warpage
Effect: flatness out of spec, stress locked in after assembly, long-term rattling or cracking; dimensional nonconformance leads to rework or scrap.
Cause: uneven cooling, differences in molecular orientation, unbalanced ejection forces, sudden ambient temperature swings. Warpage is the most complex injection-moulding failure — stress freezes during cooling and releases after ejection, sometimes not reaching its final shape until 24 hours later.
Current controls: cooling time locked into the process parameters, dimensional checks 2 hours after moulding, post-ageing measurement for critical parts.
▌ Practical Points: Dig Causes into the Process, Don't Blame the Operator
- Never write "operator error" as a failure cause — dig into process parameters and equipment conditions, where corrective actions can actually be executed;
- Every failure mode needs at least one current control, stating "how it is measured, how often, and what happens when out of spec";
- The value of a PFMEA is not the form itself, but capturing shop-floor experience so new hires avoid the pitfalls and the same problem never recurs.
One line to remember: the quality of an injection-moulding PFMEA shows in whether the causes stop at "people" or dig down to "parameters".
Want to turn your injection-moulding failure experience into a reusable failure library? CINO's FMEA failure analysis and failure library system does exactly that. Contact CINO Information Technology Consulting.