A part came out wrong — what do you change first? Pick the defect, add where it sits on the part, the resin family and when it started, and this tool returns a ranked list of likely causes and concrete fixes. Each one is tagged process, tooling, material or design and labelled with who holds the dial, so you change the cheapest reversible thing the evidence points to — one variable at a time.
Engineering estimate and a starting-point checklist, not a root-cause diagnosis or a quote.
Every molding defect has a long list of possible causes, and most published troubleshooting tables print them flat — twenty causes in no particular order, so you end up changing three settings at once and learning nothing. This tool does the opposite. It ranks the causes for the defect you picked, keyed to two questions the part answers for free: where the defect sits and when it started. Then it tags each fix by who holds the dial, because the molder, the tool shop, the resin handler and the part designer own different corners of the problem and you want to call the right one.
The one rule. Change the cheapest reversible thing the evidence points to, and change one variable at a time. A press dial or a quick clean-and-check is cheap and instantly reversible; cutting a vent deeper or opening a gate is neither. So process and cleaning changes usually sort to the top. But physical evidence overrides the rule, and that is where most of the value is: a burn mark at the unfilled end means trapped air, so you vent before you touch pressure; a witness line on the parting surface points at the mold, not the molder; brown streaks that follow the flow are thermal degradation while silver streaks are friction or moisture. The tool bakes those overrides in, so when you tell it the streak is brown it moves "check melt temperature and residence time" up and pushes "dry the resin" down.
Why the tags matter. A fix tagged [P] process is a dial on the press you can try this shot and undo the next. [T] tooling is a change to the mold — clean it, or send it out to be cut. [M] material is in the resin and how it is dried, reground and colored. [D] design is in the part geometry and takes a drawing change. The order the tool returns respects reversibility: a [P] dial outranks a [T] metal cut outranks a [D] redraw, all else equal, because you try the thing you can undo first. The "who changes it" label — molder, tool shop, resin handling, part designer — tells you who to walk over to.
Resin suppliers do not fully agree, and pretending they do is how bad advice spreads. Three real disagreements are built into the rankings:
Splay, gas bubbles, internal voids, burns and black specks get confused constantly because the surface tells you little. The tests that separate them are worth more than any ranking:
When flash shows up, one of the ranked checks is clamp tonnage versus the projected area times the cavity pressure. Here is that check with real numbers so you can see whether the press is the problem. Clamp force needed equals the projected area of everything inside the parting line (parts plus runners) multiplied by the average cavity pressure:
So a 60-tonne press running this tool at 500 bar is under its own flash threshold and no amount of parting-line cleaning will fix it — you need a bigger press, a lower cavity pressure (slower, cooler, or an easier-flow grade), or less projected area per shot. But notice the order: you only reach this check after the free ones (clean the parting line, confirm the shot size, check the resin lot), because if the press held yesterday at the same tonnage, it did not get weaker overnight. Work the clamp math on the injection molding tonnage calculator; it uses the same A × p relationship with resin-specific pressure factors.
The ranking logic is checked against three real situations. Load the splay case with the button above.
| Situation | Top three the tool returns | What NOT to touch yet |
|---|---|---|
| Flash at the parting line, PP, started after weeks of a steady run | 1) Clean parting line, vents, shut-offs [T] 2) Confirm shot size / move pack earlier, check pack wasn't raised [P] 3) Check retained prior lot / melt-flow drift [M] | Don't raise clamp past the press rating or grind vents deeper before measuring. Tonnage comes later — the press didn't weaken overnight. |
| Splay, nylon 6/6, near the gate, silver parabolic streaks | 1) Measure resin moisture — dryer dew point, time, open-box exposure [M] 2) Reduce shear at the gate — land, size, sharp edges, speed [T/P] 3) Check melt temp & residence [P] | Don't just drop the barrel temperature before ruling out wet resin. Silver + parabolic + hygroscopic resin points at moisture first. |
| Short shot at end of fill with a burn mark, ABS, first shots after a cold start | 1) Clean & verify vents at the last-fill point [T] 2) Confirm shot size / stable cushion [P] 3) Raise melt then mold temp; let the tool reach equilibrium [P] | Don't raise injection speed/pressure — the default short-shot fix makes the dieseling burn worse when a burn mark is present. |
Some fixes are metal: a worn parting line, an over-deep vent, a gate in the wrong place, an undercut with too little draft. When you reach them, stop dialing and document, because a clear package gets a right-first-time correction and a vague one gets a guess. Hand over: shot photos including a short-shot series that shows how the cavity fills, the defect location marked on a part drawing, witness marks photographed, the resin and lot, and the setup sheet (temperatures, pressures, speeds, cushion, cycle). We coordinate between the molder and the tool shop so the change is cut once and verified against the same shots — JCH works a vetted manufacturing network across Taiwan, Vietnam and the US; we do not claim to own the press or the tool, and we will not quote a cut sight-unseen.
Sources. The causes, fixes and ordering draw on published resin-supplier and training troubleshooting guides — BASF (Injection Molding: Problems in Engineering Thermoplastics), Eastman (copolyester troubleshooting), Formosa Plastics (PE/PP process guideline) and RJG training material — dated between 2009 and 2025. Resin-specific numbers (vent depths, residence times, barrel setpoints) in those guides apply only to their own resins and are not published here as universal values; use the linked calculators and your resin supplier's data sheet for setpoints.
Reviewed by the JCH tooling team · updated November 2026
Flash is a thin fin forced into the parting line, a vent or a shut-off, from either the mold not sealing or the cavity pressure being too high. Start cheap: clean the parting line, vents and shut-off faces, because a resin film or dirt holds the halves open a few thousandths. Then confirm the shot size, move the pack or transfer earlier and slow the end of fill. If flash appeared after a resin, color or regrind change, suspect a lower-viscosity or degraded lot before touching the press. Check clamp tonnage against the projected area times the cavity pressure only after the free checks — if the press held yesterday it did not get weaker overnight. Inspect the parting line and shut-offs for wear and vents for being too deep last.
Splay has five look-alike causes: moisture, heat, shear, trapped air and a water leak. Appearance alone cannot tell you which, so test each. For hygroscopic resins measure moisture first — parabolic silver streaks pointing downstream are usually wet resin flashing to steam. Brown streaks following the flow are degradation, so check melt temperature and residence time. Silver right at the gate is high shear, so look at gate size, land, sharp edges and injection speed. Air splay is too much decompression or low back pressure. Last, rule out a cracked cooling line or a sweating feed throat. Change one cause at a time.
A short shot is a part that did not fill, usually at the far end. Causes: not enough melt metered or a leaking check ring, melt or mold temperature too low, injection pressure or speed too low, a plugged or undersized gate, runner or vent, or a flow path too long and thin. Check shot size and cushion first. If a burn mark sits at the unfilled end, trapped air is stalling the front — clean the vents before touching pressure, because raising speed with a burn present worsens the dieseling. Raise melt temperature before mold temperature, and let a cold tool reach equilibrium before judging early shots.
Black specks are carbonised polymer or foreign matter. The most common source is contaminated or dirty regrind, so shut off the regrind and run virgin to prove it. Next purge the barrel and hot runner and check the actual melt temperature, because previous material or an over-hot barrel chars. Reduce residence time by running a shot that is 25 to 75 percent of the barrel. Inspect the hot runner for dead spots and the check ring and screw for wear. Clean the dryer, hopper and lines. Specks that come from a carbonised deposit in a vent are a tooling cleaning job, not a material problem.
Change the cheapest reversible thing the evidence points to, one variable at a time. A press dial or a clean-and-check is cheaper and faster to undo than cutting metal, so those usually come first. But physical evidence overrides: a burn mark at the unfilled end sends you to the vents, a parting-line witness points at the mold, and brown versus silver streak colour tells degradation from moisture. If the defect started after a steady run, ask what changed rather than assuming the press weakened. Never raise clamp past the press rating or grind vents deeper before measuring.
They are the same unfed-shrinkage problem with two outcomes. A sink forms when the skin is still soft and gets pulled inward; a void forms when the skin has already frozen and the shrinkage tears an internal vacuum instead. To tell a void from a gas bubble, drill into the thick section under water — water drawn in means a vacuum void, no water means a trapped-gas bubble (look at venting). Both sinks and true voids are fixed by feeding more volume: more pack pressure and time, a bigger or better-placed gate, and thinner thick sections. The mold-temperature lever is contested and opposite for the two, so leave it low in the order.
Stuck between two causes, or not sure whether your streak is moisture or heat? Describe the part and what you see. Our AI assistant answers from this website, and our tooling team follows up if you want a second set of eyes on the shots.
AI assistant. Answers can be wrong; nothing it says is a quote or a promise.
This troubleshooter gives an engineering estimate and a starting-point checklist. When a defect survives the process dials, our tooling team reviews the short-shot series and witness marks with you, decides whether it is really process, tooling, material or design, and coordinates the correction across a vetted manufacturing network. No quote without seeing the shots.
Book a 30-min call Send photos and a drawing