Regrind is the ghost in the attic of injection molding. Designers write “virgin material only” out of a vague fear, and molders quietly run regrind on jobs where it’s perfectly fine — and occasionally on jobs where it isn’t. Both reflexes skip the actual engineering question, which has a real answer: what does running this material a second (or third) time actually do to it, and does that matter for this part?
The honest position is that regrind is neither the villain the spec sheet implies nor the free money the cost accountant hopes for. A polymer that you don’t burn, hydrolyze, or color-shift can make acceptable parts after several heat histories. One that you mishandle on a single pass can be scrap before it ever fills the cavity. The difference is process discipline, not the word “regrind.”
This article doesn’t reproduce historical WJT Associates material. It applies the same practical reasoning — understand the mechanism, then manage it — to one of the most superstition-driven topics on the floor.
What actually degrades regrind
A second heat history doesn’t ruin most thermoplastics by itself. What ruins them is one of three specific, avoidable mechanisms. If you keep all three under control, regrind behaves; if you ignore them, even virgin material suffers.
| Degradation mode | What happens | Materials most at risk | How to control it |
|---|---|---|---|
| Thermal degradation (burning) | Excess heat and residence time break polymer chains; black specks, brittleness, splay | Any resin held too hot too long; worse for heat-sensitive grades | Correct barrel temps and residence time; don’t run regrind in an oversized barrel at low throughput |
| Hydrolysis (moisture) | Water plus heat chemically cleaves chains — irreversible loss of properties | Hygroscopic resins: PET, PC, nylon, PBT, PLA | Dry regrind to spec every pass; never assume reground material is dry |
| Color/contamination shift | Yellowing, off-color, foreign material from floor scrap and mixed grinds | Light colors, optical and cosmetic parts | Keep grinds segregated and clean; tighten ratio on cosmetic jobs |
The takeaway is that “can I use regrind here?” is really three questions: Will it get too hot? Is it dry? Is it clean and the right color? Answer those, and the heat-history fear mostly evaporates. Properties do drift down gradually with each pass — molecular weight steps down, impact and elongation soften first — but for most non-critical parts that drift is well within the working margin for several cycles.
The honest economics
Molders have a complicated relationship with regrind, and it’s worth being clear-eyed about why.
On the upside, regrind is material you’ve already been paid for. Run a job properly and you look for parts that can consume your sprues and runners — you charge the first customer for the regrind you couldn’t reuse on their job, then run it into a second job as if you’d bought fresh resin for it. That’s not a trick; it’s recovering value from material that would otherwise be waste.
On the downside, regrind that has nowhere to go becomes a liability. If a tool’s sprue-and-runner weight exceeds the regrind allowance on the part, you accumulate ground material faster than that job can consume it. If you can’t mold it into something else, it piles up — and disposal can cost more than the plastic is worth, because shipping scrap to a recycler sometimes exceeds the value of the scrap. The material doesn’t vanish; it sits on a skid as a problem.
The point is that regrind is a flow to be balanced, not a free input. The shops that profit from it design jobs so the generation rate and the consumption rate match.
Controlling the ratio
The single most useful control is a documented, enforced regrind percentage — and the discipline to actually meter it rather than eyeball it.
- Set the ratio by part criticality, not by convenience. Structural, optical, regulated, and high-fatigue parts warrant low or zero regrind. Internal, non-cosmetic, non-critical parts tolerate more.
- Meter it; don’t dump it. “A scoop of regrind in the hopper” is not a ratio. Blend to a target percentage with a blender or a measured charge so the property drift is bounded and known.
- Treat regrind as a process input. If you change the ratio, you’ve changed the material — which means a documented process and, for anything sensitive, a check that the window still holds.
- Cap the cumulative passes. Track how many heat histories the material has seen. Properties step down each pass; set a limit appropriate to the resin and the part rather than recycling indefinitely.
A regrind ratio that lives in a setup sheet and gets metered is an engineering control. A ratio that lives in an operator’s judgment is a variable you’re not measuring — and unmeasured variables are where the intermittent quality problems come from.
The lot traceability problem — and a clean solution
In regulated industries, the real objection to regrind often isn’t properties at all. It’s traceability: the dread of not being able to say exactly which raw-material lot a finished part came from. Mixing yesterday’s grind into today’s parts seems to break the chain.
There’s a straightforward way to keep the chain intact: use the regrind immediately, within the same lot it came from. Grind the runners from the job you’re currently running and feed them back into that same job, in that same production lot. The material never leaves its own lot, so traceability is preserved — you’re not blending an unknown history into a new lot, you’re closing the loop inside one. This handles the compliance concern without abandoning the recovery, as long as the resin tolerates the heat history and you’ve controlled drying and cleanliness.
For parts where even that isn’t acceptable, the answer is virgin-only — and that’s a legitimate engineering decision, not a default to reach for out of habit.
The granulator decides regrind quality before you ever blend it
Everything above assumes the reground material is clean, uniform, and uncontaminated — and that assumption lives or dies at the granulator. A neglected grinder quietly poisons the regrind stream no matter how disciplined your ratio control is downstream.
- Keep the cutting chamber healthy. Worn blades produce inconsistent particle size and excessive fines that feed and melt unevenly; replace them on a schedule, not after the grind quality has visibly dropped.
- Don’t overfill it. Feeding a granulator faster than it can cut overloads the chamber, generates heat, and degrades the very material you’re trying to recover.
- Mind the mounts and guards. Check anti-vibration mounts and the chamber curtains/seals — they keep the unit running smoothly and keep stray flakes from escaping into the surrounding area.
- Treat the grinder area as a contamination zone. Stray pellets and flakes from a previous job are exactly how the wrong color or grade gets into the next one. Keep the area clean and segregated, the same way you segregate the grinds themselves.
- Track scrap by cause. A simple scrap-analysis sheet — logging why parts were scrapped — makes both the regrind volume and any quality issues visible, and ties scrap rate directly back to part cost through fewer good parts per shot.
- Respect the safety basics. Never reach into a running granulator for any reason.
The point is that regrind quality isn’t only a blending decision at the press. It’s set upstream by how well the grinder is maintained and how clean the area around it is kept.
A practical regrind checklist
Before you run regrind on a job, walk this list. It takes a few minutes and prevents the kind of intermittent scrap that costs hours to chase later:
- Is the resin hygroscopic? If so, the regrind gets dried to spec every pass — no exceptions, no “it was dry yesterday.”
- Is this part criticality-sensitive? Structural, optical, or regulated parts get a low or zero ratio.
- Is the grind clean and segregated? No floor sweepings, no mixed grades, no color contamination.
- Is the ratio metered to a documented target, and is it on the setup sheet?
- How many passes has this material seen? Respect the cumulative-history cap.
- For regulated work, is the regrind staying inside its own lot to preserve traceability?
- Does the generation rate match the consumption rate, or are you building a skid of material you can’t use?
For deeper, vendor-neutral coverage of resin drying, recycling, and the science of polymer degradation, the materials archives at Plastics Technology (ptonline.com) are a strong industry reference.
FAQs
Does regrind always lower part quality?
No. A thermoplastic you don’t overheat, expose to moisture, or contaminate can make acceptable parts through several heat histories. Properties do drift down gradually with each pass — impact strength and elongation soften before anything else — but for most non-critical parts that drift stays well within the working margin. Quality problems usually come from the three avoidable mechanisms (burning, hydrolysis, contamination), not from the simple fact of being reground.
How much regrind can I safely use?
It depends entirely on the part and the resin, which is why the ratio should be set by criticality rather than convenience. Structural, optical, regulated, or high-fatigue parts often warrant low single digits or zero. Robust internal parts can tolerate more. The non-negotiables are that the ratio is metered to a documented target rather than eyeballed, and that you cap the cumulative number of heat histories the material accumulates.
Why is drying regrind so important?
Because hygroscopic resins — PET, PC, nylon, PBT, and others — degrade by hydrolysis when moisture and heat combine in the barrel, and that chain scission is irreversible. Regrind has often been sitting exposed to shop humidity, so it can be wetter than fresh resin, not drier. Assuming reground material is dry “because it was dry yesterday” is one of the most common ways a job that ran fine starts producing splay, brittleness, and inconsistent parts.
How do I use regrind in a traceability-controlled (medical or regulated) job?
Keep the regrind inside its own lot: grind the runners from the job you’re currently running and feed them straight back into that same production lot. Because the material never leaves the lot it originated in, the traceability chain stays intact — you’re closing the loop within one lot rather than blending an unknown history into a new one. Where even that isn’t acceptable to the customer or the regulation, virgin-only is the correct call.