There’s a particular kind of call that comes in every few months: intermittent short shots, processing that won’t hold steady, sporadic micro-voids or void clusters that show up and disappear with no obvious pattern. The caller wants a setting — the one magic adjustment that makes it go away. The honest answer is usually a question they don’t want to hear: when did you last pull the screw, check it for wear, replace the non-return tip, or even clean it?
The reliable response to that question is a thousand-yard stare. Pulling the screw means downtime, and downtime on a busy press feels like the enemy. So the injection unit — the one part of the machine that physically meters and delivers every shot — becomes the last place anyone looks, long after it should have been the first.
This article doesn’t reproduce the historical WJT Associates maintenance material. It uses the same applied reasoning — diagnose the injection unit before you blame the process — to explain how screw, barrel, and non-return valve wear actually presents on the floor, and how to decide when it’s time to pull the screw.
A note on terminology: the non-return valve goes by many names and designs — ball checks, sliding rings, wipe valves. To keep this readable, “NRV” below covers all of them. What matters isn’t the mechanism; it’s whether it seals.
Why the injection unit gets ignored
A worn NRV doesn’t announce itself. It fails intermittently, and intermittent problems are the ones operators learn to live with. The part is short this shot, fine the next five, short again. The cushion reads stable on average but wanders shot to shot. Somebody bumps shot size to cover the short ones, the average looks fine again, and the real problem goes underground — getting worse the whole time.
By the time the symptoms are impossible to ignore, the plant has usually spent days chasing the process: melt temperature, fill speed, pack, drying, even the resin lot. All of it is wasted motion if the valve isn’t sealing, because none of those settings can fix a check ring that lets melt slip backward during pack.
Reading the symptoms
Injection-unit wear has a signature. Once you know it, it’s hard to mistake for a true process problem.
| Symptom | What it points to | What it is not |
|---|---|---|
| Cushion varies shot to shot, no setting changed | NRV leaking back during pack — the seal isn’t repeatable | A pack-time problem; pack can’t fix a valve that won’t hold |
| Intermittent short shots on a stable setup | Melt slipping past the check ring before it seats | A fill-speed problem |
| Sporadic micro-voids or void clusters | Inconsistent pack pressure reaching the cavity as the valve leaks | Always a drying/material problem (though rule that out first) |
| Slowing or inconsistent recovery/plasticizing | Worn flights, worn barrel, or both — clearance is gone | A back-pressure setting problem |
| Black specks or degradation after a clean resin lot | Material hanging up in scored barrel or behind a worn valve | A contaminated-lot problem |
| Process that “used to be easy” now needs babysitting | Gradual clearance loss between screw and barrel | Operator skill |
The tell that separates wear from process is inconsistency on an unchanged setup. A process problem is repeatable — same setting, same result. Mechanical wear gives you a different result from the same inputs, because the hardware isn’t doing the same thing each shot.
The cushion check anyone can run
You don’t have to pull the screw to get strong evidence the NRV is the culprit. The check ring’s whole job is to hold a seal during pack and hold; a simple test exposes whether it does.
- Run the process to a normal cushion and let it stabilize.
- Watch cushion over 20–30 consecutive shots and record it — don’t trust the averaged display.
- A healthy valve holds cushion within a tight, repeatable band. A worn one wanders, with occasional shots that lose cushion as melt slips back.
- For a more direct test, after the screw recovers and the shot is metered, hold position without injecting and watch whether the screw creeps forward — a valve that won’t hold static pressure won’t hold dynamic pack pressure either.
Twenty minutes of watching real cushion data tells you more than another day of changing settings. If cushion won’t stay put on an unchanged process, stop adjusting the process.
When to pull the screw
Pulling the screw is downtime, and the decision is a judgment call — but it’s a judgment that should be made on evidence, not avoided on principle. Pull it when:
- Cushion wanders on an unchanged setup and the check-ring test shows the valve won’t hold.
- Recovery time has crept up and back pressure no longer behaves the way it used to — a sign of lost screw-to-barrel clearance.
- You’re getting degradation, black specks, or color hang-up that survives a proper purge, suggesting material is parking in scored barrel or behind worn valve components.
- The same press has quietly become “the difficult machine” that needs an experienced hand to keep running — wear makes a process narrow and twitchy.
When the screw comes out, inspect and measure rather than eyeball. The point is to know which component is gone, not just to clean it and put it back.
| Component | What to check | Common verdict |
|---|---|---|
| Non-return valve | Seat condition, ring/ball wear, seating face, clearance | Replace the tip assembly — it’s cheap relative to the scrap it causes |
| Screw flights | Wear on flight lands, scoring, lost hard facing | Re-flight or replace if clearance is out of spec |
| Barrel | Bore wear, scoring, glazing | Measure clearance; a worn barrel defeats a new screw |
| Check-ring seat | Wear and impact damage where the ring seats | Often the actual leak path even when the ring looks fine |
A frequent mistake is replacing only the NRV when the barrel is also worn. A new valve in a worn barrel still leaks, because the clearance the valve depends on is gone. Measure the barrel before you celebrate a new tip.
Build it into preventive maintenance
The reason injection-unit wear causes so much firefighting is that almost nobody schedules for it. Molds get preventive maintenance schedules; screws rarely do. A basic standard fixes most of it:
- Track recovery time, cushion stability, and back-pressure behavior as running indicators — a slow drift in any of them is the early warning.
- Set a clean-and-inspect interval for the NRV based on material abrasiveness and hours, not on “when it finally fails.” Glass-filled and flame-retardant resins eat valves and barrels far faster than unfilled commodity grades.
- Keep a spare NRV tip assembly for each machine class on the shelf. The cost of the part is trivial next to the cost of the scrap and downtime it prevents.
For deeper, vendor-neutral coverage of screw and barrel wear mechanisms and measurement, the technical archives at Plastics Technology (ptonline.com) are a solid industry reference.
FAQs
How do I know if my problem is the non-return valve or the process?
Look for inconsistency on an unchanged setup. A process problem is repeatable — same setting, same result every shot. A worn valve gives you different results from identical inputs: cushion that wanders, short shots that come and go, sporadic voids. If you haven’t changed anything and the parts won’t behave, watch cushion over 20–30 shots before you touch another setting.
Can I keep running with a worn non-return valve?
You can, and many shops do — by bumping shot size to cover the shorts. The problem is that it masks a worsening condition and gives you a process that swings between short and flashed once the extra shot size can’t compensate. It also quietly raises your scrap rate the whole time. Running on a worn valve is borrowing against downtime you’ll pay back later, usually during a rush order.
How often should the non-return valve be inspected?
It depends far more on material than on a calendar. Abrasive resins — glass-filled, mineral-filled, flame-retardant grades — can wear a valve in a fraction of the hours an unfilled commodity resin would. Set the interval by material abrasiveness and running hours, track cushion and recovery as early-warning indicators, and inspect before failure rather than after.
Does a new valve fix a worn barrel?
No. The non-return valve seals against clearance between the screw, ring, and barrel. If the barrel bore is worn, a new valve still leaks because the surface it seals against is gone. Always measure barrel and screw clearance when the screw is out — replacing only the valve in a worn barrel is a repair that doesn’t repair anything.