A cluster comes off the press with flow lines running down the third cavity. The operator checks the nozzle temperature, then the injection pressure, then the hold time. Everything sits inside the card. He shoots again and gets the same lines.
The wax is the problem, and it has been for two weeks. Nobody caught it because reclaimed wax still looks like wax. Same colour, same hardness at room temperature, same snap when you break a sprue over your knee. What changed is not visible in the bucket. It is in the water content, the fines, and how many times that batch has been through the tank.
Recovery is not one operation. It is four, and each one removes a different thing.
What recovery actually has to remove
Pattern wax leaving the press is not dirty in the way sand is dirty. Three things accumulate in it:
- Water. Enters with the charge, with condensate on cold returns, and through humid storage. It does not sit on top as a layer. It is carried in the melt and comes back out as foam, spatter and short fills.
- Fines and ash. Gate stubs carry shell dust, investment residue and floor grit. These do not melt. They stay suspended and end up in the pattern surface.
- Heat history. Every pass through the tank costs the wax something. This one cannot be removed at all, only prevented.
The four stages, with the numbers on them
| # | Wax tank | Temperature | Time | What it removes |
|---|---|---|---|---|
| 1 | Settling tank I | 85–90 °C | 6–8 h | Coarse sludge, grit, shell dust |
| 2 | Dewatering tank | 110–120 °C, stirred | 10–12 h | Water |
| 3 | Settling tank II | 80–85 °C | >12 h, filtered through <60 mesh | Fine ash, remaining solids |
| 4 | Holding oven | 54 ± 2 °C | >24 h | Nothing. It tempers the melt to injection temperature. |
Stage 2 is the one that gets skipped, and it is the only one where the endpoint is a judgement rather than a clock. The instruction is to stir until the surface stops foaming. Foam at 110–120 °C is water leaving the melt, so “no more foam” is the actual completion signal — not the hour count. If you pull the batch at 10 hours because the shift ends and it is still foaming, you have moved water into stage 3, where 80–85 °C will not drive it off.
The temperature window is narrow on purpose. Below boiling, water stays in. Much above it, the wax starts to age. There is no safe “hotter is faster” setting here.
Two settings that never make it onto the card
The written schedule gives temperatures and hours. It also gives two maintenance items that are routinely dropped because they are not product parameters:
- Drain water and sludge from every tank on a fixed interval. Not when it looks bad. The sludge layer at the bottom of a settling tank does not stay put once you start transferring; a full tank drawn down to half will pull the bottom layer forward into the next vessel. If stage 1’s sludge is not drained, stage 2 spends its twelve hours dewatering a suspension instead of a melt.
- Keep the heat transfer oil level about 200 mm below the top of the vessel. An under-filled jacket gives uneven wall temperature, and the wax at the top runs hotter than the wax at the bottom. You will read the setpoint and believe it.
New wax is an input, not a repair
The rule is 3–5% fresh wax per batch, adjusted to how the reclaimed material is performing. It is a top-up, and it is the standard excuse for not fixing the process. If a shop is running 8% or 10% fresh wax to hold quality, it is not running a recovery system, it is buying new wax slowly and paying for four tanks and the labour to do it.
Watch that number. It is the cheapest early warning you have. When the required percentage starts creeping up month over month, something upstream changed and the tanks are where you notice it.
Overheating is the one failure you cannot recover from
Aging is named in the work rules as the consequence of running the wax too hot, and it is different in kind from water or ash. Water comes out in stage 2. Ash comes out on the 60-mesh screen. Aged wax stays aged, and adding fresh material dilutes the symptom without restoring the batch.
The practical signature is that the press stops behaving like itself. Settings that produced a clean pattern three weeks ago now produce flow lines and incomplete fills, and adjusting the press does not fix it. If you have already chased nozzle temperature up and hold time out and the defects are still there, stop tuning the machine and pull a wax sample.
Containers: no carbon steel
The rule in the shop documentation is blunt — containers that contact wax must not be carbon steel, on the grounds of FeCl₃-type compound formation. It matters most in acid treatment of low-temperature wax, where the batch is deliberately boiled with 25–35% water plus 3–5% industrial hydrochloric acid (or 2–3% concentrated sulphuric acid) and held until the white soap disappears. Acid plus carbon steel in the same vessel is a bad combination for reasons that have nothing to do with wax quality.
For medium-temperature wax the mechanism is milder, but the rule is worth keeping as a blanket one. Iron that gets into the wax does not stay in the wax. It goes into the pattern, into the shell, and comes back out after burnout as the kind of surface contamination that gets logged against the wrong department. If you have ever chased scattered pitting or black speckling on castings and found nothing in the melt, check what the wax has been sitting in.
Where bad wax shows up downstream
None of this fails at the press. It fails three weeks later, in the shell and in the casting:
- Water and volatiles carried into the shell leave carbon residue and soot after burnout, which reads as gas porosity and scattered pitting on the casting. The burnout schedule gets blamed. See carbon residue and soot after burnout for how that failure presents.
- Ash and fines raise the residue in the shell and contribute to inclusions and surface speckling.
- Flow lines and cold shuts in the pattern itself carry straight through to the casting surface.
The diagnostic habit worth building: when a shell or casting defect appears in a scatter pattern with no positional logic, ask what changed in the wax before you ask what changed in the slurry.
A different schedule, for reference
The defect atlas appendix gives a shorter variant for medium-temperature wax: static dewatering at 100–110 °C, decontamination at 70–85 °C for 1–2 hours, then holding at 48–52 °C for 8–24 hours. Different wax grades and different tank hardware take different numbers. It is a reference, not a substitute — the plant process card governs, and mixing the two half-way produces a schedule that satisfies neither.
One technique worth knowing about
Japanese practice literature describes a quench-and-temper variant that runs against the usual “melt gently at 60 °C and hold” approach. Instead of slow melting and holding, the wax is melted, then quenched to room temperature or below, then blended at the specified temperature and time. The reported result is a noticeably denser, smoother pattern surface, with the mechanism compared to rapid cooling refining grain structure.
We have not run this ourselves and are not presenting it as plant data. It is included because it explains something otherwise puzzling: two shops with identical tank schedules and identical press settings can produce patterns with clearly different surface quality. Wax thermal history before the press matters, and it is invisible on the process card.
What to check on Monday morning
- Is the dewatering tank actually reaching 110–120 °C, or is it reaching the setpoint on a probe that sits against the jacket wall?
- When was sludge last drained from each tank? Is there a record, or is it “when it needs it”?
- What percentage of fresh wax went in last month, and what was it six months ago?
- Does anyone look at the 60-mesh screen when it is changed, or does it just get changed?
- Has any wax-contact container been replaced with carbon steel because it was cheaper and the old one cracked?
Four of those five cost nothing to answer. The fifth is the one that explains a defect nobody can find.