Extrusion Line Acceptance from China: The Days Before Shipping Are the Ones That Count
In 2024 I watched a China-based supplier commission a 90mm single-screw line that had passed its factory acceptance test — and 12 days after installation, the pipe coming off it had a variation in wall thickness that could not be dialled out. The factory acceptance test had run for four hours on pellets the buyer never specified, at a haul-off speed the factory liked, on the factory's own mould. Acceptance was signed on a set of data that did not match the production contract.
A FAT for an extrusion line is not a formality. The section below is the practical list of checks I use.
Match the test to the production conditions
The FAT must run under the conditions in the contract. Not an idealised version of them. That means: the resin specified in the purchase order, the die from the contract, the extruder screw from the contract, and the line running at the contracted output. If the contract specifies 250 kg/h at 0.8mm wall thickness with a particular compound, then the FAT should hit both targets simultaneously and hold them for a minimum of 8 hours.
Any longer is better. Four hours proves the machine runs; eight hours is where thermal drift shows up. Twelve hours is where the operator's attention starts to slip, and that is when quality falls. If you are going to invest in a FAT, pay for two shifts of it.
Measure the line, not just the machine
On-line measurement is the first check. Wall thickness scanner, diameter gauge, and temperature control across all zones. Every reading you accept in the FAT will become the baseline for the commissioning crew, and they will fix the machine to that baseline. If the FAT scanner is the supplier's own and is uncertain, bring your own gauge for a second reading.
On-line gauging for pipe typically uses ultrasonic wall scanners. Confirm the scanner is calibrated with a sample of known wall thickness and that a printout is produced, not just a verbal confirmation.
Hold the stability test
An hour of stable production is not stability. Run long enough to see the barrel temperature settle, the melt pressure settle, the haul-off align with the puller, and the line's throughput stay within ±5%. If the line requires constant hands-on adjustment to stay within tolerance, it is not stable, and every shift on your site will need that operator.
Document what the operator does during the test. If you see the operator tweaking the die or the haul-off every 20 minutes, ask why. There is usually a reason, and it usually means an hour of run time will not be enough to find it.
Check the periphery
- Cooling tank: temperature, water flow, and the vacuum tank seal. A leak that only appears under full load appears at your site.
- Haul-off belts: the clamp force and the synchronicity. Wrongly synced belts cause ovality and marks on the surface.
- Winder or cutter: check the cut quality at full speed and the coil tension.
- Control cabinet: check the PLC program is the same version you will receive, that backup is on delivery, and that the recipe set is included.
- Spare parts: confirm the spares list matches the contract and is physically on the truck before it leaves the factory.
The samples you take home
Take samples of every product the line will run, in each size. A pipe sample with a label showing date, line settings, operator, and the exact wall and diameter reading is worth more than a promise. Also take samples of the raw material the line was fed. When something goes wrong at your site in month three, the samples are the comparison point.
Do not sign acceptance at the factory
Signing acceptance at the factory transfers risk of transport damage back to the buyer. The acceptance certificate should record what was tested, the materials and speeds used, the measured results, and the defects found. Sign it as "acceptance subject to satisfactory commissioning at site", and hold a portion of the balance against that commissioning.
Commissioning at your site should include at least 36 hours of continuous running at the contracted conditions. In my experience, this is where the real surprises appear: a line that ran at 250kg/h in the factory runs at 220 at your site, and you discover the factory's ambient temperature was helping. Or the water supply at your factory is warmer, and the cooling tank no longer holds the wall thickness.
None of this is unreasonable to ask for. Most suppliers are comfortable with it, and the ones who are not are telling you something useful.