Home / Valve Test Benches / FAT checklist
The factory acceptance test — the last look before the crate is closed
Once a test bench is packed, every problem it carries becomes your problem at the dock, in customs, and on your workshop floor while your production waits. A factory acceptance test (FAT) is not a ceremony; it is a structured few hours in which a defined list of things is checked and the results are written down. This page is that list — six groups of checks for a hydrostatic valve test bench, a suggested test agenda for the day, and the paperwork to settle before the machine ships.
Quick answer
| Who runs the test | You or your inspector, with the supplier's test engineer |
|---|---|
| What valve is tested | A representative valve from your production, at its real shell and seat pressures |
| What is recorded | Hold time against the timer, the time–pressure curve, the pass/alarm result |
| What ships with the machine | Tooling for the full DN range, spare seals, English manual, wiring diagram |
| If you cannot attend | Same agenda run live over video call |
Why acceptance before shipment, not after arrival
A test bench that fails at your factory does not just cost a repair. It costs the weeks between the claim and the resolution: the part that has to be sent, the engineer who has to travel or guide by video, and the test records your own customers are waiting for in the meantime. Everything that can be discovered in the supplier's workshop costs hours there; the same thing discovered after arrival costs weeks. That asymmetry is the whole argument for a structured FAT.
It also sets the baseline for later. The records exported from the machine during the FAT run are the reference your team can compare against a year later when a reading looks wrong — pressure calibration drifts, and a baseline is how you tell machine drift from valve rejects. For what the gauge can and cannot tell apart, see the notes on clamping and sealing fixtures.
Group 1 — identity and documentation
- The machine is the machine you ordered. Model, serial number and the rated range on the nameplate against the order: DN range, test pressure, clamping force. The nameplate is also what customs and your own asset register will use.
- Documents by name, not by category. English manual, wiring diagram, PLC programme backup — each listed in the order confirmation, each physically present at the FAT. "Documents included" is not a checklist item; which documents is.
- Voltage and frequency match your supply. Confirmed in writing at order stage, verified on the cabinet at the FAT.
Group 2 — the pressure path
- Static checks with the system at rest. Oil level, hose condition, no leaks under the power unit after ten minutes running.
- Pressure build-up behaviour. The ramp to the set pressure should be smooth, without hunting around the setpoint — hunting shows up later as nuisance alarms on real valves.
- Calibration of the reading. The displayed pressure is the number every test record will carry, so it is compared against a reference gauge at the FAT. On Changxun machines this is part of the commissioning agenda — pressure calibration and acceptance settings are confirmed with the buyer at handover.
- Relief and safety settings. Where the circuit is protected, the relief setting is checked and noted, because it defines the ceiling nobody should adjust in the field.
Group 3 — the hold, the timer and the acceptance factor
A hydrostatic test is a closed volume, a set pressure, and a hold time. Two settings decide whether two benches judge the same valve the same way, and both belong in the FAT:
- The hold time is watched against a stopwatch. Commonly 60 seconds for the shell test and 15 seconds for the seat test on small sizes — whatever your standard calls for, the timer on the panel should agree with the clock.
- The acceptance factor is agreed, not discovered. On machines with an automatic pass/alarm, a run counts as a pass when the pressure stays above the test pressure times the acceptance factor for the whole hold. Two benches with different factors will pass and fail the same valve; agree yours before the FAT run, not after a dispute. What the report may state about leakage is a separate question — see API 598 and ISO 5208 leakage rate classes.
- The test pressure itself. For a PN-rated valve, the shell test is 1.5 × PN and the seat test 1.1 × PN; the full ladder from PN10 to PN160 is in the hydrostatic test pressure chart. The FAT run should reach the highest pressure in your ordered range, not a demonstration value.
Group 4 — fixture and tooling completeness
Tooling is where quotations quietly differ, and the FAT is the moment to count it. Two benches with the same tonnage can arrive with different tooling, and the missing half is bought later at spare-part prices — plus the days the machine stands idle waiting for it.
- Plates or jaws for the whole DN range. Not the first size — every band. The DFS2-600, for example, ships with upper and lower blind plates for DN50-150, DN200-300, DN350-450 and DN500-600, plus four extension rods and two rod flanges, so the full DN50 – DN600 range is testable on day one.
- Each plate pair is mounted and tried once. A plate that sits in a crate untried is an assumption; a plate that has closed a valve at the FAT is a fact.
- Spares are consumables, not extras. Spare sealing elements on the shelf are cheaper than a rejected batch traced to a hardened seal. Count them at the FAT and put them on the packing list.
- Adapters for your body styles. Wafer and weld-neck adapters are machined to a standard and schedule; if you sell wafer butterfly or butt-weld end valves, the adapters are part of the acceptance scope, and whether the bench layout is vertical or horizontal for your largest bodies should already be settled — see horizontal vs vertical layout.
Group 5 — controls, records and data export
- One full automatic cycle on the screen. Low-pressure fill and vent, changeover to the high-pressure circuit, build-up, hold timer, and the pass/alarm verdict against the agreed acceptance factor — watched end to end, once, without the supplier touching it.
- The record survives the machine. Every run stored with its time–pressure curve, and the USB export performed at the FAT with the file opened on a laptop. The curve is what an incoming-inspection department and most end customers' inspectors ask to see.
- History query. Finding a specific valve's run in the machine's stored history, by searching for it, not by scrolling.
Group 6 — safety and packing
- Interlock behaviour. On Changxun dual-station frames, all hydraulic buttons are disabled while the workpiece is still pressurised — the interlock is tried, not described.
- Emergency stop and guarding. Where guards are fitted, they stop the cycle; where a stop is pressed, the cycle stops.
- The packing list matches reality. Tooling, spares, manual, wiring diagram, lifting points and crate dimensions checked at the FAT so the crate that leaves is the crate that arrives complete.
- The arrival-day agenda is agreed before shipping. What gets connected first, who supplies power and water on site, and when the acceptance settings will be re-confirmed after transport. Machines are typically commissioned by video call; the FAT is where that call is planned.
A one-day FAT agenda that covers the above
| Step | What happens | Result recorded |
|---|---|---|
| 1. Paper check | Nameplate against order; documents present by name; voltage match | Serial number noted on the FAT sheet |
| 2. Static checks | Oil, hoses, ten minutes running, no leaks | Leaks: none / observed |
| 3. Calibration | Displayed pressure against reference gauge; relief setting noted | Readings at two points on the range |
| 4. FAT valve, shell test | Your representative valve at 1.5 × PN, hold watched against a stopwatch | Hold time, curve screenshot |
| 5. FAT valve, seat test | 1.1 × PN on the closed disc | Hold time, verdict against the agreed acceptance factor |
| 6. Records | USB export of the two runs, file opened on a laptop | Files named and archived |
| 7. Tooling count | Plates per DN band mounted and tried; spares counted | Packing list signed |
| 8. Open points | Anything unresolved written down with an owner and a date | Nothing ships on a verbal promise |
Step 4 uses your valve, not the supplier's showcase piece. A showcase piece proves the machine can hold pressure; your valve proves the machine can test your valve — which is what you are actually buying. If you cannot attend, the same agenda runs live over video call, with the USB export shown on screen.
Frequently asked questions
Can a valve test bench be accepted remotely if I cannot visit the factory?
Yes, with structure. The supplier runs the agreed FAT agenda live over a video call: your representative valve is tested at the shell and seat pressures, the hold timer is watched against a stopwatch, and the test record with its time–pressure curve is exported to USB and shown on screen. Changxun standard benches ship with commissioning support by video call covering pressure calibration and acceptance settings, so remote acceptance follows the same format.
What is an acceptance factor on a test bench?
It is the tolerance applied to the hold: a pass is recorded when the pressure stays above the test pressure multiplied by the acceptance factor for the whole hold time, and a drop below it raises an alarm. Agree the factor before shipment — two benches with different factors will judge the same valve differently.
What tooling should ship with a valve test bench?
Sealing plates or grip jaws covering the whole ordered DN range, not only the first size — on the DFS2-600, blind plates for DN50-150, DN200-300, DN350-450 and DN500-600 plus extension rods and rod flanges — together with spare sealing elements, which are consumables. Count both at the FAT.
Which documents should I receive with the machine?
As a minimum: an English manual, a wiring diagram, and the PLC programme backup. Ask for them to be listed by name in the order confirmation rather than described generically as "documents", and verify each one is physically present at the FAT.
What valve should be used for the FAT test run?
A representative valve from your own production: the size that dominates your order volume, tested at its real shell and seat pressures — for a PN-rated valve, 1.5 × PN for the shell and 1.1 × PN for the seat. Testing a showcase piece proves the machine can hold pressure; testing your valve proves the machine can test your valve.
What should be settled before the machine is packed?
The open points from the FAT closed in writing, the tooling and spare-part list checked against the order, the FAT records exported from the USB port kept as the acceptance baseline, and the commissioning agenda for arrival agreed — what connects first, who supplies power and water, and when the acceptance settings are re-confirmed after transport.
Benches this agenda applies to
DFS2-600 · Dual station
- DN50 – DN300 / DN350 – DN600
- 0 – 7 MPa water
- Pass/alarm against acceptance factor, USB export
SYJ-600SQ · Water + air
- DN50 – DN600
- 40 MPa water / 4 MPa air
- 90 t clamp, one clamping for both media
ZZC2-200 · Automatic line
- DN50 – DN200 gate valves
- Strength / lower seal / upper seal sequence
- Printed test records
For the selection path from pressure to leakage class to layout to fixture, start at the valve pressure test bench buyer guide, then the test pressure chart, the leakage rate classes, and the clamping methods reference.
Send your valve types, DN range and test pressure — the proposal includes the clamping method, the model and the FAT agenda, within 12 working hours.