1. The Field Problem: The Certificate Said Pass, The Bearing Failed

A distributor in Southeast Asia bought sixty backstops over two years from three suppliers. All arrived with a one-line inspection certificate: "rated torque tested, passed". Failures occurred in roughly one unit in twelve, always within the first six months, always with the same two signatures — uneven sprag contact marking and race pitting. The rated torque test had been passed by every one of them, because a short static reverse-hold at rated torque proves almost nothing about a backstop.

Rated torque is one property out of about ten that determine whether a backstop survives. A factory acceptance test (FAT) that verifies only torque transfers all the remaining risk to the buyer. This article sets out what a meaningful FAT looks like and what records should travel with the shipment.

2. Mechanism: Why the Torque Test Is Not Enough

A static reverse-hold at rated torque confirms only that the sprags can wedge and that the housing does not burst. It says nothing about:

• Whether the sprags share load evenly — a unit with 40 percent of sprags idle can pass a torque test at 100 percent while sitting at 1.7× local contact stress.
• Whether it will still work after 10,000 lock cycles — contact fatigue, not static strength, is the service failure mode.
• How much heat it makes in freewheeling — drag torque is invisible to a reverse-hold test.
• Whether it will engage at running speed — centrifugal lift-off speed is untested.
• Whether the geometry will fit the shaft — bore size inside tolerance, runout, squareness of the mounting face.
• Whether the material is what the certificate claims — hardness and case depth can be off without any visible sign.

3. Worked Example: A FAT That Actually Screens

The table below is a practical acceptance programme for a 6 kN·m class sprag backstop. It is written so that each item produces a number that can be compared with a specification, and a record that can be filed and audited.

Sampling logic: items 1, 2, 3, 6 and 11 on every unit; items 5, 7, 8 and 9 per batch (define the batch, for example 50 units or one heat); items 4 and 10 as type tests per series, repeated when design or supplier changes. This structure keeps the cost per unit reasonable while leaving no unverified assumption in the reliability case.

4. Engineering Solutions: Making the FAT Enforceable

• Put the test list in the purchase order, not in an email. Acceptance criteria must be numbers with tolerances, and the record format must be named.
• Require actual values, not ticks. A report that shows the measured torque, angle, hardness and Ra is auditable; a checklist of ticks is not.
• Witness or third-party inspect the first article and any unit above a defined value. The cost of a witness day is trivial against a failed conveyor.
• Define the batch so that sampling means something, and state what happens on a failed sample — retest of the batch, or rejection.
• Keep the dossier. Store the traceability documentation with the machine records; it is the only way to resolve a failure claim two years later.

5. Series Selection and Verification

Red flags in a supplier's answer: "we test every unit at rated torque, that is enough"; refusal to state the shock basis of the rating; a hardness value quoted as a range wider than 5 HRC; no serial-to-heat traceability; a type-test report with no date or specimen identification; and any offer to send the record "after shipment".

6. FAQ

Q: Is a FAT at rated torque sufficient for a critical conveyor?
A: No. It verifies strength, not reliability. Add torque-angle, drag, contact pattern and material verification to cover the actual failure modes.

Q: What is the single most valuable extra test?
A: The contact pattern blue-check. It exposes alignment, sprag seating and manufacturing quality in one photograph, and it is cheap.

Q: Should every unit be dynamically cycled?
A: Normally no — dynamic cycling is a type test. But require it whenever the series, design or manufacturing source changes, and hold the report.

Q: How do I handle a supplier who cannot provide hardness data?
A: Treat the material claim as unverified. In backstop duty, unverified hardness is the most expensive saving available to a manufacturer and the most costly risk to the buyer.