Thickness-Range Trap in a WPS Revision
A welding engineer is revising a GTAW WPS for 6061-T6 aluminum aerospace brackets. The original PQR was qualified on 0.250 in (6.4 mm) plate with a single-pass butt joint. Production now needs a 0.375 in (9.5 mm) joint and a 0.100 in (2.5 mm) joint, both in the same alloy and temper. The engineer plans to keep the existing PQR and simply edit the WPS thickness range to 0.100-0.375 in, arguing that the process, filler, and joint type are unchanged. The customer's drawing invokes D17.1 and requires WPS qualification per the standard.
1. Evaluate whether the existing PQR supports the proposed thickness range, and explain the governing logic.
Read the worked answer
The existing PQR does not automatically support 0.100-0.375 in. D17.1 qualification is built on qualified thickness ranges derived from the test coupon, and the range is not symmetric around the coupon thickness. The 0.375 in joint exceeds the upper limit the 0.250 in coupon supports without additional qualification, and the 0.100 in joint is a separate thin-section case that may fall outside the qualified lower bound. The engineer's argument that 'process and filler are unchanged' is the classic wrong reasoning: thickness range is a qualification variable precisely because heat input, dilution, and pass count change with section. The correct path is to check the PQR's stated qualified range against the D17.1 thickness rules, and if either production thickness falls outside, qualify a new coupon (or a coupon covering the needed range) before revising the WPS. Editing the WPS range without PQR support is a documentation nonconformity even if the welds pass NDE.
2. What should the engineer do if the 0.100 in joint must ship before a new PQR can be completed?
Read the worked answer
There is no compliant shortcut that lets an unqualified thickness ship under a D17.1-qualified drawing. Options: (1) qualify a thin-section coupon first, which is the only fully compliant route; (2) if the contract permits, obtain customer engineering approval for a deviation with supporting data, but that is a concession, not qualification; (3) re-examine whether the 0.100 in joint actually falls within the existing PQR's lower qualified bound, because if it does, no new qualification is needed. The engineer should not ship on the assumption that thin sections are 'easier' and therefore covered. The judgement the exam tests is refusing to let schedule pressure convert an unqualified WPS into an accepted one.