70 mm cruciform joints: FCAW or SAW under a tight delivery
A bridge fabricator must produce 24 cruciform joints from 70 mm thick S355J2N plate (CEV 0.43), each with 1.2 m of full-penetration weld on both sides. The shop has one column-and-boom SAW station and three FCAW sets running 1.2 mm rutile-cored wire. The only existing WPQR covers 20 mm plate with FCAW. Delivery is due in eight weeks, and the SAW station is booked for another contract for the first three weeks. Winter shop temperature falls to 5 °C. The specification requires preheat per EN 1011-2, maximum 350 HV10 in the HAZ, and 27 J at -20 °C in the weld metal. Distortion is a known problem: the previous batch needed 15 mm of flame straightening on the arms. The QA engineer wants to stay with FCAW to avoid a new procedure qualification; the production manager wants SAW because it is faster.
1. Which process and joint sequence would you specify, and why? Address the 5 °C shop temperature and the distortion history.
Read the worked answer
The decision is driven by thickness, toughness and hydrogen, not by speed alone. At 70 mm with CEV 0.43 the joint is highly restrained and hydrogen-sensitive, so the process must deliver low weld-metal hydrogen and reliable -20 °C toughness. Rutile-cored FCAW typically runs at H8–H16 and is the least certain route to 27 J at -20 °C; it can be made to work, but with a basic flux-cored wire and strict baking, not with the rutile wire on the floor. SAW with an agglomerated basic flux (H5 class) gives high deposition, low hydrogen and good toughness, so I would use SAW for fill and cap after a qualified root. The three-week booking is a planning constraint, not a technical one: schedule the 24 joints after week 3, or subcontract. The FCAW argument also fails on qualification: a 20 mm WPQR does not cover 70 mm, so a new WPQR to ISO 15614-1 is needed in either case. For the 5 °C shop, preheat to at least 100–150 °C and measure it 75 mm either side of the joint; do not rely on ambient temperature. Sequence the welding balanced about the neutral axis, back-gouge to sound metal, use back-step passes, and measure arm distortion after every second pass rather than straightening at the end.
2. What essential variables and hydrogen-control measures must be in place before the first production weld?
Read the worked answer
The WPQR must cover the full 70 mm thickness, the process and consumable combination actually used, the preheat and interpass range, and the position. The qualification test coupon must include weld-metal and HAZ impact testing at -20 °C, a hardness traverse against the 350 HV10 limit, and macro examination of the full penetration. Thickness is an essential variable: a 20 mm test does not extend to 70 mm. The welders need ISO 9606-1 qualification covering the thickness and position. Hydrogen control is a chain, not a single measure: buy flux with a certificate stating H5 or better; bake it at the flux manufacturer's temperature (typically 300–350 °C) and hold it in a heated quiver at 100–150 °C; keep the wire coil clean and dry; remove rust, paint and moisture from the groove; maintain preheat until the joint is complete; cap interpass at about 250 °C; and let the joint cool slowly under a blanket. For a highly restrained 70 mm cruciform, consider a post-weld hydrogen soak at 200–250 °C for two hours. On the procedure coupon, do the hardness survey after the same cooling practice planned for production.
3. How would you control distortion and verify the finished joints?
Read the worked answer
Distortion in a cruciform comes mostly from unbalanced shrinkage about the neutral axis. Control it by welding alternately on both sides and from the centre outward, using back-step technique, keeping the arms free to move until the root is complete, and, where the design allows, pre-setting the arms against the expected shrinkage rather than correcting afterwards. Do not use strongbacks indiscriminately on a 70 mm joint: they convert shrinkage into residual stress and raise the cracking risk. Record distortion after each joint so the next one can be pre-set from measured data. Flame straightening is acceptable only under an approved procedure with controlled temperature and cooling, since it can alter toughness. Verification should combine volumetric NDT of the full-penetration weld (UT or RT, according to the specification) with MT or PT of the root and cap, and a dimensional check after welding and again after any straightening. Because the previous batch needed 15 mm of correction, add a dimensional hold point before the joint is released.