Fatigue crack at a crane runway bracket weld
A 12 mm thick S355J2 bracket is fillet-welded with an 8 mm leg to the top flange of a crane runway beam. The bracket carries a reversing load of 60-120 kN, about 2 x 10^6 cycles over 14 months. A crack has initiated at the weld toe on the flange and grown 25 mm along the toe. The client proposes simply increasing the fillet leg to 10 mm and re-welding the cracked area. The bracket plate is set square to the flange, with the weld toe in the high-stress region of the flange. The weld shows slight undercut and a poor toe profile. The WPS is qualified and the welder is capable. The client wants a quick fix to avoid changing the bracket position.
1. Is increasing the fillet leg from 8 to 10 mm likely to solve the fatigue problem? Explain the mechanics.
Read the worked answer
Increasing the leg from 8 mm to 10 mm increases the throat thickness and reduces the nominal shear stress on the weld throat. That may help if the throat were the governing fatigue location. In this case the crack started at the weld toe on the flange, so the critical location is the toe, not the throat.
Fatigue at a weld toe is driven by the local stress range and the stress concentration at the toe. A larger fillet can actually make the toe geometry worse if it increases the weld toe angle or creates a sharper transition. It also adds stiffness at the bracket-to-flange connection and may attract more load into the detail. Unless the original 8 mm fillet was undersized for the static load, a 10 mm fillet is unlikely to remove the fatigue crack cause.
The visible undercut and poor toe profile are also significant. They act as crack initiators. Simply re-welding over the cracked area without removing the crack and improving the toe will leave a similar detail in the same high-stress region. The proposal treats the symptom, not the mechanism.
2. Propose a design and fabrication corrective action that addresses the toe crack, including how to treat the existing crack and what NDT is needed.
Read the worked answer
Design change
The best solution is to move the weld toe out of the high-stress region or reduce the stress concentration. Options include changing the bracket to a bolted connection, introducing a gusset or transition that spreads the load, or using a full-penetration weld with a smooth transition and grinding the toe. If the bracket position must stay, a local design change such as a longer bracket with a tapered end can reduce the stiffness mismatch.
Crack treatment
The crack must be removed completely. Gouge or grind out the crack to sound metal, then inspect the excavation with magnetic particle or penetrant testing before re-welding. Ultrasonic testing can help determine crack depth before excavation. Re-welding over a crack is not acceptable.
Fabrication controls
Use a qualified repair WPS, clean surfaces, controlled heat input and a smooth weld toe. Where permitted, grind the toe to a generous radius or use TIG dressing to improve fatigue life. Avoid undercut. Record the repair and the final toe profile.
NDT
After repair, perform visual inspection and magnetic particle testing on the toe. If the design change is structural, use UT or PAUT on the load-bearing weld. If fatigue life is critical, consider a post-repair inspection after a short service interval.
3. What would you tell the client about the relative cost and schedule of your proposal versus the 10 mm fillet?
Read the worked answer
I would explain that the 10 mm fillet is cheaper initially but carries a high risk of repeat failure. It does not change the toe stress concentration, does not remove the existing crack properly if the shop simply welds over it, and may require another shutdown within a similar service period. The cost of a second repair includes crane downtime, inspection, scaffolding or access, and lost production, which usually dwarfs the difference in weld size.
The design and fabrication correction costs more up front because it requires engineering review, crack excavation, NDT, possible bracket modification and a controlled repair. It should be scheduled with the crane out of service and the load removed. The benefit is a detail that addresses the fatigue mechanism rather than only increasing the weld throat.
If the client insists on the 10 mm fillet as a temporary measure, I would require full crack removal, toe grinding and a documented inspection interval. I would not present it as a permanent solution.