Two Mixed Bundles of Flat Bar: Bending Cracks and a Hard Zone Beside the Weld
A fabrication unit receives two bundles of 25 mm × 6 mm flat bar to make 40 bench legs. Bundle A is stamped 'MS IS 2062 E250' and bundle B is stamped 'EN8', but the storekeeper has mixed them because the tags faded in the rain. Each leg must be cold-bent through 90° on a hand bender and then arc-welded to a 5 mm MS base plate. Ten legs from the same bundle crack on the outside of the bend; those same legs, when welded, show a hard zone and a hairline crack beside the bead after cooling. A file slides easily across the cracked legs' surface but skates over a sample from the other bundle. The supervisor wants to 'soften them with a gas torch and carry on' so that the 4 pm dispatch is not missed.
1. Identify which bundle is which and set out the evidence chain that supports your conclusion.
Read the worked answer
The cracking bundle is EN8 (approximately 0.35–0.45 % C); the other is IS 2062 E250 (about 0.20 % C).
Evidence chain:
- Cold-bend cracking on the outside of the bend. The outer fibres are in tension and must stretch. Higher carbon means higher strength but much lower ductility and a higher ductile-to-brittle transition, so EN8 cracks at a 90° tight bend radius while E250 deforms.
- File test. Hardness rises with carbon content; a file cuts E250 but skates over EN8. This is a quick, shop-floor confirmation, not a substitute for a hardness test.
- Hard zone and hairline crack beside the bead. EN8 heated into the austenitising range and then cooled quickly by the cold base plate and the surrounding mass transforms to martensite - hard, brittle, and prone to hydrogen-assisted cold cracking. E250's lower carbon gives a much softer, tougher heat-affected zone.
All three symptoms point the same way, which is why the identification can be accepted without waiting for a laboratory report.
2. Explain why the crack appears beside the weld rather than in it, and evaluate the supervisor's 'torch-anneal and carry on' plan.
Read the worked answer
The bead itself is high-carbon diluted weld metal, but the critical region is the heat-affected zone: it is heated just below melting, then chilled by the cold base plate and the jig. In a 0.4 % C steel that produces untempered martensite plus a hydrogen-rich, restrained joint - the three classic ingredients of cold cracking. The crack therefore runs in the coarse-grained HAZ a millimetre or two from the toe, often hours after welding, which is why it seems to appear 'later'.
The torch-anneal plan is not acceptable as written:
- A gas torch gives no temperature control or soak; you cannot hold 600–650 °C uniformly along a 25 mm × 6 mm section without a pyrometer and a soaking time.
- Cooling rate is uncontrolled - a soft patch beside a hard patch is worse than a uniformly hard zone because it concentrates strain.
- It leaves scale, distortion and residual stress, and the bender will simply crack the softened-but-embrittled bend zone again.
- It treats the symptom. The real fault is material selection for a cold-bend operation.
A controlled stress-relief in a furnace with a chart recorder would be engineering; a torch in the yard is guesswork.
3. Recommend a material and process plan for the 40 legs, and say how you would separate the mixed stock permanently.
Read the worked answer
Material: use E250 (IS 2062) for all 40 bent legs - it is designed for cold formability and welding, is cheaper, and is what the drawing realistically assumes. Keep EN8 for parts needing wear resistance or strength after heat treatment (keys, pins, shafts), not for cold bends.
If EN8 must be used anyway: bend hot (850–900 °C, slow cool), or bend with a generous radius, then weld with a qualified procedure - preheat 200–250 °C, low-hydrogen electrodes dried at 300–350 °C, small stringer beads, controlled interpass temperature, slow cooling in sand or vermiculite, and a delayed crack check after 24–48 hours.
Stock separation: do not rely on faded tags. Cut a 100 mm sample from each bundle, stamp the grade on the sample and on the bundle end, colour-code with paint (a different colour per grade), and keep a bin card with heat number. Record the grade on the issue slip so the next operator does not repeat the mistake.
Reject the 'weld one of each and see' test: it produces one good and one cracked joint and tells you nothing you did not already know from the file and bend results.