In the off-highway and mining sectors great use is made of hardfacing alloy. This case study is a good example of the need for ensuring that welding procedure tests are carefully considered.
In the manufacture of large capacity ore grabs, the bucket lips, made from BS4360 Grade 43A steel, were hardfaced on both sides of the bevelled edges in a lattice pattern. Six grabs hardfaced in this manner were put into service and in four of these cracks appeared in the parent metal and weld metal, varying in length from 30-150mm.
Inspection of other welded bucket lips before being put into service revealed isolated cases of cracking which were considered to be unacceptable due to the risk of propagation. It was not possible to view the cracked bucket lips, which were situated in Malaysia, but four samples of new bucket lips were made available to TWI for welding procedure tests so that a procedure could be developed to produce welds free from cracks.
To assess the tendency for cracking of the parent metal, bead on plate welds were made with 5 and 6mm diameter low alloy steel electrodes giving martensitic weld deposits with hardness of 400HV. Welds were made with low and high currents with constant travel speeds so as to vary the energy input. No preheating was used in this case. Visual and magnetic particle inspection 48 hours after welding did not reveal any cracking in either weld or parent metal.
Since some of the cracks found had occurred at positions where the edge of one weld had been remelted by the adjacent weld, it was suspected that the high hardenability of the low alloy steel weld metal may have initiated cracking which had propagated through the weld and parent plate during service. To investigate this, double layer welds were made as well as two weld runs side by side, and macrosections were examined for the presence of cracks in the remelted parts of the first welds deposited.
No parent metal cracking was observed and no cracking occurred in welds remelted by adjacent weld beads. The hardness of the heat affected zones of the first welds deposited was lowered by 100 points by the tempering effect of adjacent weld beads. It was concluded that some additional causes of cracking was present that had been absent in the procedure tests.
One possibility was that when the bucket lips were hardfaced on one side the contraction of these welds may have caused sufficient distortion to increase the restraint on the welds deposited on the second side. A further test was carried out in which the bevelled part of the bucket lip was hardfaced with the welds deposited in a particular sequence.
The first side was hardfaced and without interruption the reverse side was then welded. A short time after welding, transverse cracks were seen in welds numbered 1 and 4 on the first side welded. Therefore, it appeared that there straint caused by contraction of the second side welded was sufficient to cause hydrogen induced cracking of welds on the first side which by this time could have cooled to the temperature at which susceptibility to cracking occurs. If the problem had involved solidification cracking which occurs at higher temperatures, it would have been welds on the second side that had cracked.
Additional tests showed that cracking could be avoided by two alternative procedures:
- Weld first side and allow the welds to cool to room temperature before welding the reverse side.
- Preheat the parent plate to 200°C before welding, which would allow welding of the reverse side without interruption.
This case illustrates why welding procedure tests should replicate precisely the welding sequence that occurs in production.