Reduction in hardness due to stress relief (again!!)

[andyenergy]

The earlier post on this subject was specific to a different steel alloy. I have experienced a reduction in hardness on parent material after a stress relief.

The background - An 8" diameter shaft (BS 970 Grade 080M40 - .36 - .42% C) was machined down and overlayed with weld metal about 2" thick for a test piece. The test piece was then furnace stress relieved at 620 C for 7 hours, with a heating rate of 50 C/hr and a cooling rate of 50 C/hr. When we did hardness tests of the parent material (the thickest section) the hardness values obtained were about 176 HV about equivalent to 549 N/mm2. This is perfectly within the specification of the base material, BUT the mill cert states that it had an original UTS of 706 N/mm2 (equivalent to about 221HV). Why the big reduction in hardness? or why the very high UTS in the first place?

I've had the original material hardness checked and the values are about 225-230 HV which is suggesting that the mill cert is correct. I also (prior to this test) carried out a stress relief test on some En8D material (0.45%C 0.72% Mn)and although the entire heating and cooling process was not as long they soaked for about 6 hours and 13 hours. There was no significant difference between the as recieved and the two stress relieved samples.

The only thing I can think of is that my 080M40 material was hardnened and tempered and that the stress relief has undone the tempering. However the section thickness (8&quot suggest that this was not the case due to LRS.

Anyone any ideas how this might have happened?

Andy

 

[unclesyd]

First check and see what was the thermal history of the bar as received.
Look at where the hardness checks were located physically in respect to each other.
There can be a considerable difference in hardness across the diameter of of a large bar, though I can't say for your material.

 

[Carburize]

The key may be the original tempering temperature - as unclesyd has noted - check the final tempering details for the production material and the piece on which you did your trial, if the sample which you weld overlayed was tempered close to 620C then a reduction in hardness would be expected.
Other areas to check
1)The actual furnace temperatures - is it possible that the weld test sample was actually heated above 620C?
2) Effects of the welding process itself - if it was only a small sample it is possible that the base material was "tempered" by the welding before it was put into the furnace - try a hardness traverse immediately after welding.

 

[andyenergy]

Thanks guys - good responses.

I haven't yet been able to check the actual furnace temperature, but micros of the material indicate pearlite islands that in some locations are just beginning to spherodise - there is plent of lamellar pearlite left and the spherodised areas are still in long strings - they have just begun to ball up. On this evidence, i don't think that the temperture has been excessive (given the extended time). I recognise I need to find this out, which is why I have requested the records and a section of the original unheat treated material.

The mill certs are very difficult to read (faxed and in Czech). The material was supplied as hot rolled turned bar and there are no heat treatment temperatures quoted. Grain size is quoted as 7 according to ASTM E122, which seems about right to me. The stockholder, who supplied a certificate of conformity claims that the material was tested in the normalized condition. I don't know where the original test pieces were taken from, but the UTS still seems quite high for a medium C/S. The yield was reported as 426 N/mm2 and the hardness 179 HB (equivalent to about 189HV, which is about 608 N/mm2) so this seems reasonable.

As I noted in my original post the section was quite thick - actually the overlay ranged in depth and at the shallowest area was perhaps only 10 mm depth, with about 40 mm of parent below it. The HAZ is about 4 mm wide and gives a hardness of 221HVish. The HAZ structure is tempered martensite, as expected. The parent hardnesses have been done on a conventional Vickers machine within about 5 mm of the edge of the HAZ and by portable hardness testing on the end of the sample where the parent is thickest and least susceptible to heat effects from the welding.

I can't really get my head round why the parent material if it was normalised in the first place was so strong and why such a large drop in UTS should have taken place even given the extensive stress relief and the appranetly small amount of spherodisation.

(that turned in to my met report didn't it!)

I'll keep you informed of progress.

Many thanks

Andy

 

[unclesyd]

If I understand you last post I think you may be too close to the HAZ for a good hardness test. As "Carburize" states traverse hardness testing is warranted.
Carbon steel 0.40 can achieve 82,000 # T.S. with a RB of 165-170 with Normalizing at 1650̊F and air cooling. This material can be Spheroidized, but I see no reason for any spheroidization.

0.40 C steel is a little tough to be welding on without preheat. Was any preheat applied to your sample prior to welding?

 

[stanweld]

Assuming that the original mill tests were done on the material in the hot rolled condition or normalized condition, your stress relieving heat treatment will lower hardness and strength. And the lowered hardness is not unexpected.

When post weld stress relieving heat treatments are required and hardness/tensile properties of the base material are desired to be maintained, you should order your material to be tested with all anticipated thermal processing.