Stress Relief Before Welding

[tc7]

As part of our overhaul of large sheave blocks (frames that secure a double row of 20-inch sheaves), the frames usually require weld repairs of numerous cracks that reside either in some of the existing original welds or cracks that may have propagated into base material. The material is typically 4130 plate stock and the original design drawings for these frames show that the material at time of new construction was to be normalized. The structure sees high dynamic loads during it’s operation.

Our typical repair procedure requires complete grind out to sound base material, MT inspection of the grindout and then preheat to 400 deg F. We use an 8018-C3 rod for all repairs. Stress relief is usually by local means (torch) and we hold 1100 deg for the typical 1 hr/in.

The problem: after repair, new cracks may show up adjacent to the previous repair, OR may show up inches away from the repair. And this pattern will continue frequently for 4-5 iterations or more until cracks all go away.

It now seems obvious to me that these frames come in with tremendous residual stresses that somehow get aggravated during our repairs and cause new cracks. I am thinking that we should perform a full oven stress relief PRIOR to the beginning of any weld repair. My intent is to even exceed our typical stress relief temps (1100 deg F) and maybe go up annealing temps (1450 deg F ?) with a subsequent furnace cool.


Well that’s the background and plan – I would appreciate any concerns or comments. Will annealing at 1450 deg F degrade the overall characteristics of what was originally normalized material?

Thanks

 

[metengr]

tc7:
I concur with your pre-repair stress relief for this application. I would not stress relieve at 1450 deg F. Also, make sure you use low hydrogen electrodes for all weld repairs.

Perform all defect removal before pre-repair stress relief and upon cool down to 500 deg F, immediately begin weld repairs using local preheat maintenance.

 

[tc7]

Hi Met-

I would like to know the reason that you caution against the 1450 deg F stress relief (in favor of 1100 deg F)?


Also, I will not be able to maintain your suggeted 500 deg heat from oven to repair as I need to send the frame out to a not-so-local vendor with a sufficiently large oven. It will necessarily need to cool off for transportation - what is you concern with this practice?

Thankyou again.

 

[stanweld]

1450F is too low a temperature for annealing. Annealing will degrade the properties of both the base metal and previous deposited weld metal with E8016-C3. The cracking described with the methodologies that you are using for repairs of fatigue cracks are not unexpected.

 

[metengr]

tc7;
I thought that you could take advantage of being at temperature from a pre-repair stress relief operation to keep the entire component warm, and engage weld repairs using local preheat maintenance. This would result in as low a thermal gradient as possible.

Agree with stanweld where multiple repair efforts are not that uncommon because of fatigue damaged material located adjacent to repair locations. Cracks can develop under thermal tensile stresses (local heat). I also believe that some of the cracks you observed could have been caused by hydrogen cracking from in adequate preheat combined with residual stresses (before your stress relief operation).

 

[MikeMet]

I'm not sure a stress relief before welding will help, but it can not hurt. I would do the SR at 1150F for up to 2 hrs. Are you sure the cracking after repair are not delayed hydrogen cracks? Can occur especially very near the weld. I have also had good luck welding 4130 with a higher preheat of 550 - 600 F and then holding that preheat for 1 hr after welding and slow cool to RT. You don't mention how much of the base metal around the weld you preheat, but you should preheat to as far away as practical to ensure there is not heat sink caused by the surrounding base metak. I also think you could also consider shot peening the areas that typically crack after repair to try and prevent additional cracking

 

[unclesyd]

The mention of cooling rate by MikeMet might be very important depending on the constraints and thickness. We weld 4130 in 2" and 3" sections highly strained configurations where we always ensure that the cool down rate is very slow. We have the luxury of having an oven to preheat and cool the shop fabricated and repaired components. Even without preheat we still try to cool the weldment slowly.
We would tend not to preheat with a torch depending on the component configuration one might get some bad temperature gradients in the wrong places.
In a lot of field repairs of both 4140 and 4130 we will use Allstate 275 or Incoweld A or in cases of a wear areas we use Hastelloy C.

May I ask if you have the oven why can't you have the welder come to you?

Can you post a picture of the offending component?

 

[Tmoose]

So, the originals develop cracks somewhere between manufacturing and first overhaul, and repairs crack right away? Sounds like the design may not really be "right" for 4130, or vice versa.