Warping of Tubesheets and Flanges

[GinBread]

I've been doing engineering design of ASME Sec VIII pressure vessels for about 7-8 years now and I've noticed a recurring topic over all these years that I'm having trouble finding in depth literature or discussions about. As the title suggests, I am looking for any and all information, guidance, personal experience, etc... that pertains to how to prevent warping when welding a tubesheet or a flange to a pressure vessel. This happened so often at one of the previous companies that I worked for that we actually designed our heat exchangers to have stub ends so that when we welded the tubesheets to the shell stubs, we could take the stub and machine the tubesheet flat again. It was just built into our production process, which just feels very wrong to me. I no longer work at that company, but a different one now and I am noticing that same trend happening here as well. I've gotten little tidbits of half-baked ideas on how to prevent this sort of thing from occurring, but I've never seen a fully fleshed out practice put into play. So, if you have anything at all to share on this topic, I would greatly appreciate it!

 

[EdStainless]

The welding sequence is important in this.
I have worked with a lot of HX manufacturers and have never seen this be an issue.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[DVWE]

It would be helpful to know material types, sizes, thicknesses, diameters, etc. you’re working with. I am in agreement with Ed, warpage is a less talked about issue.

The devil is in the details; she also wears prada.

 

[GinBread]

At my previous company, we saw it with the worst effects in diameters exceeding 12 NPS, where the tubesheets were anywhere from 3/4" to 1" thick. Amazingly, we did a vessel that was around 30" diameter and tubesheets that were closer to 2" thick... and they managed to warp the tubesheets by at least 1/8".

All typical carbon steel materials, SA-516-70, SA-106-B, etc...

Edit: There were also some very small diameter (5" pipe) vessels that had custom ring type flanges that were around 7/8" thick that always got warped so badly that the studs wouldn't line up properly with the mating blind.

 

[DVWE]

Most tubesheets I deal with are 2-1/2+ inches thick. Hasn’t really been an issue.

One thing you might want to consider is a “chill block” for the thicknesses you mentioned. It is the same or similar thickness and diameter as the tubesheet, and bolts to the back side or opposite side of welding. It acts as a major heat sink, and also a strong back of sorts. I’ve seen them with multiple bolt circles drilled into them for use on different sizes. They work really well.

The devil is in the details; she also wears prada.

 

[GinBread]

Is it better to use the same material or a material with a higher heat transfer coefficient? The previous company I worked for had something very similar to this except that it was made of solid aluminum.

Feel free to point me towards some literature instead of just straight up giving me an answer. I genuinely want to understand this issue better.

 

[NRP99]

This book or book on welding might help in understanding distortion.

Analysis of Welded Structures: Residual Stresses, Distortion, and Their Consequences by Koichi Masubuchi

 

[r6155]

See WELD DISTORTION in

http://www.lincolnelectric.com

Regards

 

[GinBread]

Good stuff, thank you r6155 and NRP99

 

[EdStainless]

We used steel or SS blocks when needed to prevent possible contamination from other metals.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed