Stress Relief vs. Normalization

[teb1]

I am welding a 3" thick x 70" OD tube x 48" long. This piece will be machined afer welding on the OD and ID. This is not a code pressure vessel part. Since I want to machine it after it is welded and I don't want it "walking around" during the machining I am debating whether to specify a normalization vs a stress relief post weld heat treatment of the tube.

I would appreciate any thoughts and recommendations.

Thanks.

 

[vesselfab]

if this was a code part

if you normalize you would have to redo all mechanical properties testing after normalization (tensile pulls)

 

[teb1]

As I said this is not a code vessel part. I just want it stablized so that it stays dimensionally stable during the machining process.

I didn't know about having to re-test for the mechanical properties after normalization thanks for that info. I wouldn't be able to re-cert the material material since I don't have any scraps from this material to normalize and re-test.

Does anyone in the audience have experience along these lines? Will a high temp stress relief 1150-1200F for a 3-4 hours accomplish what I want?

 

[metengr]

Yes, a subcritical stress relief is what you want for this application. A subcritical stress relief thermal treatment means that the temperature for obtaining stress relief does not exceed the lower transformation temperature of the steel.

If you specify a normalization thermal treatment this will alter the bulk properties of the steel, and will void the mechanical properties reported on the MTR.

 

[bklauba]

Another option you have is to vibratory stress relieve it. No loss (or gain) of phys properties. See

http://www.vsrtechnology.net

BK

 

[teb1]

Metengr,

What would be the lower transformation temperature of ASTM SA-36 material?

Bkluaba,

I have researched vibratory stress relief in the past. I never could find definitive proof that it worked. There are testimonies and statements but I have never seen hard data such as before and after strain data or residual stress data from strain gages that would prove that it worked.

Thanks for the input though.

Tommy

 

[vesselfab]

1100*F min to 1200*F max is what is normally done

2.25 to 2.5 hours holding time for full pwht

yours may not need to hold quite that long, but while you have it up......

 

[metengr]

What would be the lower transformation temperature of ASTM SA-36 material?

About 1330 deg F for this material. I would remain 50 deg F below this temperature, which is where most construction codes settle on 1100 deg F min for subcritical thermal treatment.

 

[teb1]

Thanks everyone.

I really appreciate the input and advice.

Tommy

 

[unclesyd]

Depending on the amount of machining,criticality of the dimensions,and thickness I might consider a second stress relief after rough machining.

 

[dbooker630]

Have you looked at Meta-Lax (

http://www.bonal.com,

http://www.meta-lax.com)

vibratory stress relief? I have heard a lot lately about this technology for weld stress relief although I have not seen any residual stress values either. They do report 500% fatigue life improvement on one of their brochure though!

 

[teb1]

dbooker630,

I tried the "blank" equipment a couple of years ago on a large gear that I was doing a weld repair on. The kindest thing that I can say about it is that I wasn't satisfied. Believe me they make some wonderful claims, but if method really worked as well as they say, I think the ASME would recognize it as an alterantive to thermal stress relieving. If I thought it worked I would try it again and might buy one but I would have to see proof, not claims.

Thanks,

Tommy

 

[bklauba]

reb1,

Look at the website I sited (

http://www.vsrtechology.net

). Open the link and go to the technical library. Probably the best work that might surprise you is that of Dawson and Moffat. They took numerous samples of 3 different alloys, and succeeded in removing 90% of the stresses initially present, using resonant vibration.

Dawson was at Perkins Engine at the time; Moffat (who sent me a clean, hard copy of this work for inclusion in the VSR Library) was at the U. of Liverpool.

This work was first published in the UK, later by the ASME. It still is considered the most fundamental (which its title states) work in this area.

BK

 

[unclesyd]

I go along with teb1 on this one as this large of a part is not the place to run lab tests.

This is not saying this process doesn't work but based on my experience there has got be more definitive and measurable results for it to be a viable alternative to the thermal process of stress relieving.

 

[bklauba]

Here is the URL of the work by Dawson & Moffat. It now works.

http://www.vsrtechnology.net/wp-content/uploads/dawson-moffat.pdf

BK

 

[DMiller01]

I have a similar part I am considering either stress relief or normalizing.

It is a 62" long a992 10wx88 wide flange I-beam with 3/4" thick gussets every 9 inches. It is heavily welded with many 100% welds.

It will be used in a press application and will be subject to cyclic loading (every 3 seconds). As a result I am concerned about fatigue.

Being an I-beam it has some brittleness in the "K-area" from the steel mill. Also due to the extensive welding I am concerned that it might be considered to have heavy constrained welding.

The I-beam is .08 carbon and the gusset material is .18 carbon. I guess you would consider it a mild carbon steel that had no heat treatment from the steel mill.

I think that normalizing it will
1. remove brittleness at "k-zone" and around welds
2. Reduce the "over strength" of the welds as compared
with the base metal
3. And remove welding stresses.

I am concerned that I don't significantly weaken the part. Am I over thinking this?

Picture attached.

 

[bklauba]

DMiller01,

You might want to download the paper found at this link:

http://www.vsrtechnology.net/influence-of-vibratory-treatment/

Thermal stress relief fared rather poorly against vibratory, in enhancing fatigue life in this study.

Your workpiece would also lend itself well to the VSR Process. Where are you located?

BK