ASME VIII-Div 1 Tube-to-Tubesheet Joint Shear Load Test 1

[jcat4]

Hi All,

The history docket for a tube bundle we received from our manufacturer demonstrates that several tubes fall outside of the manufacturer's specified percent wall reduction range. We are now concerned that the tube-to-tubesheet joint strength is not adequate for these tubes and we may have to reject this tube bundle.

Can a shear load test (A-3) be used to confirm if the tube-to-tubesheet joint strength is adequate?
i.e. if L (test) > Lmax (calculated as per A-2), then OK?

Additional Information:
- tubes and tubesheet material: titanium
- % wall reduction range specified: 4% - 8%
- roller expansion: expanded, enhanced with two or more grooves
- designed using the no-test joint efficiencies
- successfully tested in accordance with UG-99

Any help would be greatly appreciated.

 

[SnTMan]

jcat4, my experience is that test strengths are usually far above the strengths calculated per Appenix A. You can certainly have a sample joint tested and accept or reject on those results.

Are you more concerned with joint strength or leak tightness? I am not sure the former directly says much about the latter.

I believe my concern with an over-rolled joint would be related to the ability to re-roll the joint in the future if it were to start leaking. That, in turn, might depend on how much it was over-rolled.

In any case, it would normally be your manufacturers' responsibility to prove the joint in a way you can accept.

Regards,

Mike

 

[jcat4]

Thanks for the quick response Mike.

We are more concerned with leak tightness (sorry I forgot to address this in my initial post).

Our shell and tube heat exchanger has oil on the shell side and water on the tube side. The water pressure is greater than the oil pressure. Therefore, if a leak occurs at the tube-to-tubesheet joint, water will contaminate the oil and may potentially result in a multiple bearing loss at our steam turbine.

So, does testing for joint strength not necessarily address leak tightness?

Also, does the fact that the tube bundle was successfully tested in accordance with UG-99 mean that it is actually OK - regardless if some tubes fall out of the manufacturer's percent wall reduction range?

Jennifer

 

[SnTMan]

jcat4, last first, successful hydro really only means the hydrotest was successful, that is, no leaks were found at hydrotest conditions.

I would say that joint strength does not directly prove tightness.

If it is feasible you might consider a more sensitive tightness test such as an air test or even a helium test of the joints. One of these, provided you can identify a realistic acceptance criteria, can more directly tell you what you want to know.

For an air test with bubble soultion, no visible bubbles is pretty common, for helium you will have to identify an acceptable leak rate.

For the fluids in your service an air-soap test should be adequate.

Regards,

Mike

 

[jcat4]

Again, thank you very much for your advice Mike.

 

[SnTMan]

jcat4, you're welcome, I should have added, your manufacturer may have already done an air-soap test, it is very common to do so. You might check with them.

Regards,

Mike

 

[chicopee]

Your OP did not state if the tube ends were rolled or surface welded?. If rolled, couldn't you have the manufacturer surface weld as extra reaasurance?

 

[jcat4]

chicopee, the tube ends were rolled. They are not surface welded.

The tube bundle is already sitting at our station. (Note: to avoid confusion from my OP, we still have the opportunity to reject this tube bundle even if it has already left the manufacturer's shop.)

So, it would be difficult to ask the manufacturer to perform surface welding since it would take several weeks to ship the tube bundle back to their shop and then back to our station.

Nevertheless, it is still an option and we will take it into consideration.

Thank you.

 

[SnTMan]

jcat4, if you bought a bundle only, the manufacturer probably did not do an air-soap test, as there would be no vessel to pressurize. You could perhaps do so at the site by installing the bundle in the shell, and pressurizing at 25 psi or so.

Regards,

Mike

 

[rmw]

Will this bundle go into a shell that can be pressurized, or is it a condenser bundle?

rmw

 

[jcat4]

Mike... sounds good. Thanks again.

rmw... yes, this tube bundle will go into a shell that can be pressurized. It is not a condenser bundle.

 

[rmw]

jcat4,

Then you are good to go with some of the suggestions already made.

I can't imagine why your vendor didn't sell you welded tube to tubesheet joints. Were they asleep at the switch or something, or did they lack the capability?

rmw

 

[jcat4]

Sounds good rmw. I only joined the project when the tube bundle was already completed and shipped to site. So I don't really know what happened with the specification or the tendering process.

Again, thank you very much everyone for all of your help!

 

[Duwe6]

jcat4: think hard before revising your spec to welded tubes. Rolled is much better if you ever plan on retubing and reusing the tubesheel. Typically, rolled & welded, or welded only is just used for Lethal Service, or service where the corrosion rate [including erosion due to flow] is 0.0001" per year or less. One additional thought; if the bundle is operated at a temperature at or above 300-deg F, the tubes will tighten up in service. Thus a 'weep' leak at room temp will almost always result in a tight tube at working temp & pressure.

 

[jcat4]

Thank you for your input Duwe6. We will not be revising the spec for welded tubes. It was just an option to be considered for the future.

That is an interesting point about tubes tightening up during working temperature and pressure. Thanks again.

 

[rmw]

I don't know what duwe6's background is, but I would take exception to most of what he/she stated. I agree with the lethal part. But that's about it. I have dealt with precious few lethal services (although the ones I did have experience with were deadly lethal - phosgene) but I have had plenty of experience with welded tube to tubesheet heat exchangers in a wide variety of services and metallurgies. Spent the whole day just today in meetings dealing with just the issue of tube welding for a Hx on cooling water service.

I will state on the other hand that of all the Ti tubed Hx's (with either all Ti or overlayed Ti tubesheets) were welded. Seal welded, that is, not strength welded.

I'm sitting here puzzling on how a Ti tube in a Ti tubesheet is going to tighten up at temperature when the coefficient of expansion is identical. I'm willing to learn, however.

rmw

 

[gr2vessels]

It sounds paradoxal, but the hole in the tubesheet will not necessarily expand/increase the diamenter with the increase of temperature. It would rather tend to decrease, as the material expands towards the void. The tube on the other hand is so thin, that the ID/OD will increase, tightening the joint in the tubesheet. I believe that was Duwe6 refering to.
Cheers,
gr2vessels

 

[jcat4]

Thanks again everyone for all for all of the great feedback.

Now just for my own curiosity...

It seems that there is no correlation between percent wall reduction and joint shear load strength. Or percent wall reduction and leak tightness. (Or so it seems through some quick research).

If this is the case, then how do manufacturer's know what percent wall reduction will give adequate joint shear load strength and leak tightness?

 

[SNORGY]

It is my experience that 8% to 10% tube wall reduction will usually give rise to optimum "pull strength" (i.e., the force required to pull the tube out of the joint) for carbon steel and austenitic materials. I have seen the number as high as 12% in some reputable manufacturers' tube rolling procedures. It is not so for materials that work-harden easily; I once had a pull test done for martensitic stainless steel (air cooler bundle) and the optimum ended up at around 4.5%.

It will depend on the tolerance between tube OD and tubesheet hole ID as well, and also on whether or not grooves are present (and how many).

Regards,

SNORGY.