tube to tubesheet joint for U type heat exchanger 1

[pastyl]

Hi everybody,
Facing the problem of what method should you follow for having the required tube to tubesheet joint the following codes are applicable.
1) ASME VIII Div 1,UW-20 gives the required weld lenght for four acceptable weld types-fillet,grooved or combination welds - full strength or partial strength.
(Logic, since for these types of welds can anyone calculate it's strength).
ASME VIII Div 1,App. A gives the weld load for different (and many others from UW-10) weld combination with expansion or not.But App. A does not apply to U tubes.
.So for U tubes with joint method different from these shown on UW-20, ASME VIII Div 1 applies or not ?.
2) At HEI standard for Power Plant Heat Exchangers is written that if an acceptable weld joint has lenght from t to 1.4t (t=tube thickness) is strenght weld joint. And at HEI acceptable welds are all which are refered to ASME VIII Div 1,App. A .
At HEI Standard the fillet weld of the tube behind the tubesheet has equal strength with the autogenous weld, which happen if the tube part which is extended behind the tubesheet is being melt from a TIG welding process on the tubesheet (fig. f at ASME VIII Div 1,App. A).
So for a type f weld joint of U type tube, from which ASME does not have( as I think) any response, HEI considers it as a strength weld.
Any opinion of the differences in practice of the above mentioned weld types, fillet of type f autogenous, and for its strength will be appreciated.
In any case, apart from the weld, a 3 ring grooves strenght expansion is done.
Any opinion of what process, weld as described above or expansion, should become first would be highly appreciated.
Because both methods have advantages and disadvantages and the theory does not give clear results, I would prefer an advise for people who are heat exchangers manufacturers and have tested these methods.

 

[MrBTU]

Pastyl,
First, you must decide between welded or roll-expanded only joint. Our first choice is roll expanded joint. However, service conditions, tube and tubesheet materials and gage, and customer requirements sometimes dictate welded.
When welding there are two kinds of weld - seal weld, for sealing against leaks only, and strength weld, per UW-20. Then, for strength weld there is partial and full, and options for the weld detail as you have described. Even with a full strength weld we do a contact roll so as to eliminate the crevice between tube OD and tube hole, and thereby minimize the corrosion attack there. The grooving is not necessary for a strength welded joint, since it only increases the pull-out, and the pull out of a full strenth weld already equals the tensile of the tube.
You didn't mention appendix AA
So - can you give us more details as to your geometry, materials, and service - then we can give you more information.

 

[pastyl]

MrBTU,

As per AA the case is Fig AA-1.1(d),U tube type gasketed both sides, tubesheet not extension.
Tubesheet material is A105, tubes material is A179.Heat exchanger size is 20" .shell design pressure 5barg, tempe/re 200 C, tube design pressure/tempe/re 30/150 C. tubesheet thick. 86mm.
My question is first what is the difference between joint contacts as per Fig A-2(6) and Fig A-2(1) i.e.fusion with filler metal weld and fillet weld respectively. It is better to use many Amperes and melt the tube, extending at the back of tubesheet after rolling, to tubesheet, creating a Fig A-2(6) weld, or it is better to use low amperes in order to achieve a fillet weld as per Fig A-2(1).
My second question is why APP. A is not valid for U tube and how according ASME is characterized a joint as per Fig A-2(6) for U tubes, because according HEI is a strength weld.
Finally, nevertheless the weld joint, according HEI tubes must first welded and then expanded.But the above process creates problems to welds and the opposite process ( first expand then weld) to expanding portion.So, which is the preferred sequence of the process.
We use first expansion ( not only rolling but full expansion) with 3 ring grooves at 50mm lenght and then we make the weld according Fig A-2(6)or Fig A-2(1).

 

[DG72]

FYI--You might check the material specification for SA-105. It states that it is not intended for tubesheet material.

 

[rmw]

While HEI does recommend welding and then strength rolling and the serious Hx manufacturers that are the members of HEI and whose top technical people are on the rules writing committees, it does not prevent contact rolling prior to welding. Those companies do that as a regular course of manufacturing.

In addition to what MrBTU has stated about contact rolling, it also centers the tube in the hole which aids good welding.

But a contact roll is not tight enough to prevent the escape of the weld gases past the tube which prevents blowing out the weld at the completion point.

If you are doing autogenous welding, you may get away with it. If not, have a method to check for weld blow outs.

Then you full roll, but not all the way to the end of the tubesheet. Your crevice corrosion potential will drive how far you should roll on the full roll.

DO NOT roll over the top of the welded end of the tube. Begin the roll inside the tube ~1/2 inch past the weld to keep from cold working the weld and damaging it.

I can speak for 3 of the HEI companies and state that as a general rule they only use autogenous welding on Ti tubes to Ti tubesheets (or overlays.) But they are not usually doing that by hand either.

I fully expect Unclesyd to weigh in when he discovers this thread with a contrarian view, which he and I have gone back and forth on previously in this forum, but I do respect his experience and expertise and welcome his comments. He can vouch for the other method. He has done a lot of unusual things in his time and done them successfully according to him.

rmw

 

[pastyl]

First to make a correction on my previus thread.
Tubesheet material is SA-350LF2, not SA-105.
Although ASME and TEMA-as TEMA material requirements arise form ASME- does not prevent A105 for using as a tubesheet material, paragraph UCS-66 requirements does many time forbidden for this material as a tubesheet.
So, other standards as HEI does not include A105 as a tubesheet material.
Now, I would really appreciate a response for the differences in weld strenght between the two types I mentioned before-Fig A-2(6)or Fig A-2(1) of Ap.A.
It is understood that the strenght of a fillet weld can be calculated as its lenght and area are specific, but this cannot be done from an TIG weld with filler metal when tube melts behind the tubesheet,creating a Fig A-2(6) shape to the joint.
Why HEI puts these two types of weld to the same category while ASME not?. And how you can weld a very thin tube to the very thick tubesheet without tube totally melting, creating thus a fillet weld?.And how you can be sure that both tubes and tubesheet have been partially melt for ensuring thus a right a weld joint ?

 

[SnTMan]

pastyl, first, both UW-20 and Appendix A present methods for calculating the strength of a tube joint. Neither directly addresses leak tightness. Second, neither directly applies to U-tube exchangers because there is generally considered to be no loading on the joint. Either can be used, if desired, to calculate the joint strength.

In the context of Appendix A, weld dimension "a" > 1.4t is generally considered a strength weld, although I don't believe the term appears in the appendix.

The difference between the two sketches you mention are the difference between whether the weld dimension "a" is greater or less than 1.4t. The Code, TEMA, HEI do not mandate the type of joint used, it is based on may things including MOC's, service, tolerance for leakage, fabrication capabilities and preferences, etc.

My company welds a lot of light wall tubing, almost exclusively by just fusing it to the tubesheet. It
would be classified as a seal weld per UW-20, as there is no appreciable weld leg. However when considered as an Appendix A joint, a strength can be calculated based on factors other than the weld.

Joints are tested in several ways to indicate the quality, such as pressure and penetrant testing.

Regards,

Mike