Hi all,
I hav seen there are lots of methods/opinions discussed in this forum for modeling air coolers. I hav gone ahead with the method mentioned in below link
http://65.57.255.42/ubbthreads/ubbthreads.php?ubb=showflat&Number=77#Post77
In our project, piping is qualified for codal requirement with a seismic acceleration of 0.23g
Pl. note seismic stress/loads are qualified for non friction case (as a project requirement, I understand that there may be difference of opinion for many of us on this).

My question is should we impose seismic acceleration at the rigid elements of the header box? Is there any factor that can be applied upon 0.23g?

Unlike other pressure vessel/ other equip since there is provision for lateral movement of tube bundle, this is becoming crucial.
The huge mass of air cooler with seismic acceleration makes it tricky.
Upon Imposing seismic acceleration, it is difficult to maintain a 6mm gap (API 661 Para 7.1.1.2).
I agree option of increasing this gap or thrust block can also be worked.

The philosophy of imposing seismic acceleration. Is it right?

If it is not required to impose, how difference in structural response of cooler and structure can be modelled?

Any opinion on this?

You should consider the acceleration, but you do not have to maintain the 6mm gap. The API 661 para 7.1.1.2. state that the vendor shall make provision for lateral movement (regardless the source, can be thermal or other). It is like a guide with gap.

Regards,

Thanks Dan,
Agreed.
Due to Non friction sesimic movement, it is difficult even to maintain a 15mm gap (Provided it is agreed by vendor).
After movement of 15mm, the header box elements impose a load of 28 KN on the thrust box (if provided).

Is this realistic?

Yes, if you do not have other supports for the pipe header. It will be 0.23 times Weight. But you have also a guide on the other side of the cooler, so you may want to refine your model to reflect this. About 0.23 times half of the cooler weight will act on that one.

Regards,

The stiffness of piping is has little effectiveness in restrainting the lateral movement.


Im a bit skeptical on the usual approach.
Is there any alternate way of modelling this?

other supports = Limit stops or guides (pending on the piping layout)

stiffness of piping =piping configuration with limit stops and guides

In this case, this is what is expected to have. 0.23 times weight.

regards,