Nozzle to shell diameter ratio 1

[ANK72]

Hello all,
Can someone please advise what calculation procedure to use for evaluating stresses at nozzle attachments larger than half the vessel diameter?
There is an existing heat exchanger that was designed to ASME VIII Div 1 in the fifties. It will experience uplift in a seismic event thus causing dynamic loads on the nozzles.
The shell is made of DN500, XS SMLS pipe (508mm OD, 12.7mm thk) and the product outlet nozzle is DN250, XS SMLS pipe (273mm OD, 12.7mm thk).

My understanding is that the given D/d ratio is outside the scope of PD5500 and WRC107. Can you please recommend an alternative procedure?

The nozzle loads (both static and dynamic) have been established from an AutoPIPE stress analysis.
Many thanks

 

[codeeng]

You can use NozzlePro which is a module of FEPipe. This is an FEA programe using Div 2 criteria. You also get spring rates of the attachment which can be incorporated into the pipe stress analysis to reduce loads.

 

[ANK72]

codeeng,
Thanks for your input. I just want to know what method caters best for large round attachments. If PD5500, Bijlaard or WRC 107/297 don't cover d/D>0.5 then what does? FEA is not an option as we don't have any in-house FEA software. Have to do it the good old manual way.

 

[StressGuy]

FE/Pipe Nozzle Pro is something you should really look at if your company does much vessel engineering. The program is not a true generalized FEA (i.e. build something completely from scratch). It has a collection of templates that you basically plug parameters into. Nozzle Pro in particular is ideal for reviewing nozzle loads. As far as cost goes, if your company can afford AutoPIPE, it can certainly afford FE/Pipe and it is at least as useful as a pipe stress program.

Did WRC-107 even exist in the 1950's?


Edward L. Klein
Pipe Stress Engineer
Houston, Texas

"All the world is a Spring"

All opinions expressed here are my own and not my company's.

 

[codeeng]

A lot of design software use end-of-curve values for d/D ratios outside WRC but d/D > 0.7 requires analysis per Code. I've used this criteria in the past but you should use good judgement. Otherwise its better to use FEA.