What should be the RF pad width to be considered for pipe to pipe branch connection, if pad is required only to reduce the stress at the junction from Stress analysis not for internal pressure.

Code B 31.3

Please can someone clarify?

Thanks

Read B31.3,304.3 Branch Connections and APPENDIX H.





There are certain thumb rules that can be followed in case of dummy/Trunnion RF pad. (Caution:Check with your supervisor).

D-Header dia, T- Header thickness
d-branch/dummy dia, t-branch thickness

Then RF pad can be of dia 2*d and thickness of T.

I presume the same can be used in your case.

Read this also:
http://65.57.255.42/ubbthreads/ubbthreads.php?ubb=showflat&Number=19266#Post19266
http://65.57.255.42/ubbthreads/ubbthreads.php?ubb=showflat&Number=31163#Post31163

Dear Stressguy,

Thanks for your reply,

'B31.3,304.3 Branch Connections and APPENDIX H' calculations deals with connections to sustain the internal pressure not for induced stresses due to thermal expansion.

Please correct me if i am wrong.

This is my opinion!

When RPad is required only to reduce local stress generated by external piping on a vessel (WRC 107) or piping, the width of Rpad should be enough so that the local stress generated at edge of the pad are within stress allowable limit.

Above check is done while doing WRC 107 hence width of the Pad is iterated to a standard value and finalised.

In case of pipe, I am a not sure that we can proceed with WRC calc. (Even if geometric parameters may be within limits as specified by WRC107, the boundary condition is different- here we have just support/guide rather than an anchor). You may go back to Prof. Bijlaard's paper can come to a conclusion.

Presuming d/D>0.15, Having a Rpad dia greater than 2*d might not be effective for analysis purpose (which means stress generated is no longer a local stress) and taking practicality of welding it.
Hence, above thumb rule of " 2*d and thickness of T" should be good.

Minumum pad width sqrt(r*t) and round up (2" min practical width). Maximum width r. Another high tech choice 2.5*sqrt(r*t), (vessel code ~1.1*sqrt(d*t)). Typically trunnion pad widths should be << r.