Hello fellows!!

What part of the code that permits the use of FEA in lieu of B31J, to obtain SIF's and flex factors for branch connections.

Thanks,

I think if 45 deg. lateral connection is not available in B31J, then it's time to use FEA or physical tests.

Warm Regards,

Hello Borzki,

I think you have to use codes for BPV. And it is same as for ASME and EN codes. The B31J it is piping code and applicated for beam models only.

But we can try to compare results obtained by both types of codes. And do not forget that in piping codes SIF for a butt weld is included already (1.8-2). Also SIF in pipe code some times conservarive and for some pair D/t it is not realistic. Becouse one formula should cover all variations.


See codes: ASME BPV Sec. III (NB-3200 DESIGN BY ANALYSIS) or BPV Sec. VIII Div2 (Part 5 DESIGN BY ANALYSIS). Or EN 13445.

Best regards Pavel

For B31.3, see paragraph 300.c.3.

Thanks Richard for that valuable information.

Warm Regards,

Very strange. the only thing that nobody can answer is what is going on with the 45 degree branch.
It looks to be protected by some magic and SIF can not be foreseen.
What is so special of this angle that makes impossible to get this SIF?
If FEA can determine the values, then some benchmark can be done to see what is the range of these SIF respect to the 90 degrees. Then to integrate into the classic formulas. Lets say 1.5 or 1.2 or whatever factor it will go out from the results.
And then to close this matter.

If these 45 degrees branches are so mystical and there are dangers associated with this, then why not to exclude them and replace with latrolets or angled tees?
Why to continue to say we use them but we do not know anything about them?

Doesn't seems to be logic. Common sense should prevail.


Some companies accepted a 60 degree angle instead the 45 as being acceptable. So the problem was solved somehow. The 45 degree branch was mainly requested by process for the flare systems to facilitate the flow. Not all the companies are considering this approach.
Point is that 45 degrees branches are also very difficult to fabricate, even more difficult when there are reinforcements pads.

for me is a mystery.

Hello fellows,

I have experienced before to use 90 deg. welding tee. I have seen also a study to use the 90 deg. connection since it is favorable in terms of strength considering also the flow in the study, in which this is the reason why lateral connection is normally been the practice.

Anyway, my senior engineer before just multiply the 90 deg. SIF by 2 to be used for lateral and design the piping system accordingly, instead of going thru the pain of complex mathematics.


Warm Regards,

The "magic" that "protects" the lateral is likely the complexity of variability.

Laterals are not covered dimensionally by ASME, thus it's difficult to expect a piping code to try to keep up with something that isn't standardized.

I would expect that at some point it might be possible to develop a function that would cover SIFs and flexibility factors that would be a function of OD1, OD2, WT1, WT2, crotch radius1, crotch radius 2, crotch radius 3, crotch radius 4, lateral offset, and approach angle.

However, until then, we do what we must to move forward. I have also seen the doubling of SIFs.

I've attached images depicting CL/CL intersection, weld center, and OD/CL intersection. At a minimum, we should be incorporating an offset to not take credit for the length between red and green, but likely red and blue makes more sense from a conservative standpoint.

But for those 2 sharp corners whose location are dependent on the angle. Or that long, sharp edge where the two cylinders are coming together. Are we certain for what range a SIF + safety factor of 2 is accurate?

Hi Michael,
It's not like we send a rocket to the Mars.

This is one thing that it did not have any progress in about how many years, 60? 80?

We had progress in any field but not on this.
Still I do not get it.

Hello Fellows,

If the tools is available in the desktop, might as well use it and compare it to doubling of SIF.

What is your opinion in the torsional and axial SIF's which is now available in this tools.

I am doing one analysis and it has an impact (it qualifies without it and fails if I include it).

What would be an alternate fitting instead of providing excessive flexibility. In my previous experience, it's good that the client approves the use of 90 deg. connection.

Warm Regards,

Why, Dan, have you not progressed this? Sheesh!




My opinion is if you have better information to indicate failure, then it fails.

If someone strong arms you into doing it a certain way, then you denote such.

"By stamping this document, I am authenticating the adherence of this piping configuration to the codes and standards as laid out by the end user, and am not qualifying its actual suitability, whose responsibility belongs to the end user. I note the suitability is called into question based on the calculations as determined by..."

If progress mean to use expensive software just to determine the SIF, then no.

I do not always have the luxury to have a FEA license.
But next time when I will need it I will ask here for values.

By Occan's razor... the best explanation would be we don't like the result. For different reasons, of course.

Thanks Michael for the very good advice. I have recommended a lay-out that would get rid of torsional moment but it has access issue. In the end, all agreed to use that sif's from the tool and just optimize the lay-out.
Better safe than sorry.

Cheers,