Dear Mr. Richard and Mr. Dave,

I have a case when the tie-in point flange is connected to branch connection.
Is it possible to evaluate stress (EXP)at branch connection by input of force and moment as boundary condition in the tie-in point rather than displacement?

L1: W+P+T1+F1 (OPE)
L2: W+P+F2 (SUS)
L3: L1-L2 (EXP)

where;
F1- tie-in point force and moment during OPE(T1) case
F2- tie-in point force and moment during SUS case

Thank you in advance.

If F1 provided in the operating load case gives your the correct operating position and if F2 provided in the installed (or SUS) case gives you the correct installed position then these force sets would be proper.
But how do you know what these forces are? We observe position, that's a measured value. Typically, we do not easily measure (internal) load. So, you iterate - adjusting those 6 forces and moments until the forces produce the observed or expected position.
I assume you have some complication that does not allow you to simply enter those known(?) deflections instead of these approximate load sets.

Dear Dave and Ron.

It seems like the connected piping system is Vendor or other's scope.
I believe sometimes considering the operation mode (control valve On-Off, etc), stress engineer should verify the stress and flange leak, etc. for combined piping system including the Vendor's.

When stress engineer want to verify aboves, vendor provide the interface information such as displacement, generally.

But, considering reversely, stress engineer want to assess the piping system is safe or not, at that time, we send the interface load information to vendor, because we already incorporate the displacement information in our stress analysis system and found the interface loads.

I believe vendor or the other side can verify his piping system using provided interface load (SUS and OPE)... and the input of these forces is same concept like calculated spring support (Variable) will be applied in some operation load cases and SUS case (not rigid, as designed).

Regards,
Ince Kim

I think if we connect the pipe to pipe by weld, the only way to calculate its safety is stress result specified at the B31 codes not inputting displacement or force&moment. That's for checking local stress I guess.

Also, inputting displacement or loads without piping's stiffness(or spring rate), can leads us to have different and non realistic values.

Thank you for your reply.

I agree with Mr. Ince. Actually the exact scenario is that we calculate the forces and moment (SUS/OPE) in the interface flange connection using the displacement from the vendor as boundary condition. The Vendor has to verify their piping system using the interface loads from us and incorporate it to their system as external loads.

Dear Mr. Dave,

Regarding your reply " But how do you know what these forces are? We observe position, that's a measured value. Typically, we do not easily measure (internal) load. So, you iterate - adjusting those 6 forces and moments until the forces produce the observed or expected position."

Mr. Dave did I answer your question about the forces? These forces and moments are taken from the calculation of connected piping system. Could you please explain more when you say "we observe position"? Do you mean piping deflection at SUS and OPE case when you say "position"? Thank you.^^

You say "These forces and moments are taken from the calculation of connected piping system." So this "connected piping system" is the portion that you are NOT including in your CAESAR II analysis? If so, how was this connection point modeled in that analysis that produced the forces and moments that you have?
When you break a system at an arbitrary point and independently analyze the two subsystems, the deflections and rotations of this connecting point in both models must match (and the forces and moments at this point must be in equilibrium) for you to consider the analyses acceptable.

If the break point is not equipment to piping, you'd better model vendor's piping into your model. It's easier, and more reliable.

Then, as I said, stresses are only things you have to consider(if there are flanges, leakage could be concerned). In my poinion, piping model shall be equipment to equipment, even if some piping is not your scope.

As Mr. Dave mentioned, too much variable caused by NOT modeling piping will lead you to get wrong results.

So, you would request their drawing for piping as well as support in advance.

Thank you Mr. Dave and Mr. MK for valuable inputs. I fully agree with your replies.