Hi, in regard to spring constant at a specific point, i have a question. a sample case is as below.

(pipe of BOP) ----------|---------- (pipe of Boiler side)
welding connection & T.P (1)

in this case, first this analysis is performed separately at T.P(1).
Movement data(2) at T.P(1) from boiler side is considered in analysis for BOP part. and then calculated force and moment(3) from BOP analysis is considered in analysis for boiler side again.
In here, I`d like to consider actual stiffness(spring constant) data at TP(1) in Boiler side analysis.(to reduce stress value)
so i approach to use basic formula F=& x k.(F/& = K)

calculated force from BOP analysis(3) / movement data from Boiler analysis(2) = spring constant at T.P(1)

finally i input calculated spring constant at T.P(1) in divided boiler side modeling and confirm reduced stress values.

Is this a wrong approach or correct?

Please show your experiences and theory.
thank you.

It sounds to me as if you wish to remove an artificial boundary condition that separates your two models. Your current method is to iterate between both models until the loads and displacements between them equalize.
Why not include them both in the same model and let the software balance this internal junction?

There are some reasons to have to divide into two modeling.(division, company, program)
Anyway, in this case between two models, i am wondering whether my approach to apply stiffness values is right or not.

And if two models are able to be unified into a single model.
Is follow steps right?
1. input unit force at a specific point such as TP(a),
2. Calculate stiffness at the point based on displacement
3. Apply the stiffness to TP(a) with force and moment in boiler part modeling(not unified)

The purpose of my approach is to reduce stress values for boiler header piping.
Thank u