Dear all,
I m trying to compute Stress intensification factor for various piping fitting usign FEA (ANSYS not NOZZLE FE) to study the factors influencing SIF Eg D/d , T/t / Pressure , etc. Recently I came across a paper stating explanation of the same.
Link : http://www.cbi.com/getattachment/82f9d3d9-e1f8-47a6-8570-b1379bc4c0f3/A-Finite-Element-based-Study-on-Stress-Intensifica.aspx
My Question is as below on how to compute ,
If suppose im using 2 model,
Shell model:
Compute Equivalent Von-mises stress (Top/Bottom) = Pm + Pb --------1
Calculate Stress in straight pipe = (Applied load * Dist from load pt to Mx stress pt) / Section modulus -------2
FEA stress = (Pm + Pb) * FSRF / 2 Eq. 2{9} reference ---------3
FSRF = 1 for pipe , 2 for weld
SIF = Eqn 3 / Eqn 2
Or
SIF = Eqn 1 / Eqn 2
Solid model:
Compute Equivalent Von-mises stress to find Max stress point
Draw Stress classification line across thickness -----------------a
Compute Pm + Pb + Q ------------------1
Calculate Stress in straight pipe = (Applied load * Dist from load pt to Mx stress pt) / Section modulus -------2
FEA stress = (Pm + Pb + Q) * FSRF / 2 Eq. 2{9} reference ---------3
SIF = Eqn 3 / Eqn 2
Or
SIF = Eqn a / Eqn 2
Should i linearize the stresses before computing or take as such and compute the ratio.
Any sort of help will be appreciated!
Thanks in advance !!
Hi
You find a long in depth explanation on the use of Stress Intensification Factors in document DNV Recommended Procedure C203 (DNV RP-C203)
https://rules.dnvgl.com/docs/pdf/DNV/codes/docs/2011-10/RP-C203.pdf
Simply put, with a coarse model, you use the SIF to estimate the peak stress using rules that are based on analytical models or interpolation between numerical cases. Alternatively, you apply the loads of the coarse model as boundary conditions to a refined local FE model and directly compute the peak stress. It is then up to you to assure model convergence.
Incidentally, the French Code Aster users have a charming name for the latter - they call it a structural zoom.
For combined loads, you must pause and consider what you are doing as the peak stress location may not be at the same place for the two different load cases. You can of course just sum peak stress using SIF but this may then become unnecessarily conservative. Simply combining load cases using structural zoom is then less conservative.
Sincerely
Claes
- Badges
- Science topic