setup of free-shape optimization/right direction of changing
Hi guys,
I'm trying to optimize T-connection of aluminum tubes (photo for better understanding only, later I want to optimize existing mountain bike frame). The goal is to obtain changed shape(section) of tubes that have nonlinear shape on length. This must be a kind of automatic optimization, not the one with predefined shapes. I couldn't find a good tutorial or informations.
First I have a question how to optimize this type of connection. Is free-shape optimization good for this kind of optimization?
Second: I modeled some kind of free-shape optimization and i would say it is going almost well. I mean the iterations/convergence are/is reached but model is infeasible. When i take a close look at result I see that my model is not changing shape in direction of element normals but along them. It seems like a burning paper roll. That is the thing i want to fix. Additionaly model is optimizing on the free edge that is far away from stresses. Connection is not changed.
My current setup:
Responses: mass, stress from non design
objective: min mass
constraint: stress upper bound 280 (MPa)
shape design variable: created and added sidecon barrier with bmface elements (barrier is a elements offset of one tube)
design space: first tube
non design: second tube
Maybe someone did this kind of optimization a can help me? I need this for my thesis.
For more information ask me here.
<?xml version="1.0" encoding="UTF-8"?>