Discussion Forum

Hi all,

I was having some problems getting my shape optimization for a microfluidic channel
working. The idea is that the channel should have constant strain rate through its center.
The strain rate equations and constraint is set up correctly, however the shape
optimization on its top and bottom walls is not working. I am trying to set it up similar
to the beam optimization problem here:

www.comsol.com/blogs/designing-new-structures-with-shape-optimization/

The entrance height is Wu and the exit height is Wc, I am trying to use a berstain polynomial
to model the top and bottom walls. My current iteration is really close, however the right side
sags a little for some reason. the channel itself is centered on 0 and shifted in the x so that it begins at x=0. I create a prescribed deformation and outline the rectangular channel. The equation I enter for the displacement in the y is "(Yg/(Wu)-1)*((Wu)*((A2)*Xp^1*Xn^2+(A3)*Xp^2*Xn^1))+(Yg/(Wc)-1)*(Wc*Xp^3)" with Xp=Xg/Lc and Xn=1-Xp. This is the berstain polynomial. Under optimization, I control the length, a2, a3, and the entrance height (the entrance height is coupled to the exit height through and equation) while constraining the constant strain rate through the center. Attached is a picture of the output, for some reason it sags on the right side, as mentioned previously.