Dear Sirs,
I would like to have some informations regarding the Flux/Source BCs in Comsol 4.2. I have read that they correspond to the Neumann BCs. I am using two different modules (General PDE and Structural Mechanics) and compare the results. I am solving a pure traction problem on a rectangle in 3D.
I am modelling the material as incompressible and using Neo-Hookean law. Using the Structural Mechanics module, the imposition of the traction bc is easy using the Boundary Load option. In that case, I define a force per unit of area [N/m^2] and I get the proper results. When I use the General PDE module, I define my traction BC (Neumann type BC) using the Flux/Source BC. In that case, the correct general formulation is
-n flux = g - qu
where I set q=0 and g should be my forcing term. In that option the value is measured as N/m so I guess that I should insert the traction value I used in the Structural Mechanics module divided by a length. Is it right? How would you usually impose this kind of BCs?
Supposing that my approach of imposing the traction value divided by a length is correct, since the boundary face where the traction is applied has two length (in my case it's a planar surface) for which length I should divide the traction?
Many thanks,
I hope to have been clear.
Kind regards,
Paolo Tricerri
I would like to have some informations regarding the Flux/Source BCs in Comsol 4.2. I have read that they correspond to the Neumann BCs. I am using two different modules (General PDE and Structural Mechanics) and compare the results. I am solving a pure traction problem on a rectangle in 3D.
I am modelling the material as incompressible and using Neo-Hookean law. Using the Structural Mechanics module, the imposition of the traction bc is easy using the Boundary Load option. In that case, I define a force per unit of area [N/m^2] and I get the proper results. When I use the General PDE module, I define my traction BC (Neumann type BC) using the Flux/Source BC. In that case, the correct general formulation is
-n flux = g - qu
where I set q=0 and g should be my forcing term. In that option the value is measured as N/m so I guess that I should insert the traction value I used in the Structural Mechanics module divided by a length. Is it right? How would you usually impose this kind of BCs?
Supposing that my approach of imposing the traction value divided by a length is correct, since the boundary face where the traction is applied has two length (in my case it's a planar surface) for which length I should divide the traction?
Many thanks,
I hope to have been clear.
Kind regards,
Paolo Tricerri