Discussion Forum

I am modeling a deposition process using Tom Moffat's CEAC physics (see Electrochimica Acta 53 p145-154) which includes a pair of surface coverage ODEs defined on a deformed geometry. The ODE takes a standard form as

d theta_sps/dt = "curv_spatial" *v*theta_sps + additonal terms

where "curve_spatial" should be the COMSOL defined function available on deformed geometries, v is the velocity of the moving boundary (used also on the deformed geometry node) and theta_sps is the coverage fraction of the surface by chemical species sps. I have this ODE implemented in a "Boundary ODEs" node.

If I use the function "curv" all is fine and the time dependent model runs to completion. But my question is, according to the sparse documentation of the "curv" function, it would appear I should be using "curv_spatial" instead:

"If the model includes a deformed mesh, curvature variables are available both for the deformed configuration and for the undeformed configuration. Curvature variables for the deformed configuration (spatial frame) are formed by appending the suffix _spatial to curv (in 2D) or curv1 and curv2 (in 3D). The curvature variables in the undeformed configuration (material frame) are just curv (in 2D) or curv1 and curv2 (in 3D)."

If I use "curv_spatial" (all physics nodes are set to spatial) I get an unknown variable error. Is the documentation wrong, or do I need to do something different to get the current value of "curv" for the spatial frame (deformed)?