On 03/05/2018 09:09 PM, Ben Melosh wrote:

SfePy does not handle units at all, so the units of the results are determined by units of the input parameters (material constants, loads, etc.). For >example, SI units on input -> SI units on output. Note that all the input units need to be consistent. Some consistent unit sets: http://www.eng->tips.com/viewthread.cfm?qid=355753

I do not understand what you mean by the strain rate, because the solution of your problem does not depend on time. Maybe you just need the >ev_cauchy_strain applied to the velocity u?

Try using stiffness_from_lame() (see examples/linear_elasticity/linear_elastic.py) - in linear isotropic elasticity, the stress is, given the Lame parameters >lambda, mu:

sigma = lambda tr(e) I + 2 mu e

and in incompressible fluid:

sigma = -p I + 2 mu e(velocity)

so I guess you could set lambda = 0, and mu to the viscosity to get the shear stress part using ev_cauchy_stress.

r.

Thanks Robert,

I guess my main question regarding strain rate is weather there are equations in place to solve for non-newtonian flow?

I forgot about that one :)

It is not there directly, but you can update the nonlinear parameter depending on the solution as in the examples [1, 2]. The examples do not deal with flows, but by analogy, the viscosity could be made dependent on the velocity from a previous pseudo-time step.

Does that help?

r.

[1] http://sfepy.org/doc-devel/examples/diffusion/poisson_field_dependent_materi... [2] http://sfepy.org/doc-devel/examples/linear_elasticity/material_nonlinearity....