long duration of calculation

[NataCalc]

Hello,

In my FEA there are viscoelastic+2nd order Ogden material models, hex meshed single bar and boundary conditions extracted from 4D CT scan which capture the extension-flexion-extension of the wrist. To make sure that boundary conditions reproduce well the physiological movement, the rigid bodies with two bones model and applied boundary conditions from 4d CT scan was run. In actual model boundary conditions are applied to the end of bars. The question is it normal that simulation is running for months? The deformation is relatively large. The interpolation does not help significantly to optimise the convergence therefore the run time.
Following Febio parameters are used:

max_refs=150; %Max reforms
max_ups=0; %Set to zero to use full-Newton iterations
opt_iter=150; %Optimum number of iterations
max_retries=150; %Maximum number of retires
dtmin=(1/numTimeSteps)/100; %Minimum time step size
dtmax=1/numTimeSteps; %Maximum time step size
tStep=0.3; %Duration of step
min_residual=1e-15;
Mag_val=1;

[NataCalc]

The problem was solved. The large deformation causes long run. By refining the deformation improves not only the convergence but CPU time as well.

[maas]

Hi,

I'm glad to hear you were able to make the model run faster. I just wanted to add a comment on the max_retries parameter.

In general there is very little gained from increasing this parameter from its default value. The bigger you make this value, the smaller the changes to the time step size when a time step fails. When a time step fails, FEBio will retry the time step with a smaller step size. The new step size is calculated from:

new_dt = old_dt - old_dt / (max_retries + 1)

So, when setting the max_retries to a large value, you are decreasing the step size by very small amounts when a time step fails. It is very likely that FEBio will continue to fail for many tries, until the step size has decreased sufficiently. In general, I advice to leave this parameter to its default value of 5 and instead try to avoid failure of time steps by adjusting other parameters in the model.

Cheers,

Steve

[NataCalc]

Thank you Steve. I will take this into account.