Time Steps and Step Times

[lexiedm14]

I am working on simulating an anterior drawer test for a knee model. I have prescribed a 140N Rigid force (Rigid constraint) to the tibia, but I am unsure what is the correct time step and step time to use. I would like to have data points for every 10N but when I have done two different time steps and step sizes, my data gets different results.

Here is what I have so far:
#Data : Y - displacement

1 0
2 -0.277363956
3 -0.554273546
4 -0.829988301
5 -1.10360467
6 -1.37464225
7 -1.64258349
8 -1.90781415
9 -2.1707046
10 -2.43138742
11 -2.69004941
12 -2.94677401
13 -3.20156789
14 -3.45448184
15 -3.7055347

Time steps: 14
Step size: 0.01
Max Step size: 0.01
Min step size: 0.0001


#Data : Y - displacement

1 0
2 -1.28666353
3 -2.29551077
4 -3.12529516
5 -4.18544292
6 -4.79165077
7 -5.41136122
8 -5.99449301
9 -6.94237328
10 -7.85927868
11 -10.9846926
12 -11.7048731
13 -12.36339
14 -14.7815981

Time step: 10
Step size: 0.07
Max step size: 0.07
Min step size: 0.0001


I am not sure how to get the correct data points as they both produce different displacement data. How do I know what would be a good time stamp to choose?

Thanks,
Lexie

[maas]

Hi Lexie,

I assume you have a load curve assigned to your rigid force. If this curve is defined over a time range, for example, from 0 to 1, and you want output at every 10 N, you would set the time steps to 14 and the step size to 0.07143 (= 1 / 14). In your cases above, you get different answers because the time settings span different time ranges (0 - 0.14 in case 1, and 0 - 0.7 in case 2).

Cheers,

Steve

[lexiedm14]

Hi Steve,

This helps a lot. Thank you for your help.

I have noticed that when I simulate two different knee models with the same rigid constraints, material properties, forces, etc., I get vastly different displacements (4.5 vs 12.6 mm). In both cases, the auto time stepper was turned on and they had the same time step and step size values however, each modeling scenario had different time steps in Postview. The 12.6mm model has a total of 17 time steps while the 4.5 mm model only has 9. Could this play a role in why I am getting completely different data?

[maas]

Possibly. Although all the physics is similar, I assume the geometry of the two knees differ and that can have a dramatic effect on the results. Perhaps one knee had some negative Jacobians at some point that slowed down the progress, but the other knee was able to avoid negative Jacobians. It is difficult to say without seeing the models, but geometry can definitely have an impact on convergence and even results.

Cheers,

Steve