Converting forces from axisymmetric to 3d/symmetric ?

[adlh01]

I am trying to simulate a prototype on ANSYS in 3D. It has symmetries in the XY and ZY plane so I am only going to simulate 1/4th of the prototype. The prototype is an artificial lens. It has small 'legs' in the edges where it can be 'pulled', which is where the load will go. So the load is perpendicular to the axis.

The load is obtained from literature studies on the human lens. In these studies the load is usually reported as the total load required to stretch the lens by a certain amount in the x-axis, and values are from 0.08 N to 0.1 N. Usually these models are axisymmetric; I cannot do an axisymmetric model for my application because the 'legs' in the edges remove the axial symmetry.

In the 3D model, I wish to apply this load as a pressure on the surface of the 'legs', not a point force. My questions are 1) Should be pressure be calculated using the surface area for the application point in the entire model (the prototype's 4 legs), or just the 1/4th part that I'm modeling?

Right now I am doing it for 1/4th of the model and the reaction forces in both symmetry planes add up to 0.08 N.

Thank you.

 

[AlephZero]

The best way to check you have the correct loads and symmetry boundary conditions this is to make a simple "complete" model with of something the same symmetry as your lens, and a "quarter" model using symmetry.

If the results are the same, you are doing it right.
If the results are different, the looking at the differences will probably tell you what you did wrong.

In "real-world" engineering, people do checks like that all the time, if they are using an option in the software for the first time. It's much better than not being sure that you did the right thing - or even assuming the user manual doesn't have any errors in it!

 

[nvn]

Should the pressure be calculated using the surface area for the application point in the entire model (the prototype's four legs)? ...

I did not understand why there is only one "application point in the entire model (the prototype's four legs)," instead of four application points in the entire model (one on each leg).

... Or just the [quarter] part that I'm modeling?

Yes. In other words, for this particular situation and load orientation, first pretend the entire model is there, and compute the pressure (p1) you would normally apply to each leg. Then, apply exactly that same pressure (p1) to the leg in your quarter model.