Applying Dirichlet and Neumann boundary condition on a sphere

[shenrilin]

Dear all,

     I'm modeling a spherical problem with clamped boundary condition on the lower part and traction boundary condition on the upper part as shown in Figure 1 and 2.    But I'm  not sure how to apply them.  It will be very grateful if anyone can give me some advice or an input file example.

    Thanks,

[JStorm]

In case you no the relation of the node distribution on the surface, you can use the LOOP command to calculate the coordinates of each node. Than you can apply the bc with e.g. CLOAD, CBOUndary and CDISplacement.

[shenrilin]

Thank you very much for your kind reply.

I want to apply uniform pressure on the spherical cap. If one applies this pressure load on the nodes directly,  one need to get the surface area and then distribute to the nodes, which is difficult for the irregular mesh.  So I wonder if there is a command or other straightforward way to realize it. If not, your method is also an option.

Looking forward to other advice from anybody else.

[Prof. R.L. Taylor]

At present 8.4 does not have general loads on regions.  We are working on forms that could allow use of pressure load on element faces.  If you can post a mesh (not too big) that we can use for testing we can try to see if the method could work for you.   

A new release is coming soon that could have some of the features you need.

[shenrilin]

Dear Prof. Taylor,

Thank you very very much for your reply.

I  attached the mesh of sphere, one with 519 4-node  tetrahedron elements (Figure1 and ISPH4) and the other with 519 10-node tetrahedron elements(Figure2 and ISPH10) for your test.
For your convenience, I describe the model a little.  A sphere model with radius 5 and lower cap is applied clamped BCs and the upper one is applied pressure loads. the height of sphere cap is 1.

By the way, would you please give me some guide such as which subroutine I can refer to  so that I may write my own code?

Thanks,
Rilin

[Prof. R.L. Taylor]

At present there is no simple way to do what you want.  If you have the surface mesh (attached are two for the inputs you provided) you could write a user module to apply loads on regions you define.  For curved boundaries this is not trivial -- but for spheres could be attempted.

[shenrilin]

Dear Professor,

Thank you very much.

Could you please tell which file define the function of command "*CSUR"? I think this file may help me to extend 2D distributed line loads to 3D  surface loads.

Best,
Rilin

[Prof. R.L. Taylor]

the routines affected are:

pecmes.f:        surfl = .false.  ! Calls ploadc
ploadc.f:      if(surfl) then       ! Does calls for load functions (including CSURf)
pmesh.f:        surfl = .true.    ! Does input of CSURf data --> saves to plinka -- which is a function that saves deferred data to disk.

The "ploadc" has several parts including the CSUR parts based on the "surfl" check.

[shenrilin]

Dear Professor,

Thank you very much. I will try what you told me.

Best,
Rilin

[shenrilin]

Dear Professor,

Do you have any published papers or books about how to convert pressure(surface load) to nodal forces theoretically and numerically?  I search the internet but cannot find much useful information.

Thank you very much.
Best,

[Prof. R.L. Taylor]

I think any basic book on finite element of elasticity problems will cover the applicaiton of pressure loading.

In our book by Zienkiewicz, Taylor & Zhu, "The Finite Element Method: Its Basis & Fundamentals", 7th ed, Butterworth-Hienemann, 2013, the topic is covered in Ch 7,  see pp 217-219.