load in axisymmetric analysing from load in 3d

[sanshi2000]

Dear all

I am using the axisymmetric elements to analyse the deformation of a thick-walled tube applied load (nodal force or pressure) at end, but the results are different with those from corresponding 3D solid elements. I was guessing the load for axisymmetric cases need to be change to some values related to radius or 2*pi, but I can not figure out the exact relationship. Could you give me some suggestions?   

The input files for axisymmetric and 3d under nodal forces and pressure are attached.

Thanks a lot in advance.

Lei

[FEAP_Admin]

When applying loads at the nodes you need to remember the factor of r in the r dr part of the consistent nodal force term.  Integrals are performed on a 1-radian sector; so no factors of 2*pi are needed.  When you use the pressure element (depending on your version) it should take care of the axis-symmetric case if you ask it to do so; for example

  mate,2
   pressure
   load -fs
   axis symmetric

look in elements/frame/presld.f.
 

[sanshi2000]

Thank you very much. The load for axisymmetric analyse can work well as it in 3d now, but axisymmetric analyse undergoing nodal forces still give different results with corresponding 3D analyse.

For example, apply constant nodal force at the top end of a tube fixed at the bottom and get displacement in 3D analyse. In order to get same displacement in axisymmetric analyse, how to calculate the nodal force in terms of the constant nodal force used in 3D analyse?     

Thanks a lot.

[FEAP_Admin]

Attached are two files. 

Iax3d shows how to do this with pressure elements in 3D the other options are CSURface loads or nodal loads.  Nodals loads are the hardest and are only recommended if you have a separate program that generates them for you.

Iax2d shos how to solve the exact same problem in 2D (axis-symmetry) with nodal loads (the least recommended method).  Better options are CSURface loads or pressure elements.

The problem is thin-walled pressure R/t ~ 10 and pressure of 10, so the hoop stress should be ~100 as is easily confirmed by the plots.

[sanshi2000]

Thousands of thanks. They are good examples to learn axisymmetric and 3d analyse. But what I am trying to do now is to calculate the corresponding nodal forces in axisymetric analyse from cosistant nodal forces in 3d of a same problem. Any suggestion and even references are very appreciated.

[FEAP_Admin]

In 3D you need to compute $\int_A N_A^T p n dA$.  Get a copy of Zienkiewicz and Taylor vol 1 and/or 2.