We make models

[matteo.dardari]

hello to all, use nx6 for meshare models to the finite elements of hip prostheses cups (semisfere hollow). for what concerns mesh I started from a 2d and I carried out a revolution to pass to 3d and cleaned it from some double knots. Now I would like to understand how I can bind solid bodies to mesh without mesharli and I would like to know if the technique I used to make mesh and optimal or there are better methods (regarding the program's help is not very exhaustive)! ! ! !

 

[IPdesign]

Good morning matteo and welcome

First of all, an earmark to not be presented in the appropriate section . (you can always make up )

for the specific question on the fem I hand to whom is definitely more experienced than me.

While I believe that the non-exclusive can be synonymous with course not done? correct me if I am wrong!

 

[cacciatorino]

Im rizzoli? :smile:

I don't know nx but one way could be:

1) for the creation of mesh: Revenue the shell of the internal or external semisphere, and make a mesh 2d, and then extrude it according to the thickness of the material. In this way you should have elements of homogeneous size across the entire semisphere area, which I think you don't have when you make a revolution. I recommend at least 4-5 elements in thickness.

2) for connection to rigid elements, there are those who use beam elements. or more simply you can impose that certain areas of mesh are infinitely rigid, simulating in this way the great difference of stiffness between the titanium or stellite support or what it is and the real acetabulus in polyethylene or ceramic. should know uin better the problem (materials, loads, condizoni of bond, etc...).

It also checks if you can use symmetry conditions, in this way you solve many grains.

A colleague told me here:
http://it.groups.yahoo.com/group/fem-analysis/people are very prepared.

 

[matteo.dardari]

Hi.
while I thank you for the availability.
ipdesign would tell me where the section is to introduce me and I assure you that I do it because I didn't find it and thought I had completed the registration and apologized if I didn't do it the right way.

My university rapporteur is Professor Cristofolini who works at the rizzoli but I am doing the thesis at a company of boycott. My problem is that I have to make a fem with nx and I think it is very different from ansys (the rapporteur at the company knows a bit 'ansys and told me that it is different from nx, first worked at the rizzoli). indeed my mesh starts from a section from which I echo the 2d and after the revolve of the latter I make the mesh of a central cylinder not put before with tetrahedra because otherwise I got it with too sharp corners.
basically I have to study the behavior of inserts(liner) in polyethylene in physiological load condition, with support cup and femoral head in titanium. the whole system is hyperstatic the only thing I care about and demonstrate what the load effects on the support (liner).
Thanks again.

 

[IPdesign]

ipdesign would tell me where the section is to introduce me...

There she is. http://www.cad3d.it/forum1/showthread.php?t=14379

 

[cacciatorino]

My university rapporteur is Professor Cristofolini who works at the rizzoli but I am doing the thesis at a company of boycott.

He was also my rapporteur, say hello to me!

My problem is that I have to make a fem with nx and I think it is very different from ansys (the rapporteur at the company knows a bit 'ansys and told me that it is different from nx, first worked at the rizzoli). indeed my mesh starts from a section from which I echo the 2d and after the revolve of the latter I make the mesh of a central cylinder not put before with tetrahedra because otherwise I got it with too sharp corners.
basically I have to study the behavior of inserts(liner) in polyethylene in physiological load condition, with support cup and femoral head in titanium. the whole system is hyperstatic the only thing I care about and demonstrate what the load effects on the support (liner).
Thanks again.[/QUOTE]You should see a scheme of the apparatus. from how I understand, you can consider infinitely rigid support, and impose unilateral contact relationships between the inserts in pe and the metal part. I imagine that these pe inserts will be stuck or strung in the metal matrix, but in any case I don't think you can consider a bond between metal and polymer. I believe that at the metal-polymer interface the only correct relationship is a unilateral contact, possibly with friction.

In general, in these cases it starts from a very simplified model, and once the problem is developed at coeff level. of friction and fineness of mesh, you pass to real geometry.

I don't know if the nx fem treats big shifts, do you have to simulate wear?

 

[matteo.dardari]

Hello hunter,
I thank you again for the availability, for the moment I solved the problem of mesh and if I can get a good picture in the next few days I attach a file to make clear the explanation of the problem. the outside of my liner has a double conicity is shaped like a pyramid for the inacast in the cup so it is not meshable from the ester towards the inteer using the thickness.
Today I tried to simulate a model I got, it seemed to be all right but approximately at the end of the simulation gave me two mistakes that I did not understand well!!! I don't know if you're nx experts but if you could help me there would be very grateful!! !

the two mistakes are these:

_
for statistics decomposition of matrix kllcom .
nastran msg: user fatal message 9137 (phase1d)
^^ run ended two to overdose pivot ratios in matrix kll.