Modelling a stent

[Joanne]

Hi all Modelling experts!!
<?:namespace prefix = o ns = "urn:schemas-microsoft-comfficeffice" />
I

 

[Israr]

First of all you have placed your question in the wrong sub-forum You should have placed it in analysis sub-forum.


I have checked your model. First you should have use shell technology for your assembly and then create volume regions for load applications.


Israr

 

[james.lynch]

Hi Israr,


could you expand a little on this? would you onlyuse Shell elements to simulate the plaque or would you also useshells for the stent (some typeof beam element perhaps?)


Also, is it not possible to simulate it using solids? granted shells shouldreduce processing time..


Regards,


James

 

[Joanne]

Israr,


Thanks for your reply - i took your advice about the other forum and have posted it there too


I dont understand what you mean about about the shell technology. do you think i should be using shell elements only? could you explain how i should incorporate volume regions?


thanks a lot for your advice


Joanne

 

[Brian_Adkins]

Moderator note:Moved posting to correct forum anddeleted duplicate posting.


-Brian

 

[Joanne]

Thanks to all of you who helped me with my problem on my last post. I got rid of the "Confirmation" message which previously deleted the load, by using the FACE/SURFACE tool to select the exact contact regions. However it stills appears when I constrain the stent... but the simulation runs now anyway.


I am still having great difficulty creating a mesh fine enough to study the contact area in question (inside curved face of cylinder in contact with metal struts of stent)

. I thought I had a breakthrough earlier today when I was messing around with the model in Mechanica FEM mode. When I held the max element size to .1mm on the curved surface, it created a "beautiful" mesh - exactly what i have been dreaming about for the past fews months...<?:namespace prefix = v ns = "urn:schemas-microsoft-com:vml" /> </v:stroke></v:f></v:f></v:f></v:f></v:f></v:f></v:f></v:f></v:f></v:f></v:f></v:f></v:ulas></vath><?:namespace prefix = o ns = "urn:schemas-microsoft-comfficeffice" /><o:lock aspectratio="t" v:ext="edit"></o:lock></v:shape></v:stroke></v:f></v:f></v:f></v:f></v:f></v:f></v:f></v:f></v:f></v:f></v:f></v:f></v:ulas></vath><o:lock aspectratio="t" v:ext="edit"></o:lock></v:shape></v:stroke></v:f></v:f></v:f></v:f></v:f></v:f></v:f></v:f></v:f></v:f></v:f></v:f></v:ulas></vath><o:lock aspectratio="t" v:ext="edit"></o:lock></v:shape>
But the dream sooon faded when I realised, from looking through this forum, that the meshing is not interchangeable between FEM mode and Native because one uses h-elements and the other uses p-elements.

I have tried creating Volume regions (Local Mesh Refinement with SPA is no good). This has helped to some degree but I read in another post that
"the element edges must have a length less than half of the length of the contact line (or smaller)" - and my contact area is tiny (around 0.01mm) so I'm not sure if Volume regions will help me.

My question is:
Are such small elements really necessary if all I want to get is the Maximum Normal Contact Force on the cylinder due to a radial force from the metal stent and a good representation plot of the stress distribution?
Also any suggestions for achieving small elements efficiently on this curved surface (i.e. pattern??) would be greatly appriciated!



I have upload the latest model (suppressed items and converted for commercial access) here with some volume regions set up.


2005-09-30_150134_Zip_con_5A2.ZIP


At this stage I'm not sure if Mechanica can do the analysis for me and think I might have to go to another package... I hope I'm wrong...

 

[Israr]

Use autoGEM control to refine youe mesh.


and tell you that Pro/Mechanica is powerful enough to solve your problem. It will be your incompetency that you can't do it.


By now I don't have my computer nearby, so will download your part and fix it.


Israr

 

[Joanne]

Israr,


Thanks for restoring my faith in MECHANICA. I have only started using it during the summer and I wasn't aware of how powerful it is. I'm looking forward to the fixed model.


I have tried using the Autogem control to refine the mesh myself but without success. The edge distribution method it only created very "long" elements.


Meanwhile I have created volume regions in the cylinder part (Plaque) now using extrusion methodand I am happy with the mesh density that this has acheived. However when I try to create the mesh in the assembly this message comes up:
"Could not complete creating elements on all of the selected entities"
the selected entities are the outer cylinder part as far as i can see.


I have tried seeding points and creating more coarse volume regions in the outer cylinder part (Artery) because I thought the reason for the error was that the mesh wasn't fine enough at their mating surfaces. however this didn't help.


I have attached the model here 2005-10-01_115217_G24-17_zip_con.ZIP
(zipped and with many items suppressed to reduce file size)


Thanks again for the reply
Joanne

 

[Joanne]

Israr,


I have been working on the model since the last post. The constraints are only on the cut surfaces of the artery and plaque - fixed normal to the cutting plane only. When I try to constrain thestent in the same waythe followingmessage stillappears
<BLOCKQUOTE dir=ltr style="MARGIN-RIGHT: 0px">


Confirmation:
The highlighted loads or constraints will be deleted
For one or more of the following reasons:
- on a face or edge that was split,
- on a face or edge that was merged,
- on a face that was collapsed to an edge during compression,
- pressure load on multiple surfaces having inconsistent normals (to fix, create separate loads for each surface)
- constraints on two edges or faces that collapsed to the same edge or face during compression. </BLOCKQUOTE>
on clicking "confirm" the stent constraint ia deleted.
The SPA simulation completed, however there are many warnings relating to the contact area "contact area is small compared to size of adjacent elements" - this is whatprompted me to refine the mesh.
Most worryingly, the maximumdisplacement magnitude is96m! the model has outside radius of only5.5mm.
At this stage I just want a reasonable result - I am no longer worried about mesh refinement just as long as the solution converges, i don't think i need it.
Could you take a look at this latest modelI have attached please?2005-10-04_141847_G24-17_con-sup_zip.zip

 

[Joanne]

Hi everyone,


I have managed to get the model running and using SPA with Local Mesh Refinement has indeed given me a very good plot of the inside suface of the plaque showing the stresses caused by the stent - but it's still not right.


Since my previous posting I have made a few changes to the stent. I've added rigid connections between each of the wires in the stent and can only manage to constrain the edges of the cut surfaces of wire normal to their cut surfaces.


However, running the simulation with quick check initially gives me a max displacement of over 100mm - the outside diameter of the artery is only 11mm! I am currently running it with SPA LMR


What I'm asking for at this stage is for somebody just to look over my model to check that the connections, constraints and loads are selected correctly.


2005-10-07_101319_zip_supcom_2417_RC.zip


thanks a lot


Joanne

 

[bem]

What version of pro are you using?