Material assignment for cortical and cancellous bones seperately
Hi all,
I have two relations for material property assignment, one for cortical and another one for cancellous bone (Am modeling a humerus). The problem is i made two separate masks for cortical and cancellous bone. Then did 'unite' boolean to make a new mask and developed 3d mesh. When i assign materials i chose 'limit to mask' option and chose cortical mask and assigned cortical set of eqns and i assigned 10 sets of materials. Next i did the same for cancellous with 'limit to mask' to cancellous mask and i assigned 10 sets of materials.
The issue is i should have 20 sets in total (10 for cortical and 10 for cancellous) but i am not having 20 instead i have 10 only. Can you please guide me how to do separate material assignment for cortical and cancellous bone. thanks.
Dear patricia and all,
The previous problem was solved, ur non-manifold assembly worked fine. Now my issue is i am modeling a humerus and my application is such that i must have the medullary cavity preserved in my humerus FE mesh. When i segment the CT data with the preset CT bone HU valve (and even after tweaking) i get lots of discontinuities on the surface of the bone (especially on the condylar side) so i have to do manual "edit mask" which is very boring. So came up with the following procedure.
1. I generated a mask with the default HU setting and consequently generated a surface mesh using "calculate 3d" and ofcourse the surface mesh had lots of discontinuities as always.
2. I wrapped the whole surface mesh and now i lost the cavity but i gotta nice mesh without discontinuities
3. Now i went reverse , I calculated mask from the 3D with the wrapped surface mesh which gave me a nice mask back in my mask tab.
4.Now i created a seperate mask for the medularry cavity and subtracted this mask from the "reversed" whole bone wrapped surface mesh mask.
5.Now with thus obtained new mask i.e (whole bone minus medullary cavity) i calculated 3D and everything looked good as there were two surface in the diaphysis region. One representing the outer and the other the inner hense preserving the cavity.
Now the issue is when i take it to the remesher (3-matic) and do fixing using fixing wizard in the process 3matic deletes (or unifies) the inner surface for some unknown reasons. Even if i skip some procedures in the fix wizard and try to generate volume mesh the mesher totally neglects the inner surface and fills the entire bone with tet elements without preserving the cavity. On the other hand if i directly generate a 3d mesh from the default set HU values (i.e without goin thru the 5 steps memtioned above) and if i try to generate volume mesh then things are fine.
What is the problem with the procedure? and where i am making the mistake and why 3 matic refuses to preserve the medullary cavity. Thanks for your time.
Added: I did some more experimentation. I xported the surface mesh to Abaqus and converted it into volumetric mesh. No problems ! i gotta nice mesh with cavity preserved. But here the issue is when i import the volumetric mesh from abaqus to mimics the mesh is shifting away from its actual position. So when i press materials command its giving an error stating the mesh is not inside the image area. However i checked the co-ordinates of the surface mesh generated by mimics and the volume mesh generated by abaqus both are exactly same other than the formatting. Problem again !
regards,
mike
Dear Mike,
It is good to know that the previous problem is solved. 1 problem less to worry about. Let's focus now on the new one :-)
The procedure you describe to obtain a smooth surface with the Wrap function is perfectly correct and should not cause any problems. The issue is, when you use the Fix Wizard in the Remesher, it will fix the part taking into account that you want to go to Additive Manufacturing and not to Finite Element Analysis. For this reason, one of the first operations it performs is the removal of all shells except the largest one, in this case, the humerus without the medullary cavity. We do not always recommend the utilization of the Fix Wizard, because in some cases, what is interpreted as being an error, is not. You can unselect the option "Shells" from the list and then it should be fine. In the end you should have 2 volumes, one for the bone and one for the medullary cavity. Just make sure you delete the second one.
You might be wondering why the unwrapped model did not behave the same way. The reason is that in the unwrapped model, there is probably a connection between the medullary cavity and the external surface. Therefore, you have only one shell.
Regarding the second issue, could you provide us with some more information? Which version of Mimics and Abaqus are you using? Have this issue happened more than once, or only this time, for this specific mesh? What are the original images (Axial, Coronal, or Sagittal)?
Looking forward to hearing from you!
Patricia
Dear Patricia,
Thanks for your lightning fast replies !! Not very surprisingly ur strategy worked .. again ofcourse. But one small point i observed is only if i turn OFF the "group by materials" option in the "create vol mesh" i am geting 2 volumetric mesh out of which am deleting the inside one. But if i leave it ON then the "create vol mesh" command itself is deleting the inner shell and results in only one volume! (removed subvols= 2). What "group by materials" means? It is not clear in the documentation also. Is there any harm if i turn OFF "group by materials"?
I hope you dont mind fixing the second issue also. For the second issue am using Mimics 13.1 and Abaqus 6.10 (am using "convert tri to tet" command in edit mesh). I have attached a scrShot of the shift away problem i was mentioning you. It always happens like even if i do the meshing in hypermesh 9.0 then also i face the same problem. Finally i din get ur question "What are the original images (Axial, Coronal, or Sagittal)?". Thanks for ur time.
Regards
Mike
Dear Mike,
Good to know that the previous issues have been solved.
The "group by materials" option of the volume mesh generation is a tool used when you have calculated a non-manifold assembly based on masks. To give you an example, imagine a binary dataset where material A is black and material B is white. When you segment this dataset, you will have a green mask for material A and a yellow mask for material B. When calculating a non-manifold assembly based on masks, you will end up with one surface called Green, one surface called Yellow and a third surface (the interface) called Green-Yellow. In some cases, e.g. if you are dealing with rock samples, each surface is composed out of multiple parts. When you generate a volume mesh, however, you want to have only two volumes: one for material A and one for material B, even if each of these volumes is composed out of multiple rock fragments. This is what the "Group by materials" option does: it groups all the shells belonging to the same surface/material so you end up with only one volume. When you calculate a non-manifold assembly based on Boolean operations (the one present in the Remesher), this option will remove one of the volumes, so it should be kept OFF.
Regarding the positioning of re-imported meshes, would it be possible to install Mimics 14.0 (just released) and verify if the problem persists?
Dear Mike,
If I understand correctly your description, after you have generated a volume mesh, you have obtained only one volume, comprising both the cortical and the trabecular bone. If that is the case, then you will not be able to use the Uniform or the Look-Up file methods as you described. This is because you have united the two masks with a Boolean operation before the calculation of the 3D object. You have two different options:
1. If you continue with one volume, you can use the Mask method and then select the masks you have created for cortical and cancellous bone. In this case, you will end up with two different materials.
2. You can create a non-manifold assembly between the cortical and the cancellous bone before generating the volume mesh. In this case, you will end up with two volumes and then you can assign material properties to each volume separately. This means that you can use the Uniform or the Look-Up File methods with 10 materials each, limited to the corresponding mask. In the end, you should have 20 materials as you expected.
I hope this answers your question!