Our self supporting community for image based modelling

[Lee Margetts]

In just two months, our community has grown to have more than 30
members!

Our small community is already very diverse, comprising students,
academics, technicians and professionals in industry. Our backgrounds
include Materials Science, Engineering, Geology, Medicine,
Biotechnology, Computing and Palaeontology. This is difficult to
appreciate as very few members have added even basic details to their
personal profiles! Luckily, I know many of you myself and can point
this out.

One of the strengths of this forum is that it allows us to build a
self supporting "virtual community". As well as hosting debates, the
Google Group technology allows us to share useful documents and create
new ones online. This is excellent for research, teaching and
learning.

At Manchester, there's been a little debate about which package to use
for creating finite element meshes from tomography images (we have
Avizo and Simpleware). I've therefore uploaded a document highlighting
some differences between the two packages (for meshing). I hope you
find it useful! If you would like to comment on it, please do.

[Andreas Herchenhan]

As a new mamber, I would like to introduce myself here.

I´m Andreas, a german MSc student in Biomechanics at the UoM. My
background is a BSc in Biology, which I completed at the RWTH Aachen
(Germany).
At the moment, I am working on a project about tendon. Our lab,
headed by Karl Kadler, developed a method to produce a tendon-like
tissue from embryonic chick tendon cells. These constructs show
astonishing simillar mechanical properties to embryonic chick tendon.

To establish the conection to image based modelling, we are very
interested in building a FE-model to discover more about the force
distribution on the ultrastructure-level in our constructs.
Therefore, I am very grateful for any advice or suggestions. Of
course, I will be more than happy to answer any questions about this
for a better understanding.

My personal interest is to learn as much as possible about image based
modelling as an opportunity to discover biological tissues and
structures in a "new way".

Best regards,
Andreas

[Lee Margetts]

 

The document I posted originally seems to have disappeared. I've copied and pasted the text below. It may be completely out of date now as I have not revisited this theme in a while. Happy to hear opinions from others.

 

Meshing: Avizo only has a Delaunay mesher- Simpleware offers choice  of Delaunay or Grid based meshing (have different advantages). Grid based is 100% robust and fast which means that for the large complex models typically obtained from micro-ct data you get meshes of good quality, very quickly and that work without any correcting- in contrast Delaunay meshing on complex domains can take, hours or even days and *may not* result in a viable mesh.

 

Multiparts: If you notice none of the 'parts' they segment touch one  another - This is critical for any applications with multiple domains (typical of materials, and indeed many, image based meshing  FE /CFD applications). You can do it in Avizo but really not robust- ie mesh that comes out frequently poor or just won't run.

 

Smoothing: Simpleware smoothing and ati-aliasing is topology and volume preserving and user controlled-even when dealing with multipart smoothing (junctions between 3 or more domains). Avizo can  not handle this Surface/volume reconstruction: Simpleware much more  robust and accurate at surface and volume meshing (better quality surface triangulation.

Material properties: Can't assign material properties based on  background (e.g. Bone or foam or two phase micture properties based on Hounsfield values)

 

Contact surfaces: Can't define boundaries between different domains as contact pairs (e.g. to model slip between phases, femoral head and cup..)

Shell elements: Can't define shell elements - (e.g. on the surface of a volume to model a coating or instead of the volume to model rigid inclusion or boundary.

Nodesets: Can't export nodesets