Hi. I am currently a biology graduate student at the University of Mississippi. I am interested in using CFD to answer some questions for a potential thesis project regarding biomechanics; more specifically, biomechanics of bodies moving through water (marine fish and their body parts). Here's my first question:

How difficult would it be for someone totally new (like myself) to generate a 3d model of some species of fish (based on morphometric measurements, etc.) and then model that body's hydrodynamic properties (i.e. dead drag) as it moves through an incompressible fluid (such as water) using some CFD software?

Is this just crazy or is this how this stuff works? What software programs would such a feat require? Finally, please let me know if there's any easy way to do this using programs you know of (and suggest a few if you don't mind). Any help is greatly appreciated. Thanks!

Matt Gaylord Biology Dept. University Of Mississippi

If the fish is dead (can't tell from your post), measurements of its shape would I think be a matter of developing a system of coordinates describing the fish surface. That could be used in several of the commercial codes to calculate a drag at various water velocities. I don't know if that's what you're wanting.

To model a swimming fish fits into a category usually called 'fluid structure interaction'. You'd somehow have to describe the motion of every point on the surface of the fish as a function of time and feed that into the simulation code. I know these are more difficult computational problems, but I don't know enough to make suggestions for a code to do the job.

If you're doing an advanced degree, will your advisor expect you to understand the underlying math and physics? Without that, CFD is just numbers and interpreting them is the heart of the work. I'd think you should count on taking some advanced courses not in the biology curriculum.

Likely someone has looked at this problem and published something. Maybe you'll get a better answer from others.

Good luck!

Thanks for the input. I have been looking into this further today and have learned that a laser scanning platform may be what I need to generate a 3d model. The fish would be dead. actually, I was considering casting a plaster model of the fish to standardize for differences in body surface across individual sample subjects. The numbers I am ultimately interested in would be those associated with lift and drag (assuming completely laminar flow). These I would then use to extrapolate regarding differences in energy budget for fish with slightly different morphological features. In fact, because the primary difference this project would be concerned with is the influence of head shape on drag etc., the "body" of the fish could just be some generic shape such as a cylinder. The overall focus of the first phase of this project would just be to get at the relative differences across different species. Then later on to tackle and quantify what implications these differences may have energetically for the fish throughout their life history, etc.

Thanks again for the help. I still think at this point that CFD may actually be what I need, but any further advice/comments are appreciated. Thanks.

Matt Gaylord Biology Dept. University Of Mississippi

Interesting project!

We seem to agree that dead fish are easier to compute than live fish. I think the calculations you describe are quite doable, probably several of the commercial codes will allow the meshing required.

Again, I don't have the experience to make recommendations for specific software. Quick thought: is a fish symmetric about a vertical plane through its tail and backbone? [answer yes and cut the problem size in half].

PS: Sounds as if you have the math and physics needed in hand already. Not usual for a life scientist but commendable! (We'll not discuss my 'grasp' of anatomy here).

Again, good luck!

Matt take a look at

ftp://ftp.adapco.com/pub/outgoing/hammerloop.gif

It was done with a combination of star-ccm+ and star-cd4. The meshing took 1 day starting from an STL file. The morphing took another day. Its not trivial, but far from impossible. Although you can do it, I don't believe you need Fluid-Structure Interaction for a first approximation. You can assume to start that the fish is just a solid body that moves and is not affected by the water.

"You can assume to start that the fish is just a solid body that moves and is not affected by the water."

That would certainly be the assumption we're planning to work from. Virtually everything will be held constant (flow will be assumed entirely laminar, the actual body of the fish will be some constant, generic form, flow speed, etc. will all be held constant). The only variable will be head shape. The animated image you linked to was actually a quite prescient one as well. We are actually planning to look at differences in shark head morphology (i.e. various hammerhead species and a few carcharinid, or more torpedo-shaped species). The idea would be to determine what makes the hammerhead special in terms of its hydrodynamic properties. I'm hoping to elucidate a bit more regarding it's evolution, adaptive significance, or even perhaps uncover some maladaptive properties (i.e. increased drag for the animal) to better understand why it arose.