CFD analysis of a starfish?

  • Thread starter Physalia
  • Start date
  • #1
4
0
Hi all,
I am working on a project in where I have to find drag and lift forces on a 3D rendering of a starfish. See below for a picture of the organism in question.

http://www.piscoweb.org/files/image...onitoring/Rocky-Shore-Monitoring/pisaster.jpg

This particular species can withstand breaking wave velocities of up to 20 m/s and very high Reynolds numbers. I will attempt do this using a CFD analysis on a 3D rendering of the star. My problem is that I am not familiar with this type of analysis at all and am wondering what the most straightforward way of doing this is? Also does anyone know if it would be possible to pay for someone to do this for me, or is this quite expensive?

Thanks
 

Answers and Replies

  • #2
Danger
Gold Member
9,647
251
Welcome to PF, Physailia.
I don't really know anything about hydrodynamics, but the basis of your question confuses me. Do starfish not cling to the bottom and walk along, rather than swim?
 
  • #4
Danger
Gold Member
9,647
251
Wow! That's a wicked appendage. I wouldn't know whether to use it to groom my cat or to treat my girlfriend.
 
  • #5
boneh3ad
Science Advisor
Insights Author
Gold Member
3,279
958
Starfish also swim using a sort of water jet system. It is actually quite neat.
 
  • #6
Danger
Gold Member
9,647
251
Starfish also swim using a sort of water jet system. It is actually quite neat.

I was aware of that for mollusks such as squid, octopi, and clams, but didn't realize that starfish had the capability. That's cool.
 
  • #7
AlephZero
Science Advisor
Homework Helper
7,002
293
There is commercial CFD software available (and probably some freeware as well) that should be able to handle this size of problem on a PC. FLUENT is a reputable commercial program.

I'm not a CFD expert, but the obvious problem is that unlike typical CFD applications (aircraft, IC engines, etc) starfish don't have a smooth surface. I think you need some expert advice on how to include that in your analysis, without trying to model the shape of every individual "wart".
 
  • #8
minger
Science Advisor
1,496
2
You're answer can only be as good as your boundary conditions and assumptions. What condition are you going to be running this simulation? What position will the starfish be? What flow conditions? Are these real? What are the actual flow conditions the starfish will see?

I have seen CFD results of swimming fish. That makes sense to me, although I really don't understand why you're attempting to do this for a starfish.
 
  • #9
Danger
Gold Member
9,647
251
There is commercial CFD software available (and probably some freeware as well) that should be able to handle this size of problem on a PC. FLUENT is a reputable commercial program.
Is that available for computers, or only Windows-burners? I'd be glad to obtain one if it's available in a Leopard or Tiger version.
 
  • #10
boneh3ad
Science Advisor
Insights Author
Gold Member
3,279
958
You will be hard pressed to find much of any engineering software that runs on Mac. It is mostly Windows or Linux.

On top of that, Fluent is incredibly expensive. I don't know. This seems like a ridiculous project to be assigned in class unless the school has fluent available or unless the students are expected to simplify the problem to make it hand-solvable.
 
Last edited:
  • #11
Danger
Gold Member
9,647
251
It is mostly Windows or Linux.
I have no experience with, nor knowledge of, Linux other than learning from others on this site that it a superior OS. I rather liked DOS III, but beyond that my experience has been with only dozens of variations of Windows and Mac OS.
I hate Windows with a burning passion. That is based upon experience, not prejudice.
 
  • #12
boneh3ad
Science Advisor
Insights Author
Gold Member
3,279
958
Regardless, a lot of software just doesn't run on Macs simply because it is too uncommon in the business world. Most of the major codes I am familiar with work on Windows and Linux, and most of the actual calculations are done on Linux systems due to the extremely low overhead and stability (and other factors of course).

Just for the record, the base price for a single-seat Fluent license is typically on the order of $20,000 if I remember right. I think Fluent also has a service where you can submit jobs online and they charge you a few bucks per CPU hour for them to run your simulation for you.
 
  • #13
4
0
Starfish also swim using a sort of water jet system. It is actually quite neat.

I was aware of that for mollusks such as squid, octopi, and clams, but didn't realize that starfish had the capability. That's cool.

They don't actually have the ability to swim. Unless you count when they're blown of the rocks by a wave.
 
  • #14
4
0
There is commercial CFD software available (and probably some freeware as well) that should be able to handle this size of problem on a PC. FLUENT is a reputable commercial program.

I'm not a CFD expert, but the obvious problem is that unlike typical CFD applications (aircraft, IC engines, etc) starfish don't have a smooth surface. I think you need some expert advice on how to include that in your analysis, without trying to model the shape of every individual "wart".

You're answer can only be as good as your boundary conditions and assumptions. What condition are you going to be running this simulation? What position will the starfish be? What flow conditions? Are these real? What are the actual flow conditions the starfish will see?

I have seen CFD results of swimming fish. That makes sense to me, although I really don't understand why you're attempting to do this for a starfish.

The Starfish will be attached to a flat surface and experiencing uniform flow horizontally across its body.

I have some estimations of boundary layer thickness and I understand that I may have to make some assumptions of skin friction and surface roughness. Although I am using this to compare two very different morphs found in a single species to test their differential adaptation to flow. I simply want to test for hydrodynamic properties of the two.

I will look at FLUENT thanks.
 
  • #15
Mech_Engineer
Science Advisor
Gold Member
2,575
172
It could be interesting to see the difference in required suction to stick to a rock in different flows given different skin roughnesses.

Look into ANSYS CFX as well, hopefully you can get a student license though because CFX or FLUENT (or both) are quite expensive. We use ANSYS Multiphysics at my work, which includes CFX.

Note the difference between the two solvers:

ANSYS.com said:
The ANSYS CFX solver uses finite elements (cell vertex numerics),
similar to those used in structural analysis, to discretize the domain. In
contrast, the ANSYS FLUENT solver uses finite volumes (cell-centered
numerics). Ultimately, though, both approaches form control volume
equations that ensure exact conservation of flow quantities, a vital
property for accurate CFD simulations.

http://www.ansys.com/products/fluid-dynamics/default.asp [Broken]
 
Last edited by a moderator:
  • #16
Danger
Gold Member
9,647
251
Just for the record, the base price for a single-seat Fluent license is typically on the order of $20,000 if I remember right.
Crap! That's over 2 years of income for me. :grumpy:


I have some estimations of boundary layer thickness and I understand that I may have to make some assumptions of skin friction and surface roughness.

That's one that would hopelessly confuse me. Those are pimply little bastards. I can't begin to think of whether that would result in ridiculous turbulence or a laminar flow. I'll be very interested to see what you come up with on that.

edit: Sorry, Mech. I wasn't ignoring you; you sneaked in while I was composing. I'm sure that your link will be great for anyone who can understand it. Unfortunately, that club does not include me.
 
  • #17
Mech_Engineer
Science Advisor
Gold Member
2,575
172
That's one that would hopelessly confuse me. Those are pimply little bastards. I can't begin to think of whether that would result in ridiculous turbulence or a laminar flow. I'll be very interested to see what you come up with on that.

He should be able to simplify the starfish's "pimply skin" with a surface roughness coefficient. This should allow his to vary the roughness with the same model, looking at different effects.

I'm quite sure it will end up being turbulent flow, but there's a lot to a CFD problem. Hell there are 3 or 4 different ways to model the turbulent flow in the CFD solver! I hope the OP knows what he's getting himself into...
 
  • #18
Danger
Gold Member
9,647
251
I hope the OP knows what he's getting himself into...

It probably won't be pretty, but I'm pretty sure that it will be interesting.
 
  • #19
183
1
I was aware of that for mollusks such as squid, octopi, and clams, but didn't realize that starfish had the capability. That's cool.

I think that's a misnomer. Their locomotion is indeed hydraulic, but via internal pathways, not external jets. http://en.wikipedia.org/wiki/Starfish#Locomotion".

As for modeling the skin, I think the best approach would be to test actual starfish, then modify roughness parameters until the results match. However, once you've tested actuals, you have the data, and since you're not actually designing starfish, why proceed with a modeling program?
 
Last edited by a moderator:
  • #21
boneh3ad
Science Advisor
Insights Author
Gold Member
3,279
958
Surface roughness aside, it may still be laminar given the low speeds in the ocean and the small size of the starfish. There isn't really a way to know without testing or DNS though.
 
  • #22
Physalia said:
The Starfish will be attached to a flat surface and experiencing uniform flow horizontally across its body.

I have some estimations of boundary layer thickness and I understand that I may have to make some assumptions of skin friction and surface roughness. Although I am using this to compare two very different morphs found in a single species to test their differential adaptation to flow. I simply want to test for hydrodynamic properties of the two.

I will look at FLUENT thanks.

Even if you are only interested in a comparison between two competing designs, I strongly recommend you to go for a full scale test, specially because it is possible to do. This will give much more substance to your research and will probably save you time and criticism.
 
Top