Author Topic: Lobster Model Advice  (Read 1909 times)

0 Members and 1 Guest are viewing this topic.

Offline SulaTopic starter

  • Beginner
  • *
  • Posts: 1
  • Helpful? 0
Lobster Model Advice
« on: October 05, 2007, 08:22:02 AM »
Hi, I am a marine biologist working on the effects of handling on caught lobsters. I am trying to make a model lobster to test how much damage is being caused by specific procedures on board the boat. The most important features I want to test are lateral compression of the head/carapace and strain on the joint between the head and tail, and the joints between the head and claws. I don't really know too much about this kind of thing but am hoping someone here may have some ideas of the parts I could use - I've contacted a couple of companies but after showing initial interest in this crazy idea, they've all backed off without giving any real help. I don't know if this means that such a thing cannot be done, or they just don't know how to do it??

Any advice that anyone can give would be really gratefully appreciated!

S

Offline Admin

  • Administrator
  • Supreme Robot
  • *****
  • Posts: 11,653
  • Helpful? 169
    • Society of Robots
Re: Lobster Model Advice
« Reply #1 on: October 05, 2007, 10:43:54 AM »
I think the main problem with the companies is return of investment. Engineers are expensive . . . You probably just didnt offer enough $$

So if I was to tackle the problem, I would try two different methods (both of them).

First, I would run a stress analysis simulation on a computer model. Developing the model will be difficult . . . you will need to know things such as the material properties of the lobster - for example, the elastic modulus, hardness, etc. This will require what is called a DMA - dynamic material analysis. I believe there are companies/universities that will do this for you if you supply them with samples.

Then plug those values into your computer simulation.

This is an example of a computational stress analysis I ran on my robot:
http://www.societyofrobots.com/robotforum/index.php?topic=1218.msg8215#msg8215

The second method would be to build an actual model. Probably the best way to do this is with 3D rapid prototyping. You will need to make your own computer model, then send the files to a company to print it. Will probably cost ~$200.
This might help you some:
http://www.societyofrobots.com/robotforum/index.php?topic=461.0

You should embed force sensors into your printed model:
http://www.societyofrobots.com/sensors_forcetorque.shtml

hope this helps some . . .

Offline jsmoker

  • Full Member
  • ***
  • Posts: 118
  • Helpful? 2
Re: Lobster Model Advice
« Reply #2 on: October 05, 2007, 10:46:02 AM »
One solution is the strain gauge.
It's a small paper-thin device about 1/2 by 1/4 inch.  Superglue (or strain gauge glue if you want to go expensive) to any surface and you can measure the strain along the surface.  If you know the elasticity of the surface, you can also get the stress applied.  you'll need to make a Wheatstone Bridge circuit to get a signal though.  Just google the term and you'll find out how.   

here's some info:
http://www.omega.com/literature/transactions/volume3/strain.html

Here a site where you can get it as well as look into other strain sensors:
http://www.omega.com/pressure/psc.html

I have to admit, what you're trying to do is gonna be pretty tough though.  You'll probably also need to model the lobster using some kind of finite element procedure.  Also, replicating biological joints of that kind I'd say is going to be tough as well.  And as mentioned before you'll need to recreate the elasticity (ie the bendy-ness if you want to get technical) of the shell.

Hope this helps, good luck!

Offline Admin

  • Administrator
  • Supreme Robot
  • *****
  • Posts: 11,653
  • Helpful? 169
    • Society of Robots
Re: Lobster Model Advice
« Reply #3 on: October 05, 2007, 10:47:32 AM »
Oh and I forgot to mention, I work on building a robot fish. So I do lots of stuff similar to what you are asking for.

Check the science papers listed here:
http://www.societyofrobots.com/robotforum/index.php?topic=889.0

It could give you ideas.

If my above method sounds too complicated, jsmoker's idea will work too. You just attach the sensors to a dead lobster and data log it. But using strain gauges themselves can be a bit tricky, too . . .
« Last Edit: October 05, 2007, 10:50:00 AM by Admin »

 


Get Your Ad Here

data_list