ZOMG!!! Wooo! My project works!

[Fenixgoon]

Link (youtube)

So for my numerical methods class, my group project is simulating the impact of a motorcycle helmet (I kinda chose that for the group ). I'm the primary coder, so I've spent a hell of a lot of time trying to figure out how to get things work. Today we finally finished our "first approximation" model.

There are a lot of improvements to make, but I'm just excited we've gotten this far

If anyone wants the gruesome details I'll list them, otherwise I will spare ATOT

We used ABAQUS, which is a commercial finite element program.

 

[Fritzo]

It looks like an MRI of a fat guy and his penis is showing.

 

[halik]

That's impressive, I don't even wanna think how much physics went into the thing. What school?

 

[Jeff7]

What software? ANSYS?

Project for Vibrations: A plastic car model, impacting with a wall. The professor says the solve times for ANSYS could be around an hour, or closer to 5 hours if the car starts spinning after impact.

For Finite Element Analysis II, we simulated a ball hitting a beam after falling 100 inches. One setting caused a problem: The ball would hit the beam, bounce twice more, and then launch at high speed nearly 1800 inches high. Much weirdness.

 

[Fenixgoon]

Originally posted by: Jeff7
What software? ANSYS?

Project for Vibrations: A plastic car model, impacting with a wall. The professor says the solve times for ANSYS could be around an hour, or closer to 5 hours if the car starts spinning after impact.

ABAQUS.

Vibrations sound pretty messy, if my friend's vibrations class midterm was any indication. I'm sticking to stress and strain

Originally posted by: halik
That's impressive, I don't even wanna think how much physics went into the thing. What school?

Drexel University, Materials Engineering.

The hardest part is getting the coding down. All of the physics are pre-defined in ABAQUS and you just use the different "tools" (keywords) to tell it what you want to happen.

 

[Jeff7]

Ewww, that software sounds nasty. Coding? Icky.

There are some settings to stash into ANSYS, or Workbench, but not really any "coding."
It's certainly got its share of quirks though.

 

[Epic Fail]

You got waffle owned in the comments.

 

[KillerCharlie]

When I was an aerospace engineering student we had a lot of required numerical methods classes (more than any other major), and I took a few electives. I absolutely loved writing CFD code. I was about to go to grad school for CFD when I realized I didn't want to do it for a living. Using existing codes is fun, but not as fulfilling as writing your own. It made me go slightly crazy though.

 

[vshah]

ahh nice, its such a good feeling when shit finally starts working. Don't get cocky now though. After a big breakthrough I'd always think of 5 more nifty features I wanted to add...

 

[Fenixgoon]

Originally posted by: vshah
ahh nice, its such a good feeling when shit finally starts working. Don't get cocky now though. After a big breakthrough I'd always think of 5 more nifty features I wanted to add...

yeah, like brittle fracture modes

requires experimental data though

 

[TallBill]

Originally posted by: Epic Fail
You got waffle owned in the comments.

Who's that robertthebob jerk?

 

[Born2bwire]

Man, I had to write my own FEM, MOM, and FDTD codes...

 

[Fenixgoon]

Originally posted by: Born2bwire
Man, I had to write my own FEM, MOM, and FDTD codes...

you mean like defining the physics of the systems? that would suck


FDTD isn't too bad though. but yeah, doing complex FEM from scratch would be killer

 

[Ballatician]

Very nice. I used ABAQUS when modeling an improved spinal implant but wish I knew how to use the program better.

 

[Born2bwire]

Originally posted by: Fenixgoon

Originally posted by: Born2bwire
Man, I had to write my own FEM, MOM, and FDTD codes...

you mean like defining the physics of the systems? that would suck


FDTD isn't too bad though. but yeah, doing complex FEM from scratch would be killer

Yeah. My FEM project was an electromagnetic simulation using vector basis functions. I decided to model a vertical Hertzian dipole antenna (not something that can be modeled very well via FEM) above a layered medium which made it easier since the problem can be a 2D simulations since it is axially symmetric.

EDIT: The best thing about all these projects are the pretty pictures and videos you get out of them. Nothing makes a paper or presentation that much more interesting than visuals.

 

[Fenixgoon]

Originally posted by: Born2bwire

Originally posted by: Fenixgoon

Originally posted by: Born2bwire
Man, I had to write my own FEM, MOM, and FDTD codes...

you mean like defining the physics of the systems? that would suck


FDTD isn't too bad though. but yeah, doing complex FEM from scratch would be killer

Yeah. My FEM project was an electromagnetic simulation using vector basis functions. I decided to model a vertical Hertzian dipole antenna (not something that can be modeled very well via FEM) above a layered medium which made it easier since the problem can be a 2D simulations since it is axially symmetric.

EDIT: The best thing about all these projects are the pretty pictures and videos you get out of them. Nothing makes a paper or presentation that much more interesting than visuals.

sounds really complicated E&M was the part of physics i hated most, lol

 

[Howard]

Originally posted by: Jeff7
Ewww, that software sounds nasty. Coding? Icky.

There are some settings to stash into ANSYS, or Workbench, but not really any "coding."
It's certainly got its share of quirks though.

FUCK YOU ANSYS

 

[zinfamous]

Originally posted by: Epic Fail
You got waffle owned in the comments.

 

[Fenixgoon]

Originally posted by: zinfamous

Originally posted by: Epic Fail
You got waffle owned in the comments.

as i replied, i for one, welcome our new breakfast-food overlords

 

[Dirigible]

That brings me back. My first finite element project about fifteen years ago was to model the impact of a piece of wood with a helmet that loggers use. I think I even used ABAQUS. I could be wrong, but it sure sounds familiar.