Finite Element Analysis is cool.

[AtomicLabMonkey]

Check it out; I've been playing around with rollcage models in a FEA program, and it's pretty sweet. You can apply loads to the structure and not only see stress levels in all the tubes, but it calculates displacement - so you can see how the cage will flex under your given load as you add & subtract different bars from the structure. It can even animate it for you.

[AtomicLabMonkey]

And here's a picture of how it illustrates displacement... you get a numerical output too of course, so you can see exact numbers as well.

[tysonK]

ok that's pretty cool

are those models for your stang?!?!

[AtomicLabMonkey]

Originally Posted by tysonK

are those models for your stang?!?!

I wish. Maybe someday it will have a full cage like that. It might have a more bare-bones 6-pt. bar in its future though...

[AtomicLabMonkey]

Sweet...

[Dean]

Anybody haev good pictures of Gary's cage? I'm really leaning towards 6 points, but with a front hoop.

[sperry]

Originally Posted by Dean

Anybody haev good pictures of Gary's cage? I'm really leaning towards 6 points, but with a front hoop.

I can't find any great pictures but here's what I got at the last Sears Point USTCC race I went to:

[img:e55f9a35d6=600]http://www.seccs.org/gallery/Race%20Events/2003-10-26%20NASA%20USTCC%20Race%20at%20Sears%20Point/USTCC%20WRX%20Before%20the%20race%206.jpg[/img:e55f9a35d6]

And here's one from Reno-Fernley:

[img:e55f9a35d6=600]http://www.seccs.org/gallery/Race%20Events/2003-08-09%20NASA%20USTCC%20Race%20at%20Reno-Fernley%20Raceway/2003-08-09%20Day%201/USTCC%20WRX%20Interior.jpg[/img:e55f9a35d6]

[Dean]

Wo, that looks completely different than the interior on his web site...

http://www.teamsmr.com/html/car.html

The web page cage does not penetrate the dash like the one in your pictures amungst other things...

[sperry]

Originally Posted by Dean

Wo, that looks completely different than the interior on his web site...

http://www.teamsmr.com/html/car.html

The web page cage does not penetrate the dash like the one in your pictures amungst other things...

Those look like early pics of the car... it looks much better now!

[AtomicLabMonkey]

Solid model of a hypothetical adjustable swaybar... I'm developing some mad sk11lz here...

[Kevin M]

That's a really good idea, but it's missing one small detail- how do you ensure that both sides are exactly the same? Graduated markings, or maybe ball-bearing detents?

[AtomicLabMonkey]

Originally Posted by BAN SUVS

That's a really good idea, but it's missing one small detail- how do you ensure that both sides are exactly the same? Graduated markings, or maybe ball-bearing detents?

Actually that's not an original idea, it's fairly common on racecars. If I was using one, I'd just measure the clamp position from the end of the bar; like you said, if you wanted you could scribe markings on the bar itself for quick changes.

[AtomicLabMonkey]

Originally Posted by Dean

Anybody haev good pictures of Gary's cage? I'm really leaning towards 6 points, but with a front hoop.

That's what I would recommend, if you can afford it. The FEA analysis I've done for different cage types clearly shows trends on how they affect chassis stiffness. In a frontal torsional load case (like I showed above), even going from a regular 6-point bar to a 6-point with full upper structure with A-pillar bars results in less than half as much chassis twist/displacement. Also get a good, experienced cage builder to do it who will tie the cage into the unibody at as many places as possible or allowed.