Jon's 2 year roll cage saga...

[LLave]

When you put in the tube between the cage legs I would add a tube underneath (say 5/8 ID) that allows you to slip in a remote jack point. You'll need to notch the rocker and you can then back this up with plates on the inside of the rocker to the tube and on top of the rocker. If this isn't clear I'll attempt some Cary CAD to make a visual version. Just an idea. I'm going to a go a little farther and do this at the ends too so I can put the car on stands with nothing under it very much like a rally car.

 

Cary

 

I would like to see the Cary CAD visualization of this if its not too much trouble.

[JMortensen]

[sarcasm]I feel much safer now[/sarcasm]

 

 

Onto another subject: I've got a box of about 20 of those little Joe's taco gussets. I'm just wondering if I should take a page from BJ http://forums.hybridz.org/showthread.php?t=109864&page=6 and go crazy with these suckers, or whether I should call it a day. I'm getting ready to do door bars, and once that is done it will be a lot harder to get in there to weld anything, which is why I'm wondering now.

 

I've read a couple things here lately that seem to indicate that it might be worth a couple hours to get them installed.

[proxlamus©]

you have to weight that thing when your done lol

[johnc]

I'm on the fence regarding gussets. Roll cages have worked well for decades without them especially if the tube junctions are a tight fit and welded properly.

[srgunz]

I was going to do a jackpoint straight through the rocker with a plate on either end and a tube through the middle. How would you suggest incorporating a jack point into the tube above the rocker? Seems like the jack would have to go right through the door...

 

 

Jon. I took these two photos in June of this year. This is a EP 240Z from Colorado. It has a jack point on the roll cage as you can see. I do not see a hole in the door from this picture and I cannot remember if it did. It would seem that it did.

 

 

 

 

[thehelix112]

I'm on the fence regarding gussets. Roll cages have worked well for decades without them especially if the tube junctions are a tight fit and welded properly.

 

 

A couple of things spring to mind.

 

I don't pretend to be `in' with the racing crowd, but a look at some professional teams view of gussets might be an indication one way or another.

 

Also, it would seem that having the tubes slightly away from the chassis (pillars are what I'm thinking of primarily) and gussets inbetween provides yet more material to crumple/absorb energy in the event of a crash. Something hits pillar, pillar crumples gusset against tube, tube absorbs the rest. Anything wrong with my train of thought here?

 

I think I'll be using them.

 

Dave

[johnc]

A gusset is probably needed if a bar in a cage is spanning a failry long opening in the chassis. The upper cage bars that span the top of the door opening in a BMW E36 chassis is a good example.

 

Be careful about blindly copying what Pro teams do. They are usually building cars to handle much higer speed impacts then the typical SCCA regional racer encounters. Our cage designs should focus on a 70 to 90 mph impact, not 140 to 160.

[johnc]

Also, it would seem that having the tubes slightly away from the chassis (pillars are what I'm thinking of primarily) and gussets inbetween provides yet more material to crumple/absorb energy in the event of a crash. Something hits pillar, pillar crumples gusset against tube, tube absorbs the rest. Anything wrong with my train of thought here?

 

A roll cage should not be designed to crumple in any way. The rest of the car can and should crumple to the cage structure, but the cage itself is the last resort and should not give. A cage is designed to transfer load through itself to other areas of the vehicle. It should not absorb load by buckling, bending, or compresing.

[thehelix112]

I think I understand what you're saying John, but surely you don't think that any cage, anywhere will not deform if the metal involved is pushed beyond its yield point.

 

I would think it prudent to take the failure of a cage in the event of an impact beyond its intended use into account when designing it no?

 

Dave

[johnc]

I would think it prudent to take the failure of a cage in the event of an impact beyond its intended use into account when designing it no?

 

If a single impact exceeds the design strength of a roll cage (with or without gussets), how do you predict the failure? Given an accurate prediction of cage failure from a single impact, how do you control that failure?

[JMortensen]

Still looking for opinions on the gusset issue...

[JIM73240Z]

i would add the gussets, but john is right. there is no way to predict the damage of a single impact and design around that. if you do, you affect something else in the design. i think the gussets are a good add but not ness.

 

jimbo

[thehelix112]

Hehe.. I have no idea how to do it, but that doesn't mean it shouldn't be done.

 

I would be surprised if there is no way to predict these things. I'm sure there is some FEA software out there somewhere that incorporates elastic deformation, yield point and plastic deformation behaviours.

 

Sorry to sidetrack Jon.

 

Dave

[LLave]

This isn't exactly scientific by any means but if you already have the gussets, I say what the hell weld them in. You aren't talking about any substantial weight gain and it is not that much work. Personally if I had them already I would just burn them in.

[bjhines]

I played around with some misbent tubing sections and a few gussets to figure out how to weld them and how to deal with the warpage they can cause. My dash bar is critical that it fits exactly where it does.

I found that the gussets add considerable stiffness to tubing junctions. It is much harder to cause a complete failure of the gusseted junctions. They also make it much harder to bend the junctions the way they would in a T-bone impact.

The process of welding the gussets can also cause considerable warpage if you rush things and get the area too hot.

 

 

I primarily reinforced the front seciton of the car with 14 wrap around gussets. I used 4 massive 14gauge gussets in the double door bars. I have only a few gusstes in the rear section. I have more tubing junctions and triangulation in the rear. I decided not to use more gusstes in that area.

 

There are serveral dozen other types of gussets throughout the chassis. I was going to take pics of every one like this,

but it got too time consuming.

 

 

 

I used a different style for attaching the front A-pillar bars to the rocker and floor.

[JMortensen]

Thanks John for the pics. I was in the shop starting on the door bars today (got the tube right above the rocker in on both sides) and I just don't see too many places where I can put the gussets. The rear is triangulated 6 ways to Sunday, the dash bar already has large tube gussets in each corner, same at the top of the windshield. About the only thing I came up with was in the V where the front strut bars intersect at the dash, but that area is already tied into the firewall and the cowl, so it just seems redundant.

 

I did install some long shear panels at the base of the A pillar bars. I'll take some pics tomorrow of what's been done, I just don't know where I can really use these gussets. I just don't know if there's any point in gusseting the shoulder bar. It doesn't seem that important from a chassis stiffening perspective, and I'm not worried about the joint in a T bone collision.

[johnc]

In the few roll cage failures I've seen, the tube junction failures (which were rare) were all caused by poor welding. At the Lincoln Electric welding school they had a roll cage out of a Joe Gibbs Winston Cup car on display. Three tube junctions in the driver's side door bar failed from lack of penetration and almost impaled Tony Stewart.

 

Every fabrication team member from Joe Gibbs racing was sent through the two week Lincoln Electric Motorsports welding school after that incident. They also go back periodically for a refresher course.

 

If the tube junction is fit up properly and welded correctly, its unlikely to fail before the joined tubes themselves fail. A gusset won't help in that situation. A gusset can work as additional insurance against a poor welding job, but if the welder screwed up the joint weld, what's to keep him from screwing up the gusset weld?

[JMortensen]

If the tube junction is fit up properly and welded correctly, its unlikely to fail before the joined tubes themselves fail. A gusset won't help in that situation. A gusset can work as additional insurance against a poor welding job, but if the welder screwed up the joint weld, what's to keep him from screwing up the gusset weld?

The reason that I was considering doing all the gusseting was to stiffen the structure, not to prevent joint failure. I'm really not worried about joint failure, I did a really good job on my tube joints and I'm confident in my ability to weld the 1/8" tubing together.

[johnc]

IMHO, FWIW, the stiffness gain from a small taco gusset is more then offset by the additional weight of the gusset, the weld filler, and the extra paint needed to cover it.

[JMortensen]

I think I'm in agreement with you John. Here are some pics of the places that I could put these gussets along with my thoughts:

 

I give this one a maybe. It's easy to access, but the tube to the strut tower that is being gusseted is already a short span.

 

Gusseting the shoulder bar seems like a waste of time, and will interfere with the door bar installation:

 

Putting a tiny taco gusset under the large tube gusset seems retarded:

 

In the V intersection at the dash. I'd need to grind this sucker down quite a bit and by the time you do that, is it really worth it? Plus the V intersects the cowl and the firewall, so its not a long unsupported tube:

 

 

Here's the gusset from the door hinge area to the A pillar tube:

 

 

Bottom door bars are 3/4" above the rockers:

Here's how I welded them in:

 

In the back, the longest span of tubing is the strut tower bar. I don't see a need to gusset any of this: