kurt M

I myself would not start cutting just yet based on this set of models. Gussets have value for resisting loads (EDIT duh) not perpendicular to the gusset flat. The one in the model was a single thickness example being acted upon such that it had no geometric strength. The models did not approach the fail values of the welds as well.
A well made gusset might increase the welded joint value as well. Two of the 4 arms of an X are welded and the flat plate more so if on the opposite side of the load might increase the weld fail load point. Even a somewhat thin taco might act to spread load and prevent a stress riser or load concentration failure.

gbaker

You're capable of doing it, you just don't have the horsepower at your disposal.

Dynamic FEAs are massively complex and require that one pull out the big guns. The software that Wayne State uses to model our product's function during a crash leases for $25,000 per year. When they used it to drive a new dynamics model of the human head, it took a week to solve.

Yeah, you can do it, but we're talking some major juice.

Geo

I'm not so considering removing the gussets because I think they don't do anything (I know they do). I'm thinking about it because I'm thinking of adding three tubes to support the center of the X and I think they will do more. In fact, if I do that I'll probably weld plates over the little triangles they create.

The gussets are not fully welded in just yet.

bruinbro

I'll post an example of higher and lower loads to show you the consistency of the results for one of the designs. You'll have to trust me on the others; it's too much of a pain to post them all.

Remember, the gussets have other purposes. You may not want to scrap them just yet. Look at the head on crash results, especially the shape, not the magnitude of the deformation. This tells me it is a good design for stiffeneng the doorway.

Bro

bruinbro

You are correct, except that a thicker gusset did not change the out of plane performance. My belief is that it helps the in-plane stiffness and would obviously help out the weld joints.

Bro

kurt M

Oops! And you are as well. I left out one word while pecking away this a.m. -not-, should read "not perpendicular"

When will spel cheke be good enough to catch stupid based errors too? Or would that prevent too many internet posts?

M758

It is my feeling that if you really want to see the benefit of the taco gusset on the x take a look at the stress plots. Really if you look at the gusset it won't do much to stiffen the structure in that direction. What it does do however is extend the load and puts less load in the welded joints in the case of an impact. The gusset's greatest asset is when things go non linear or approach yeild.

Geo

Understood.

Again, I'm not so much considering scrapping the gussets for nothing as much as replacing them with additional tubing as I had in a picture in a previous post.

95m3racer

I'm confused by this comment. I've never, ever...ever seen ANY rule set in any series mandate interiorcage structure gusseting or bracing. Its merely part of the acge design, not affecting # of or location of mounting points, which is the only restriction on any cage in all classes (and mounting plate,tubing size, etc). Specifically in ITS i've seen some very complex cages, including "supernodes" tying in every pickup point to the center of the vehicle.

I think you misunderstood what I was stating. Putting a gusset at all bends in the cage is legal in every sanctioning body, it does in no way or form impede on any rules.

These sheet metal or tube-truss "gussets" can provide a significant change in overall rigidty and deformation characteristics.

And, as someone else stated, how much does this actually apply to what is REALLY happening - a very very small fraction of a fraction of a decimal percentage, unfortunately. Again, data can look great on paper (or computer in this case), but this analysis is only good for designing 3d-virtual objects, it does not affect what is actually being designed because there are so many key variables not being accounted for.

Everyone has to realize we're not designing for optimal perfect structure here, we're designing something that we are actually going to be building, and will in real life be struct by who knows what, at who knows what angle, at who knows what force. This is something that is common, testing for something thats not realistic. While it might seem proper assumptions are being made, and in this case VERY intelligent assumptions and great knowledge is being applied, it does not change the fact that, I garuntee, there will be no conclusive result to these test after you factor in the largest REAL variables that we are dealing with.

While someone might say an "X" is safest after these FEA tests, this might not include the information that the test had the bars as X angle with X contact area on each tube, for this specific situation...again, too many important unknowns that are mandatory in determining what would really save someones life.

And as said before, software to accurately AND precisely (they ARE different critieria) perform these tasks is extremely expensive and requires extremely hardcore machineS (plural). I do admire Bro (and everyone else for thier comments/suggestions) for going about this in depth analysis, however, if you guys are really trying to find the best answer to conclusively say X is better than Y, you need to take other routes, which include existing crash data and R&D performed on a much higher level that accounts for all the real life variables. IMHO if this much energy and time is being spent on something, it might as well be towards something that can help the real issue at hand, which is what physical structure is going to suit best for this application, which will require simple design analysis followed by advanced crash testing and dynamic testing of real cars and thier cages.

Just WAY too many variables to use any of this data that would prove anything more than a simple engineering 101 knowledge application would.

Fortunately for us, these tests have already been performed and the results are out there for us to see on high level professional unbibody and tube frame based "production" cars. I have seen prodominantly "X" bracing with sheet metal gussets for door bracing. I dont think I have ever really seen the "club level" type "nascar-ish ladder bars" on any of these cars. And as carbon semi-monocoque chassis/tubs get more popular for high level competition, its going to factory out the tube-structural design aspect, probably just keeping them in for roll over and side impact driver protection (anti protrusion). I've posted exampled previously on this thread. I'll try to find some more as I come across them.

P.S. There are always too many Steve's

Geo

You're right Steve, I misunderstood. I thought you were talking about tying the cage to the coachwork.

Sorry about that.

Cory M

Here are some good pictures of the BMW motorsport cage that 95m3racer mentioned. This is probably way overkill for the majority of us but it is still worth checking out:

http://e30m3performance.com/tech_art...46-1/index.htm

95m3racer

Thats the standard E46 Motorsport 2dr chassis.

The GTR chassis is better in a couple ways, its a bit more "complex" as it ties into different areas of the car. Its built to a certain spec.

The normal Motorsport cage you linked to is designed to be a basic universal chassis that can be homologated for whomever, whether it be Rolex or Belcar, it can be modified for each teams exact use. Usually they dont even come with a harness bar, as thats upto the team to decide on.

As far as Complexx's cage, it's currently being built