magnetic1

Agreed. We have no way of telling what that same incident would have done to a "NASCAR" bar. Iceman's setup is actually more than what you typically see already. Most "NASCAR bars" in CR are just two bent tubes with a couple verticals.

True NASCAR door bars have a TON of mass.

The safest door design, IMO, would be a combo. Have a X, then "Nascar" outside of that. The NASCAR bars would bend, absorb some of the impact, then the X would protect you. Or you can get fancy foam stuff, but it is $$$

Greg Smith

That is what I would call a "simple X," which is the example fatbillybob was using. Yes having taco gussets, especially on the top and bottom, would help greatly. Joints will generally fail right at the edge of the weld near the HAZ, which is where the tube failed. The weld itself did not fail. Maybe too much heat was used in the weld, or maybe it was 'perfect.' My bet is that either way it would have failed the same way.
My post was not meant to say that an X will fail, simple to show how it can fail(going back to one tube in the middle surrounded by 2 HAZ's). There's more info on the wreck in this thread(post 113+), you'll probably have to join to see the attachments though.

EDIT-First 2 pics are the X car, last 2 are Nascar. According to the forum "The impact was very similar. The lack of damage to the rocker is due to the bars absorbing most of the impact and preventing intrusion. I think this is what we all are hoping for. Check out the photo below of where the energy went, the front was similar. I am just hoping we all can learn from these examples." I don't believe it was as a severe/similar wreck as the X car though, just posting to post...

RedlineMan

Ouch...

Nothing against your post at all, Greg. That's a pretty good hit. No way that material should have failed like that. Something strange there.

Now, I want to know if that was a flat X or a geometric one. The latter has TREMENDOUS advantages at absorbing and diffusing energy in different directions and into different parts of the chassis/cage, and smart builders (including Porsche) use them. Did I mention TREMENDOUS?!?!?

Agreed on Iceman's bars. They are better than a lot I've seen, but they still essentially rely on strength of material, not diffusion of energy.

fatbillybob

Greg,

You are welcome to your style of bars as you see fit. But you are correct you need to trim down the variables here. Bars will always fail if the hit is big enough and you can always find somewhere to some one where they did fail. If you would do an FEA on the two structures the math would prove the value of the simple "X". But just to keep on your point about HAZ and not knowing anything about the cage or it's builder or materials you can say from a purely scientific point of view that suppose this was a Cromo "X". While it is known that stress relieving/mormalizing /tempering is not needed on less that 1/4" tubing we do know from science that the TIG or MIG heat mess up the marsenite structure of the steel and increases the brittleness of an already brittle cromo 4130 tube. So we also know from science that you can head that tube to 1200F for a couple minutes and let it air cool and that will "temper" the steel. Yes temper is the proper term not stress relieve, not normalize. Now if you add that with a thinner bar on the passenger side because it is far from the driver to save some weight well you increase possible fracture. So there are alot of things we don't know about this failed structure. We can start however with a fresh piece of paper and do the FEA and it will tell you what could and could not work. Then you have to crash test some structures.

fatbillybob

This is so key to building a cage. Understanding like this is the difference between a guy who can weld a cage and a pro who can design and fabricate a safety structure. The cages look similar when you write the check but who would you rather be writing that check out to?

Rich Sandor

John H, can you please explain the difference between a flat X and geometric X ??

kurt M

Flat X is two tubes that cross each other in the same plane. The geometric X is a pyramid shape with the point facing out. The idea behind the geometric X is that it will want to spread open when pushed inward transferring energy into the front and rear body. When things work right this should resist force better than just the tubes resistance to bend from sideways force input alone. It relies on its shape as well as its mass. The X shape in a door hole has a side effect of adding good geometric rigidity to weight in the door hole area of the body as well.
Good cage design means using nodes as well as good tube angles, welds and materials. One of my pet peeves is looking at tube layouts that don’t work together to gain geometric strength when they could have. Take post #17 picture #3. The bent Nascar door bar is a clear example of this. Note the crushed/deformed tube with the support tube above it rather than directly behind it. That is an example of poor design IMO. Small detail but all that was needed to make the structure stronger was to line the two tubes up across from each other. No added weight to gain geometric strength. From a glance at the examples it looks like the X bar took a sharp pointed hit while the Nascar bar took a flat hit that pushed in the whole door area. I bet the same shaped hit would have folded in the Nascar bar but not have seen the tubes break at the HAZ as seen in the X bar.

chrisp

I'm curious....assuming a perfect hit at the tip of the X, the geometric would be better, but in general wouldn't it be weaker because upon impact the tubes are in compression rather than tension? If the geometric X sees a load not perfectly centered on the peak of the X won't the leg(s) of the X seeing the point load crease and then start to fold inward until they come under tension? That would mean that the geometic X would have to collapse fully inward (mirror image) before it tensions whereas the flax X would be under tension the entire time. Of course as the X is collapsing inward it's absorbing energy which is good...but compromising the saftey cell which is bad....which is better?

RedlineMan

Chris;

To some extent, you are correct. It aint a perfect world, is it? Still, the NASCAR bar is ALWAYS in the realm you mention. It relies on tremendous bracing to a huge frame sill (heavy tube steel, not flimsy uni-body) and numerous cross braces to be truly effective. The kinds of NASCAR bars you usually see are NOT back braced like a real ones, and are completely dependant on the strength of the material to resist bending, NOT on sending loads elsewhere in different tangents than the direction of the hit.

A flat X is in tension from a static state, and will remain so throughout any incident. A geometric X is in immediate linear compression, and the hope is that it sends enough loads in different direction that it does NOT BECOME a tensioned member. A good NASCAR bar is like a paper bowl glued on a smooth table. If the glue were not there, the sides would slip out and it would just squash flat. With the glue (which gives it far more geometric strength), it will not squash flat, but buckle eventually, and take a lot more load doing it. A poorly built NASCAR bar is not in tension or compression, really. It simply relies on the tube's resistance to bending, and that is not very much really.

---------------------------------------------------------------------------------------------------------------------------------
And no... I don't consider myself a pro by any means. There are many that could fab rings around me, I'm sure. If I'm lucky, I just have a good intuitive, native sense about how something should be. I just do what I do, and try to be successful at it.

stownsen914

There was a thread over on the Pelican board a couple months ago about knee bars (or knee breakers as some prefer to call them), and whether they're worth having in a cage. A couple posts advocated not having knee bars in a cage, citing that you can injure your legs in an impact. They also cited evidence that FEA analysis indicated that knee bars did not contribute to stiffness of the cage. I think this may ignore the fact that in a substantial side impact on a car without a knee bar, the cage could collapse laterally. Thoughts?

Scott

chrisp

I think that in order to substantially support the sill and the foot of the cage from collapsing inward you need a tube in compression running the width of the car. Otherwise no matter where the knee bar is your just relying on the bending of the A-pillar tube. Obviously running a tube down around your feet is a bad idea but there are was to tie the sill and center tunnel together that will no expose you to harm. For example a seat mounting structure with tubes (and requisite backing plates to prevent punch through) running between the sill and tunnel will help keep the driver seating area from compressing sideways.

RedlineMan

Hey;

Knee Knockers DO NOT increase cage stiffness. They improve side intrusion protection. THAT'S IT. Belt bars are the same, save for giving a place for the harnesses to mount.

kurt M

Tension is likely what caused the pictured X bar to fail in the HAZ. The job of the door bars is to prevent intrusion. side hits are a hard one. As FFB pointed out there is a short hard to defend distance between the door skin and your skin. I will post what I did to the car I am building with regards to the drivers side door area and seat.

PogueMoHone

I don't know if these (cell) photos will help or confuse, they are out of my Works Race car. Therefore one can assume that the cage is professionally designed. It is certainly more substantial than many I have seen and is welded to the frame of the car as well as the floor.

Hopefully the Engineers amongst you can figure out the different views. I tried to photograph the various joints and welds. the way they did the center section is very different than one might expect.

What is amazing (to me) is the strengthening at the various corners.

Here goes with the photos:

Jeff Lamb

PogueMoHone, thanks for sharing the pics. There are definitely some very interesting things going on with your cage. I see where a lot of triangulation was used and also the use of gussets and sheet metal "tacos". The door bars look like the "simple X" design with use of sheet metal tacos. It was cool that where the cage builder could not triangulate in the area of the two opening doors, they reinforced the top tube with what appears to be extra sections of tubing blended in with sheet metal gussets.

Does the main hoop of your cage have a full diagonal bar running from the top of one corner to the bottom of the opposite corner? I couldn't figure that out from the photos. Plus, with the absence of a horizontal cross bar in your main hoop, the shoulder belts unfortunately have to be rather long to reach their mounting points (although, the shoulder belt mounting location does look very solid).

I'd love to see some more photos of your cage if you have the chance to shoot some cleaner images with a camera.

Jeff