Door Opening "X" Bars as Side Protection

[dickita15]

To Chris’s comment about the drivers wanting weight savings of the X bars, I saved plenty in gutting the right door to make up for the small about of extra tubing.
It is somewhat bothersome that this incident took the force in the door alone rather than using the crush structure of the rocker and floor. Given IT cage rules I do not know how to change that.

[Wreckerboy]

I'm glad that Rich Hunter is okay. Does anybody know anything about the driver that hit him? Chris built the cage in my Miata as well (which I unfortunately crash tested a few weeks later at LRP). I can't say enough good things about the quality of his work.

Earlier it was mentioned about drivers wanting the weight savings of x bars in the right side doors. When we discussed this as Chris was putting my cage together last summer, I thought (please correct me if I am wrong) that he mentioned that one of the advantages to NOT doing so was the straight load path gained from front to rear.

I understand that everything is a tradeoff, but but perhaps the best solution is a combination of the two or three approaches: the traditional X bar style arrangement as well as the NASCAR anti-intrustion design, as well as tacos. Yes, this will impart a weight penalty. The other thing to consider is the expanding foam mentioned here. The NASCRAP COT uses attenuating foam in the doors now. Does this sort of thing exist for our adaptation? I'm thinking that spritzing the inside of the tubes with the conventional expanding foam available at Lowe's is not a good idea...

[Greg Amy]

I'm re-quoting Sandro's photo, because this is a good illustrative example of Richie's/Chris' build. Note the driver's side full-up "NASCAR" on the left side; right side is a single-tube bar bisecting another bar to create an "X". Richie did not have a horizontal bar at the bottom.

I think something here is being overlooked. We should be looking at the car as a sacrificial anode in that the structure successfully absorbed the force of impact which allowed the driver to survive.

By your description of the force of impact, it sounds like the cage did a good job of absorbing the energy. Is there really a design that's practical in weight and material cost that can withstand any possible impact ? And how should it look after an incredible impact ? I wouldn't expect it to look like it did before the accident.

I suspect you don't understand the full scope of what happened to Richie's car. If Richie had been 180 spun around with the x-bar on the driver's side, he would be dead today.

Period.

There was absolutely ZERO space between the passenger side door and the exhaust tunnel. Zero. The passenger side door was halfway into the car, and only inches from hitting the driver, thankfully on the other side of the car.

A lot of people like to get into a discussion of "crush zones" and "energy absorption" and the like, but we don't have the tools to be able to quantify that. As such, it becomes more an "art" than a science. And without a survival cell, absorbing energy kinda doesn't matter. Granted, dying of a detached brain is just as dead as being crushed, but given the choice I think I'd rather take my chances with g-forces.

Finally, while this discussion centers around a crushed passenger side, my focus is how we should design a proper driver's side crash structure.

The design you show does appear to be a stronger way to make an X. I'm sure that at an 80 mph direct impact there would be some failure somewhere in that construction too.

Of course. But this particular door structure failed in tension, and failed in a way we cannot accept.

Let me make this clear: I am not hammering Chris' work. In point of fact, my Nissan NX2000 has a (highly more structural) x-bar design on the driver's side. But, my point is that the design as illustrated above in Sandro's photo can withstand exactly half the tensile forces of a true double bar.

Think of it this way: let's say you cut that "X" out of there and hang a car from it; how much force can it withstand? Only as much force as one tube. Why only one tube, you ask? Because its weakest place is in the middle, where there's only the cross-sectional area of one bar. If, however, you were to weld a plate across the face of that "X" it could withstand twice the load, because it's now being distributed among two tubes.

A "taco" gusset would not do the same thing, unless it were also welded across the face. The purpose of gussets is not to increase tensile strength but to support the tubes in a bending moment. For the purposes of the discussion at hand, two plates welded across the back and face of the "X" would be just as effective.

It is somewhat bothersome that this incident took the force in the door alone rather than using the crush structure of the rocker and floor. Given IT cage rules I do not know how to change that.

Well, as you know, the best thing would to be allowed to weld to the rocker panel. But trying to get that IT rule changed would be Sisyphian. As a compromise, I'm envisioning using the plate are rule to extend the plates as far towards the center of the door as possible, then welding tubes to that. Kessler always added a longitudinal bar at the bottom on my cars (similar to Sandro's photo, above) but I think we might expand upon that a bit more.

Does anybody know anything about the driver that hit him?

I saw Richie talking to him in impound, looked fine. Front-end damage on the 'Ta didn't seem significantly worse than other bad SM wrecks...that car seems to handle hits pretty well.

I am the one who built the cage in Richies car. I wish Rich had let poeple photo the damage as that is the only way we all learn.

I chatted with Richie about it, but he wanted to discuss it with you first. If you can talk him into letting one of his guys take some photos to post here (or you take some when you see it), I think we can all really learn from it.

Honestly, Chris, there was really nothing you could have done differently to the passenger side to affect a significantly different result; it was that hard a hit. All I'm offering here is a detailed discussion of the failure mode and how we might apply that to our driver's side going forward.

This is timely, as we've got a new car build going to happen this winter. I'm personally no longer a fan of the pure "X" bar on the driver's side, though I'm not clear exactly how far towards the "NASCAR" I'm wanting to go...

[lateapex911]

I looked at it closely. The bar (as Greg described) sheared, as did all the other failure points, near to, but not at the welds. The door pillar/upper rocker area was pushed in nearly a foot. The cage had significant deformation at the cross car horizontal.

The points being made regarding the car deforming so that the energy being transferred to the driver is less are good ones. When we see a cage sheared, we get all amazed. In and of itself, the shearing isn't the end of the world, unless it happened at very low velocities. I'd GUESS this impact had a car that was initially traveling at 80, then locked them up, and hit in the 50MPH range, which is significant as Richie was not moving at all.

On the passenger side, I think "weaker" isn't bad, as it allows more deformation, which takes energy out of the driver impact. On the drivers side, such intrusion would be catastrophic. That said, I think a lower bar wouldn't be a bad idea, with a vertical bar linking it to the center of the X.


EDIT: I see Greg posted as I was writing, making my comments redundant. Oh well!

[tom91ita]

how are the "taco" gussets being formed for around a bar? i see the formed hole as more optional for further weight savings.

i think i will add some sheet metal to my driver door bars as some insurance. not for this type of incident but there are still suspension pieces, etc. that can penetrate the door/cage.

and my car/logbook is old enough that the plate rules do not apply, iirc.

edit: that went away, i guess. haven't looked lately cause i was not changing anything. here is an excerpt from 2008 GCR.

3. Mounting Plates
a. Mounting plates welded to the structure of the car shall not be less than .080 inches thick. The maximum area of each mounting plate in the American Sedan, Improved Touring, Showroom Stock, Spec Miata, and Touring classes shall be 144 square inches. Plates may be on multiple planes but shall not be greater than fifteen inches on any side.

[chuck baader]

I see the addition of door bars on both sides as extra intrusion protection. I designed my own cage and I like to be as far away from harm as possible. Attached are the pictures of both sid[IMG]file:///C:/Documents%20and%20Settings/Charles%20Baader/Desktop/TEMP/DSC_0027.JPG[/IMG]es of my car. Chuck

[RexRacer19]

This is timely, as we've got a new car build going to happen this winter. I'm personally no longer a fan of the pure "X" bar on the driver's side, though I'm not clear exactly how far towards the "NASCAR" I'm wanting to go...

Hey Greg,

This thread has lured me out of lurking...

That is a tough question/quandary that you have posed there. I am not sure if there is an ultimate solution here that is not a compete tank of a roll cage. My thoughts have always been "build it as best you can within reason, but there comes a point when, no matter what you do from a safety standpoint, the crash is just going to be too bad."

From the two schools of thought...The X style bar is structurally more efficient and increases the overall strength of the cage IMO. There is however, less metal there right next to the driver. The NASCAR style bars add more metal to the area adjacent to the driver, but are generally full of dead load paths. In nearly every example of a NASCAR bar shown in this thread, there is a nice mesh of tubes forming the side impact. BUT, that mesh of tubes gets attached to a single vertical tube in the front, and a single vertical tube in the rear mid-span. When having to take on an impact such as Richie's, The tubes that the NASCAR bars are attached to will be the first to deform/fail in a massive impact. Just like in Richie's crash, they will be in tension, which is less than ideal.

-Jeff

[Speed Raycer]

IThat said, I think a lower bar wouldn't be a bad idea, with a vertical bar linking it to the center of the X.

Possibility of becoming a spear in the same situation.

I've been toying with going with a 1 bend X lately. Single straight bar ala the traditional X then the second bar bent out around it, then gusseted together.

I have to wonder if the car had had a NASCAR setup on the side that took the hit and had the same results what the discussion would be.

Why isn't anyone bringing up 4130 as an option to the 1020 used? 4130 isn't all that much more expensive anymore. Maybe 1-1.50 pf

Edit... reason I bring up 4130 is because this seems to me to be a case of the materials properities being exceeded during the impact... not necessarily of the "design"

[ekim952522000]

<----Mike who is very interested in the discussion is getting a cage put into his ITR ITR next month and has to be worried about getting hit by 3200lb Supra's

I am thinking that a 3 bar door bar setup plus lower bar is going to be necessary on the drive side.

[tnord]

Possibility of becoming a spear in the same situation.

I've been toying with going with a 1 bend X lately. Single straight bar ala the traditional X then the second bar bent out around it, then gusseted together.

I have to wonder if the car had had a NASCAR setup on the side that took the hit and had the same results what the discussion would be.

Why isn't anyone bringing up 4130 as an option to the 1020 used? 4130 isn't all that much more expensive anymore. Maybe 1-1.50 pf

Scott -

is 4130 just a type of steel "alloy?" i'm not terribly knowledgeable in this area, but i am interested in this topic as i'll likely be building a new car this winter that should have plenty of room for extra weight in the cage.

is chromoly any stronger than the "regular" stuff? if you were to strategically place some stronger tubing (being chromoly or 4130 or whatever) that might cost and/or weigh more, where would you suggest that be used? the main hoop and the driver door bars perhaps?

[ddewhurst]

***The bar (as Greg described) sheared, as did all the other failure points, near to, but not at the welds.***

Very ^ interesting. Jake, how far would you remember the failures are with respect to the edge of the welds. Do we know what tube material was used & what weld process was used?

[chuck baader]

4130 has greater ultimate strength than DOM, but it brakes whereas DOM bends more before it brakes. Chuck

[RexRacer19]

4130 has greater ultimate strength than DOM, but it brakes whereas DOM bends more before it brakes. Chuck

I would add that the welding process would be more difficult as well. For this application, I think that TIG welding would be the only good way to do it to keep it from becoming too brittle at the joints.

[Speed Raycer]

(Random ramblings while I'm at work....)

First, I'm not a metalurgist but I did get some sleep last night
2nd, I'm a much bigger fan of 102x in wheel to wheel cages than I am of 4130 due to the nature of the two materials.

It doesn't surprise me at all that the Mild Steel material failed *away* from the welds. The impact exceeded the materials properties. The welds *when done properly (and it sounds like they were well done)* end up being stronger than the base material. Its the nature of the material.

4130's (chromoly) yield (bending) point is greater than 1018's (mild steel) yield point (75kpsi vs 70k). Ultimate strength is 95kpsi vs 82kpsi. It's tensile strength is higher which leads to the "break before bend" reputation. For the same tubing size and thickness, the two materials weigh the same.

There's a ton more work/time involved with using 4130. It likes to be welded slowly. Same joint, a MIG weld might take 2-3 minutes, a TIG joint can take 3x as long. The pre/post weld heat treatment process adds time. I won't get into the debate on whether or not the heat treatment is necessary at our required thickness' nor will I get into the TIG/MIG/OXY debate either. 4130 & 1020 shouldn't really be mixed as the materials "wash" together during the weld making wierd things happen to the material properties of the 1020 tube.

A good free read from our buddy Mr. Smith:
http://books.google.com/books?id=5a8...esult#PPA66,M1

I liken 4130/10XX to a Hershey bar. 4130 is a bar kept in the freezer. 1020 is one kept in a cabinet. Feel free to do the experiment on your own

[Sandro]

The NASCAR style bars add more metal to the area adjacent to the driver, but are generally full of dead load paths. In nearly every example of a NASCAR bar shown in this thread, there is a nice mesh of tubes forming the side impact. BUT, that mesh of tubes gets attached to a single vertical tube in the front, and a single vertical tube in the rear mid-span. When having to take on an impact such as Richie's, The tubes that the NASCAR bars are attached to will be the first to deform/fail in a massive impact. Just like in Richie's crash, they will be in tension, which is less than ideal.

-Jeff

agree with what you are saying, but when I designed mine, I thought that bending, of the main hoop and a pillar bar would allow for some absorption.(the main hoop only has about a 2ft span where it isn't reinforced in another direction, and even less on the a pillar as I have a bar going to the firewall.) If I had added additional bars in line with the nascar bars to say the rear or the opposite b pillar, their would be zero absorption and in a hard blow my brain would probably fly out the passenger window.

My nascar bar also only has two bends( /-----\ ) on each bar not four as I see many do ( _/---\_ ), so a blow on the nascar bar will push the a and b pillars at 45*, not even relying on the welds to hold them in place, cause the bar will be forced into the a and b pillars, as opposed to the bars with four bends that you are now relying on the welds to hold the nascar bar to the pillars, putting the welds in shear when hit. Which as someone mentioned even when the welds are done properly the metal around them ends up shearing.

Please don't think I am trying to say my cage is perfect, just trying to give some insight to the thoughts I put into it, especially as I said my biggest fear is being T-Boned.

I personally think the passenger side should be able to give a little, so it reduces the g load in an accident but on the driver side you don't have that much room for it to give, until the driver begins to be crushed. So as Greg said I think on the driver side you need to accept the higher g load, instead of being crushed. That is also the reason I have a fully bolstered seat to reduce the chance of snapping my neck.

[jerestahl]

A friend of mine did a steel tubing failure research program 29 years ago. I believe the results were published in Stock Car Racing Magazine at the time. The bottom line he says unless you pre-heat every weld to at least 300 degrees you will have parent material failure when the loads are high enough. I have some frightening photos taken of a SM at Summit Point last year that has multiple tubing failures. The attached photo was my interpretation of how to build some protection into a door bar and to meet the idiotic SCCA 8 point cage attachment rule which is the 1st rule that needs to be changed to improve driver safety.

[Dave Patten]

Guys,
4130 (chrome moly or for the old timers,”Shelby Tubing”) is stronger than mild steel (1020) but if not properly welded and the welds “normalized” after welding it is significantly weaker than mild steel.

The welds on 4130 steel need to be normalized once completed. Basically this means the whole weld area must be heated and allowed to cool slowly, all at the same rate. This is very hard to do and something an average hobbyist does not have the talent to do. I never have used and will not use 4130 in anything structural that I weld just because I lack the expertise required to do the job properly.

On another note, welding creates a change in the metallurgic structure of the metal from the heat of welding. The fact that the weld joints didn’t fail in Richie’s cage is a good sign that it was properly welded. The breaks adjacent to the welds are as a result of the weld processes and not a fabrication failure. These weak points are just a matter of the process, nothing more. To limit the amount of heat during the welding process, tight fitting tube joints are needed. Filling gaps with filler rod or MIG wire will make the metal adjacent to the weld joint even weaker because of the added heat it introduces.

Also when doing a NASCAR style door bar, if you stagger the vertical tubes so they are offset from the bars above and below the whole assembly will be stronger because the weld process heat will be dispersed on the horizontal tubes and not concentrated top and bottom at the same point as when they are directly aligned. See this article in Stock Car Racing magazine http://tinyurl.com/544ogk

[RexRacer19]

Please don't think I am trying to say my cage is perfect, just trying to give some insight to the thoughts I put into it, especially as I said my biggest fear is being T-Boned.

I don't want to come across as a know-it-all either. As I have told some others, there is more than one way to skin a cat when it comes to these things. Both design types have merits and drawbacks. The important thing is to build/buy a cage that YOU will feel safe in and meets the standards of the GCR. These discussions are generally healthy and provide some insight for folks going through that process.

[spnkzss]

I don't want to come across as a know-it-all either. As I have told some others, there is more than one way to skin a cat when it comes to these things. Both design types have merits and drawbacks. The important thing is to build/buy a cage that YOU will feel safe in and meets the standards of the GCR. These discussions are generally healthy and provide some insight for folks going through that process.

I have to agree. I find it humorous that the first cage I had in my C car worked just how I needed it to. Then when I built the A car I had a whole different mentality of what I that was right and modified accordingly. THen I saw what you and Xian were doing and have another completely different mentality.

[Xian]

I have to agree. I find it humorous that the first cage I had in my C car worked just how I needed it to. Then when I built the A car I had a whole different mentality of what I that was right and modified accordingly. THen I saw what you and Xian were doing and have another completely different mentality.

Very true... I saw some nifty things on Jeff's cage that I liked

FWIW, back to the question about side impact penetration... here is "my" solution to it:




By "my" I mean "Chris Schimmel at Competition Cages" Similar to Kirk's cage, my door bars project out into the door cavity and should end up in compression in the instance of a side impact.

Christian