Geo

A couple of things about this.

1. I've never seen anyone build door bars this way with a regular curving arch. I cannot imagine anyone having the dies necessary to do so, especially since this arch would vary from car to car. NASCAR bars are not a true arch. They are straight tubes with curves on the end. Much less than ideal.

2. I'm sure the bending mentioned above is true for a square (axial) hit. What about an angled hit? Will it be as strong? I'm imagining not. Also, if one of the tubes the arched tube was attached to was hit, I think you'll find the tube will bend easier than a straight tube. For instance, use a tube like you show above for a dash tube and take a hit in the front support tube and the tube will keep bending (don't know if it would be stronger or weaker than the compromise I made).

One of the common misconeptions of NASCAR bars is that they are an arch. They are not. Sice they are already bent, the failure mode would be to bend inward before having to stretch. A straight tube between the main hoop and front tubes MUST stretch in order to bend. This takes considerably more force than just bending an already bent tube.

Mike S.

Geo et als...my "1/7" statement pertains to substituting a 3" offset gentle arc for a straight tube behind the driver on the horizontal tube in the main hoop. It was a followup to Larry's post. I selected an arc to make the calc easy for me It'll get you in the ballpark of a die bend at the ends tube that is typically done. Yep, just for a pure colinear axial load. Numerical support that reveals that "big bows blow" for colinear axial loading. Nothing here applies to doorbars.


Good points, info, and question..."bend or no bend?" I'd say it debends. I'll make some numbers at lunchtime to look at the effect of the bend amount. It'll give us some qualitative "meat" out there to chew on (aka masticate per John...I had to look that one up in my dictionary). I'll use 1.75" OD X 0.095" wall 1020 steel DOM tube. Question...do you cage building guys use hot finished or cold drawn tube?

Mike

RedlineMan

Regarding Door Bars;

I have only intuition to go on. That is my CAD. To my mind, the effectiveness of a "NASCAR bar" relies on two points.

1) What the bars attach to. Like any structure, if the foundation is weak, the structure is by default. In this case, the foundation is the vertical side areas of the main and side hoops. If these lengths of tube are free standing, only being attached at their base plates, they will not have enough strength to support the pressure applied to the door bar assembly. If some form of bracing in compression exists to transfer/resist door bar load ("knee knocker," "Belt bar," or other transverse brace), then the door bar assembly itself can do its work.

2) Being massively over built. An arch is a very strong shape, but these bars are not linear arches. Because they are essentially flat along their length, it is my intuition that they possess very much less strength than a true arch. Further, the disance that these bars project out is just the distance they will project inward if they fail as an "arch" structurally, until they again take tension and start to pull on the members they are attached to. The NASCAR guys must be aware of this, because they do not rely on one, or even two bars, but 4-5. Then, not only are there many tiers of tubing, but each is tied to the rest by multiple sets of vertical stringers. If you hit one bar, you get the strength of the rest activated as well.

My gut tells me that you lose quite a bit of the overbuilt strength of a NASCAR door bar as you reduce the number of members. I feel it is rather obvious that even an overbuilt NASCAR door bar assembly would not help much if the transverse bracing of the hoops was left out. Therefore, my feeling is that you need to maintain a healthy percentage of these two aspects to gain any significant effect from any door bar assembly of this type.

Thoughts?

M758

remember also a NASCAR weight what 3500lbs. So there is plenty of weight than can be added in extra tubing and it won't cause a car to over weight. In otherwords they have the luxury of adding 50lbs of door bars because their cars weight so darn much.

If weight is an issue you need to use good design to get strength at a low weight. If weight not an issue you can "just add metal" till it is strong enough. You run into problems if you take "heavy design" and reduce the weight and expect it to work.

tinman944

Agreed, agreed
With NASCAR bars it is the whole package.
Honestly, And I am guessing here as I have not seen a rule book or built a Busch car for ten years.

I think when you see how close there are to the steering wheel is because they have shortened up the door area.(cockpit area) you are really in a safety pod if you will.
So in essance the door bars are shorter that would make them stronger.( does that make sence??)
Plus as John said they really stack those bars in there I think they are up to five high now.
I think the last car we built was four high and yes the vertical bars are built into the equation.

tinman944

The other part that NASCAR has that most do not.
All the cars are the same.
When one car hits another,Rare that they do.
They know the impact zones that they need to make tough.

Geo

Also remember that not only are NASCAR cars heavy (allowing for all the mass in the door), but that they want as much weight on the left side of the car that they can get (I would assume this is regulated by NASCAR however. In any event, having the left side heavy on an American oval track car (do they run circle tracks clockwise in other parts of the world?) is not a disadvantage. Having the left side heavy on a left/right car isn't the best idea, especially when a fat@$$ like me is doing the driving (making the left side even heavier).

Mike S.

I think you hit a home run with you interpretation.

And another home run.

Good design/engineering looks at the system, not just individual components. Another hard part is envisioning all the possible situations that may occur and pragmatically solving them (to the extent they have sufficient value).

Thanks guys...all this is educating me and broadening my thinking re 944 cage construction.

Mike

tinman944

John,
On your cage are you intending to put front bars in ?
I think I understand you dont like knee crushers.
Is it your height that causes this?

RedlineMan

Hey Bob;

Some time here... I'm gonna have to earn some money! I am fully prepared not to have time to finish it into a full cage. I also don't see doing a dash tube, although I have nothing against them... necessarily.

I really am not convinced of the cost/benefit for the driving I'm doing right now (DE). If I finish things, I will be integrating the side hoop into the cowl and a-pillar as I have the sill and main hoops, and hope to get decent intrusion protection that way. 944s have pretty rugged cowls to start with. If I ever did a dash tube, I think it would have a bend in it (Thanks Mike!), or maybe even a beam structure like my rear truss.

I honestly feel I stand to get hurt from smacking the knee bar more than a t-bone intrusion... again, in the driving I do.

fatbillybob

I'm thinking we are loosing focus and overthinking the problem. Now when I do these things I always remember what you told me once that " I was thinking too much." I think the design excecise is great to kick around some ideas but for example integrating the sill tube fully in the OEM sill plate is really just making a larger sill plate and I am not so convinced that this is then the foundation of a good cage. I propose that the "cage" should be within the car structure and attached to it but not be part of it. I have been thinking about the bend issue brought up by Mike. I think Mike has a great idea but I think it is the wrong approach that math alone does not have a solution for. When things get complicated and advanced I always go right back to the basics. The basic reason you have a cage is as a last line of defence because something bad is hapening and the structure around you is falling apart. Therefore, the cage needs to remain intact and be as strong as possible while the car is best designed to disintegrate to shed off energy. I would like the oem sill plate to bend like a pretzel but never have that energy actually reach and deform the cage or its sill tube. If you make a sill plate "one" with the sill tube any impact instantly compromises the "Cage". Not withstanding the NASCAR view of the cage world, the "proper" cage has already been thought out in most of the rulebooks out there and mostly that means straight "X" doorbars and straight dash tubes and as few bends as possible main hoops and triangulation where possible. If my dash tube bends and hits my chest I don't think it will be the dash tube that killed me. Max integrity is the goal with a cage. That means straight tubes, minimal bends, and triangulation.

As a sidenote, when someone uses light 18G steel with lightening and stiffeningholes and tacks it from say an A-piller to the A-piller tube that 18G steel will deform before the 0.95 wall or 0.120 wall tube and helps to maintain cage integrity. If you uses 1/8" plate to attach tube to A-piller you have a stronger structure but any force is directly transmitted to the tube and compromises it because the 1/8" plate transfers 100% of the energy into the tube while the 18g deforms. I got no math it is just how my head it thinking right now. Am I way off base?

kurt M

(Caution uneducated opinion to follow) FBB, If the force is able to compromise the A pillar area and then be resisted by the cage inside would it not have been resisted if the two were originally connected? One gets pressed into the other. There ether is enough force to bend them both or not. There might be odd conditions where this would not hold true but for the most part it should. You might also gain structural strength by having the two working in lockstep. If you have two tubes in a plane and you web between them you now have an "I" beam of sorts and it might stronger than the two tubes by themselves.

I agree on the tube on the sill plate. Inefecent use of materal and no added strutrual triangulation. Any added streinght is only via the tubes side load streingth and that is a poor use of a tube. There are other lighter weight methods to make a tube root weld point. If the design parameters change then the design changes but if you are putting a cage in a car that 5000+ other folks have as well for the same uses as they did you have the power of learning from their mistakes rather than making your own.
IMHO Simple systems that work are the good systems. Building a tube based safety system based on looks or any criteria other than strength to weight is not good build practice.
Sometimes things have been reduced to the lowest common denominator by others that came before you. Copying a good commonly used design is not bad or demeaning or cheating.
I like to use the 4 wheels saw. Why do cars have 4 wheels? Perhaps it is that 3 or 5 just don’t work as well for most uses. You can build a car with 6 or 3 wheels but will it be better for the job at hand? Big trucks have many wheels not for the best design of the truck but to spread the load out on the road. If the roadbed is not a problem then 4 really big wheels makes for a better more efficient truck. Look at the off road dump trucks. They do not have to be built to fit on common roads and in turn have the same ratios as many cars just much much bigger.

John, please take no offence to this or the above as I don't want to be part of the pissing match but you gotta remove the bondo brother. There is NO place for bondo in any safety system, period.

RedlineMan

You both raise some good points.

The webbing between tubes and unibody are indeed to create beams. Engineered beams do not rely on mass of material, but layout, to achieve superior strength-to-weight characteristics. They can be at once stronger and lighter than a more massive part.

I am tieing into the car where the occupants are most vulnerable. This is at the door posts, side hoop, and header tube. I have seen more wrecks than eveyone else here combined, I'd wager, and I know how cars are built and how they deform. That's from 25 years fixing them and 10 years insurance appraising them.

Rocker panels have torsional and longitudinal bending strength, but little else. Looking at Kurt's statements made about my sill tube will prove that. But, a tube is a lot a stronger than tin in that regard. The only part of a stock chassis that has any large side intrusion protection is the same part of any cage; the transverse members, mostly. In a stock vehicle, those are the door bars, the pillars, the cowl, the rear seat base structure, and the front seat mount area that is usually a beam of some sort. A rocker panel assembly itself offers very little intrusion protection. It's just thin gauge sheet, after all. If it were that strong, and 911 would not need a reinforced jack point, and 944 rockers would not buckle when using the center jack point.

Yes, simple is good (assuming smart too). That proves nothing else. That does not prove that other ideas do not work, or might even be better. Where racing rules are concerned, as someone said earlier that is often to remove "creative" engineering. If simple were the only way, and always better, there would exist no room for innovation and improvement. Things are improved constantly by people who don't limit themsleves by "simple" or "conventional."

"NEWS FLASH - Driver survives horrendous crash only to die from bondo poisining!"

I don't think so. If anyone questions how the stuff under the bondo was done, I've got lots of photos, and there are plenty of other areas that will not be so finely finished for them to look at. Sorry, but the bondo stays! It's going to look soooo nice when it's painted!

tinman944

Myself I would rather use the 18 GA and let it absorb the energy before it got to the cage.
Now this becomes a fine line when getting a cage approved for racing.
Some allow 4,6,8 point of mounting and give a square inch area to do it in.
I have never tried to get one past tech in this way ,I have only heard stories of having to saw cut between the two.

fatbillybob

Kurt's I beam make complete sence to me and I agree. I'm just not sure that's what you want. I also Agree with John that there is not much strength in the mentioned places. But I am not sure we want more strength there. Tinman actually put what I was thinking in simpler terms..." I would rather use the 18g and let it absorb the energy before it got to the cage." I think we need this ride down in energy and we need the car to fall apart. In fact I do not agree with people triangulating past the front shock towers on 911's which I often see because doing so makes for a rigid car but the driver gets 100% of the impact (but maybe you need to make that compromise to protect the fuel cell). We are so caught up in having no build-up of kinetic energy in our seatbelt systems wanting low stretch belts etc... but forget that we want all that stretch in the form of crumple zones in the car shell. At least I think this is what we want.