RedlineMan

I agree to some extent;

Within the framework of what these cars were originally designed for (use at less than 100mph), the deformability is well designed. However, depending on the speeds we will put them to, I feel you need to ratchet up the protection commensurate with that. A 911 chassis is pretty soft up front, and it does not take a whole lot to throw the roof up in the air in a front end collision. If we're talking stock class speeds, I would not be overly concerned. If I'm in GT3 territory, I'm putting tubing ahead of the shock towers. The 55MPH bumper, as I call it.



Same for my side intrusion setup. I agree you'd like some deformability, but again relative to speed and potential impact. I still have all of the original sheetmetal meeting the blow first, with the added support of a longitudinal member behind it. I feel it is a distinct possibility that this may ride down the higher energy of a track biff while keeping intrusion to a safe minimum. Speculation indeed, but seemingly sound.

M758

Well herein lies one of the biggest issues with cage design.

The crumple zone.

Modern cars are design for performing well in 50 or so mph impacts. This means the car will crumple and absorb energy before putting that into the driver. The factory can make cars much stiffer, but this hurts the driver more than helps.

Now the roll cage. Roll cages do a nice job of just that. Keeping the roof above your head. However tube frame the entire car and you can make it too stiff.

This is partially the Dale Earnhardt problem. The Nascar tub is designed for major high speed impacts and 50 mph bumper cars. So impact in the wrong way mean the tub does not deform, but put all the energy in the chassis and then in to the driver. So then things like a BSF becomes that more regualar.

Make the tub soft and it absorbs the energy through metal deformation. The trick it is very HARD to optimize a tub for a 50 mph wreck and 100 mph and a 150 mph wreck. Even if it is a simple "full frontal" type collision. The structure that works at 50 may be crush like pretzle at 100mph. The one that is strong at 100 may be too stiff at 50 cause major deceleration to the body. Thus the driver feels a the same energy as 100 mph wreck because the cage did not absorb.

The biggest down fall of the cage builder is the lack of proper analysis and testing. A factory can analyize the deformation of the stock tub and they use design and material selection to tune the chassis for impacts over a certain speed range. As soon as you add in a cage you change that tune. Without doing detailed analysis and knowing the properties of the base car you can't ever do as good a job at tuning as the factory. In fact most folks simple "eyeball it".

Thankfully years of cage building and race car wrecks has provided a certain body of knowledge about what things work well and what does not. Even so John's "55 mph" bumper can only be described as "something that seems like a good idea" I doubt he (nor 99% of all cage builders) can provide hard numbers on what the modification does to driver energies in even simple impacts. We can only guess that it is a good thing.

Now if you think about any wreck you may be involved in you need to estimate the speeds.

Most car to car contact occurs at lower realtive speeds. Mostly since while both car may be going fast the impact is result of the differecne in speeds. Clearly what is more dangerious are the two wrecks in the 24 hrs of Datyona this year. DP's coming full bore around the kink in the infield wackign at car that got turneded around. Even then however you have one fast moving body non rigid body hitting a stationary, but not fixed rigid body that can move. Better tag the side of a car 130mph that a solid wall.

So therefore the worst impacts are often single issues. So figure if your car has a max speed of 140 mph and comes in at 2800lbs you can begin to calcuate the worst case energy dissipation needed. Then figure once you spin your speed slows and has is vector. This means the direction if speed is very important. For example a glacing blow is much less impact energy as compare to a more solid hit even if the car is going the same speed.

So as you can see you can probably do some calcautions to determine the mostly likely forms of impact and get some rough numbers on impact energies. Form there you could build a cage around those probabilies. Factories do it for road cars, and certainly they do it for F1, Indy car, probably NASCAR and high even Road race cars (Prototypes etc). Not so for most of us club racers. however.

tinman944

M758,
The average shop can not afford to do the testing needed to give all of your answers.
Believe me If I could crash test cars I would be doing it. Mostly for my own personal satisfaction.
I have been to the Gm factory and watch crash testing. There is more to it than just wrecking cars( As you all could guess)
I am not sure how most do it, but when I build anything out past the shocks on either end of a car I always use a little lighter tubing. In hopes to absorb the energy.

Larry Herman

Sounds like you could borrow from bicycle technology a little. Large diameter thin-walled tubing. Extremely stiff, yet quite crushable.

kurt M

To absorb the 140 mph to hard surface hit you would likely need a 30 foot long crush zone. As Tinman points out go thin wall or dia past the susp points. Saves weight too.

I like to tinker as much or more than the next guy and have a shop full of fun stuff to mangle metal with but when it is time to tube up my 911 project car I am looking at and coping the pro builds.

Ski

not to oversimplify this, but through twelve pages:
YOU build cages on your gut and intuition, your CAD is in your head, and based on instinct? People actually trust their lives to this thought process?

I actually had to go back and look at the first photo again of that harness bar.

RedlineMan

Bret;

You are making some assumptions.

1) That convention and sanctioning body rules are always the best way, or are the only way. It is true that a lot of that is tested and proven to work under certain circumstances. Usually at best, this means it works most of the time, or is the most appropriate compromise known at the time. However, testing and modeling are only as "accurate" as the number of trials and differing situations you put something through. These "conventions" have also proven to be less than perfect many times over history, and if these rules are not constantly rewritten to adjust for modern knowledge, they become a detriment.

2) Just because something you see does not look "conventional" does not mean that it is not fundamentally sound, or further that it does not contain the actual underlying properties of the covention if you dissect it carefully.

For instance, I take it you do not like my belt bar? You think it is weak or otherwise unsound? Do you know that attaching the belt bar to the rear braces by simply running a straight tube between them is recommended by the FIA as a legitimate means of setting up shoulder harnesses? Let me give you an example of why this could be a very smart way to arrange this.

The "convention" says that you need a transverse horizontal tube between the sides of the main hoop, intended to give side impact protection. Most times this bar is also utilized for the shoulder belt mount. What happens if the car sustains a hard side impact and that tube buckles? If it buckles forward, the driver's belts just got loosened. Is that a good scenario? A belt bar mounted to the rear braces is fairly isolated from a great deal of deformation a cage might see, certainly more so than a transverse tube in the direct line of fire.

Let me carry this out further for the sake of argument. Let's say that you now will change your mind and agree that if the FIA says it is OK, that it must be a good way to do it. My design is far stronger than a simple straight tube between the rear braces.

People here have already said that the rear braces would just buckle since the load was pulling on them right in the unsupported middle. That is not likely to happen in the first place because that tube is under extreme tension in a static state. ANY tube or assembly will fail given enough force, but at some point you have to step back, take a wider view, and pay attention to actual need.

In regards to mine, if you study the diagram of forces I posted later on in the thread, you will see that on top of that brace having to pull everything it is attached to into the center, there is also a triangle created by my layout that prevents the rear tube from buckling down. In a 40G hit, that belt bar bar would see somewhere in the neighborhood of 3000lbs of force through the shoulder straps. Does the longer tranverse belt bar layout withstand more force than the rear brace belt bar? Why? Are the side hoops less likely to be pulled in than the rear braces pulled forward? If the former is "acceptable", why not the latter?

People are free to think what they want. Their ideas will be based on their ability to comprehend what is going on. If someone walks away from this thread with one wrong-headed convention questioned and exposed, or maybe a new idea for improvements in equipment... or thinking, then it has been valuable.

kurt M

John, are you questioning conventional thinking or physics?

APKhaos

Amen, brother.

Geo

Yes, but the main hoop is then supported by 2 diagonals. The main hoop does not go unsupported internally

Pretty much the driver dies. To make that tube buckle would involve such force that you can forget the door bars holding.

Again, I don't think the horizontal tube across the main hoop is likely to deform and if it does, you won't need the belts. Personally, I wouldn't race a car w/o that horizontal tube and I have two in my car for added strength in a critical area (upper and lower).

I don't see it. You have NO protection from a side impact collapsing the legs of the main hoop.

Sorry John, but your design has critical nodes hanging in mid air with little support. You're also not very well triangulated.

Sorry John, I know we've butted heads on occasion, but I have always respected you. I've stayed out of this, but I must say, if someone walks away from this thinking this is the way to design a cage, that would be wrong. FWIW, Tinman was harsh, but I think right on the money. As I tried to say in my best diplomatic writing, this is a case of being too clever, which is where I think too many cages go wrong. You can come up with reasons in your head why it should work, but that doesn't mean it will.

FWIW even the pros get too clever. Over the years I've seen a lot of fads come and go in cage building, but they always seem to come back to the basic standard and for good reason.

RedlineMan

Yes - To the extent that conventional thinking stops most people from thinking beyond. This thread highlights that quite effectively.

RedlineMan

True. I have debated the inclusion of a transverse bar from the start of this project.
BASED ON MY NEEDS AT THIS TIME I have resisted it, and yet I am still on the fence about it.

All speculation. I'm not racing this car either. If I ever do, it is a moot point because I will have to add one. I've never said they don't do anything, I just disagree that it is the only way.

You're right, you don't see it. You are a proven coventionalist, Goerge, and you don't see it because it is uncoventional. Not saying it's better or worse than the cage convention, but it's there. Maybe not to the standards of convention... or maybe so? Who can tell me?

Show me.

G, you are free to think with just as much conviction as I am, and you do have more conviction than most. I also see a very fine line between convention and narrow-mindedness. You likely see me as a bit of an anarchist, and I see you as a bit simplistic. That's just the dynamic between people that think differently. PoTAAtoe, PoTAHtoe.
-----------------------------------------------------------------------------------------
As usual on Rennlist, there are a lot of people attributing things to me that I have never said. I have never said I was definitely going to race this car. I have never said that my ideas were better than anyone elses. I have never said that the convention does not work. I have never said that this is THE way to build a cage. Everyone is free to think what they want. I can't control if they can think, or how, or about what, but IF they think at all, then this thread has been worthwhile.

I did not set out to build a conventional, bland cage because that represents absolutely no challenge to me. I do THOSE kinds of cages all the time. I wanted to force myself to THINK and CREATE. Half of me is tempted to cut it all out and start over. The other half likes a lot of what I have done. For those that don't recognize it, that is the creative process.

If you never try anything, then you are just copying. If that fits the bill, or is mandated, then fine. But... anyone can do that.

fatbillybob

I think this is a great leraning thread. Thanks to John for sticking his neck out. Geo, do you think the hortizontal belt bar in the main hoop is really a big deal? I think if the internal of the main hoop should be triangulated at least once with the node by the driver's head if done this way. It offers nice triangulation to the main hoop for the primary job which is rollover. Secondarily the main hoop supports the doorbars. Alot has got to bend to deform the mainhoop from side impact. Also, I have seen many harness bars that are curved between the mainhoop side legs so that there is more room for tall drivers. So those bars are weak to useless except as harness bars. That is caused by the compromise of making the rollhoop maximum height for head clearance on that same tall driver as most rooflines slope back and down.

Geo

And that's cool John. It's your car and you don't need to meet the requiements of any sanctioning body, so that's none of my business. My only concern is what people come away from this thread with when you brought up the point.

John, you can label me all you like. I see a cage as doing two things: 1) Protecting the occupant, and 2) stiffening the chassis. Bonus points for designs that do this with the least amount of material. I consider myself a pragmatist when it comes to cages. No points for style from me.

OK, lame ASCII art, but you basically have designed braces that look like this:

>--<

Nothing is supporting those nodes at the points. No straight tubes. One can justify all they like, but there are to triangles attached to two trapezoids where 4 triangles would be much stronger. But, like you, I'm an arm chair engineer so I can't give you all the technical reasons and calculations behind it. I just know that tubes that are the shortest distance between two points here will be stronger.

I don't see you at all as an anarchist. Sorry you see me as simplistic. Actually, in many ways I see us as quite similar. We both have an absolute passion for safety (we obviously disagree about cages though), and we are both arm chair engineers who don't have the sheepskin, but have a solid understanding of the concepts generally (referring to both of us).

Geo

Purely my personal opinion here, but yes, I think they are a big deal. That is why I added an additional tube at the base of my main hoop as well. The 944 has a somewhat unique design (911 as well from what I have seen) to be able to tie a straight tube across the bottom of the main hoop. Most cars don't have this and I debated it. It's the one tube in my cage I consider optional. I installed it because of the strength it would provide in a side impact.

Think about a cage w/o the belt level horizontal tube. A strike by a car at the the shoulder (seated positon in the car) level w/o the horizontal tube there. The car will strike at or around the middle of a long expanse of tubing and a tube with a now unsupported bend. It becomes a weak point of the cage.

I'm not professing to be a master cage designer here. As John and I have both mentioned at times, cages require compromises. I've put mine out in the open in this thread and in the past. I'm perfectly happy to discuss what I know of cage design and critique any cage, including my own.

As John suggests, I have a relatively simplistic view of cage bulding. I think designing cages is somewhat like porting heads in that it's easier to screw them up than to improve them.