tinman944

I have to agree with James on this one...dom is stronger and more reliable because there is no seam..when you bent seamed tubing the failure is in the welded seam.It may look great when bending and when finished but on impact the seam will split and the rigidness of the tube is shot and will fold in half.
Titanium??? The only thing is does fairly easy is weld ,it does not bend shear or machine very good..
The best all around material for a light cage is 4130 tubing..it weighs the same as steel but you can get away with thinner wall and still be strong.
The bad part is it should be welded by a certified welder and be annealed

Adam Richman

John, I am no expert but it seems to me that the load path on the left or right side of the main hoop considering a car with only a rollbar vs. a full cage would make the main hoop corners pre-disposed to deflecting. I am curious as to why you wouldn't remove the interior tubes on the main, run a diagonal through the x-brace from the d. side top to the psgr. side footer and another diagonal from the psgr. side top to the d. side rear shock tower kicker. In my mind's eye, thinking of a tub w/ rollbar only, I'd want a.) my head to be as close to the main hoop as possible and b.) not see any crushable plane that would comprimise where I sat (looking at the bend on the main hoop to an imaginary plane through the psgr. a pillar in one scenario and seeing both side bars deflecting inward in another) and c.) a easy place to locate the harnesses. I think by doing the diagonal, you could provide a horizontal from the diag. to the d. side b pillar and allow the driver to drop the seat back further (or bring the hoop more foreward).

As I said though, I am far from an expert.

RedlineMan

James & Tinman -

I here what you are saying loud and clear re: DOM -vs- ERW. However, I don't think you are offering an apples to apples contrast here. If you take an equal length of both types of tube and subject it to stress, it is assumed and accepted the ERW will fail first, at the weld seam. GIVEN. That's apples to apples.

If you have a roll bar design such as the Autopower that is a plain open hoop with two rear bracing legs that is made of DOM, and then an equivalent design made from ERW, it is assumed and excepted that the ERW bar will fail the same as above. GIVEN. That's apples to apples.

If you take a piece of paper, it does not have much compression strength. If you take that paper and make corrugated cardboard out of it, its strength increases MANY hundreds of times. That's an apples to oranges argument, but the point is the weak paper can be made stronger through design.

If you take that simple ERW roll bar and add internal bracing similar to my 944 bar, you can easily exceed the strength of the plain DOM version many times over. GIVEN?

My point is not to argue that DOM is not superior, as it certainly is. My point is to say that within a given framework and utilization, and adding in some more esoteric parameters (like not exceeding the strength of the mount point utilized, and so on), ERW can be effectively utilized to make a decent product. I feel that my designs offer more structural support than other products offering less triangulation.

A good debate to be sure, and I value your comments. Does anyone get my point, and then agree with it?

RedlineMan

Adam;

HUH? I value your input, but my brain started smoking after the first couple sentences. Draw me a picture, eh? I'm a visual learner

James Achard

John, I think we are all slightly missing the points of others arguments here, myself included. To clarify your point(I think), you are of the school that a proper design is just as important as the materials used. I don't think anyone here is arguing agaist this philosophy. There are many subtle nuances that make one design better than another which is another whole discussion in itself. The other argument here is the choice of materials, give or take. My feeling is that we need to separate these two issues at first before we can proceed on. As I said before, DOM tubing is slightly more money in exchange for more consistency and strength. In my opinion, this makes using this tubing a win win situation. If ERW tubing was as good as DOM, everyone would be using it. The fact that it isn't and for good reasons as stated in the info above. I am still looking for the AWS(American Welding Society) paper on this, just so you know. With the weight of DOM and ERW the same for arguments sake, why go through the effort of saving a few bucks on a very expensive safety investment. When you start of with inferior materials your end product will never be as good, after all, alchemy was never a successful career.

Cheers, James

RedlineMan

Hi James;

Your concern is appreciated. I am not a learned fellow in many respects, but neither do I miss much...I quite hope! The point you just clarified was never an issue in my mind, as it is quite obvious. And therein lies the rub for me.

If I continue to combine the two aspects of this topic (that you so properly separated; design -vs- material) in ignorant stubbornness, I could say that I make my safety equipment from Iconel, and therefore every other product out there that was made of DOM was inferior because my material was better than your material, REGARDLESS of my design.

To further stretch my point, say I sell swing set kits, and my swing sets come with grade 8 fasteners. They are better than yours because you only use garden variety grade 5 stuff. Does that make my swing set better? It's only a swing set! Who cares? Within its design parameter, the grade 8 fastener gains you virtually nothing in reality.

You and I know that DOM is "better" than ERW. We also know that design is a mitigating factor here, and that "better" materials do not make up for "poor" design. Further, the "poorer" materials used wisely or more extensively may render two products equal in the final analysis and in reality for their intended purpose.

I make niche products. Bolt-in rollbars for guys with non-dedicated track cars that don't want to cut up their car, but want some more "protection." I am using existing factory bolt points to do it. As such it seems to me there are other pressing concerns involved.

I have had people argue with me that they wanted grade 10.9 fasteners instead of the grade 8 that I supply, since that would make the mounting "stronger." True in theory and to the narrow extent that they are thinking. However, the BEST designs are ones that consider far more parameters than the end users will likely ever realize.

The BEST design for a piece of safety equipment is one that distributes load evenly, and does not create undue stress in any areas that cannot handle it. I would guess that everyone would jump on the "grade 10.9 bandwagon" without thinking further because they know that these fasteners are stronger than grade 8, and that this would be "better." However, "upgrading" to the harder fastener would be detrimental if that extra strength overpowered the mount point and caused it to fail.

For that reason, it could also be the case that I could take my design, render it from DOM, and create a product that would fail catastrophically because it overpowered the mount points! I'd have to come up with a weaker design to mitigate the effect of the stronger material!!!

I just have a problem with "bandwagons." "A 2.7 911 engine is a piece of junk." "You can't buy a 944 because every one of them is going to lose the timing belt." You have to have a sub strap because you will slide under the lap belt if you hit something." You can't use a snell "M" rated helmet for our usage because they are for motorcycles."

These are all popular myths that turn into legends, when in reality only some of them contain a grain of truth, and if you never dig deeper you never know the real story, where the myth strayed from the facts of the matter, or which ones are BS.

I think what you and I are doing here is finding out we likely agree on most everything! To the extent that either of us - or any other readers - picks up on something they hadn't before, it is a very worthy discussion. DEEP WATERS!

Adam Richman

John, warning, to rival my uneducateditude on the subject is my skills photo-chopping images but here's my best stab. The design *to me and my untrained eyes* looks like it would do a great job of supporting a vertical load with the car on its roof (the ceter area). My concern is that cars have to get on their roofs first. It may very well just be the angle of the image but, it looks like the central vertical bars and the legs (img 1. A, B, D, E) both angle toward the center of the car. On an impact to the upper b pillar, it strikes me there is not a force acting against it colapsing to the inside (and it leaves a larger gap exposed to deform between the b pillar and the upright - between C-A, D-A). My comments about this case w/ only a main hoop vs. a full cage is that the A pillars are only so strong so pancaking the top (say psgr. side A pillar) gives me visually a plane from that point to point C (img 1) and seeing that as having no force against its colapse is asking that bend to do a lot of work. I am not an expert by any means but have seen first hand (or first elbow I guess ) that tubes with bends have a lesser resistance to bending some more - even if going against the bend.

I was suggesting something like Img 2 (and yeah my artwork sucks) where you would have a diagonal supporting both left and right B pillars and you could use the diagonal between A and B (Img. 2) to hang a tube for mounting harnesses (it may also allows the seat to go back further). Personally, in an ideal world, the main hoop and the main diagonal would tie in to the rockers but I realize this is a bolt-in application.

Img 1.


Img 2.


if it's not clear from the image, I mean for the main diagonal to span the main hoop, the second diagonal to go to the rear shock tower mount point.

RedlineMan

Hey Adam;

Actually, to me, your artwork - flawed as you think it might be - is clearer than your descriptions. Make of that what you will!

You have made a couple of fatal assumptions about the design of my bar that render your ideas impossible. However, trying to decipher your description in the last note set me to thinking. I realized a weakness that I had not seen before, so you are not a total loser!!

Fatal Assumption. The ends of the main hoop ARE NOT attached to the rear body cross member, they are merely resting on it. (NOTE - This bar bolts in with no drilling or cutting, except for the carpet at the end of the hoops. This is under the seat base and does not show when it is reintalled)

If you factor in this bit of info, the other aspects of the design (all the ones you removed) will start to make more sense. By removing them, you removed necessary bracing for the hoop to remain intact if compressed. You are correct about the vertical support I have engineered. Further, I have created an interlocking trapezoidal design that has a pretty fair amount of rigidity from many angles, and spreads load in many directions.

But again... you did set me to thinking about another aspect that I had not previously. That is why we call them PROTOTYPES!

RedlineMan

Hey;

You sparked me to think of these mods. You at least triggered me to thinking about impact to the top corners of the b-pillar area, which is a concern. I realized I had no x-bracing directly to the lower mount, which is the rear outer seatbelt mount.

So we go from this...



To this...



I would prefer it be simpler, but then there is not much simple about the interior parameters of ANY 911.

Did I mention I don't like stuff visible in the rear view mirror?

keith

Hey John - I think that is essentially what Adam was saying - there was no bracing against deflecting from those corners during a roll...

Now I'm wondering why you wouldn't tie the three uprights that hit the main hoop on each side into a closer arrangement - less visual clutter, and distribution of forces down all three legs in the event of a hit up there?

Also, with the additional X, I'm wondering if the little 45degree brace from the main hoop to the lower seatbelt mounting hole is necessary...

RedlineMan

Hey Keith;

And this is where we get into things that people don't think of, or couldn't know if they don't build this stuff.

- Going to the same point on the hoop would put the x-braces RIGHT in the rear view. See my last comment in my last note!

The point it inersects in my rendering does increase "visual clutter" to some degree, but has two advantages.

- 1) It is in direct line with the corners of the roof for stiffness.

- 2) It is a simple end cut to fabricate. The sharper or more complex the end cut, the more time it takes to fit. This CANNOT be underestimated when you are building dozens of them!

keith

It's hard to tell from the image - seems like the center of the X would be below the line of sight from my desk

And 2) - John, I'm certain you have a good tubing notcher, don't you?

Carry on!

karlooz

hey redlineman, is your 996 roll bar in production? have you switched the tubing to DOM? is it PCA and SCCA legal? looks really good especially since you say its easy to remove. that will allow me to remove it for street use.

kurt M

(One persons semi-educated opinion) All bolt in bars are a study in compromise but might offer increased safety. (You could also say all bars and cages are a study in compromise too)

The main hoop base is a floating point and geometrically an L shape even with the added diagonal. It is not attached except by a remotely placed bolt that is both in single sheer and at the end of a flat plate. I would think that this is a design compromise as there is little to keep the bar in place during a (unlikely in real world) side loading event. Is the base of the main hoop prevented from side movement in any way at the base? If not when you load it will want to move and transfer energy to the single sheer bolt. The rear diagonal will help some but if the base of the main were attached somehow everything would improve. Adding X bracing diagonally from top corners of the main hoop to rear bolt point does not help much as in reality you are attaching to an extended flat plate that will just bend and you are still loading the same single sheer bolt.

Strive to make a good well though out product but listen to the potential customer too. Personally I prefer to use a race seat and then not have to think about the straps slipping but if folks want guides add them as an option. You can make them all the way across or make them bolt on into already tapped holes so you can pick the location that fits best. A long flattened hoop with a hole at each end. Provide one for each strap. Drill a row of corresponding holes along the harness bar so the purchaser can select optimum placement. This way you can mass produce and only have to jig drill and tap the holes if requested. This can even be done after paint if you offer stock paint or prior if sending out for paint. Now you have an added sell point and an additional income item. Needed? If it is not unsafe you let the purchaser decide.

Bottom line. People will buy what they feel good about regardless of line drawings, words in caps, exclamation points and bold print. If you are trying to SELL something and folks are hesitant about the material used, or want something added, listen to them. You will make up the somewhat increased material cost in volume. Asking questions is only the beginning, the next step is listening to the answers and utilizing the information gathered.

BTW I like Ti, it makes real purdy bright white sparks when I cut it with the plasma cutter. I just can’t bend it worth a damn.

mpaton

I don't want to hijack this thread, but I would like to a a component discussing the strength or roll bars. John, you talk several times about strength, and I agree it is important. However what I never hear about these days is energy absorbtion. I believe this should still be an important consideration in roll bar/cage design and I KNOW that it was in the 1970s. It's still a big component in F1 crashworthiness today.

When I was at college in the early 70s, I had a part time job at Safety Devices. I never spoke for them and I don't now, but I did learn a lot from the owner and chief designer. I owned 2 of their products, and I tested each of them. I was studying engineering, and I know what a stress/strain graph for steel looks like. In those days, the view was that you could make the bar as strong as you liked, but if you made it too strong, the bar would never absorb any energy by deformation. If you don't get energy absorbtion, then in the extreme case you get rolls and somersaults that seem to go on for ever as the car bounces down
the track, with each bounce being a significant deceleration/impact for the driver.

If you make the bar too weak, then you absorb energy all right, but you deform the occupant space, and you can't do that too much. So like most engineering, there are tradeoffs.

On the materials aspect, I'm prepared to believe that seamless tubing may be more consistent in its deformation at high strain/deformation levels.

Since those days, cars have become stronger in most directions, and faster, and significantly heavier.

But I'd still bet that you can build a roll bar that is so strong that it doesn't absorb energy in some impacts and also increases deceleration forces on the occupants, and gives them more impacts. And I'm also still certain that you can still build a bar that is insufficiently strong. Especially as today's cars hit harder.

So that leaves my engineering mind as unimpressed with protection in dirt oval crashes, and with roll bars whose only touted virtue is strength. Not intending to pick on John; I see may amateur built cages whose owners are very proud of their strong cages.

Opinions?

Michael Paton