Billet Aluminum Chassis Cobra

[masraum]

Pretty cool.

http://www.kirkhammotorsports.com/book_aoe/

I'd heard of this company that makes Cobras in an old Polish aircraft factory, but I didn't know that they made the chassis components out of billet Al.

[Racerbvd]

Sweet!!!!

[911pcars]

Apparently, only if your name is Ellison and you own a company called Oracle.

Sherwood

[911pcars]

Originally Posted by 911pcars

Apparently, only if your name is Ellison and you own a company called Oracle.

And I'm wondering what the total weight is of this one-off? Realize this is an exercise in creating an exotic hot rod. If they wanted to use state-of-the-art building techniques, they'd probably use other building methods and materials.

Ellison will probably drive this once, then park it in the lobby of his corporate headquarters; a good tax write off.

Sherwood

[Schumi]

For some reason, those pieces on the floor aren't quite screaming 'torsional rigidity' to me.

[MattKellett]

Really sweet chassis.

But really, can anyone else say Mclaren F1 rip off?

[Shuie]

Why billet instead of DOM tubing?

[masraum]

Originally Posted by Schumi

For some reason, those pieces on the floor aren't quite screaming 'torsional rigidity' to me.

The book (various links on the page above) references Porsche several times.

http://www.kirkhammotorsports.com/book_aoe/aoe_03.pdf

They told me about the 1971–1972 Porsche 917 chassis that were made out of aluminum and prone to cracking failures. To predict the failures, Porsche welded Schrader valves into their chassis tubes and mounted a gauge onto another bung in the chassis. Before a race the team pressurized Chassis components are doweled together—like an engine’s connecting rod. The bolt then passes through the dowel.
the chassis with air; every time the car came into a pit stop, they checked the pressure gauge. If the chassis lost pressure, they knew they had a fatigue crack somewhere in the chassis. Porsche engineers are very bright; if they thought aluminum could save them weight, then I reasoned I should be able to use it as well.

By analyzing an original chassis, we discovered the original 427 Cobra chassis had a stiffness of 1450 foot pounds/degree of deflection. Analysis of the billet aluminum chassis showed a stiffness of close to 4500 foot pounds/degree of deflection, or a 300% improvement over an original chassis (actual stiffness is a little lower because we did not perfectly model the bolted-together joints).

I find it interesting that they talk about Porsche in the "Chapter 11: Half Shafts" section too.

[gtc]

Bling.
Not especially light weight or high tech. Just looks kind of neat.

[masraum]

I've only perused a few of the "Chapters". I'm looking at the "steering rack" chapter now. They bought a rack and pinion rack for the car and found several problems that they started trying to fix, but eventually just trashed the rack and decided to make their own from scratch.

[Schumi]

Originally Posted by masraum

The book (various links on the page above) references Porsche several times.

http://www.kirkhammotorsports.com/book_aoe/aoe_03.pdf




I find it interesting that they talk about Porsche in the "Chapter 11: Half Shafts" section too.

Wrong wrong wrong. Not the porsche part, but the part where he thought that there is some similarity between his machined, non-efficiently loaded aluminum frame could be compared with a aluminum tubular space frame of the 917.

The tubular space frame connected suspension and engine nodes with welded aluminum triangles, ensuring wonderful load paths.

Having chassis that is large, squareangulared chunks of billet aluminum bolted together, no matter how beautifully pocketed they are, is no where near as efficient.

[911pcars]

I must agree.

According the the quote:

"By analyzing an original chassis, we discovered the original 427 Cobra chassis had a stiffness of 1450 foot pounds/degree of deflection. Analysis of the billet aluminum chassis showed a stiffness of close to 4500 foot pounds/degree of deflection, or a 300% improvement over an original chassis (actual stiffness is a little lower because we did not perfectly model the bolted-together joints)."

Here's a chart I cobbled together a couple of years ago on a topic on 911 chassis strength/reinforcement.



Looking at the numbers, one can see how flexible the Cobra chassis is (1450 ft.lbs/deg. before and 4500 after) compared with even a modest passenger car like a Dodge Neon @ 6,000 ft.lbs/deg. much less state-of-the-art supercars like the Veyron at 44,136.

Yes, this was an exercise in manufacturing prowess (very effective) and producing a one-off collector's piece. I'm sure the company and Ellison are happy with the end results.

Sherwood

[Schumi]

One very important thing I see, when looking at that chart, is that since a lot of those numbers are factory specifications, they are most likely all measured in different ways.

Real torsional rigidity of a chassis is measured as the centerline of the axles, with the loads applied at the hubs. This would require rigid shocks, or, on our formula cars, we use 'shock bricks'- steel rods in place of the shocks- which are mounted inboard, on rockers to pull/pushrods. On a road car this isn't very plausible to test, so alot of numbers are pulled out of either simulation or thin air.

The key is that just putting two solid mounts on the chassis and twisting it is not loading it in the way that it is in real life- through suspension components. So a lot of numbers that people quote are just total BS.

[jyl]

Since this was a cost-no-object one-off, why didn't they make a triangulated space frame? Would have been way cooler (think Birdcage Maserati) than an old ladder frame. It's not like there's anything "stock" about that Cobra anyway.

[Schumi]

Back on the Cobra-

I looked at some more detailed pictures and found even more things that I would not consider particularly great ideas. It looks like a fair amount of those aluminum frame bits are bolted together... bolted joints are inherently not the greatest means of joining chassis bits- this is why aluminum monocoque racers, like the 956/962, are riveted together, like airplanes. Bolted joints can fail in a multitude of ways, and add to an already lacking torsional rigidity. But, if done properly, can still function at a high level- airplane bulkheads made of aluminum or titanium are bolted together.. right?

The main flaw here is that the Cobra's structures- where bolted together- are tapped holes. Tapped aluminum holes. Wherever bolted joints through aluminum are used on jet aircraft, the holes are through holes, and tight tolerance, self locking NAS hardware is used. This Cobra is designed with socket head cap screws torquing into tapped aluminum holes. Socket head cap screws do not have nearly the tight tolerances that NAS hardware has- thus attributing to more compliance - again with the rigidity. And also, they are not able to be torqued to apply a high amount of preload- necessary to press the surfaces together tightly and use clamp loading to lessen compliance in the fixtures. Because they are threading into aluminum. Where a non-locking nut or, in this case, no nut is used, then it is imperative that at least the bolts are head-drilled and safety wired together to prevent them from backing out due to vibration- vibration that this Cobra will surely be producing.


Again, it's a wonderfully machined work of art, but not a performance engineered machine. I need not get into how Solidworks is not a suspension parameter design tool, either...

[herr_oberst]

Originally Posted by jyl

Since this was a cost-no-object one-off, why didn't they make a triangulated space frame? Would have been way cooler (think Birdcage Maserati) than an old ladder frame. It's not like there's anything "stock" about that Cobra anyway.

Or a bonded honeycomb / aluminum sandwich torsion tube ala Lotus. Or even a riveted aluminum / honeycomb sandwich monocoque ala any F1 car from 40 years ago. This just screams look at me. Sort of like the Audi some Arab had covered in diamonds a while back.

[jim72911t]

The best design for a chassis? Obviously not. But as a manufacturing engineer who does this sort of stuff for a living, I gotta say I like it.

Some of the solutions they came up with on the manufacturing end are pretty cool. (see the section on the main frame rails) I've done this kind of stuff before, using similar techniques, and believe me, it ain't easy.

$4500 for a book though? The binding would have to be cnc'd from solid gold before I paid that!

[Shuie]

It all made sense when I went back and read who commissioned the build

[RWebb]

Originally Posted by herr_oberst

Or a bonded honeycomb / aluminum sandwich torsion tube ala Lotus. Or even a riveted aluminum / honeycomb sandwich monocoque ala any F1 car from 40 years ago. This just screams look at me. Sort of like the Audi some Arab had covered in diamonds a while back.

- Those would be great for stiffness/wt. but I bet it's cheaper to pgm a machine to carve one out a chunk of Al than to create more complex fabricated structures like that.

I bet they next offer a limited series of 500 or something...

[javadog]

You guys may be missing the point, here. Think of it like you would a '32 Ford hot rod. Someone had an idea to do something that hasn't been done before, it looks cool and it won't get driven much anyway. The point wasn't to make the best frame they could, or the strongest, or the most technically advanced. These guys build things out of aluminum, not carbon fiber. Tube frames have been done, ad naseum. Sheet metal frames, ditto. Carbon fiber chassis are available too. But, billet (the magic word for non-technical types) aluminum is new for this application.

I like it for what it is. I wouldn't build one but I can appreciate the thing. I don't like how it's put together either but then I also don't like the standard Cobra they build, for the same reasons. I can tell you, I'm in a minority. Most people take one look at the polished body and matt finish stripes and start drooling...

JR