GSpeed Bolt-in Roll Bar
 

We're getting ready to begin production on our bolt-in roll bars for BRZs and FR-Ss. As many of you are aware, we had the interior of our 2015 BRZ 3D scanned. This has allowed us to create a roll bar design that maximizes interior space while providing a high amount of strength and safety.

Before we get into the specifics, let's go over the design goals for this project.

A bolt-in roll bar:

-is NOT as strong or safe as a welded-in cage.
-provides additional strength in event of a roll-over incident.
-is not designed to be used with OEM seats or airbags.
-relies heavily on strong landing plates.

With that in mind, here's where our design sits currently:

http://gspeed.com/wp-content/uploads...7ia0b52pyg.jpg

You can see the main hoop has to come forward of the "shelf" that forms the rear seats over the fuel tank. In order to get a good "sandwich" joint through the chassis, we have to mount to the lower part of the floor pan. This requires a more reclined main roll hoop than is common to main roll hoops. Because this is NOT a welded in cage, and this is NOT a convertible, this is not subject to the same hoop inclination angle rules as a proper cage.

The harnesses will attach to the rear horizontal bar, and we're planning on including threaded hardpoints for industry-standard harness eye bolts.

Here's a close-up of the main landing plates:

http://gspeed.com/wp-content/uploads...m02hdwoe5g.jpg

And a view from the rear:

http://gspeed.com/wp-content/uploads...egsui92cby.jpg

By mounting the rear landing plates to the wheel wells, we can eliminate many of the problems associated with passing through the seats. We'll still come up with a good block-off plate, but that problem is largely bypassed.

This design shortens the hoop supports, which makes it impractical and largely unnecessary to run a diagonal brace between the two. The rear view mirror would be practically useless with a bar in the way.

You can see the joints at the top of the main hoop supports that allow the roll bar to fit through the doors. This interlocking joint is vastly superior to the "bolt and formed clevis" style from a structural standpoint. It does add some cost, but we don't half-ass things and we're not going to compromise strength to save a few bucks.

With 1.5" x 0.095" mild steel tubing, the final design should be well under 50 lbs.

We are trying to keep cost down, but not at the expense of quality. Our goal is cost-effective, not cheap. This is designed and manufactured in the US, specifically for this platform.

As the design and prototyping progresses, we'll update this thread.

Jake

Interesting design and attempt to keep some of the OEM interior but why not make the main hoop vertical by sacrificing (removing) the rear seat bottoms which are almost useless in a stock Twin and even more so with even this roll bar in order to keep the stock front seats/airbags? It would cheaper for the end user to remove the back seat than to buy new aftermarket seats.

The way the unibody is built on this car, it's impossible to access the backside of that face to bolt through it. There's a large cavity between the top of that "shelf" and the bottom of the car you can't access without cutting.

To add to that, the OEM seats shouldn't have any problem fitting with this roll bar. It's not intended for OEM seats from a safety perspective, but they will fit.

What about anti-sub and lap belts mounting? I assume people running these will want to run 6-pts.

I would assume anyone with sub straps would have fabbed mounts on the floor of the car, no?

Bingo. Done properly, sub straps and lap belts are mounted directly to the chassis with reinforcement plates. Something like this.

I like it so far.
It looks like it retains a nice cargo space when used with no rear seats.

Why no cross bar on the main hoop? I believe that is required with most racing organizations when it comes to tech for cages.

Good question.

Because the main hoop landing plates have to be so far forward, there would be very little room for seats. Like the challenges we ran into when designing our full roll cage, this car's compact design makes for some interesting shapes.

You may notice this shares nothing with our weld-in roll cage. It's not a roll cage, and shouldn't be considered as such. Due to the way the car's built, you can't put a bolt-in main hoop in the same location you would put a weld-in main hoop without significant modifications to the chassis, negating the whole point of a bolt-in. Others have asked if a bolt-in roll bar could be welded in place later as part of a main cage build, and no, this could not. The designs are too different.

Like any other bolt-in four point roll bar, this would not be considered a roll cage by any racing organization. This is meant for HPDEs and Time Trials where you need to run a seat and harness, but don't want to (or can't) install a weld-in cage.

Does that make sense?

Jake

AutoPower and others have solved this issue with the main loop needing to be up on the seat ledge with a welded in plate and bolting to that.

I'd be uncomfortable using any bar that bolts to the floor in these cars as it is extremely flimsy. It would take too much convincing that it won't punch though...

True. At the risk of being too blunt, that's not the problem we're trying to solve here. If you weld in plates and bolt to them, now you've got the permanent chassis mods of a weld-in cage, with all the safety and stiffness of a bolt-in.

With a large enough landing plate distributing the load, you can get some surprisingly large forces.

Hypothetically, let's say we've got a 4" x 6" landing plate. That's got a perimeter of 20". If our material is 0.020" thick (just a worst-case guess, I'm not at the shop right now), that's a cross-sectional area of 0.400 sq. in.

Mild steel typically has an ultimate yield strength of 36,000 psi, which means its shear strength is somewhere around 20,500 psi.

20,500 psi * 0.400" = 8200 lb to punch that one landing plate through, almost 4X the weight of the entire car on that one plate. A 3" x 3" landing plate would only take 4920 lbs to yield.

Now, that's not entirely accurate given the myriad of variables involved, but it does give you a ballpark estimate of the forces involved.

We've all seen those pictures of the Mustang that had the Autopower punch through the floor pan at Hallett. I couldn't find up-close pictures of the landing plates with a 5 minute google search, but the consensus seemed to be those landing plates were too thin, and too small. That's a totally solvable problem, and why we're putting so much emphasis on getting a good landing plate design.

Jake

One has to add in the kinetic energy of the car in motion.

KE in ft-lbs = .5 * mass/gravity * velocity ^ 2

KE of a 2700 lb car hitting something at 50 mph = .5 * (2700 / 32 ) * 73.5 fps ^ 2 = 227,907 ft-lbs

So, if I got that right, just like the bar in that Mustang video, this bar will pop through the floor pretty easily.

You're not supposed to absorb the longitudinal 50 mph kinetic energy with the bolt in cage. Only the kinetic energy of the upside down movement, which is massively less.

Most of the longitudinal energy will be lost with the car sliding and rolling in the gravel.

Like the Alonso crash at Melbourne this year.

MaximeT is right. You're looking at the total kinetic energy in the system, and applying that to one single point in the car all at once. Your numbers aren't wrong, that's just not the right equation for this situation.

Jake