Bolt in cage vs bolt in roll bar? A cage typically extends to the front cabin so a Targa so equipped, IMHO, would be stiffer than a same year coupe. A bolt-in roll bar will typically provide less support than a cage.

Let's try to keep this thread clear of auxillary pipes and tubes as it was meant to address modifications to the chassis w/o using a cage/roll bar.

Regards,
Sherwood

Agreed.

And I would also ask people with non-911 questions (996 or 997 based cars) to start their own thread if their questions are no immediately answered...

Blue72, weather was terrible and work was worse so didn't get to Silverstone. Am at Donington with them early next month instead. Anyway, I don't think the 996/997 RSs continued the seam welding, but I have emailed Porsche Engineering for clarification.

I hear what Randy is saying, but I might post the answer here when I hear from the factory, as I think/hope it will be both relevant and interesting. Might take a while to get an answer though...

Hey Guys,

Wondering if someone can explain to me (type slowly and use small words :)) the purpose of the added brace between the pickup point on the torsion tube and the rear seat area of the tub. I stole the pic below from an earlier post on this thread.

I am not questioning the value of this mod, and plan to do it to my own car since I am doing lots of welding on the torsion tube anyway. But, I'm having trouble picturing what forces are at work here and how the added piece helps.

Also, anybody have a tip on what thickness to use? I have some 3/32" plate sitting around, which seems reasonable considering what it would be welded to on either side.

Thanks
Scott

the added brace you mention is to strengthen the torsion tube. the best way to think of it is that the tire is exerting the force on the chassis, the suspension mounting points need to be reinforced so the forces are absorbed by more of the unibody and not just the formed sheetmetal immediatly connected to the trailing arm. chassis failures always start and the point of most impact/fatigue. to some degree the suspension pickup points are usually this point. if these areas are to weak (or not designed for stiffer springs, sticky wide tires etc..) the rest of the chassis really is not doing enough to add strength. the max force the rear tire is putting on the chassis is usually the tire pushing into the chassis which if the spring plate acts as a pivot, the force is pushing in on the torsion tube pickup point. that means the added plate in the picture is acting in shear. the seat bucket is fairly sturdy due to the shape of the stamping so it is a good spot to anchor to. i would use metal that is close to the gauge of the sheetmetal you are welding to, i don't think 11ga will be stronger than 18ga in this application because you are depending on the 18-20ga seat bucket for strength.

Thanks, just what I needed to set me straight.

I figured that the plate would have to be in shear to do any good. When I think of the torsion tube I always focus on the twisting forces from the action of the torsion bars themselves. Obviously those aren't the only forces at work.

Cheers
Scott

Just for fun windshield related trivia, aluminium and glass have almost identical Young's modulus & density. Glass has a yield strength about 10 times higher than plain jane Aluminium (though obviously less allowable plastic strain...). As my extremely nerdy and smart ex-boss Keith told me as an aside (with a perfectly straight face), that's why they have fiber-glass and not fiber-aluminium ...

Now aluminum is 1/3 the stiffness of steel ... so a 5mm windshield is equivalent to an almost 2mm steel sheet. Back in my test days, if a buck had no windshield, we'd always coarsely rivet 2mm sheet steel in place. From my experience, the windshield contributes significantly if it's glued in.

Measuring torsional stiffness - everybody does their own method, even different groups in the same corporation ... 2 ways seem to make sense. Either pick axle centerlines, or pick where the vertical loads enter the chassis. The 911 vertical loads come in at the front steering rack crossmember and at the rear torsion bar tube (that's why those bushings always get off center), unless the car has coilovers. The rear torsion bar tube stiffness also has a small contribution, multiplied by the distance to the wheel centerline.

Mine creaks a lot going over steep driveway ramps, so I'd guess the 911 is pretty soft.

Outside of glued in glass, a bigger closed-section tube down the center bulkheaded into the front steering rack crossmember mounts and the rear torsion bar tube seems like the easiest way to increase stiffness for the determined fabricator... and the stiffness increase at least is easy to calculate using classical mechanics.

My $0.002 (inflation takes its toll)

Yep, sounds about right.

To update where I got to since suggesting this a couple years ago in this old but good thread:

The early cars don't have enough area around the windshield to glue them in properly and make this a drop-in mod. Instead, you need to weld in some extra steel to have enough "landing area" to glue the 993 windshield in.

The risk, and the reason we ended up not going that way (yet) on my car then becomes cracking. The later cars, with structure specifically designed to glue the windshields in, sometimes still have this problem when you race them. The thing that talked me out of doing this to my '69 (so far) is the fact that some aluminum-framed newer Ferrari's were cracking windshields every other race because of excessive chassis flex, and of course it's a PITA to R&R them after they are glued in. In other words you need a very significant amount of stiffness before you can safely glue the windshield in.

Even with FEA I don't know how much stiffness that is, so we've elected to start without a glued in windshield and measure how much flex I get in that area. We currently estimate that my '69 chassis is up into the 10k-15k ft/lbs per degree range, which is maybe (probably?) enough, but it all depends on where the flex is, etc.

I'm now convinced that applying this mod to a stock or lightly stiffened early 911 (~2k lbs per degree), while a nice idea, would be a recipe for cracked windshield if the car has stiff suspension and is driven hard. No evidence to back this up, though (yet)...

-Pete

"a bigger closed-section tube down the center bulkheaded into the front steering rack crossmember mounts and the rear torsion bar tube seems like the easiest way to increase stiffness for the determined fabricator... and the stiffness increase at least is easy to calculate using classical mechanics."

- It would be nice to see something that is stamped out by the 100's or 1,000's that can be welded to the existing unit body and provide a nice stiffness increase...

Doesn't a roll bar, or better yet, roll cage do most of this stiffening?

I don't think one would want their car too stiff; I can just imagine cracked windshields, and all sorts of harmonic disturbances that could rattle apart transmissions, engine internals, and even wreaking havoc on Webers/PMOs.

Nonetheless, I've heard acid dipping is a good way to stiffen the body. I don't know how or why, but it's what I once heard.

David - look at the top of this page (or last one) - the whole thread is devoted to non-roll bar or cage items.

acid-dipping sounds weird as it removes material (even a thin layer of metal) - maybe it slightly alters an intensive property and became an old wives tale...

"I don't think one would want their car too stiff;...."

With all due respects, tell that to the good engineers at Bughatti:

http://youtube.com/watch?v=9hEZc4JMMsk

....or Brilliance:
http://www.youtube.com/watch?v=7e9XuG0WYS8&NR=1

For a comparison, the chart I posted on page 1 of this thread contains this data for chassis stiffness:
Veyron, 60,000 NM/degree
Porsche (early), 2,719 NM/degree

Sherwood

Oh - OK. Sorry, Randy - my bad.

I will look into this, though. My roll bar ads 50 lbs to my car. TRE's cromoly bar is a scant 30 lbs.

But just to touch a bit on roll bars, I've often thought mine does nearly nothing to increase my car's structure, and with that in mind, is dead weight - save, of course, for the protection it provides. It's an AutoPower that bolts to the floor, not the door sills.

Can you or Sherwood add to my thought floor-mounted roll bars don't strengthen the structure as well as those that mount to the sills?

Thanks.

Not for the faint of heart but a 964 floor would do the trick.:eek::D

http://forums.pelicanparts.com/uploa...1199761611.jpg


http://forums.pelicanparts.com/uploa...1199761637.jpg

wow!!! no kidding, that is a serious upgrade, Teo's car?

Yes.

very cool, did it raise the stiffness of the car as expected? I'd love to be able to have you guys build me one in the future, wow, what a blessing to have guys with passion building a ride for you. From what I've heard from him since he moved over the pond he is happy with her.

This was not done to make it stiffer, but to accommodate the G50/31.

Yes it is stiffer but......

To be honest it would be easier to backdate a 964 and get all the other benefits.

Scott

In a more or less stock 911, the ends of the rear torsion bars attach in the center of the torsion tube. When the rear tires go over a bump, the bars try to twist the whole torsion tube. The main attachments for the torsion tube are over at the sides where they poke through the unibody side rails. Over time (usually a long long time) cracks can develop in the sheet metal over there. At some point Porsche welded a plate forward from the top center of the torsion tube to the sheet metal about where the rear seat passengers' innside knees would be (if there were any such passengers). I assume this was to reduce the twisting out at the frame rails. This reinforcement is hard to see, but it is up there on cars after some production year.

Or the Factory might have intended to reduce flex in the torsion tube induced by the inner trailing (bananna) arm mounts.

I guess both would count as chassis stiffening, though the winding up of the torsion tube under bump loading would be more like having a slightly less stiff torsion bar, and thus ought not to be as destructive to good handling as the usual issue with flex - unwanted movement of the attachment/pickup points, with unwanted - or at least uncontrolled - changes in geometry.

Now the car in the picture looks like it needed more help here than most, since it looks to be set up to take a 930 trailing arm, which is shorter, which means the mounts have to stick out farther from the torsion tube, and thus can more easily twist it. Those rear seat buckets are strong enough to accept a seat belt mount, so they can do some good. Plus, the torsion tube in the picture looks like it has been cut out and rewelded back in. A guy might do that so he can remove the rather heavy steel slug in its center if he is using coilovers. But maybe that slug stiffens up the tube as well, so some compensation is in order. I'd like to get rid of that slug on my coilover track car, but that's a whole lot of work.

I am dubious that any stiffening of significance (I think this thread was started at least in part by owners of the rather more flexible Targas and Cabs, who really can benefit by more stiffness) will be achieved on a stock or near stock coupe by adding something in this area.

DD - apologies accepted. Now write on the board 1000 times: you can never have a chassis that is too stiff. Mind you, the handling of Sherman tanks was not enhanced by their extreme stiffness, but that's a different matter. Springs and bushings and tires and seats are what should do the isolation of lifes bumps from our backsides, not flexy chassis. (Is chassis the plural of chassis?).

Walt Fricke

Good eye Walt.
This was in fact a G50 torsion tube being grafted into a '72 targa tub which was being converted to a speedster. It was also going to coilovers. The trailing arms were SC however not 930.

These mods were done on many RSR's first by the factory and then added by the teams that did not have them from the factory.