"The crossbrace over the gas tank would seem most effective for a torsion bar based car where the loading is along the longitude and less effective for a coilover based car. I would think that the coilover car benifits most from the bracing between the shock towers."

In retrospect this statement I made is almost total BS. It only takes into account the loading from the springs. Obviously when cornering the chassis is loaded at all three pickup points per side in the front suspension.

Sherwood, I take it that your looking for ways to strengthen the chassis with out cosmeticaly altering the car hence no roll cage. Is a roll bar acceptable? If it is then you can tie it to the B pillar at the seat belt bung. Tyson Schmidt did this and he said the results were diffenently noticable.

Also, how about a triangle welded into the A pillar where it meets the roof? You'd have to peel back the roof liner and pillar vinyl to install but it might be worth it.

Keep in mind that I'm no engineer which should be obvious by now. I'm just throwing out ideas and the engineers on this board can feel free to point out the hairbrain ideas when they see them.

Geoffery, always does things halfa$$ed :) Its amazing what he accomplishes in that small garage.

Bobby,
Thanks for the suggestions. A roll bar or a roll cage will greatly improve the torsional rigidity of most any chassis. However, in a street-driven car, I'm not convinced the security of a roll bar or cage is worth the potential hazard of a possible helmetless head/roll bar collision. There are two sides on this. Rennlist.com has the most current discussion, but we can save that for another thread.

Thus, I'm looking at some DIY methods that are not visible or not quite invisible. Strengthening the B pillars is an idea, as is some floor reinforcement and some stitch welding along some selective chassis seams. Intuitively, A trunk brace in the space over the fuel tank seems like a deserving area, yet with adequate access to spare tire and battery.

Sherwood

Coming into this thread a bit late. Is there a difference between a 964/993 C2 and C4 tub? Seriously, aren't they the same chassis? I thought so...

Sherwood,
would this help? Download the image draw onto it and repost where the stifness would be added.
Bob

http://forums.pelicanparts.com/uploa...1137041054.jpg

I'd be interested in hearing exactly which seams one would seam weld over the spot welds... just the longitudinal ones in the middle of the car? or a lot more than that?

I can't predict any of the following will accomplish the objective of adding minimal weight for maximum effectiveness, but here are my best guesses (maybe WAGs), to improve chassis rigidity.

Red lines indicate chassis reinforcement plates. The trunk brace is somehow removable for spare tire and battery access. The rear strut tower reinforcements are the typical RSR mods; maybe not needed with torsion bars:
http://forums.pelicanparts.com/uploa...1137508707.jpg

Do you see any spot welds here? Stitch or seam weld them for add'l strength. BTW, letters and arrows are point-to-point chassis measuring points. These specs determine if the chassis is straight:
http://forums.pelicanparts.com/uploa...1137508975.jpg

Here are individual floor pan pieces. With the exception of the bolt-on pieces, are all these pieces seam-welded by the factory? If not, stitch or seam-weld for add'l strength. Later torsion bar tubes, for example, have add'l bracing:
http://forums.pelicanparts.com/uploa...1137509028.jpg

Here's the inner rocker (no. 4). Part no. 7 is a reinf. piece from Stoddard's catalog:
http://forums.pelicanparts.com/uploa...1137509063.jpg

... and a couple of more Stoddard replacement pieces; could be used to reinforce these areas:
http://forums.pelicanparts.com/uploa...1137509162.jpg

These mods might total less than 100 lbs. Some of the above can be accomplished on a finished tub. Accessing the bottom might be easier on a rottissorie.

Some areas to think about. Comments?

Sherwood
http://members.rennlist.org/911pcars

That about covers it - except for the roof structure. It would be nice to know what the most critical areas are... I doubt we will ever know, unless somebody has access to a finite element force vector program designed for cars...

TRE, Pdoc and others might post what they usually do... that will be better than a WAG.

I've got the FEA software . . but getting the surface info could be tough. ;)

I wonder if the 993 chassis was still basically the same as earlier ones including inner fenders?

Are the outer fenders interchangeable between 993 and pre-89?

"I wonder if the 993 chassis was still basically the same as earlier ones including inner fenders?"

Basically the same, yes; insofar as rear engine, RWD. I suspect Porsche chassis engineers weren't sitting on their keysters drawing the 993. That they are faster and have more speed potential is not only because of a larger engine and better suspension.

"That about covers it - except for the roof structure. It would be nice to know what the most critical areas are.."

Cabriolets used somewhat the same rocker panel reinforcment, along with perhaps thicker or higher strength metal to compensate for the lack of a roof.

Just to remind - I'm aware that the addition of a complete roll cage makes all this discussion moot. I just want to explore what it might take to strengthen the chassis (and thus improve handling) without resorting to visible race car-type add-ons.

Sherwood

What about cutting some box tube or rollcage tube in half and welding it onto the bottomside of the car from front suspension to rear suspension? More trouble than its worth or something viable? Then you could weld some bosses onto the chassis and bolt a flat panel on the bottom of the car from the front splitter to the engine/trans area.

Getting ahead it seems.

Short of tube framing substructures into the car, seam welding, adding reinforcement plates and some triangulation on the front end seems to be the best bet. I really do wonder how much some of the cages on the market short of a custom weld in that ties to the front and rear actually stiffen a car.

I just did some research. You are wrong. Only 996RS were built on C4 tub, not 993RS nor 964RS..

How easy (or difficult) is it to do this? Is it just a straightforward swap?

Structurely -- adding material to the center of the car will have relatively little affect compared to adding it to the paremeter. It might help some with regards to bending motion, but in regards to torsional flex, it won't do much.

I seem to agree with you John, but no one ever seemed to tell Chapman that. Have you ever chased a lotus around the track :)

I think a strengthened tunnel would help longitudinal strength a lot if not doing much for torsion. A large box is harder to twist than a smaller one though.

Luke;
There are a number of successful cars (Lotus Elan, TVR's both come to mind) that used a "center backbone" chassis. In the case of a TVR, the outriggers are key to the stiffness. Also, if you make the car light so the stresses are less, that makes a big difference too. But if you've got a 911, and a 911 with a tunnel (as opposed to comparing a 911 with a Lotus), I don't think that the tunnel is going to make a huge difference.

I should update this thread to show an important piece that can be added for improved chassis rigidity.

Thom Fitzpatrick (Widebody) added these pieces to his track car, but I see the advantage in using them in any 911. The additional weight is about 18 lbs.

http://forums.pelicanparts.com/uploa...1143309256.jpg

http://forums.pelicanparts.com/uploa...1143309288.jpg

As it turns out, the factory RSRs used a similar treatment (except double thickness), in addition to a full cage.

Sherwood

First, I'm glad this thread is revived...

I agree that stiffening the perimeter of our 911s appears to be the way to go. Also, some bracing across the car might be good.

What we really want to figure out here, is for a car with no roll cage, what changes make sense in terms of time, cost & trouble (effort).

There are basically three ways to evaluate the effects, the improvements:
1. educated guesses, esp. if made by people with a background in statics and in dynamic stresses. obviously this includes automotive engineers of various types, but also includes people like some civil engs., architects, etc.
2. following factory changes, e.g. the RSR above, and the things done to cabs and targas
3. Modelling efforts, either analytical or numerical, the latter including Finite Element (aka distributed parameter) models
4. Subjective - seat of the pants driving experiences. For this, one would want to try and reduce side effects such as placebo effects, time lag between driving the unmodded vs. modded car, etc. Also, a careful narrative despcription would be helpful of the car and the road.

There are several directions involved:
1. Folding - front to rear
2. Twist - e.g. from the LF corner to the RR - includes a folding component of course...
3. side to side folding - probably not a big issue except as it relates to #2

For the feasability issues there are several factors also:
1. What can be done from inside the car - e.g. seam welding the places where P AG did spot welds. This is nice b/c you don't have to remove all the undercoating and buy a rotisserie.
2. What can be done inside the trunk. Easiets thing of course is a simple tie bar type strut linking the shock towers.
3. What can be done with the engine out. The most common thing is triangles or a bar between the shock mounts. I've asked Tyson in other threads if he thinks this will help on a torsion barred street 911 - and he said no. So I didn't do it when I had my engine out.
4. What can be done if the shell is stripped of subassemblies and its undercoating (street cars, remember) and placed on a rotisserie.

Finally, I'll just remind that there are some old tests where the drivers like the targas best b/c the body was more flexible... apparently the car's handling was more forgiving. IIRC, the tester was a race car driver too. But maybe I don't recall that right....

I'm surprised no one mentioned the windshield. The late, glued in 993 style windshields turn the glass itself into a stressed member, contributing significantly to torsional stiffness. It helps a lot:
Coupes are stiffer than cabs because the chassis effectively works like a tube rather than a U section- and everyone knows a tube is a much stiffer structure. BUT the only thing completing that tube are two thin A pillar posts in a weak parallelogram arrangement with the windshield glass in between.
I've done FEA on my 911 when I designed my cage, and these A pillars are clearly the weak link, as others have correctly identified. Anything you can do here; welding stiffeners to them, etc, will have a dramatic effect.

So a 993 style flush mount windshield update would be my first suggestion. And you getter better aero as part of the bargain...