964 convertible: how to add stiffness

[FreudeamFahren]

Guys, after years of lurking around and reading numerous threads, all filled to the brim with excellent information, (thanks for that!) I think I finally need your help. I took a look at this beauty earlier today. It's for sale at a trader near Amsterdam. The previous owner traded it in on a Cayenne, as he needed the space. These pictures were made by him, on the day the decals were put on.





It was built in 1990 and has done around 250.000 kms (155.000 miles). I personally like the color / decal / wheel combination, it's got heated seats for the misses and optioncode 220 for me ;-) I can see myself doing a run to the Alps in it next summer already. The paint is not original on all panels, but the chassis shows no sign of any accident damage. The engine's been rebuilt at 200.000 kms (125.000 miles) and it shows. Starts well, runs nicely, pulls strongly and best of all, it's as dry as a 911 engine can be It may not have a stack of bills to go with the maintenance booklet, but a quick call to the specialist who performed the work confirmed continued mechanical TLC since it was imported to the Netherlands in 1997.

There are a few smalls projects around the car: the sill rubbers have seen better times, the rear window has two small tears, the gas struts under the front lid don't work anymore and the hoodcover is missing. Also, the biting point of the clutch was very high and during the test drive, the clutch didn't feel very strong, so I got the feeling it will need a new clutch some time soon. A proper pre purchase inspection will tell if that's really the case. I also hope to find out then, if the suspension is an after market setup, or just a matter of standard shocks with lowering springs. Last of all, the paint shines reasonably nice, but could do with a thorough detailing session. All in all, interesting projects to do, with no real financial "bumps", except the clutch.

Anyway, my question is this: is there a way to add some stiffness to the chassis of a convertible? It does without a front strut brace, so that's point of attention number 1. Are there any other options to make the tub feel tighter, more rigid?

[Wil Ferch]

Here are some threads for inspiration-->

Targa Chasis Flex. See Widebody911 post #4, solution "A" and "B".

Crazy idea #5672 - front stiffening .... especially post #9.

Notice that a Targa or cab is essentially a front "box"....and a rear "box"....attached by(almost) flat sheet metal in between ( see again post #9). Not much torsional or "Twisting" resistance. The "real" improvement comes from adding a stiff box section along the length of the threshold door rockers....but that is extreme (1) Welding and (2.) resulting high step-over. See the first link by Widebody911.

In my humble opinion....Targas and Cabs are meant for nice weather cruising, and not terrinbly serious fast road work ( others here are sure to disagree), so a Coupe should be seriously considered if this is a concern.

Adding the rear "Targa bars" and the trunk strut braces ( or even "X" brace)...help....as would adding a bolt-in rear roll cage....but the real "need" is in the section between the front and rear "boxes".

[RSTarga]

Actually the 964 is pretty stiff in comparison to the earlier 911 chassis. It shouldn't need any additional stiffening for street use.

[Wil Ferch]

I've heard that said too....

A challenge to *everyone* here.....show me numbers from a credible source. Seriously. Been looking and only found an off-handed remark from Mark Donahue in his book about early 911 being 2000 ft/lb-degree.....I find that number incredibly and unbelievably low.... even without super computers used "in the day". Even early 911's were considered very stiff compared to most other cars.

[RWebb]

you want quantitative test data?

or what?

I have Sherwood's chart in Excel now and have a few additions, but it does not include the former beyond Donahue's info, which is in a lot of old threads on this. Email me if you want the xls. file.

[Wil Ferch]

I don't need charts and graphs....just a credible quote or source beyond the ONE quote by Donahiue what the chassis stiffness of an air-cooled 911 coupe is. Thanks. If you have something of interest...pls email me at

[ratpiper71T]

I had that flexible problem, too.

I installed these in the cab recently and the stiffness problem was solved.













Okay, sorry I don't really have a cab...I don't have these either...Sigh...I'll shut up now...

[RWebb]

I don't think there is any question that the 964 is much stiffer than a 911, Wil.

Exactly how much is the question.

[Wil Ferch]

Randy:

There may indeed be a question if a 964 open cab is stiffer than a torsion-bar 911 coupe.....regardless.....I'd like to see numbers from credible sources. Not anecdotal or subjective text. Maybe SAE published something on this topic?

The reason is two fold. During the 80's and 90's...whenever a new model ( of any car brand) was announced....it was typically accompanied by the statement....."new model is 40% stiffer than last year's model". Really?.......year-after-year?, spanning many years? You have to wonder.

Numbers....we need numbers....

And also this--> just how is "torsional stiffnesss" determined and measured?. As torque applied thru an apparatus using the mounting points of the suspension?...as that would be where the forces are applied in real life? It would seem any car can test only thru the actions of where the wheels are. Shorter wheel base cars ( all else equal) should show better numbers. I recall reading about the development of the 550 , that wheelbase control ( keeping it short) was an early, crude way to keep chassis stiff.

[tcar]

Wil, your posts are excellent.

I saw some tests (where? - don't remember) re: cabrio stiffness.

Interesting that the cab was measureably stiffer with the top up than down. I'm assuming that it adds some tension to the front-rear 'boxes'. (Cannot add compression).

This cannot happen with a targa, by the way as you only have the pin connections at the bar.

I restored a Ghia convertible and I could actually feel the added stiffness when the top was up.

[Wil Ferch]

KG.... my first car ( I always thought of it as a "poor-man's Porsche)..... but a coupe...

[RWebb]

Originally Posted by Wil Ferch

... [During the 1980's- on... whenever a new model ( of any car brand) was announced....it was typically accompanied by the statement....."new model is 40% stiffer than last year's model". Really?.......year-after-year?, spanning many years? You have to wonder.

I find that completely unexceptional and to be expected. Here is why: it coincides with the widespread use of Finite Element modelling used for development of unit bodies on cars, along with much faster computers. I used these techniques (for heat exchange modelling) in the late 1970s on animals and they just crushed an Amdahl 470. I used FORTRAN and paid no attention to optimization of the code. I later found out that big car co.s like GM were using such models and had vectorized their code. Later, they had their own little CAE packages and etc.

So eventually, one could just buy the software off the shelf and run it on most computers. Ease of use means greater ability to simulate different solutions, and increased familiarity with the techniques should mean the continued incremental improvements you cite.

I'd bet the Germans have specified methodology for these measurements, tho it may not be the same as what SAE had.

A suitably motivated person - such as yourself - could easily search the SAE journals, and also perform comparative measurements.

For me, I will just save up some $$ and let Dave at TRE stiffen my '73 coupe (someday). He has posted a bit about this before - IIRC, it was the Framboise/Strawberry car he was modifying & I posted to him on that thread.

[Wil Ferch]

40% stiffness increase ......y-o-y......for (say) 5 years yields a front-to-back overall increase of ~ 540%

Even with FEA ( and understanding its power)...I find this rather incredible. I mean , the original cars were "engineered" too....maybe with sliderules....but sliderules got us to the moon and back too !! What were the first-gen chassis made of ?...spaghetti ?

Back to the original poster. Unfortunately all the bolt-on stuff will be things like front tower strut braces ( maybe even the more elaborate "X" brace, taking up some trunk room), and then maybe a rear DAS type roll bar. The unfortunate truth is that what he needs is something like what Thom Fitzpatrick ( widebody911) did in the earlier link I posted.....either a massive "tube" welded into the door opening at a higher level...or an inverted "U" sheetmetal also welded into place. This is drastic and there is no bolt-on equivalent....but is the most "necessary" in terms of where the weakest points are, that could use come beefing up.

I searched the SAE library...and couldn't find a mathematical number for the air-cooled 911 torsional stiffness numbers....maybe longitme SAE members read this and can help.

[RWebb]

Originally Posted by Wil Ferch

... What were the first-gen chassis made of ?...spaghetti ?

...

white ash IIRC

thx for the search of the SAE library

now, would whomever does the torsional stiffness measurements also determine the CG?

"Inquiring minds WANT to know..."

[RSTarga]

the 964 has quite a bit of added reinforcement, starting with the central tunnel. Even the last generation torsion bar 911 (89) was much stiffer than our early cars. The side structures were sturdier from 86 on etc. When the factory offered a turbo Targa it was because they had added quite a bit of reinforcement.

[efhughes3]

I'd buy it, and enjoy it. Then decide IF it needs something.

There are a lot of "well-tracked" Targas and Cabs running around. Unless your name is Pat Long, Marc Lieb, or similar, you won't suffer in your experience. They do make nice weather cruisers, but you can get in some seriously fast road work too.

[FreudeamFahren]

Thanks for all the advice guys. There's trouble in (potential) paradise however. The trader has stopped returning my phone calls and replying my emails, while we were planning a PPI. Which says enough. The search continues.