Okay everyone, so my 914 tub was stripped out and sandblasted. Removing the passenger-side rocker cover revealed a hellhole of biblical proportions, but I was neither surprised nor terribly disappointed, as my plans were to chop up the car anyway. My plans for this car are to increase its torsional stiffness by almost an order of magnitude (looking for a target stiffness of ~5-6000lbft/deg) , without really adding any appreciable weight to the chassis. Tubular reinforcements are out, as they provide nowhere near the necessary structural efficiency.. I should add that the car is to be a speedster, with no targa bar, windshield frame, or functional doors.


First order of business is to replace the rotted factory longs, one at a time, with custom sheetmetal longs approximately the same width as the original pieces, but significantly higher, reaching up to the top of the inner door post. These would tie in to the inner door post and the rear of the front wheel housing up front, and to the rear frame rails out back. The idea is that the greatly increased cross-sectional area, combined with the diagonal sheetmetal X-bracing within the new longs should greatly stiffen the chassis where it is most flexible.

A very helpful, albeit simplified way of estimating the deflection of a frame is to figure out the stiffness of all its longitudinal and lateral elements (the same way you calculate the equivalent resistance of parallel resistors in a circuit). This means that if our longs are super-stiff and our-lateral connections aren't up to snuff, then that's where most of the deflection will take place. So the next step will be to completely box in the rear shock tower area. I will be cutting out the rear trunk floor, so I will probably
be using that sheetmetal to do the job. Much attention will be paid to the connection between the frame rails and the shock tower area, as it is a bit of a weak spot the way it is.

The front-end is also due to receive some attention, specifically where the bulkhead meets the longs, but I still need to figure out the details..

Of course, I will also be adding localized reinforcements as-needed (suspension ears, front trunk, etc..), as well as a roll-hoop behind the driver.


I'd like to hear your thoughts on this, specifically on the best way to remove the rotted-out longs and replace them with my "superlongs" without mangling the chassis. It is a given that I will be installing an adjustable bolt-on brace from the seatbelt mount on the targa bar to the upper door hinge mounting, but I was also thinking of installing another bolt-on brace from the 3-bolt outer suspension ear mounting to the lower door hinge mounting, then add another brace between the upper and lower door hinge mounts. Should this be enough to keep the chassis from moving when the longs are removed (one at a time of course)?

Check the classic threads. There are many pioneers that have blazed this trail for you.

Good luck and take pics!

do you have access to a celette bench? if so, that would give you all the suspension pickup points. with some creative bracing, you should be able to take out one long at the time without changing the overall geometry ...

Andy


Click to view attachment

This thread will probably turn into a debate and so be it but essentially what you are trying to do is create a tube chassis with a only a few tubes ,box's and the factory sheetmetal connecting everything else. The 914 would not fare well to the modifications you are looking at.

Not an attempt to discourage you but at least enlighten you on what to expect.

I think you have a good idea of what and where to strengthen a chassis but with a unibody to much focus on the center section will tear the front and rear sections apart. Need to look at the 3 sections (rear, center, front) of the chassis and the dynamics involved with each independently and then all 3 combined as a system.
Begin at the center section and work outward.

The tunnel would need to become a structural backbone in proportion (height, width) to the longs. This is not hard to do but the weak point would be the forward and rear bulkheads where they would need to tie in to the tunnel. At the center connection at the tunnel/ bulkheads the load would need to fan out and be distributed outward to the longs, suspension pickups and towers. Not so easy to do with a 914 due to component packaging. It is easy to do with an 8 point cage.

Try to picture an hourglass laying flat and looking at it from above. Then drawing two parallel lines from on each side chamber to chamber. You end up with a rectangle with an "X" through the middle. If the front or rear sections do not flex (controlled) in harmony with the center section they will fatigue. To overcome this you would need to make all 3 sections equally rigid laterally, longitude, shear, and torsion. This can be done but you start gaining weight fast. The more weight that goes higher also changes the center of gravity and completely changes the handling characteristics that were designed into the suspension. Are you beginning to see where this is going?

I actually considered something similar to what your idea is. Fabricating 10x4x 1/8 box section in the longs, firewall and front bulkhead. Basically a big rectangle. Modifying the tunnel to 10 inches wide and 8 inches high with 2x2 square tubes. At that point I knew the front and rear sections of the chassis would not survive this marriage without significant structural modification. Adding an 8 point cage the next factor in the equation was weight. To strengthen the other parts of the car I had to add metal and remove where metal was not needed. At this point I was looking at a monocoque or full tube frame. After much thought the concept was a diminished return. I have a friend who worked some figures on CAD chassis software and the numbers were fine. The weight was the killer.

The fork in the road was restore my longs and add an 8 point cage or go full tube frame. If I had the budget I would have gone full tube.


I will share some photos that hopefull will give you a decent perspective
on tubes and box's. If you really absorb the photos you will begin to understand chassis dynamics.

The first photo is Hari's 914 chassis. This car weights 1730 lbs ready to race.
Lot of small diameter tubes criss crossing and this car is stiff. All that remains of a 914 is the hinge pillars and windshield frame.

The rest are all mid engine tube chassis.

Here are some more pics of 914's in tube format. Some more extreme than others.

Here is an excellent example of a semi-tube frame 914 with major modifications to the front and rear sections. Not much remains of the original car.

Not a 914 but a full tube Fiero chassis. Similar dimensions to the 914 and packaging. This guy did a first rate build - - I think was built at home. Excellent metalcraft skills. He considered the center section route and ended building an entire chassis from scratch.

I do see some weak areas between the firewall and rear suspension mounts.
The bay that surrounds the motor could use some more tubes and reinforcement.
Visually to the eye it is clear that something is missing in the design.
500 pounds plus is going to be caged in that area. Needs more meat from the roll hoop to the rear of the chassis.

Doesn't it really matter most what the final intended purpose of the vehicle is for? I can see full tube chassis for road racing, Baja , etc. Hydra doesnt say what he's going to do with this vehicle other than make it into a "hop over her door" speedster. Do you think "normal" driving will cause the front and rear boxes to snap off?

First off I'd like to thank you all, Jeff in particular, for your responses.


To answer your question Andy, no I do not have access to a celette bench, but I was thinking of fabbing up a 4x4x0.120" frame connecting the outer rear trailing arm mount to the front crossmember mount, and going across the car from side to side.
This frame would bolt on to the chassis, and when used with an adjustable seatbelt mount to door jamb brace, should be enough to support the chassis with the longs out (one at a time), This would all be done with all 4 wheels on the ground. Is my methodology correct or would I end up twisting the chassis into a pretzel?

Oh and Jeff,
I'm well aware that the local stiffnesses need to be matched. This is most significant at the front end, like you said. If the longs are stiff enough and their connection to the front bulkhead isn't stiff enough, the spot welds holding the two will quickly fatigue and rip right off. Much attention will be levied to this area, as I believe it to be the second weakest/most-flexible part of the chassis after the longs. As for the center tunnel, it need not be anything more than a decorative cover if the longs are stiff enough...


On a slightly unrelated note, check out the picture of the Alfa Romeo 33 Competizione , which is conceptually similar to what I wish to achieve here with the 914..

Oh, and I should have mentioned that this is to be a "clubman" race car, to be used in autocrosses, low-speed circuit racing, and maybe the occasional hillclimb.

Great posts and photos/drawings!!!

I can't wait to see how this all turns out!