"Sub Frame support connector" from Z-Specialties

[jmortensen]

Haven't worked on too many convertibles, so I can't really speak to this. I can say that the 944 convertibles have a double floor pan like a sandwich. I have also seen the bottom brace for Miatas which connects the frame rails underneath.

My take is this: On the top side, any opening in the cage that is very large should have some sort of gussets. The larger the unreinforced hole, the more likely it is to flex. On the bottom of the car it should be no different. There is the vertical panel above the rear control arm mounts which is part of the torque box that is back there, and the rocker panels and that's pretty much it until you get to the front crossmember in the stock vehicle. The trans crossmember doesn't connect the subframes together. Even if you add the subframe connectors, there is still a very long span of unconnected frame members. I could go out and measure, but I'll guess 6' long and 3' wide with no cross braces or anything on the whole span.

In Herb Adams's Chassis Engineering he shows how an X brace across a ladder frame improves torsional rigidity, and also shows how a NASCAR style roll cage on top does not necessarily add much torsional rigidity to a ladder frame. P89 fig 12-12 and p90 fig 12-15. This is all done with balsa wood modeling, but it proves the idea of the X on the bottom is sound.

[John Coffey]

In Herb Adams's Chassis Engineering he shows how an X brace across a ladder frame improves torsional rigidity, and also shows how a NASCAR style roll cage on top does not necessarily add much torsional rigidity to a ladder frame. P89 fig 12-12 and p90 fig 12-15. This is all done with balsa wood modeling, but it proves the idea of the X on the bottom is sound.

That's all well and good for a tubular or finite element structure, but we are talking about a monocoque structure where shear panels comprise most of the strength of the structure. Even a roll cage relies on these shear panels for its strength at the bottom. You Porsche 944 convertible is aperfect example of this. They just added a second layer to the floor pan shear panel.

A 240Z floor has 3 longitudinal structures (rocker panels and trans tunnel) and one lateral structre (front seat mounts) with shear panels connecting all of them. Add in two more longitudinal structures (SFCs) and now you have longitudinal members with one semi-crossmember all reinfoced by a shear panel.

Again, as I've said before, adding two additional tubular structural members in an "X" to a structure that has 5 longitudinal members and a shear panel is just adding weight. That weight is better applied in at least half-a-dozen other places on the chassis.

[jmortensen]

It sounds like what we're disagreeing on is how effective a shear panel the floor and trans tunnel is. I would suggest that it isn't very effective, because it is so thin, and the trans tunnel part makes it a lot weaker in shear than if it were a solid floor all the way across. It's not a bead rolled in the floor for strength, it's a major curve in the very thin sheet metal. The seat mounts don't connect across the tunnel, if they did that would make the floor much stronger. After much hammering on my trans tunnel I can say with some amount of conviction that the tunnel is easy to flex and easy to bend. I would want a shear panel to be SPRUNG so that it could be effective in shear, and the tunnel just isn't very sprung in my estimation. Shear panels can't be effective if they're "loose", so far as I've experienced them.

The subframe connectors, like the rockers, are primarily adding beam strength, not torsional strength. A small amount of one comes automatically with the other, but they are very different things and to get both you need to design for both in my opinion.

You may ultimately be right that the weight might be better applied somewhere else. I am just saying that I would like to see it tested to know for sure. I've seen a procedural manual for torsional and beam testing, and it is not very easy to do, so I'm not holding my breath for someone to go out and test it. Instead I'm saying that in THEORY, the X should be very useful. I don't know what people are going to make the X from, but the one that I had planned was going to weigh about 7 lbs. You just can't get weight lower, and the 7 lbs would be quite a bit lighter than the strap that the guy is selling in this thread.

Again, I don't think it will work as planned due to ground and exhaust clearance issues, but if I find that I can make it work I would definitely bolt it on and try it before poo poo-ing the idea.