The XJ Unibody Chassis
Compiled from posts on the NAXJA Modifications Tech Forum by "Alyn" and Ed A Stevens. Edited by Jeff Veach
Your XJ frame is far stronger than the Chevy frame when it comes to torsional stiffness. Torsional stiffness is measured in ft/lbs per degree of twist. A typical steel ladder frame (ie Chevy)
would be rated around 500 to 1,000 lbs per deg. The XJ unibody chassis is likely two to three times that stiff, or several thousand ft/lbs per degree.
The best way to increase the stiffness would be to add a well designed roll cage. The more diagonal bracing, X-members and corner gussets, the better. Also, the more places you tie the cage into the unibody, the greater it will add to overall stiffness. This shouldn't be limited to the floor, but also the door pillars and fire wall. Nay's point is also very important. Beef up the mounting points. They localize the stress transferred from the movement of the wheels. That is the best place to concentrate your efforts. While rock rails have a job to do, they don't add much to the chassis stiffness (they're like that Chevy frame)
That axle, even a D-60 hanging over the edge with or without the sway bar is a minor load compared to high velocity impacts from hitting bumps both on and off the road. Just as a rough example, figure that your XJ weighs about 3500 lbs. About 1,000 of that is the wheels, tires and axles. That leaves 2500 lbs distributed over the four corners of the vehicle supported by the springs. Since the front is heavier, figure about 700 lbs per wheel on the front end. If you run over a rock in the road, say 3" high, that bump compresses your 150 ft/lb springs and 50/75 ft/lbs sway bar roughly 3" which adds another 500 lbs of load to that corner of the vehicle. The shocks which are velocity sensitive can easily add twice that, or 900 lbs more for a total impact on that corner of 2,100 lbs or over 1 ton of force acting agaist the chassis.
Ever since GM sent a bunch of engineers down to Texas to help Jim Hall design his Chaparral race cars in the 60's, chassis designers have found that the stiffer a chassis is, the easier it is to engineer the suspension to behave is a predictable and controlled manner. Off-road suspension design is about the last form of auto sport to get serious about this, but things are changing rapidly.
In a modern race car the forces can exceed five g's (five times the force of gravity) on any one wheel. Take a one metric ton race car (~2200 lbs), a FIA GTO coupe, stick it in a corner and encounter a bump. Five times the total car weight, over eleven thousand pounds, can be acting on the tire & spring on one corner. That is mucho torque on the frame, and can make tires point all over the road (or off the road).
This kind of load sounds a bit excessive until you learn the aircraft guy's have been designing structures like this for fifty years. It was only in the mid-60's that Engineering and large tires entered racing. Jim Hall recruited GM Engineering to counter Shelby and Richie Ginter's aircraft Engineer buddies at Ferrari and their structures experience learned in WWII (along with aerodynamics). Later, when Ford explored using a carbon graphite composite frame (very early 80's) they built a Probe race car out of shear panels. The car was glued together out of plastic panels and was designed to withstand ~10 g's at any corner. It used up some major Cray computer modeling time and resulted in a very competitive car.
How does this relate to the XJ Cherokee? In the early 80's finite element analysis of structures became a usable tool with computer simulation programs. The partnership of AMC with the French (can't spell that company's name) allowed the AMC engineers a crack at using the French company's supercomputer and modeling program to help design the XJ body. The XJ was one of the first American body/frame designs to be modeled using finite element analysis on a supercomputer. The goal was to allow only 1/2 of a degree in bending with a load of ~three g's (I cannot remember the exact load). The structure came out very well, except a compromise (or rushed mistake) was made to use a less expensive fiberglass hatch (that was addressed in 1996).
On the issue of tying the frame and the unit body together, it is a common practice in racing VW based vehicles. A year or so back Greg Freidman used this advantage to build the cage in his white 89 XJ. The resulting structure is very rigid. The task of tying the XJ roof to a cage Halo loop, and then rigid roll bars, is not as difficult as most believe. The grab handle mounts provide a pilot locator to set up a bolt-in or welded structure that accomplishes a exceptional combined structure. The combination is strong enough to allow some of the cross-bracing to be excluded for a street vehicle.
On the issue of Off-road design evolution, take a look at Walker Evans' ARCA rock crawler S-10.The trailing three link front end (and rear suspension) is designed to counter the ledge obstacle problems common to the XJ four leading links. Coil over shocks, turning brakes, rear steer, front-mid engine placement, and more. It is a good study for an enthusiast (and reflects the serious nature of the driver and competitor, beyond the shear expense to build the thing).
NAXJA, and the author of this tech article disclaim any and all liability associated with any "do it yourself" vehicle modifications and/or repairs.
Content © 1999 North American XJ Association