I can't remember the exact numbers, but if I remember correctly, every .004" increase in bar diameter gives an increase in stiffness on the order of 14%. Correct me if I'm wrong guys, but I know it's over 10% and that number sticks in my head. Remember though that shorter bars of the same diam will be stiffer. That said, the small block B-body bars were .088" I believe, followed by .092", then .096". Current offerings by mopar performance show the .096" max size offered is the big block upgrade size, while stock the big blocks came with .092"s if that gives you some frame of reference.
All of the aftermarket bars by Firm Feel, Hotchkis etc that are greater than .096" were not offered stock on our cars, but some small increases can give huge gains. Too big though and you can get somewhat of a rough ride like you say. Many of these big bars are better suited for racing the twisty tracks, but somewhat bigger can give you improvement on the street.
It's all personal preference though and how the rest of the car is set up. I have a '68 Plymouth Satellite and it came with .088" bars. Pretty soft. I upgraded to .092" bars and while better with my small block, I wish I would have gone to .096". Now I'm upgrading my motor to a big block so I'm going to probably go to 1.000" bars and call it good. (incidentally, the AREngineering article I posted below says that 1.000" bars are somewhat ideal for my car.)
The Hotchkis 1.1" bars would be a pretty huge jump from the 1.000" bars (assuming every .004" increase gives a somewhat noticeable increase in stiffness.) Guys, correct me if I'm wrong here, but I think those Hotchkis 1.1" bars are not necessarily meant for the street, but more for road racing or autocross where you have a smooth flat surface to drive on. On 2nd thought, I hear some are running bars bigger than 1.000" with no complaints, so I wouldn't know for sure unless I road in the car myself I guess!
Here's a great read from the experts at AREngineering:
http://arengineering.com/tech/torsion-bar-tango/
A great excerpt posted from that site:
A handy rule of thumb is to pick a bar that has a wheel rate that is 1/10thof the front-end weight. For instance, a 3600 pound car with 50% of the weight on the front end has a front end weight of 1800 pounds. A good starting point for the wheel rate would be 1800/10 or 180 lbs./inch bar. Closest bar available for the B body is the 1.00 inch bar with a wheel rate of 186 lbs./inch. For an A body, the 0.920 bar at 150 lbs./inch is probably the best choice. The next size up is the 0.990 bar at 200 lbs./inch which is probably a tad too stiff for most folks. Remember, the A body bars are shorter than the B body bars which makes them stiffer.This rule of thumb calculation is derived from the formula for natural frequency of sprung bodies. This 1/10 rule of thumb provides you with a frequency of about 1.40 cycles per second which is typical for high performance cars. For a more in depth discussion of this topic, check out “How to make your Car Handle”, by Fred Puhn. Be careful with this concept though, since more is not always better. A super stiff front suspension will have a very high natural frequency. If the natural frequency gets high enough, the car will actually be painful to drive.An interesting side note: The 1.00 bar in a 3600 pound B body provides a ride that is noticeably stiffer than the factory springs. Yet, this same wheel rate of 186 lbs/inch was the standard spring rate in the cushy riding Chrysler Imperial. How could that be? The difference is that since the Imperial weighed at least 5000 pounds, the natural frequency of the system is lowered back down to about 1.1 from 1.4. This just goes to show that matching the torsion bar rate to the car weight is the important factor. That is why the rule of thumb of spring rate being 10% of the front end weight works so well. By following that formula, you account for the car weight in your decision.
Drag Race Bars
A fairly common question people have is if they can use the drag race bar on street driven cars. Obviously, this low rate bar is going to have difficulty controlling a heavy car during cornering so a person wouldn’t want it for any type of daily driving. But the real reason to avoid such a bar on the street has to do with fact that this bar could be dangerously overloaded in such situations. Remember, the torsion bar is nothing more than a spring. And much like a valve spring that is overloaded, an overloaded torsion bar can fail. With a spring rate of only 92 pounds per inch of travel, the drag race bar has to be severely twisted to support the nose of a heavy car. Such a dramatic amount of twisting sends the internal stresses sky high. If the car hits a large bump and bottoms out, this overloaded bar could snap from the stress. Even if it doesn’t snap right away, the life of such an overload spring is going to be fairly short.The importance of matching the weight of the car to the bar size is shown by calculating twist as a function of force.Bar size Twist required to support 1000 lbs Stress in bar
.840 52 degrees 112,000 psi
.920 36 degrees 85,000 psi
1.00 26 degrees 66,000 psi
As you can see, the .840 diameter bar must be twisted twice as much as the 1.000 bar in order to support 1000 pounds of weight. This severe twist increases the stress in the bar by almost a factor of 2. Bottom line is that the drag race torsion bars are just too small for street use on the heavy B body. Minimum bar size should be the 0.920 bar in order to keep the stress level within reason.
One item to consider when using drag race bars is to use a heavier bar as the car gets faster. You’ll want to look at this as a way to control the front end of the car and prevent it from wheel standing or hanging up in the air at the fast end of the track. A heavy car probably needs a softer spring in the front to get some weight transfer to the rear tires but as your car gets faster and faster, you’ll want more control up front.
