|
Dear Philip, Thanks for both your messages! I appreciate the time you took to answer in such detail as well as the effort you put in previously to actually make the changes and generate some empirical data. I thought the list would be interested in this information as well. [Philip] "I've done the steering rack conversion you've been discussing on the Europa list. I saw your posting asking for first-hand experiences and thought I'd drop you a line. At the risk of stating the obvious I'll take you through what I've learned. My research into the changes made by Lotus points to them occurring when the Twincam was being developed. From the sources I could find, it seems Lotus added extensions to each end of the rack and eliminated the tie-rod extensions found on the S1 and S2. By moving the inner tie-rod ends (inner balljoints) further apart and shortening the tie-rods, the tie-rod ends (outer balljoints) move in arcs more closely conforming to the arcs made by the upper ball joints and trunnions of the uprights during suspension movement. ..." [Kevin] I have the S1/S2 manual as well as the TC supplement. I see the drawings illustrating the extensions and see what you mean (though the newer rack is not pictured). From previous comments on the list I thought that Lotus narrowed the center section of the rack rather than widened it (and lengthened the tie-rods ?? -- maybe I misunderstood). However, as you explain below, the changes you made do work when making sharp turns that significantly compress the suspension. [Philip] ... Prior to the modifications, my S1 had what I thought was a bad case of roll oversteer. I have since read that that term refers to the rear suspension steering the car while the same effect caused by the front suspension is called bump steer. Why there's no oversteer/understeer differentiation with bump steer I don't know. Anyway, I used to have to unwind the steering wheel after turning in to a corner or the car would tighten its line very noticeably. (This was not apparent at track days where I would adjust the Spax to be very stiff, causing the roll rate to be much slower than when using the softer street settings.) With the changes I made, that tendency is very much reduced. I suspect that if I had done a better, more informed job of relocating the rack I could have eliminated the trait entirely. As to the specific changes, I had intended to acquire a Twincam steering rack but couldn't find one before I became impatient to just be done with it. I had quite a few pieces left over from rebuilding my original rack so I had some welded together to move the inner ball joint out to match the length lost in eliminating the tie-rod extensions. Instead of the pieces (using terms from the shop manual) being assembled on the rack in the order of locknut, sleevenut, tie-rod, cup nut, I made it locknut, sleeve nut welded to cup nut welded to sleeve nut, tie-rod, cup nut. I also had an adjustable jig made to use in place of the rack mounts while finalising everything. To determine how far to thread the new cup and sleeve nut together before welding them and how much to raise the rack I trial assembled everything and adjusted it all while checking the bump steer. A clever person could probably have calculated this. The issue I never resolved in my mind was what bump steer characteristics I should aim to have beyond the obvious answer of "less". I found that the bump steer curve will change as the wheels are steered. (This appears to me to result from caster causing the tie-rod end of the upright's steering arm to move vertically as well as horizontally in relation to the ball joint and trunnion when steering the wheels.) [I need to add a clarification to my [previous] message. When I said I thought the caster angle causes the bump steer curve to change when turning, I should also have mentioned kingpin inclination, which probably has a greater effect than caster.] I tried to focus on getting the best set of bump steer curves for the wheel on the outside of a turn as it steers in, since it will have load transferred to it from the opposite wheel. [Kevin] Yes, exactly. This load transfer (with the softer stock springs) is what I was thinking about. I suspect any modifications are a measured/considered compromise of one area over another. With an adequate computer program, I guess you could run through any change and predict the dynamic results in 3-D. Similarly, with a mapped and computer controlled hydraulic steering system (costing many times the value of the car!) you could probably eliminate bump steer entirely. ;-) [Philip] ... I used the completely arbitrary range of 0 to ¾ turn at the steering wheel. This resulted in a less than optimal curve when the wheels are straight ahead. I did this in part because of experiences my friends and I have had with Elans. They use shims under the rack mounts to set proper rack height, the amount of shimming required being marked on the frame by the factory. To get the best bump steer curve when the wheels are straight ahead always requires setting the rack higher than is marked on the Elan frame. But if straight ahead is not the steering angle to check, what is? I still don't have a good answer to that and I'm sure it's why I still notice a slight bit of bump steer now. I had wanted to have adjustable rack mounts in case my first attempt was off a bit, but the quote from my machinist was too high. Instead, I had him make a non-adjustable pair with the rack moved to a setting I thought would be best but would be unable to road-test beforehand. Oh, well. Another modification Lotus made that I copied was to make a rack tube extension. This moves the bushing opposite the pinion further out. The extension is made the same length as the existing tubular spacer controlling steering lock, which is then eliminated. The bushing will take much longer to wear out with this mod. Again, I made the extension from a leftover rack. As for the 47, because the brake master cylinder was moved out of the way of the rack, the steering components could be placed more optimally, at least according to a magazine article of the period. Beyond that the only evidence I have of any differences are cutaway drawings from those magazine articles. Philip Jones |