Announcement

Collapse
No announcement yet.

Bumpsteering An S13 @ Multiple Steering Angles

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Bumpsteering An S13 @ Multiple Steering Angles

    I have measured & adjusted the bumpsteer on the front of my S13. I have measured the bumpsteer at the straight ahead position, 1/4 turn of the steering wheel, & 1/2 turn of the steering wheel position. The idea is to have the optimum bumpsteer setting on the outside (loaded) tire during cornering.

    The straight ahead position bumpsteer curve effects the stability during braking, and corner entry or initial steering response while braking or during suspension dive. This is a good curve to know in order to properly set your initial or static toe setting. For example at the striaght ahead position my optimum setup tends to have a bumpsteer curve that toes in during brake dive, so I'd want to compensate for that with static toe out for a crisp turn in.

    For the steering angle bumpsteer curve on the outside or loaded tire, I've chosen to have it setup with minimal toe change, or the best possible setting during 1/4 steering wheel angle. This gives me the best compromise for the bumpsteer curve in the straight ahead and 1/2 steering wheel angle curves. I rarely turn the steering wheel further then 1/2 of a turn.

    I've also measured the inside tire bumpsteer curves but this tire is neglected. There is always a compromise, and this tire does not have a nice curve with my optimal "outside tire setup".







    For my optimal setup I ended up with...

    Straight ahead position bumpsteer reading .020" to .025"/inch of travel. (Its a pretty linear curve). With the it toe-ing in during front suspension dive, or bump, & toe-ing out during droop.

    1/4 steering wheel angle bumpsteer curve, from full droop, in .5" increments
    Through 5" of travel. these numbers are the total toe change.
    +4, +6, +5, +4, 0, 0, -2, -8, -14, -19

    So I initally had up to +.006 toe in then it stabilzed at 0 toe change, and then toe out up to about .020".

    1/2 steering angle curve changes .015" to .020" per inch of travel (this is a pretty linear curve through 5" of travel) The toe however tends to toe out under bump, and of course toes in during droop.

    The inside tire's bumpsteer curve is pretty horrible (most likely the curve is just accelerated to to my testing method due to ackerman), but @ 1/4 turn it toes out under bump .050"/inch, and @ 1/2 turn of the steering wheel, it toes out .075"/inch of bump.

    By the way, spacing the toe arm up & down .250" makes all the difference, and completely shifts the sweet spot. I need to machine some .050" spacers, but I can shift the curve from one area of interest to the others with a .125" spacer.

    Interesting points of topics are the arc of the toe arm. It's angle changes when caster and/or negative camber are adjusted. The spindle rolls forward and rearward when caster is adjusted, and is tilted in & out when camber is adjusted. Also, the toe arm not only pivots but moves up & down on a different arc with different caster & SAI angles.

    My next goal is to positivley change the trasitional bumpsteer curves by changing SAI & Caster, or at the least completely understand & verify the effects of those settings on the bumpsteer curve.
    Last edited by Jason M; 09-01-2012, 06:08 PM.

  • #2
    Always love reading this stuff. I myself would like to start checking this stuff out on my car.
    I have one question regarding measuring bumpsteer that I'm having a hard time wrapping my brain around. How does the single guage setup account for camber change vs toe change? With the two gauges measuring the difference makes since, but with only one gauge, I'm not seeing it. Sorry if that's a silly question!

    Comment


    • #3
      Re-read "Engineer To Win" today to get Carol Smith's opinion on bumpsteer. He sets up his cars to toe out no more then .060" through 2" of travel. Usaully runs about .030" toe out for 2" of bump depending on what the car wants to help it turn in. His theory is that toe out helps the inside tire develop slip angle on initial turn in. Running toe out during bump lets you run less static toe out. (less scrub down the striaghts)

      Also got a chance to measure the distance of my inner & out pivot points on my LCA's & tie rods. The LCA is about 14" center to center where the tie rod was about 12" center to center. I always thought that the pivots for the LCA & the tie rod should be the same distance center to center, & have the same insant center.

      Plan is to shorten the LCA & mod my spindles, then re-check everything. I may look into relocating the rack if I can't get things right.

      Comment


      • #4
        I think a lot of people prefer the single gauge setup that I have.(I was pleasently surprised to read this while researching bumpsteer) There are a couple of great circle track related articles on bumpsteer that address the different guage setups. My LongAcre setup pivots at the base (notice the rod ends in the pic's) to account for track width & camber changes. One side has a roller & the other has the indicator that measures the toe plates movement. I can get very repeatable numbers with my method, so it proves the gauges accuracy.

        Originally posted by CoolBlue View Post
        Always love reading this stuff. I myself would like to start checking this stuff out on my car.
        I have one question regarding measuring bumpsteer that I'm having a hard time wrapping my brain around. How does the single guage setup account for camber change vs toe change? With the two gauges measuring the difference makes since, but with only one gauge, I'm not seeing it. Sorry if that's a silly question!

        Comment


        • #5
          This is the article that inpsired me to check my bumpsteer at different steering angles.
          http://www.circletrack.com/chassiste...d/viewall.html

          Comment


          • #6
            At the SCCA national convention a few years back I sat in on a talk given by Dan Binks.

            He mentioned that they do toe sweeps on the alignment rack at the shop for every ride height they plan on running for the next race to verify the toe curves ahead of time for a given setup.

            That way they can get the exact alignment specs with the toe curve they want dialed in the shop so if they need to do a ride height change at the track they can quickly adjust everything to spec.

            Make sure to accurately account for your desired ride height and factor that in along with the other standard alignment adjustments to find the sweet spots!
            The Nerd shall inherit the podium for knowledge is power.

            Comment


            • #7
              Originally posted by Jason M View Post
              Re-read "Engineer To Win" today to get Carol Smith's opinion on bumpsteer. He sets up his cars to toe out no more then .060" through 2" of travel. Usaully runs about .030" toe out for 2" of bump depending on what the car wants to help it turn in. His theory is that toe out helps the inside tire develop slip angle on initial turn in. Running toe out during bump lets you run less static toe out. (less scrub down the striaghts)

              Also got a chance to measure the distance of my inner & out pivot points on my LCA's & tie rods. The LCA is about 14" center to center where the tie rod was about 12" center to center. I always thought that the pivots for the LCA & the tie rod should be the same distance center to center, & have the same insant center.

              Plan is to shorten the LCA & mod my spindles, then re-check everything. I may look into relocating the rack if I can't get things right.
              While i was sorting the front bumpsteer early on, i spoke to an aquaintance whom clarified a few points, he also aked me to measure front LCA & complete tie rod arm length; my lengths were about the same as yours; the difference in lengths causes different arcs which is why my bump could not get down to 0, for an inch or 2 of bump or droop, he suggested a shorter rack, he said its not a good idea to shorten LCA very much ( I'll have to ask him why; he just mentioned it starts bringing other issues to light, he implied that most pass. cars have LCA lengths of about 14inch.)

              I was also told i can't just "eyeball toe arm level to ground" & assume bump steer is good, because different caster settings lower & raise toe pickup point

              Comment


              • #8
                I need to take a closer look at my Rack & pinion, but I think the easiest solution to have the tie rods & LCA's be the same length is going to be three fold.

                #1 Cut the Rack shorter & if necessary drill & tap the holes for the inner tie rod ends to bolt onto. This may decrease steering angle a little, but who needs full lock on a race car anyways...

                #2 Shorten the front LCA .5-1". This will reduce SAI, & negative camber. I can gain back the camber with tilting the knuckle in the strut ears, by slotting the strut. Decreasing SAI will decrease the required steering effort, and limit the necessary caster & associated jacking effect, which should help keep the inside rear tire planted. (How many race cars run 14 to 16 deg's SAI anyways???) I had previously lenghtened my LCA to compensate for spacing the ball joint pivot down, becuase the angle of the ball joint is towards the outside of the car. (the spindle moves inward with the oem length LCA when raising the roll center or spacing the ball joint pivot down)

                #3 Change the position & angle of the tie rod arm on the spindle. Reducing ackerman would increase the tie rod length, but I don't think I want to reduce ackerman. The angle of the tie rod pin is toward the center of the car, so when you space the tie rod down, the tie rod needs to be shortened to maintain the proper toe.

                Comment


                • #9
                  Looks like lots of folks are shortening there racks. http://www.webrodder.com/article.php?AID=394&SID=16

                  I would not recommend sectioning the rack & welding it back together like most of the Locost guys are doing. Who knows what alloys the racks are, and how any heat treatments are efeected. That seems sketchy to me.

                  My plan is to remove about 1 to maybe 1.5" of rack length on each side, then drill & tap. I'm not sure how deep the rack is threaded, but we could probably take off a little length with a dye grinder with the rack still in the car, and still have usable thread engagement. I'am going to remove a inner tie rod & take measurements tonight.
                  Last edited by Jason M; 09-06-2012, 11:59 PM.

                  Comment


                  • #10
                    I spent 4-5 hours bumpsteering tonight, & I learned alot. I want to make a few corrections to some of my earlier posts.

                    #1 shortening the LCA does not effect the ratio of lenght between the tie rod & LCA, I think shortening the LCA has very, very little effect on bumpsteer. Obviosly if you shorten the length of the LCA, you have to shorten the tie rod to maintain the proper toe setting.

                    #2 Having the LCA & Tie rod the same length (when measured pivot point to pivot point) does NOT give you the ideal bumpsteer curve.

                    You have to account for spindle roll, which is generated by antidive, caster, & also spindle tilt generated by camber gain/SAI. The spindle rolls rearward under bump, & causes the toe arm to move only 7/8" per 1" of LCA/Hub travel. This means if both the tie rod & LCA are the same length, you will have a toe curve that always tends to accelerate towards toeing-in.

                    I tested combinations of tie the tie rod being +2", +1.5", 1", & .5" longer then my initial settings, which were LCA 14", tie rod 12".

                    The rack can be shortened 1/2" on each side, safely, & with out drilling and tapping. Shims can then be added between the inner tie rod & the rack, in order to change the length of the tie rod.

                    #3 The length of the tie rod is important, but needs to be adjusted to a length that compensates for caster, SAI/camber gain, and antidive settings, which all effects spindle roll & tilt during articulation.

                    Thats all for now. My next step is to change the caster & SAI settings & retest. Also need to measure spindle roll in degrees. It seems to really accelrate during the last inch of bump travel. I have still not found the ideal combination for my setup, at multiple steering angles, but have made a significant inprovement on paper, & a;so have a much better understanding of my front suspension. I used to think Mac struts were simple. Kinematics in 3d are hard to visualize!

                    Comment


                    • #11
                      give a read. quite a few ppl have been making knuckles already (stateside)

                      http://www.stewartleask.com/build/32...cal-balljoints

                      Comment


                      • #12
                        I've verified that equal length toe rod & LCA's is not the awnser for a good bumpsteer at multiple steering angles. So far the best I can do is optimize 1/4 turn steering wheel anlge, have a small amount of toe in under bump in the striaght ahead postion, but I end up with .020 to .040" toe out per inch of bump travel at the 1/2 turn steering wheel angle.

                        I would like like 15-30 thousandths of toe out per inch of bump in the striaght ahead postion, & zero toe change at either the 1/4 or preferably the 1/2 turn steering wheel postion.

                        I think optimizing the 1/4, 1/2, & 3/4 turn steering wheel angles, or dynamic bumpsteer curves is where the improvement needs to be focused upon. What does it matter if the bumpsteer curve is optimized in a striaght line, if while cornering, the tire has a bumpsteer curve that moves .100" per inch of travel. I hit the brakes two to 3 times on an autocross course but I almost always have the steering wheel turned.

                        The toe arm/outer tie rod end moves in an arc dicated by a combination of SAI/KPI & caster. What must be done to compensate for this arc?

                        My current thinking is that the rack needs to be moved either forward or rearward to compensate for the caster and SAI settings, in order to achieve a near zero bumpsteer curve while turning the steering wheel.

                        Comment


                        • #13
                          Jason, any thoughts on this ?

                          http://www.facebook.com/photo.php?fb...type=1&theater

                          Comment


                          • #14
                            I think:
                            #1 They beat me to the punch
                            #2 It's about time the 240 has a proper spindle available.
                            #3 I WANT THEM!
                            #4 I don't like that the lower ball joint position is fixed. With my spindles that I'am making, I will be able to move the outer pivot of the lower ball joint to change the scrub raduis & pin lead. Once I build my non vented lightweight rotor brake setup, I'll be able to move the ball joint in towards the center of the wheel to optimize the scrub raduis, if I want to . And i'll also be able to mess around with pin lead.

                            I've been watching them develope for awhile now.

                            Moving the tie rod pickup point towards the ball joint, to increase the steering ratio will effect the bumpsteer curve. I believe that shifting the rack rearwards or the tie rod pickup point forwards should help us dial in a sweet spot at mulitiple steering angles. It's certainly something that I need to test though. I'am to the point where I need to invest into the performance trends software program to develope the optimum Macpherson strut setup.

                            I've dialed in & tested a setup with 10 degs SAI, & 5 degrees caster. My camber gain/loss is zero, & my bumpsteer curve improved a very small amount, but it's not worth the loss of caster. I have to test this setup at different toe rod legnths, but it does intially seem like reducing caster improved things a bit. I can dial in -3 deg's camber at the knuckle with this setup, so static camber is not an issue, but my strut angle is now past 90 deg's and actually tilted slightly towards the outside of the car. Reducing SAI does increase positive scrub raduis, but the camber gain curve is basically not even changed when compard with my 14 deg baseline setup, & dynamic camber when turning the wheel will be improved when I dial the caster back up to 7 deg's. SAI can also increase steering effeort.

                            I'll mess around with caster a bit more, & try a couple more tie rod length settings, but I need to finish my spindles before I start moving or shortening the rack.

                            Comment


                            • #15
                              Don, is it possible to dial in the bumpsteer at different steering angles? Have you measured and adjusted the bumpsteer curves at different steering angles?

                              Comment

                              Working...
                              X