Announcement

Collapse
No announcement yet.

Working on roll center correction solution

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

  • Working on roll center correction solution

    I was inspired by a thread about this on Migweb a few weeks ago and the idea couldn't leave my mind ever since.

    http://www.migweb.co.uk/forums/other...-adjuster.html

    It made me finally realise what Benh was on about a while ago making his whishbone mods. I am now in the phase of seeing plenty light at the end of the dark tunnel the whole idea went through. The complicated parts is I want it done on a road legal car, for motorsport purposes there are tons of off the shelve solutions available, off course at a price. The results can make a stunning improvement to you car handling, , at least I expect the same improvements on a MK2 as I've experienced on my colleagues lowered Civic.

    For those who think, what's going on here and how come I've never heard about this before read this first:
    http://www.meganracing.com/tech/faqs...ect=Suspension

    And to the relations between Caster, Scrub and KPA (kingpin inclination angle)
    http://www.motoiq.com/MagazineArticl...and-Scrub.aspx

    For information, as many will know my Astra has already gone through some changes in geometry, like increased Caster from 1.5 to 4.7 degrees. I'm not planning to make the car a sportscar for competitive use, but purely want to see where I can improve the handling, as hobby project and for the fun of it. The improvements already made also make it taste like more.

    In short the effect is as follows, the stock suspension will have the wishbones positioned in a good angle to prevent roll. The car will have the roll center fairly near the center of mass. The center of mass acts as a lever over the roll center, so if these are becoming wider spaced the lever effect will increase and the car will tend to lean over much further in corners. The only means fighting it then is using stiffer springs, with all the negatives that come with it. These negatives are usually far greater as the advantages of lowering. And the result of lowering is, guess what?? Right, vastly lower roll center, far lower as the lowered point of mass...

    Stock suspension:


    Lowered:


    Roll center corrected:


    Even though the suspension geometry doesn't quite match a MK2, the basic effect is clear and also applies to a single wishbone suspension.

    Now comes the hard part. Just shimming the stock balljoint won't fix the problem as it will only cosmetically place the wishbone back in to stock position. The real solution is moving the center of the balljoint lower with it. The stock steering knuckles don't really lend themselves well for a good solution and the off the shelve solutions for instance offered by superfwd are a dangerous thing to be used on a road car, and besides may void any form of insurance when the part would fail and cause an accident. And the chance it fails is definitely big as the moment of force will increase by the increased length of the balljoint pin.

    Here's a good example of HOW NOT TO:


    Here's how it can be done on some cars, resulting in a safe and road legal solution:


    I've thought about things like making the single hole stub at the bottom of the knuckle a two holed one, and modify the wishbones to accomodate an S2000 style upside down balljoint:




    I've also thought about welding, but the problem is you can never tell if it will hold for sure. You can weld cast steel with the proper methods, but when one tests good in a stress test it doesn't mean another will too. The risk of cracks and the resulting complications of fabricating it with identical specs more than once would mean for me that I would never sit behind the wheel without worrying about it, even if it would hold a lifetime...

    So I've been spending day and night exploring the net to find a knuckle that might fit the bill, and I think I've succeeded! I found knuckles that I also found cheap in used condition, so cheap I couldn't resist to just order them and see later.

    While at it I gathered wheel bearing size specs, brake size specs etc. etc. to compare to the stock GM strut/knuckle and already saw many similarities and things that weren't too hard to adapt. I also posted about my plans on the Dutch Opel forums and got a couple of offers to pick up struts for free to be able to do off the car measurements and some fabrication experiments.

    I've also ordered a pair of Honda Civic wishbones, just to cut the ends off with the balljoint holes, which will save me some fabrication to make life a bit easier. It will then also fit well for certain.

    The knuckles that should make it all possible are Honda Civic mk7:


    Which have no steering arms as this Civic also uses a mc Pherson strut. They have an overall larger wheel bearing, but the inner size differs precisely 4mm from the GM hub, so these can be pressed in using a metric standard size piece of tube with 2mm walls. these come in 42.4mm outer size, so the do have to be heated before pressed. Then brakes... There are tons of ways to get this done, simplest is going by 262mm rotors and get the stock Honda calipers with the knuckles. The GM hub should be no real problem as the line up almost perfectly with the Honda hub.

    I'm not willing to sacrifice the great break performance I already have with my 288mm kit, so I searched on and found the Honda Accord mk8 uses ATE calipers with identical piston size (57mm) that are almost identical with the GM ATE calipers, the main difference being the caliper brackets fit the Honda knuckle. Then the set up goes to 300mm rotors, which are available for a Renault Grand Scenic in 4x100, so will fit a stock GM alloy wheel. Mind you though that 300mm brakes will probably only fit 16 inch alloys! My Vectra alloys should present no issues for as far as I can see....

    Roll center adjusters come in many sizes, shapes and prices for the Civic, just type it in to google and many pop up...
    http://www.ebay.com/itm/Megan-Racing...79ccaf&vxp=mtr


    I plan to remove the MK2 knuckles from the strut tubes and weld brackets to them to bolt to the Civic knuckles, which is about the same construction as the Nova/Corsa/MK4 and more GM cars have. Bonus is when I give them one slide hole that camber can be adjusted as well!

    There's still a lot to measure and figure out, but so far so good.
    Last edited by Vinci; 27-10-2013, 08:49 AM.

  • #2
    Re: Working on roll center correction solution

    very interesting subject, and some good easy to follow info there.

    whilst i understand the words written here, and their application, my understanding doesnt extend to whether these changes are better.

    there is one point in there though, when it says lowering a car will make roll worse unless spring poundage is increase. absolutely right, but i couldnt name any lowering springs that keep standard spring rate, they all go up some 30-50 (ish) pounds. i wonder how this factors into it, does it remove some or all of the bad effect?

    Comment


    • #3
      Re: Working on roll center correction solution

      Originally posted by DarrenH View Post
      very interesting subject, and some good easy to follow info there.

      whilst i understand the words written here, and their application, my understanding doesnt extend to whether these changes are better.

      there is one point in there though, when it says lowering a car will make roll worse unless spring poundage is increase. absolutely right, but i couldnt name any lowering springs that keep standard spring rate, they all go up some 30-50 (ish) pounds. i wonder how this factors into it, does it remove some or all of the bad effect?
      Well, I don't want to pretend to be the suspension expert here, and many aspects depend greatly on the purpose of the car. One thing I am certain about though is that lowering springs are higher poundage because most customers that buy them haven't got a clue about roll center and the lowering springs manufacturers don't mind to keep it that way. The higher poundage is both a necessity, and a means of ensuring the product will not fail to convince the customer day 1.

      A non roll center corrected, lowered car with stiffer suspension will not handle that much poorer as the stock set-up, and in some smooth pieces of road might even feel much improved. The down sides will only reveal themselves under certain circumstances, when the road is a bit bumpy, then the less active stiffer suspension will make the car bounce quicker, making it loose traction and in effect decrease cornering speed.

      For high cornering speed, even with bumps you are best off with active suspension that enables the best possible traction and when as a bonus there is less roll the angles at which the geometry is designed are best maintained. Don't forget that at the limits, when there is a lot of roll the camber of the front wheels will be at the least optimal, leaning to positive camber (wheel hanging out the arch)

      There is not only roll, hence the link posted above about the relations between caster, scrub and roll. A lot of it relates to the design considerations that the designer of the car had to deal with. For instance, a MK2 only had power steering for the top of the range models. Most didn't have it. The tyre profile height, most were sitting on 13 inch wheels with balloon thick tyres. To enable light steering the MK2 was designed with little positive caster and minimal scrub radius. Todays cars have the strut tops pointing very much inwards and backwards, which would make the steering wheel impossible to move without power steering, but does make the car handling much better. Todays cars always have power steering, so the designers are more free to focus on maximum handling capabilities.

      There's yet so much more to it. For instance brake, bump and torque steer. Too much of the good bits will eventually create negatives at the other end of the scale, such as the above named...

      I don't think I can explain it any better than the links at the start of the topic, so if you want to know it all in much detail I can only point towards these pieces of information and even wikipedia, which have quite good articles on for instance scrub radius and more suspension geometry subjects....
      Last edited by Vinci; 27-10-2013, 09:15 AM.

      Comment


      • #4
        Re: Working on roll center correction solution

        i wonder also if lower springs being high poundage is necessity to stop suspension bottoming out, now that bump travel is reduced 30mm?

        do you think (collectively anyone reading this) that near to standard ride height with increase spring poundage would be better than lowered car, i guess it must be if standard roll centre is retained.

        Comment


        • #5
          Re: Working on roll center correction solution

          Ideally you'd have progressive spring rate off course, such as an air spring, but yes, stiffer helps to prevent bottoming out when lowered as there's less travel, but it is not the main reason they make them that way.

          Which is better depends solely on what the road surface is like and how fast the car is driven over it, lower spring rate will make the suspension more active as long as travel is used optimally, the stiffer the rate becomes the less usable travel remains. If you can still handle your car well though with the tyres flying half the time that could make it different for you. If you ride offroad and want lift off and have the dukes of hazard feel, that's your choice.

          Generally speaking stiffer springs than what can be rendered optimal, which is the stock rate, will only result in less usable travel, less active suspension and thus decreased traction.

          If you lower, and more or less maintain stock spring rate, place some good bump stops and correct roll, you should see it all working out for the best. Mind you that the wishbones and steering arm angles were designed at angles that were optimised at the suspension and doing a modification, such as lowering has consequences, when really analysing it are worse than the gains.
          Last edited by Vinci; 27-10-2013, 09:52 AM.

          Comment


          • #6
            Re: Working on roll center correction solution

            Originally posted by DarrenH View Post
            i wonder also if lower springs being high poundage is necessity to stop suspension bottoming out, now that bump travel is reduced 30mm?

            do you think (collectively anyone reading this) that near to standard ride height with increase spring poundage would be better than lowered car, i guess it must be if standard roll centre is retained.

            Indeed, but they still bottom out I have the destroyed bump stops to show. Since sorting the struts out on the MK3 it's so much better, with the MK2 eibach's really not that stiff at all but has a lot of traction. I think most of the stiffness on that was bumpstop rather than spring!


            I'd also say yeah a near standard ride height could well be better, more so when you think of all the other problems that lowering cause's. Lowering the outer balljoint does help to keep the wheel's in negative camber, more so with the extra Castor for example mine stay fully negative now rather than going positive. Even as standard they did not do that, but it's not by much that they do.


            Personally I have a feeling that the scrub radius is what make's higher powered cars dart around under load as one side lose's traction, as it create's a sideways force so as one wheel slips the other side pulls it over. It's something I'm looking into at the moment tbh, but the Astra SAI is 14.5' and the castor is only 1.5' as standard that's quite a difference. Makes the scrub radius a little too high I think.

            Just to add this I cant remember how correct this screen grab is, or how far along my workings are but some might be incorrect..

            Click image for larger version

Name:	GTE more anti dive less bump1.JPG
Views:	1
Size:	149.5 KB
ID:	472796
            Returned GTE to default settings : On axle stands

            Comment


            • #7
              Re: Working on roll center correction solution

              That's just started to dawn on me as well, funny you should write the same findings just after it! As they are mcPherson struts they have negative scrub by nature and the Astra may well have it a little too much as it is rather aggressive with self centering too and has quite a bit of torque and brake steer by nature... I should do some measurements soon to see where it's all really at and where it needs to go from there.

              Also want to have a look at the wheels and tyres I will want to drive in the future as my present 16 inch alloys with 205x55 tyres have quite an influence on the scrub as well. May want to convert to a little bit slimmer tyre when these are worn out. That would mean I keep the wheels at 16 inch, but perhaps will fit 195x50 or 195x55 on them...

              Any ideas on that welcome!

              Comment


              • #8
                Re: Working on roll center correction solution

                Here's a good drawing showing how roll places your outer turning wheel in to positive camber with a lowered car:

                Comment


                • #9
                  Re: Working on roll center correction solution

                  Originally posted by Vinci View Post
                  That's just started to dawn on me as well, funny you should write the same findings just after it! As they are mcPherson struts they have negative scrub by nature and the Astra may well have it a little too much as it is rather aggressive with self centering too and has quite a bit of torque and brake steer by nature... I should do some measurements soon to see where it's all really at and where it needs to go from there.

                  Also want to have a look at the wheels and tyres I will want to drive in the future as my present 16 inch alloys with 205x55 tyres have quite an influence on the scrub as well. May want to convert to a little bit slimmer tyre when these are worn out. That would mean I keep the wheels at 16 inch, but perhaps will fit 195x50 or 195x55 on them...

                  Any ideas on that welcome!

                  Yeah going by my data chart the scrub radius is about 46mm, but I'm not sure I had the correct offset ect entered in as I was looking at other things when I took that screen grab. I'll have to dig out the measurements and double check that one tbh.
                  Returned GTE to default settings : On axle stands

                  Comment


                  • #10
                    Re: Working on roll center correction solution

                    Well that seams okay, in the article posted at the beginning of the relations between scrub, caster and KPA they mentioned 50mm as the maximum for a FWD car.

                    I do suppose my larger wheels decreased it a little bit then?

                    Comment


                    • #11
                      Re: Working on roll center correction solution

                      Crap, really bummed out at the moment....

                      Just been out measuring the Honda knuckles and compared them to the GM knuckles and the natural roll of the two is already so much different, and the wrong way around. There is roughly 40mm difference between the two when I measure from the center of the hub towards the center of the balljoint.

                      Honda:



                      GM:



                      Difference:


                      I doubt there are roll center adjusters for the Honda that do more than 40mm...

                      Back to the drawing board I guess...

                      Comment


                      • #12
                        Re: Working on roll center correction solution

                        I do guess plan A is not such a bad idea, the Honda S2000 balljoints that mount with two bolts. These can be greatly extended with no real strength penalty. I can mill two plates with a spacing that fits around the stub and a cylindrical piece sticking out of it that can be shaped on to it by a lathe. When correctly machined it should be pressed in to the knuckle and then the two bolts will lock it up further. No welding should then be needed....

                        Comment


                        • #13
                          Re: Working on roll center correction solution

                          Further to the pm, the only hub (possible useful) pictures I have on my computer to hand are astra mk6.

                          Click image for larger version

Name:	astramk6 (4).jpg
Views:	1
Size:	631.0 KB
ID:	472799Click image for larger version

Name:	astramk6 (5).jpg
Views:	1
Size:	245.4 KB
ID:	472800Click image for larger version

Name:	astramk6 (6).jpg
Views:	1
Size:	295.7 KB
ID:	472801Click image for larger version

Name:	astramk6 (17).jpg
Views:	1
Size:	336.7 KB
ID:	472802

                          I'll see if I have anything else lurking on my phone and what turns up at work on the ramps.
                          Returned GTE to default settings : On axle stands

                          Comment


                          • #14
                            Re: Working on roll center correction solution

                            Those are aluminium aren't they? Nice to save some unsprung weight, but how strong would it be when using extended balljoints?

                            Comment


                            • #15
                              Re: Working on roll center correction solution

                              Yeah they are alloy, not sure on how strong in all honesty. I was thinking of them for mine so I measured some, but I was put off by the price of wheel bearings! That and the 5 stud only, so I went for Vectra B instead.

                              I think your on the right track with the Honda stuff, those bottom ball joints are press fit which is easier to work with. As you say it could be pressed out and press in an extension to take the 2 bolt other ball joint, but then I'd be tempted drill and dowel pin sideways to be sure it's well fixed in the hub.
                              Returned GTE to default settings : On axle stands

                              Comment

                              Working...
                              X