Anybody experimented with longer wishbones on a McStrut?
Discussion
Hello all!
As the title says, really! Just wondering if any of the knowledgeable chaps on here had experimented with lengthened wishbones on a McStrut / Chapman Strut?
It seems that the longer the wishbone, the less roll centre migration. Well, according to my little model made in cad anyway.
Also seems to benefit from increasing the KPI, but I wouldn't want to take this too far as I think you'd have to counter the negatives with more castor. Could move it so far though, till the steering got too terrible...!
Anyway, just a thought! Hoping for some interesting reading appearing below this post!
Cheers! - Mark
As the title says, really! Just wondering if any of the knowledgeable chaps on here had experimented with lengthened wishbones on a McStrut / Chapman Strut?
It seems that the longer the wishbone, the less roll centre migration. Well, according to my little model made in cad anyway.
Also seems to benefit from increasing the KPI, but I wouldn't want to take this too far as I think you'd have to counter the negatives with more castor. Could move it so far though, till the steering got too terrible...!
Anyway, just a thought! Hoping for some interesting reading appearing below this post!
Cheers! - Mark
I have put 3cm longer wishbones on the front of a Porsche 944 rally car. To keep the scrub radius the same I also moved the top of the strut outwards by 3cm using a negative camber top mount fitted the 'wrong' way around. The mod also required longer track rods.
I haven't completed testing yet, but initial impressions are good.
Geometry checks haven't revealed any unexpected problems.
I haven't completed testing yet, but initial impressions are good.
Geometry checks haven't revealed any unexpected problems.
Good work Iain, nice choice of car to rally!
Sounds like a good idea there, increasing the front track. Was it the 944 that suffered from "falling over" onto the front wheels due to a larger rear track as standard? Sounds like you are aiming to alleviate this? Good luck with it all!
What I am meaning with my original question, is leaving the track width the same, but sending the inboard lower wishbone mounts further inboard. The nearer the middle the less the roll centre migrates according to my model, a problem which possibly gives the strut type suspension some of its "wooley" feel.
Sounds like a good idea there, increasing the front track. Was it the 944 that suffered from "falling over" onto the front wheels due to a larger rear track as standard? Sounds like you are aiming to alleviate this? Good luck with it all!
What I am meaning with my original question, is leaving the track width the same, but sending the inboard lower wishbone mounts further inboard. The nearer the middle the less the roll centre migrates according to my model, a problem which possibly gives the strut type suspension some of its "wooley" feel.
Moving the mounting points further inwards will give you a different roll centre. If my brain is working correctly ( its early monday morn ) it will give you less RC.
I guess when you say whooly feeling you mean the car is slow to react and moves about a lot ? Have you tried some harder springs and different shocks ?
I guess when you say whooly feeling you mean the car is slow to react and moves about a lot ? Have you tried some harder springs and different shocks ?
Thanks for the reply!
Yes, I do kind of mean that. I have driven a few different combinations of shock and spring from very firm to fairly firm, and none have ever given the feedback and control of my mk2 Honda CRX which runs double wishbones, and not too stiff springs.
After analysing the suspension movements using Solidworks I found that the roll centre migrated ridiculously with any suspension motion. Is this what you mean with "less RC"? That it is moving about less?
I had a play with arm lengths etc and found that lengthening the lower wishbones reduced the migration considerably. I am talking keeping the track the same and sending the inboard mounts further under the car here.
I realise I can simply limit suspension movement with more roll stiffness, but this has always been described as a crutch for a bad setup, I believe Chapman said this? I always like to have the suspension as supple as possible whilst remaining controlled.
Top marks for the early morning off the top of your head roll centre analysis BTW!
Yes, I do kind of mean that. I have driven a few different combinations of shock and spring from very firm to fairly firm, and none have ever given the feedback and control of my mk2 Honda CRX which runs double wishbones, and not too stiff springs.
After analysing the suspension movements using Solidworks I found that the roll centre migrated ridiculously with any suspension motion. Is this what you mean with "less RC"? That it is moving about less?
I had a play with arm lengths etc and found that lengthening the lower wishbones reduced the migration considerably. I am talking keeping the track the same and sending the inboard mounts further under the car here.
I realise I can simply limit suspension movement with more roll stiffness, but this has always been described as a crutch for a bad setup, I believe Chapman said this? I always like to have the suspension as supple as possible whilst remaining controlled.
Top marks for the early morning off the top of your head roll centre analysis BTW!
Mark_r33 said:
Thanks for the reply!
Yes, I do kind of mean that. I have driven a few different combinations of shock and spring from very firm to fairly firm, and none have ever given the feedback and control of my mk2 Honda CRX which runs double wishbones, and not too stiff springs.
After analysing the suspension movements using Solidworks I found that the roll centre migrated ridiculously with any suspension motion. Is this what you mean with "less RC"? That it is moving about less?
I had a play with arm lengths etc and found that lengthening the lower wishbones reduced the migration considerably. I am talking keeping the track the same and sending the inboard mounts further under the car here.
I realise I can simply limit suspension movement with more roll stiffness, but this has always been described as a crutch for a bad setup, I believe Chapman said this? I always like to have the suspension as supple as possible whilst remaining controlled.
Top marks for the early morning off the top of your head roll centre analysis BTW!
Yes if you have a lower rc it means the less the car will roll in the corner. Genrally less roll will give less grip also. Yes, I do kind of mean that. I have driven a few different combinations of shock and spring from very firm to fairly firm, and none have ever given the feedback and control of my mk2 Honda CRX which runs double wishbones, and not too stiff springs.
After analysing the suspension movements using Solidworks I found that the roll centre migrated ridiculously with any suspension motion. Is this what you mean with "less RC"? That it is moving about less?
I had a play with arm lengths etc and found that lengthening the lower wishbones reduced the migration considerably. I am talking keeping the track the same and sending the inboard mounts further under the car here.
I realise I can simply limit suspension movement with more roll stiffness, but this has always been described as a crutch for a bad setup, I believe Chapman said this? I always like to have the suspension as supple as possible whilst remaining controlled.
Top marks for the early morning off the top of your head roll centre analysis BTW!
In my experience changing the rc will give a bigger change in feel that springs and shocks they are more of a fine tuning type of thing. Lowering the car will also help the car to feel more direct.
I should point out my experience is of racing RC cars at a fairly high level not
real cars but the priciples work the same.
From a practical point of view id think it depends on what your doing with the car. Is it a race car or road plodder ? I would imagine to change the mouning points of the wishbones would be far from straight forward and you might not like the feel of it when its done.
theshrew said:
Yes if you have a lower rc it means the less the car will roll in the corner. Genrally less roll will give less grip also.
Incorrect. A lower RC means a larger roll couple between the CG and RC, which results in MORE roll.OP, when you say it migrates, where does it go? Are we talking lateral or vertical motion? Under what suspension motions?
Cheers for your input Shrew! I have many fond memories of my nitro RC car, sure I've still got it somewhere...
I concur with Kozy though, a lower roll centre increases the leverage the CofG has over the suspension, therefore increasing roll / reducing roll stiffness.
Hello Kozy! When I say it migrates I mean it moves all over the shop! I am taking the roll centre as the point of both sides projected lines, from centre of contact patch to virtual pivot point, crossing over one another. Here are the results from a vauxhall astra;
Vertical sprung mass motion only, numbers are ride heights, stays central at all times as no roll, everything is metric;
200 - 78.55
190 - 52.7
180 - 26.58
170 - .123
160 - -26.72
In degrees of roll, z is height, x is lateral motion from centreline, at 180mm rideheight;
0 - z- 26.58 x- 0
1 - z- -9.822 x- 584.79
2 - z- -42.25 x- 1203.2
3 - z- -136.00 x- 1896.7
Now I shall extend the lower wishbones from the original 335mm length to 685mm, leaving just 50mm in between the inner mounts!
Vertically;
200 - 72.42
190 - 60.06
180 - 47.6
170 - 35.05
160 - 22.38
Roll, again at 180mm ride height;
0 - z- 47.6 x- 0
1 - z- 47.21 x- 60.62
2 - z- 46.15 x- 122.06
3 - z- 44.38 x- 184.98
As you can see, it makes quite a large difference!
I concur with Kozy though, a lower roll centre increases the leverage the CofG has over the suspension, therefore increasing roll / reducing roll stiffness.
Hello Kozy! When I say it migrates I mean it moves all over the shop! I am taking the roll centre as the point of both sides projected lines, from centre of contact patch to virtual pivot point, crossing over one another. Here are the results from a vauxhall astra;
Vertical sprung mass motion only, numbers are ride heights, stays central at all times as no roll, everything is metric;
200 - 78.55
190 - 52.7
180 - 26.58
170 - .123
160 - -26.72
In degrees of roll, z is height, x is lateral motion from centreline, at 180mm rideheight;
0 - z- 26.58 x- 0
1 - z- -9.822 x- 584.79
2 - z- -42.25 x- 1203.2
3 - z- -136.00 x- 1896.7
Now I shall extend the lower wishbones from the original 335mm length to 685mm, leaving just 50mm in between the inner mounts!
Vertically;
200 - 72.42
190 - 60.06
180 - 47.6
170 - 35.05
160 - 22.38
Roll, again at 180mm ride height;
0 - z- 47.6 x- 0
1 - z- 47.21 x- 60.62
2 - z- 46.15 x- 122.06
3 - z- 44.38 x- 184.98
As you can see, it makes quite a large difference!
It does rather seem to, doesn't it!
The roll centre migration in bump shouldn't matter, however the migration in roll certainly is worth looking at. I personally go with Mark Ortiz's take on RC which is not that of the mid plane intersection between the 'n lines', but that of the midpoint of a vertical line, drawn between the two 'n lines' at about 75% of the track width from the inside wheel. This is supposed to closer represent to force based RC according to the tyre loads, and eliminates later migration from consideration as he considers it irrelevant. Indeed, only the height is considered when calculating load transfer, so this seems to work.
Try mapping the RC migration as described my Mark Ortiz in the linked article below, and see if vertical migration is still as significant.
http://www.shock-dyno.com/Ortiz/Ortiz-8-2004.pdf
The roll centre migration in bump shouldn't matter, however the migration in roll certainly is worth looking at. I personally go with Mark Ortiz's take on RC which is not that of the mid plane intersection between the 'n lines', but that of the midpoint of a vertical line, drawn between the two 'n lines' at about 75% of the track width from the inside wheel. This is supposed to closer represent to force based RC according to the tyre loads, and eliminates later migration from consideration as he considers it irrelevant. Indeed, only the height is considered when calculating load transfer, so this seems to work.
Try mapping the RC migration as described my Mark Ortiz in the linked article below, and see if vertical migration is still as significant.
http://www.shock-dyno.com/Ortiz/Ortiz-8-2004.pdf
Right, apologies for the delay, that has been quite intensive!
Many thanks for posting that article Kozy, feel free to post anything else relevant to anything else suspension wise!
So, to try and solve this in Ortiz's manner I have consulted my trusty old spreadsheet constructed from Gould's chassis dynamics equations in the book Race car suspension, by the late great Alan Staniforth.
Incidentally, I have also realised that I was closer to the correct roll centre when I used to use Alan's method of roll centre location (on the vehicle centreline) than using the method I learned from the Milliken brothers' book Race car vehicle dynamics. But I digress!
So, with my example astra vaguely entered (My spreadsheet is configured for the mid engine rear drive astra a friend and I have built, so weights are out, but the front suspension is the same only proportionally stiffer than the rear.) And ignoring the effect on roll angle that the roll centre movement would have, I shall attempt to find the roll centre for .33g, .66g, and 1g, which correlate to 1, 2 & 3 degrees of roll respectively (I softened it up on the ARB lever arms to get more roll to amplify the effect).
So, for the normal setup I get;
Roll - RC Height - Force position centre
1 - 33.196 - 491.421
2 - 11.423 - 982.842
3 - -38.056 - 1470 (100% track width)
And the extended wishbones;
1 - 32.662 - 491.421
2 - 29.933 - 982.842
3 - 23.748 - 1470 (100% track width)
The normal suspension measures a 2.6 on the Mitchell Index, and the modified version just 1.25.
It just so happens that, with my arrangement of weight and roll rates, I get about 33% front end weight transfer at .33G, 66% at .66 etc, so hence the 100% track width at 1G.
So there is still quite a large amount of change, even calculated this way. It is nice to know that the ridiculous amount of lateral migration can basically be ignored though!
Please do inform me if you can see that I have not picked this method up correctly.
Very interesting way of working it out is that, cheers Kozy!
Many thanks for posting that article Kozy, feel free to post anything else relevant to anything else suspension wise!
So, to try and solve this in Ortiz's manner I have consulted my trusty old spreadsheet constructed from Gould's chassis dynamics equations in the book Race car suspension, by the late great Alan Staniforth.
Incidentally, I have also realised that I was closer to the correct roll centre when I used to use Alan's method of roll centre location (on the vehicle centreline) than using the method I learned from the Milliken brothers' book Race car vehicle dynamics. But I digress!
So, with my example astra vaguely entered (My spreadsheet is configured for the mid engine rear drive astra a friend and I have built, so weights are out, but the front suspension is the same only proportionally stiffer than the rear.) And ignoring the effect on roll angle that the roll centre movement would have, I shall attempt to find the roll centre for .33g, .66g, and 1g, which correlate to 1, 2 & 3 degrees of roll respectively (I softened it up on the ARB lever arms to get more roll to amplify the effect).
So, for the normal setup I get;
Roll - RC Height - Force position centre
1 - 33.196 - 491.421
2 - 11.423 - 982.842
3 - -38.056 - 1470 (100% track width)
And the extended wishbones;
1 - 32.662 - 491.421
2 - 29.933 - 982.842
3 - 23.748 - 1470 (100% track width)
The normal suspension measures a 2.6 on the Mitchell Index, and the modified version just 1.25.
It just so happens that, with my arrangement of weight and roll rates, I get about 33% front end weight transfer at .33G, 66% at .66 etc, so hence the 100% track width at 1G.
So there is still quite a large amount of change, even calculated this way. It is nice to know that the ridiculous amount of lateral migration can basically be ignored though!
Please do inform me if you can see that I have not picked this method up correctly.
Very interesting way of working it out is that, cheers Kozy!
That sounds about right, though I can't really offer any insight on the RC movement of a MacPherson assembly as I have only modelled a double wishbone setup myself, which does tend to keep RC movement to a minimum when done properly. I suppose the migration you are getting is entirely possible and probably a contributing factor as to why MacPherson suspension never feels as good as a good DW suspension!
Are you sure about the load transfer though? 100% at 1G does not sound right. The model for my Civic used to give about 75% at 1G. At 100% you'd effectively you'd be up on two wheels (assuming rear roll stiffness is higher) which doesn't sound right at all.
Are you sure about the load transfer though? 100% at 1G does not sound right. The model for my Civic used to give about 75% at 1G. At 100% you'd effectively you'd be up on two wheels (assuming rear roll stiffness is higher) which doesn't sound right at all.
Edited by Kozy on Monday 28th January 11:22
Have you considered the effect on camber gain? Losing any significant amount of camber gain in roll is definitely not desirable on a McPherson as it's usually quite limited anyway.
When you say 'wooly feel', are you sure that this is a roll centre problem? With a standard road car I'd be more inclined to start with extra bracing as the setup tends to put a lot of load through the shell and it's a quick and easy fix, then I'd start looking at bushes, rack mounting and general compliance.
If you're talking about track work then you may well also find the car has a lot of bump steer built into it, worth playing with outer ball joint and rack heights to reduce it a bit if you can.
When you say 'wooly feel', are you sure that this is a roll centre problem? With a standard road car I'd be more inclined to start with extra bracing as the setup tends to put a lot of load through the shell and it's a quick and easy fix, then I'd start looking at bushes, rack mounting and general compliance.
If you're talking about track work then you may well also find the car has a lot of bump steer built into it, worth playing with outer ball joint and rack heights to reduce it a bit if you can.
Edited by The Wookie on Monday 28th January 11:56
The details in my calculator are for an Astra coupe that we have put a v6 turbo engine in the back, so the front end roll stiffness is proportionaly higher. At 1g it is just about taking the weight off of the inside front wheel.
I don't know how much of a real effect it has either, but surely it can't help!
Hello Wookie!
It's not as bad as I've maybe made out, just lacks the feel that my double wishbones do is all. We have fitted spherical bearings all round, and that's sorted it a good bit. The crx still "feels" better though, even without the sphericals at the moment.
Bump steer is on the agenda, but from measuring it up there isn't a rediculous amount anyway.
The strut top maybe another point of concern, so we have some of those in development too.
Back to the original point though, it does seem worth a try, and on our rear engined astra it wouldn't be too hard to try it out. May give it a go eventually.
Oh yes, last bit, camber gain. Utterly terrible on the astra suspension at any reasonable ride height! So we won't be missing much there.
I don't know how much of a real effect it has either, but surely it can't help!
Hello Wookie!
It's not as bad as I've maybe made out, just lacks the feel that my double wishbones do is all. We have fitted spherical bearings all round, and that's sorted it a good bit. The crx still "feels" better though, even without the sphericals at the moment.
Bump steer is on the agenda, but from measuring it up there isn't a rediculous amount anyway.
The strut top maybe another point of concern, so we have some of those in development too.
Back to the original point though, it does seem worth a try, and on our rear engined astra it wouldn't be too hard to try it out. May give it a go eventually.
Oh yes, last bit, camber gain. Utterly terrible on the astra suspension at any reasonable ride height! So we won't be missing much there.
Good stuff, definitely worth looking at making sure those strut tops are solid and well braced!
Also worth pointing out that roll centre height to camber gain is quite a significant compromise on a McPherson, so if it's got particularly rubbish camber gain and a bigger than necessary roll centre height then it probably wont hurt to drop it a bit for the sake of more gain!
Also worth pointing out that roll centre height to camber gain is quite a significant compromise on a McPherson, so if it's got particularly rubbish camber gain and a bigger than necessary roll centre height then it probably wont hurt to drop it a bit for the sake of more gain!
Yeah I know what you mean buddy!
If I set it to gain any camber in roll then my roll centre is at about axle height!
At the same ride height for each I get roughly .1 degrees gain per degree roll. The longer arms actually give just slightly more, but we are talking .01's.
I guess that's why everyone uses so much static Neg and high roll stiffness in these things!
Kozy, just reread this and noticed you mentioned a Civic. What you got?
I love my oldies and have a '90 16v CRX with mugen kit that is in bits, awaiting some love, and a '90 Crx SiR with glass roof, LSD box and B16a2 with a t28 from a Pulsar GTiR. Awesome little cars, a very capable chassis to say it is 25ish years old!
I haven't managed to measure either one up yet though, been too busy learning about Astra's due to a new business venture. So would be nice to know what you know if you have measured up an ED, EE, EF, EG, EK, DC2 etc chassis, if you wouldn't mind and it's not too cheeky of course!
If I set it to gain any camber in roll then my roll centre is at about axle height!
At the same ride height for each I get roughly .1 degrees gain per degree roll. The longer arms actually give just slightly more, but we are talking .01's.
I guess that's why everyone uses so much static Neg and high roll stiffness in these things!
Kozy, just reread this and noticed you mentioned a Civic. What you got?
I love my oldies and have a '90 16v CRX with mugen kit that is in bits, awaiting some love, and a '90 Crx SiR with glass roof, LSD box and B16a2 with a t28 from a Pulsar GTiR. Awesome little cars, a very capable chassis to say it is 25ish years old!
I haven't managed to measure either one up yet though, been too busy learning about Astra's due to a new business venture. So would be nice to know what you know if you have measured up an ED, EE, EF, EG, EK, DC2 etc chassis, if you wouldn't mind and it's not too cheeky of course!
I had an EK4 with lots of EK9 spec stuff, B16B, Quaife, Spoon suspension, massive rear ARB and lots of other bits. I measured all the suspension and modelled it up so I could see where the wheels were pointing on corner exits, I could then make sure I was running just the right amount of camber so that the contact patch was used nicely when I needed it most. Seemed to work, I had a bit of an issue with oversteer as I never optimised the rear!
Superb little car and really did show up some more expensive machinery anywhere corners were concerned! Had to sell up once my nipper arrived, swapped this and a Mondeo shed for an ATR which I'm not allowed to race. Got an MX5 for that now. No doubt the same level of setup will be undertaken in time!
Happy to give you measurements but can't gurantee their accuracy!
Superb little car and really did show up some more expensive machinery anywhere corners were concerned! Had to sell up once my nipper arrived, swapped this and a Mondeo shed for an ATR which I'm not allowed to race. Got an MX5 for that now. No doubt the same level of setup will be undertaken in time!
Happy to give you measurements but can't gurantee their accuracy!
Sounds superb Dan (< only just realised how good the profile bit on PH is!)
I put a B18c4 with pr3 head in an EK4 for a lad about a year ago, went really well! Was the same colour as yours too!
Yeah, whatever you have got will be better than nothing for a bit of analysis mate. Fire it over to mreilly533 at gmail dot com if you could please!
I know a few people who are venturing into MX-5 territory, they seem a very capable car too. When we measuring that up...?!
I put a B18c4 with pr3 head in an EK4 for a lad about a year ago, went really well! Was the same colour as yours too!
Yeah, whatever you have got will be better than nothing for a bit of analysis mate. Fire it over to mreilly533 at gmail dot com if you could please!
I know a few people who are venturing into MX-5 territory, they seem a very capable car too. When we measuring that up...?!
Not until the weather improves! 'tis not the season for spending the day underneath a car with rulers and bits of string!
MX5 is running in standard class this year so no need to do anything in a hurry.
I will have to pull up the Pro/E file for the Civic and jot down the dimensions, will get them over to you when I find some time.
MX5 is running in standard class this year so no need to do anything in a hurry.
I will have to pull up the Pro/E file for the Civic and jot down the dimensions, will get them over to you when I find some time.
Mark_r33 said:
Yeah I know what you mean buddy!
If I set it to gain any camber in roll then my roll centre is at about axle height!
At the same ride height for each I get roughly .1 degrees gain per degree roll. The longer arms actually give just slightly more, but we are talking .01's.
I guess that's why everyone uses so much static Neg and high roll stiffness in these things!
Kozy, just reread this and noticed you mentioned a Civic. What you got?
I love my oldies and have a '90 16v CRX with mugen kit that is in bits, awaiting some love, and a '90 Crx SiR with glass roof, LSD box and B16a2 with a t28 from a Pulsar GTiR. Awesome little cars, a very capable chassis to say it is 25ish years old!
I haven't managed to measure either one up yet though, been too busy learning about Astra's due to a new business venture. So would be nice to know what you know if you have measured up an ED, EE, EF, EG, EK, DC2 etc chassis, if you wouldn't mind and it's not too cheeky of course!
Not great but not the end of the world. If I set it to gain any camber in roll then my roll centre is at about axle height!
At the same ride height for each I get roughly .1 degrees gain per degree roll. The longer arms actually give just slightly more, but we are talking .01's.
I guess that's why everyone uses so much static Neg and high roll stiffness in these things!
Kozy, just reread this and noticed you mentioned a Civic. What you got?
I love my oldies and have a '90 16v CRX with mugen kit that is in bits, awaiting some love, and a '90 Crx SiR with glass roof, LSD box and B16a2 with a t28 from a Pulsar GTiR. Awesome little cars, a very capable chassis to say it is 25ish years old!
I haven't managed to measure either one up yet though, been too busy learning about Astra's due to a new business venture. So would be nice to know what you know if you have measured up an ED, EE, EF, EG, EK, DC2 etc chassis, if you wouldn't mind and it's not too cheeky of course!
Having said that, on the Golf this year, we had crap body control and poor grip with high roll centres and soft springs and the car was miles better when we dropped the roll centres, increased the static camber and fitted stiffer springs.
The only difficult part of it is that you need to make sure your damping is spot on or it'll make it a b
d to drive!ETA - Sorry just realised I was arse about face with camber gain and roll centres before, raising gives more!
Edited by The Wookie on Tuesday 29th January 09:38
Gassing Station | Suspension, Brakes & Tyres | Top of Page | What's New | My Stuff