Spring rates

[Jerry B]

This subject may have been covered in this forum before, but I can`t find a thread so could some of you technical types out there please tell me what spring rates you have used in competition. Mine are fine for road use but clearly too soft for competition use as confirmed by my front wheel lifting off the ground in Dave`s video nasty.

My Hawk 3.0l V6 weighs about 960kg.

Thanks guys.

[Stratos]

Jerry,

I know my car is an Allora, not a Hawk so not directly comparable at the fornt, but should be the same at the rear, and my weight is about the same as your car.

I use:-

Front 450 lbs with Avo Dampers
Rear 350 lbs with Leda Struts


Yes, your car was interesting to watch. I remember commenting on it to you at Abingdon last year. Now you know what I meant.

Think about it now, maybe I shouldn't post this information until after this year's event in case it makes you even quicker :rolleyes: :rolleyes:

[roger001]

All depends on what you want to do with the car - Matt tells me Graham Scott was running 650 F and 800 R !!! I've opted for 600F and 500R for circuit use but only time will tell as I,m some way off completion, due to current commitments.

[rutthenut]

At the front of my car I have been using 550lbs/in springs, which could be stiffer still for circuit racing. They are not very good on wet and bumpy public roads though, where a softer spring setting would be better. They do help to keep some of the body roll in check, which is one of the problems found when racing (and anti-roll bars could be stiffer for that too).

The rear of the car has gone to 400lbs/in and I find that is still ok on the road - well, comfortable but still a bit too hard for uneven/wet road surfaces as traction can suffer.

This is on my HF3000, which has the (standard) angled front shockers that noticeably reduce the actual wheel rate (by about 50%). The Corse has very different front spring mounts, so cannot be used in direct comparisons.

I did do the maths on all of this some time ago, including suspension measurements, leverage ratios, front and rear vehicle weights, desired 'harmonic frequency' and the ratio between front and rear. This all came from the Race and Rally Car Source Book and was quite interesting. The only problem with coming at it from the theoretical angle is that I couldn't be sure what numbers to look for - the maths just allows the spring rates to be worked out to meet desired targets. Trial and error is still needed to come up with the requirements for your own needs.

The bit about lifting the front wheel is actually related to the relative roll angles of the car, front and rear. Having softer suspension clearly allows the car to roll more and show up anything of this sort. But it looks good on a Lotus Cortina, why not on a Strat too?

If you change the inner mounting of the lower wishbone on the rear suspension, this will change the theoretical roll centre at the rear and result in more or less likelihood of lifting a front wheel. That is probably the least important reason for changing the rear suspension geometry though. I also cannot remember off-hand whether the higher or lower mounting point would be the cause.

[lpriestland]

Hugh supplied me the 'i' kit with springs at 350/400 f/r, I think.

However last weekend at the Donnington Track Day, David Carson's car had 450/500 and was v.impressive. It certainly put a big :D on my face.

There was just enough roll to let you know what was happening and to be able to load it into a turn. But not too much to loose traction/etc. He said that it was much firmer on the road but not too excessive. One man's 'firm' is anothers ????!

He'd previously tried 400/450 but felt that there was a marked improvement with at the higher rate.

The Eagle F1's had been replaced with some form of Dunlop near slick which were well warm after a few laps. Apparently they rolled over much less than the earlier Eagles which were wearing the sidewall writting off.

She runs on an Alfa v6.

PS: David, I hope you don't mind mate but I couldn't find you as a registered user...

[Stratos]

Originally posted by lpriestland
Hugh supplied me the 'i' kit with springs at 350/400 f/r, I think.

However last weekend at the Donnington Track Day, David Carson's car had 450/500 and was v.impressive. It certainly put a big :D on my face.

There was just enough roll to let you know what was happening and to be able to load it into a turn. But not too much to loose traction/etc. He said that it was much firmer on the road but not too excessive. One man's 'firm' is anothers ????!

He'd previously tried 400/450 but felt that there was a marked improvement with at the higher rate.

The Eagle F1's had been replaced with some form of Dunlop near slick which were well warm after a few laps. Apparently they rolled over much less than the earlier Eagles which were wearing the sidewall writting off.

She runs on an Alfa v6.

PS: David, I hope you don't mind mate but I couldn't find you as a registered user...

Just need to clarify here that Lee is talking about his own Corse I, and David Carson's Corse I.

Both cars have twin wishbones at the rear, and different angled spring/dampers at the front, so a direct comparison with a Hawk is not possible. But the information is useful to Corse owners with the I chassis, and, I suppose, as some sort of general indiaction of spring rates.

[Stratos]

Originally posted by rutthenut
The rear of the car has gone to 400lbs/in and I find that is still ok on the road - well, comfortable but still a bit too hard for uneven/wet road surfaces as traction can suffer.



On most of the rallies that I have done with my car, I have found the rear rate to be about right (I think, cos it's better than it was with the 325 lbs that I previously had on the back), but on very bumpy, uneven sections, it's a bit too stiff, and, as John mentions, you loose a bit of traction.

But, overall, I'm fairly satisfied with the rear spring rate on my car and have no plans to change it.

[lpriestland]

I guess that if it'd been a 'Hawk i', then we'd have been on the set of Mash...with the help of Radar looking for Hot Lips.

Right?

[Jerry B]

Thanks guys, all useful information, keep the options coming.

What do I want to do with the car? Well, obviosly I want the perfect setup for the 2004 Stratos chalenge!

[ANDYG]

Jerry,
All this talk about spring rates, etc, etc will only go towards increasing your handicap for the series, I didn't think you did too bad with three wheels but if you want four wheels on the ground - well, have you tried racing blindfolded!!:rolleyes:

[chris.richard]

Originally posted by Jerry B
My Hawk 3.0l V6 weighs about 960kg.


And it'll have to be 1060kg to get a logbook....

[Jerry B]

Chris, I had an interesting chat with John Ryan the tech. bod at RAC MSA who said that there is going to be a rule change soon altering the weight to be linked to engine size. Sub 3000cc will then be 1000kg.
Here`s hoping.


There`s also the hope that the gross weight will include the driver, in which case I, along with one or two other Statos drivers will be looking for a rebate.

[rutthenut]

Originally posted by Jerry B
There`s also the hope that the gross weight will include the driver, in which case I, along with one or two other Stratos drivers will be looking for a rebate.
:)

[Stratos]

Originally posted by Jerry B
There`s also the hope that the gross weight will include the driver, in which case I, along with one or two other Statos drivers will be looking for a rebate.

:cool: :cool: :cool:

[Stratos]

Originally posted by Jerry B
Chris, I had an interesting chat with John Ryan the tech. bod at RAC MSA who said that there is going to be a rule change soon altering the weight to be linked to engine size. Sub 3000cc will then be 1000kg.
Here`s hoping.



Jerry,

Did he mention a timescale on that, as I thought that these sort of rules only changed for the start of each new season, so that would mean 2004.

[Jerry B]

I think he said they were looking at it at the moment, so I guess it may come in next year Fatty.

[chris.richard]

Are they talking of reducing all the weight limits, or just the one for engines bigger tham 2500cc with more than 2 valves per cylinder?

[Jerry B]

Didn`t ask specificaly, but he did say that the weight was going to be more linked to engine size.

[Andrew Way]

What spring rates would you guys recommend for road use on my HF3000?

Andrew.

[Jerry B]

I worked out that my road spring were about 300F and 185R, I`ve increased these to 525F and 400R for competition use and I will find out soon if they are OK.

As an aside, Leda have made me some front adjustable shocks to match the rears. They are nice bits of kit and I know that Gery Hawkridge is going to be selling them as an option.

[Andrew Way]

I'm having the Leda front adjustable shocks to match the rears and their supplying 200lb for the front springs and 250lb for the rear. I was a little worried these may be too soft.

Cheers.

Andrew.

[JohnB_SPY8808053]

Glad to hear we can get matching Leda fronts now.

What would you guys reccomend for a Hawk for use in tarmac rallies?


John B.

[colin artus]

Originally posted by Andrew Way
I'm having the Leda front adjustable shocks to match the rears and their supplying 200lb for the front springs and 250lb for the rear. I was a little worried these may be too soft.

Cheers.

Andrew.


I think you will find 200lb on the front too low. The actual wheel rate will be much less than this because of the angle of the damper and the rate falls as the wheel moves into bump. I think Gerry supplies 325 lb/in springs with the Hawk and that could be increased a bit depending on the car.

Colin

[Andrew Way]

All,

If 325lb/in is a more sensible figure for the front, will 250lb/in be OK for the back with the V6 lump?

Thanks.

Andrew.

[Stratos]

When I originally converted my car to Leda at the back end, I got 250lb springs from Leda with them, as that was what Ldea suggested I use. However, I found them a bit too soft.

I now use 350lb at the back, but, bear in mind, this is fro a rally car.

If I was using it as a road car, I think I'd use 300 or 325 on the rear.

[MudSnow]

Has anyone ever seen or had a Stratos go up on two wheels while using racing tires on tarmac? Is it possible to roll one (on a hairpin or sharp corner) or will they drift and stay stable?

[Martin K]

Has anyone ever seen or had a Stratos go up on two wheels while using racing tires on tarmac? Is it possible to roll one (on a hairpin or sharp corner) or will they drift and stay stable?

John Rutter may know different - but I never got my car on two wheels - three but never two.

I think if you hit a kirb/rock or run up a bank with the inside/unloaded tyres in a corner, or maybe hit a rut with the outside wheels when in a high G drift - that could tip the car onto two wheels or even make it roll over, but under normal circumstances either is very, very unlikely to happen as to do either the centre of mass of the car has to be outside the track of the car. Stick a pair of heavy wheels and tyres on a roof rack and it might be different!

[rutthenut]

Agreed, three-wheel cornering is possible, but two-wheel antics aren't usually the reserve of any track-going car.
Elk tests with high C-of-G cars, yes.
Grip so high as to go into a roll when cornering, doesn't really happen unless there is contact with a kerb or a rut of some sort. (not this sort of 'rut', I might add)

[Stratos]

The centre of gravity is so low, it almost impossible to get it on 2 wheels. It will spin long before it digs in to go onto 2 wheels.

As Martin says, running over something could upset the balance.

Alternatively, running wide from tarmac to loose might cause the wheels to dig into the ground.

I've tested the situation a few times :eek: , and it just slides or spins.

[Cloggie]

Dave is absolutely right, but again Jerry has proved that you can get a Stratos on two wheels...

[chris.richard]

So Jerry, you've had your Strat on 2, 3 & 4 wheels, when are you going for the final one? :p

[Stratos]

Dave is absolutely right, but again Jerry has proved that you can get a Stratos on two wheels...

I have a picture like that - on the same bump :)

[Arthur]

Hi guys

Just been messing around with a spreadsheet I wrote aeons ago intending to go hillclimbing in a home-build. I actually wrote it while lying supine for 3 months when me first disc prolapsed.

It only does double-wishbones, so I regret not too much comparison between the Corse I and and the "S" and Hawk motors, but never mind.

After a happy morning rolling around under a bog-standard Corse I, and noting that :-
1) I run the platform level at just under 5 inches clearance, which means :-
2) my upper rear wishbones are about level, and the lower rear angles down half an inch to the wheel, and...
3) All the measurements are "best I could do", especially guesses at vertical CofG heights, so
4) Cut me some slack here; inches are hard enough to measure, never mind thou's
5) For the vertical CG heights, I used 17 inches front, and 27 inches rear. If you know any better, pls advise.

I find, in very broad terms, that :-

Front wheel rate is 143
Rear wheel rate is 214
Front wheel frequency (cpm) is 118
Rear wheel frequency is 103
(although while Mr Staniforth calculates frequencies, Carroll Smith does'nt see any point in them. The sadly-missed Mr Smith advised that he baselined the car on a stopwatch with a load of bars, springs, and dampers, then when he had a set-up, he made alterations to it, whatever it might have been, by percentage of wheel rate front and back, which he claimed ... and I believe him... stiffened the car while keeping the fore/aft balance. He also used to trade bars off and put spring on whenever he could).
The calculations I used are shamelessly blagged from Allen Staniforth, and through his publications, David Gould.

The only comparison I have with the Chapman Strut brigade is the effective spring rate (as seen by the wheel) which with my 300 front, 450 rear springs gives 207 front, 310 rear. Front depends on your particular installation; rear should be directly comparable to s strut.

I note here I assumed 980 kilos converted to pounds, split in the ratio 360/540 as per the design Kilos weight. Doesn't make a whole heap of difference - ballpark will do for me.

On the camber recovery front, the outer fronts have pretty near linear camber recovery, at 0.3 degrees per degree of roll.
Rears stay damn near upright whateve the roll angle, going just a tad negative.

With the standard springs, the predicted roll angle at 1 G turn is a horrible 3.4 degrees.

So in a 1 G turn, the front loaded tyre gains just over 1 degree of negative relative to the vehicle, plus 0.5 degree static negative, leaving 2 degrees positive relative to the road. Which is doubtless why the outer shoulders of me tyres disappear in marbles and smoke.

I note that as standard, in that 1G turn, the car only dumps about 35 pounds from the rear onto the front loaded tyre - not too shabby at all. This is improved by more rear spring - with 550 rears and the original fronts, the platform goes round about neutral.
It also means David May's 3 degrees of negative all round static is a damn good place to start.....although upping the front spring rate again will dump more weight to the front. Probably not the way to go with a basically understeering car, but at the end of the day, its the times that count, rather than the engineering.

Going to 450 fronts and 550 rears gets us down to 2.5 degrees potential roll angle in that 1G turn, puts about 45 lbs onto the front from the rear, gets about 0.8 degrees neg camber recovery, and still leaving us wanting in the static department by about 2.5 degrees.

I haven't played around with theoretical roll bars yet - I have pet theories about why they don't play the game as you might expect.

Hope that all is of some interest or indeed use. I'm rather pleased that my purely geometric spreadsheet gives results borne out by actual experience, so I'll sign off with a question and observation session. Any actual answers will be treated with reverence.

Question - if the car rolls around the roll centre, and the roll centre is the intersection of the instantaneous swing axles, what on God's name happens when you pull a wheel off the ground? Really, I haven't a clue. I can't model it try as I might.

Question. Does the car roll around the roll centre? I know it does with the paper models, but then you spend time putting the wheels back above road level, and end up with exactly the same result as just tilting the platform to however many degrees and measuring the result. So what use is the roll centre? Well, swing axle jack for a start. I have this nasty notion that I cannot prove, that as well as the upward force generated, unloading the wheels at whichever end (or both) it also adds an outward component adding to the centrifugal force. After all, the weight cannot go nowhere - it cannot just disappear as a mathematical concept. Gravity, after all, is still looking very closely at the mass, so if it's gone from somewhere, its appearing somewhere else. Of course, mathematical plus and minus signs are wonderful things, and may explain the current F1 practice of parking the roll centre underground, where the jacking force works in their favour, and again, it can't come from nowhere, so I think comes out of the cornering centrifugal force. As I say, I can't bloody prove this; Carroll Smith's drawing is, I think, wrong, and I can't find any other help.

Observation. You can calculate a roll bar to death - lever arms, angular stiffness, all of that - but..........
At the loaded side, the force is applied at the wheel, compressing the spring and twisting the bar. The spring and the bar can therefore be considered as two springs in parallel. At the other side --- well, of course, the force is fed in through the bar, making it two springs in series, with a deadweight hanging on them. Add to that that the bar can't twist unless there's some resistance, and most of that resistance will come from the other side damper, which will be a variable depending on how fast you try to move it. I haven't been able to model this, either - too many variables, all interdependent. Probably why a bunch of bars and the widest adjustment range you can build in is the way to go. (I don't doubt that they are useful little buggers, just I don't subscribe to any notion of mathematical precision concerning their use).
Add to that, if you use something like a 5/16 solid seel rod (road Mini rear bar) it doesn't twist anyway - it tries to adopt an elongated S-shape, and add to that the lever arms bend instead of twisting the bar - well, you doubtless catch my drift.
I draw your attention once again to the illustrious Mr Smith's preference of replacing bar with spring wherever he could.

Last thing - spring resistance. Particularly the reduction of same with inclined dampers. Best bet is not to have any, for which mounting the damper top above and outboard of the upper inner suspension pivot will work just fine and dandy. Which is how the Corse I is, so I haven't bothered with theoretical reductions in spring resistance.

And finally the Cors I itself. I believe the work was done by Steve Greenwood, and this is what I think he had in mind when doing the concept work :-
Fronts - should be stiffer than the rears, should have as much camber recovery as he could build in. Braking - not really a problem with Radial tyres, and the front tyre widths not excessive, so generation of negative in bump and positive in roll not major concers.
Rears - should be pretty damn well pinned down with very little camber recovery, or indeed positive camber generation, I suspect mainly to suit the huge (345's ?) tyres popular on the Strat. Should be reasonably soft to accommodate the power.
Basically the car was aimed to be a fast, driveable roadie. And I have to say, those aims were pretty well achieved. It is stiff at the front, almost a requirement, cos the driver tends to drive to the stiff end of the car. The neg camber generated at the front does a great deal to temper an otherwise very understeery car, and the upright rears minimise any tendency to oversteer into the weeds on an ill-judged passing of wind. And the wheel frequencies are even staggered to prevent any pogo-ing. The rates are on or about the advised starting points for reasonably sporty machinery (generally held to be about 100 to 120).
The front roll centre is about 2.3 inches static, the rear is about 0.8 inches, so if it picks a wheel up at all, it should be a rear.

Lastly, if anyone has accurate dimensions for suspension pick-up points, or even the original design data, I'd be very interested to hear. For the suspension points, I need distance from centreline, and height above ground level to define each point. Where there's anti-anything (squat or dive) either split the difference in height, or my preference would be to take the height of the end that's not been moved, if you see what I mean - on the Corse !, the front suspension forward arm has been dropped a tad, so I used the aft, short, arm. At the rear, again the forward long arm has been raised half an inch, so I used the shorter rear arm location.

All the best, Lads
Arthur.

[chris.richard]

A rough estimate on my parameters produced 148 cpm rear and 125 cpm front :eek: :( .

(Thanks Martin)

[rutthenut]

I'll have to try and hunt down the spreadsheet I created some years ago with Hawk figures.

Plenty of comments there 'for discussion', I see Arthur...

One point I would particularly question is the importance of keeping tyres close to vertical for maximum grip under braking or accelerating in a straight line. Having three degrees of static camber means reduced grip in those conditions, which can be an issue in the wet - much more so on large Group 4 rear tyres.

[Martin K]

A rough estimate on my parameters produced 148 cpm rear and 125 cpm front :eek: :( .

(Thanks Martin)

Perhaps you are not using the wheel rate, as I calculate 124 cpm front and 120 cpm rear.

Arthur - have you allowed for the unsprung weights in your calculation of wheel frequencies?

Two points I have noted from what you have written. One. A car will be understeery with a stiffer front than rear, and the wheel frequencies show this to be the case with your car. Staniforth (and others) reckon that to avoid porpoising the the rear should be 10cpm higher than the front wheel frequency. Though there is calculation based on wheelbase to get this a bit more theoretically accurate, 10cpm is a good starting point. The stiffer front springs, together with the higher roll centre at the front mean that in cornering load is being transfered from the rear inside wheel to the front outside wheel. This is bad for traction AND bad for understeer.

Two. In an ideal world the roll axis would run from higher at the back to lower at the front - your setup gives the reverse of this.

I would try to balance the positions of the front versus rear roll centres. Together with stiffening the rear to 134 cpm, or preferably softening the front to 110 cpm, will give a much better balance and more suitable for the road. If my guesstimates of your unsprung weights are correct then a front wheel rate of around 112 lbs/in will achieve this on your car.

On the subject of roll bars, they should be used to control roll! I do not agree with Caroll Smith when he says use springs instead of bars, as springs have a clear primary function - to maintain the tyre contact patch. If that fundamental is overlooked in persuit of a secondary function - controlling roll for example - then it is clear that their performance in their primary function will be compromised. If you do prefer to use springs instead of springs and bars then you will end up with a car that feels hard and quick, but is just hard.

If you chose to follow my suggestions you would end up with a car that rolled more at the front than the rear, but by adding a front anti roll bar the front roll could be reduced probably enough to give the suspension a chance to give proper camber recovery - I am sure that proper camber recovery would have been designed into the car by Mr Greenwood. Play with the calculations and I am sure you can arrive at zero weight transfer to the front (i.e. no transfer from the inside rear wheel) and good camber control.

I think what you have currently is a safe, understeery setup that Mr Greenwood thought would be most suitable for people driving their front wheel drive saloon in the week and then jumping into a Corse at the weekend. If you want more fun, and a better balanced car for performance, then you will need to change a few things in line with what I have said above.

You are correct in thinking that loads in roll when the roll centre is higher than ground level are split between the loads that pass through the springs/bars to the tyre and loads that pass through the suspension links to the tyre. Loads that pass through the links are effectively side forces rather than vertical ones. Vertical loads are good as they increase the vertical load on the tyre which increases tyre grip. Side forces add to the centrifugal force which is trying to slide the tyre across the track, and so they are BAD as they are working against us achieving maximum cornering speeds.

Lastly, when you lift a wheel I BELIEVE that the roll centre at that end of the car becomes the tyre contact patch. i.e. at ground level and at the outside track dimension of the car. As this is where the roll centre will end up, it is important that it arrives there progressively and not instantaneously from it's designed static location, as sudden changes in roll centre location give adverse handling charateristics. That is another reason for having a low static roll centre as part of the suspension design.

I hope I have been able to answer some questions and given some food for thought and further debate!

Martin

[Martin K]

Question - if the car rolls around the roll centre, and the roll centre is the intersection of the instantaneous swing axles, what on God's name happens when you pull a wheel off the ground? Really, I haven't a clue. I can't model it try as I might.

Question. Does the car roll around the roll centre? I know it does with the paper models, but then you spend time putting the wheels back above road level, and end up with exactly the same result as just tilting the platform to however many degrees and measuring the result. So what use is the roll centre? Well, swing axle jack for a start. I have this nasty notion that I cannot prove, that as well as the upward force generated, unloading the wheels at whichever end (or both) it also adds an outward component adding to the centrifugal force. After all, the weight cannot go nowhere - it cannot just disappear as a mathematical concept. Gravity, after all, is still looking very closely at the mass, so if it's gone from somewhere, its appearing somewhere else. Of course, mathematical plus and minus signs are wonderful things, and may explain the current F1 practice of parking the roll centre underground, where the jacking force works in their favour, and again, it can't come from nowhere, so I think comes out of the cornering centrifugal force. As I say, I can't bloody prove this; Carroll Smith's drawing is, I think, wrong, and I can't find any other help.

All the best, Lads
Arthur.

Can I refer you to Competition Car Suspension, page 201 "Weight transfer via the roll centres"

Factor that into your spreadsheet and you will have your answer!

I would be interested to know your front and rear track dimensions so I can play with them, but I reckon with bars you could reduce your roll angle to 1.3 degrees at 1G and increase maximum cornering from 1.3G to 1.6G before lifting a wheel, with zero weight transfer and only 2.1 degrees of roll at the same cornering G.

[Arthur]

Martin,

Front track is 57 inches, rear is 61,5 across the cetntre of the tyres. Wheelbase 85.5 (more for comparison, really - these are after all hand-fabbed frames and may come out diffently car-to-car).

My main concern is the guesses involved with the ball joints at the front - are they true balls, or are they (as per the Mini) half a ball or less? In that case, the true pivot centre could be a way away from my estimate.
Anyone out there have access to the design locations?
Same for theoretical height of CG front and rear - as I say, I used 17 inches front and 27 inches rear - based on back-of-fag-packet guesses at masses and positions of same, with full tank and driver in car. Again, if anyone knows for sure, pls post the info!

As for weight transfer - I allowed 100 lbs per corner unsprung, using 2156 lbs total weight for the car. Maybe a bit high - but that would mean more sprung weight, and lower actual frequencies.
For comparison purposes, I measured my damper leverages as 0.69 (or rather, 1 divided by 0.69 = 1.45 per 1 inch wheel travel. Pretty linear as far as I could measure, but may get wayward if you have a lot of damper travel. Same leverage front and rear.

Roll centre inclination - well, the car will tend to skew around the line for sure. Escorts (low front, axle centre height at the rear) kicked up a front tyre. Minis (on the ground at the rear, low but above ground front) kick up a rear tyre.
Can't see that matters particularly, unless you can skew the car far enough to pull a loaded wheel off the deck, which is in any case usually the province of restricted droop travel.
In fact, on the spreadsheet, it only transfers 35 lbs forward in that theoretical 1G corner.
If I stiffen the rear, I can get that to almost zero without bars (300 lb front springs, 550 lb rears).
I suspect, as you do, I think, that Mr Greenwood sacrificed an ideal roll centre inclination to keep the rears as upright as he could through full travel. And yes, it means they go hugely positive in cornering, as borne out by the wear patterns. But radial tyres generally don't worry overmuch about camber under braking - or at least you can trade camber grip for brake without too much going wrong, if you can live with the inner shoulder wear. On a wet road, I'd guess riding on the shoulder of a 345 tyre would give more grip than having it flat down (much higher point load on the rubber) but it would be highly likely to aquaplane. Don't know, and have no intention of trying to find out.

I can get it to zero with bars as well.
Problem with bars is that if you soften off the dampers a couple of clicks, the resistance at the unloaded end of the bar falls, and the bar looks softer to the suspension. Up a few clicks, and it becomes a solid axle conversion kit (as the man used to say). So what I meant was that if you always find yourself running some theoretical minimum value of roll bar at one end, then the chances are you're short of that much roll resistance at that end, and you may as well replace it with spring. You have a more predictable set-up, and can use lighter bars when you need 'em. Again, lighter bars have less scope for doing weird things with every tweak.

Suspension hurts my mind, so I'mm off. By the by, I've been advised that If I want paying any more, I'd better get my backside to Brisbane come Friday, so I'll be off again for months in a day or two. I'll take me spreadsheet and calcs away, and see what I can come up with that doesn't involve re-building the car!

Thanks for the input
Arthur.

[chris.richard]

Perhaps you are not using the wheel rate, as I calculate 124 cpm front and 120 cpm rear.
Martin

Was that aimed at me Martin?

If so, these are the figures you quoted me, so I sort of assumed they were correctly calculated..... ;)

[Martin K]

Was that aimed at me Martin?

If so, these are the figures you quoted me, so I sort of assumed they were correctly calculated..... ;)

Yes Chris it was aimed at you - but I was a little confused as to what rates you were quoting but I thought you might have worked out Arthur's rates from his data somehow. Sorry!

[chris.richard]

Martin what factors do you use for converting spring rate to wheel rate for the standard Hawk front & rear setups?

[Martin K]

Chris,

The following apples to the Hawk only!!

The rear is easy. There is no lever acting on the spring, so the wheel rate is the spring rate. As the is the strut is not vertical, (i.e. inclined) then the wheel rate is a little less than the spring rate, but not much as the angle is small.

The actual wheel rate is given by spring rate x cos(angle of inclination). If the strut is inclined at 5 degrees, then the wheel motion is 0.996 of the spring motion, so wheel rate = (0.996)^2 x spring rate, or 0.992 spring rate.

At the front it is more tricky, as you have a lever due to the damper being mounted on the track control arm inboard of the lower ball joint. The damper/spring is inclined which also reduces stiffness.

Wheel rate is given by:

spring rate x lever arm^2 x (cos(angle of inclination))^2

From memory (and approximately):

The lever arm is 260 mm / 360 mm, so the lever arm squared is 0.522
The damper is at 45 degrees and the cosine of 45 is 0.707, so squared is 0.499.

So the wheel rate with a (standard?) 325 lb spring is:

325 x 0.522 x 0.499 = approx 85 lb.

A 450 lb front spring gives wheel rate of 117 lb.

Wheel frequency is a function of the wheel rate and the load on the particular 'corner' (i.e. the sprung weight) .

The Corse has a different lever arm ratio and a different inclination to those shown above, so although the same principles apply, the numbers above don't.

[Martin K]

Just thought this thread was re-visiting previous topics covered in this other thread, so take a look at

http://www.stratossupersite.com/forum/showthread.php?t=1925&page=1&pp=15

[rutthenut]

The rear is easy. There is no lever acting on the spring, so the wheel rate is the spring rate. As the is the strut is not vertical, (i.e. inclined) then the wheel rate is a little less than the spring rate, but not much as the angle is small.

I don't entirely agree with this point, even though this is how Stanniforth and others describe it. Some of my views are expained below, bear with me.

My bone of contention is around the position of the tyre contact patch, and the amount of offset that may exist in the hub/hub carrier assembly.

In reverse order, the hub carrier tends to sit 'outside' of the strut centre line, so could be viewed as providing some leverage against the spring, perhaps measured against the length of the lower wishbone/radius arm. The hub itself is off to the side of the hub carrier, which may introduce more offset into the equation.

With the Hawk suspension, the Chapman/McPherson strut design uses a reversed lower wishbone, so that could put forward a ratio of say 12:10 against the spring.

That can be further confused when including the forward-facing tie bar, which effectively creates a wide-based wishbone, still having an inner pivot point close to the lower chassis rail. Probably best to try and ignore that point for spring rates, and only look at it for control of camber/castor and changes in roll/dive/squat where I think it would have more effect.

I think that a lot of fwd cars use McPherson struts that will also have an amount of offset between the hub and the strut centreline, but they will also tend to run with wheels having corresponding offset (or inset?) so that the centre point of the tyre contact patch will be close to the point where the strut centre line would meet the ground. Moving these points inboard or outboard of each other can give different steering reactions and scrub radius (may also be a consideration for ABS systems).

Anyway, point of that is that if the strut centre line corresponds to the centre of the tyre contact patch (width-wise) then I would agree that the spring rate and wheel rate are the same. I guess that's what most of the equations are based on for these calculations.

But on the rear of the Stratos (Hawk suspension, at least) I think that the tyre contact patch centre will be outboard of the strut centre line by an inch or two for standard-sized wheels, depending on ET figures.

With Group 4 wheels and tyres of perhaps 12 inches width, then the centre of the contact patch will definitely be moved outboard by a number of inches. I am sure that just using the centre point would be enough to introduce an element of leverage over the spring. That would change the wheel rate when seen from the perspective of the vehicle travelling over a bump or experiencing cornerning force and [potential] body roll.

I also think that the centre of pressure or load will move dynamically from side to side across the tyre, either due to uneven road surfaces or body roll. The wider the tyre, the greater movement can be expected across the contact patch, and the greater variation in effective spring rates.

In conditions of body roll, I would expect that suspension leverage could be calculated against the spring from a point on the outer shoulder of the tyre. At least that is the area with the greatest effect (in reducing spring rate, or leading to greater body roll). I don't think it is quite as extreme as that calculation may show, but somewhere between that and steady-state figures.

The textbook calculations effectively work with a disc for geometry figures, rather than a tyre of some inches in width. Real world tyre sizes could play a part in throwing the calculated figures into a semi-random mix.

So, to try and sum up:

I think that strut-type suspension calculations should take into account any difference between the tyre contact patch (middle, or inner/outer extremes) and the projected position of the strut centre line.

FWD cars tend to have hub/wheel geometries on McPherson struts that reduce this effect.

Stratos-type Chapman struts - especially with Group 4 rears - will have an amount of leverage over the plain spring rate, giving a reduced wheel rate.

Other opinions on all of this are more than welcome.


However, the calculations are intended to provide figures as a means of comparison, which will still be true even if the numbers aren't quite 'correct' in real terms. At least if looking to change spring rates at one end of the car, before-and-after figures will make sense.

If trying to create a particular resonance difference between front and rear wheel rates, it might mean you get slightly different results to that expected, but driving the car in the target conditions will be the final arbiter.

Enough already.
J.R.

[Martin K]

John

I have often pondered over this problem also, but have always deferred to the professionals on this one!

The only way to check your theory is to measure some movements - which I confess I haven't done:

If the spring is removed from the strut and the wheel moved through, say, 50mm of vertical travel, what happens to the gap between the spring seats?

If it measures the same 50mm then the motion ratio is 1:1, so the wheel rate:spring rate ratio will be 1:1.

If the movement between the spring pans about 42mm as your suggested 12:10 lever arm ratio would give, then you are correct and the wheel rate would be only 0.69 of the spring rate.

You could also try two further measurements, 1) jacking the outer edge of the wheel and 2) jacking up the inner edge of the wheel and see if there is any difference in the resulting gap in the spring seats.

I believe the track dimension to the middle of the tyre is used in calculations, rather than the measurement to the outer edge of the tyres, because with camber recovery the centre of pressure should still be in the middle of the tyre.

One thing to note is that if the wheel rate does fall with wider wheels on a Chapman strut suspension, the roll resistance will never the less increases with the increased track they will give. I'll play with some numbers to see if these two factors cancel each other out! However, your point that the wheel frequencies would change is valid on the basis that your measurements reveal a lever arm at work.

Martin

[CorseChris]

Can I add my (probably erroneous) thought to this. I believe that the spring rate = wheel rate is valid, since the upright/strut/stubaxle is a solid unit that pivots at the bottom with no freedom between wheel/spring, unlike a wishbone say, where the pivot uncouples the spring & wheel. There is no lever ratio between wheel and spring. Increase in offset would certainly increase bending loads on the strut/piston rod and side loads on the top strut top mounting though as the assembly tries to rotate around the strut bottom joint.

Lots of sketches convince me this is right. But I'm crap at drawing ;)

[Martin K]

Can I add my (probably erroneous) thought to this. I believe that the spring rate = wheel rate is valid, since the upright/strut/stubaxle is a solid unit that pivots at the bottom with no freedom between wheel/spring, unlike a wishbone say, where the pivot uncouples the spring & wheel. There is no lever ratio between wheel and spring. Increase in offset would certainly increase bending loads on the strut/piston rod and side loads on the top strut top mounting though as the assembly tries to rotate around the strut bottom joint.

Lots of sketches convince me this is right. But I'm crap at drawing ;)

That's more or less my thinking too. As there is only one inner and one outer pivot point, which is both the lower ball joint and the effective connecting point of the strut (spring if you like) to the wishbone, there is no lever effect via the wishbone to give a differential in movement. Therefore as the motion ratio is 1:1 the spring:wheel rate is 1:1 also.

[chris.richard]

Don't think I agree! :eek:

The presence or absence of a pivot at the bottom of the strut doesn't affect the leverage. in the diagram below, the force F2 on the spring , will be 3x the force F on the wheel .

If not, why not? :confused:

[CorseChris]

In the diagram you have drawn, I agree with your figures. But that doesn't show a strut configuration. That's a swing axle (I think). Where you have the inner pivot about which your system rotates and about which leverages can be calculated, in a strut, this point is moving freely up and down and can't be used to generate a fulcrum point. It can't transmit any torque.

As Martin said, simple test. Jack up a wheel, see how far the spring pan moves. If it's not the same as the distance the wheel moves then there is a ratio to consider. Personally, I'd be worried if there was any relative movement between wheel and strut body.........


As another approach (to add yet more confusion) consider a wishbone setup. The ratio is calculated between inner pivot, spring attachment point, outer pivot. No consideration is given to the offset of the upright/hub/rim assembly for exactly the same reasons - the lower outer balljoint prevents any transmission of torque.


What offset will do is to change bending loads in the strut. This will change stiction & friction but that's a tough one to deal with in our simple model. I think we'd have to consider it fairly constant? There is also the small angle through which the strut moves to consider as well, but it's pretty small in a well set up system so can be pretty much ignored I think.

[chris.richard]

Where you have the inner pivot about which your system rotates and about which leverages can be calculated, in a strut, this point is moving freely up and down and can't be used to generate a fulcrum point. It can't transmit any torque.
Yep, I see the error, I'll get my coat...... :o

[CorseChris]

I just hope I'm right!! :))

[Arthur]

Just checking my publications.

Seems that Carroll Smith addressed this in his "tune to win" book, in the bit about "pneumatic trail". He estimates there that the effective contact point of the tyre footprint is aft of centreline, outboard of centreline, and moves further outboard and aft with increasing load.
Didn't used to matter (and still doesn't) with skinny wheels, but looks as though it has the potential to be a right bugger to evaluate with wide and soft tyres.

As above, the only way to find out if you have leverage built-in is to :-
A) measure wheel travel against spring platform movement
or the engineering answer B) - do some calcs using figures either side of actual, and see what margin of error you're talking about, and whether or not it's going to matter. If lack of brakes, say, is your first problem, no point sodding about with srpring rates til you get braking right. And strictly speaking, if you start winning, there's no point til someone catches you up.

i.e. Mr Staniforth reckons there's little value in worrying about an inclined spring/damper if it's less than 30 degrees. Still softens, but not enough to worry about.
Carroll Smith reckoned when tuning to change something by not less than 10% of wheel rate at a time (wheel rate, not frequency, and not spring) or you might not feel the change well enough to evaluate. Which more-or-less gives you a size for the window to work in - if you can calculate reasonably to +/- 5% of what's happening, it's close enough to work with.

Night, lads.

[Arthur]

Mind you, sometimes my brain starts working.
It's a worry.

The leverage of the lower arm over the spring/damper will be fixed regardless of the wheel, offset, or anything else, so the calculation for rate is valid as previously stated.
For a MacP strut, you'd have to project a line from the strut top centre through the mounting to the upright, to the control arm centreline. If this is offset from the outer control arm pivot point at the upright, then you have a leverage to consider. Since this leverage is squared to get the rate, it will have some effect. But I predict it will be a sod to measure, as are small changes in spring pan distances.

BUT - the weight transfer (which is what we're trying to estimate) is a function of mass, CG position, and track.
Wider track = less weight transfer.
So if the effective track moves around with wheel load, the actual weight transfer will change with wheel load, and that will affect the actual frequencies even though the leverage and the calculated rates remain fixed.
I think........

I want to get back to the day job. This is causing the remaining grey matter to melt.

Arthur.

[David May]

I recon that any change (sprng, dampers, brakes) under 20% is impossible to evalute under normal conditions without complex telemetry. I usually go for 30%, with the proviso that I may have to step back if I overdo it - never happened yet.

[chris.richard]

OK, here's some figures.
Hawk rear suspension. Measured travel from static compression to full droop or vice versa :rolleyes: (coz I was jacking the back wheels off the ground).
Spring pan travel = 24mm
Wheel travel 25mm
The 1mm discrepancy would be within the bounds of measurement error, I'm sure, And it's only 4% anyway. So the 1:1 spring to wheel ratio looks correct.

[rutthenut]

Certainly shows a few different interpretations of all this spring rate geometry, doesn't it?

I can see how having strut base/hub carrier/hub/wheel bolted together would give the same movement wherever it is measured (and similar points would indeed apply to upright/stub axle/hub assemblies on wishbone suspension setups). However, something in my mind still wants to consider plain leverage from my schoolboy physics. If the lever is longer, I would expect some sort of mechanical advantage. Maybe the fixed joint is what makes the difference?

Happy to concede that this does not necessarily affect the strict maths, but still have a suspicion that wheels with especially large offsets would have some input to the numbers. Be that as it may, they are only numbers for comparison against 'ideals'...

J.R.

[Martin K]

If I have managed to attach my sketch, the upper one shows a strut arrangement with it's three pivots and the wheel mounting spindle. It is clear that the spindle is not an extension of the lever arm, as there is a pivot point at the bottom of the strut.

So, if the outer end of the spindle is raised by one unit, then the inner end of the spindle, and the lower spring pan, will rise 1 unit.

The only way there could be a difference is if the spindle bends and some motion is lost.

In the lower drawing (ok, it' s line!) there is only one pivot and moving the outer end one unit will result in a movement of < 1 unit all the way along the arm towards the pivot. I must get myself a CAD package.....

[Martin K]

I recon that any change (sprng, dampers, brakes) under 20% is impossible to evalute under normal conditions without complex telemetry. I usually go for 30%, with the proviso that I may have to step back if I overdo it - never happened yet.

I think the amount of change that can be detected by a driver depends a great deal on the experience of the driver, and the amount of time he has spent driving a particular car. Experienced, sensitive drivers are SAID to be able to detect a 10% change. When the car is balanced to the drivers liking experienced race engineers will work with the driver - and the stopwatch, to make changes of less than 10% which improve performance, perhaps without the driver feeling any change in the handling at all.

There can be situations where a 50% change - or greater - makes no difference to the feel of the car. I know from personal experience! If a car is so badly out of balance then sometimes huge changes are necessary to get the balance back.

The purpose of this discussions, for my part, is to get the cars into reasonable front/rear balance with the ball park figures for wheel frequencies and the spring rates that will achieve them, rather than guessing what is going on. If on a Hawk you increased both the front springs and rear springs by, say, 10% then you wouldn't stiffen the car by 10% all round and keep the balance you had: You would be increasing rear stiffness at a greater rate than the front and therefore the car would be more over-steery than before. To keep the balance the fronts would need to be increased by about 14% rather than 10%. David, if you are adding stiffness in 30% chunks to your spring rates all round then this will clearly change the understeer/oversteer balance, which I am sure you have noticed.

The purpose of looking at wheel frequencies so closely is that there is an obvious link between wheel frequency and grip/feel/comfort and surface type - a link that is not so easy to see in spring rates. For example, to maximise performance on smooth tarmac, a 10% increase in wheel frequency might be required, but this can require a 40% increase in spring rate on the rear of a Hawk. To change a Hawk from a tarmac track car to a forest car may require a decrease in wheel frequency of 30% which would require reducing the spring rate by 50%. To make these changes in 10% increments before arriving at an ideal (for the driver) setting would be a time consuming and expensive business if a new set of springs has to be purchased and tested every time, so far better to get some reasonable initial rates and have a reasonably good starting point for tuning from.

[David May]

I think we both agree - I was referring to the road setup of a road car, not a professional track-test. I also have learnt the hard way never to change any more than is essential at a time. I prefer to do springs one end at a time, to get a feel for that change, usually playing with the dampers to suit that end first. I know there's a valid question of overall balance, but if you change all 4 springs and 4 damper settings, its very hard to know what caused or cured what.

[chris.richard]

OK, here's some figures.
Hawk rear suspension. Measured travel from static compression to full droop or vice versa :rolleyes: (coz I was jacking the back wheels off the ground).
Spring pan travel = 24mm
Wheel travel 25mm
The 1mm discrepancy would be within the bounds of measurement error, I'm sure, And it's only 4% anyway. So the 1:1 spring to wheel ratio looks correct.

Which on a 400lb spring makes my rear nearside cornerweight 175kg :confused: = approx car weight 581kg :eek: :confused:
Or are springs not linear and increase once preloaded?

[Martin K]

Which on a 400lb spring makes my rear nearside cornerweight 175kg :confused: = approx car weight 581kg :eek: :confused:
Or are springs not linear and increase once preloaded?

Doubt it!

Any pre-load does affect the calcuklation. What is the free length of the spring, and what is the compressed length at ride-height?

[chris.richard]

It's 9 inch free length, but it's compressed to allow spring pan to clear tyre. So I can't do a proper unloaded to static compression measurement.

[Martin K]

It's 9 inch free length, but it's compressed to allow spring pan to clear tyre. So I can't do a proper unloaded to static compression measurement.

Working out the corner weight is still quite straightforward.

If the free length is 9 inches and the compressed length is, say, 7.5 inches, then the corner weight would be spring rate x deflection, which equals 400 lbs x 1.5 which is 600 lbs or 272.1 Kg.

By deduction from your figures, the change in length of the spring between ride height and jacked up with the wheel clear of the ground is 0.96 inches? If so, then given the weight I have guessed above, the preload would be .54 inches or 216 lbs.

[chris.richard]

I don't understand the physics of this I'm afraid.
1) I thought the springs were linear, i.e. it takes x pounds to compress it one inch, whether it was at full extension to start with, or if it was already compressed one inch.
2) How can the fact that the spring is held partly compressed alter the corner weight of the car? It seems from what you say, the higher i wind my spring pan to raise the ride height, the less the spring will compress when I drop it back on its wheels., yet the jack still supported the same weight.

Sorry - thickness abounding

[Martin K]

I don't understand the physics of this I'm afraid.
1) I thought the springs were linear, i.e. it takes x pounds to compress it one inch, whether it was at full extension to start with, or if it was already compressed one inch.
2) How can the fact that the spring is held partly compressed alter the corner weight of the car? It seems from what you say, the higher i wind my spring pan to raise the ride height, the less the spring will compress when I drop it back on its wheels., yet the jack still supported the same weight.

Sorry - thickness abounding

Your assumption in 1) above is wrong.

If the spring is pre-loaded (i.e. partially compressed, which it will be if it measures less than 9 inches when the car is jacked up) then already the spring 'thinks' it has a load on it. Any load upto the pre-load amount will not compress the spring further. Load above the pre-load amount will.

2) The corner weight is not changed by the spring being pre-loaded. I think my previous post shows that.

However:

If your car weighs 600 lbs on the rear corner and you wound up the lower spring collar by 1.5 inches, so, that with no weight on the strut, the length of the spring was 7.5 inches, then when the car is lowered the spring will not compress further. However, this does not mean that no weight is transferred to the ground! Of course the full 600 lb will be transferred to the road, but as the spring is already pre-loaded by the amount it would compress if 600 lbs were loaded onto it, then it will not compress any more when the car is lowered onto it's wheels.

Is that clearer?

[chris.richard]

CLINK



(sound of penny dropping) :o

[chris.richard]

So for the frequency calculation do we use the calculated rate for the preloaded spring rather than the nominal poundage? :confused:

I get this horrible feeling I'm going to regret asking this..... :rolleyes:

Creak, creak..... the sound of a can of worms being opened. :)

[Martin K]

Don't panic!

You use the nominal spring rate, because the calculation takes into account the load on the spring.

The only exception to this would be if the pre-load exceeded the corner weight. Unlikely ever to be seen, but if that was the case you would use the pre-load weight rather than the corner weight in the calculation.