46 replies to this topic

[Supercar]

Okay, I am now brave enough to ask my first question. It has been puzzling me for quite some time and I think I have exhausted all the printed resources...

I know how to select a set or springs to make the car balanced and drivable. I can then add sway bars to fine-tune it, or to add roll stiffness. If I have too much roll stiffness already I could install 3rd springs (they act like a pro-roll bars of a kind by working only in jounce but not in roll). Where do I go from here? Immediately there are a lot of questions in my mind:

If I want to make one end stiffer, would I add spring rate or sway bar rate? Does the ratio of the front/rear suspension natural frequencies matter for high downforce race cars to keep the car from pitch oscillations? How about those weird cars with no sway bars on one end and 50 mm bars on the other? Is it all about aero, or is there some basic vehicle dynamics to this? What about no-downforce race cars?

I found some partial explanation unexpectedly in Carroll Smith's "Drive To Win". But it does not seem very logical. It goes something like this... "If you get wheelspin on a bumpy corner exit you need to install a softer rear sway bar to reduce the lateral load transfer and then install stiffer rear springs to balance the car." or "Install stiffer front sway bar to load the inside rear and balance the car with softer front springs or stiffer rear springs".

To me it sounds like the roll stiffness and the balance do not change. It even seems that the rear jounce stiffness increases in this example. So where does the all the *magic* come from? Any comments?

Philip

[Lukin]

If I want to make one end stiffer, would I add spring rate or sway bar rate?

Good question. Unfortunately I dont know the proper answer, and it's something I have been wondering the last few weeks.

Taking a shot in the dark, I think that in terms of overall roll stiffness (say if you want a specific roll gradient), it doesnt matter what resists the roll, be it springs or ARBs. However, you have to take into account the effect each has on other parts of the car balance/performance. Like that passage from the Carol Smith book alludes to, to get the desired overall roll resistance, you could have a combination of soft springs and stiff ARB or stiff springs and soft ARB, however softer springs give better traction so the first option is better in that case.

That's an easy example I guess, but overall Im at a loss to know which is the best method to achieve what you want, and which steps will have you (and the driver) chasing your tail.

[Antoine]

I think you have to consider the entire lap!
because if you put a stiffer spring on the front you will have a lower pitch angle, more stiff in roll!
if you change the roll bar, you increase the roll stiffness but you keep the same pitch angle!
so you increase turn in and under/steer in mid corner.

[Lukin]

Cheers Antoine I didn't think of it like that.

[Supercar]

Thanks for the answers, guys. But it seems like there is more to it. If you think that there is nothing else then please see if you can answer *all* the questions from my original post. I had like 7 question marks there!

Note that in his example Carroll Smith is not looking to make the car under/oversteer. Apparently he is just changing the way he is achieving the same roll stiffness on each end of the car. That may take us to discussing camber curves, ride heights, natural frequencies and such.

Why do Porsches run ridiculously stiff springs in the front? How did hmm... Nascar end up with no rear bars and 50 mm front bars? Would I ever want to have no front bar on a car and a very stiff rear bar? I hope there is someone here who has done something like this before and can put it all together.

Philip

[red300zx99]

Usually you try to make the suspension as soft as possible, as stated before for 'traction' or 'grip'. Using all availiable jounce travel. Taking that into consideration you try to create desirable camber curves through that range of motion. So if your in a corner at max G your about at the limit of travel on the outside wheel. Different things are on the inside wheel. Make it common sense to leave alittle slack for a bump or two, you dont want to bottom out or hit whatever is restricting your travel. Another thing to watch for how much the car pitches, which is usually only controlled by springs, but arbs can have their effect. You don't want the car pitching over to far and scraping the nose if that is an issue, I've seen that before and it doesnt sound good. You can also control that with some antis and such.

Porsches, stiff springs, I would guess it has something to do with the long cg location, but I dont know, never worked on one.

The NASCAR question brings up an important point. Sure there is no arb in the rear, but the location of the RC is easily moved giving you an easy way to adjust the roll couple, roll stiffnesses tell me little without the moment arm. 50mm? 2 inch bars and bigger are found time and time again, different schools of thought all around though. I've seen 80lb springs in some cars. I may try a paper clip someday.

[Engineguy]

Originally posted by Supercar
Why do Porsches run ridiculously stiff springs...
Philip

When you're stuck with really bad suspension geometry... you don't want the suspension to move

[Antoine]

I think you have to consider the roll center location before to compare two car and their anti roll bar, no?

I'm agree with Engineguy about Porsche suspension and I can add that If your car is pitch sensitive you need to put this kind of stiffness! even if it's not logical!

[Supercar]

Thanks, guys! So if I do not have any suspension "defects", would I then want to stay close to the 1.1-1.2 rear/front natural frequency ratio, or this is never a consideration in high-downforce cars?

And what about Carroll Smith? In his example to reduce the rear wheel spin he seems to prefer stiffer rear springs and a softer rear sway bar to keep the car balanced.

Do stiff rear sway bars reduce traction? I speculate that if there is a bump under one wheel then it would disturb the other wheel's contact patch through the sway bar. Is that a real concern or just my imagination?

We all know that softer springs provide better traction. That is if you can allow softer springs and not exceed the available suspension travel. Or not cause an adverse camber. Or not be forced to run large ride heights sacrificing the downforce. That part I understand, thank you guys.

Philip

[Ben]

Originally posted by Supercar
Okay, I am now brave enough to ask my first question. It has been puzzling me for quite some time and I think I have exhausted all the printed resources...

I know how to select a set or springs to make the car balanced and drivable. I can then add sway bars to fine-tune it, or to add roll stiffness. If I have too much roll stiffness already I could install 3rd springs (they act like a pro-roll bars of a kind by working only in jounce but not in roll). Where do I go from here? Immediately there are a lot of questions in my mind:

If I want to make one end stiffer, would I add spring rate or sway bar rate? Does the ratio of the front/rear suspension natural frequencies matter for high downforce race cars to keep the car from pitch oscillations? How about those weird cars with no sway bars on one end and 50 mm bars on the other? Is it all about aero, or is there some basic vehicle dynamics to this? What about no-downforce race cars?

I found some partial explanation unexpectedly in Carroll Smith's "Drive To Win". But it does not seem very logical. It goes something like this... "If you get wheelspin on a bumpy corner exit you need to install a softer rear sway bar to reduce the lateral load transfer and then install stiffer rear springs to balance the car." or "Install stiffer front sway bar to load the inside rear and balance the car with softer front springs or stiffer rear springs".

To me it sounds like the roll stiffness and the balance do not change. It even seems that the rear jounce stiffness increases in this example. So where does the all the *magic* come from? Any comments?

Philip

What Smith means is that although the balance is the same using a simple linear load transfer calc, the improvement comes from not having the ARB lifting the inside wheel when the outside wheel hits a bump. An ARB effectively removes the independance of the two wheels and on a bumpy track this can cause the inside wheel to loose a lot of grip because the average load on the wheel is lower because it's oscillating with a dependancy on the outside wheel.

The best way to look at these problems is to consider the RMS load on the wheel over a time history rather than a single load at an instant in time.

Also, ignore Smith when he talks about having equal front and rear roll moments. The mass centroid axis is a fiction and cars don't roll about the kinematic roll centre. Remember, Smith wrote his books before multi-body systems analysis became common. This is one area where his work is in error.

Ben

[Supercar]

Originally posted by Ben
What Smith means is that although the balance is the same using a simple linear load transfer calc, the improvement comes from not having the ARB lifting the inside wheel when the outside wheel hits a bump. An ARB effectively removes the independance of the two wheels and on a bumpy track this can cause the inside wheel to loose a lot of grip because the average load on the wheel is lower because it's oscillating with a dependancy on the outside wheel.

The best way to look at these problems is to consider the RMS load on the wheel over a time history rather than a single load at an instant in time.

Thank you, Ben. Ben, you are the Man! I had all the bits in pieces in my head already but they finally all came together after your comment about the disadvantages of stiff anti-roll bars and the RMS load on the wheel. I will be able to calculate it once I install strain gauges on those pushrods!

So it looks like on rough tracks I will have more of my roll stiffness come from springs, and on smooth tracks from bars. For rough tracks I will have softer bars. I wonder if I should have stiffer springs or just raise the ride height and keep the same springs... or even have softer springs? Need to run some math and test it. That aero... can't predict it over bumps, need to test!

Now the NASCAR setup also makes sense. Their rear suspension is a live axle, so they are trying to make it as independent as they can by not installing a bar there. They also do not want either of the rear wheels to move and upset the other one, so stiff springs and no rear bar is the answer for the rear. The front, on the other hand, is decent. But over bumps that front would outperform the rear, so they put a 2" bar on it so that over bumps it performs as bad as the rear and on a smooth track it performs as well as the rear. This keeps the car balanced regardless of the bumps!!!

Also, ignore Smith when he talks about having equal front and rear roll moments. The mass centroid axis is a fiction and cars don't roll about the kinematic roll centre. Remember, Smith wrote his books before multi-body systems analysis became common. This is one area where his work is in error.
Ben

Yes, I know, he wrote his books over couple decades, so he contradicts and sometimes corrects himself on several topics. Where could I read on multi-body systems analysis? Is this different from slicing the car like a loaf of bread and running a string (centroid) though it?

Thanks again!

Philip

[McGuire]

They do run a rear ARB In NASCAR...and more all the time given the current trend in chassis setups: extremely soft springs and a wide spread in spring rates across the rear.

[Supercar]

Originally posted by McGuire
They do run a rear ARB In NASCAR...and more all the time given the current trend in chassis setups: extremely soft springs and a wide spread in spring rates across the rear.

Like the inside (left) rear is much softer than the outside rear? :

[Ben]

Originally posted by Supercar


Where could I read on multi-body systems analysis? Is this different from slicing the car like a loaf of bread and running a string (centroid) though it?

Thanks again!

Philip

Multi-body systems analysis means programs like ADAMS or DADS. The best reference is this book by a Prof from Coventry University and Damian Harty who's the chief dynamics engineer at Prodrive. We have it at work and it is highly recommended.

http://www.amazon.co...5231895-8880447

Ben

[Supercar]

Thanks for all the comments! I just have the last little question I need to know, then I will know everything!

Do I want to stay close to the 1.1-1.2 rear/front natural frequency ratio, or this is never a consideration for high-downforce cars? Thanks again.

Philip

[Fat Boy]

The 1.1-1.2 frequency ratio suggested by Ollie was for good ride. It really doesn't have anything to do with racing.

On an aerodynamic car, I tend to run with a lower ride frequency in the rear for a couple reasons.

1. I always tune my cars to have good corner entry characteristics (at the expense of mid-corner balance...I've been through the reasoning before). This means relatively stiff front springing and relatively soft rear springing.

2. As the aero load is applied to the car it will tend to reduce the amount of rake in the car. If you are reducing the amount of rake in the car at high speeds, then you are moving the CP back as speed increases. The driver will thank you for this in a fast corner.

-----------------------------------------------------------------------------------------

You need to take the chassis tuning of NASCAR racecars with a pinch of salt. They are extremely hampered by the rules and their particular form of racing. Because of this, they tend to do things that other racers don't do. They have minimum ride heights which is why they run the ultra soft spring / stiff bar setups. As the car rolls, it compresses the inside spring which drops the ride height. It's not a roll couple thing...it's the only way available to get the car on the ground.

They will stiffen the rear spring to _reduce_ oversteer. It keeps more dynamic rake in the car which keeps the rear spoiler more in the air stream. That reduces oversteer.

They do a lot of stuff that doesn't make obvious engineering sense, even though it is the correct thing to do. That is why engineers are less important in NASCAR than a crew chief with 30 years under his belt.

--------------------------------------------------------------------------------------------

Ben said, "The best way to look at these problems is to consider the RMS load on the wheel over a time history rather than a single load at an instant in time."

I don't really see it that way. If you take a long enough time history, all setups have the same average wheel load. I make an effort to not have the big unloaded instant over a track irregularity. The reasoning is this. Once you break a tire free, the sliding friction is much less than the rolling friction. So, if we are trying to apply throttle out of a corner and get the tire unloaded, then it will start spinning. Once it starts spinning, it's real tough to get it to stop. I find I'm better off to cure that one particular bump rather than worry about the broader scope of the corner. This is analogous to removing material from an axle shaft to eliminate a stress riser. You make the ultimate strength of the piece lower, but by eliminating the trouble spot you are better off.

---------------------------------------------------------------------------------------------------

Some tracks respond better to more anti-roll bar and some tracks respond to less. There are no universal answers. In general, tracks that have a lot of single wheel bumps will want soft bars or no bars. In particular, I'd say tracks with chicanes and curbing that you have to go over the top of.

Some tracks are very bumpy, but since the bumps are 2-wheeled bumps, they respond well to stiff A/R bars. I tend to control excessive roll with the front ARB. Again, this probably goes back to my basic tuning approach. A stiff rear bar will invariably hurt power application. It makes the diff work harder and the rear tires work harder. It's really tough to make a car consistent over a long run with a stiff rear bar. If I were to make a general statement on bars, I'd say try to run the rear ARB as soft as possible or not at all if you have any low speed corners corners where power application is an issue. It depends on the size of the car, a sedan will need something just because it's so damn big. A single seater will give you more of an option.

-------------------------------------------------------------------------------------------------------

Nearly everyone wants to make a racecar super stiff. I don't know why. Stiff cars are not necessarily hard to drive, they just have no mechanical grip, and there are no amount of aerodynamics in the world that can make up for a bad mechanical setup. Even heavily aero cars like Champ Cars have there limits on the amount of spring that they will accept (although, admittedly, that level is pretty high). I find I'm much better off running on the low side of the window spring-wise and getting the driver to slow his hands down enough so that the car will accept it.

--------------------------------------------------------------------------------------------------------

Carroll Smith had probably the best quote about that heralded German marque when looking at a Lemans winning vintage 917, "All Porsche's are ****-boxes."

[Supercar]

You have many insightful comments, Fat Boy. Thank you. But I have to ask you for some clarifications.

You are saying that the 1.1-1.2 frequency ratio suggested by Ollie does not apply to racing because it is just to achieve a flat ride. Yet, all race engineers, suspension designers and aerodynamicists in particular, are concerned about cars being too pitch sensitive. Also in rally a car will not fly straight after a jump if the front/rear spring ratio is "just right". So isn't any deviation from the Ollie's rule of thumb really a compromise between flat ride and a corner-specific handling setup?

Also, you mentioned that you like to run both soft bars and soft springs in the rear. Soft bars for better traction and soft springs for better aerodynamics and turn-in. How are you achieving a balanced car then?

Ben said that he tunes for a smaller RMS wheel load. I agreed because I thought what he meant was a smaller standard deviation of the vertical wheel load. This is how oval racers in the US tune their shocks, springs and bars, I was told. That would go well with your observations about tracks with large single-wheel bumps. If there is a large bump under just one wheel, even if the rest of the track is relatively smooth, then the StdDev will be large on both wheels if the bar is stiff. It will be small if the bumps are under both wheels, even with a stiff bar and on a rough track.

I love this forum, guys. Glad to be part of it!

Philip

[Greg Locock]

No, the flat ride criterion is a subjective desire, to make passengers more comfortable. It did not arise from racing. It does not have much to do with pitch sensitivity, beyond the fact that they are both affected by spring rates and wheelbase.

Having said that I did use it on the solar car, because I wanted to minimise the drag caused by changes of angle of attack - so perhaps you could argue that your aerodynamics would be better served by a flat riding car.

One big problem is that you can only get a 'flat' ride at one speed. It's pretty easy to sort out what you need using a sprung bicycle model, ie 2DOF, with a linear damper. Run it over a bump, then stuff around with the rear spring rate until you get the minimum sum of pitch^2 over time. You can do that in Excel or MathCAD.

To be honest we don't bother with it on our faster production cars. They end up with huge front springs and (sometimes) huge front sta bars. Milliken knows why.

[Fat Boy]

Pitch sensitivity is generally just front ride height sensitivity. The aero balance on most racecars is hugely dependant on the front ride height. Anything with a flat bottom is particularly bad, but tunnel cars have the same issues. The trick to getting rid of the pitch sensitivity is to keep the front ride height operating in a window that doesn't get you into trouble.

So if I'm running stiff front springs / bars and soft rear springs / bars how do I balance the car out? Well, the short answer is that I don't. At least not in a steady state sense. My cars tend to have mid-corner understeer, but they are stable on corner entry, which allows a lot of entry speed, and they are good at putting the power to the ground, which is good for corner exit and straight line speed. The compromise is the mid-corner. Did you ever see when one of the car magazines tests a real racecar on their skidpad? It invariably has massive understeer. I think the magazine skid-pad numbers are forcing manufacturers to make cars that have high skidpad numbers, which means neutral-to-oversteer at steady state, then using stability control to make sure people don't kill themselves when they drive the cars in real life. Greg, am I right about this?

What do I do to reduce the mid-corner understeer? Well compared to with a car that has a different approach, I can usually run more front aero percentage. In theory, you should be able to run a front aero percentage equal to your front static weight percentage. In practice, you run somewhat lower. The corner entry phase of the corner generally determines how close you can get. If a car has soft front springs they might run the front aero percentage 5% below their front static percentage so the car doesn't oversteer on corner entry. Because I'm running stiffer front springs, I might be able to get within 2% of my static front weight. I can run 3% more front aero because my car is less pitch sensitive. This might not seem like a big deal, but it is.

Here's another thing. The classic driver thing to do with a car that oversteers on entry is a soft, early turn-in and then tighten the line at the apex. When the driver tightens the line at the apex, he will invariably pick up understeer. With a stable car at entry, the driver can make an aggressive, late turn-in and be starting to unwind the steering at the apex. This makes a massive difference. If you're able to take a larger radius around a corner, for a given grip level, you're going to go faster. Also, the guy that can get the steering wheel straight first is also going to be the guy that can get on the gas first.

Hope that clears up more questions on my approach. There are many different ways to skin this particular cat. This just happens to be the approach that has worked the best for me.

[Greg Locock]

" Did you ever see when one of the car magazines tests a real racecar on their skidpad? It invariably has massive understeer. I think the magazine skid-pad numbers are forcing manufacturers to make cars that have high skidpad numbers, which means neutral-to-oversteer at steady state, then using stability control to make sure people don't kill themselves when they drive the cars in real life. Greg, am I right about this? "

I'd have guessed that race cars have massive understeer (a word I'm trying to stop using) because they can't afford to have tank slappers - that is, they are generally in some danger of approaching the critical speed (ie yaw reversal) of the steering system.

The ONLY mags I know of that obsess about skid pan numbers are the American ones. The Corvette and the Mustang are locked into a death struggle of awful ride and ever wider tyres just to play up to those people. As John Miles once observed "That bloody G analyst was the worst thing we ever bought" at Lotus R&H.

So, we don't get directly involved in that silliness. For SUVs and trucks we are mostly worried about the fishhook test, and understeer (and electronics) is the safest route there. For our bread and butter car we have so many other requirements that it is hard to believe that limit state grip even gets a look-in. I don't get too involved with that - the tyre models are way beyond their ragged edges at that point.

[derstatic]

Someone mentioned 3rd springs in the topic. That are devices that have been confusing me for a while. What do they do? From what I've read they are some sort of "opposite" ARB but how do they work and what's the advantage of having one? What can you do with the cars handling by changing the 3rd spring?

[Supercar]

Originally posted by Fat Boy
Pitch sensitivity is generally just front ride height sensitivity. The aero balance on most racecars is hugely dependant on the front ride height. ... So if I'm running stiff front springs / bars and soft rear springs / bars how do I balance the car out? Well, the short answer is that I don't. ...

This is the sad part about the modern high-downforce race cars. This is why old-school engineers and some fans believe that downforce is the worst thing that happened to auto racing. The cars are now faster but there are no more spectacular power slides, and it is even hard to see any sideslip on the cars - they just drive fast and straight like on rails, do not roll and do not slide.

There is no excuse to making a car that understeers and plows all the time... unless that car has more downforce and is faster around the track. That is the current reality...

Philip

[Supercar]

Originally posted by derstatic
Someone mentioned 3rd springs in the topic. That are devices that have been confusing me for a while. What do they do? From what I've read they are some sort of "opposite" ARB but how do they work and what's the advantage of having one? What can you do with the cars handling by changing the 3rd spring?

Cars with high downforce must have stiff springs to support all that vertical load. But then the springs end up being so stiff that the car has no traction left. The different thing about the downforce is that it always points straight down and it is equal for both wheels on the same axle. Some smart people realized that you do not have to have stiff springs on each of the two wheels. You can have soft springs on each wheel, like you would want for better traction, and a stiff spring in the middle that would work just when both sides of the car want to compress or rebound.

With 3rd springs you can tune your ride height, aero ride height, pitch stiffness, pitch damping (3rd damper), vary roll stiffness independently of ride rate (similar to ARB)... You can probably also tune your 3rd damper to jack up over bumps and set a higher ride height for bumpy sections that way with no ill effect on handling or downforce in the smooth sections. This is all I came up with. Anyone has something to add or knows any secrets?

Philip

[derstatic]

Bringing this topic back up. I have found a new suspension question bothering my head.
I was talking to a friend yesterday and he had seen a tv-program about racecars and he said that when a racecar is turning i.e right, the left hand wheels move in bump. So far i'm all following, he also said that the ARB forces the righthand side of the car down, due to the linking of the left and right suspensions. I'm still following. So here comes my problem. Say this is done at the front of the car. You have a loaded left tire... a right tire that is pushed to the ground by the ARB and less roll that disturbs the aero on the front of the car and still all this cause understeer, and more understeer the stiffer the bar. Why is that?

Hope my question us understandable.

[Greg Locock]

The reason is (if I follow your question) that for a given slip angle the lateral force generated by a tyre does not increase in proportion to the vertical force on it. Typically it is proportional to the vertical load^0.7

Say the coefficient of friction at a given slip angle is 1.5, then with no load transfer the lateral force for the axle would be 1.5*W*(1^.7+1^.7)=3W.

Suppose we transfer 50% of the load from the inside tyre to the outside one.

The new lateral force would be 1.5*W*(.5^.7+1.5^.7)=2.91W, ie a 3% loss

So the car would steer wider, hence understeer. As you can see it is quite a small effect.

The bigger the a/r bar, the more load transfer you get for a given latacc

[Hoop-98]

I somewhat disagree with the previous post.lateral weight transfer is a function of track, CG height and lateral acceleration. Changing one roll bar would increase load on that outside tire while reducing the load on the other outside tire.

Nothing can change weight transfer, besides CG height and track. When we raise the rear roll center we are decreasing the lever arm in roll and increasing roll stiffness. A rear ASB is another way to increase rear roll stiffnes without increasing ride harshness and limiting travel.

Today attitude, pitch, dive, roll, control and it's effect on aero has taken precedence over mechanical grip in many race series, but really doesn't apply to a street car.

Springs and ASB 's are all part of the roll rate, while springs alone, manage travel.

jm2c

rh

[Supercar]

Originally posted by derstatic
Bringing this topic back up. I have found a new suspension question bothering my head.
I was talking to a friend yesterday and he had seen a tv-program about racecars and he said that when a racecar is turning i.e right, the left hand wheels move in bump. So far i'm all following, he also said that the ARB forces the righthand side of the car down, due to the linking of the left and right suspensions. I'm still following. So here comes my problem. Say this is done at the front of the car. You have a loaded left tire... a right tire that is pushed to the ground by the ARB and less roll that disturbs the aero on the front of the car and still all this cause understeer, and more understeer the stiffer the bar. Why is that?

Hope my question us understandable.

The answer has to do with the primary function of an ARB, which is to tune the car balance. When the front bar is made stiffer the front weight transfer increases. The front end loses its traction (like Greg said). But the total weight transfer remains the same because it is a sum of the front and the rear weight transfer. The rear weight transfer decreases and the rear end gains traction. The car understeers.

The aerodynamic effect is secondary. Just because the front ARB is now stiffer, and the car is now running flatter, the slightly increased total downforce (increased front and also increased rear!) will not increase the front traction much more than the rear traction. Within reasonable limits, the aerodynamic effect of such change is not likely to be large enough to overcome the weight transfer effect from the ARB.

Philip

[Hoop-98]

Philip,

Perhaps some of the pitch data from series would change your mind, of course if you have none, well, that would skew your vewpoint. So do you have some of the lola pitch, ride height data, to share?


rh

[Greg Locock]

I did not make clear that I was talking about the front axle, but then the original poster had already said that.

So, just to reiterate. I was talking about the front axle. I'm not quite sure what point you are making, the load transfer on that axle definitely takes place, and changing the sta bar will affect it.

However, I did forget to mention that the same will happen in reverse at the rear, so you'll expect another 3% (roughly) from that - bringing the total change in required front slip angle to 6% if you are in the linear range, which you shouldn't be.

[McGuire]

For some reason my views on chassis setup are often considered heretical around here, generating some teeth-baring and table-pounding at times. Just my $00,000.02:

Well-sorted race cars will just naturally come out with their natural frequencies higher in the front and nobody worries about it a great deal.

The antiroll bar is a tuning device in the mode of attenuation; that is to say, of negative capability. It subtracts total grip from the tire pair on the end of the car to which it is attached, by transferring load from the unladen to the laden tire in roll; the uneven distribution of load will reduce the total grip produced by the pair.

Here understeer is our friend. Understeer is a naturally stable condition, normally requiring only more wheel input to correct. Oversteer is an essentally unstable condition, requiring constant winding and unwinding of opposite input in order to keep the car pointed in an approximately proper direction. They don't call it 'loose" for nothing: with oversteer there are always too many things happening at once, most of them bad.

The other thing ARB's do is install wedge in the opposite end of the chassis. For example, front ARB will put wedge into the rear of the car in roll, loading the inside rear tire for better forward bite. On a rwd sedan with a beam axle and a big V8 it's pretty hard to get too much front roll resistance.

There is a wrinkled old rule of thumb that ARB rates should be somewhere around 10-20% of the total rate, for whatever that's worth... not so much. We typically want more roll resistance front than rear, while front/rear wheel rates and distribution can depend on a great deal of things, including aero, weight distribution, chassis design and basic intent. On road cars they are not really seeking overall balance or grip for the chassis but a portfolio of selected characteristics. When Aunt Millie stands up on the brake pedal with both feet at 70 mph on the freeway, we would like her to stay in her lane. So sedans traditionally have massive amounts of understeer even in a straight line.

So in my humble opinion: the question of ratio of spring rate vs. bar rate with respect to total roll resistance, wheel rate or frequency is not really one of grip but "balance." And unlike grip, balance cannot be calculated, quantified or data-fied. At the end of the day, balance is by definition the subjective call of the driver based on his feel and requirements. This is what makes "understeer," "oversteer," and "balance" such interesting, multi-dimensional words.

So two qualified, professional drivers can set up identical cars on the same track and arrive at remarkably different spring/bar (and damper) combinations, but nearly identical total grip and ultimate cornering force...and in the end, surprisingly similar overall balance, especially in the center of the corner. It happens every day. Where the two cars will differ is mainly in their transient responses: in other words, how the car communicates to the driver what it is about to do. Engineers don't like like this as they prefer to deal with pure science, but a major part of chassis setup is giving the driver the feel he wants.

[m9a3r5i7o2n]

Topic; Springs vs. Sway Bars (F1 and similar);

This is the most interesting writings about ARBs I have the privilege of reading and the first mention of the lifting of the inside wheel. Now I am beginning to wonder just when who and where and why the first ARBs were used or invented. What was his intention in putting one a car in the first place? Since it sounds as if the original intent was to be able to put softer springs on the car to make it ride better and yet maintain a higher cornering speed it was likely on a car such as a Cadillac or similar type vehicle.
Anyway very enjoyable and almost a dissertation! M.L. Anderson

[Earthling]

Amen to that!

[desmo]

And nice to have you back, Marion!

[McGuire]

Originally posted by m9a3r5i7o2n
Topic; Springs vs. Sway Bars (F1 and similar);

This is the most interesting writings about ARBs I have the privilege of reading and the first mention of the lifting of the inside wheel. Now I am beginning to wonder just when who and where and why the first ARBs were used or invented. What was his intention in putting one a car in the first place? Since it sounds as if the original intent was to be able to put softer springs on the car to make it ride better and yet maintain a higher cornering speed it was likely on a car such as a Cadillac or similar type vehicle.
Anyway very enjoyable and almost a dissertation! M.L. Anderson

You have that exactly right. The first modern-type (SLA) independent front suspension in general use (Cadillac, 1934) used antiroll bars front and rear. (At the time, called a "torsion rod roll stabilizer.") The development of IFS was not motivated by handling but the desperate need for a better ride, espeically in luxury automobiles. At that time the rear seat passengers were perched over the rear axle so pitch was a nightmare -- the driver was getting a much better ride than the passengers which would never do. Maurice Olley and crew saw that to get a "flat ride" the front frequencies would have to be lower than the rears, hence IFS, hence the antiroll bars.

At that same time, GM also introduced the Dubonnet-type ("knee-action") IFS for the low-priced Chevy and Pontiac...less said about it the better, perhaps.

...trying to avoid the argument as to "first" IFS, but that award may go to Sizare-Naudin back in nineteen ought-something. Nothing like modern wishbone IFS though: beam axle fixed to the frame, with long sliding kingpins suspended on a transverse leaf spring.

[Greg Locock]

Ah, Morgan's inspiration!

Was the Sizare-Naudin recognised as being a step forward at the time, or did the practical problems, combined maybe with a lack of understanding of what they wanted, prevent that?

Motto 1 Everything worth doing in cars had been tried by 1920. Usually by Lanchester.

[BRIAN GLOVER]

He was way ahead of Henry Ford and Cadillac with the first car to have interchangible parts.
I saw a Humber Super Snipe Station Wagon today. Never even knew they made one. That Humber dude was way ahead of his time also.

Originally posted by Greg Locock

Motto 1 Everything worth doing in cars had been tried by 1920. Usually by Lanchester. [/B]

[wegmann]

Originally posted by McGuire
At that same time, GM also introduced the Dubonnet-type ("knee-action") IFS for the low-priced Chevy and Pontiac...less said about it the better, perhaps.

Yup, '35 Chevy Master edition, now with knee-action suspension! I only know because that's the car my dad is rebuilding (which I mentioned in another recent thread). It has a TON of unsprung weight, but Greg really doesn't care that much

[wegmann]

Originally posted by McGuire
So two qualified, professional drivers can set up identical cars on the same track and arrive at remarkably different spring/bar (and damper) combinations, but nearly identical total grip and ultimate cornering force...and in the end, surprisingly similar overall balance, especially in the center of the corner. It happens every day. Where the two cars will differ is mainly in their transient responses: in other words, how the car communicates to the driver what it is about to do. Engineers don't like like this as they prefer to deal with pure science, but a major part of chassis setup is giving the driver the feel he wants.

Yup, and this helps form the answer to a question much higher in the thread: Why do racecars understeer so much around a skidpad? It's because they don't race around skidpads! Racecars are designed for the transient responses of braking, cornering, and accelerating out. Skidpads are nice for an engineer's steady-state analysis (assuming the tires aren't overheating too much), but not very useful for an analysis of a car's overall handling.

[Greg Locock]

"Greg really doesn't care that much" Well, I'm still waiting for data for the effect of unsprung mass on handling.

"Skidpads are nice for an engineer's steady-state analysis " Thing is for predicting limit handling reliably you need to be sure that you can predict steady state handling, so you get the constant radius stuff sorted, then you might be in with a vague chance of looking at the limit handling and getting that right.

It would be possible to leap right in and go for the limit state stuff, equally you could do so without knowing how much the car weighs, what its toe curve looked like, etc, but having to back all that stuff out from the limit state handling is going to be very long winded and time consuming and error prone.

After all, when you set the car up in the first place you do an awful lot of easier stuff before trying to get the perfect result at one corner.

There are three bits of data I usually have to get from the test itself - the actual shock absorber characteristic, the actual effective coefficient of friction modifiers of the tyre, and the true wheel travel characteristics in jounce and rebound (ie where and how those bumpers are working). You'd have thunk two of those would be a matter of record, wouldn't you?

[wegmann]

Originally posted by Greg Locock
"Greg really doesn't care that much" Well, I'm still waiting for data for the effect of unsprung mass on handling.

For the record I'm not in any position to argue (at least at the moment) ... but those knee-action units were a LOT of mass - just SEEMS wrong.

Originally posted by Greg LocockThere are three bits of data I usually have to get from the test itself - the actual shock absorber characteristic, the actual effective coefficient of friction modifiers of the tyre, and the true wheel travel characteristics in jounce and rebound (ie where and how those bumpers are working). You'd have thunk two of those would be a matter of record, wouldn't you? [/B]

Well we know that tyres are really complex, so I assume that's your odd-man out? But shock absorbers and their hysterisis aren't always that simple either.

[Greg Locock]

That's right. At least the test guys care about the shock characteristics, they just can't measure them properly. But can we ever agree on how bump stops are set up? no. Perhaps I should explain that we use the long foam cylinders, aka constant contact bump stops. So they come into play very early in jounce.

[m9a3r5i7o2n]

McGuire states;

Trying to avoid the argument as to "first" IFS, but that award may go to Sizare-Naudin back in nineteen ought-something. Nothing like modern wishbone IFS though: beam axle fixed to the frame, with long sliding kingpins suspended on a transverse leaf spring.
http://info.detnews....yrides/story/...
That is a lot older then I thought it would have been!

When I rode in the old 1930 Oakland I was surprised how well it rode as compared to my old 1933 Ford V-8 which in my opinion had a choppy ride, those old “shocks” just didn’t control the oscillation very well but the Oakland had single action Lovejoy hydraulics.

One of the odd things about the Oakland-Pontiac was the front springs that were not parallel to each other and not parallel to the main longitudinal axis of the cars by one degree and 34 minutes on a side.

M. L. Anderson

[Supercar]

Perhaps, the topic "Who invented the suspension first" should be continued on the Nostalgia Forum.

[McGuire]

Originally posted by m9a3r5i7o2n
One of the odd things about the Oakland-Pontiac was the front springs that were not parallel to each other and not parallel to the main longitudinal axis of the cars by one degree and 34 minutes on a side.

M. L. Anderson

You are right on topic sir, for that is done to tune roll resistance. The purpose of canting (and tilting fore/aft) parallel semi-elliptic leaf springs was to adjust roll under/oversteer among other things.

We tend to think of leaf springs as crude and primitive, but when you really get to looking at them they are very complicated indeed, as they serve as not only the springing medium but the locating linkage as well, while the actual geometries are not at all what they may first appear. Their damping modes are interesting too. For years the SAE produced a book called "The Leaf Spring Manual," which is well worth a look if one is interested in the old hardware, or the history of chassis development in general.

[Greg Locock]

We still use leaf springs

[Fat Boy]

Originally posted by McGuire
We tend to think of leaf springs as crude and primitive, but when you really get to looking at them they are very complicated

The same can be said of women from Siberia. That doesn't mean I want to know anything about them!

Leaf springs are heavy awful things that do a poor job of damn near anything performance related. If you are going to give them credit for anything, I suppose you'd have to say that they are good at carrying heavy loads. That's kinda like saying, "She don't sweat much...for a fat chick."

[McGuire]

Originally posted by Fat Boy


The same can be said of women from Siberia. That doesn't mean I want to know anything about them!

Leaf springs are heavy awful things that do a poor job of damn near anything performance related. If you are going to give them credit for anything, I suppose you'd have to say that they are good at carrying heavy loads. That's kinda like saying, "She don't sweat much...for a fat chick."

Hey, I like fat chicks. Or at least a little on the curvy side. Also pushrod motors, aircraft with propellers (but not as airliners I have to travel in) and leaf springs. True, we won't see F1 cars with beam axles and parallel semi-elliptics anytime soon (not to give Max any ideas)...still, that doesn't mean leaf springs aren't interesting. So interesting that GM had to send Maurice Olley out to invent the science of ride & handling. So we can at least give the leaf spring credit for that, hmm.