Handeling Thread

[MightyMike]

How a fellow seven owner expresses his passion for his car by making upgrades is of no concern to me. :grouphug:

[Guest Terry]

As has been said many times. The handling abilities of these wonderful machines far out guns the driving abilities of the majority of us, regardless of what our youthful minds tell us. What we do have, are toys that allow us to indulge ourselves in an elite world every time we strap ourselves into the silly seat. Whether you propel yourself along in the humble 1.6 or a fully blown twin cam, lets all do it with respect for the machine underneath us, and, the trouble it can get us into when we overstep our capabilities. The sheer joy of piloting my seven and the response of the general public is buzz enough for me. Nothing wrong with upgraditis, we all do some of that, lets all be mindful of where the road ends and the track begins.

Happy blatting guys...

 

YES:party:

[AndyB]

As has been said many times. The handling abilities of these wonderful machines far out guns the driving abilities of the majority of us, regardless of what our youthful minds tell us. What we do have, are toys that allow us to indulge ourselves in an elite world every time we strap ourselves into the silly seat. Whether you propel yourself along in the humble 1.6 or a fully blown twin cam, lets all do it with respect for the machine underneath us, and, the trouble it can get us into when we overstep our capabilities. The sheer joy of piloting my seven and the response of the general public is buzz enough for me. Nothing wrong with upgraditis, we all do some of that, lets all be mindful of where the road ends and the track begins.

Happy blatting guys...

 

Well said that man …. regardless I would not say no to a few more horses:driving:

[yellowss7]

My car is just about exactly as I assembled it almost 14 years ago. I have not YET, done any suspension mods, Power upgrades or anything other than safety mods. (roll cage being built as we speak). My car is street legal, and I enjoy Sunday afternoon Blats and I'm not out there burning up the corners. Just a nice drive with the wind thru my scalp:driving:

I've autoxed for 8 years before being introduced to the track. The track is the ONLY place you can actually get anywhere close to the limits of our cars Safely. I leave all my desire to speed and corner on the track.

 

And now, after about 4 years of tracktime, have just gotten to the point where I'm starting to use the potential of the car, and I'm still nowhere near wringing the last bit of grip out of it. I have as much fun out there as anyone else, with the exception of Croc and Blubarisax as they go blasting by me on the straights. (They no longer lap me though) I'm torn right now as the only place that I really need more power is on the straights. I try to keep telling myself that the skill is in the turns, but :banghead::banghead::banghead: I hate getting passed after working so hard to keep them off my butt.

 

Maybe another year of driving skill improvement before I reach for the wallet. We'll See.

Tom

[HOTTTCAR]

I'm torn right now as the only place that I really need more power is on the straights. I try to keep telling myself that the skill is in the turns, but :banghead::banghead::banghead: I hate getting passed after working so hard to keep them off my butt.

 

Tom

 

That is exactly what got me into this HP thing. Getting eaten by race prepared Vipers etc. after out cornering them. Weight to power ratio below 3 lbs/ hp. solves the problem. Controlling it at first is a bit challanging but as with flying, sailing, diving and even video games your brain begins to adjust to the challange and the initial violence of it soon begins to be very comfortable and controllable.

I will never reach Terrys skill level even if i were to work at it.

Age creates a real dead ass. And as Lauda says, with out a sensitive ass you can't be fast. :-)

 

Gale

[super7guy]

Found this site on types of diffs.

 

http://www.taylor-race.com/pdf/understanding_differentials.pdf

[AndyB]

Found this site on types of diffs.

 

http://www.taylor-race.com/pdf/understanding_differentials.pdf

 

That makes an interesting read, thanks.

[JohnK]

John,

I find it interesting to watch dirt track cars. Their inside front fender is miles above the tire, but they are flying thru the corner.

Keep it coming.

Thanks,

Carl

One of the best reference sources I've run into involving steering is Woodward Precision Power Steering, which sells to oval/dirt track folk. The catalog has descriptions in it on R&P setup and steering that are the best I've run into, including Carrol Smith's. It just takes a bit of thinking to translate it to sportscars from what people do on an oval track.

There's a really great couple of paragraphs explaining why really quick steering is so valuable. When I got to the point where I understood what he was saying I wound up not only with quick steering (1 1/3 turns lock to lock) but power (hydraulic) assisted too. Had anyone told me that steering that quick was useable in anything other than some sort of extreme race car I wouldn't have believed it. But after a couple of thousand miles out on some really twisty country roads whose surfaces are less than ideal it's a revalation what such a setup allows you to do. But you do have to have all the pieces working right (lousy tires made the car a hassle until I changed to relatively ordinary sporty tires).

But, yes - you can learn a lot from dirt track folk. There's a lot of experience and knowledge in that realm.

Edited May 25, 2015 by JohnK

[JohnK]

Staying with https://www.dropbox.com/s/6cesllwg4xmm6qq/Original_Ride0Roll0.jpg?dl=0

 

--Being a curmudgeon at heart I'll gleefully introduce a topic that's really important, that reveals something really important to understanding what the suspension is doing, and is just really involved. (Note that this figures heavily in the fix I wound up making to fix the behavior of my rear suspension – so it's not just academic.) On the WinGeo3 diagram, at or near the vertical centerline, are four boxes: one is a bit above axle height, two are nearly on top of each other about ground-level inside a containing box, and the third is off to the right a bit with a diagonal cross in it. The highest one is the CG, the vertical value of which is a guess on my part because I've never been able to successfully measure it although I've gone through the procedure that's supposed to deliver it. In addition, I've never even thought about measuring the CG's lateral location because I've never been in a situation where I needed to see the the consequences of adjusting its lateral location. (Ignorance is bliss?) Unfortunately, what I do know is that there is NOTHING that improves a car's handling more than lowering its CG and I can't find a doable way to measure it so I can fit it into the spec and obtain an accurate picture of what's going in. Still, as I build things, I always work to put things as low as I possibly can on the chassis. (The battery placement I did is a fabrication that's pleasing – see build pages)

-- With respect to these other 3 boxes - Now, there is a measure that locates on the rear suspension that point at which the car rolls about as it is cornering. You can feel (or develop the ability to feel) this point with the small of your back or your butt as you're driving 'round a corner, and it lets you know what the rear end of the car is doing. This is called the roll center, and we're pretending that it's as simple as the single point identified in this diagram (shown as Roll Cen. 7.252, 0.339, meaning that its 7.252” right of the long axis and 0.339” above ground with all the suspension components at their shown locations). SAE defines it as “The point in the transverse vertical plane through any pair of wheel centers at which lateral forces may be applied to the sprung mass without producing suspension roll. “ And this is why you spend hours re-reading and re-imagining things and re-running your suspension mode – so you can get a sense of what this means and how it affects your car. One facet of this is “Push on this point an none of the members of the suspension will move.”

--Now depending on the design of the suspension, the roll center may move. If it moves only a little, your butt may not notice anything's out of the ordinary, i. e., you know right where the rear end of the car is as it's in motion going around the corner. If the roll center moves a lot, your butt will likely inform you that whatever the rear end of the car is up to, it has just dropped out of the picture and you're on your own. Relevant, huh?

From the suspension gods: “A high roll center usually means more jacking force, which tends to raise the chassis, but a high roll center minimizes body roll. A low roll center, which can even be below ground level, minimizes the jacking effect but creates more body roll.” Jacking force is the overturning force I spoke of earlier. Look at the distance between the CG and the roll center and think what's happening there and you can understand the above – that distance is a moment arm (a lever) – the longer it is, the greater leverage the car (sprung weight) has to turn the chassis through the roll center. Thus if you could make the CG and roll center identical, nothing would move as the car cornered. Along with this, the roll center determines cross-ways weight transfer – a low roll center results in less weight transfer from the inside to the outside wheel. When both wheels have similar loads, they provide more traction than if the weight is moved from the inside to the outside wheel and the outside wheel is heavily pressed to the road. Smith does a nice analysis of this (Listening HOTTCAR?.) Nett: low roll centers give you better traction by keeping both wheels pressed on the ground.

--Staniforth writes that years of poking about at the pits and analyzing photos and picking brains has shown him that the wizards design their race cars with roll centers that stay where they're put within a couple of ten-thousandths over their range of suspension movements. (I'll guess that the guy who designed my suspension thinks 'roll centers' are something to do with a Bugler Cigarette rolling machines.)

--The reason there are these different boxes on the diagram having to do with the roll center is that there are different ways of deriving where the roll center is. The conventional value (Kinematic roll center) is gotten by intersecting lines on either side of the car, looking at a figure drawn using the Instantaneous Swing Axle and the tire contact patch The Force-Based Roll Centers (Force Application Point, F.A.P values) are the product of a second means of satisfying the SAE definition, and more realistically represent the SAE description of the roll center. Mitchell spends quite a bit of time explaining how F.A.Ps are derived and I'll compare the two if anyone's interested. His WinGeo3 even includes a separate application just to illustrate step by step how the loads are transferred through the suspension. So, this is a really significant result of how your suspension is deisigned.

--Consider that the take home here is, for the price of measuring your car very accurately, you can enter the data and SEE the things that determine what the car's going to do as it's driven- without having to take anything apart or buy new parts.

 

Next, what happens when things move.

Edited December 28, 2016 by JohnK

[JohnK]

One thing I should add here. I spent a lot of time fiddling with, reading the manuals of, reading the on-line documentation of and in general doing a whole lot of cuss'n at WinGeo3. BUT, this IS a very well developed program and allows you to do all sorts of very sophisticated analyses and see what the car's doing from many many different perspectives. It's used by more than a few big guns out there. Just don't expect it to take you by the hand and lead you through everything you need to know. Which is another way of saying that understanding suspension is one complicated issue and don't expect that answers are easy to come by.

-

All the terms that Mitchell uses are SAE, or have SAE counterparts, so they have very explicit and clear definitions. He provides definitions and explanations with his Win Geo 3 software, some in hardcopy, lots in the application itself.

-

Also, if you have a car whose handling you're really happy with and don't fancy all the work req'd to measure the car, but are interested in learning about how it works in this regard, you may find it worthwhile getting a copy of the pgm and exploring what it does. I think there are some light-weight versions and version that focus on specific areas of a car's suspension. While I haven't looked, it may be possible to find the measurements for your car on the Web, which would make the pgm more than academically useful. WindGeo3 can reveal lots of different things that are neither obvious nor intuitive, and the addition of Rowley's application allow you do do things like a WinGeo3 specification and determine what will happen to the suspension when you put that spec in a corner with, say, 1.2 G lateral force and +0.3G acceleration. There's an interface to WinGeo3 that allows for recording displacements as the car is driven, so you can watch what all the bits are doing as the car goes around a track, even multiple laps! This allows you to analyze all the curves generated by the different suspension members so you can optimize their adjustment into the 'sweet spots' where they travel through their most linear ranges. (I need another dozen years or so added to my lifetime – and a lottery win or two).

-

All of which is by way of preparing for what's coming next...

-

Definition: Virtual Swing Axle. This term is used to understand Camber change and to calculate other things, like roll center. Imagine you've got a wheel standing upright on the floor and it's attached rigidly to an axle, and the free end of the axle is held roughly horizontal and attached to a pivot. If you step over to the pivot end and lift the pivot location up and down a few inches, kind of like it would be were it attached to a differential in a car and the car was rising and falling on its suspension, you'd see the top of the wheel move a few degrees outward or inward. Now if you did exactly the same thing but used an axle that was really long, you'd see that the top of the wheel moved less in response to the the axle pivot's up/down movement. You can think of this wheel's movement when the pivot moves up and down as camber change, and understand that the amount of camber change is going to be less when the axle is long and more when the axle is short. And of course small changes in Camber during changes in Ride are a very good thing.

-

Looking at the rear suspension in its resting position -

https://www.dropbox.com/s/6cesllwg4xmm6qq/Original_Ride0Roll0.jpg?dl=0

----- Consider the left side, take the upper and lower suspension links (brown) and project them out laterally until the top and bottom projections intersect, and the point of intersection is called the Instant Center. Measure from the Instant Center back to the the upright's upper and lower ball joints and that length is the Virtual Swing Axle or Instant Center. With regard to Camber change, at that Instant Center length the suspension acts just like an actual axle would act if it were that long, as per the above paragraph. If the upper and lower suspension links are perfectly parallel, the Instant Center would be infinitely far away and the Camber change would be zero. In most cars the links are not parallel so there is Camber change, and the Virtual Swing Axle 's length changes when the upright (pivot) moves up or down – which means that your car's Camber changes not only with upright movement, but with position as well as the wheel (suspension) moves and the Instant Center changes.

-

The measurements entered into WinGeo3 show that the links are nearly the same length and are nearly parallel. This gives the suspension really long Virtual Swing Axles (lns. Cen. In the graphic) on the order of 5,000 inches – so this tells you that when the wheel moves through its 5” travel in Ride, the Camber change at this distance from the pivot of the axle, is really small. This is a property of all suspensions that have links that are roughly parallel and roughly equal. For the results this design produces when it works, look at

https://www.dropbox.com/s/zh5tmcmej4cumi7/Original_Ride-2Roll0_0_4.jpg?dl=0

which shows 2” of bump and

https://www.dropbox.com/s/jykmynkd6tsbvwq/Original_Ride%2B2Roll0.jpg?dl=0

which shows 2” of droop. Thus you can observe that, with this design, there are very small changes in Camber associated with these changes in Ride height. However, you may also notice that the Instant Centers change value a lot, but they're all relatively large so there's little effect on Camber change.

N. B. That none of the numbers/values come out even here because the data that the program performs all its calculations with is from the physical measurements taken from the car, which is not perfectly true and by a human who is neither accurate nor precise at times. With practice and patience you can make pretty consistent (precise) measurements, and by using good quality laser builders' tools like plumb bobs and levels which, in combination with a collection of 1/8” wall 2 x 2's, you can develop a methodology and some technique that'll yield pretty good data. Also, using the program, you learn which measurements have to be dead on and which can give useable results even if off a bit.

But then again, it never hurts to be obscenely wealthy in this department. https://www.dropbox.com/s/0xmzscr9q7bt62f/phoca_thumb_l_phoenix.jpg?dl=0

Besides providing a true platform on which to set up the car and adjust corner weights, notice the lengths of the links: since they're really long and nearly parallel, they provide very long virtual swing arms which gives very small changes in Camber in Ride, but note the difference in the positions of the links' pickups on the chassis – this is one way to solve a problem that I'll identify that my configuration has In the next several JPGs. (Can you find the ARB in this picture? WinGeo3 has a template for modeling this particular front suspension configuration. The Rack & Pinion (red and blue) is the same as what I designed into my car, 'cept mine is power assisted. N. B also how the engine coolant heat exchanger is positioned, catching dirty (turbulent) air following the front suspension links – Colin Chapman was the first to recognize the value of placing this exchanger in this location – doing so not only allowed the car to have a smaller frontal area (better aero / less wind resistance and better penetration than when there's a 'barn door' of a radiator at the front of the car). His design also moved weight toward the middle of the car lowering the polar moment of inertia of the car allowing the car to turn more quickly)

-

THIS is why Newton invented Physics!

Edited December 28, 2016 by JohnK
Changes mandated by DropBox

[Guest Terry]

John. Thanks for your HUGE contribution to this post. Great stuff.

[JohnK]

Now when I got the car on the road I thought I had it all covered. The program let me figure out what spring rates to at least start with, and what rates of front and rear ARBs would give me balanced front/rear weight transfer when I put the car into a corner that generated 1.0G lateral force and a bit less than 2 degrees chassis roll. Koni Racing (after I supplied them with a bunch of measurements that defined the setup that the shocks would be used in) had delivered a set of shocks that were dialed in across the board for the car I described. I was ready to rock and roll (after a few minor adjustments).

At the end of the summer (and having made those “few minor adjustments”), I found myself with a car that was really great in many respects, but whose rear end really wanted to come around on me when I pushed it. I had spend a lot of time changing springs and ARB bars and rates and the car was felt comfortable and well balanced when tossed into a corner – but the rear persisted in its goal of trying to be the first one out of the corner. Back to the books that I had thought at the time I understood perfectly, and to the program that showed me what was happening (IF I asked it the right questions.) Re-reading stuff I found that Smith illustrates and describes what I was going through clearly, but whoever you read says essentially the same thing: parallel links of equal or unequal length links don't work in roll – if you want to keep your tires perpendicular to the road surface when the chassis rolls. Regardless of what you do with your springs, shocks and ARBs, the harder your turn the worse the oversteer gets.

 

Dragging out the program, after the euphoria of thinking it was going to deliver wonderful Camber behavior everywhere because I could see that it'd do in in bump, we get the proverbial pie in the face when we ask the question, “And by the way, what happens when the chassis goes into ROLL? . . .”

 

Asking the right question, the program shows us, and allows up to measure, how much the tops of the rear wheels move toward the outside of the corner when the chassis rolls. It turns out the degrees of traction-destroying positive camber created when the chassis rolls are nearly the same as number of degrees that the chassis is rolled. The more the roll the greater the camber change, and of course the change is in the wrong direction if you want to generate traction (a degree or a degree and a half of negative camber change would improve things, especially with race tires.)

 

This is shown in the diagram

https://www.dropbox.com/s/6cesllwg4xmm6qq/Original_Ride0Roll0.jpg?dl=0 which gives the values calculated when there are two degrees of roll in the positive direction ( generated by turning left). This is shown by looking at several values and remembering that you're looking at the car from th rear (unless you've changed it from the default view to any one of several): at the center, “Roll” has an arrow next to the value of 2.00 pointing to the right which indicates the direction of the force, to the right the “Travel” value is negative and shows an arrow pointing downwards showing that the chassis has moved downwards, further to the right the third value down in the column of parameters on the right of the wheels shows that the Spring has been compressed (two arrows with their heads facing toward each other next to a negative value showing the spring compression in inches (changeable default setting), and below that is Load which is positive and greater than Load on the left side of the diagram (it is negative indicating load removed > from the at-rest value. You can see the degrees of roll illustrated by vertical lines anchored in the center of the tire footprint next to true verticals at each wheel.

---The reason I'm going through all of this is that it's easy to get confused as to which direction things are happening. Mitchell's convention is that the positive roll we see here comes from making a left turn which is indicated by negative steer. Keeping oriented is a continual hassle due to the nature of the problem.

-

Looking at

https://www.dropbox.com/s/av1bgulujv9ext0/Original_Ride0Roll-2.jpg?dl=0

shows you the results of 3 degrees of chassis roll,

and the next two diagrams show you the results of turns to the right generating 2 and 3 degrees of chassis roll.

https://www.dropbox.com/s/av1bgulujv9ext0/Original_Ride0Roll-2.jpg?dl=0

https://www.dropbox.com/s/2398kdw4sps2si2/Original_Ride0Roll-3.jpg?dl=0

 

So we ( I ) learn that this is a show stopper. The more the chassis goes into roll, the more the outside tire goes into positive camber. The only way to stop it is to dial up the ARB at the rear so the chassis doesn't roll which is stupid for more reasons that one needs to write about.

-

Back to the books and several readings of Smith and others brings us back to the punch line I quoted earlier (and apparently didn't think applied to ME):

 

From Carroll Smith's "Tune to Win". page 54: “ ...While it is possible to control wheel camber either during vertical movement or during chassis roll, it is not possible to achieve very good camber control under the combined conditions --- we have and "either --or" situation. ... “

 

So now what? Fortunately, although at the cost of reading Smith et al. even more closely and having my hair begin to hurt as a result of the mental effort, the answer is there, in the few pages leading to Smith's quote above - and you have to really understand to what the quote is telling:

-you you can't have it all,

-you can only achieve a compromise,

AND

-there's no yardstick you can use to measure what a good compromise is.

-

---What I KNOW is that 'as is' the car has really wonderful camber control in bump and just plain awful camber control in roll, and as verified by driving the car, THAT just does not work. So thinking to myself that the car's ALWAYS gonna roll, and that bump is a transient condition (and paying Madam Zelda to recite the proper incantation while she sacrifices a live rooster at the dark of the moon over the the car's differential, just for good measure), I think to myself “how about we focus on good camber control in roll and then see what happens to bump as a result?” (Hmmm. Is it better to be shot or hung?)

-

Getting to work, there were two phases to this: the first was to use the program to test different configuration of the links and find one that seemed workable, and then figure out how to actually BUILD that configuration (you may have noticed that there's not a lot of free space In a Se7en that you have where you can put extra stuff). Smith led me through the sequence of relationships that described the fix (although getting to the point where I understood it still made my hair hurt). In essence, you shorten the top links and then make it so that the shortened top links (I. e., the links that form the upper A arm) are placed higher than before at the ends where they attached to their upright – you want the top links to run downhill toward the center of the car. Figuring out how much to shorten and how much to raise took a lot of trials with the program and then assessing what each compromise was going to result in – the operative word here being “compromise”. And then I had to go back and forth between what the program showed as good, and the car said was possible, making prototypes (out of all the sorts of things one has lying around) to test that there was room for the design (these parts have to be able to move through a area in the car and not run into anything - duh!) and on top of that, verify that what I was doing with the program actually reflected reality (it is really easy to make mistakes here).

-

I was able to come up with something that looks hopeful. I found a configuration where camber change is pretty flat, and also reduces the movement of the roll center (recall this is important for giving feedback to the driver, and is generally is associated with good suspension behavior.) There's quite a bit of Camber change in bump, but it's in the negative direction, which is less bad than positive change, and the amount of movement of the wheel in bump is usually not that much, particularly on a track - while chassis roll of 2 degrees is quite a bit. Compromise.

 

Next time I'll post the graphs that show the values this configuration delivers. The parts are fabricated and I've installed them temporarily (need to clean and paint) on either side and aligned the rear and verified its movement and behavior (not chassis roll, however. Think about that. What's happening with roll is very different from bump, even though it would seem to be the same – I mean, in both circumstances, the wheel's going up and down, right?) Anyway, physically, it works without anything running into anything else, the wheels go where they're supposed to, and the hardware and software agree with each other.

 

 

A reader alert here. At some point I'm going to have to post pictures of the “deliverables”. So far my fabrications have been sound (welding just keeps getting better and better) and robust, in that nothing I've designed has proved too weak and failed. But just the same, to those out there with either refined sensibilities in matters of mechanical/automotive construction, or weak hearts that might be overstressed by a bout of hysterical laughter, I'll offer a heads-up to you that viewing pictures of my solution may be offensive or hazardous. And to those with perhaps more insight than mine who can see that this solution can only fail catastrophically, you'll be able to say “I KNEW that couldn't possibly work!” should the results of this project produce a picture in the media showing smoking remains.

 

Next, assessment of the results.

Edited December 28, 2016 by JohnK
Changes mandated by DropBox

[yellowss7]

John, after reading that post, I'm damn glad I bought a Caterham and don't have to play around like that. :rofl:

 

Tom

[Guest Terry]

John are you running a watts Linkage? I witnessed a Storker at the Track I work at, the thing was all over the place, yes it was quick in a straight line. You are quite correct Tom about the Caterham compared with some of the other kits.

[Guest Terry]

Why the hostility?

[Kitcat]

Really! Here's a straight forward thread, with a legitimate discussion of handling issues and suddenly this personal attack?

[Guest Terry]

Really! Here's a straight forward thread, with a legitimate discussion of handling issues and suddenly this personal attack?

 

I guess there's always going to be people that get offended and respond in a defensive and hostile manner. Sad really!

[Guest Terry]

John

 

Have you had your car corner weighed?

[Guest Terry]

Last year we had a guy at one of my schools with a late model Birkin. He had put the car together himself, (made a beautiful job of it) and fancied trying it at the Track. After a few laps in the 1st session he asked me to drive it and give him some feed back. The car understeared one minute and then over steered the next, the car felt nervous and down right dangerous. With no access that day to a corner-weight set up it was always going to be limited what we could achieve.

 

We started by looking at all the tire pressures and reset them to 16F - 18 R, I then softened up the dampers F/R to the softest setting. Drove the car, way better, but not right. We then disconnected the front sway bay and increased the front tire pressure by 2lb. The car was now at least drivable and all the nervousness had gone, how long did it take? 30 minutes. The moral of the story is you don’t have to spend a fortune and you can work with what you have if you understand some of the basic science that revolves around suspension.

 

John is doing a great job of informing us all of what’s possible if we are prepared to put the work in.

 

Keep it up John :hurray: I for one am learning a lot.

[xcarguy]

Terry,

 

You talk a good line for a guy who I have never seen around a seven of any kind of his own at an autocross or track. I also like the way, any chance you get you slag anything that is not a Lotus or Caterham.

 

I will leave the owner of the seven you wrote about to respond himself if he so desires.

 

Just my 2 cents worth.

 

Really! Here's a straight forward thread, with a legitimate discussion of handling issues and suddenly this personal attack?

 

I guess there's always going to be people that get offended and respond in a defensive and hostile manner. Sad really!

 

Okay, let's all grab a Snickers. :boxing: