Toe Changes for Track Use

[JMortensen]

Forgot about that thread. And I still stand by my basic conclusion: The ROI isn't there for Ackerman on a production based race car.

For 99% of the people, I agree; just run the static toe out, and it accomplishes basically the same thing. For the super anal though. I think there is an argument for Ackerman in there. In my case, having cut and hacked just about every other part of the car, modding for Ackerman wasn't that big a deal.

[Dan Baldwin]

FWIW...

 

 

Maximum inside and outside front wheel turning angles for a 2010 Corvette GS: 30.6* inside, 30.3* outside

0.3* total difference at full lock, .01* difference per degree of road wheel steering angle

 

Maximum inside and outside front wheel turning angles for a 2010 BMW M3: 36.4* inside, 27.9* outside

8.5* total difference at full lock, 0.26* difference per degree of road wheel steering angle

 

So which approach is "right" and which is "wrong" between the two? IMO, neither, it's gotta be more of a feel thing than any meaningful difference in performance potential.

 

I wholeheartedly agree with the philosophy of minimizing scrub and going with minimal front toe. I run zero front toe and minimal rear toe (~.1* per side) on the Z and the S2000.

Edited January 3, 2011 by Dan Baldwin

[JMortensen]

I don't know that you can make a statement like "it's just for feel" very accurately with this chart as your backup. To me it's a bit like saying: "Some racers run softer springs in front to reduce understeer, others use a stiffer rear bar. So which approach is "right" and which is "wrong" between the two? IMO, neither, it's gotta be more of a feel thing than any meaningful difference in performance potential."

 

There is undoubtedly more than one way to skin a cat. I think there is a distinct difference between running toe out or Ackerman as there is with the softer spring vs ARB example. Maybe not as pronounced as that, and as I said before I think that most people can run some toe out and call it a day. On the other hand, you yourself seem to suggest that static toe settings matter from a drag perspective, and I think the advantage Ackerman has over toe out is that it minimizes the static toe setting, then becomes more active the tighter the turns get. Probably the most useful thing one can take away from your chart is that in this limited sample 6 of 7 of these performance cars use Ackerman to some measurable degree or another. It seems doubtful that they would do so if it tended towards understeer. Why does the Vette not use it is the next logical question. I don't have an answer for that other than maybe they decide to use another method to skin the cat, but I think it is interesting to look at the top speeds of the cars in question. The Vette is the only one without a speed limiter and should do 190, the Lexus has a limited speed of 170 and has the next lowest amount of Ackerman. The rest of the cars are limited to 155 or lower. This also is not "proof" of anything in and of itself, but it does conveniently fit with the idea that for more stability (understeer is more stable) and turns with larger radii like the ones you might encounter at very high speeds, less Ackerman makes sense.

[Dan Baldwin]

I don't know that you can make a statement like "it's just for feel" very accurately with this chart as your backup. To me it's a bit like saying: "Some racers run softer springs in front to reduce understeer, others use a stiffer rear bar. So which approach is "right" and which is "wrong" between the two? IMO, neither, it's gotta be more of a feel thing than any meaningful difference in performance potential."

 

Not sure exactly what you're trying to say here, my only point is that:

There is undoubtedly more than one way to skin a cat.

Indeed. I'm saying that I'm not losing sleep over whether my car has too much ackerman or not enough, I set it up to respond like I want it to with minimal toe and don't worry about modding the ackerman.

 

For sure the chart you posted on the previous page would seem to indicate that you do want a bit more slip angle for the inside tire, but the inside tire is pretty insensitive in the region of maximum lateral grip.

 

Probably the most useful thing one can take away from your chart is that in this limited sample 6 of 7 of these performance cars use Ackerman to some measurable degree or another. It seems doubtful that they would do so if it tended towards understeer.

I'm sure that none of them chose an Ackerman setting independent of other suspension settings. I.e., if more ackerman *did* tend toward understeer, that would be addressed with spring rates and/or antiroll bar rates and/or etc. etc.

 

Why does the Vette not use it is the next logical question. I don't have an answer for that other than maybe they decide to use another method to skin the cat, but I think it is interesting to look at the top speeds of the cars in question. The Vette is the only one without a speed limiter and should do 190, the Lexus has a limited speed of 170 and has the next lowest amount of Ackerman. The rest of the cars are limited to 155 or lower. This also is not "proof" of anything in and of itself, but it does conveniently fit with the idea that for more stability (understeer is more stable) and turns with larger radii like the ones you might encounter at very high speeds, less Ackerman makes sense.

 

I don't have any idea why the Corvette doesn't have any Ackerman, but I seriously doubt that is costing it much if anything in terms of outright cornering capability. That was my point.

[johnc]

Ackerman is a perfect example of something that's of some relevance to low powered Formula cars (Formula Vee, Formula Ford) and Sports Racers (Spec Racer Ford, S2K) where momentum and low rolling resistance is important. Some of the low powered Production racers (HP, GP) also spend some time optimizing steering angle on their cars.

 

But its unfortunate that amateur IT and GT level road racers discovered the term in the racing books. Its taken on a life on Internet wholly out of proportion to its value to a a production sedan based racer. Pages and pages are devoted to the trivial minutia and optimizing steering angle to minutes of arc based on the turn radius, steering angle, and estimated tire slip angle.

 

This thread has basically become what I call an Internet Engineering Masturbation exercise. You guys have lost all perspective.

[JMortensen]

Found the article by Zapletal. It is well worth the read. The figures discussed in the article are all the way at the end, so I suggest printing the 23 pages so that you can reference them as you read it.

 

Going to pick up some graph paper and a compass tonight and see just what effect moving my rack back had.

 

http://forums.hybrid...846#entry914846

Edited January 6, 2011 by JMortensen

[Dan Baldwin]

This thread has basically become what I call an Internet Engineering Masturbation exercise. You guys have lost all perspective.

 

I agree that it's not anything to worry about, that was my point exactly.

 

But I'm usually up for some Internet Engineering Masturbation

[JMortensen]

So I'm guessing I may be the only person here who has tried the same car before and after Ackerman mods (friend's 510 where it really made a big difference). Why do I say that? Well, I just did the graph paper test, and found out some interesting things:

 

The stock Z has very slight anti-Ackerman, moving the rack back where I have it gives perfect parallel steer on the graph paper all the way through the rack motion.

Angling the steer knuckle out 10 degrees which is about 3/4" (I don't think there is room even if you wanted to) gives the following results:

 

Rack movement vs steering angle outer and inner wheel:

 

.5" = 7 degrees

 

1" = 13.5 , 14

 

1.5" = 20 , 22.333

 

I did this full scale, so I rand out of room here, but 1.5" of rack movement is almost a whole turn of the wheel, so going farther than that is kinda pointless, but you can see that the Ackermann effect is minimal.

 

Lengthening the steer knuckle to 5" (about 5/8" longer than stock) gives the following:

 

.5" = 6

 

1" = 12

 

1.5" = 17

 

2" = 23 outside, 24 inside

 

2.5" = 29, 31

 

Lengthening the steer knuckle to 5" and bending the steer knuckle 10 degrees:

 

.5" = 6.25 degrees

 

1" = 11.5, 12.5

 

1.5" 17, 19

 

2" = 23, 26

 

2.5" = 29, 33.75

 

 

So the question is now what? Do I give up and let it go, or try to figure out how to make it better, I think that paper I uploaded really shows why it might be worthwhile to work on it. Couple of thoughts are to make a new crossmember, or try the Z bar steering, although that looks like you need a rear mount rack mounted up front and it would involve a lot of fabrication. Maybe a shorter steer knuckle could be angled out farther. If the knuckle was really short this might work I think, and then just rely on the ps and get a slower rack ratio to make it all work. Rear steer would allow a lot of options, but would no doubt have a lot of interference with the TC rods I worked so hard on. Not sure I would want to try steering off of the strut housing up higher, but that would also be an option.

 

Dammit! Well, at least I figured it out before I ordered the new rack...

Edited January 6, 2011 by JMortensen

[Dan Baldwin]

Found the article by Zapletal. It is well worth the read. The figures discussed in the article are all the way at the end, so I suggest printing the 23 pages so that you can reference them as you read it.

 

Going to pick up some graph paper and a compass tonight and see just what effect moving my rack back had.

 

http://forums.hybrid...846#entry914846

 

Look at Figure 1 in your link which shows lateral force vs. slip angle for different normal force or load on a tire. This would argue for anti-ackerman.

 

Now look at the plot of lateral force vs. slip angle for different tire loads you posted previously in this thread. That one argues for pro-ackerman.

 

Who ya gonna trust?!

 

I think you're knocking yourself out for very very little, EVEN IF you did testing to figure out what the grip vs. slip angle vs. load relationships are for the exact tire you're going to be using under the exact conditions and range of loads you'll be using them under.

 

Without knowing what those relationships are for your specific tire and load range (and additionally what the *actual* range of steering angles is for your those tires under the conditions you're trying to optimize for), there's no way for you to know what to aim for.

[JMortensen]

Look at Figure 1 in your link which shows lateral force vs. slip angle for different normal force or load on a tire. This would argue for anti-ackerman.

 

Now look at the plot of lateral force vs. slip angle for different tire loads you posted previously in this thread. That one argues for pro-ackerman.

 

Who ya gonna trust?!

 

I think you're knocking yourself out for very very little, EVEN IF you did testing to figure out what the grip vs. slip angle vs. load relationships are for the exact tire you're going to be using under the exact conditions and range of loads you'll be using them under.

 

Without knowing what those relationships are for your specific tire and load range (and additionally what the *actual* range of steering angles is for your those tires under the conditions you're trying to optimize for), there's no way for you to know what to aim for.

I'm going to trust that pointing the tire the right direction to go around the corner is more important. Why? Because I've personally felt the change it made.

 

You're doing the same thing Leon is doing, focusing on the peak grip. That's fine if all you're worried about are large diameter, steady state corners. For an autocross car where the steer angles are sharp and steering wheel movements fast, I think the dynamic toe in and the negative effect it has is a much much much bigger deal. I think it also might be a pretty big deal on the road courses, and I think that can be proven out by looking at the clips I pointed out before and how many times the wheel passes 90 degrees. I did look and found a spec on steering ratio for the Z. 17.8:1, so that means at 90 degrees you've got ~5 degrees on the outside wheel. I came to realize that I was quoting the Ackerman specs in CCS, but that was for the particular racecar they were modeling, and a different wheelbase car might need more or less than that figure, but I think we can say that there probably should be "some" and there is none. I'm going to try and get to the second diagram exercise today.

[Dan Baldwin]

I'm going to trust that pointing the tire the right direction to go around the corner is more important.

The problem is that you don't have a basis to say what direction is the "right" direction.

 

Why? Because I've personally felt the change it made.

You *FELT* the change it made? For one thing, it is possible to *feel* a change that doesn't exist (placebo effect). For another, what *feels* better isn't always faster.

 

You're doing the same thing Leon is doing, focusing on the peak grip. That's fine if all you're worried about are large diameter, steady state corners.

I would think you might want to maximize lateral grip for smaller-diameter lower-speed corners as well!

 

For an autocross car where the steer angles are sharp and steering wheel movements fast, I think the dynamic toe in and the negative effect it has

You are *assuming* that it has a significant negative effect. The engineers who are responsible for the Corvette's steering geometry might beg to differ...

 

 

OK, I just did the geometry, and for a 30ft radius corner (that's pretty tight!), assuming zero toe at the rears, outside tires at 5deg slip angle (assuming for a moment that that is optimal for the loaded tires), and with zero ackerman (parallel steer) and zero static front toe-in, I'm only seeing 3.5degrees of slip angle at the inside front. That would be less than optimal even assuming the chart that suggests anti-ackerman.

11.6 degree steering angle, btw, or 206 degrees at the steering wheel.

 

So, I will concede that *some* ackerman *should* be better for maximum lateral grip.

 

FWIW, with 100% Ackerman, I get 5.9 degrees of slip angle at the inside front for the same cornering conditions, which would be close to "optimal" going by the chart on the previous page of this thread that suggests pro-ackerman. (fwiw, 12.75 degrees steering angle average at the road wheels, 227deg at the steering wheel)

 

The tire you are running will determine what is the theoretical optimum amount of ackerman, but it looks to me like it will generally be between zero and 100% Ackerman.

 

 

But *STILL*, I don't think the Corvette guys necessarily screwed up TOO too bad going with parallel steer. I still doubt that there is going to be all that much in it one way or the other.

 

 

The M3 would have inside tire slip angle of about 8 degrees for the above cornering case (30' turn radius, outside tires at 5deg slip), which would be significantly GREATER than "optimal" for the more pro-ackerman case of grip vs. slip angle vs. load.

 

I also don't think that the M3 guys necessarily screwed up too too bad by having what appears to be entirely TOO MUCH ackerman.

 

Again, it appears to me that there's not much in it, particularly for our cars.

Edited January 6, 2011 by Dan Baldwin

[JMortensen]

I think you're still neglecting the other aspects of Ackerman other than optimizing the lateral grip in a steady state corner. Did you read the paper?

[Dan Baldwin]

I think you're still neglecting the other aspects of Ackerman other than optimizing the lateral grip in a steady state corner. Did you read the paper?

I glanced over it. I understand the concept of more ackerman generally helping with turn in on lower-speed corners. I can appreciate that for a given amount of turn-in respones, more ackerman would reduce the amount of static toe-out.

 

At road courses, for higher-speed corners, whether ackerman helps or not will depend on how the tire lateral grip vs. slip angle curve and how it varies with load. If at lower loads the tire develops maximum grip at a lower slip angle, I want anti-ackerman. If at lower loads the tire develops maximum grip at a higher slip angle, I want pro-ackerman (generally speaking, of course it depends on the curves).

 

In any case, the stopwatch is the ultimate arbiter of what works and what doesn't. But if you were able to do back-to-back testing of stock ackerman (apparently parallel steer) vs. 100% ackerman vs. M3-level over-ackerman, you'd still have to optimize each individual setup in order to draw conclusions.

 

This isn't terribly likely to happen...

 

I have to say I still haven't seen anything that indicates to me that optimizing ackerman is going to gain much if anything in terms of speed at the track. Maybe for autoX...

[260DET]

This resolved yet

Edited January 7, 2011 by 260DET

[Dan Baldwin]

This resolved yet

It's not the kind of thing that's going to be "resolved" via discussion on this forum.

 

JMort's link gives some insight:

http://forums.hybrid...846#entry914846

 

Here's another link I came across last night (which references the author of the above link):

http://cgsuspension.com/ackerman.htm

 

From which:

My takeaways:

1) What you set toe for and whether you want pro- or anti-ackerman is going to be driven by how the tires you're on respond at the load levels they're subjected to (rough estimate, a decently set-up 240Z should be loading the outside tires to on the order of 1000 lb. and the insides at on the order of 250 lb.)

2) The unloaded inside tires' grip is a weaker function of slip angle than the loaded outside tires. So the effect of toe-out or pro-ackerman is more to do with additional DRAG at the inside tire giving a yawing torque in the "right" direction, helping to rotate the car on corner entry. Of course this additional drag is slowing you down on corner exit, but then you're unwinding the steering anyway. In this regard I would definitely prefer more ackerman to more static toe-in.

3) It would be good to know if the data accounts for the inside tire having adverse camber...

 

Seems to me that ackerman is only really going to have an effect in very low-speed corners. The amount of toe-out you get for smaller steering angles (~6degrees or less) is going to be in the tenths of a degree, even with fairly full ackerman or even "over-ackerman". Toe-out with steering angle isn't linear, and at zero steering angle the change in toe-out with steering angle is zero. It "ramps up" with increasing steering angle.

 

For reference, cornering at 1-g around a 100' radius equates to 39mph, a very low-speed corner for a road course. In that situation, assuming 5 degree slip angles for the outside tires, "full Ackerman" would give you 5.26 degrees slip at the inside tire. 0.22 degrees of the difference is from ackerman steering (.04deg difference in trajectory angles of front tires).

 

It turns out that the aforementioned BMW M3's relatively extreme amount of ackerman is a smidge less than "full" ackerman.

 

So for this *quite* low speed steady-state cornering situation, the difference between the extremes of parallel steer (Corvette, 240z), and full ackerman (greater than the high-end-of-ackerman outlier BMW's), is .22 degrees, or 3/32" of toe-out.

 

So, I guess if that saves 3/32" of static toe-out for the same amount of turn-in responsiveness in very low-speed corners, I'm all for it.

 

In particular, if the turn-in at medium to high-speed corners is good, and greater responsiveness is desired for very low-speed corners, more ackerman would be the preferred way to do it in the interests of keeping static toe-out minimized and keeping the same stability for medium-speed and faster corners.

 

BUT, if the tire grip vs. slip-angle vs. load curves look more like the "anti-ackerman" case shown above, parallel-steer might be "better" overall.

 

 

Long/short, I no longer think it's necessarily a waste of time/effort if the handling is good at higher speeds but turn-in responsiveness sucks at low speeds. But I wouldn't go by feel alone.

 

 

Whew...

[JMortensen]

That's a pretty good synopsis Dan. I would say that as far as "feel" goes, when a car goes from understeer to oversteer with one change, it doesn't take a lot of "feel" to recognize it. I asked my friend yesterday and he said he heated the knuckles with a torch and bent them so that the lines through the ball joint and tie rod intersected 1' in front of the rear axle. He then sent them out to be heat treated. So he also wasn't running "full" Ackerman.

 

I've always been one to improve the end of the car that doesn't handle, rather than loosen up the one with traction. After he did this change he had to work on the rear to match the front, and when he got that done I was never quite able to catch up to him anymore, unless we went to the big track where I cleaned his clock with a big hp advantage. You can call that feel if you like, and while it wasn't a scientific, repeatable test, neither was the one that determined for me that getting slapped in the face doesn't feel good, but I still trust the results of both.

 

I also went out to the garage and took actual measurements instead of extrapolating the figures that I had gotten from another thread, so I'm going to do more graphing and will post the results. Looks like a bit more Ackerman may be there than I previously thought, still not thinking that it will be very much. I'm also going to see what moving the front wheel forward an inch via lengthening the TC rod does.

Edited January 7, 2011 by JMortensen

[JMortensen]

I've learned a lesson again. Maybe it will stick this time. Never use extrapolated info like "the rack and tie rods are parallel to the LCA" when you haven't verified the measurements. As it turns out, the rack is set back a bit. I also was surprised to find that the tie rods are .5" shorter than the LCA. I went through and made measurements of my setup and graphed them out per .5" of rack movement.

 

Did some more graphing, here's what I came up with.

 

Stock (as close as I could figure, I've changed the LCA length so I guessed at 11" for the stock length from pivot to pivot, and I'm not sure that the control arm is directly perpendicular to the crossmember):

6.5, 6.5

12.5, 13

19, 20

26, 27

33, 36

 

With my current setup with the rack moved back 7/8":

6, 6

12.5, 13

20, 21

26, 28.5

33, 37

 

If add 1" of wheelbase to my setup:

6, 6.5

12, 13.5

19, 21

25, 29.5

31.5, 38

 

I also found that changing the tie rod length by +/- .5" has almost no effect. It appears that adding caster helps about as much as moving the rack back.

 

With a 5 degree slip angle and my wider than stock track, the "right" steer angle on the front wheels would be:

 

5,6

10,12

27, 30

30, 49

40, 63

 

So you can see that I don't have enough Ackerman, but the difference probably isn't enough to justify Z bar steering or anything else when you're looking at the lower steering values where it is more likely that the car will be driven. I'm going to add as much caster as possible, and for autox I may just run a heavy static toe out setting, and then I could probably run close to 0 at an open track and do reasonably well.

Edited January 7, 2011 by JMortensen

[260DET]

Ran across this discussion on Ackerman, may be of interest http://forums.corner-carvers.com/showthread.php?t=3846

[heavy85]

Continuing the discussion ... does this mean changing back to the stock length steering arms (vs the shorter one's I have now) would be a good change?

 

Cameron

[JMortensen]

Honestly, after looking at it I think the difference in terms of Ackerman would be next to nil. You could graph it out and not have to guess, but I tried making the arms longer and that didn't have much effect, and I think the quick steer knuckles are only 10mm shorter.