Ackerman through bump and steer

[rcx194]

I've come to wanting to check how steady my ackerman is through bump and steer. My static ackerman is around 30%. This is about as good as I can get as my steering rack is quite far forward of my outer tie rods. But for off road race use low ackerman seems to be favoured. I've been using Lotus Suspension Analysis.

The ackerman appears steady throughout steering:

However, through bump it goes nuts:

I spoke to Lotus and they said to me that as ackerman definitions rely on the difference between angles of the wheels they can become unstable when the wheels are both facing forwards. Their exact words:

Ackermann as it is one of those parameters that seems to attract a number of differing definitions. In addition as most of the algorithms tend to rely on differing steer angles between the left and right wheel, they all have the potential to be unstable as you approach the straight ahead case, (i.e when both steer angles are the same).

Unfortunately my tech support with Lotus has expired so I can't show them my graph to get any comments on the graph. To me it looks like a problem with infinities around the zero point, i.e. a problem with the formula. When I animate my design the wheels appear quite parallel with each other:

Any ideas how to check my ackerman further?

 

[GregLocock]

I suggest you plot the difference in steer angle (toe) between each wheel vs bump/rebound. This should allay your fears and confirm Lotus' diagnosis. By the by, who were you talking to at Lotus?

Ackermann at straight ahead is dominated by the static toe setting and is unhelpful.

Parallel steer (ie zero Ackermann) has been used successfully by circuit racers, as has 100% Ackermann. An experienced and successful race engineer told me it is just about the least used tool in his tuning box. If he had the time he would investigate the ideal Ackermann by running different toe settings and comparing slow and fast corners. Toe is fake Ackermann.

For production cars the main use of Ackermann is to minimise the turning circle at full lock, essentially by preventing the outer wheel from fighting the inner wheel. That is not the same as 100% Ackermann incidentally. That is not really a consideration for a circuit car.

For a circuit car there is theory that excess A may help to rotate the car into a corner, because the inner wheel will drag the car around. As theories go, plausible but not confirmed. If I was less lazy I would run that in ADAMS. I am lazy. Given that a race car should be operating with each tire on the limit of its friction circle at all times, (for a hot lap), I don't think you can really be prescriptive like that for a kinematic effect.



Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[NormPeterson]

Greg - I think I've even read where anti-Ackermann (negative A?) has some supporters. Something about the peak "mu" of the lightly loaded inboard tire occurring at a smaller slip angle than is the situation on the outboard tire.


Norm

 

[GregLocock]

That's another plausible hypothesis. Hmm, that could be a nice little project for a vehicle dynamics student. I'd run a swept steer to establish max latacc at a few speeds, and some sort of step steer and look at the details of the build up of yaw rate. Do that for a few different cars and tires and see if there is a general trend. You might get a different answer with trail braking as well. Here's a nice paper on the latacc part

https://www.google.com.au/url?sa=t&...7E09e0cP6Qx3JIKJA&sig2=Ys723YsUdudRmAe1RLtL-w

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[GregLocock]

There's a good thread on F1T and a relevant thread on the FSAE forum.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[cibachrome]

The use of anti-Ackermann is very easy to document and analyze. Simple road test procedures reveal the effects of using pro vs. parallel vs. anti Ackermann and the reasons for its specification:
Yes, the inside tire can be seen dragging down the front grip levels. However, this characteristic is highly dependent on the tire characteristics themselves as well as the load transfer distribution. In other words, its not always the case for a fixed steering geometry setting. Even tire pressure gets into the recipe.

One point of confusion is usually that 'correcting' the Ackermann either with toe or geometry will loosen the car up because it reduces understeer if you gain front grip (reduce the front axle sideslip gradient). This effect will usually be misconstrued as a 'bad' thing because the potential for higher max lat is there, just not the reduced control sensitivity necessary to attain it. Toe (out) changes of 1 to 2 degrees are commonly needed to get this extra sidebite. This can be a little 'tireing' on the tread on long straight runs. Change tire brand or construction on the same brand and it can be a whole new ball game.

Needless to say, the effect is the same on the rear, with the same tendencies: extra rear grip adds understeer and this will be viewed as 'worse' unless the front is also rebalanced to bring the control sensitivity back up. This is done (in the rear) by toe and roll steer alterations.

If I put my 'Student' hat on, I'll produce some tire carpet plots with 4 wheel tire force & moment traces on them. The effect(s) are pretty obvious, graphically. It makes a wonderfull excercise for simulation when you have tire data worthy of engineering the car instead of wrenching it.

Happy Holidays to all !

 

[cibachrome]

Here's one typical response example taken from analysis of a small student competition car subject to rules and engineering constraints. Tires are race slicks run on a belt surface tire test machine and fitted to an appropriate, non-Pacejka tire model. Tire surface asymmetries have been removed to make the 4 tires have no conicity or plysteer effects (Not that this makes a big difference for the example purpose)..

The car is still able to achieve steady state conditions in a constant speed increasing steer angle test, but does better with some toe-out added. Is not driveable with same amount of toe-in because of an oversteering level beyond its negative Ackermann gradient.

Put a different brand of tire on the car, different results in the limit. One key aspect of this analysis is the high amount of tire load reserve in play (The wheel weights are maybe only 1/2 of the tire's rated Fz load at specified pressure). This analysis is not limited to small rear weight biased race cars. Its a useful analysis for all high performance (as in race prepped track) cars.