Trying to make my Sports Saloon ride and handle like a rival one

[Marquis]

Hello everyone I have a desire to have my (1998-2002) X308 XJR ride and handle like an (2000-2005) E39 BMW M5. Let me preface this by saying that I’m NOT interested in buying an E39 M5: The Jag stays, I love the looks and style of the Jag and it represents a special time for me- from when I worked for Jaguar. The BMW rides very well over the terrible broken surface mid western roads over here but remains very composed and taut. I never cared for how the E60 M5 rode and IMO I thought BMW hit a highpoint in terms of R and H compromise on the E39 M5. I drove one again recently to see if my memory of driving one (during the development of the Jag XJR at Whitley) had drifted but alas I found it as incredible as I found it then. The Jag is very soft, -which is fine, but it seems to lose composure over bad roads and at 9/10ths, in addition, if I boot it hard to kick the tail out and then catch it, the ‘wallowness’ makes it degenerate into a drift in the other direction. Very annoying. The only other car that I remember that used to do this was my 1970s BMW E21 323i.
X308 XJR steering is delightfully quick but lacks ‘feel’. The BMW has GREAT steering, its slower ( about 2.7 turns LTL vs 3) and it also still utilizes a steering box! Someone suggested a slower rack may transmit more steering feel- but going to a slower rack is not an option for me!
I managed to change the servotronic ZF relay in my Jag to one that lowers the assistance level more at higher speeds. The difference is barely perceptible

I had the E39 M5 spring rates sent away to be measured. The values were supringingly soft. When I factored in the vehicle corner weights, unsprung mass, the motion ratios- I’m getting 1.34 Hz ride freq at the front and only 0.95 Hz at the rear. This is contrary to flat ride theory so it has me miffed.
Torsional rigidity comparison: 13,000 Nm/deg vs 15,000 (BMW claimed 22,000 at the time but we tested this at Whitley and found different results)
Jag less than 52% front weight, BMW close to 50%
Unsprung masses surprisingly similar! (100-120 lbs)
Dynamic index: This is a concept a colleague and vehicle dynamics/R and H expert taught me about. He said that all the vehicles well tuned for great dynamics on the street and track have a dynamic index of 0.9-1. Ive attached a diagram of what this is. Suprisingly both the Jag and the BMW are inline with each other.
Roll centres – I will be looking at that next and comparing

Toe settings- I’ve heard that increasing tow out can increase on centre steering feel
Castor, traditionally BMWs have used a lot of castor- between 5-8 degrees, while Porsches have used up to 11 degrees. The Jag is at about 4.5 (between 3-6 degrees). I could target a value of, say 9 degrees . I don’t know how adjustable this is- and may have to resort to custom parts.
Raise the front roll centre- this is probably unwise as it has other handling implications….
I am unable to get the OEM damper rates- but I was able to get hold of SOME for the previous Jaguar X306 inline 6 XJR shock data (bump and rebound at various velocities). I know for a fact that the previous inline six XJR was set up stiffer than the X308 V8 XJR I have.
I have found a place that will make fully adjustable dampers (for low speed and high speed bump and rebound) for my Jag to my baseline curve.
In light of the almost unbelievable spring rate measurement data I got about the M5, I may look to get the X308 XJR springs measured also.

I tried estimating the rate using various methods- using the properties of steel, number of coils- (assuming number of active coils etc) looking at the front to rear motion ratios, looking the amount of spring travel while knowing the corner weight. I worked out the unsprung mass component by component. None of these methods correlated to one another and for the BMW none correlated to the measured data.

Next steps will be trying to assess the roll centres front to rear- comparing the Jag to the BMW (although with my terrible results in trying to calculate the E type roll centres (see my other recent thread) I’m a bit worried, and then get the stock Jaguar XJR spring rates measured.
How do I measure the roll stiffness of my Jaguar XJR?
I’m open to any other ideas.

http://www.auto-scape.com

Sideways To Victory!

 

[cibachrome]

Couple of things (I actually prefer the E46, BTW):

By messing with the ZF to reduce the boost, you added to the net steering compliance because you lowered the gear stiffness. So, you lowered the steering gain because you increased the understeer contributed by the steering system. This adds load/torque (and thus even more compliance) because other parts are worked harder. Since steering parts are always nonlinear (soft then hard), I don't 'feel' you accomplished what you want. In fact, gone the other way.

The Bimmers use caster offset as well as caster to accomplish their steering feel goals. I'll let you guess what their knuckles/uprights look like. This can make front driveshaft action quite a mess if you also desire 4wd.

BMW uses tires with VERY SPECIAL FY and MZ traits. Their real OEM tires have a circled star on the sidewall. Replacement tires installed by an official BWM dealer may not even have the star tires if your experiments require several sets of replacements. The car(s) will be different even though you changed nothing in the suspension, steering, chassis, pressures or rims. BMW's are best on the Contis, BTW. You will need to guess why.

You need to know that Jags are lame because they need to have large (i.e. stiff) tires for their total vehicle weight because of all the other sponges in the suspension. This gives them poor transient response because they even have to mount the steering gear in Jello to give it some undertseer for GVW load, replacement tire issues and spare tire use, you know, SAFETY reasons.

Get a VBOX or a yawrate gyro and a speed signal and run a fixed steer test over a speed sweep and you can measure your car's steady state properties in as many conditions as you can think of. Heck, I've even done it with my BOAT !

BTW: you may need a good 'haptic' driver to make sense out of all this. Greek word for FEEL. If you don't have a driver who is "better than the car", move on to the next project.

 

[GregLocock]

"I’m getting 1.34 Hz ride freq at the front and only 0.95 Hz at the rear. This is contrary to flat ride theory so it has me miffed."

Yes, that's right. For performance cars the ride balance is ignored. The priority is to maintain traction at the driven axle, ie low spring rates. The dead axle is then set stiffer to help provide a stable base for the live suspension. I'm pretty sure Milliken discusses this.

For practical reasons I seem to have ended up with roughly equal front and rear frequencies on several production cars, certainly way adrift of the 1.15-1.20 target ratio from flat ride. The result of that is needing a smaller sta bar at the front, and more work for the damper tuning guys to get a good ride.




Cheers

Greg Locock


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

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

 

[Marquis]

Cibachrome, thank you. I do actually prefer the E46 over the E39 as more of a sporty car but the E39 R and H to me is great as a daily driver on these dreadful roads.
Thanks for the info on the vbox or the yawrate gyro- I'll certainly look into it.

I had a friend who worked at COnti, I'll dig into the OEM tyre aspect.

I don't fully understand the steering assistance comments, I'll have to re-read a couple of times to understand.

Greg, thanks for that insight. I was beginning to question the results. I dug into older platform 5 series BMWs and they seemed to have a similar F:R set up, where as the smaller 3 series (like the E46) does not. Im not brave enough to copy the 5 series set up on the Jag without damping theory expertise.

http://www.auto-scape.com

Sideways To Victory!

 

[cibachrome]

A couple of other comments based on your own tactics:

2.7 vs. 3 turns LTL means nothing to anyone because you have no measured the overall steer ratios. Yo can't presume that the steer arm lengths are the same or that the uprights turn the same amount. Plus you may not realize how BMW can regulate an overall steer ratio by phasing the intermediate shaft and by having column and pinion shafts that are not co-planar. As for Jags, you may find that they have no concept of these steering geometry influencers as shown by the high variances in overall ratio measurements ON THE SAME MODEL. Some also have a different ratio turning left vs. turning right. Vehicles which exhibit these traits demonstrate that their production processes are out of control.

These cars don't have "power steering". They have power ASSISTED steering. This means that they have a mechanical connection from the pinion to the gearset that is usually a teeny tiny little shaft (the T-bar) just a few millimeters in diameter. As the road load builds up, the T-bar twists. Attached to this shaft is a rotary valve that admits pressure to the assist chamber(s) (left or right hand turn). The rotary valve is nonlinear, meaning that the more the T-bat twists, the pressure increases disproportioally. So, your steering wheel torque comes from tire and suspension forces and moments attenuated by the assist pressure reaction. By cutting back the boost (assist) force as you said, the T-bar carries more load (and thus more displacement and more understeer). Meanwhile, the higher pinion torque increases the torque on the steering column parts and that adds some understeer, too.

As for adding toe-out to improve 'feel', that depends a LOT on tire properties. I would doubt that toe-out helps you at all in this regard. These settings drive relaxation response characteristics, so, once again, you tire brand, size, construction, tread depth, pressure and rim width all have a say in your 'feel' recipe.

Lastly, monkeying with caster certainly is a common tactic in amateur racing venues. But, few recognize the changes that accompany such a 'caster change'. You also change the outter tie rod ball height and thus the ride and roll steer (bump steer) by a large amount and a few other parameters, besides these. Funny how doing this type of caster change actually un-does any benefit you were hoping would improve the car. Kingpin friction is usually the first thing complained about.

 

[Marquis]

"The Bimmers use caster offset as well as caster to accomplish their steering feel goals. I'll let you guess what their knuckles/uprights look like. This can make front driveshaft action quite a mess if you also desire 4wd."

They probably do that to reduce mechanical trail.

"You need to know that Jags are lame because they need to have large (i.e. stiff) tires for their total vehicle weight because of all the other sponges in the suspension. This gives them poor transient response because they even have to mount the steering gear in Jello to give it some undertseer for GVW load, replacement tire issues and spare tire use, you know, SAFETY reasons."

So in your opinion what is an effective way to tackle secondary ride for a comfort- like an E39 M5 does?

http://www.auto-scape.com

Sideways To Victory!

 

[cibachrome]

Best way to take a stab at ride is to look at the transmissibility of the seat. This includes arm, back, thigh and leg support. You might find it interesting that you can take the seat out of high end MB and put it into your Jag and the ride will not be so great. Same for the BMW seat. The seat alone does not make a good ride score. And, Jag seat in a MB is also a big disappointment. The seat elements need to be tuned to the specific coherence of the chassis and steering wheel interface.