E type Roll Centre and Handling

[Marquis Rex]

Greetings,
I have been digging deep into vehicle dynamics in general and Roll centres in particular.
My OEM background has had me deliver various vehicle projects where Ive had a good overview but Im now digging in technically deeper. The aim is to learn mainly. If I can benefit the community by sharing what I have learned I'll be happy to do so. I'd be happy to hear from people here if this is too technical and needs to be watered down or vice versa. This forum is the sort of technical audience level I'm aiming for, I'll cover what my key take aways have been and then share the roll centres I have. The E type is a nice car to apply some of the calcs to because, it generally has a simpler set up than modern multi link suspension cars- so is easier to understand.

These are some of my take aways-

•The Steering axle benefits from low Roll Centre , long equivalent control arm lengths result in steering predictability due to minimal track change, minimal RC height change , minimal camber change etc

• High Roll Centre at the rear controls roll/yaw coupling and having a higher RC at the rear stabilizes the vehicles by causing yaw in the opposite direction to the direction of turn as the car rolls

• Total Lateral Load Transfer Distribution or 'TLLTD' at the front builds progressively with body roll. Higher rear RC means lots of nearly instantaneous geometric weight transfer from the rear end, causing transient overtsteer that compensates for other perceived sources of steering sluggishness (improves turn-in). When driving the E type- I find even at modest cornering speeds I can feel the steering load up nicely. When it was happening, I was thinking of the roll centre at the rear being higher than the front and transferring load progressively.

• Raising front roll centre improves on-centre steering feel-The E type for me seems to have a lot of 'feel' and feedback compared to my X308 XJR and even my E24 BMW M635CSi. The X308s front roll centre is under the ground at -50mm. My Porsche 993 Turbo has the best steering feel of all of my cars.


- A School of thought- as shared by Gordon Murray and practiced on the Maclaren F1- was to have the distance between the C of G and Rc the same front to rear, and keeping the Rcs both low.
By having Rcs higher rather than lower (or zero) a larger part of the TLLTD will be geometric . With lower Rcs more load will be transferred via the suspension system itself.With zero Rcs the load transfer will transfer immediately via the tyre contact patch while the rest of the Wt will pass through the springs and shocks. In addition zero Rc allows the lionshare of the vehicle dynamics tuning to be done by fine tuning the suspension. Im not sure how I feel about this one- I have no direct experience of a zero Rc (frt and rear) vehicle and suspension tuning it.

And finally



The E types roll centres seem rather high- 280 mm at the front and 146 mm at the rear. This seems weird to me-with high Roll Centres in general and the rear being lower than the front. I was corrected by a colleague in the industry, who had the E type front and rear roll centres measured- on a Kinematics And Compliance Rig. He got 60mm at the front and 150 mm at the rear. This makes more sense to me.


Ive never pushed my E type to the limit, I know it has close to equal weight distribution, and have never heard of evil handling characteristics. The steering is very 'talkative'.
I hope this post is of interest to people.
This shows a list comparison of the Jaguar E type and its roll centres compared to other modern contemporaries:

[abowie]

The E types roll centres seem rather high- 280 mm at the front and 146 mm at the rear.

Very informative, thank you.

How did you arrive at the 280mm figure, or is the maths too complex?

[Marquis Rex]

Its not really complicated- you get pictures, like the one I posted, and take dimensions and angles.

Im posting my diagrams of how I worked out the X308 Jaguar roll centres









The 280 mm number for the front of the E type is just so far out of the norm. However you can see how I arrived at it, if you project/extend the LCA arm lines and do the same for the UCA toward the centre of the car, and then project where these lines come together. Its at this projected point where it comes together, where you extend ANOTHER line to the vertical centreline of the car. Its where this line meets the centreline that is the point of the roll centre.

I do think that the roll centre importance is sometimes over stated and also bear in mind that the E type was designed when cross ply tyres were commonplace.

[PeterCrespin]

Thanks for that. My D uses mostly E-type front suspension plus a heavy solid back axle on 4 unequal length trailing arms with a Watts linkage pivoting off the diff cover. The car rides on correct Dunlop peg drives with very tall 16” Blockley cross-ply tyres 6.00” and 6.50” front/rear. To get the ride height in D spec the front suspension is lowered, with 1/4” packing behind the UCA mounts and eccentric spindles to keep camber sensible. What would be your guess as to the effects of these changes on handling?

BTW should I call you Rex or Marquis? I’m guessing Rex, (which is also the name of my youngest grandson)

[Marquis Rex]

Please call me Rex.

By lowering the UCA inboard mount you would probably raise the roll centre by between 8-12mm (you're changing the UCA angle effectively).

I don't envisage huge changes to handling by doing this. You've done the right thing by controlling the camber.

Although roll centres DO effect handling , in isolation, not significantly, they're part of an ingredient for a larger cooking recipe. In this case the finished dish is the cars handling characteristics

[PeterCrespin]

Thanks Rex. I deliberately chose to avoid the common use of IRS to upgrade the rear suspension. Am looking forward to a 1955-ish feel when sampling my creation next spring, with luck...

[Marquis Rex]

Thanks Rex. I deliberately chose to avoid the common use of IRS to upgrade the rear suspension. Am looking forward to a 1955-ish feel when sampling my creation next spring, with luck...

Sounds like an exciting authentic project

[CliveR]

Hello Gentlemen - is this discussion still active? I've just joined the forum having agreed to help an old friend update the suspension on his E type. This might seem like sacrilege but we think there's a lot of room for improvement as the world has moved on in sixty years. As you're leaping to your keyboards I should add I was, until my retirement a few years ago, employed as a vehicle development engineer for forty-plus years, starting a few miles down the road from Browns Lane and specialising in ride & handling for the last twenty. I promise we'll respect the starting point but I do think we can squeeze more performance than the Jaguar engineers were allowed within the constraints they were clearly given.
I'd like to add comments to some of the discussion on roll centres in due course, please excuse me starting off with a request - Rex mentioned having K&C data for an E type, is any of that in the public domain? It would be very useful to correlate the model I'm building.
I'll be very happy to share thoughts and conclusions.
Regards
Clive

[mgcjag]

Hi Clive...welcom to the forum...any old post can be started up again and im sure this will be popular...please try to put your name and car model in the signature area...Steve

[Gfhug]

Clive, you could ask Rex for any data he used by clicking on his user name and either sending a PM (private message) or even find his email address is shown.
Keep us posted and be prepared to duck below the parapet as there’ll be those who shout sacrilege
You’ll see lots of posts about so-called upgrades, wider tyres, stronger springs or dampers. Many have been tried and found wanting, but if you do find an inexpensive improvement, we’d be pleased to hear.

Geoff

[bitsobrits]

Clive,

Would very much like to see a more developmental/empirical (vs anecdotal) discussion on this subject.

Utilizing 60 year old suspension settings with modern tires seems to leave much on the table, handling wise.

[Durango2k]

Not really E related, but Citroen did a lot in 1934 with the 11CV and even more in 1955 to get the centre of roll as much up as they could. Sadly I have no data, but their goal always was a low centre of gravity (ladder frame in DS, SM, and thin window pillars, one spoke steering wheel, thin glass, plastic roof, alloy bonnet front / rear, plastic rear window in early DS). This way they could get away with their ultra soft suspension with long ways the car body could move up / down.

They used twin control arms on the front axle (tecnically like the E‘s A-shaped parts), and long drawn arms on the rear.

Then, in 1990, Peugeot dictated that the XM was to be a version of the 605, somehow, so they tried a McPherson adaption of the hydraulic system, and that car had neither a low centre of grav nor a high roll centre, hence it did not drive nicely- then they invented a third suspension sphere plus some electronic trickery which failed every day, so the cars went into safety mode which gave a harsh ride, so the whole hydropneumatic system not really made any sense at all in these.

So- yes a high centre of roll is good, and if you then also go for a soft suspension with long ways, and inboard brakes for low unsprung weight, you‘ll get a very comfy yet safe car (the SM being capable of 230 km/h).

Carsten

[Durango2k]

… oh and I am somewhat convinced that the E- Type‘s front suspension setup is a bit a derivative from the 11CV. Both have triangle arms, and both have torsion bars. Or is there any other (british) car setup that would have been „closer“ ?

[christopher storey]

Well, the XKs and the Marks V, VII, VIII and IX used a very similar torsion bar setup, but I think that the similarity to that of the Traction Avant was noted even in the late 1940s when it appeared

[CliveR]

First, thanks to Geoff, I've sent a email to Rex hoping to get some of that precious K&C data.
Following that I feel I owe some explanation since I appear to have started the discussion going again. The project I'm working on started, as so many do, with musing on what we might do to upgrade a car that needed major rebuild work. My partner in crime, the car's owner, shipped it from the US to Australia and is very much open to modifications to update the performance to match the appearance. I spent my entire working life as a car development engineer so while I respect history I generally feel compelled to find the weak performance areas and improve them. (I have serious plans for my old TR5, not relevant here). To get to the point, after a bit of study we're going to fit new suspension front and rear. While acknowledging the efforts the Jaguar engineers made to fit reasonable systems under that beautiful body, there really are better ways if we relax the cost constraints and allow modern parts and systems. One example has already been covered in this forum - the front roll centre. As Rex mentioned, it's very high. More significant, it moves a lot as the suspension rises and falls. Undesirable in itself as it will tend to give a lack of certainty or connection in steering feel; even more so when compared with the relative stability of the rear roll centre. I estimated the roll centres very crudely from the drawings posted on this forum, I've attached the plot to show how roll centre heights vary as the suspension moves up and down. My figure agrees quite well with the figure of 280mm mentioned by Rex, at "kerb" loading condition, but it drops severely as the car moves into a bump situation - under braking for example. Ignoring all the other variables for the moment, a high front roll centre will tend to give understeer so conversely the car will tend to move towards an oversteer balance as it pitches under braking. If the rear of the car is rising at the same time, the graph shows us the rear roll centre rises giving another cue towards oversteer. None of these is a guaranteed spin, they're indicators of a tendency. I believe we can do better with current tools and materials. For illustration, I recall the hours it took a designer at Triumph to produce a plot like the attached, with ruler and pencil. It was impossible for them to explore the full range of possibilities in all the parameters we can study in minutes with very affordable software. I don't think it's immodest to say we can do better.
Wish you all good health

[christopher storey]

Another reinvention of the wheel . What a shame Norman Dewis wasted so much of his time and got it all wrong

[Gfhug]

Agreed Christopher. Perhaps Clive and his partner in crime should read Norman’s autobiography and appreciate the work he put into the handling of all the cars he was involved with. Significantly more than it would appear were done by other manufacturers.
If new suspension parts start being designed and fitted what would be the response of certification authorities in, say, Australia? From comments elsewhere they seem reluctant to have anything other than original design parts fitted.
Also, what happens to the character of the car? That’s a big reason for buying them, they need driving not just sitting in and having the work done for you.

As I said before, if Clive can see a simple, inexpensive “improvement” that won’t detract from the character of the car then all well and good.

Geoff

[mgcjag]

Hi Clive....A good starting point would be to actually drive an E type...have you?....road or track?.....A standard E type is more than adequate and will easily keep up with modern traffic......even on a "slow"track day it will handle well....but nothing like a modern "hot hatchback".....a well prepared "race" E type would be a better track car.....if its track performance your after then Rob Beere Raceing would be a good starting point..they do plenty of handeling mods.......as you say its now fairly easy to plug and play with figures in a program.....but nothing like actually making changes and test driving. ....whats your actual thinking....just changes to the current suspension...or a radical change like an E type body on a modern set up......Steve

[CliveR]

And good morning to you! Unfortunately I never met Norman Dewis in person but I do appreciate the efforts he must have made to get respectable results from that curious rear suspension. Skilful development can take you a long way. Having worked in the same field as Mr. Dewis, though, I'm confident he would have been very happy to dump it all and go round again with a more favourable layout if the opportunity had been presented to him.
I won't address the certification question here, my partner in crime (as he would be delighted to be known) will take care of that.
As to the character, if you mean the ride, handling, stability and steering response, it will be improved. Admittedly we might lose some vagueness and uncertainty along the way, we might have to disagree on whether that's a desirable aspect of character.
To answer Geoff's point, after a minimal amount of study I suggest you might try lowering the front of the car. That's a pretty obvious response but I can support it with a chart! According to my chart this will not only lower the front roll centre, it will bring it closer to the rear. The significance of that is the roll axis will not be so steeply sloped. When the roll axis slopes steeply the car can feel as if it's rolling in a conical manner (wider at one end than the other), which in turn gives a disturbing feeling that you're not quite sure where it's heading. These are subtle factors that depend on other factors, damping being a major one. By my experience you don't need to identify the feeling to find it unsettling, you just feel life got better if it goes away. Like noisy kids. That indirectly addresses Geoff's point - personally I don't get a lot of pleasure from struggling to direct a car, I get the enjoyment from identifying the engineering root cause, fixing it and enjoying the result. I've worked with hundreds of design and development engineers on four continents and never known one who chose to make a car hard to drive in order to give the customer more challenge.
With all of that said, I recognise my views and activities will raise some eyebrows among this fraternity. i hope you can tolerate a range of views, I will try to contribute in a positive manner and be sensitive to those with greater reverence for originality.
Moving along, I've updated the roll centre chart I posted earlier to be more informative. I've added the grey line at the top to show the suspension moving 120mm, with the roll centres also moving 120mm to suit. In other words, the roll centres move with the suspension. Ideal? Not necessarily, but somewhere between that and zero movement (imagine the grey line horizontal) would be pretty good. It's obvious the rear suspension comes a lot closer to that than the front, so I would address the front first to get a more precise feel. Ten minutes doodling suggested you could get the roll centre down appreciably by moving the inboard upper wishbone pickup but I didn't go as far as checking the slope of the curve, which is what really matters. Moving the lower wishbone would be a lot more effective but there's that sort-of-torsion bar complicating matters. This is where we came in, how we moved to major mods.
I hope some of this is helpful.

[Tom W]

The lightweight E-types had the front upper wishbone inboard pivots positioned lower than those on the standard road cars. There were also changes to the rear geometry. They had a wider track thanks to longer lower wishbones and a spacer between the diff output flange and the driveshaft which has the effect of moving the rear upper wishbone pivots outboard.