E type Roll Centre and Handling

[nichmoss]

Hi Clive

Yes, these are major changes and I hope it works out. I can see why you have decided to change both front and rear suspension so you can tune the two together. You are the expert for suspension design so should know what you're aiming for and will, likely, achieve.

For the rear modifications, I can see how the beams are extended aft to protect the fuel tank (I have no idea how you will fit the tank as the original is very tight and those beams won't allow it to go in, I suspect). The modifications to the structure in the wheel area don't seem to link the new beams very well to either the front bulkhead of the boot floor, or the bulkhead behind the seats - these are two of the primary structural elements for transferring vertical loads to the suspension from the fuel tank and the passenger compartment so, this seems odd. It's not clear where you'll put the coils/dampers but you'd expect a good load path from there to these areas. For that upper wishbone, with it's lateral loads, I'm also surprised that you haven't braced across the diff area between the two new side beams - anyway, if you have someone you trust doing the structural mods then I'm sure it will all be checked. In the same way, I'm also surprised that the diff support structure doesn't extend down the boot floor front bulkhead as that makes the joint unnecessarily weaker for the sake of a small weight increase.

I doubt I could comment in a serious fashion except to say that we wouldn't build any aircraft like this where the loads paths are not well defined (at least from the pictures) - I suppose you don't have the longevity/lightness/safety concerns in the same manner but still, it doesn't look the best solution in my opinion. You're on the spot and may be able to see how this would work but with just tiny pictures, I can hardly see what has gone on so, you can take my comments based on that.

I think my original comment related to adding structure to support the changes you're making. I probably mentioned that there is a torsional stiffness plot for the shell with frames somewhere (although I can't find it) and it obviously shows a major loss of stiffness at the front bulkhead where the frames attach. If you're modifying the front, then this is one area to consider in detail. I believe this drop in stiffness is one reason for the torsion bar suspension which concentrates some of the loads at a big heavy plate in the bulkhead area. If you put all the loads through the frames, I would be concerned about the local areas where you're attaching the coils/dampers, the rest of the frame structure and that connection. All areas you should pay close attention to. If you can add a structural panel in any of the gaps between the frames, then you win a huge amount for the frame stiffness and you might also consider extending the lower frames part way down the tunnel in the same way as a D-Type. I don't think there is room on a stock car but this would certainly help with getting the front suspension loads back to the shell and increase the stiffness of your front suspension attachments.

Clearly, as you are well aware, if you don't have the shell stiffness (and strength) then the work you do on the suspension will be compromised and you've modified the shell and frames significantly so this is one area that might upset your work. It's been interesting to have the chat but I probably won't be able to post anything else of much interest so, good luck and I will not have stopped posting because I'm unhappy with your replies.

Best wishes, Chris

[CliveR]

Hi Clive, I’ll be keen to see what you come up with. I found this interesting video of Evolution Etypes IRS, it shows some extensive modification work that you might find helpful:

Thanks, I watched the video a few times but didn't really understand the Evolution system. My understanding of the original system is that the lower transverse link rotates about the pivot on needle roller bearings, it's pretty stiff, allowing camber angle change but not steer. The fore-aft radius arm is attached with rubber bushes so it works with the subframe mounts to give some level of steer under cornering loads. (Does anyone know which way it steers? I can't visualise it). Further, the radius arm bushes take up any length or angle variation as the upright moves up and down on its rigid links and that compliance probably contributes some refinement on bumps.
It seems the system in the video replaces the compliant radius arm with an angled link, the forward end being attached to the subframe instead of the body. That will presumably affect the steer in response to lateral force, again I can't visualise direction of the change with no idea of the relative stiffness of the rubber elements. I think the angled link also gives some steer as the suspension rises and falls - probably desirable in theory but it introduces a potential geometry conflict with the original links, if the new ends are solid fixings not flexible bushes (I doubt bushes would be useful there as they would be too soft to influence the geometry). All speculation as I don't know the detail. Does anyone here have any direct experience?
We've taken the easy route and gone for a double wishbone system, with a toe link to control steer. Inboard connections to the chassis are bushes (rubber and polyurethane), outboard connections to the upright are ball joints. More detail later as it progresses.
Cheers

[Tom W]

The way the original IRS works is really quiet tricky to visualise, and trickier to describe.

Considering the IRS as a stand-alone unit, the uprights are free to move through bump and rebound, relative to the cage. Because all the pivots are bearings, not compliant bushes, there’s no option for change in toe throughout the wheel travel. Camber changes through wheel travel as the upper and lower wishbones are different lengths.

When the whole IRS is considered as it’s mounted to the car, things become more complex. The IRS is fixed to the car about 4 compliant points (V mounts) towards its top. These are relatively rigid in the vertical direction and across the car, so they can support the weight of the car and transmit lateral forces. However, because they’re above the centreline of the wheels, there would be a tendency for the IRS to rock back and forward between braking and drive forces. This is where the radius arm comes in. Fore-aft forces are transmitted through the radius arm, into the body shell. The radius arm geometry is imperfect, so it’s fitted with compliant bushes both ends.

At the static ride height, the radius arm is near enough horizontal. So, in a corner, considering the outer wheel in bump, as the wheel and radius arm moves up, the outer corner of the IRS is pulled forward. Moving to the inner wheel, as the radius arm drops, that also pulls the corresponding wheel forward. As both wheels are linked to each other through the IRS with no relative movement possible besides bump, if one corner of the IRS is pulled more than the other, then the whole IRS will steer relative to the car. Interestingly, it can be seen the movement of one wheel can have an influence on on the steering of both rear wheels.

Where things become tricky to predict is in the compliance in the radius arm bushes. With the slots in the big radius arm bushes arranged as per the workshop manual, the arm is able to change length somewhat, negating some of the effect it can have on steering the IRS.

The Evolution system works differently. The fore-aft forces are reacted directly into the extended IRS cage assembly, through the angled radius arms. The arms have to be angled to get the forward pivot in line with the inner lower wishbone pivot. The IRS bottom plate is extended to meet the factory radius arm pickup points on the shell. Conceptually, it’s a double wishbone suspension mounted on a subframe that’s isolated from the body by a compliant mounting. No toe change occurs through suspension movement, and the movement of one wheel has no influence on the geometry of the other (anti-roll bar aside)

[CliveR]

The way the original IRS works is really quiet tricky to visualise, and trickier to describe.

Got it, thanks Tom. I didn't know the role played by the V-mounts, my interpretation from your description is:
- the Jaguar system will pull the wheel and subframe assembly forward with wheel travel in either direction from the neutral (longest) position
- the wheel with greater travel will tend to generate greater steer angle on that side
- steer would not occur in the rare situation where the inner and outer wheels move up or down by equal amounts
Interesting that there are situations where the wheels behave as not actually independent. I can imagine times when this may not generate the same predictable wheel motion (and corresponding driver confidence) as when steer depends only on the motion of a single wheel. Could there be situations where this condition is less comfortable than a solid axle, where the relationship between the wheels is fixed?
The question of radius arm bush compliance is interesting. If it has enough compliance to reduce its steer effect, I'm thinking it could introduce a feeling of vagueness or delayed response in some situations. Perhaps one of the compromises in making a suspension work across a range of luxury saloons and sports cars.
The above "coulds" are general observations, not based on direct experience with the system we're discussing.

[Tom W]

Yes, that’s my understanding of it. The only point that may be not quite right is my assumption that the static position for the radius arm is truly horizontal and therefore longest. I am writing this from the comfort on the house, rather than going out to the garage to properly check.

There are so many variables in how the system could behave depending on what the assumption is for the various levels of compliance between V-Block bushes and radius arm bushes. I think in reality it’s tuned to behave in a predictable, repeatable manner. Any rear wheel steering must effect both wheels since the solid pivots in the IRS don’t allow for any relative change of toe between the rear wheels. So in this respect perhaps it does behave like a solid axle?

In my practical experience, the standard suspension feels very controlled. There’s more body roll compared to a modern car, but once settled into a corner the car is very docile. The balance is very neutral and with small changes of throttle or steering the trajectory of the car can be adjusted. There isn’t the tendency to understeer that’s prevalent in more modern cars. The limit of grip is much lower than a modern car, but the feel of where that limit is communicated very well. The limit is reached gradually, rather than an all or nothing snap change. I’ve not encountered any unpredictable handling anomalies, so they’re either not there, or they occur outside of my driving envelope.

[CliveR]

Good point about the solid pivots restricting steer, although I wonder whether the tubular section of the lower link is stiff enough to do that? As an aside, is it stiff enough to resist bending from the spring / damper loads? Just curious.
While we're on the subject of the IRS pivots, is their axis nominally horizontal? That is, the wheel rises and falls perpendicular to horizontal vehicle plane (however we define that)?

[christopher storey]

The lower link would resist the bending loads in landing a jumbo jet !! here is a picture of it - it is a shame you have disposed of the originals because it would answer so many of your questions

[CliveR]

Hi Clive
Yes, these are major changes and I hope it works out. I can see why you have decided to change both front and rear suspension so you can tune the two together. You are the expert for suspension design so should know what you're aiming for and will, likely, achieve.

Thanks Chris,
You’re generous but I don’t actually have much background in pure suspension design. Most of my working time was spent in vehicle development, or vehicle integration as it’s known these days – matching and blending the properties that affect driver experience and satisfaction. Ride and handling are naturally towards the top of that list along with such things as powertrain performance, noise and vibration, heating / ventilation and a lot of intangibles such as squeaks and rattles, door operation, control feel….etc. It's the best job in the industry as far as I’m concerned. I picked up some suspension theory along the way, while tuning suspension and working with the clever people who do chassis design. The best part of the job, for me, is damper tuning. Those little tubes have amazing power to define the character of the car.

Thanks for taking the time to put your comments together. I absolutely agree your point about needing well-defined load paths, although it’s not an aircraft I share your conviction about longevity / lightness and safety. I should say the car in the pictures is a work in progress, still lacking some critical pieces. This is where I should confidently affirm we’ve covered all the areas you mention, but we won’t take that for granted – I passed on your thoughts to ensure we haven’t missed anything. I can’t be more specific as most of the structural planning for the rear was done in Australia by my partner in crime, Peter. He’s currently in the middle of a two month business trip across three continents, he has people working on the structure but I won’t see the results for a few weeks

One detail I can confirm is that the fuel tank will be a flexible bladder, hence the reduced space. Another is that from their proportions the upper wishbone loads will be relatively low, I’ll be surprised if the analysts tell us we need structure in addition to the new side beams and existing floor. If they do, we’ll add it. The rear diff support wasn’t defined when the photo was taken, it will certainly need some substantial structure for the torque we expect to deliver there.

The centre and front sections are more advanced in build, so easier to discuss. We’ve added structure in the centre section to ensure the whole structure is coherent, with high overall stiffness. The sills have full length internal beams connecting front and rear structures, the transmission tunnel is a tubular & panel structure joining the front and rear bulkheads. (picture attached). Moving to the front, we had the same concern as you about loss of stiffness at the front frame / bulkhead attachment. The upper outboard point looks particularly vulnerable in that respect. Considering the starting point as two long slender pylons which are imperfectly triangulated, supported somewhat insecurely at the bulkhead, our direction was to improve the triangulation of each pylon, connect them to form a single structure across the full width of the car and carry the loads back from the bulkhead into the centre section. We’ve taken the liberties that a one-off private project allows – triangulation under the engine, front box extended with tubular and panel bracing as appropriate, modified connections through the bulkhead. The new front box will handle the loads from the combined spring /damper. I’ll follow up with some pictures, further detail and some questions. To answer one vital question – the car will be lighter overall than the original, due mainly to the mass we’re saving in the powertrain

[CliveR]

The lower link would resist the bending loads in landing a jumbo jet !! here is a picture of it - it is a shame you have disposed of the originals because it would answer so many of your questions

You're right, thanks to those who've responded anyway. My questions and musings are really just curiosity as we committed to the new system long ago.
I've attached a crude markup to show roughly how our new structure is located relative to the original, it might be easier to understand than my recent words. It's very diagrammatic, not to be taken too literally in the details. The rear is still a work in progress, not complete as shown here.