[brian_]

Since this thread has been revived...

The rear end will need to be redesigned entirely. The engineer didnt seem to worried about that so he has boosted my confidence. I'm thinking a trailing arm set up at this stage. Double wishbones could be tricky to fit around the motor.

You can go from ideas and problems straight to a suitable and fully developed design in one salvage purchase... of a complete Tesla Model S (or X) rear subframe with suspension. Yes, the same thing that motor came out of.

Chris did this in his Tesla Powered Nissan r32 skyline. It's a major project structurally, but doesn't require suspension design or fabrication.

 

[brian_]

The rear end will need to be redesigned entirely. The engineer didnt seem to worried about that so he has boosted my confidence. I'm thinking a trailing arm set up at this stage. Double wishbones could be tricky to fit around the motor.

Independent suspensions are challenging, because the wide drive unit limits the length of the lateral arms. That can be worked around, but it means that any production multi-link suspension - other than the one built specifically to fit for Tesla - is unlikely to fit; double wishbone suspensions will be just as unlikely to fit.

Because the Tesla motor is behind the axle line, there is more space for suspension arms ahead of the axle line. That means that trailing and semi-trailing arm designs are most likely to fit. Trailing arms don't work very well and are only used in designs that link the arms with a twist beam, which is not easy to home-built properly and might be difficult to retrofit to a car not designed for it. While the semi-trailing arm design as been obsolete for decades (it was popular in the 1970's), it's the most common design used with Tesla drive units in projects in this forum... which are generally older BMWs and Porsches.

Semi-trailing arm suspension geometry is crude and can be problematic, but there is a multi-link variation used by the Fiat Doblo which has better camber control and still puts all of the arms ahead of the axle line:

There have been other clever multi-link variations of the semi-trailing arm layout; this is just one that I have seen that I could think of offhand. Of course to use this from a production car you would need one which was rear wheel drive or four wheel drive (the Doblo is front wheel drive).

Building a good independent suspension from scratch is not easy. Certainly the easiest independent suspension which will fit around a Tesla drive unit would be semi-trailing arms. You could build a custom crossmember to mount the suspension of an older large BMW sedan (such as a 7-series up to the 1994 E32) to the mounting points in the XJ6 structure, although even that isn't guaranteed to fit.

... yes it was tempting to get the whole setup but one regulation here in order to get the car registered is that you cant widen the car more than two inches. Tesla is much wider than the jag unfortunately.

For some cars it might be possible to use narrower and/or further inset wheels to reduce the overall width, but there are mechanical limitations on that, and the width difference in these cars is indeed substantial:

1969 Jaguar XJ6: 1768 mm / 69.6 in (from automobile catalog)
Tesla Model S: 77.3 in (1,964 mm) (from Wikipedia)​

That's 204 mm, 154 mm more increase than allowed.

The Jag apparently came with 205/70R15 tires, so just using stock-width tires would knock off some width, but I don't know if the whole difference could be made up.

 

[brian_]

I dont underestimate the challenges involved in creating the suspension from scratch however at this point i have benefit of ignorance and optimism .

I understand Damian used some form of trailing arm suspension on his BMW. as you said Porsche has used it for years and thats a good performance car so i know its possible and performance can be good. what level i can achieve remains to be seen.

Damien converted a BMW 8-Series (E31). Although this car came with a relatively modern multi-link suspension, he couldn't make that work with the Tesla drive unit, so he swapped in a semi-trailing arm suspension from an E34 car (a 5-Series up to the mid-1990's): Tesla Powered BMW E31 8 Series post #17

The history of Porsche is built on semi-trailing arm rear suspensions (because that's what an upgraded air-cooled VW had, and because it fits easily with a rear engine), but they haven't used it for about a quarter century. The 911 conversions in this forum are all older models. The multi-link systems in newer 911's would likely work, too, since they are designed to fit around the transaxle and rear engine, but that's not a cheap source of parts! On the other hand, the 993 generation of 911 (993 is the internal type identifier; 911 is the marketing model name) has the suspension mounted to an interesting aluminum subframe system which has separate right and left parts, which could be excellent for mounting to a tubular steel framework around an electric drive unit.

A rudimentary design i had in mind was what i'd seen in offroad camper trailers.

Excellent... for an off-road trailer. It can work for a car, but is no longer used in quite that configuration.

That's a pure trailing arm design, which means that the lines through the pivots run straight across the vehicle. With reasonable bushings this leads to the rear wheels steering undesirably when under cornering force; setting the pivot axis at an angle somewhat rearward toward the inboard end causes both toe inward and negative camber on the outside of the turn, which helps. Since the arms then don't purely "trail" (point rearward) the result is called a semi-trailing arm suspension.

Even for the trailer, semi-trailing would be better. There are lots of trailing arm trailer suspensions (almost all independent light trailer suspensions are of this type), but there are a few which are semi-trailing for better stability.

Fitting something OEM would be my preference because than someone smarter than myself has designed it. then it would also have torsion bars and proper bushings etc.

I agree - this sort of design is something that is best done by someone with more experience and resources than a typical DIYer, although the semi-trailing arm system is at a level which can be home-designed successfully.

Few modern suspensions use torsion bar springs; coils work better in most cases. If you mean a stabilizer bar... yes, having that already built is one of the advantages of using production pieces.

 

[brian_]

That's a pure trailing arm design, which means that the lines through the pivots run straight across the vehicle. With reasonable bushings this leads to the rear wheels steering undesirably when under cornering force; setting the pivot axis at an angle somewhat rearward toward the inboard end causes both toe inward and negative camber on the outside of the turn, which helps. Since the arms then don't purely "trail" (point rearward) the result is called a semi-trailing arm suspension.

In case this wasn't clear without an illustration:

(from Autozine - Suspension Geometry)

 

[brian_]

Semi-trailing arm suspension geometry is crude and can be problematic, but there is a variation used by the Fiat Doblo which has better camber control and still puts all of the arms ahead of the axle line:
...
There have been other clever variations of the semi-trailing arm design; this is just one that I have seen. Of course to use this from a production car you would need one which was rear wheel drive or four wheel drive (the Doblo is front wheel drive).

Okay, for those thinking that I've missed the obvious... the very common multi-link variation which looks like that one from Fiat, but is for rear wheel drive, is BMW's Z-axle. It appears in many BMW models starting with the E36 generation of 3-Series.

(from Brent Ford Racing)

Assuming that you would use a different subframe to mount the inner ends of the lateral links, it looks like this would leave the entire space behind the axle line clear, and would likely work well with a Tesla drive unit. It's not as good as more sophisticated multi-link designs, but at least it isn't just semi-trailing arms (which are inferior to the original Jag IRS).

 

[brian_]

Thanks for the link to the autozone page, the specifics of 50-70 degree pivot angle are useful to have and i've kept the link in my document of suspension information.

The pivot axis may also be a bit off of horizontal, and that angle matters as well.

Looking at the car, I can imagine some sort of multilink design, but it fitting and performing well dont necessarily go hand in hand. The semi trailing arm was easy to fit and function adequately.

Very true!

The original jag suspension uses the half shaft as a axially loaded member which I cant do with the tesla unit.

Yes, that's a common characteristic of obsolete IRS designs, and not acceptable for modern components.

Complicating things, the chassis design and motor size dont lend themselves to the introduction of an upper control arm. Its possible to have one now but it would be ridiculously short resulting in wild changes in camber. It may be possible to have one roughly above the axle but the chassis limits travel.

The half shaft as suspension link does eliminate the need for a separate upper arm, and the body design takes advantage of that.

A common technique to improve this interference issue is to make the upper lateral arm boomerang-shaped, so it curved under the frame rail. This makes it less stiff, so careful design is required. This aftermarket upper A-arm for a Ford Mustang rear suspension shows a curved upper arm in tubular steel: IRS Rear Upper Control Arms for 1999-2004 Mustang SVT Cobra
Many curved upper links are cast and nearly a plate, such as the upper link in the rear suspension of Ford's "Control Blade" design used in the Ford Focus (and the many other cars of several brands which copied it) for many years. Ford built a display for the introduction of a new Mondeo generation showing very different old and new rear suspension designs... with a boomerang upper control link in both of them.

 

[brian_]

Strut suspensions are often incorrectly called Chapman struts when used in the rear, where they don't steer. A real Chapman strut uses the half shaft as a lateral suspension link, like the old Jag IRS... I don't think anyone wants that, for either the loads on the final drive, or the mandatory use of U-joints, or for the geometry limitations of the lateral link location controlled by the differential output height.

A MacPherson strut suspension, using some combination of lower links to locate the bottom of the strut and control toe, has been used at the rear on many vehicles, although not frequently with rear wheel drive. This design is now uncommon at the rear, although even the most recent Porsche Boxster/Cayman still uses it (likely to keep this model cheaper and less capable than the 911). It does avoid the need for a lateral upper link, but requires a tall strut. That strut usually carries the spring, making it bulky, but it is also possible to put the spring on the lower control arm... a design typically called a modified MacPherson strut. The spring for a modified MacPherson strut can be a torsion bar (as in Porsche and Honda designs) or a coil (as in the Ford Fox chassis); a coil spring on the lower arm could be positioned similarly to the coil-over shock of the original Jaguar IRS.

 

[brian_]

World Rally Car regulations for the last decade have been based on the Super 2000 spec, which mandates MacPherson front and rear, despite WRC cars being 4 wheel drive. Super 2000 is also used for the World Touring Car Championship, where BMW were series winners several times with their E90 320si.

I was aware of other racing applications, in which the suspension was unrelated to the production vehicle, but I had not looked at the details of these two series.

These are good examples of situations which are controlled by competition regulations, rather than technical design factors. A full-cage race car is a relatively easy structure in which to mount a MacPherson strut, and packaging around a rear seat and cargo compartment is not a factor (interfering panels are just cut away); in both respects, this is a very different situation from a street-use sedan with a production unibody and full interior.

I can see why these series might specify the design. They want a single design to level competition, an inexpensive but capable design to keep competitors cost down, and perhaps even spec parts from a sponsoring supplier. This is done with many parts of competition cars, which often have minimal relationship to any production car. Super 2000 is a good example mechanically, whether in WRC or WTCC, although production-related bodies are used. A WRC or WTCC car also doesn't keep a functional rear interior.

A real BMW E90 3-Series doesn't have MacPherson strut rear suspension (and probably none of the cars in these series do) - it has a modern multi-link.

None of this has much to do with a classic Jag conversion. If a strut suspension fits easily that's great, but it seems unlikely.

 

[brian_]

MacPherson struts in a non-steering application can be arranged a little differently from usual, to address the packaging problems. Normally the strut is nearly on the axle line, for the desired steering geometry, and with a driven axle that forces the strut to be very high, above the axle. In the rear (where it doesn't need to steer), the strut can be behind the axle line, so the bottom of the strut can be lower than the axle shaft. This approach is an option, if it helps, but I suspect that it won't help with the frame rail issue.

 

[brian_]

Strut suspension is easy to design and make - and is easy to set to achieve the suspension results
The fact that the angles are set by the position of the strut top means that you can get precise angles

The only issue that I can see is that it prevents some body shapes - but only at the front - the back is almost always high enough to mean that you can fit your struts

It's no worse to design than other non-trivial (not single arm) suspensions, but Duncan wisely chose not to design or make one, but instead to incorporate a production Subaru suspension to the Device.

True, there's lots of height in most bodies, including the XJ6. Of course, there's that little detail of an existing body structure...

I used Strut suspension on the front and rear of my "Device"...

The Device is a tube-framed vehicle designed to fit the components, including the suspension. That's a substantially different situation from the XJ6.

 

[brian_]

Brian
You are completely WRONG in this...

That's strange... because I didn't disagree with any of the points which followed this, except...

Almost all current production vehicles use Strut suspension...

No. A large fraction (probably the majority by sales volume) of current production cars (not light trucks) do use MacPherson strut front suspension, but few have used it at the rear in the last decade or two.

We're supposed to be discussing a suspension to make a Tesla drive unit work in the back of the Jaguar XJ6. In the current Jaguar lineup only the E-Pace (the smallest and least expensive model) uses a MacPherson strut front suspension, and none uses a MacPherson strut rear suspension. The rest of the front suspensions are all extended hub carrier double A-arms, while the rears are all multi-link (generally Jag's Integral Link design) or double A-arm.

 

[brian_]

I had looked at McPherson Struts (the difference between them and Chapman struts is not immediately obvious, perhaps how they mount to the wheel hub?, i'll look into it further).

The difference is that a Chapman struts use the axle shaft as a suspension link (so cornering force is carried through the differential bearings), while MacPherson struts have normal shafts with CV joints and separate suspension links. Few cars ever had Chapman struts, and none have had them for decades; they're a historical curiosity.

Holding all the wiggliness of a 400kw car with a single control arm and strut seemed like a lot to ask so I had perhaps incorrectly, dismissed the idea early on for lack of understanding.

There's no problem with strength at all. The strut does take bending stress (unlike the shock in other designs), but they have a larger rod diameter and this isn't a problem. Where MacPherson struts are currently used (at the front of unibody cars, particularly those with transverse engines), they're usually supporting the heavier end of the vehicle and they are often controlling the drive axle. The current 6- and 8-wheeled vehicles armoured vehicles used by the Canadian army have strut suspensions for the forward axle(s)... and they weigh 16 tons or more.

After our conversations so far, I had another look at the rear suspension today. The front upper control arm is about 20cm long, the lower, 30cm long, not as long as I thought. I had dismissed the idea double wishbone suspension would be possible in the rear. With some clever positioning, perhaps I could get something similarly long into the rear, I had expected it would need to be much longer to get reasonable wheel travel. The current rear lower control arm is about 45cm long and may need to be shortened, I imagine there is a sweet spot for the length of upper vs lower control arm. All that said, this train of thought has just landed me in the poo pile that is lining up u-joints and suspension pivots and I can feel a cold sweat coming on

As I said I had not put much thought to suspension design and now i'm suffering some errors as a result. For now I should put all options back on the table and educate myself for all options their strengths and weaknesses. Regrettably there is no perfect, easy and cheap option for my handling/ performance goals.

There are a lot of factors in suspension design. I wouldn't consider playing with arm lengths without a substantial understanding of the geometry, which is a good reason to use a production design. For instance, there is no ideal ratio of upper to lower control arm length.

The strut design on the back of my device is (each side)

Two in/out links - one with length adjustment - both effectively simple tubes with eyes for the rubber bushes

One fore/aft link - in the Subaru it goes forwards but I swapped the rear hubs right to left so that it went backwards

These three nicely constrain the bottom of the strut - then all you need is the top of the strut...

This is a common design of the 1980's; I had a 1984 Toyota Tercel with the same design. The "in/out" links are usually called lateral links; the fore/aft link is usually called a longitudinal link.

 

[brian_]

The biggest challenge of using a strut suspension at the rear is that - completely unlike the original Jag IRS which is obviously designed to be low and flat across the top - it fundamentally requires two towers to accommodate the struts.

...- you could cut the turret off another car (like a Subaru - or almost any car with struts) and weld it into position...

This is true, but those towers (or turrets) would go through the existing structure, and into the interior.

I'm not clear on the rules yet but am working on the assumption I cant cut and move the chassis rails.

This could be one problem with fitting in strut towers, depending on frame rail spacing and strut spacing.

The other problem is what strut towers would run into in the interior, but I noticed these comments:

The transmission tunnel will be housing four modules, plus another two stacks of four on the front suspension. Plus four on the shelf above the IRS...

There us heaps and heaps of space for struts...

I've driven an XJ6 (I think a Series III, rather than this Series I), but I didn't look behind the rear seat and I don't recall the details of the trunk. It sounded at first like this body had a feature which was common in the 1960's, in which the trunk is behind base of the rear window, the rear seat back near the leading edge of the rear window, and in between them is a wasteland of awkward "parcel shelf" over the rear axle and suspension, within the passenger compartment volume. Looking at online images of bare XJ body shells, it seems more likely that these are references to the forward section of the trunk, where the floor is stepped up over the rear axle and suspension, ahead of the trunk opening and below the rear window.
(an old bodyshell image is attached to the end of this post)

In either case, that's a great battery location for some cars (as long as you package the modules safely), and if strut towers (for MacPherson struts or just springs and shocks of another IRS design) were to poke into there it wouldn't interfere with seating space or the main trunk space (although of course now battery and suspension are competing for space). There are just those structural issues to consider... struts outboard of those frame rails (visible in the bodyshell image) would require a very wide track.

While looking for information on the XJ body, I discovered that if you have buckets of money Jaguar Classic might still produce a complete new bodyshell for you (although only for the Series III, not earlier).

 

[brian_]

IMHO you do NOT want your batteries high up at strut level - the best place for all the heavy bits like batteries is down on the floor - as low as possible

Ideally, yes. Unfortunately the reality of a car body not designed for an under-floor battery box is that low locations are not likely available. The engine compartment and transmission tunnel go to the bottom of the car, and are planned locations; even the top of the engine area stack will be higher than that over-axle shelf. For the rest, there is the portion of the spare tire well behind the Tesla drive unit, but that's well behind the axle and so it is undesirable as well. Stacking them on the floor in the rear seat space might not be popular with the rear seat passengers...

The stock "saddlebag" fuel tank locations are not very low, either... and not shaped to accommodate big flat modules.

 

[brian_]

Fuel tank spaces

I believe a strut tower would fit. The fuel tanks sit above the wheels so without them leave plenty of height. The bulge locates under the filler cap but there is still space forward of that.

Just to be clear, with or without fuel tanks doesn't matter to struts, because the front of the tank is still well behind the axle line, and so behind even a large spring on a strut. The tanks are also behind the tires, and so mostly outboard of a strut location.

The tank locations look like good places for components such as chargers or coolant pumps, or even just storage bins. And one of the fill cap locations is just begging to become the charge port.

 

[brian_]

Spare tire well

In the rear, i am losing most of the spare wheel well to fit the motor might check space again though. The 4 module stack is 34cm high...

It appears that a 1969 XJ6 came with E70-15 tires, which are 26" or 660 mm in diameter. The spare tire well appears to be entirely behind the differential housing, so it must extend at least 26" or 660 mm past the diff. The front side of the Tesla motor will sit further forward that the back of the Jag diff, so the rear mount of the drive unit should end up about mid-way back in the spare tire well.

Does that leave enough space for modules (only stack of two due to height) in the spare well behind the motor? Is the spare well width (which looks like it is barely enough for the tire) enough for the length of a module? Both dimensions look tight, but especially the width, so I wouldn't be surprised if a module won't fit there.

This could be another space for various ancillary components (electronics, coolant pumps...), or just a storage cubby for the air compressor and tire puncture-repair stuff that one needs when one doesn't have a spare tire.

 

[brian_]

There is not enough space in the trunk for a module to fit width-wise, it is about 1" shy of being doable without cutting into the old muffler compartment (which is not the end of the world if this much fabrication work is already happening).

The mufflers flank the spare tire well, so I can see how cutting out one or both sides of the spare well would allow the use of the muffler space, but isn't there a structural ("frame") rail between them? The structure has to be somewhere, and forward around the axle it is in two rails essentially lined up with the sides of the spare tire well. Does the structure run outboard of the mufflers, or above them? The mufflers look like they might be under the fuel tanks.

 

[brian_]

View attachment 123741

Looking good
Custom-fabbed control arms and uprights... are you using Tesla hub-bearing-carrier units so that you can use (shortened) Tesla halfshafts (and brakes, and wheels)?

 

[brian_]

No, I am having driveshafts made for my application. Tesla hubs/brakes/wheels are too big and I needed standard Jag wheel fitment View attachment 123771 View attachment 123772 View attachment 123773

Nice.
So what hub/bearing units are you using? An old Jag wouldn't have come with a unit like that.

 

[brian_]

Indeed, thankfully the germans decided to use a stud pattern that can be made to work with little (to zero) modification

33411093371

7316577660657

I was going to ask "which Germans?", meaning "which German car does the hub belong to, which has a bolt pattern that works for your car?"... but from the part numbers it looks like a common BMW hub, used for multiple models.

I had guessed it was a sealed hub-bearing-carrier unit, but it appears to be a strange (which shouldn't surprise me from BMW) sealed bearing in a carrier plus a separate hub... which means that the stub of the drive axle presumably retains the hub in the bearing. The all-in-one units (including Tesla) are typically staked together, so while they can't be rebuilt, they don't need the axle shaft to hold them together. Not a problem, just an assembly factor to keep in mind.

The BMW bolt pattern is 5x120mm, and it looks like the Jags were 5x4.75". 4.75" is 120.65 mm; while that's very close to 120 mm, it's not the same - I wouldn't settle for that. I would expect there to be a hub-bearing assembly for an older GM model that would actually be 4.75", given that this was a very common GM bolt pattern and that GM has used unit bearing assemblies for decades, and while they are most common as non-driven hubs for the rear of front wheel drive cars, GM part #7470005 (from the front of a Chevrolet S-10 or 1979-1985 Buick Riviera, for instance) is a driven hub-bearing-carrier unit with 5x4.75" bolt pattern. 1984 and newer Corvette rears might work, too. There are other 5x4.75 vehicles, but many would not use these bearing assemblies, and I don't think Aston Martin or Lamborghini parts help much. At least they're European, which might be better than looking for parts for GM North American models in the UK.

Do the Jaguar and BMW centre bores match? I don't know exactly which Jaguar and which BMW are being mixed, but it looks like Jaguar used 73.8 mm bore while BMW used 72.6 mm. I don't know if a centering ring rigid enough to ensure concentricity (i.e. not plastic) is practical for that combination. The GM hubs won't match the Jaguar centre bore, either, but as long as the hub is smaller than the wheel a centering ring can be used.

It would be handy if the splines matched between the Tesla side and the hub, even though custom-length axle shafts are needed... but I have no idea what splines any of these bits have.