Last week David Tracy and Jason Torchinsky showed us some pictures of the upcoming Slate EV, which is going to come equipped with a MacPherson strut front suspension and a De Dion setup in the rear. The Slate EV is meant to be very inexpensive, and in trying to keep the price low, we will likely see many examples of design decisions made in the interest of cost. The choice of suspensions is a good example of this. Let’s talk about it.
MacPherson front suspensions have adorned cheap cars for decades. They are easy to manufacture, cheap to assemble, and relatively easy to maintain and service. The low number of parts involved in their design helps keep cost low. Considering the fact that this design has been in use in cheap FWD cars for the past 30-40 years, you’d think they were just that: cheap, but not very good. You’d be wrong though.
A MacPherson strut is cheap to make but can also perform very well if designed properly. The BMW 3 Series has used a MacPherson strut since time immemorial and the Porche 911 has used one since 1989 when they switched away from the independent torsion bar system the car was originally introduced with. Nobody would say the suspension on the BMW 3 series and 911 don’t work well!
So, the choice of a MacPherson strut as the front suspension in the Slate EV seems like a logical move. Cheap to make but can work very well if properly designed. We’ll reserve judgment on the second part of that sentence once we get a chance to drive it.
But what about that rear suspension? Why choose a De Dion? My guess is that most Autopians have never heard of it. In fact, the chances of seeing one under any car built in the last 130 years are pretty slim. Very few OEM’s used it. Let’s have a little history lesson to bring us all up to speed on what a De Dion axle is.
What Is A De Dion Suspension?
The De Dion axle takes its name from Jules-Albert de Dion who founded the French automobile company De Dion-Bouton in 1883. And while the axle has the name of the company founder, it was actually invented around 1894 by co-founder Georges Bouton’s brother-in-law Charles Trépardoux. At that time, De Dion-Bouton were building steam powered cars which had the engine and driveline mounted at the rear of the frame. Since the engine was mounted in the rear, there was no room for a driveshaft between the engine and the differential. Many cars and trucks of that era with similar powertrain layouts used chain drive to connect the powertrain to the wheels, but perhaps Trépardoux felt this was not adequate for their cars.
Whatever his reasons, he designed a suspension that solved the problem of getting the power out to the wheels without the need for chains or a traditional driveshaft.
The design consists of a left and right hub, rigidly connected to each other by a large metal tube, a center mounted differential, and two short halfshafts connecting the left and right hubs to the differential.
Patent drawing via: De Dion Suspension History – Historic Vehicles
Since the hubs were rigidly connected to each other via a connecting tube, the whole assembly acts very much like a solid live axle. The wheels are kept at a constant camber angle and maintain their relationship to the ground. Of course, the axle also has the same issue as a solid axle does where when one wheel hits a bump, it causes the other wheel to tilt. It’s not an independent suspension
However, the fact that the differential is not part of the axle assembly means you need two halfshafts to transfer the power from the differential out to the wheels. In this way it is similar to an independent design.
In the image two above, the axle is shown mounted on leaf springs, but the beauty of the design is that it can be mounted just like any solid axle can: with leaf springs, coil springs, 4 links, Panhard rods, Watts links, etc. Any way you can think of to mount a solid axle can also be used to mount a De Dion. But no matter how you mount it, one difference always remains: While a solid axle includes the differential as part of the suspension, the De Dion takes the differential and mounts it to the frame, and that is one of the great advantages of this design and why it is such a winner for an EV.
The reason this is so important is that differentials are heavy, and when they are part of the suspension, they have to move up and down with it as well. It’s called un-sprung mass.
Un-Sprung Mass
When a solid axle encounters a bump in the road, the wheels and suspension have to move up and over the bump. When the differential is part of the suspension, there is a lot of mass that has to be forced up and over the bump. All that mass then has to be slowed down by the springs and dampers and pushed back down to let the tires maintain contact with the road.
In the old days of luxo-barge American cars, springs were quite soft to give a plush ride. This meant there wasn’t a lot of force coming from them to help push all that mass back down after a bump. The result was that the axles would “dance” and tramp, losing traction in braking, accelerating, and cornering on rough roads. Having the differential disconnected from the axle means far less un-sprung mass and a much smoother and more controlled ride as well as better handling since the tires could better maintain contact with the road. It’s interesting that what we now see as the primary advantage of the De Dion design wasn’t even an issue in the early days. I can pretty much guarantee that un-sprung mass and tire traction over rough roads did not even enter Trépardoux’s mind when he invented it.
Regardless of the reasons why it was invented, the design proved quite useful and saw a lot of use up through the 1920’s. It fell out of favor for a few decades but was then used extensively by Alfa-Romeo in the Alfeta, GT, GTV, GT6, Alfa 6, 75/Milano and others.
It was also in use on all Aston-Martin cars from 1967-1989. It is now used on several electric trucks as well as the Mercedes-Benz EQG.
Why Is This So Good for EV’s?
So, why does this design make so much sense for an EV? If you watched the video David Tracy and I made about the Ford F150, Raptor and Lightning (see above) you will remember how the package of the battery in the Lightning meant there couldn’t be a traditional driveshaft running down the center of the vehicle. All that space was needed for the battery and it forced the electric drive motor to be packaged in between the rear wheels right where the live axle needed to be. Ford’s answer to this problem was to ditch the live axle and design an independent rear suspension using two trailing arms.
They could have used a De Dion but decided against it, I believe, for package reasons. Look again at the early patent drawing above and the image of the Alfa suspension. In the patent drawing, the tube runs underneath the differential. However, in later cars where ground clearance is at a minimum below the diff, there isn’t room for the tube. The solution is to place the tube behind the differential. In an F-150 Lightning, however, this is where the spare tire is and making space for a De Dion tube would have meant a longer rear overhang. Since Ford’s desire was to reuse as mush sheet metal as possible, a longer rear overhang was out of the question. It meant the De Dion would not have worked for them.
So, the short of it is that, in an EV, a De Dion tube works so well because it’s simple, rugged, offers good articulation, keeps unsprung mass down (which would be a big problem given that, unlike on an ICE, a solid rear axle “eBeam” includes the electric motor), and still allows for rear motor packaging, which is necessary, because the battery takes up all the space in the middle of the vehicle. There’s a reason why well-known EV-conversion specialist EV West sells these in bulk.
Specifically, EV West sells the De Dion tube axle that came in the rather range Ford Ranger EV. Here’s a look at a stack of Ranger De Dions:
As EV West notes, this simple suspension makes mounting a rear drive unit in a gas vehicle that already has a solid axle easy, making for a clean conversion:
These are De Dion axles from the original Ford Ranger EV that was produced in 1998-2001. These are a great path to mounting a Tesla motor in the rear of your pickup or van project to free up the engine bay for the battery pack. These will also maintain the full efficiency of the Tesla drive by allowing a transverse mounting. Ford engineered and manufactured. These are limited to quantity on hand.
The Slate De Dion
Slate, however didn’t have the Ford F-150 Lightning’s packaging problem. The new company mounted its (optional) spare tire on the rear tailgate so there is nothing behind the differential. There is plenty space to package a De Dion tube. That doesn’t mean they didn’t consider other option, though. Based on a conversation David Tracy had with Eric Keipper, Slate’s head of Engineering, we know they looked at numerous other suspension architectures, including independent suspension and beam axles.
Beam axles for EV’s, sometimes called E-Beam’s, are under development by several companies and are designed to replace the live axles currently found in almost all pickup trucks:
These axles are designed to swap in place of a live axle with minimal modification to the truck’s design and can be mounted using any of the usual live axle spring and link systems used today. But while a traditional live axle contains the differential as part of the axle, an E-beam includes the entire electric motor as well. All that mass is now moving up and down with the wheels over bumps in the road. As we learned earlier, a live axle isn’t great from a ride and handling perspective due to high unsprung mass. Now imagine adding the weight of the electric motor to it. Also, the majority of e-beam type axles under development are designed for larger pickups since that is where the market is. They are not meant for a low-cost small vehicle.
A beam axle was therefore ruled out early in the process since no designs were available for the class of vehicle Slate was developing. This left the De Dion as the natural choice.
Slate’s implementation of this design can be seen here:
As I mentioned earlier, De Dion axles can be mounted in any of the many ways live axles can. Slate chose to use a classic 4-link design with a Panhard rod. Cheap, easy to make, and easy to assemble. The two lower links are clearly visible above as well as the Panhard rod. The upper links are a bit buried in this image, but you see the bolt holding the left side link to the De Dion tube. It’s a very good solution and should work well for them. The only thing that does stand out to me is that I would have preferred to see the Panhard rod be perfectly horizontal. The angle shown above will cause some side-to-side motion of the rear end over bumps and undulations in the road which could easily be prevented with a more horizontal installation angle. Perhaps this could be a future ride and handling upgrade (Note: This is a preproduction unit).
All things considered though, the design looks quite decent. A De Dion makes perfect sense for a low-cost truck-like EV. It’s relatively cheap, easy to make and due to lower un-sprung mass, will work better than a live axle or e-beam.
Kinematically, how is a torsion-bar 911’s front suspension not a McPherson strut type? Or are you considering it definitionally to have a coil spring mounted quasi-concentrically with the damper? I don’t see how the spring type or location (see e.g. not only 911s but also Mustangs and other Fords) matters.
As motors become more of an off the shelf commodity, I’m expecting EVs to have one per wheel via CV half shafts and double A arm, coil overs at all corners. Then some fancy controllers for torque vectoring.
I’m must be one of the few other nerds who got excited reading it had a de Dion out back! I’ve thought they are really cool since I learned about them on old (like really old, 100 year old) Indy cars like Millers and the like.
I wonder if it would make lowering the truck more difficult? I was picturing nabbing one of these in the future with the fastback roof added and lowering it to make it look like a hot hatch lol.
Damn, Slow Joe beat me to it by a few seconds
https://youtu.be/VLUD6RwO-AA?si=5ha97TRjcegLehXz&t=45
The Rover P6 also used a De Dion rear axle with the added complication of inboard brakes to further reduce unsprung weight.
I half expected them to use Chapman struts, since they offer the same advantages as McPherson struts
DeDion is a good example of tech that snobs dismiss when the disadvantages against their preferred “better” tech are invisible to most users in most cases and rarely make a big difference if direct comparisons are available. Some things are mostly only better on-paper. For this type of vehicle and its purpose, the cost advantage of DeDion is far more important and I would suggest this is even a more upscale design than I would expect (namely, a live axle, not that live axles are bad, but the advantages/disadvantages don’t stack up as well for this small truck of modest payload not intended for serious offroad use).
One of the reasons often cited for the Alfa Romeo Montreal’s failure was the “outdated” DeDion vs independent rear ends of contemporaries when those independent designs were not as good. As the Alfa and its contemporaries are all outdated classics now, people have more reasonably re-evaluated them on their merits rather than spec sheets and the Montreal rates highly for handling.
“the chances of seeing one (De Dion axle) under any car built in the last 130 years are pretty slim”
Wouldn’t the rear axle of the AWD Mazda 3, CX-30 and CX-50 be considered De Dion? The differential is bolted to the subframe with CV shafts going out to the wheel hubs which are connected by a torsion beam rear axle?
No it isn’t. The Mazda system, like all torsion beams, rely on the twisting and flexing of the center part of the beam assembly. If there is no flex in this part of the system, then the whole design doesn’t work. A De Dion, on the other hand, does not rely on flex. In fact, the axle could be perfectly rigid, and it would still work.
Worth mentioning, the (now-Chevrolet) Brightdrop van also uses a De Dion rear axle (in AWD configuration), so another point for usage in EV applications.
I have known of their existence for many years because Caterham used to use them until just a few years ago. I never bothered to look at what they were or how they worked though, and was planning to do so after seeing Slate plans to use them, so I was very glad to see this here saving me the effort!
Was going to say the same, seemed like I was aware of them in the context of a the super7.
Same. First learned of their existence from this cool book “Roadster”, about a fairly mechanically un-inclined guy who decides to build a Seven himself from the kit.
The Slate just got me talking about my dream minitruck conversion with a Tesla drive unit in a DeDion setup with the original leaf springs, and here you’re telling me I can get original Ranger EV DeDion rear-ends from EV West? With a Toyota bolt circle?
Umm, my bank account’s not gonna like this…
It’s funny to see two defunct automobile companies names used in the same rear suspension. De Dion and Panhard
Earle S. MacPherson is also defunct 🙂
As is Colin Chapman, who tried to rename the MacPherson strut as the “Chapman Strut” when there was a halfshaft added to it to transmit power.
How is this different from a really really stuff anti sway bar?
A stiff anti-sway bar would not allow one wheel to move up without the other moving in the same direction. Both a live axle and a De Dion allow free movement up and down as well as in rotation (as viewed from the back). An anti-sway bar tries to stop this rotation from happening.
A sway bar also doesn’t prevent camber squat on an independent suspension system at full compression. A Di Dion would, just like a solid axle.
That all makes sense to me. There’s been a lot of negativity about this little trucklet, but I hope they pull it off. I’d love to buy one in a few years.
You must be in an adjacent universe. All I am hearing is that it’s the greatest thing since things were a thing.
Plenty of it out there. Not here, but out there. Don’t go out there. It’s not very nice. Stay here where we love trucklets.
I dunno man, I saw a ton of negativity about it here last week. I’m into though.
I pretty much confine myself to here and http://www.opposite-lock.com
The commentariat at both places, have been none too impressed.
Been saving up all my Amazon gift cards for this truck.
Free same-day delivery if you have Prime!
Now if I could just get them to unlock my “temporarily locked for your safety” account…
So outside of unsprung weight for E-axle applications, what benefit does a Di Dion have over a solid rear end? Like what prompted Alfa Romeo to adopt it so heavily after it initially fell out of favor?
The E-axle sprung weight argument makes total sense, and I completely get the benefits (especially in Slate’s case where a Di Dion setup opens up a lot of different EV transaxle options while retaining the same suspension), but for typical ICE uses, what advantages does it have over a solid rear axle?
It seems like it wouldn’t have full articulation ability that a four link solid rear end would have, while at the same time constraining more of the suspension travel than an independent suspension, possibly binding up those CVs if you push it too hard.
Generally speaking, a De Dion can have the same travel as a solid axle. There is no reason why it would need to be less unless your talking about serious rock crawlers that have extreme axle travels. The limiting factor becomes the inner CV joints because they have to plunge in and out as the suspension travels and there are limits to how much they can do this.
The advantage of a De Dion over a solid axle is the lower unsprung mass and the improved ride and handling this gives. That’s the reason for doing it and why Alfa Romeo did it, without incurring the cost of a fully independent system.
I’ve owned both Alfettas and a GTV, and I remember the handling was great (I’ve driven the GTV in anger many times), much better than some cars with more modern suspension. I was always very impressed by how Alfa managed to do that with a DeDion rear (I guess the transaxle and inboard brakes helped a bit with unsprung weight and f/r balance)
More than a bit – that was the whole point of both of those features. Live axles actually have advantages – they keep the tire contact patch flat on the ground since there is no camber change, and no toe changes with suspension travel either. But on a relatively light car, the unsprung weight becomes a real problem.
-Fellow former GTV6 owner
And avoids the downsides of cheap independent rear suspensions (camber and toe changes with suspension travel). IRS can be good or cheap, but not both. Live axles are actually pretty good, other than the unsprung weight issue, which the deDion arrangement does a good job of mitigating.
Can’t you use universal joints with a De Dion instead of those finicky CV joints?
My favorite example is in the 1969 Chaparral 2H Can-Am car. I bet there are no CV joints there!
John Surtees was not impressed however.
I don’t know what the suspension design on those Can-am cars was but I would guess they are similar to the C2/C3 Corvettes and 60’s-80’s Jaguars where the halfshafts were also the upper arms of the suspension. No plunge needed in those designs. IN those cases, yes, you can use universal joints.
Also, the Triumph TR6 used universal joints but then had a sliding spline in the middle of each shaft to allow the length to change.
A Di Dion can eliminate that 90 degree gear set in solid axles which help efficiency. I mean I think you’re on to the reasons why it isn’t common. Independent is better from a ride handling standpoint and solid live axle is better from a durability standpoint. An EV needs to avoid the 90 degree gear set (hypoid) at all costs in the name of efficiency so that’s 1 more positive in the decision matrix.
Rear transaxels
Yes, thank you for mentioning this. Solid axles are not possible with a transaxle.
Aha! That makes perfect sense, thank you. I assume then that a Di Dion kept the sprung weight low while being cheaper than an independent rear end?
Correct
As Arch Duke Maxyenk posts: Rear transaxles prevent using a solid axle.
Beyond that, on a smooth surface a De Dion rear suspension always keeps the wheels vertical to the road – no matter how much the car leans. For this reason they were once considered superior to an independent suspension for road racing.
There is actually another advantage to a De Dion which I didn’t mention in the post. Solid axles are limited to zero degrees camber angle. A De Dion on the other hand can be designed to have some camber angle built into the axle. You could design it with -1 or-2 degrees of camber like you would an independent suspension. Like a solid axle, the camber angle will always be the same, but unlike a solid axle, it doesn’t HAVE to be zero.
Front solid axles can have camber designed into them since they already need an axle joint for steering, but yes, point taken, especially for a rear end.
My YJ solid front axle has a tiny amount of camber that can be adjusted with the steering knuckle bolts.
Well the NASCAR guys put a little negative camber into their solid axles. But I also get the impression that axles are considered a wear item there.
I believe they did that by making the splines on the inner end of the axle shafts barrel shaped and bending the axle housing. Lots of wear and tear on the splines, but as you said, they are a wear-out item.
A couple things I’ve been wondering about with the De Dion axles since the electric G-wagon reveal is do each of the axle shafts need to be able to lengthen/shorten as the suspension cycles? Maybe only for a longer travel suspension? And do the axle shafts limit articulation? Seems like for a suspension with low amounts of travel, there are would be no issues, but for applications with longer travel or articulation, they might bind and such.
Yes, the axles need to be able to change length. This is usually accomplished in the inner CV joint with a Tripod design or similar which allows for plunge to happen. The shaft doesn’t actually change length, but the location of the inner joint is able to move inboard and outboard as needed. There is of course a limit to how much this joint can move and that will limit how far the suspension can move up and down. It’s not normally an issue for most vehicles, but it probably wouldn’t work for a desert race truck, for instance.
Interesting, thanks. Sounds like something similar in concept to a slip yoke for a fixed length driveshaft, vs. a driveshaft that changes length then.
Exactly. The problem with slip yokes is that they tend to bind under torque. A good example of this are the halfshafts in the Triumph TR6. Under acceleration, the splines in these shafts would bind and prevent the suspension from moving. It made the ride much harsher under acceleration than otherwise. I used to own one when I lived in England and had to regularly grease these splines. Didn’t help much though!
Jeep had an issue with the rear propshaft slip yoke on the WJ body Grand Cherokee. The car would develop a “stop bump” characteristic wherein the driver would experience a bump feeling (it felt like somebody tapped your rear bumper) just after stopping. The cause was the slip yoke sticking, binding and then releasing just as the vehicle stopped. The TSB fix was a nickel-plated slip yoke, which was apparently expensive, but it did the trick.
In addition to Hubert’s answer (that tripod extension is key to some suspension work being possible without a ton more disassembly), there are aftermarket CVs that do have extending shafts for lifted applications.
I haven’t used them, but the concept makes sense.
At some point, for offroading/lifting at least, seems like the complexity of the CV shafts isn’t worthwhile vs just dealing with the downsides of a live solid axle, but that’s certainly more of a niche application.
If you’re off-roading, you certainly don’t care about ride and unsprung mass. And if you’re lifting, you DEFINTELY don’t care about ride and unsprung mass!
De Dion axles with portal gear hubs are a pretty cool way to get lots of clearance, even with massive differentials, and relatively little unsprung weight.
Of course if I were designing the optimal EV rock crawler, I would try trailing and leading arms with hub motors and no axles at all.
Actually, for a rock crawler, why not this?
Mmmm. For rock crawling, hub motors are not the answer. Let’s say you are teetering on two diagonally opposite wheels. Hub motors would only send 1/4 the total vehicle torque to each of the two wheels that have traction. If you had dual motors (one front and one rear) with locking differentials, each motor could send 1/2 the total vehicle torque to the two wheels that have traction. Double what hub motors can do.
Rover went another way with the P6. The driven axle shafts were fixed length, as plunging CV joints were expensive at the time, and a sliding spline joint was put in the middle of the De Dion tube.
I remember seeing the De Dion suspension in the flesh for the first time on early SxSs (the utility ones with a bed, before they made sports ones). It was pretty neat to watch it work.
MacPherson strut assemblies are available for dirt cheap from a ton of OEM and aftermarket suppliers, built to whatever specs you want, if you order enough of them
For giggles, I looked up an entire replacement front strut assembly (strut/spring/top bearing) for a 1993 Toyota Corolla on Rock Auto. Pricing starts at under $50 retail! Imagine what the wholesale price would be at quantity of 10,000 pieces!
I replaced all 4 strut assemblies on my ’99 Corolla for $176.99 plus tax and shipping last summer.
In the UK the Rover P6 (2000 and 3500) is fondly remembered for its DeDion rear axle. It was generally considered a well riding and sound handling car of its day, hence its popularity with the police.
I’ve read that these Slates have a pretty decent payload capacity for the size of vehicle they are, but it looks like there’s not much rear suspension travel.
Does this mean that the rear is sprung very stiff? Compare to, say, a 1990’s Ford Ranger and it looks like the Ranger has a lot more travel in the rear.
I don’t know much about the suspension travel but in general, more payload means less suspension travel and a harder ride because the springs have to be stiffer unless you go with fancy spring designs like variable rate but then that adds cost.
They could have very progressive rear springs, super soft with no load, super stiff for the second half of the travel. That would have been easier with leaf springs, but an add-on air bag would be a good solution. Nice and progressive.
Payload I have seen ia like 1500lbs is all, so for the size of truck it’s pretty decent, but it’s not a ton.
It is, however, 3/4’s of a ton. (badum-tss)
Now I’m the type of spring who will never settle down
Where VCs have some cash, well, you know that I’m around
I wine ’em and I dine ’em cause to me they’re all the same
I cut some corners here and there, the parts don’t have a name.
They call me The Wanderer; yeah, the Wanderer; I bounce round and round and round and round
–(de)Dion