De Dion rear end design..

Shocks, Springs, Brakes, Frame, Body Work, etc

Moderator: Team

Brian P
Guru
Guru
Posts: 1611
Joined: Wed Jan 02, 2008 7:35 pm
Location: Toronto, Ontario, Canada

Re: De Dion rear end design..

Post by Brian P »

Don't forget the one BIG bad disadvantage of using de Dion suspension: the advertising folks can't shout out "LOOK! 4 wheel independent suspension!"
englertracing
Guru
Guru
Posts: 1547
Joined: Thu Sep 29, 2011 8:55 am
Location:

Re: De Dion rear end design..

Post by englertracing »

The testing was done in the 60s I believe.
Also regarding the full independent....
Usually the trophy trucks and truggys finish the Baja 1k before the irs buggys :D
lemons racer
New Member
New Member
Posts: 8
Joined: Sun Mar 27, 2011 10:07 pm
Location:

Re: De Dion rear end design..

Post by lemons racer »

The De Dion was used allot back in the 50s-60s mostly by european manufactures that could not afford to produce a full independent suspension which is MUCH more expensive to develop. The advantage of the De Dion is less unsprung weight compared to a live axle and that's about it.
The last car I worked on using this suspension was a Montaverdi Hi (forgive my spelling, this was back in the mid 80s)) built in 68, 1st of 2 built. 426 Hemi, ZF transaxle, the car looks something like a Bora. When under that car it was less than impressive although I was told when new it was clocked at 189 mph.
pdq67
Guru
Guru
Posts: 9841
Joined: Thu Mar 04, 2010 8:05 pm
Location:

Re: De Dion rear end design..

Post by pdq67 »

^^^^^^^^^^^^^^^^^^

"When under that car it was less than impressive although I was told when new it was clocked at 189 mph."

I wonder how fast my next door neighbor's 440/4-speed equipped Jensen interceptor is?

pdq67
Ratu
Expert
Expert
Posts: 740
Joined: Sun Oct 02, 2011 7:58 pm
Location:

Re: De Dion rear end design..

Post by Ratu »

Beware the assertion that IRS is truly independent for it is not. A bump at one wheel affects the opposite wheel. To avoid that completely the spring rate would need to be extraordinarily soft indeed, certainly very soft in roll, i.e. one wheel bump (theoretically offering no roll resistance at all!). Either that or the car itself would have to be extraordinarily heavy. A force delivered from one wheel and affecting the sprung mass necessarily means that there is an effect on the opposite side suspension. If there is an anti-roll bar then the loss of independence is increased even further. Beware the potentially misleading term "independent".

De Dion is interesting in that it allows lower unsprung mass than does a live axle. Apart from that it can be made to exploit most of the advantages of the beam axle. There is also one extra possibility. During cornering one may arrange for the wheels to camber inward (towards the apex). That is not something I have ever seen actually accomplished with IRS (although theoretically possible with a very elaborate system of links). Years ago the late Spen King delivered an interesting lecture about suspension where he mentioned this possibility. It could be achieved by splitting the de Dion tube into two portions wherein there was a rotational joint between the portions allowing for freedom for rotation of the tube portions along their major axis. Clearly the geometry of the locating arms would need to be correct, but with pencil, paper, compass, set-square and a little time it does not take much time to figure it out. You can even make a model with cardboard, using drawing pins for the pivots. Of course CAD is the most convenient way to experiment with this, still the old fashioned ways have a charm of their own and they reward the experimenter with a tactile, physical insight.

Later I discovered that the same effect as this can be obtained with a torsion beam axle so long as the axis of the torsion beam passes behind the centreline of the wheels. This very scheme was developed for the rear suspension of the Australian Ford Falcon as an alternative to IRS. It was a whole lot lighter and much cheaper to manufacture, but despite excellent potential, reliability and functionality it was dropped and the IRS was adopted instead due to a strong intervention from the marketing department...

On a related note, I recall the Donovan V-8 powered Alfa Romeo Alfetta raced by the Algie brothers in New Zealand, of all places. It featured a de Dion rear suspension just like the standard car. It was very competitive and in the wet it was unbeatable. The drivers mentioned the car was very forgiving in its handling and provided an excellent tactile feedback to the driver. Interestingly in the modern mid-engine era Ferrari built a F-1 car which could be readily swapped from double wishbone to de Dion and back again. Drivers who tested it ran identical lap times with double wishbone and de Dion. They stated that the de Dion felt better, giving more feedback and being more forgiving. In the end Ferrari stayed with double wishbone due to attributes of easier adjustability in the pits on race day, as well as for aerodynamic reasons.

Meanwhile over in Australia there have been people experimenting with beam axle front suspension for a road race car. I first heard about it early last year and have been asking around about it since. So far it is confirmed it was used on a Super Sedan with some success. I find this intriguing and have been keen to know more about what it is like and the reasoning behind its use.
Leftcoaster
Pro
Pro
Posts: 454
Joined: Sat Jun 04, 2011 5:46 pm
Location: Canada

Re: De Dion rear end design..

Post by Leftcoaster »

Ratu wrote: Mon Jan 22, 2018 1:24 pm De Dion is interesting in that it allows lower unsprung mass than does a live axle. Apart from that it can be made to exploit most of the advantages of the beam axle. There is also one extra possibility. During cornering one may arrange for the wheels to camber inward (towards the apex). That is not something I have ever seen actually accomplished with IRS (although theoretically possible with a very elaborate system of links). Years ago the late Spen King delivered an interesting lecture about suspension where he mentioned this possibility. It could be achieved by splitting the de Dion tube into two portions wherein there was a rotational joint between the portions allowing for freedom for rotation of the tube portions along their major axis. Clearly the geometry of the locating arms would need to be correct, but with pencil, paper, compass, set-square and a little time it does not take much time to figure it out. You can even make a model with cardboard, using drawing pins for the pivots. Of course CAD is the most convenient way to experiment with this, still the old fashioned ways have a charm of their own and they reward the experimenter with a tactile, physical insight.
The 1963 - 1977 Rover P6 employed a rear differential with integral disc brakes mounted to the chassis, and a De Dion suspension with trailing arms, a transverse beam with halves that slid and rotated within a sleeve as you describe, and fixed length half shafts which maintained track within their arc of travel

The geometry was successfully tweaked by club racers but the differential unit proved incapable of supporting the torque and horsepower outputs of developed alloy BOP V8 engines, and was most commonly replaced by Jaguar IRS units

To create a wide engine bay without intruding Mac Pherson strut style towers, an ingenious disc braked front suspension placed its shocks vertically and the springs horizontally; all exterior panels were bolted to the steel frame, and with alloy fenders, hood, trunk, and roof, the UK 3500S with V8 and auto trans weighed around 2850lbs
Brian P
Guru
Guru
Posts: 1611
Joined: Wed Jan 02, 2008 7:35 pm
Location: Toronto, Ontario, Canada

Re: De Dion rear end design..

Post by Brian P »

I suspect that modern tires are insensitive enough to camber that the bad side effects of having camber change fully compensated or over-compensated for body roll outweigh the good.

While it is theoretically possible for a twist-beam-axle rear suspension design to cause the tires to lean in when the car body leans out by having the twist-beam aft of the wheel center-line, in practice this is never done. Some twist-beam axles have the twist beam in line with the wheels (chrysler K-car, gm X-car, early VW Jetta Mk6 before they switched back to IRS, and some older Toyota and Suzuki small cars were like this) which gives "beam axle" geometry - the wheels stay straight up and down regardless of body roll - but the far more common design has the twist beam aft of the chassis-end pivots but in front of the axle centerline, which only partially compensates for the camber (depending on how forward the twist-beam is of the axles). Cars with twist-beams like this include VW Golf Mk1-2-3-4-7, Honda Fit, GM everything newer than the X-cars and A-cars, Ford Fiesta, Fiat everything that uses a twist-beam, Renault everything that uses a twist-beam, Citroen everything that uses a twist-beam, etc.

Large camber changes (on both sides!) just because you go over a one-wheel bump are just another way of upsetting the chassis.
Ratu
Expert
Expert
Posts: 740
Joined: Sun Oct 02, 2011 7:58 pm
Location:

Re: De Dion rear end design..

Post by Ratu »

Hmmmmm. Bumps are more transient than roll displacement....
Brian P
Guru
Guru
Posts: 1611
Joined: Wed Jan 02, 2008 7:35 pm
Location: Toronto, Ontario, Canada

Re: De Dion rear end design..

Post by Brian P »

A customer of mine builds the rear suspension module for the current Honda Civic and they have a complete module sitting in the lobby. It is a multilink IRS design with a main trailing link attached to the spindle, upper and lower lateral links just ahead of the wheel centerline which locate the spindle in camber, and a lateral link aft of the wheel centerline which is the toe link. The upper and lower lateral links are very close to the same length and are almost parallel ... I'd estimate that the instant center is about two or three track-widths on the other side of the other wheel (so perhaps 15 or 20 percent camber correction for body roll). The toe link aft of the wheels is much longer than the lateral links ahead of the wheels, which is a recipe for slight toe-in both above and below nominal ride height. In top view, all of the lateral links are parallel near as I can tell visually, which means fore-aft wheel movement due to bump impacts (the trailling link has a very soft bushing mount to the bodyshell) doesn't produce much if any motion in toe.

- Compensating for camber change as a result of body roll is a low priority with this design.
- Toe control in response to body roll is a high priority. (Toe-in in response to body roll = slight roll understeer and it puts the inside wheel, which wants less slip angle due to less lateral load, in a better situation where it can retain more lateral grip instead of being forced to slide sideways)
- Preserving the wheel alignment (minimal toe change) in response to a bump impact, while still cushioning the bump impact from being transmitted directly into the chassis, is a high priority.
- As with all trailing-arm-based designs, "anti-lift" in response to braking forces is an integral feature. Minimizes pitch during braking and keeps suspension travel (and the resulting toe changes) in a manageable range.

This suspension layout is pretty similar to a number of other late model designs and they're all copies of the Mk1 Ford Focus which as far as I can tell was the first one to use this design.

Can't do all that with a twist-beam, a de-Dion, a rigid rear axle, etc.
peejay
Guru
Guru
Posts: 1946
Joined: Tue Aug 10, 2010 9:16 pm
Location:

Re: De Dion rear end design..

Post by peejay »

Brian P wrote: Fri Apr 06, 2018 10:02 amThis suspension layout is pretty similar to a number of other late model designs and they're all copies of the Mk1 Ford Focus which as far as I can tell was the first one to use this design.
The Focus suspension looks similar to the one Mitsubishi used in some of their Lancers, and both in turn look very similar to the setup Honda started using in Civics in 1988, minus the odd swinging link in the front that necessitated a bizarre floating bushing.
Brian P
Guru
Guru
Posts: 1611
Joined: Wed Jan 02, 2008 7:35 pm
Location: Toronto, Ontario, Canada

Re: De Dion rear end design..

Post by Brian P »

Not knowing much about Mitsubishi prompted some digging, and the recent Lancer does indeed use a similar design. I know Mitsubishi used trailing arms in the early days of front wheel drive, and they may have gone through the twist-beam phase at some point ... I don't know when they started using the multilink design that seemingly everyone uses nowadays if they're not using a twist-beam.

Honda has gone through a few rear suspension designs, from MacPherson (gen 1-2) to twist-beam (gen 3) to the funky trailing arm with toe link mentioned above (gen 4-5-6) to a lower "H-arm" with bizarre non-parallel pivot angles and upper link (gen 7-8-9) and finally to the current design described above, which finally properly separates the NVH bushings from the "handling" bushings.
peejay
Guru
Guru
Posts: 1946
Joined: Tue Aug 10, 2010 9:16 pm
Location:

Re: De Dion rear end design..

Post by peejay »

Brian P wrote: Sat Apr 07, 2018 6:35 pm Not knowing much about Mitsubishi prompted some digging, and the recent Lancer does indeed use a similar design. I know Mitsubishi used trailing arms in the early days of front wheel drive, and they may have gone through the twist-beam phase at some point ... I don't know when they started using the multilink design that seemingly everyone uses nowadays if they're not using a twist-beam.
The Mitsubishi suspension I am thinking of was used in the Lancer Evolution I/II/III (1992-1996). This same chassis with a slight wheelbase adjustment was also used on the 1998(?)-2004 Volvo S40, as Volvo found it to be more expedient to buy a pre-existing small car than to design their own.

"Oh cool, I have a LanEvo", I thought, when I got my S40. Then I discovered that exactly zero suspension upgrade parts exist for these cars, since they are 20 years old and, since it's a Japan-only car, the majority of them have been scrapped, so no aftermarket exists either.

You can see the heritage in design, though, comparing it to later Evolutions.

Image
(my S40. Wheels are some takeoff Lancer OZ Rally wheels I bought for my RX-7, which has been down since 2015. Funnily enough doing a reverse image search on Google Images suggests "mitsubishi")

Image

I ended up replacing the S40 with an S60R. It's kind of like having an Evo except it is reliable, and I don't have to worry about it getting stolen when I to to the movies or something.
Brian P
Guru
Guru
Posts: 1611
Joined: Wed Jan 02, 2008 7:35 pm
Location: Toronto, Ontario, Canada

Re: De Dion rear end design..

Post by Brian P »

Neat car. And it reminds me of something related to the original topic of this thread!

The predecessor of the Volvo S40 can be considered the Volvo 340-series which itself was a successor to the DAF 66 ... and if I remember right, the Volvo S40 was built in DAF's historical plant ... Volvo bought DAF and inherited the plant and the vehicle design, the changeover to the Mitsubishi platform happened after the historical Volvo rear-drive designs became obsolete.

How is this relevant, you might ask?

Why, de Dion rear suspension, of course! Leaf springs and a dead-beam axle.

http://www.volvo300club.nl/berichten.ph ... u=techniek
grandsport51
Pro
Pro
Posts: 274
Joined: Sun Jan 14, 2007 11:47 am
Location:

Re: De Dion rear end design..

Post by grandsport51 »

And of course the famous (not) Chaparral 2-H
With the submerged cockpit and De Dion rear suspension.Image
You do not have the required permissions to view the files attached to this post.
LIGHT 'EM UP
clshore
Expert
Expert
Posts: 703
Joined: Sat Dec 12, 2009 4:48 pm
Location: Beverly Hills, FL

Re: De Dion rear end design..

Post by clshore »

THANK YOU!!
I've been searching for photos of the Chaparral 2H DeDion.
I had read a Jim Hall interview where he talked about the center pivot, but could never find any drawings or photos.
So I could only conjecture about the details.

That photo is great, and helps explain a lot about what Hall said.
Post Reply