Suspension Thread.

[Steve Thomas]

convert the X1/9 to MR2 uprights, realize all of those benefits and improve the X1/9s suspension geometry for performance applications without the effort required to convert to double A-arm.

This is the most sensible suspension upgrade suggestion I have heard. Take it a small step further and include the whole strut and strut top and benefit from more choice in shockies too?

 

[Steve Hoelscher]

Absolutly - and they are cheap to mass produce. Ally is a very nice production material but do not try and do ally suspension parts at home children!

I used to do consulting work for major auto manufacturers. In that capacity I would meet with a project team to review my report. There were always engineers from manufacturing, design, NVH, Ride & Handling, etc... The manufacturing guys, who were also the 'cost' guys, could overrule everybody. It was always interesting to review the design priority standards matrix and try to interpret that to design criteria.

Given that, if steel were a significantly cheaper option for manufacturing, that's what those parts would be made of.

 

[Steve Hoelscher]

I also have one of these in the shed. My stock X never bit me but the stock MR2 has. And it is much twitchier in poor weather and on broken road. Much more tiring to drive. I commuted in both, couple of hours a day. I like the MR2, I love the X!

The wider sportscar world doesn't appreciate the sophistication of the X1/9's chassis. Its a strut suspesion and the car is underpowered by the standards of the casual observer. The chassis is remarkably good and responds very well to performance upgrades. Consider that a basically stock X1/9 chassis, with a simple spring/shock/wheel/tire upgrade comfortably accomodates a K20 making 3X the original factory rating of 75hp and the resulting performance/handling is on par with modern supercars. Yes, the chassis is that good.

When I started building the MR2 my evaluation of the chassis architecture and suspension geometry showed that the chassis is simply a stretched/widened X1/9 with Corrola rear suspension. Toyota got the front end correct as it is a virtual copy of the X. The rear is where they erred. They used the basic Corrola front suspension layout (packaged with the Corrola 4AGE engine and C52 transaxle) as a cost savings measure but failed to understand the X's rear geometry. The X's rear roll center is below the front and the bump steer is very consistent. Like most front engine rear drive chassis the AW11's rear roll center is above the fronts. This makes the car more prone to rotation on corner entry, and coupled with the rear weight bias and bump steer issue, considerably less stable and predictable than the X.

 

[Steve Hoelscher]

This is the most sensible suspension upgrade suggestion I have heard. Take it a small step further and include the whole strut and strut top and benefit from more choice in shockies too?

I have made this suggestion before. Yes, the change to AW11 or SW20 (second gen MR2) uprights allows for an easy converstion to a more commonly available strut.

If I were building a very high performance X1/9 (and not doing a full tube chassis), I would basically duplicate the suspension I did for the MR2 including the MR2 struts (although my MR2 uses a very custom unit given its requirements).

 

[MikeHynes]

Here are some pics I gathered. These may be parts an Italian hill climb team makes? Looks like a lightweight upright.

 

[Steve Hoelscher]

Here are some pics I gathered. These may be parts an Italian hill climb team makes? Looks like a lightweight upright.
View attachment 78007View attachment 78008View attachment 78009View attachment 78010View attachment 78011

Custom uprights are pretty common on racecars. Those, obviously billet aluminum, are setup to accomodate a racing type bearing and hub flange with a center lock, pin-drive, wheel arrangement.

 

[Rupunzell]

From a previous discussion on dynamic roll centers..

"What is often not appreciated, traction or mechanical grip capability of the tire does depends on the dynamic roll center, it is where force is effectively applied when cornering. One way of visualizing this, take a broom and put one hand on the top of the broom stick. Put the other hand about center of the broom, then push the broom with that hand and note the amount of force needed to move the sweeper bristles of this broom. Repeat the same with the hand near the sweeper bristles of the broom, it will take more force to move the broom. This is a gross illustration of where the force applied makes a difference in traction-mechanical grip. If the dynamic roll center moves below road or ground level, the force on that tire will tend to cause that tire side wall to roll reducing the tire's ability to produce traction.

The dynamics involved is a LOT more complex than lowering the overall center of gravity, there is suspension geometry, dynamic roll centers that move and LOTs more. Dynamic roll center can be used to tune the mechanical grip of the suspension along with a long list of other various other chassis-suspension-wheel/tire combos that are interactive with no specific absolute for a given chassis-suspension and driver preference-style.."

Scuderia Ferric Oxide

There are low cost threaded collar kits known as "LeMons lowering specials" found on eBay for not a lot of $. When these first appeared en-mass on LeMons cars, the Lemons folks went looking to discover they were very much low-$.. They work ok, sort of ok. What is often not appreciated, traction...

xwebforums.com

~Essentially altering the roll center is a chassis tuning/setup tool. This is why many WRC cars have movable suspension points, to allow this adjustment and others that are needed based on the challenges and needs of a given rally stage. It comes down to suspension/chassis geometry and not just "double A-arms" "MacPherson-Chapman strut" or many others..

While on this topic of strut suspension, Mazda Miata was heavily based on and tried to copy the Lotus Elan.. which has all "strut" suspension. MacPherson (FORD patent) in the front Chapman struts in the rear (or Colin moved the MacPherson idea to the rear). It remains one of the best performing front engine chassis/suspension from that era and in many ways to the day.. Often not appreciated or understood, the dynamic roll center of the Lotus Elan is about 1" above the road.. Chapman understood this business of dynamic roll centers good.
Or why simply lowering the car, wider wheels, en all that can produce horrid suspension/chassis behavior.

While on the topic of front engine two seaters, the Honda S2000 is a FAR better overall design than the Mazda Miata. Honda S2000 is likely one of the very best front engine two seaters from that era.

Notable, engineering and similar knowledgeable car folks admire the exxe chassis suspension. From Colin Chapman, Dallara, the engineers that created the first reproduction of the FORD GT-40 at FORD and many others.. Yet, in the populous sports car world the exxe remains a greatly under appreciated design to this day. It also illustrated the divide between understanding good design -vs- desirability due to marketing.
~Or why one of the reasons the exxe was so often copied by so many over the years..

It is the sum of all parts involved that produces the overall net results..
Bernice

 

[Rupunzell]

Note the slip angles of the tires on Steve Hoelscher's MR2 in this corner, yet the tire patch remains essentially flat.
This is how high grip from the tires is achieved.

LeMons rotary X corner apex, note the flat tire contact patch front & rear, inside& outside tires. Tire slip angle slightly visible.
Open diff, no wheel spin or traction issues at corner exit. It is much about chassis/suspension set up. Rotary powered X has just over
200 whp on tap. Open diff cars are easier to drive.

Bernice

 

[Rupunzell]

and remember - aluminium always fatigues - always. Every alloy, every application. It is a fundamental aspect of aluminium. Then we have lack of stiffness. Ally is one third the stiffness of steel. In general terms, redesigning a steel part in ally gives a 50% weight reduction although ally is one third the density of steel. This is due to beneficial geometric effects - the ally part must be larger due to low material stiffness and as you increase size, your geometric stiffness rises [non-linear]. The product of material and geometric stiffness usually allows 50% weight reduction. But - ally is not good at temperature and the stronger alloys suffer more. The summary is that if you want to use ally in highly stressed and or hot areas, the design is non trivial and the new part will look very different from the original steel part. I have not looked, but I suspect that advice for improving the various production cars that come will ally arms is to replace them with steel to improve stiffness

Ally suspension arms have been used as suspension arms for a long time.

Saab when presented with the GM (ick) chassis ditched the OEM GM suspension bits then completely re-designed the suspension for the NG900 with ally suspension arms in front, circa early 1990's. Replacement is a non-issue as they are replaced as complete units bushings and all and easily-readily available to this day.. even when the NG900 went out of production decades ago..

As for ally as a structural material, like any material it comes down to how it is used and the specific application and a very long list of other factors involved... including the overall cost of applying any given material to be used..

Bernice

 

[Kendall]

The X's rear roll center is below the front and the bump steer is very consistent. Like most front engine rear drive chassis the AW11's rear roll center is above the fronts. This makes the car more prone to rotation on corner entry, and coupled with the rear weight bias and bump steer issue, considerably less stable and predictable than the X.

Please let me know if I'm understanding this correctly: the MR2 tends to rotate more into a turn because the higher roll center of the rear applies a greater force than the front. Essentially it's like having a longer lever to apply torque from away from the point of contact with the ground. Is that a correct interpretation? What makes the roll center of the X rear so low? This is a function of suspension geometry?

 

[Steve Thomas]

Ally suspension arms have been used as suspension arms for a long time

They have. But when it comes to modifying suspension by enthusistic amatuers without the full support of an OEM or even pro race engineering team - aluminium is to be avoided.

 

[Steve Hoelscher]

Please let me know if I'm understanding this correctly: the MR2 tends to rotate more into a turn because the higher roll center of the rear applies a greater force than the front. Essentially it's like having a longer lever to apply torque from away from the point of contact with the ground. Is that a correct interpretation? What makes the roll center of the X rear so low? This is a function of suspension geometry?

So we have now arrived at one of the most imporant aspects of how well a car turns and the chassis' behavior as it turns. Books have been written on this subject so I will try to simplify as much as practical.

A few specifics to answer some of your questions: The roll center height is a direct function of suspension geometry. The front and rear suspension each have their own roll centers. The line drawn through the front and rear roll centers is called the roll axis. The roll axis is the centerline about which the car's chassis rotates with body roll.

Bernice spoke above about the broomstick and I will use that example. The base of the broom would be the roll center. The point about which the body rolls when cornering. But what causes body roll? The lateral g forces are concentrated at the car's center of gravity and each end of the car has its own C of G. So the lateral g pushes on the chassis at the point of the C of G. The distance between the C of G and the roll center is called a "moment arm" (and is our broomstick). So the C of G uses the Momement Arm as a lever to roll the body. The more lateral g applied the more the car rolls. Also, the longer the moment arm (broom stick) the more leverage the lateral g has over the springs/swaybars to roll the body. The photo below is a double a-arm but the principle is the same.

The roll axis for the X1/9 is inclined toward the front, as it should be for a rear weight bias car. Using the Porsche as a similar example although somewhat exaggerated:

When the body rolls on the roll axis the end with the lower roll center will have slighly more traction due to slightly less weight transfer, but, most importantly, will react slower to transition (change of direction) because the longer moment arm will take longer to move for a given input of lateral force. This is because of the distance the body must move laterally due to the greater length of an arch as it moves further away from its centerpoint (the principle of Pi).

With consideration of the above: If the AW11's (MR2) rear roll center is higher than the front, the moment arm will be shorter in the rear. The effect then is the rear will move laterally quicker than the front as the car turns into the corner. And that quicker motion gets the rear end's increased mass moving much faster. And given the slightly reduced traction due to the increased weight transfer from the higher rear roll center the inertia created can easily overcome the available traction. Thus you have a recipe for a very nervous chassis.

The X1/9 rear roll center, as in the illustration above, is below the front so the turn-in is smooth and predictable, while the MR2 is much more demanding to drive. The MR2's bump steer complicates the matter while the X1/9 has very little rear bump steer.

 

[Steve Hoelscher]

One aspect of chassis dynamics that I failled to mention is a chassis' dynamic roll center. If you look at the illustration you can see the roll center height is primarily a function of the lower control arm angle. Therefore, as the control arms move the roll center does as well. That means that the roll center "migrates" as the chassis rolls. As a result minimizing body roll minimizes dynamic roll center movement.

Given the above, there are two primary reasons to minimize body roll in a strut chassis: 1) to reduce dynamic loss of camber, and 2) to reduce dynamic roll center migration.

The X1/9 has fairly long rear lower control arms when compared to many other designs. This is one of the many factors that result is the excellent handling manners of the X1/9.

 

[Rupunzell]

One aspect of chassis dynamics that I failled to mention is a chassis' dynamic roll center. If you look at the illustration you can see the roll center height is primarily a function of the lower control arm angle. Therefore, as the control arms move the roll center does as well. That means that the roll center "migrates" as the chassis rolls. As a result minimizing body roll minimizes dynamic roll center movement.

Given the above, there are two primary reasons to minimize body roll in a strut chassis: 1) to reduce dynamic loss of camber, and 2) to reduce dynamic roll center migration.

The X1/9 has fairly long rear lower control arms when compared to many other designs. This is one of the many factors that result is the excellent handling manners of the X1/9.

Position of that lower control arm is affected any time the static height of the exxe chassis is altered (often lowered). It is why it is un-wise to simply "lower" the exxe for looks or similar without fully understanding how changes in ride height impacts suspension geometry..

IMO, too many exxe owners lower their exxe for fashion/looks with no consideration to how the suspension and chassis dynamics will be impacted. Similar applies to wheels/tire choices.

Take a survey of any road/track vehicle with a known and proven GOOD suspension, this survey often reveals long_ish control arms to aid in controlling suspension geometry. The advantage of the long lower control arms on the exxe is often un-appreciated, yet one of the key factors as to why the exxe has a good rear suspension..

Bernice

 

[Steve Hoelscher]

Bernice is quite right. Some lowering is good, but that point is quickly surpassed.

What is not obvious is, due to the geometry involved, as you lower the car the roll center lowers at a higher rate than just ride height. And the lower you get the worse it gets. You can very quickly find the roll center well below ground level and creating lots of problems.

 

[Ulix]

Steve,
Have you determined a sort of „optimal“ ride height for an X1/9?

I know you had yours so stiff that it doesn‘t matter much.

 

[fastx19]

To follow up on Ulix's question, I have always been lead to believe that for the rear of the X, the control arm should be basically flat (parallel) to the ground. In this case, the inboard pickup point on the control arm lines up with the ball join on the rear upright. I always then made the front just a hair higher than the rear.

This was kind of the "this seems to be fast" setup which we all shared back in the day in the Pacific Northwest between Mike Mittlestead and the gang.

Again, though, this setup was mostly for very light street duty and mostly autocross action.

 

[Steve Hoelscher]

To follow up on Ulix's question, I have always been lead to believe that for the rear of the X, the control arm should be basically flat (parallel) to the ground. In this case, the inboard pickup point on the control arm lines up with the ball join on the rear upright. I always then made the front just a hair higher than the rear.

This was kind of the "this seems to be fast" setup which we all shared back in the day in the Pacific Northwest between Mike Mittlestead and the gang.

Again, though, this setup was mostly for very light street duty and mostly autocross action.

That's a good rule of thumb. Once you're past horizontal you quickly get into the bad side of the curve.

Of importance is the car's intended use. If you're wanting a comfortable street car then you need the ride height for suspension travel.

A few notes on lowering: First, the X1/9 is highly dependent on the bump stops. Hard cornering in a stock car will engage the bump stops on the outside corners. This is as designed. The bumpstops are quite progressive and is why you probably didn't notice this. So if you want to lower the car significantly you need to address the bump stops and the amount of available travel.

Once the control arms are near parallel with the gound the roll center is close to ground level. This is because the ball joints are located above the centerline of the control arm and why I often use the term "virtual control arm" (the line drawn between the control arm pivot point and the center of the ball joint).

Once the control arms are level the camber curve is at its crossover point. That is; as the suspension compresses further the dynamic camber begins to go positive. This is NOT what you want. So you want to keep the control arms from compressing much further with body roll.

And as noted previously, the longer the moment arm (between CofG and roll center) the more leverage the car's lateral g has to roll the body. So as you lower the car, despite the CofG being lower, it will have more body roll for a given amount of lateral g.

All of this means that for a street driven car, the more you lower it, the less suspension travel you have and the more the body will roll when cornering. So you have to stiffen the springs to control bottoming and body roll. Which of course makes the ride harsher for a street car. For a street car I tend to lower the front end until the rocker panels are parallel to the ground and leave the rear at stock height. Or maybe lower the rear a half inch and match the front. But I typically want the virtual control arm to be parallel to the ground.

 

[Rupunzell]

To follow up on Ulix's question, I have always been lead to believe that for the rear of the X, the control arm should be basically flat (parallel) to the ground. In this case, the inboard pickup point on the control arm lines up with the ball join on the rear upright. I always then made the front just a hair higher than the rear.

This was kind of the "this seems to be fast" setup which we all shared back in the day in the Pacific Northwest between Mike Mittlestead and the gang.

Again, though, this setup was mostly for very light street duty and mostly autocross action.

From a previous post:

"That said and given there are absolute geometric limitations to the suspension parts in the exxe, it makes sense to limit the suspension travel to keep the suspension part curves reasonable. This is partly due to the need for camber control and control of how much the dynamic roll center is allowed to move around. This translates to what is perceived as high spring rates, but they are not that high once cornering roll angles are considered, the spring rates are simply a product of the cornering loads involved and what is needed to keep the suspension curves reasonable.

~Typically, the LeMons chassis which is stock begins with setting the ride height with the front lower control arm about level with the road level.~

Once this is set up, put the chassis on the corner weight set up, level the rest of the chassis and get the weight per corner reasonable and cross weights near 50%. This means same tire pressures all around with typical driver weight in chassis. Beginning static camber about 3 degrees negative rear, 2 degrees negative front. The steering links have been modified to allow bump steer adjustment which should be done once the static ride height is set. Rubber bushings have been replaced with alloy steel, fiber reinforced teflon liner spherical bearings made by Spec-Line (aerospace supplier) Spherical bearings should be staked at the outside of the bearing to pre-load the bearing and lesser quality spherical bearings will not last. If crappy spherical bearing or rod ends are used, they make chassis set up awful as they will not hold a setting and be repeatable. Dampers must have the ability to control the spring rates involved. Memory notes 550lb/in front and 425lb/in rear (?) no stability bars on a 1,700 pound car. Loaded roll angles are about 2 ish degrees. Chassis easily produces 1G on track with DOT 180 tires. Rear spring/damper rates were tweaked a few times to deal with corner exit wheel spin, there is enough torque from that rotary to make this a very real problem."

Scuderia Ferric Oxide

There are low cost threaded collar kits known as "LeMons lowering specials" found on eBay for not a lot of $. When these first appeared en-mass on LeMons cars, the Lemons folks went looking to discover they were very much low-$.. They work ok, sort of ok. What is often not appreciated, traction...

xwebforums.com

What works good on the track centric car might not work at all on a car driven on public roads. It all comes down to trade-offs and a long list of other factors.

Corning on the bump stops is more common than realized. Given there has been so much discussion about the Miata (NA/NB) suspension, that suspension has been designed to corner on its bump stops as a ride corning roll trade off. Cornering on bump stops allows a lower/softer spring rates and damper ratios that deliver a ride the majority of Miata owners/drivers will accept. To limit body roll in the corners, the Miata suspension limits body roll by using the bump stops as a means to effectively increase spring rates while allowing softer/lower rate springs to be used for straight line ride quality. Don't think the Miata is the only car to do this, cornering on bump stops has been done on open wheel real race cars by shaping the bump stops on Koni race dampers.. It works and can be very effective if properly done.

It is much about goals and needs and trade offs.

Two of the most interesting suspension developments with exception to the Citroen DS/ID Hydropneumatic system would be:

~Aero via reducing the air pressure or create a vacuum under the car body effectively and greatly increasing the down force generated on the car body which can Greatly increase the grip from tires..

~Lotus active suspension.. which effectively controls suspension geometry as needed for road, cornering loading and more..

All of which brings up the question and topic of how much the motoring public values suspension/chassis excellence. IMO, majority of motorist and enthusiast to some degree do not have a good idea or broad experience with truly excellent chassis/suspension motor vehicles as many lack the driving skills and broad exposure to a wide variety of suspension/chassis motor cars.. What is easiest to judge, easy to access and very salable is straight line acceleration and the numeric value of "Horse Power"... Assessment and judgement of chassis/suspension performance is a LOT more difficult and not easy in many ways. Yet wonky suspension/chassis behavior can promote and habituate bad driving/motoring habits that are difficult to un-learn.

~This is one of the many reasons why the exxe remains an under appreciated gem by most motoring folks and motoring media.

There is another factors not mentioned yet in this discussion, location of weight CG. If the weight CG is put at the driver's lower back, it greatly aids in the ability of the driver to access and "feel" how the car is behaving. This is due to the lower back being the human balance point for standing and more. If the weight CG is placed there (as in the exxe) the car can become an extension of the driver in many ways. Cars that are not designed this way tend to disconnect the driver from the feel of the car based on the driver's lower back. Example would be the Miata NA/NB, it has a weight CG at the steering wheel, appealing for some, not workable for others..


Bernice

 

[Rupunzell]

More on active suspensions:

Banned in F1, now common on tanks and other vehicles with specific design goal and minimized marketing/cost limitations.

Citroen DS Suspension Hydrolique Fonctionnement Technique:

More Citroen...

Bernice