I’m A Former Tesla Suspension Engineer And I Need To Tell You Why The ‘Double Ball Joint’ Suspension Is So Incredible

Doubleballs Top
ADVERTISEMENT

Hello fellow Autopians and welcome to another edition of Ask An Engineer. Today we’re going to discuss the front suspension setup found on the Tesla Model S, BMW 3 Series, Dodge Charger, and an increasing number of sports cars. It’s called the “double pivot” or “double ball-joint” suspension, and its design results in the front wheels rotating about a virtual steering axis when you turn the steering wheel (instead of one defined by two physical points like you might expect). And the axis itself actually moves. Let’s look at how it works, and at the significant advantages it offers over a traditional setup.

This article is inspired by a question sent to me by a reader named Kevin. He asked about a comment I made a while ago where I stated that having a steering rack located in front of the front axle had some advantages (like steering feel). He asked why this was and whether a double pivot front suspension also benefits from the rack being in the front. The answer is a very definite “Yes.”

Before we get into that and so much more, you may be wondering what this double pivot suspension Kevin is asking about is, and why it matters when it comes to steering gear position. Let’s start by looking at some front suspension design history.

[Editor’s Note: A quick note on the headline: Huibert was indeed a Tesla engineer, but he was also a Ford engineer (he worked on the Ford GT), and an engineer for a number of other companies. We chose that headline because it’ll interest more of you, but you should know that Huibert hasn’t worked there for quite some time. -DT]. 

Front Suspension History

For the first 30 to 40 years of automotive history, front suspensions consisted of a steel beam mounted on one or two leaf springs. They were simple, and while effective, they weren’t particularly comfortable and their directional stability left much to be desired. Starting in the late 30’s and 40’s, a new kind of front suspension came onto the scene, ditching the beam and mounting each wheel independently to the vehicle using a pair of A-arms or wishbones in a configuration that became known as the double wishbone. There were other permutations of an independent front suspension; the VW Beetle had two parallel trailing arms and the Citroen 2CV had a forward facing single arm, but the majority of passenger cars were migrating to the double wishbone design. 

20220803 123834 Screen Shot 2022 08 03 At 2.19.45 Pm

Above we see a typical front suspension of the time with its stamped steel upper and lower arms attaching to a steel knuckle with a ball joint at the top and another one at the bottom. The steering axis, or kingpin axis, is formed by a line going through the centers of the upper and lower ball joints, and represents the axis which the knuckle/wheel pivots around when the car is steered.

Screen Shot 2022 07 29 At 2.54.18 Pm

A few years later saw the introduction of a new type of independent front suspension, the MacPherson strut, shown above. Invented in 1945 by Earl MacPherson, it didn’t see regular production use until the early 50’s and didn’t really become prevalent until much later. The success of front wheel drive in the 70’s saw an explosion of MacPherson struts since their more compact design was better for the small cars and transverse engine placement that became popular along with front wheel drive.

Like the double wishbone design, the kingpin/steering axis in the MacPherson strut goes through the center of the lower ball joint, but unlike the double wishbone, there is no upper ball joint to help define where the kingpin axis is. Instead, the upper pivot in a MacPherson is at the top of the spring where a bearing allows the entire knuckle and spring assembly to pivot when the car is steered. This upper pivot along with the lower ball joint is what defines the kingpin axis in a MacPherson strut.

How A Traditional Suspension Compromises Brake Size

20220803 132349

Why A Traditional Suspension Has To Move The Ball Joint Towards The Brakes To Minimize ‘Scrub Radius’

Both the MacPherson Strut and the double wishbone designs share a common feature, which is a lower control arm with a single ball joint attaching it to the knuckle. Since this lower ball joint helps to define the kingpin axis, its location is critical. At this point it is important to understand that suspension engineers try to place the kingpin axis as far outboard as possible in order to minimize something called the scrub radius. This is the distance between the point where the kingpin axis intersects the ground and the center of the tire contact patch when the car is viewed from the front.

Screen Shot 2022 08 03 At 7.23.10 Pm

You can visualize scrub radius by watching this video; you can see that the tire isn’t just spinning about a vertical axis going through its center (like a coin would if you flicked it on a tabletop), it’s making a “sweeping” motion, the nature of which can be described by the scrub radius:

Above (and pasted again below) we see a cross section of a tire with the lower ball joint and the kingpin/steering axis going through it. In the case of the diagram, the intersection of the kingpin axis with the ground happens inboard of the tire contact patch. When this happens, the scrub radius is said to be positive. If the kingpin axis were to intersect the ground outboard of the tire contact patch then the scrub radius is said to be negative.

The reason suspension engineers try to minimize the scrub radius is that, when the brakes are applied, the braking force and the scrub radius cause a torque to be applied that tries to steer the wheel. You can visualize that in the image below; if the ground is pushing the tire backwards at the contact patch, which is a few millimeters outboard of where the kingpin axis intersects with the ground, that tire is going to want to rotate, which will pull on the steering tie rod, sending forces to the steering wheel. 

Normally, when braking on dry pavement, we have equal braking forces on the left and right wheels, which together with the scrub radius create equal and opposite torques applied on the left and right wheels. They cancel each other out (they both push or pull the steering rack with similar forces). But when one wheel is on a slippery surface such as wet pavement or ice, the braking forces are not equal and therefore the left and right torques caused by the scrub radius are also not equal. If the suspension was designed with a large scrub radius, these torques could rip the wheel out of your hands or at least cause the wheel to turn suddenly. You can see why suspension engineers try to minimize the scrub radius as much as possible.

Screen Shot 2022 07 29 At 10.23.31 Am

So, if we want to minimize the scrub radius, and if the lower ball joint defines where the kingpin axis is (which defines how big the scrub radius is), then it stands to reason that we want the lower ball joint to be placed as far outboard as possible. Unfortunately, the lower ball joint is not the only thing that has to fit inside the wheel. We also need brakes, and they get in the way. Let’s look at how the brakes fit inside the wheel.

This Creates Brake Packaging Struggles

Screen Shot 2022 07 29 At 10.25.58 Am

Here we see a typical brake package inside a wheel. The caliper is packaged as close as possible to the inside face of the wheel and as close to the rim section as possible. This gives the largest diameter brake rotor. Notice also that the lower ball joint is sitting very close to the rotor and that it really can’t go any farther outboard. As it is, we have to be very careful here because the heat from the brake rotor can melt the rubber boot that protects the ball joint, so we have to stay at least far enough away from the rotor to allow a heat shield to fit in between the two parts. You can see how we are really limited on how far we can push the lower ball joint outboard.

You can also see how the dropwell of the rim limits where we can put the brake caliper and hence how large a brake rotor we can fit. In modern cars, with 18-inch, 19-inch, and even larger wheels becoming more common, this isn’t really much of a problem since the rim is far enough away from the calipers that we can fit almost any size brake rotor we want, but back in the 60’s and 70’s, 13 and 14-inch wheels were the norm while 15-inches was considered a large wheel at that time. With diameters so small, you can see how the rim section would have forced a much smaller brake rotor in order to fit the calipers inside these wheels. In fact, nine and 10-inch diameter brake rotors were commonplace while today 12-inch and even 14-inch rotors are often used. For a 15-inch wheel, the largest brake rotor anyone was able to fit back then was about 11 inches in diameter — large by the standards of the day but not sufficient for a fast car that needed to be able to stop as fast as it could accelerate.

How A Traditional Suspension Compromises Steering Feel

20220803 132231

Why A Traditional Suspension Usually Has The Steering Rack Behind The Axle, And Why That’s Bad

It’s not just braking concerns that the Double Ball-Joint suspension aims to improve, it’s also steering feel.

One thing that automakers tend to do is build understeer into their suspensions as a way to maximize vehicle safety should you overcook a turn (the front of a car is designed to crash into things; it’s generally considered better to understeer into something than to go sideways and possibly roll over). David Tracy (who edited this article) discussed how Ford does this with the Ford Bronco Raptor’s “roll steer,” but I’ll talk about how automakers do it with front suspension bushing compliance. Put simply, one way to do it is to make it such that, when you’re turning, the lateral loads on the tire deform bushings in a way that promotes the tire steering away from the turn.

What you don’t want to deform are the steering rack bushings. Ideally, you want the rack to be hard mounted to the vehicle structure, because that means the loads imparted during turns will end up at the steering wheel instead of being absorbed by squishy rubber, and you’ll enjoy the precise steering feel that everyone likes on a nice-handling sports car. Instead, you want the bushings in the lower control arm to deflect, like the ones in the diagram below do during a left turn — notice how the control arm bushings deflecting while the tie rod stays firm causes the tire to rotate away from the turn, inducing understeer.

Screen Shot 2022 08 02 At 3.13.25 Pm

This stiff steering/nice steering feedback really only works if the steering rack is ahead of the axle. If you move the rack behind the axle centerline, you have to build bushing compliance into the rack itself in order to induce understeer during cornering, and you also have to make the suspension bushings stiff, which can have adverse ride quality effects. Check out what you have to do with a rear-mounted rack to build in understeer:

Screen Shot 2022 08 02 At 3.09.19 Pm

Obviously, a bad ride and sloppy steering isn’t optimal, so you’d ideally want the rack up front; the problem is that a traditional suspension design generally requires the rack behind the wheel. Why? It has to do with a concept called Ackerman.

Ackerman Puts The Rack Behind The Axle On A Traditional Suspension Setup

Screen Shot 2022 08 02 At 3.22.43 Pm

When you turn the steering wheel of your car to navigate a turn, the inner and outer tires are tracing circles of different radii. So to smoothly navigate the turn, the inner wheel should turn a bit more than the outer wheel. You might have noticed that to be the case just looking at the front end of a car whose wheels are cranked hard to one side:

Screen Shot 2022 08 02 At 3.38.59 Pm

The way you achieve Ackermann with a traditional suspension setup is to make sure that the line from the lower ball joint through where the steering tie rod mounts to the knuckle goes directly to the center of the rear axle. Like so:

Ackermann Simple Design
Image credit: Ackermann.svg/Creative Commons Attribution-Share Alike 3.0 Unported

Here’s a little animation of how this setup yields proper Ackermann:

Ackermann2

(Note: It’s a little hard to imagine how that steering set up causes a difference in steering angle depending upon whether the wheel is on the inside or outside of the turn, especially given that the rack is moving the steering arm on each side the same amount. What it comes down to is that the perpendicular distance between the tie rod and the kingpin axis needs to get shorter on the inside wheel and longer in the outside wheel so that the inside wheel turns more for each inch of rack travel and the outside wheel turns less. Again, it’s a little hard to visualize, but that’s not really that important).

You could also build in the appropriate Ackermann in a similar way with a steering rack mounted ahead of the front axle, but the point where the tie rods intersect the knuckle would still need to be on the line between the ball joint and the center of the rear axle, meaning the tie rod would be quite far outboard, likely interfering with brakes, or at the very least compromising brake size.

Screen Shot 2022 08 02 At 3.35.25 Pm

So, with a traditional MacPherson strut or double wishbone design, in order to minimize scrub radius, provide the necessary understeer, and have good Ackermann, you end up with a lower ball joint located far outboard and a softly mounted steering rack located behind the axle centerline, possibly resulting in less-than optimal steering feel. The Double Ball Joint suspension doesn’t have these problems.

The Double Ball Joint To The Rescue

In the early 70’s, BMW decided it had enough of these tradeoffs, so they went about inventing a new kind of front suspension. The solution was to split the lower control arm into two pieces and connect each one to the knuckle with its own ball joint: It’s called the double ball joint front suspension, or double pivot as Kevin called it in his question. BMW placed the two ball joints much farther inboard than the conventional single ball joint design, and this allowed BMW to move the brake rotor inboard as well. Getting away from the dropwell in the rim allowed BMW to move the caliper radially away from the wheel center and fit a much bigger brake rotor. You can see what the effect was here:

Screen Shot 2022 07 29 At 10.30.26 Am

Although the image doesn’t show it as well, BMW was able to use this design to get a 12-inch brake rotor into the same 15-inch wheel that everyone else could only get an 11-inch rotor into. This was a really big deal that gave the company a distinct advantage over its competition, especially on the Autobahn where braking power was and still is extremely important.

The Virtual Ball Joint Allows Room For Big Brakes

Looking at the image above, some of you may have noticed that the kingpin axis does NOT go through the lower ball joint. How can that be? We already stated that the kingpin axis goes through the center of the lower ball joint, but when there are two ball joints instead of one, which one does the kingpin axis go through? The answer is neither. The two ball joints create what is called a “virtual” ball joint and thereby a virtual kingpin axis. Here’s how it works:

Screen Shot 2022 08 02 At 3.46.14 Pm

Looking at the suspension from below, draw lines through the inner bushings of the two lower links and through the corresponding two outer ball joints as shown in the picture above. Extend these lines until they intersect (the photo angle makes the virtual lower ball joint look a bit farther outboard than in reality). The point at which they intersect is the “virtual” ball joint, which is then used to define where the kingpin axis is in the same way as the single lower ball joint did in the traditional double wishbone design

In reality, it’s not quite that simple, because if you look carefully, the lower control arms mount to the chassis on a different Z-plane, so the two lines do not actually intersect, even though it might look like they do from directly below. To really find the kingpin axis (which we need to know to determine things like scrub radius, caster angle and a number of other suspension parameters related to the location of the kingpin axis), we just make a couple of planes. Remember from High School Geometry class how you can define a plane using three points? If you take the center of the upper ball joint (or the center of the upper spring mount in the case of a MacPherson strut) along with the center of one of the inner bushings and the center of the corresponding lower ball joint, you have three points that define a plane. Similarly, if you use the center of the same upper ball joint (or upper spring mount) along with the center of the other inner bushing and its corresponding outer ball joint, you get three more points that define another plane. Where these two planes intersect is the kingpin axis. With this method you can create the equation of the line that represents the kingpin axis and calculate the scrub radius, caster angle and anything else you need to know related to the location of this axis.

Screen Shot 2022 08 03 At 2.11.28 Pm

Screen Shot 2022 08 03 At 2.15.01 Pm

Notice how much further inboard our two ball joints are relative to the “virtual” ball joint. This is space that can be used to pull the brake rotor inboard and away from the dropwell in the rim. That allows us to move the caliper radially away from the wheel center and fit a much bigger brake rotor.

The Double Pivot Suspension Needs To Be Up Front To Achieve Its Ackermann Goals, And That’s Good For Steering Feel

Pretty cool, eh? Yes it is, until you start to steer the car and then things get really weird. It’s complicated and probably best to explain by showing you.

Unlike in a traditional front suspension, the steering gear in a Double Pivot setup actually prefers to be in front of the axle centerline; in fact, it’s pretty much required to be up front. Now I know there are some manufacturers that have used a double pivot suspension with the steering gear located behind the axle centerline (Acura did it with the RLX and Mazda did it with the Mazda 6 many years ago) but these cars suffer from the same distinct disadvantage that comes with such a steering gear placement, namely a large turning circle. The RLX turning diameter is 40.5-feet which is almost two feet larger than its worst competitor, the Volvo S90 and almost six feet larger than the Lexus GS. I contend this is because of the location of the steering gear that Honda chose for this vehicle. Any insiders out there wishing to enlighten me? I am all ears.

Why? Because whereas on a traditional suspension you achieve Ackermann by placing the tie rod-to-knuckle interface on a line between the ball joint and the center of the rear axle, on a Double Ball Joint suspension, the steering/kingpin axis actually moves, and it’s the movement of this axis that causes the inside wheel to turn harder than the outside one.

Screen Shot 2022 08 02 At 5.50.18 Pm   Screen Shot 2022 08 02 At 5.54.39 Pm

The easiest way to understand that is to realize that the wheel and tire rotates about a virtual axis that changes positions, moving forward on the inside wheel and backwards on the outside wheel. For simplicity, you can just imagine that the inside wheel spins about the ball joint closest to its tie rod (the front ball joint) and the outside wheel rotates about the ball joint farthest from the tie rod (the rear ball joint). And we all know that a given linear displacement at a large radius arm is going to swing a smaller arc (and thus rotate less) than that same displacement pushing on a short radius arm. Here’s a video of David at a junkyard walking you through the Double Pivot suspension’s Ackermann motion:

Still think the double ball joint design is cool? Well, so do I and it’s the reason I’ve used it in several cars that I’ve worked on in my career. There is only one downside to this design, and that is cost. Ball joints are not cheap and having two instead of one per side definitely adds to the cost of the suspension. For cars at the low end of the price spectrum, this additional cost may be too much which is why you see this design primarily in higher end cars and SUV’s.

One final word: It doesn’t matter if you have an internal combustion engine or an electric motor, the advantages of the double ball joint design are the same. In fact, it’s the size of internal combustion engines that often get in the way of putting the steering gear where we want it. With EV motors being so much smaller, I think we will see more use of the double ball joint design simply because of the advantages we’ve talked about here.

20220730 160741 (1)

20220730 160736

20220730 102109

A big thanks to Kevin for sending me this question. If any of you have any other automotive engineering type questions, please send them to AskAnEngineer@theautopian.com and I will do my best to answer them.

 

About the Author

View All My Posts

68 thoughts on “I’m A Former Tesla Suspension Engineer And I Need To Tell You Why The ‘Double Ball Joint’ Suspension Is So Incredible

    1. David, the Honda Type R design is a way of getting around some of the limitations of a MacPherson strut and making it behave more like a double wishbone. As such, it is still limited by the brake rotor on how far outboard the lower ball joint can be placed. The double ball joint is a way of projecting the lower ball joint “virtually” to a place where it would be physically impossible to place a real ball joint.

  1. Huibert Mees, is the 2017 Tesla Model X suspension a copy of the Mercedes GLS suspension? What do you suggest to eliminate the inner tire wear due to negative camber on the rear of the Tesla Model X 2017 & 2018?

  2. Thanks for the wonderful explanation. In your article, you mentioned that the lower control arms mount to the chassis (knuckle) on a different Z-plane. Could you please explain the reasons and advantages/disadvantages?

  3. Makes $440 to $500 per day Οnline work which I received $17894 in one month online performing from home. I’m a daily student and work just one to a strive of hours in my spare time. everybody will do that jΟb and online raise extra cash by simply using this link..

  4. I hate this suspension setup. I had one of those Mazda 6s from the rust belt. Taking the front suspension apart was a complete nightmare. Since both ball joints are right next to each other, it made it way harder to remove them from the knuckle. I was cursing that ridiculous suspension all day

  5. With EV’s having the motors more inboard (as far as I’m aware) would there be any reason they couldn’t move the brake setup close to the motor on the other end of the axle shaft so you have more room in the wheel?

  6. What about Longitudinal Audis like the A4 starting in 96′ up through current gen? They have double ball-joint on the lower control arms as well as double ball-joint on the upper arms. Why is this? Does it have anything to do with their rear steering rack placement? I believe they used this style across almost their entire lineup for more than 2 decades with few changes.

  7. That article was fantastic Huibert! I feel like breaking out the old CAD program and start drawing that home made Lotus seven I’ve been wanting to build for years.

    I’ll definitely keep your articles handy when working on the suspension!

  8. nice article. But i wonder how does this setup improve the road and steering feel , since the bushings seem to have so much compliance in them and allow for quite a great deal of movement on all directions. I noticed the pics and videos are of older cards where the bushings might already be shot, but even on the mustang there was quite a lot of movement. Doesn’t this impact stability and static alignment somehow?

    1. The movement you see in the bushings is rotational so wouldn’t affect steering feel. A shot bushing would have movement along the direction of the link, which is not the case in the video. That Mustang only has 36,000 miles on it so those bushings are still in excellent shape.

  9. Great article! I had been able to visualize some of this before, but there’s a lot of other things I didn’t consider. I’d also REALLY like to thank you for helping with a mystery that has annoyed me for years: why did my 1st generation Subaru Legacy have such extraordinary steering feel? I had attributed it largely to the aftermarket steering wheel, but other cars with those still didn’t match up, so I knew I was missing something. I suspected it had something to do with the rear-mounted rack because it was unusual on that class of car, but not on cars known for their steering feel—Lotus Elise, for example—I just couldn’t figure out a good reason WHY.

  10. Never heard of this before, but using the two arms to create a virtual ball joint is one of the most elegant uses for a set of 4-bar linkages that I’ve seen.

Leave a Reply