Back in 1990, Honda and Acura introduced the NSX, an aluminum sports car that, at the time, was a bit of a disappointment: The styling was a bit bland and the performance, while competitive for its day, didn’t have that visceral feeling of a Ferrari or Porsche even though you could rev the engine all the way to an 8,700 RPM redline. The Japanese sports car didn’t have the scream of a Ferrari and it didn’t shout “look at me” like a Lambo. But there were some areas where the car really shined.
[Editor’s Note: Welcome back to dynamics engineer Huibert Mees’s column, where the former Tesla Model S/Ford GT suspension designer can write whatever the heck he wants. It’s been a little while since you last read his work, but that’s not on him — that’s on me! There’s a bit of an editing backlog, but I’m being honest when I say that editing Huibert’s nerdy engineering pieces is among the most joy I’ve had as a car journalist. You’ll enjoy this one, I know it. -DT].
One example where the NSX stood out was its ride, which is the main subject of this article; I’ll get to that in a second, but for now you’ll have to indulge me as I briefly talk about the shifter. Oh.…. My….. God! That shifter! Anyone who owns or has driven a first generation NSX will know exactly what I’m talking about. It is pure magic and adds significantly to the joy of driving the car.
The NSX’s Shifter Inspired The Ford To Run The Ford GT’s Shift Cables Through The Engine
Look at that thing. It’s gorgeous! The handle is a polished teardrop shape that fits in your hand perfectly, and the snick-snick feeling of changing gears has, in my humble opinion, never been equalled in any other car. When we designed the 2005 Ford GT, our goal was to make our shifter feel as nice as the NSX and I think we got pretty close. It took an inordinate amount of effort to get there, though. Good shift feel is all about keeping the number of bends in the cables to a minimum and keeping the radius of the bends as large as possible. Anything else adds friction and ruins the feel.
In the NSX, with its transverse engine mounting, it was relatively easy to route the shift cables from the shifter to the transmission with a minimum of bends, but in the GT, with that big engine sitting between the shifter and the transmission, it was much more difficult. Going around the engine with the cables would have meant many bends and lots of friction, so the decision was made that if we couldn’t go around the engine, we would go through it. And that’s what we did. The GT’s shift cables go through the valley of the engine underneath the intake manifold. It made for an almost straight shot from the shifter to the transmission and very good shift feel.
How The NSX Suspension’s Incredible ‘Pivot Assembly’ Gave It Such A Good Ride
But that’s not what I came here to talk about today. I want to talk about the NSX front suspension, because there is something unusual about this design that is the key to the excellent ride the car has. There is an extra piece in the front suspension that no other car before ever had and no one has used since. It connects the forward bushings of the upper and lower control arms to each other. In fact, they aren’t really bushings at all, they are ball joints, which is highly unusual in itself.
So why would Honda go through the effort of putting this extra piece in there? Well, it goes back to something called “caster wind-up.” To understand what that means we have to first talk about two of the primary parameters that go into a front suspension design: caster angle and caster trail.
Caster Angle
The caster angle is the angle of the line between the upper and lower ball joint centers when viewed from the side of the vehicle. This line is called the Kingpin Axis, and like a hinge on a door, it is the hinge that the front suspension rotates around when you steer the car.
Caster angle is very important for steering stability and returnability. It’s the same as on a bicycle. The front fork of a bicycle is angled forward to give stability in steering. If the fork were perfectly vertical it would be very difficult to control and you could forget riding no-handed. It would be all over the place. The same thing is true in a car. Caster angle helps the car go in a straight line and be stable. It also helps the steering come back to center when you let go of the wheel.
Another important thing caster angle does is provide steering feel. The forces coming back from the tire act through the caster angle and the caster trail (see diagram similar to the one above below) to give the feedback forces we feel in the steering wheel (look at the image above, and you’ll see that the lateral forces between the ground and the tire are a distance from the steering axis; this means that cornering loads on the contact patch will try to create a torque about that axis that the driver will feel in the wheel; this is steering feedback or “feel,” and it’s highly dependent upon caster and caster trail). Of course, too much caster means these forces get too high and we would then need to increase steering boost to compensate which would hurt steering feel. It’s a delicate balancing act that the suspension and steering engineers have to play. We’re talking about small values though. In my experience, between 4 and 6 degrees of caster angle strikes a good balance between all the competing requirements the suspension designer needs to consider.
Caster Trail
Caster trail is the distance along the ground between the center of the tire contact patch and the point where the kingpin axis intersects the ground. The reason it is important is that during cornering, the side force and the caster trail create a moment around the kingpin axis that works to return the steering back to straight ahead. This moment also helps create understeer which is important for safety should you lose control of the car in a turn. Understeer makes sure the car is pointing forward when you lose control so that if you hit something after running off the road, you hit it with the front of the car, not the side or rear. Again, we’re not talking about big numbers here. Between 25 and 35 mm is very typical in modern cars.
Both of these parameters are very important for the functioning of a front suspension and it is important that they change as little as possible while driving.
To fully understand how all this fits together, we need to also consider that a car suspension contains rubber bushings to make them quiet and comfortable. When a tire rolls over imperfections in the road, like potholes, expansion joints, broken pavement, etc., it needs to be able to move out of the way slightly or else the impacts would be very jarring and noisy to the passengers. The bushings allow the tire and the suspension to move not just up and down with the spring and damper, but also fore and aft slightly so that the impact and energy of the tire hitting the pothole or expansion joint is absorbed as much as possible. If you’ve ever driven a race car you will know what I mean. Race car suspensions have very little, if any, rubber in their bushings and the results is very crisp handling but also a very jarring and uncomfortable ride. It’s a good thing race tracks tend to be smooth!
Unfortunately, all this rubber causes a problem during braking. When we apply the brakes, we create a braking force which tries to push the suspension rearward in the same way a pothole does. However, braking also generates a moment that tries to rotate the suspension in the same direction the tire is rolling. The rotation of the suspension caused by this moment reduces, or even eliminates, the caster angle and the caster trail. This is called the caster wind-up. The caster angle and caster trail are changed as a result of the “wind-up” in the suspension caused by braking forces. This means that if we brake during a corner, the understeer and steering returning forces created by the caster angle and caster trail are reduced or even eliminated by the caster wind-up.
For the most part, more rubber will make the car’s ride quieter and more comfortable, up to a point. Too much rubber can lead to other problems, like shake, which refers to the vibrations in the car just after the pothole or expansion joint impact. You can feel this very clearly in some older, softly sprung Lincolns or Cadillacs. Hit a pothole in one of those cars and the shaking seems to last forever. Too much rubber can also make the steering feel numb and uncommunicative and make the handling dull and uninteresting.
How Honda’s Design Prevents ‘Wind Up’
Let’s look at how rubber in the bushings causes caster wind-up. When a suspension encounters an impact, like a pothole, the force of the impact tries to push it rearward. (In the image below, the NSX is moving towards the bottom left; the rear of the car is located towards the top right)
In an ordinary double wishbone design, the impact force causes the upper and lower ball joint forces to be in the same rearward direction which tries to rotate both control arms. This rotation is resisted by outward reaction forces at the forward bushings of both arms. The rubber in these bushings deflects as a result of these forces, which allows each arm to rotate. This rotation allows the upper and lower ball joints to move rearward and allows the tire to move rearward, thereby absorbing some of the impact.
During braking, on the other hand, the upper and lower ball joint forces are in opposite directions because there is the addition of a moment created by the brakes trying to slow the car down. The car doesn’t want to slow down because of Newton’s first law of motion which states that an object in motion (or at rest) wants to stay in motion (or at rest). The motion, or momentum, of the car wants to keep the wheels rotating even though the brakes are trying to slow them down. This creates a moment that pushes on the suspension.
Since the upper and lower ball joint forces are now in opposite directions, the reaction forces at the bushing are also in opposite directions and any deflection in these bushings will cause the control arms to rotate in opposite directions. This rotation causes the upper ball joint to move forward and the lower ball joint to move rearward. Remember that the kingpin axis, caster angle, and caster trail are defined by where these ball joints are so if they move then these parameters change as well. If the upper ball joint moves forward and the lower ball joint moves reward then the caster angle will be reduced and the caster trail will also be reduced. The more rubber is in the bushings, the more motion of the ball joints can occur, leading to more change in caster angle and caster trial.
Compare what you see here with the corresponding image above and you can see how the motion of the upper and lower ball joints have reduced the caster angle and caster trail.
So, suspension engineers are left with a dilemma: Do we add lots of rubber to make the car comfortable and live with the caster wind-up, or do we keep the rubber to a minimum, sacrifice ride but keep the handling and steering nice and crisp? The answer, of course, depends on the type of car you have. It might be fine to have dull handling and steering in a Town Car but it wouldn’t be fine in a Corvette. The same goes for ride; a Town Car needs to have a soft and quiet ride but a Corvette can get away with a much stiffer and noisier ride.
But what if there were a way of getting a soft ride without sacrificing handling and steering? Let’s see what Honda did to try to overcome this dilemma. Let’s look at pothole impacts and braking again and see how the NSX front suspension handles these situations.
First, let’s look at the extra vertical link (Honda calls it a “Pivot Assembly”) Honda put into the NSX suspension and see how it’s mounted.
The vertical link, located towards the front of the vehicle (on the left) is mounted on two rubber bushings that form a vertical pivot axis which allows the link to rotate around this axis. It is also connected to the upper and lower control arms by two ball joints.
The rubber bushings allow the link to rotate around the pivot axis when the upper and lower control arms push on it and return it to center when the forces coming from the control arms go away.
Now let’s look at what happens during a pothole impact.
As we saw before, the pothole impact tries to push the suspension rearward which tries to rotate the control arms in the direction shown. This creates an outward force at the front control arm bushings, which in the case of the NSX are ball joints connected to the vertical link. Since both ball joints are pushing outward on the vertical link, it will rotate around its pivot axis and allow both ball joints to deflect outward which allows the control arms to rotate and move the tire rearward. This is exactly what we want for a good ride and impact absorption. Sounds perfect so far.
Now let’s look at what happens during braking.
As in the ordinary wishbone suspension, we have the braking force but we also have the braking moment, which causes the upper and lower ball joint forces to be in opposite directions. We also have the same opposite reaction forces at the control arm forward bushings, but in this case instead of having rubber bushings connected to the body, we have ball joints connected to the vertical link. The ball joints won’t deflect very much on their own and since the upper and lower reaction forces are pushing on the vertical link in opposite directions, it won’t rotate around its pivot axis and therefore the upper and lower control arms are not allowed to rotate like they did in the ordinary design. Since the control arms can’t rotate, the upper and lower ball joints can’t move which stops the change in caster angle and caster trail from happening and ensures the handling, steering feel and returnability are not changed when braking in a corner. Pure. Genius.
So why did Honda choose not to have this feature in the new NSX, and why has no one copied this idea? The NSX proved the concept worked and worked very well, but the extra part looks stupidly expensive and even in $100,000+ cars, component cost is still important. The other problem is that the upper and lower control arm forward bushings must be almost directly above each other which limits what you can do with the overall suspension design. Still, it seems to me this concept could have been adapted by other high end sportscars to reduce the compromise between ride and handling.
I still think this is a brilliant design and it blew me away from the moment I first saw it all those years ago.
I put pivot locks on mine to sharpen the responsiveness on track, but eventually I removed the locks as age and wisdom gave the car less challenging duty.
You’ll only notice the difference as far as responsiveness on a a track. You will however feel a more “calm” road experience just letting this little miracle piece do its thing. I do suggest lower transverse chassis rebars. The front ones like that from the NSXR is no longer made, biut Lovefab makes its own version for over $400.00. I contacted online metals a short while back and it will cost me under $140 to buy, cut, and drill to fit an even sturdier bar of light aluminum alloy.
That’s my first project for the car when the warm returns to eastern OK.
Firstly THANKYOU for actually giving an accurate representation of the contemporary reports on this car.
The cold hard facts were that journalists praised many aspects of the NSX but still -almost universally- had the Ferrari as the better driver’s car.The NSX was seen as a well built sports car that was too sterile (my words).
The NSX started gaining it’s present overblown reputation only years later.
My experience first seeing the NA1 NSX at a dealer was “I want it!” But good lord it was $65,000. Years later with money under my belt I got a 2000 NSX with 12k miles for 65k! (in 2003). I’ve owned it ever since and I agree that the shifter is the most gratifying thing to use in any car or machine of any type. It is glorious. I have the privilege of being close to Daisy Auto in Woodbine MD where Ben the NSX whisperer resides. In 2017 they gave my car a once over (including the 35 hoses in the cooling system) to return it to better than new condition. I still cannot believe the smooth silky feel of the engine at 8k RPM that sounds like it would like to run all day at 18k RPM if you’d let it. A most amazing car.
Interesting tidbit: some part of the nether region of my brain was trying to tell me, “Weren’t there people modifying or removing this special component on the NSX?” Got into researching it tonight, and it’s true.
The most popular approach is installing a clamp that ‘locks’ the NSX’s “compliance pivot”, primarily for track duty for more consistent alignment control.
Funny enough, that clamp can be seen on the Pacific Motors-sourced images used in this article. 🙂
Or you could get rid of those pesky ball joints and do like Morgan does.
But this is a majorly great article, more like this please. Except NSX ugly? Compared to what?
Good article! I just realized that Huibert might be able to shed some light on a NSX/Ford suspension mystery.
Why are there Ford stampings on some NSX rear-upper control arms?
http://www.nsxprime.com/forum/showthread.php/208945-Ford-stamping-on-NSX-rear-suspension-parts
Shared OEM who incorrectly stamped some parts? Internet prank? Somewhere in between?
OK, that bit about the shifter and linkage was some fascinating reading. It’s one of those things that seems obvious when you hear it but I never thought of it. Thanks for that.
I never got to drive one. 🙁
I got to blip the throttle and snick the shifter around when it wasn’t running, and I felt pretty special just doing that.
I think I remember a contemporary write up that punted to “which Honda engineers assure us somehow helps with steering feel without roughness, especially during braking” but I might be backfilling my memories.
This Is Awesome.
I’m curious, could you accomplish something similar with biased bushings? As in bushings that are thicker vertically to provide compliance, but thinner horizontally to resist the twist of the braking force? You’d give up some amount of compliance, but it seems like the NSX version must too since it also restricts the motion in the bushed joints. I also might not fully understand how it limits the twisting though. 🙂
The problem is that the forces from a pothole and the forces from braking are both horizontal. It is possible to make a bushing that is soft in one direction but stiff in the other but these cause other problems and they are much less effective than the link Honda designed.
Sort of what Murray did in the McLaren F1
Got it, after a bit of looking. An as-seen-from-above diagram would have helped, to better see how pothole impact force on the ball joints makes the pivot rotate, but I ain’t complaining. Presumably pivot bushings are compliant in rotation but not side-to-side.
I remember these cars being pretty in magazine photos but not out on the road. Body construction violated one of my cardinal rules; Never Trust Glue.
First off, Huibert, David, great article! David, I particular appreciate you adding the video to help explain Huibert’s details. It really helped!
Second, David, you went to the Salt Lick and didn’t stop by to say “Hi” ?!? I’m just a few miles away, assuming you got your t-shirt at the original location.
I fail to see the point, you still have bushings in the pivot, and the ball joints are going to transmit the reaction forces to them. Shouldn’t they compress in the same fashion as a standard setup? What am I missing here?
Texa, in a normal suspension, the upper and lower arm bushings can compress in opposite directions during braking because the forces in the upper and lower arms are in opposite directions. That’s what causes the caster wind-up. This link works through the rotation of its bushings not compression. This prevents the upper and lower arms from moving in opposite directions because they are tied together by the pivot link.
Yes but isn’t the couple of opposite forces coming from the control arms going to create a moment in the pivot (the truss structure I mean), which will induce some bushing compression in opposite directions?
Man I wish we had a white board to talk it out, it makes things so much easier.
Yes you are correct but the bushings would be designed to resist compression while still allowing the rotation that makes the link function. That way the upper and lower control arms can still move, but only in the same direction.
I see! That was the missing piece of the puzzle that was … puzzling me.
That’s cool! Because on a classic suspension you do not have the ability to design bushing that are only compliant when moving together, since they work in compression, but having a different stiffness in compression and torsion is achievable. That’s how they concile ride comfort and good geometry.
Thanks for taking the time to explain this Huibert!
I was following along just fine til we got to the diagrams of the NSX suspension. There’s no indication of which way is forward. As such I’m lost as to what is actually happening.
It’s stated in the article! Front of vehicle is on the left!
I’ll add a second mention to make it ABUNDANTLY CLEAR.
Your left or my left?
I’ll let myself out.
Thanks for the comment. I’ll make sure to include it in the future. For reference, front of the vehicle is ALWAYS to the left in my diagrams.
Hey, you don’t need to placate this comment. There is a clear blue arrow in each diagram indicating the front of the car.
The “This Part Here” is to the rear of the lash up.
Ok I got your point and will read the article fully later, but as an owner of a fifth generation Accord, the links that make it drive so great are also going to make it drive pretty damn well even on the brink of failure. So all I’m saying is you gotta be on top of checking out the whole suspension on a regular interval with either the simplest or most complicated Honda’s.
I came for Torch, but these sorts of technical highlights are what really has my attention! The bifurcation of articles like this and whatever Torch’s latest musings happen to emit are going to make Autopian a force to be reckoned with. Keep it coming!
I agree. I like both the in-depth technical write-ups, and the humor. It’s a good combination and you can pick what suits your mood.
Great article. And I got my physics lesson for today.
I hope the final exam is multiple choice.
Bland styling? I thought the NSX was like hot buttered sex when it came out and it has aged wonderfully.
clean & simple =/= bland
Simple ingredients prepared properly can be delicious.
Yeah, that seemed strange to me, too. 13 year old me drooled like an underfed dog looking at a fine steak.
I agree that the styling has aged extremely well. I love the look of the car but I distinctly remember the general reaction in the press was one of underwhelment (is that a word?) by the styling. We had one as a competitive assessment vehicle on the Ford GT program and there was a lot of competition for who got to drive it.
At the time there were quite a few similarly wedge-y vehicles.
Singing: “It was the age of hypotenuuuuuse. Hypotenuuuuuuuuuuse!”
I distinctly remember that Car and Driver had a comparison test in 1992 where they said the styling was bland when it was, to little 7 year old me, the most exciting looking car in the test.