Especially since automobiles became more common and affordable in the 1920s and 30s, people have been looking for ways to make their cars stand out. One common way is lowering the ride height. In those early days, suspensions were basic, and speeds were relatively low, so the changes made didn’t have too much impact on the way the cars functioned. But modern automobiles are different, and “slamming” them can have profound effects on a vehicle’s handling. Here’s a look at the engineering behind how lowering a car can ruin its handling.
Welcome to another edition of Ask An Engineer. I received a letter from a reader recently where he asked me for help with a problem he was having with his 2017 Skoda Combi. He wanted to lower the car by about an inch for a better look, but he didn’t want to mess up the suspension by doing it.
He knew lowering the car would screw up the roll centers but he wanted to find out if there was a way to fix it. It got me thinking about what happens to the suspension when we lower a car and how that might impact the way the car handles.
Roll Centers
For those of you who haven’t watched my videos (shame on you!), the suspension roll center is an imaginary point around which the body rotates when it rolls side to side in a corner (you’ve probably all felt a car “roll” when you take a turn really fast. The car is not literally rolling over, but the body shifts/rotates in such a way that it seems like you might if you’d taken that turn fast enough).
Both the front and rear suspensions have their own roll centers, and if you draw a line between them you get the axis the body rolls around in a corner, similar to how a door rotates around its hinge axis when you open and close it. In the video above I explain a few methods for determining where the roll center of a suspension is, but for the purposes of this discussion, we’ll stick with the simpler, geometric method. It works reasonably well and illustrates the point. To demonstrate we’ll use a computer model of a generic MacPherson strut, since that is the type of front suspension our friends 2017 Skoda has. (And it’s extremely common).
In a MacPherson strut, you can find the roll center by drawing a line from the top of the strut, perpendicular to the axis of the strut.
Then draw a line along the lower arm (specifically where it mounts to the body) and extend it until it crosses the line from the strut.
The intersection point is called the instant center, and it represents the instant center of rotation of the wheel and tire as they move up and down when viewed from the front. Next, draw a line from this point to the center of the tire contact patch.
Where this line crosses the centerline of the car is the roll center height, which in the case of our model is 95 mm.
But what would happen if we lowered the car by 30 mm? How would the roll center change? We can see this by moving the wheels up 30 mm which is the same as moving the body down:
If we zoom in on the suspension we can see how moving the wheel up 30 mm, or moving the body down 30 mm reduces the height of the roll center:
In fact, moving the wheels up 30mm (again, the same thing as dropping the car 30mm) has resulted in a roll center height of 16 mm. A reduction of 79 mm.
This is fairly typical of MacPherson strut designs. The distance the roll center moves will be over twice the distance the ride height changes.
So What?
The next obvious question is “ok, so what”? Who cares if the roll center drops when we lower a car?
Remember that the front and rear roll centers define the axis that the body rolls around when cornering. But the distance between this axis and the center of gravity determines how much the body WANTS to roll. It’s called the roll moment, and it is the cornering force multiplied by the roll moment arm which is the vertical distance between the vehicle center of gravity and the roll axis (that’s an axis created by a line from the front to rear roll centers).
Make this distance bigger and you make the roll moment bigger. Make it smaller and the roll moment gets smaller.
At this point, it is important to note that there are actually two moments happening at the same time here, and it is very important to understand the difference. We’ve been talking about the roll moment which controls how much the body wants to roll in a corner. But there is a second moment which I will call the cornering moment, and it determines the weight transfer that happens in a corner. They are related to each other only in the fact that they both depend on the location of the center of gravity, but otherwise they are unrelated.
When a car enters a corner, the inside and outside tires develop a cornering force that pushes the car around the turn. This force is balanced by an inertial force acting at the car’s center of gravity. Since the tires and the center of gravity are at different heights, these forces create a moment that wants to roll the entire car over. The reason the car doesn’t flip over is that the moment is resisted by the inside and outside tires through the lateral distance between them, i.e. the track width:
The weight transfer is equal to the vehicle inertial force (which is the same as the cornering force), multiplied by the vertical distance between the center of gravity and the ground, divided by the track width. Notice that the weight transfer gets smaller if the height of the center of gravity gets smaller or if the track width gets bigger.
Notice also that neither the height of the center of gravity above ground nor the track width have anything to do with the roll moment. It is strictly a function of the cornering force and the roll moment arm. Therefore, widening the track width of a car will reduce the amount of weight transfer but will have no impact on the amount of roll the body sees unless the act of widening the track also somehow changes the roll center height. Similarly, lowering the center of gravity will reduce the weight transfer but will have no impact on the roll moment unless it also reduces the roll moment arm.
Let’s go back to our example suspension. By lowering the car 30 mm, we increased the distance between the center of gravity and the roll axis by 79 mm, but we also reduced the height of the center of gravity above the ground by 30 mm. It is tempting think that because the center of gravity is lower and the weight transfer has been reduced, there must be a corresponding reduction in roll, but you see from our example why that is not true. The roll moment has nothing to do with how high the center of gravity is above ground. It only cares about how high it is above the roll axis.
Here’s another way to think about it. Let’s suppose we have a suspension design that puts the roll center at the same height as the center of gravity. In other words, the length of the roll moment arm is zero:
Since the roll moment is dependent on the length of the roll moment arm, you would expect in this case that the roll moment would be zero and the car would not roll at all. And you would be correct. Remember that the roll centers define the axis the body rolls around in a corner. It would be similar to trying to close a door by pushing directly on the hinges. It wouldn’t work. You have to push on the door some distance away from the hinges to get it to close. If the distance is zero, the door won’t rotate. Same here.
However, the center of gravity is still above ground so there will still be weight transfer happening. So, how can we have weight transfer but no body roll. The answer is in the way the weight transfer force travels from the body down to the tires.
There are two paths that the weight transfer force can use to get from the body down to the ground. One is through the suspension links:
while the other is through the springs:
Part of the job of the roll center is to control how much of the weight transfer force goes through the suspension links vs how much goes through the springs. The portion of the weight transfer that passes through the springs will cause compression in the springs since it is just responding to the extra force. This is why we get body roll. In the case of the car with the roll center the same height as the center of gravity, the lack of body roll means there must be no spring compression. The force in the spring must be staying the same. This means all the weight transfer is passing through the suspension links while none of it is going through the spring.
If we look at the opposite extreme where the roll center is on the ground, we have a situation where the roll moment arm and the center of gravity height are equal:
In that case, all the weight transfer force will pass through the spring and none of it goes through the suspension links. Remember, weight transfer is a vertical force pushing down on the outside tire and pulling up on the inside tire. There are still horizontal cornering forces that the tires have to resist, and these will still pass through the links regardless of what happens with the weight transfer force.
Roll Axis Skew
So far, we’ve talked about the roll axis and the roll axis moment arm as if it were purely determined by the front roll center height. The fact is that the roll axis is defined as a line going from the front roll center to the rear roll center and the two may not be at the same height. In fact, in my experience, the rear roll center is almost always a bit higher than the front. What this does is put a bit of skew into the roll axis. Having a skewed roll axis in this way adds a bit of understeer to the car. Here’s how that works.
If the roll axis were perfectly horizontal, i.e. the front and rear roll centers are at the same height, the car would roll horizontally side to side:
If, on the other hand, the roll axis is skewed with the rear roll center being higher than the front, the body will roll in a way that puts more pressure on the outside front tire:
Putting more pressure on that outside front tire means it has to work harder so its slip angle will be greater, adding understeer to the car.
In the case of our example suspension, lowering the car resulted in a front roll center that is 79 mm lower than it was before. If the rear roll center didn’t change, then the lowering of the front roll center would increase the skew in our roll axis considerably, adding more understeer. The reality, however, is that the rear roll center will also drop but it will likely not drop as much as the front. MacPherson struts are notorious for having a lot of roll center movement for a given suspension movement. Multilink and double wishbone designs (commonly found int he rear) are much better in this regard. They will still show some roll center change, but it will be much less.
So dropping our car 30 mm front and rear will very likely increase the skew in the roll axis and add understeer to our car.
Possible Fixes?
So, what can we do about this? Can we lower a car and then fix the roll center problems we’ve created?
There are things we can do that will help but, in most cases, there is no real fix.
Several aftermarket companies have special ball joints available that will move the lower ball joints down. This puts the lower control arm closer to the position it was in before lowering and moves the roll center back up but there are limits to how far you can go with these parts. An example are these ball joints sold by Hardrace for the Mitsubishi Lancer Evo:
These parts have an extended housing that moves the lower ball joints down 15 mm. This would certainly be help but doesn’t compensate for a 30 mm vehicle drop.
In some cases, though, it may be possible to move the ball joint down the full 30 mm. It depends on how the knuckle is designed. Toyota uses a two-piece knuckle in some of its cars and in those cases, we can install a spacer between the OE ball joint and the knuckle and get the full 30 mm change:
We need to be very careful here though, because moving the lower ball joint down has the added impact of changing the relationship between the lower arm and the steering tie rod. We can see the effect of this in our example strut. Before lowering, the lower arm to tie rod relationship looked like this:
After we lowered the car 30 mm it looked like this:
Now, if we install one of these roll center correcting ball joints, the relationship will look like this:
This relationship controls the change in toe angle as the wheel moves up and down. It’s called “bump steer” and it is a critical part of the whole understeer/oversteer behavior of the car. In almost all cars, as the front wheel moves up, it will simultaneously steer slightly outward. This has the effect of causing the car to understeer a little as the body rolls in a corner. It effectively removes some of the steering angle you put in at the steering wheel, thereby causing the suspension to steer a little less than you asked for.
Unfortunately, the amount of bump steer a suspension has is extremely sensitive to the relationship between the lower arm and the tie rod. Even a 1 mm change will have a noticeable impact on the way the car behaves, so a 30 mm change is absolutely enormous. And depending on where the steering gear is located, the change will either greatly increase or greatly decrease the amount of bump steer. If the steering gear is mounted behind the front wheel centerline, as it is on most front wheel drive cars, the change will increase bump steer. In a car with the steering gear mounted ahead of the wheel, centerline, it will decrease the bump steer and may even put the car into and oversteer condition.
[Editor’ Note: I’m including this picture of a slammed Conestoga wagon because Huibert’s first drafts referenced pioneer days and Huibert asked for it but then it got edited out but I already threw this together, so I decided I may as well throw it in, because a slammed covered wagon is funny. – JT]
Those of you who own Toyotas that use a two-piece knuckle are in luck because as you can see from the photo above, the part that holds the lower ball joint also holds the outer tie rod end. This means that the 30 mm spacers move both joints together. This avoids the whole problem I just mentioned, so you Toyota owners just read about an interesting problem that’s entirely irrelevant to your situation. For the rest of us, we still have an issue because most knuckles are single piece. Moving the lower ball joint down does nothing to fix the position of the tie rod end.
The only way to resolve the bump steer problem in that case would be to move the outer tie rod end down along with the lower ball joint somehow but this is often very difficult to do and would most likely require a custom knuckle.
The Rear Problem
Another problem you run into when lowering a car is that while there are some things you can do to help the front roll center, it is way more difficult to fix the rear roll center. In the case of our friend’s Skoda Combi, tha car uses a trailing blade rear suspension design and there is no good way to move a ball joint or two to get the roll center back to where it was.
You would have to move the entire suspension up and that would be quite impossible without cutting up the rear underbody.
The Bottom Line
What we’re left with now is a situation where, by lowering our car, we’ve increased the roll moment arm, thereby increasing the amount of body roll. We’ve also increased the skew of the roll axis thereby increasing understeer.
But then we go and install roll center correcting ball joints or spacers which moved the front roll center back up, but since we can’t really fix the rear roll center, we’ve now reduced the roll axis skew, reducing understeer. But (unless you own one of those Toyotas) we’ve changed the relationship between the front lower control arm and the tie rod which could have the effect of either increasing or decreasing bump steer and with it understeer.
The end result is a giant mess where the balance between the springs, dampers, bump steer, roll axis, and all the things that go into making a car handle properly has been upset. And no one knows where this balance has ended up. The reality is that for most people, under normal driving conditions, it won’t matter. Unless you regularly probe the limits of your car’s handling performance, like on a racetrack, you will probably never know anything has changed. But if you ever find yourself in an emergency where you have to quickly maneuver out of a bad situation, the change in handling balance can rear its ugly head and the car may behave in very unpredictable ways.
Knowing all this now, should you lower your car? If you like the way it looks and don’t care about much else, go for it. But don’t do it expecting your car to handle better. It very likely will not and may actually handle worse as a result. I hope you never get into a situation where you have to find out.
Interestingly, on cars with double wishbone suspensions (such as some higher end luxury and sports cars) lowering will have the opposite effect. With a double wishbone setup the instant center is determined as the intersection of lines extended from the upper and lower control arms. Lowering the car ’tilts’ both control arms downward, lowering the instant center and therefore the roll center as well as the CG. The relative lowering of the CG vs. the roll center will depend on a variety of factors, but they are at least positively correlated.
Maybe slammed BMW 5-series guys are onto something…
Lowering a car can make it handle improperly. Raising a car (or truck) can make it handle improperly. Putting wheel spacers on a car can make it handle improperly. Stancing a car can make it handle improperly. Putting oversized wheels on a car can make it handle improperly.
Weird. It’s almost like automotive engineers know what they’re doing when they design a car to use the components put on at the factory. Maybe we should leave well enough alone for cars that travel on public roads.
Reminds me of the 80’s when cars would be lifted for headlight regulations and suddenly they drove like shit. Can’t imagine why?!
How does the rear roll center being higher than the front contribute to understeer? I would think it would act similarly to a stiffer sway bar (overload the rear outside tire)
But it looks cool 😀 😉
I found this one particularly interesting Huibert! I’ve always been interested in the technical aspects of suspension,albeit mostly to do with motorcycles.
I knew roll center was a thing on cars and i knew lowering can do bad things but wasnt exactly sure why, so this has been great.
I learned my lesson by once over-lowering a GM H-body, causing it to pitch and bottom-out in the corners on favorite mountain roads with dips and bumps in them while a stock-height car would be pulling away from me. I would like to see more discussion sometime on the practical effects of what I would call “lateral suspension” where with a too-low roll center there is too much compliance such as with the above example, then where the C/G and roll center are aligned there is none, and with the roll center above the C/G such as with a swing-axle there is anti-suspension causing skittishness and chatter through bumpy corners. It’s the sometimes-choppy public roads such as Angeles Forest Hwy I grew up on where it would make a difference and the perfect roll moment arm amount might be a little more than on a smooth race track, made up for by sway bar and shock rate to give that lateral suspension which may be made-up intuition in my head or a real thing (ask an engineer…). Those would be the same conditions where you might want a little bump steer to compensate for sudden higher- or lower wheel/tire loading and traction, another matter.
Interesting. Now, Skoda in question (and Golf etc.) have the factory sport springs as an cheap option (or packed together with some visual sporty enhancements). The ride height is 15 mm lower, no other changes in the setup as far as I know. I wonder how this works in practice?
I’m assuming (hoping) that when a vehicle manufacturer offers a ‘sports suspension’ option, that they are adjusting the suspension accordingly to handle the modified geometry (e.g. stiffer springs / differently-rated shocks, larger anti-roll bars, etc.). But is that only peanut-buttering over the geometry issues, or is it an actual solution?
[looking at you, BMW]
That surprises me.I would have thought anything BMW make is set up better than other brands.No clue if that’s the case though
In the M cars, they have completely different subframes and control arms, but I don’t think any parts change between the standard cars and sport packages. If I was engineering them, I’d include spacers for the subframes in the standard cars that aren’t present in sport package-equipped ones, but I have no idea what they actually do about it.
I do know that for the E8x / E9x models, the ZSP / ZMP packages which have the M Sports suspension had thicker front anti-roll bars and rear anti-roll bars (non-sport didn’t get the rear bar), but aside from than that, I think everything else other than the M Sports struts / springs are identical.
FWIW, the M Sports suspension is 15(?)mm lower than the standard suspension in the E8x / E9x, so half the drop of Huibert’s scenario.
There’s no way to know for sure unless we were to measure a BMW’s suspension geometry in the flesh in both positions, for all we know they may have optimized the car’s handling at the dropped height, and then lifted it a bit for the non-enthusiast trims to improve comfort.
AFAIK, the only differences in front suspension between E8x / E9x M Sports suspension and 1M/M3 suspension are shocks, springs, anti-roll bar (same 26.5mm diameter, just with bump stops to prevent lateral motion), subframe bushings, and wishbone (which is slightly longer and has spherical bearing on subframe side). Front subframe is different due to engine mounts for V8-equipped M3, but identical to 1M. Hub carrier and radius rod are identical geometry-wise.
Rear is pretty much completely different, though (although a few lighter-weight parts can be bolted up w/out issue on E8x/E9x).
FYI – swapping E9x M3 front radius arms and wishbones into E8x/E9x is a *very* popular upgrade as you get lighter-weight control arms as well as a bit more camber range.
That makes sense, so for the E9x cars, the M suspension was just a more adjustable/lighter version of the standard?
I’m not as well versed on those, but on the F-chassis cars the M2/3/4 have different subframes and longer control arms. M2 control arms are a popular way to add camber and caster to an M235i, improving turn-in and steering feel, and something I really wanted to do before I sold mine. Instead, I opted to improve steering feel (and cut horsepower in half) by buying a Boxster.
There is A W220 in my neighborhood that’s in great shape other than it’s absolutely slammed. I can only cringe imagining how the once pristine ride is now probably pretty terrible.
very likely has blown suspension; when airmatic fails it essentially slams the car – the difference in ride height is profound.
Really? I knew about the air suspension but this seems so low that I can’t imagine the wheel wells are unmodified. Are they able to move like this? This car occasionally moves spots but I’ve never actually seen it driving.