BMW says the new 2019 BMW 3 Series, which debuted last month in Paris, “moves the game on once again in terms of driving dynamics.” So, to learn the tech behind what makes the new car handle better than its predecessor, I spoke with dynamics engineer Robert Rothmiller. Here’s what he told me.
The head of functional design and integration for driving dynamics of the new G20 3 Series broke down the main changes to the car’s handling into three categories: weight and track width, body and chassis stiffness, and technical systems. The first of those promises to give the car better grip in turns, the second could improve chassis response and comfort, and the last supposedly yields a better driving experience thanks to changes in the steering system and in damper technology.
I’ll begin by discussing technical systems, specifically steering. This is an area where the 3 Series has received significant criticism since the current generation F30 model launched with electric power steering, inspiring articles like Car and Driver’s “Steer Me, Feel Me: Exploring Why BMWs No Longer Excel in Steering Feel.” But Rothmiller says his team has made improvements.
Not long after the current-generation F30 BMW 3 Series launched to much criticism about steering feedback—a term defined as how well a car communicates grip to the driver via the steering wheel—Car and Driver hooked the car up to a $3 million kinematics and compliance machine and compared it to its predecessor. The conclusion? “The most significant change from the E90 to the F30 results from the switch to electric power-steering assist, which diminishes feel,” the magazine wrote.
In particular, the test measured feedback by looking at “aligning torque,” which is a torque created by the force between the road and each front tire, and that tends to straighten the wheel during a turn—similar to what you’ve probably noticed in a shopping cart wheel. That self-aligning torque in a turn can be translated through the steering tie rod, steering rack, steering shaft, and finally to the steering wheel, helping the driver feel what the front wheels are up to. Here’s what Car and Driver found about the aligning torque of the F30 and E90:
The K&C machine reported that aligning torque is 64 percent lower in the [F30] 328i than in the [E90] 335i. While the switch to electric power steering may improve mileage, it’s a bummer for Bimmer driving satisfaction.
This method of measuring steering feedback is similar to how BMW itself does it, with Rothmiller telling me the Bavarian automaker quantifies steering feel by reading the torque at the steering wheel at a given vehicle lateral acceleration. And in that measurement, he says, the new G20 3 Series is up to 10 percent better than its predecessor thanks in part to changes in suspension geometry.
The general suspension setup, which you can see above, is the same as that of the outgoing F30 3 Series. There’s a double-joint spring strut front suspension and a five-link suspension in the rear, but it’s not literally the same, even if it looks similar. “It’s kind of a whole optimized system. There was not the possibility to carry one single part from predecessor,” the dynamics engineer told me.
With the new 3 Series, engineers have increased a dimension called Nachlaufstrecke. In english that’s mechanical trail, also called caster trail. It’s related to caster, but while caster is an angle, this trail is a linear distance.
Defined as the horizontal distance between where the vehicle’s steering axis intersects with the ground and the center of the tire’s contact patch (see yellow in the image above), trail is an important metric because it essentially quantifies the “lever arm” that the lateral force exerted by the road onto the tire (this force would push inwards at the “center of contact patch” labeled above during a turn) has to rotate the wheel about the steering axis to create that self-aligning torque.
This quantity, the engineer says, is larger on the new car than on the predecessor thanks in part to a steeper angle between the top mount and the road surface.
That steering axis, it’s worth mentioning is “virtual” in that it extends from the top strut mount to an imaginary point, and not to a single joint. This is because there are two ball joints in this “double-joint spring strut suspension” about which the knuckle rotates.
But this “trail” lever arm between the force of the road and the steering axis isn’t the only thing affecting how much grip information gets transmitted to the driver. Rothmiller also mentioned the distance or “lever arm” between the steering tie rod end and the steering axis (which he says BMW reduced for better feedback) and also the steering rack gear ratio. He told me that the new car’s 10 percent steering wheel torque improvement is a product of optimization of these ingredients.
“The front axle is designed to give you more mechanical feedback at the steering wheel,” he told me, so that the driver can better feel if the car is understeering or oversteering, and that the vehicle’s handling remains predictable and honest.
One thing that’s worth mentioning is that we didn’t really discuss the calibration of the electric power steering assist, which is often blamed for “filtering” out the torque that the wheels would otherwise want to transmit to the steering wheel to generate feedback. One possible reason why Rothmiller didn’t mention it is that new power assist calibration might not have affected feedback as much as the aforementioned geometry changes. Car and Driver, who spoke with chassis division head Jos Van As and who got to drive a prototype 3 Series, seems to confirm this theory, writing:
Van As says one key lesson his team learned is that relying on the tuning latitude afforded by electric power steering—such as automatic self-centering—tends to mute feel. This time around, they worked more on the underlying kinematics, and to good effect.
BMW has been messing with variable steering for a while now, much of it hated by purists. The problem has been that the idea of steering that’s quick in corners but not darty while making little adjustments on the highway seems good, but the application of the idea feels weird.
As it turns out, there’s a mechanical part that’s responsible for the strangeness, one that BMW changed up.
Unsurprisingly, Rothmiller mentioned altering the new variable sport steering setup that comes with the M Sport and adaptive M Sport suspensions. Here’s how it works.
Rothmiller says the new sport steering offers 2 millimeters of additional lateral rack travel per degree of steering wheel rotation in the center, whereas the old sports steering had a similar on-center ratio as the new standard, non-sport rack.
But in a turn, well off center, the steering ratio changes thanks to an increased gear spacing in the rack. As Rothmiller told me, the previous car’s setup yielded a sudden change in ratio thanks to a sharp step in the spacing of the gear teeth on the steering rack. He drew a picture to help explain this concept:
At the center of the rack, the gear teeth are close together, meaning the ratio of steering wheel input to wheel output is relatively small, yielding less sensitive steering when driving straight at high speeds. But as the driver turns the steering wheel, it rotates the pinion, whose teeth then mesh with rack teeth farther spaced out, yielding more wheel angle change per steering wheel angle change, meaning the steering is more direct.
In the plot above, the dashed line represents the old BMW sport steering design. Notice how there’s a sharp step up from the less sensitive on-center ratio to the lower ratio in the turns. This sharp step, Rothmiller told me, could possibly cause a driver to turn too hard into a bend if they couldn’t anticipate the ratio change.
The revised sport steering setup smoothens out the transition from a high steering ratio on center to a lower steering ratio in turns by gradually increasing rack gear spacing, as shown by the dotted line in the plot above. The change, Rothmiller says, is all about precision, and creating a predictable steering setup for the driver. This is an example of how the 3 Series’ changes are not just about raw handling figures, but about reducing the effort needed to drive the car fast.
Between this and the changes in the steering geometry, BMW thinks it has improved the overall steering experience of the new 3 Series.
Rothmiller also talked about the 3 Series’ suspension system, especially touting the car’s all-new “lift-related dampers,” which make their BMW debut as standard on the regular G20 3 Series and M Sport models, and which vary damping forces as a function of wheel travel.
He even drew me a diagram of how they work:
On the left, labeled “RA” for “rear axle,” there’s essentially a smaller damper within a damper, which—when the car is unloaded—remains out of the smaller tube. But when you throw some weight in the trunk, go over an undulating road, or enter a turn quickly and cause significant load transfer, that damper can enter the smaller tube, and provide more damping.
“As soon as the additional piston dives into the smaller inner tube, of course you get a higher damping force,” Rothmiller told me. But it doesn’t have to dive entirely into the inner tube to have an effect. Rothmiller pointed out that as the piston reaches about 0.8 to 1.2 inches into its roughly four-inch downward travel, it achieves a phenomenon that he called “Staudruckeffect,” which I understood to be a stall condition in the main chamber caused by the rapidly diminishing area between the smaller piston and the main chamber walls. (As you can see, the main chamber tapers down in size.) This condition yields higher damping forces even before the little piston enters its tube.
Rothmiller drew the rear damper once more below in the normal ride height position (labeled “1b” below) and in a slightly compressed position (“1a” below), along with a force vs. displacement plot (number two)—a standard visual representation of how dampers behave.
The force versus displacement curve represents the damping force created by the shock as a function of its displacement (at a given shock velocity). Everything above the X-axis (quadrants three and four labeled in black) represents a car’s wheel in rebound—the shock is extending. And everything below is the wheel in bump (quadrants one and two)—the shock is compressing.
Imagine you’re in the new 3 Series and you drive over a rock. At normal wheel ride height, let’s say you’re at the point “1" that Rothmiller drew on the plot 1a. As you hit the bump and the wheel goes up, “displacement” goes down as the shock shortens, but instead of following the normal “football” curve that’s characteristic of a standard shock, the slope gets bigger in magnitude, and there’s more resistive force countering the wheel’s upward movement.
Chinese chassis supplier BWI group offers a similar type of damper, and shows the additional damping force via the plot below:
The shock otherwise behaves like a normal damper. Thus, this lift-related rear dampers act to increase damping forces only towards the end of compression by up to 50 percent, which is why BMW also refers to it as a “compression stop.”
The front axle dampers, shown on the right, work in much the same way as the rears, except they increase damping forces during rebound only. The shocks consist of a smaller inner diameter-tube within the larger tube, as well as additional rings around the piston rod. The idea behind focusing on additional rebound damping at the front versus compression damping in the back, Rothmiller told me, is to alleviate any concerns of front wheel lift when hitting a bump, and to help keep the rear under control as its squats. He described the idea further to Road & Track, saying:
If you’re driving on an undulating road, the front axle helps you throwing out, the rear axle helps you diving in, or pushing through. We call it the “skyhook.” You’re hooked to the sky and never lose the horizontal position.
This standard, purely mechanical lift-related damper setup on the front and rear of the new 3 Series, he told me, helps with handling in undulating roads, yielding “much more body control.”
Rothmiller told me that two important parameters that could play a role in improving the new 3 Series’ handling are track width and weight. The first is a point we mentioned in our introduction of the new G20 generation: The distance between the centers of the tires on the same axle is up 1.7 inches inches in the front and 0.8 inches in the rear.
It’s worth mentioning that it’s not surprising to see a track width increase on a family car growing more mainstream like the 3 Series. After all, one of the selling points of the new car is its increased roominess, but nonetheless it’s a dimension that Rothmiller claims will improve the car’s cornering grip over the outgoing car, so it’s worth examining here. If anything, it’s interesting to see how engineers work in terms of limitations and tradeoffs in such a jack-of-all-trades car.
The track width gains, Rothmiller told me, are generally limited by a number of factors, including parking garage sizes around the world (particularly in Germany and Japan), as well as by desired vehicle turning radius, which has to be small, but which is limited by the elastokinematics (the degree of motion factoring in material elastic deformation) of the front suspension.
“On the other hand, it should not be too heavy,” he went on. “We have to be as small as possible and as wide as necessary,” he told me, pointing out that the 3 Series in particular can’t be wider than the 5 Series, and that there are architectural bounds.
As far as weight, Rothmiller says that—despite the car being 2.9 inches longer, 0.6 inches wider, and 0.5 inches taller than its predecessor—overall heft has dropped 121 pounds compared to a similarly equipped F30 3 Series. “Almost every single part in the car is lighter,” he told me, going on to say that it wasn’t just the big parts like the aluminum hood and fenders (which together save 33 pounds), but rather an “intelligent” use of materials throughout the vehicle.
BMW’s press release uses the same “intelligent” term to describe the material mix of aluminum and high-strength steels, which yielded a 44 pound lighter body structure. It’s worth mentioning that BMW’s U.S. media site has curb weights for the automatic-only 330i and 330xi as 41 pounds and 58 pounds more than the outgoing models, respectively. I suspect this apparent increase has to do with with the new car’s additional standard equipment.
But perhaps just as important as how much the car weighs is where the car weighs. In the fore-aft dimension, the new 3 Series maintains the 50:50 weight distribution of its predecessor, though in the up-down dimension, the center of gravity has dropped by 0.4 inches. And this, along with the wider track width, could make a significant difference in how a car drives.
“The wider track and the lower center of gravity allows you to drive faster around the corner,” Rothmiller told me.
Though Rothmiller didn’t get into the particulars of how track width and a low center of gravity affect handling, and though I’m far from an expert in vehicle dynamics myself, one way that improving these two dimensions helps is in reducing load transfer—the change in vertical forces acting on the wheels due to lateral acceleration—during cornering.
Reducing load transfer can improve overall grip because of a concept called tire load sensitivity. In essence, this term describes the fact that the coefficient of friction between a rubber tire and pavement actually decreases with load, or normal force, at a given slip angle—this is an inherent property of rubber.
A tire’s overall grip doesn’t decrease with vertical load—it still goes up—but the rate at which grip increases with vertical tire load drops due to that drop in friction coefficient, as is shown in the decreasing slope of the extremely simplified plot below (there’s a similar plot on the website Racing Car Dynamics).
Here’s how tire-load sensitivity relates to load transfer: In the drawing above, case one represents a 4,000 pound car driving through a turn with no load transfer whatsoever. Two thousand pounds of vertical load are exerted on the inner and outer sets of tires (for simplicity, we’ll just treat the inner and outer sets of tires as single entities). So the inner set of tires and the outer set of tires are both at coordinate “A” shown, a point that corresponds to a traction of “2" (an arbitrary number I made up). Total traction is simply the sum of the traction of the inner and outer sets of tires: four.
Case two represents a car in the middle of a turn where lots of load transfer has occurred. It shows the outer tires having more load than the inner: 3,000 pounds versus 1,000 pounds. The outer tires—now at point “C”—have more grip than before, at 2.5 (up 0.5), but their grip didn’t increase as much as the inner tires’ grip decreased to coordinate “B” (down one point). Thus, because of the significant load transfer, overall available lateral force generated by the car in the turn is reduced to 3.5 compared with 4.0 in the case of a car with the same load on all tires.
This is an extremely simplified example with made-up units, but the point is this that, because the rate of change of a tire’s maximum lateral force with respect to vertical load on the tires (the slope) decreases with vertical load, minimizing a car’s load transfer helps maximize traction in the lateral direction.
To understand how the new BMW 3 Series’ lower center of gravity and wider track width help keep load transfer down, it’s simple enough to imagine in your head that something low and wide is less likely to want to tip, and transfer load to the outside in a turn.
If that’s not intuitive, have a look at the free body diagram below or the one on Racing Car Dynamics. It shows a car in the middle of a left-hand turn, pulling a certain lateral acceleration, a. The forces acting on the car are the vehicle’s weight, W (which acts through the car’s center of gravity and downward), normal forces Ni and No countering the vehicle weight, and centripetal forces (represented here by tire friction Fi and Fo) which are countering an apparent centrifugal force, ma (the product of the car’s mass and its lateral acceleration).
If you’re not a fan of math, go ahead and skip to the equation below.
To understand how much load gets transferred to the outside wheel in such a turn as a function of lateral acceleration, track width, and height of center of gravity, we realize that the car is not rolling over in this scenario, and thus, the sum of all the moments created by the product of the various forces and their lever arms must equal zero.
Let’s look at those moments. There’s the weight of the car trying to turn the vehicle counterclockwise about the outside wheel (with a lever arm of half the track width since we’re assuming the center of gravity is in the center of the car), there’s the normal force acting on the inside wheel (with a moment arm of T, the track width), and there’s the centripetal “force” ma (with a moment arm of h, the height of the center of gravity) trying to rotate the car clockwise.
With a bit of arithmetic, you find that the vertical force acting on the inside tires during a turn is the initial vertical load, W/2, minus the product: the mass of the car times the lateral acceleration times the height of the center of gravity divided by the track width. In other words: it’s the initial vertical load minus mah/2t. If we want that acceleration to be put in terms of G, we realize that “a” in terms of “G” is simply “a” divided by gravity. Substituting that in, we find that the load transfer is directly proportional to the center of gravity, and inversely proportional to track width.
Looking at this equation, it’s apparent that both a lower center of gravity and a larger track width decrease load transfer in a turn, and thus—due to tire load sensitivity—the overall lateral force available in a turn are increased. This is why the new 3 Series’ lower Cg and wider track could make a significant difference in handling—specifically in available cornering grip.
This same principle of tire-load sensitivity is why the car’s 50:50 weight distribution is so important in maximizing grip, as Engineering Explained mentions here.
Chassis rigidity is something you hear a lot about in the auto industry, and for good reason. A floppy chassis doesn’t respond quickly to steering inputs, nor does it tend to ride or handle well, since the body’s flex effectively acts as another uncontrolled spring as the car’s suspension hard points are displaced due to elastic deformation.
BMW says the rigidity of the 3 Series’ body has increased by 25 percent compared to the predecessor, with gains as high as 50 percent in “certain areas.” Rothmiller told me that “certain areas” refers to the top mount on the front suspension and also the front subframe mount.
The result, according to BMW, is that the car’s suspension, particularly its performance suspension variants, can “go about their work in the most efficient way possible.” Rothmiller told me that, especially in the case of the M340i model and other performance variants with sharper springs, dampers, steering, brakes, and especially bushings, the stiffer body was key. “The old components would be, simply speaking, too soft,” he said. Using stiff bushings, dampers, and springs on a soft chassis, he told me, would not be fun to drive, nor would it be comfortable.
According to BMW’s press release, suspension spring rates on the M Sport and Adaptive M suspension are up 20 percent over the previous 3 Series “without loss of comfort.” Stiffer spring rates are good for reducing body roll in turns, but a stiff spring acting against a body that can deform is hardly effective or cushy.
The way that the company was able to score these improvements, Rothmiller told me, has much to do with changes to the way BMW did its modeling.
“We learned a lot about how to model and how to simulate with the computer the stiffness of the chassis and body,” he said, telling me the team used to measure body stiffness at the jack mounts, but now it measures stiffness from the road surface, through the tires, up through the chassis, and even to the driver, optimizing accordingly.
Rothmiller mentioned a number of other changes that could help the new 3 Series handle better, like a rear axle whose toe and camber are designed to be neutral over the wheel stroke to keep the back end from essentially steering itself, new tires, a new brake booster, and new brake hydraulics for the M-Sport brakes that promise a “sportier pedal feel.” Not to mention, there’s the available Adaptive M suspension which BMW says gets electronically controlled dampers with “new valves and an optimized control algorithm.”
But based on my discussion with Rothmiller, the big story seemed to be about increased body stiffness, improved steering geometry and directness, the new stock lift-related dampers, and changes to overall vehicle dimensions.
BMW says the new G20 3 Series’ ride, handling, and steering are better than its predecessor. “It has to be the ultimate sports sedan. There was never a question on that,” the company representative told me over the phone, strongly.
We’ll find out how all of this translates into the real world when we get time behind the wheel later this year.