SAE CO414 Applied Vehicle Dynamics

I haven't heard of this course, though I do live about 15 minutes from BMW's SC location and have sent a note to my boss...As a vehicle engineer it excites me to see to see people nerd out about this stuff like I do. 

AngryCorvair - are you based in the GSP area? I think it'd be neat to have a GRM meet up at Cars and Coffee now that the weather is getting nicer...

morello159 said:

I haven't heard of this course, though I do live about 15 minutes from BMW's SC location and have sent a note to my boss...As a vehicle engineer it excites me to see to see people nerd out about this stuff like I do. 

AngryCorvair - are you based in the GSP area? I think it'd be neat to have a GRM meet up at Cars and Coffee now that the weather is getting nicer...

Nope, I'm in the Detroit area.   We run this seminar twice per year at the BMW PC in Greer.  More info on the course at SAE C0414 Applied Vehicle Dynamics

Mazdeuce-Seth wrote:

The car is close to neutral so the front and rear axle were both near saturation from a lateral grip standpoint. Adding longitudinal force always takes away your ability to have lateral force. Stabbing the brake saturated the front grip before the weight could transfer, locked the fronts, and we sailed off in understeer hell. Applying the brakes firmly but gently allowed the weight to transfer, loaded the front, unloaded the rear, and we saturated rear lateral grip and spun the car. Brake bias plays into this too, probably heavily enough that AngryCorvair is going to step in and correct me about the dynamics. The end point though is that how you apply the brakes, and the rate of weight transfer has dramatic effects on the outcome.

You nailed it, so no correction needed.  I will just throw out some book-ends to consider.   

First, consider a car with no front brakes.   cornering near the limits of adhesion, it will always get loose (ie seek a larger slip angle on rear tires) when the brakes are applied, because the longitudinal request can only be satisfied by the rear tires.

Now, consider a car with no rear brakes.   in the same situation, it will always push (ie seek a larger slip angle on the front tires) when the brakes are applied, because the longitudinal request can only be satisfied by the front tires.

The reaction of cars with brakes at all four corners will be somewhere between these two book-end conditions.   Where exactly on that continuum will depend on front and rear tire grip, foundation brake balance, jounce and rebound damping, etc.

Bold font applied to "get loose" and "push" above because these are the types of terms we use to describe transient conditions.   Remember, per SAE definitions, "understeer" and "oversteer" only apply to steady-state cornering.

In reply to AngryCorvair :

SAE definitions are 100% going to screw with how I describe things from now on. Thanks for that. 

There were 12 modules in the class. I'm going to discuss three of them in total. This the second. 

A car rolling along on a flat surface (with no significant toe issues) has a little longitudinal slip to counteract air resistance and all of that, and no measurable lateral slip in the tires. We know we can't generate lateral force until we get slip, right? So let's do a lane change maneuver. 
Turn your hands, turn the wheel, the steering shaft, rack and all that. Unless you're driving a 1968 Ford or anything Buick made prior to 2014, that is fairly quick, but some of the movement is taken up by rubber in the system and this varries from car to car. Once the movement gets to the front tires we deal with the fact that they're squishy monsters and there is a measurable time between when the car starts to turn them and when they get turned far enough to have generated enough slip to get the front of the car moving laterally. This is measurable with whole seconds or very large fractions thereof. Big amounts of time. So now the front is moving and that generates a yaw in the car. Only now, ages after we've given the signal to turn, the rear starts to do the squishy dance through the tires and generate slip and the rear of the car is generating lateral force. Finally the whole car is turning. This can take multiple whole seconds. And you can feel it! 
We took the lowly 340s out and did a lane change maneuver with the goal of trying to listen to the car. We drove and rode in the back seat while other students did it. Even in the 340, which is no Buick from a steering response standpoint, had input and phase 1 (front axle) and phase 2 (rear axle) as distinct noticeable events. This was eye opening to me for two reasons. 1. I learned to listen to something in the car that I hadn't been able to before, that's cool. 2. Slaloms. Some cars are good at slaloms, and others are terrible, just horrible. I've always attributed this to shocks/squishiness, and that's certainly part of it as does affect the steering impulse, but a large part, and maybe even a larger part is the relationship and phase lag of phase 1 and phase 2 steering. Aside from tire selection and suspension changes, there are things we can do to change/shorten the phase delay between our hands and the rear wheels generating lateral force. We can reduce the squishiness in the rack and any steering shaft parts. We can add static toe to the system so that we have tires that are generating some slip angle and carcass flex before we even start our turn, and probably other things involving bushings and other stuff. We can also take our cars out to a very very very safe spot and do a lane change and feel it. Get used to feeling it. Use that as a metric for evaluating cars and our changes to them. Ever wonder why manufacturers do a double lane change in testing? I always thought it was to simulate the "don't run into the kid" situation, and it is, but the phase 1 phase 2 steering behavior is very much tunable and recognizable characteristic of a car. Want to make your next 3 Series feel like your last 3 Series? Want you make your new Kia feel like a 3 Series? You better be thinking about this. 

I'm going to search around the web for a plot of this steering response. I'm trying not to steal images from the class, so this space may be populated with an image soon. 

Thank you for writing this up, seriously.  I'm having a great time reading it, although I have to fight the FSAE nostalgia in order to concentrate.  

 Does the class go over anything having to do with loose surfaces, like, say, gravel?  With gravel tires?  Or snow with snow tires?  I think I need to find an excuse to take the course.

In reply to ¯\_(ツ)_/¯ :

C0610 (that's C zero six one zero) Applied Brake Controls is hosted at a winter test facility in northern Michigan, primarily on packed snow (deformable) and ice (not deformable).  Neither class run on gravel or sand.

In reply to ¯\_(ツ)_/¯ :

The class is as big as the questions you ask.  From a class structure standpoint you're dealing with constant mu tupe surfaces. They do talk about differences in weight transfer on low mu surfaces and split mu like the shoulder of a road. Most of the extra discussion about weird surfaces occurred when we were talking about control with ABS and ESC and how the cars dealt with it. I will say that it would be difficult to find yourself in a room with more expertise in this area from and OE standpoint. All of these guys have countless hours both behind the wheel and behind the keyboard dealing with the full range of surfaces that cars drive on. 

An interesting/fun exercise to do that really accentuates the phase lag from front to rear axle is to steer back and forth at the same frequency as the phase lag. On tires with low lateral stiffness it's pretty easy to do, as this frequency is low, and you can actually get the car to "pivot" back and forth about the center while driving down the road. 

mazdeuce - Seth said:

Now, I always thought pitch, dive, whatever you want to call it, was part of this and in fact it's really not. Weight transfer is all geometry and the only way to decrease it (if we assume wheelbase is constant) is to lower CG either through lowering the car or taking weight off places above the CG. Pitch is important, but only so far is is makes your suspension do all sorts of wonky things that mess with your contact patch and bump steer and what not, the weight will transfer either way. Dynamics. 

This paragraph is glorious and expresses clearly an idea I've struggled to explain.

I'm not necessarily condoning this idea, but it might not hurt to have a cooler labeled "Questions for AngryCorvair" filled with a lubricating beverage at the Challenge. The only real question is whether that cooler should be larger or smaller than the one labled "Questions for stafford1500"

dclafleur said:

mazdeuce - Seth said:

Now, I always thought pitch, dive, whatever you want to call it, was part of this and in fact it's really not. Weight transfer is all geometry and the only way to decrease it (if we assume wheelbase is constant) is to lower CG either through lowering the car or taking weight off places above the CG. Pitch is important, but only so far is is makes your suspension do all sorts of wonky things that mess with your contact patch and bump steer and what not, the weight will transfer either way. Dynamics. 

This paragraph is glorious and expresses clearly an idea I've struggled to explain.

One of my biggest takeaways from the class was the idea of separating suspension movement from vehicle dynamics. There is a cool equation in our book that uses the term USG or Understeer Gradient. That throws tire cornering stiffness, camber thrust, aligning torque, lateral load transfer, roll steer, lateral force compliance steer, and the entire steering subsystem together in one term. Then you do something really cool, you define understeer and oversteer as steady state responses and all of those things collapse to one number that you can back out by driving in a circle. What that told me is that for the sake of vehicle dynamics, or rather learning vehicle dynamics, none of those things are individually relevant. What is relevant is how the car behaves as a whole. You understand that first, and then you go and mess about with the pieces. 

I vote for the larger one for Stafford1500....

However, I do share.

Also, reading along and enjoying the discussion.

As far as standards go don't get everyone confused with camber vs inclination angles... right Angry?

Wonderful stuff, thanks very much for writing it up!

In reply to stafford1500 :

We can probably get away with just one cooler, as long as there's some Gatorade in it.

I am unfamiliar with "inclination" vs "camber".  I'm only familiar with inclination as Steering Axis Inclination aka caster.  Help me out here!

Mazdeuce - Seth wrote:

Understeer Gradient...  That throws tire cornering stiffness, camber thrust, aligning torque, lateral load transfer, roll steer, lateral force compliance steer, and the entire steering subsystem together in one term. Then you do something really cool, you define understeer and oversteer as steady state responses and all of those things collapse to one number that you can back out by driving in a circle. What that told me is that for the sake of vehicle dynamics, or rather learning vehicle dynamics, none of those things are individually relevant.

The classic example of how the individual relevance of a change bows down to the overall vehicle-level effect, and what makes the phrase "all else being equal" at once meaningful and meaningless, is increasing front bar diameter on certain FWD cars.

Individually, increasing bar diameter causes an increase in lateral load transfer (ie weight transfer) on that axle.  We know that increasing weight (ie normal force) on a tire lowers the tire's efficiency at producing lateral grip (in fact, it is exactly that property of the tire which makes bars useful), that axle will have lower total lateral grip due to the unequal distribution of normal force between the two tires on that axle, and therefore (say it with me) "all else being equal", if this is a front bar, it should move the USG toward "more UNDERsteery than it was with the smaller bar."

However, in the real world, increasing the bar diameter also limits relative travel of inboard and outboard suspensions across that axle, and that might allow the tires on that axle to generate more (or less) camber thrust and/or more (or less) roll steer and/or more (or less) lateral force compliance steer, and the contributions of those things to the overall USG might result in a vehicle that is "more OVERsteery than it was with the smaller bar."

If you get the impression that I dig this stuff, you are right.   USG is my favorite module of the class.

and at the risk of being all man-crushy, i'm going to go ahead and say that Mr Deuce is doing a fantastic job of making this discussion less engineery and more gearheady.

AngryCorvair said:

In reply to stafford1500 :

I am unfamiliar with "inclination" vs "camber".  I'm only familiar with inclination as Steering Axis Inclination aka caster.  Help me out here!

There's also the kingpin inclination axis (KPI below). Similar to caster but in the xz-plane and could be confused with camber. 

morello159 said:

An interesting/fun exercise to do that really accentuates the phase lag from front to rear axle is to steer back and forth at the same frequency as the phase lag. On tires with low lateral stiffness it's pretty easy to do, as this frequency is low, and you can actually get the car to "pivot" back and forth about the center while driving down the road. 

Yes, this is highly entertaining when you're driving behind your friend doing it in a 1982 Mercedes 300D.

In reply to morello159 :

but of course!   thanks for refreshing my memory on that.  I was really drawing a blank but that figure straightened me right out.

I am loving how mazdeuce explains this so far. I have understood it myself for decades, but could never explain it in a manner that actually transmits the knowledge to a neophyte, and I have tried numerous times. If this can't, said neophyte is a hopeless case. Kudos to the max.

Aw shucks. You guys are either being super nice or you're just trying to convince me to take even more weird classes. Either way, I approve. 

The last thing I'm gong to talk about is the ABS and ESC stuff. More than any other part of the class, this is where having four smart guys in the room to ask questions paid off. 

We all know about ABS right? Not much to talk about aside from system bias and old Ford rear wheel only ABS and pulsing time vs. wheel inertia and the fast that we're not really looking at speed but the change in speed over time (a max equals mu peak anyone?) and the fact that super sticky tires may well be outside the boundary conditions that your system has. And that wasn't the interesting part, the interesting part was ESC. 

Electronic Stability Control being mandated came about because people kept tipping over and the data showed pretty clearly that ESC reduced that. Yaw sensors got cheap and the rest of the hardware was already in the car with the ABS system anyway, so by 2012 everyone had to have it. That's a good thing. For most of us ESC is voodoo. When we try and do anything fun it shuts us down like a cranky 4th grade teacher. No fun allowed. But what is it really doing? 

The heart of it is the yaw sensor. It's measuring stuff in degrees per second (or probably radians because engineer) and comparing that with wheel speeds and steering angle and other inputs and trying to figure out what's going on. If it sees you yawing in any way that is outside it's happy place it shuts things down. How it does that is the cool part. Lets say for instance you're cruising along and through no fault of your own the rear swings out to the left in an epic drift. 

The car thinks about this very rapidly and grabs some brake on the left front to induce a counter yaw moment to bring it back in line. It's probably also killing your throttle at a prescribed rate that is likely slower than chopping it. If you understeer it does something similar but it grabs the inside rear brake to induce the yaw moment. 

Why did I find this so cool? Because they let us back out onto the skid pad to play with it. We got to drive around with things off and on. We were able to feel how the car was catching our intentional stupidity and how incredibly good it was at that. Carrying this forward to real life and the car looks and what you're trying to do and helps you. Too fast into a corner and understeering? Here, let me help you turn. Oversteering because the throttle is super fun? Let me grab that and bring it back in without you heroically doing a giant tank slapper into a tree. The tire testers said the real joy is to take a car out onto a snowy surface and let it understeer. Where we've been learning for years to slow down and unsteer the wheel to gain back grip, in this case you can crank the wheel harder, let the car yaw with ESC and it will use the grip is has to turn the car more. Once is gets the car around then you unsteer and go on your merry way. When we talk about the magic of current FWD cars and how they turn when they seem like they shouldn't, this is what I think is happening. Hot (or smoking) brakes in the rear confirms that they're likely creating yaw that the steering isn't able to create by itself. Also, modern cars that use voodoo are super duper unhappy when you take away that rear wheel yaw ability with a stiff rear bar and lifting, or nearly lifting the inside rear wheel. The hardware is on the car. The OE's are using it in this way already. Who's going to step up their rallycross game with performance tailored yaw inducing braking? The possibilities are neat. 

In reply to mazdeuce - Seth :

good explanation, but i hate that "uncorrected" graphic because it shows the CG tending toward a *smaller* radius in oversteer, and we know this doesn't happen because lateral accel = V squared over R.  because a max equals mu peak, at the vehicle level it doesn't matter which end of the car loses grip.  the end result is that the radius of the vehicle's path must *increase*, ie V squared over a larger R.   i don't blame you, because i know that graphic has been floating around for years.

i know someone will say "but i've seen cars spin off to the inside of the turn".  and i would counter that with "yes, because the vehicle was simultaneously losing V, at some point in the event V dropped enough that V squared over R fell back below the mu peak value, and at that time the vehicle went exactly where the tires were pointed."

mazdeuce - Seth said:

The hardware is on the car. The OE's are using it in this way already. Who's going to step up their rallycross game with performance tailored yaw inducing braking? The possibilities are neat. 

This is pretty much what brake-based "torque vectoring" differentials like those in the Focus ST are doing already, only with the front tires instead of the rear. Even more interesting are cars that combine this with real torque vectoring rear differentials. The SH-AWD system Acura employs on the MDX, for example, has a rear differential that can selectively over-drive one rear tire while braking the front inside. Acura's hybrid SH-AWD system is interesting as well, in that it uses electric motors to selectively add or reduce torque to a given wheel (front axle on the NSX, rear axle on the MDX). 

With Ford's recent V8 hybrid patents, I've been having dreams of a V8 mustang with front e-motor torque-vectoring and an active torque-vectoring rear differential... If ever they wanted to build a GT-R killer, that would be one way to do it!

In reply to AngryCorvair :

Yea, it's hard to find diagrams that get everything right without stealing from the class book which I'm still trying not to do. But your point is valid, if you lose grip at one end then your center of mass swings out to a larger radius. That's kind of the whole idea behind why you lost grip, because you"ve exceeded the whole grip/radius/speed thing. 

In reply to morello159 :

I'm sure they're all playing with it. A couple of weeks ago I got a ride in a new Veloster Turbo-N. We went into a slow corner and my highly calibrated butt told me that there was absolutely no way he was going to come close to holding the line but the car did some serious magic and did. It's all really cool stuff and it's going to change how we think of handling in the very near future. 

Interesting stuff.  I'm starting to get a "grip" on what the info is providing.  A legend explaining some of the symbols may be nice.  I'm glad I didn't lose you guys at turn 1.