MoTeC, a Journey: Getting The Most From Data Analysis

First Release Of My Workbook!

First Release Manic ACC Workbook: Download Here
Lap Channel Report Google Spreadsheet v0.3: Link
Sections Channel Report W/ Dampers Google Spreadsheet v0.2: Link

This is my first release for my workbook! I have not finished the video explainer, but I hope to finally get it done this week (started a new job, very busy lately). In the meantime, please enjoy exploring my workbook.

As this is a first (unfinished) release, I do not have all track map sections completed, and I have not cleaned up all my math channels that I have collected over time. I will try my best to answer any questions or errors you come across, but please understand that my time is limited and I will try my best to address any and all concerns. Thank you!

 

Hi Manic, how is Manic's production . I'm so expecting with your video and your final blow mind workbook (there are a lot information which are impossible to decrypt)

Enviado desde mi M2102J20SG mediante Tapatalk

 

Video Explainer for my workspace!

Basics:

Thank you for your patience! I apologize as this was very much an impromptu recording and does not represent my best work. It is also only about the first tab, the basics tab.

Even just walking through the first workbook, it took me almost half an hour! I tried recording an explainer for every page and it ended up being 2hrs+, and I still wasn't finished

In any case, I will put these videos out piecemeal and see how you guys like it. Good luck and let me know if you have any questions

 

Seriously, thank you, manic. Everything you have posted on this thread has been incredibly helpful. <3

 

Thank you very much Manic

The only problem is i cant see Math Constant Quick Edit tab on the left
Does anyone have any idea how to get it?

Thanks

EDIT: Nevermind,i just needed to update the Motec

EDIT 2: I have it,but now it doesnt show me the thingy that you can change steering ratio

EDIT 3: I just needed to download the workbook again and it works now

 

Hey Manic, first of all, thanks for all the knowledge you're spreading
The workspace is amazing.
Can you explain a bit more what are Bumpstop_Damp, Bumpstop_Force and how to read these channels?
Also what's the difference between bumpstopup and bumpstopdown ride?

 

All of these channels are in some way describing how the forces created by the bumpstops are working on the suspension. Bumpstop Force shows the amount of force that is resisting the incoming force, which is why negative bumpstop forces occur on compression. Bumpstop damper force was probably added after 1.8, which details the amount of force created by the damping effects of the bumpstop rubber (not including the damper), and is simply added to the bumpstop force. This creates the hysteresis (or lag) in the suspension movement thanks to the velocity-dependent damping forces and makes the suspension less reactive and immediate like it was before 1.8 update. The bumpstop damp traces have a force/velocity relationship which you can try and surmise for yourself by plotting bumpstop_damp against damper velocity.

Bumpstop_up/_down indicate which direction the bumpstop is being activated after the initial suspension travel has run out. UP means compression, DOWN means extension/rebound. All of the bumpstop settings in the car setup page affect the compression side, whereas the extension force characteristics are not, for whatever reason. When the damping forces exceed the bumpstop forces, the gate closes and does not record the resulting bumpstop data ranges. This hysteresis either means that the suspension is no longer engaging the bumpstop, or some other effect is occurring which I'm not aware of.

There's still a lot of speculation here and I don't know if we'll ever get all the answers, but that's at least what I've been able to tell.

 

My confusion about the up/down bumpstop channels is that the down channel seems to activate near the end of suspension travel (extended), well out of range of the bumpstops. Does this have to do with the damping forces you mentioned?

edit: actually, looking more closely at it, i think it is tied to suspension travel. similar to how you gated the suspension data to show you where the line is where the bumpstop engages, i think you could do the same with the down channel at the bottom of the suspension. There's a consistent suspension value where the down channel activates.

edit2: edited your Susp Bump Travel expression to show the DN channel. Image updated.

edit3: im dumb. re-reading your post, this is exactly what you are saying. well, oh well.

 

Driver Performance:



Next video explainer on the driver performance tab is out! This was a rush job (and it STILL took over an hour to explain, but I managed to cut it down to 40 min because it is what you deserve), so apologize that it is not up to my usual standards. Unfortunately I will be out for the next several weeks, so I wanted to give you guys a little something to digest in that space (Attitude velocity post should be out soon).

Let me know if there's anything I can help with or if you have any comments. This stuff is complicated (as I make a lot of mistakes in this video!) and the data can sometimes be more confusing than illuminating. Take your time and don't worry if it doesn't make sense at first - drive different cars, tracks, etc., and just see how the data and the traces change between them. You'll get the hang of it soon enough. There will be more posts about car setup and the like after I return from my travels. Enjoy!

 

“If everything seems under control, you are not going fast enough.” - Mario Andretti

Yaw Rate And Attitude Velocity

So far we have talked about the understeer angle, which uses a car’s wheelbase, steering ratio, and steer angle to make some assumptions about the car’s balance. Of course, steering is not the only component that has an effect on the car’s rotation - we also have the pedals contributing as well. If we want a robust picture of what’s happening, we will need to delve into the car’s rotational speed, called yaw rate. So let’s get rotating!



Yaw is the orientation of the car around its vertical axis, like spinning a top. A car that is turning develops yaw, otherwise it would never turn. Using an accelerometer located at the car’s CoG, we can determine the change in yaw over time, referred to as yaw rate. If the car is driving in a straight line, the yaw rate would be zero deg/sec, but as the car starts turning, a yaw rate starts to develop that indicates how fast the car is turning around that vertical axis. This can be a bit tricky to think about - normally we wouldn’t think of turning to have anything to do with spinning (unless we lose control), but physically that is what is happening in conjunction with the forward momentum of the car. If you maintain a consistent yaw rate, you will produce a circle trajectory, and faster or slower yaw rates will produce smaller and bigger circles. Kunos have very kindly included the yaw rate as one of the logged channels, dubbed ‘ROTY’ (rotation of the y-axis).
We can compare yaw rate to the car’s angular velocity, which is functionally the same as the Ackermann angle but measured in degrees/second instead (just like with the Ackermann angle, the angular velocity is a theoretical value that assumes the rotational speed necessary to negotiate a given corner radius). When we superimpose both traces over each other, we have something very similar to how we compare steering angle to Ackermann angle:



Even though we make the same comparisons as we do for the understeer angle, there are some notable differences between the two. One, we see that a car builds a yaw rate in a positive or negative direction depending if it is making a left or right turn. The angular velocity follows along, and appears closer to the yaw rate than the ackermann angle is to the steer angle. It follows then that we are looking at a different kind of phenomenon here and cannot directly compare it to slip angle or steer angle. We will need to use a different metric to find out what we are looking at.
We can find the difference between these two traces the same way we do with the understeer angle, though with a slight variation. Whenever the yaw rate exceeds the angular velocity, this results in the car ‘oversteering’ - the body is rotating faster than the angular velocity. Whenever the yaw rate falls below the angular velocity, the body is rotating slower and is thus ‘understeering’. When we subtract these values from each other, we produce what is called an attitude velocity:



It should be immediately obvious that what we are looking at is fundamentally different from the understeer angle. Whereas the understeer angle looks rather benign and predictable, our attitude velocity looks like chaos and a lot of noise! Don’t worry though, we can still make some important observations using this channel.
First thing to note is that oversteer is in the positive direction, while understeer is indicated in the negative direction, the opposite from our other channel. The other important distinction is that what we are observing is not just the effect of the steering input, but all of the driver’s inputs, and then some. Hit a nasty kerb? The attitude velocity will show you, generally, whether the car starts losing front or rear grip. Throttle oversteer can also be observed in this channel - there is usually a jump in the attitude velocity trace alongside the throttle input. While these effects can be seen in the understeer angle, it is much easier to see the magnitude of these events with attitude velocity.
But why does this look so chaotic? Mainly because of that interaction I mentioned earlier, where the driver induces a yaw moment that then ‘pulls’ the car in that direction. But because of a variety of factors, the car does not always come along cleanly, and it may catch up quicker than expected, or it might hit a bump and start pulling the car in the opposite direction, and if that event is significant enough, the driver will need to compensate in order to counteract that movement. Ever notice how good drivers make very small steering adjustments while they are turning? In addition to testing out the limits of grip, they are also counteracting that ping-ponging in the car’s attitude to a manageable state so that the car can navigate the optimal line at the highest speed possible. This is why a lot of elite drivers tend to prefer a twitchy car - it magnifies this oscillating behavior and makes it easier to ‘feel’ what the car is doing before they commit to a course of action (Fernando Alonso utilizes this technique of making micro inputs before committing to larger inputs in order to gauge how the car will behave, something that probably helped him develop the necessary skill to drive the early 2000s Renault car to its limit).
Another important consideration is paying attention to how the car develops angular velocity. As I described earlier, positive attitude velocity indicates ‘oversteer’ and the negative values ‘understeer’, but this is somewhat of a misnomer. In order to develop the angular velocity necessary to corner, you have to create some oversteer, and in order to exit a corner, you have to create understeer to start reducing the angular velocity. It follows then that when looking at attitude velocity, the most obvious indicator that we are experiencing perceptible oversteer/understeer is the magnitude of the trace. Larger, fast and/or sustained spikes are a good indication that the car is experiencing significant imbalances. What is causing the imbalance is another matter - kerbs, puddles, flat spots, etc., can all create spikes (or lack thereof) in the attitude velocity. It requires an extra bit of judgment to determine if it is a setup issue.
We can utilize some additional terms to differentiate between ‘normal’ over/understeer versus problematic over/understeer:

Mechanical Oversteer: Positive attitude velocity
Mechanical Understeer: Negative attitude velocity
Perceived Oversteer: Conventional view of car imbalance, marked by the driver’s perception of too much turn-in and sudden increases in attitude velocity
Perceived Understeer: Driver’s perception of car not turning in enough, and indicated either by significant negative attitude velocity or lackluster positive attitude velocity.

As such there is no easy metric (like with understeer angle) to characterize the car’s balance using attitude velocity. Take a look at the difference between average understeer angle vs. attitude velocity:




As you can see, the regressions/trend lines are different - we’re not getting as much of a clear idea of what’s happening to the car’s balance over the course of a stint. We can try and better understand attitude velocity by looking at average POSITIVE values and average NEGATIVE values:




First, let’s examine the differences here. If we remember our traction circle, we can recall that it’s not a perfect circle, instead looking more like a squashed circle. This tells us that there is more grip being produced in the corner entry phase as opposed to the corner exit phase. This is reflected in the attitude velocity - more average positive values, less average negative values. We can then try to utilize these two measures to get us an idea how the car is behaving in these two phases. Mid-corner will not be as obvious, which is better observed by looking at the average attitude velocity in steady state cornering/coasting phases (easy enough to do with gating).

Here we examine one driver in the M4 and a different driver in the McLaren, including one setup that has a damper change. Notice how the balance of the car (at mid-corner) is uniquely different and of different magnitudes between them. This may suggest to the driver/engineer what direction they should consider for car setup/driving technique.​

What we see develop here is a much more nuanced and complex picture of car balance. Setup changes will always be an exercise in compromise, and it is important when making particular changes to our setup that we are aware of where these compromises will occur. Increased stability in one section of the track may lead to other imbalances elsewhere, so we can investigate these effects both on a micro level (yaw rate, attitude velocity, understeer angle, etc.) and at a macro level (lap time comparisons). More complex utilizations of yaw rate and attitude velocity involve comparing relative rates of stability (measured by a theoretical ‘smooth’ data trace) at all transitional stages, such as braking, turn-in, and so on. This advanced utility of the yaw rate is beyond the scope of this article but you can read about one such use here.

Simplifying The Attitude Velocity Channel

While you can garner all the information you need looking at the attitude velocity channel, it also helps to visualize how the velocity develops by integrating the channel. What’s produced on the other end very much resembles the understeer angle.



Due to how integration works, we only integrate for the individual sections. Therein we can observe how the yaw angle develops throughout the cornering phase, giving us a clearer picture of just how smoothly we developed our yaw angles. This is also another way to visualize the difference between corner entry and corner exit performance - more yaw is utilized at the start of a corner than during its exit, typically. If our trace falls below or above this threshold by the end of the corner, we know that significant issues occurred during this part of the stint.
Final thing I would like to mention is that attitude velocity is very sensitive to everything that happens on track. One of its closest corollaries is steer speed, so greater variance in our steering smoothness will inevitably show up in the attitude velocity trace. Aside from that, observe what happens on corner exit when applying throttle - there is a very sudden spike in negative attitude velocity as the car attempts to “straighten out” and reduce its angular velocity. Notice how the integrated attitude velocity channel moves when going over kerbs - does the trace slope downward, suggesting mechanical understeer, or upward, suggesting mechanical oversteer? Also take a look at wheel speeds, and how the attitude velocity is affected as various wheels slip. A lot of the car’s behavior in these circumstances can be observed in the attitude velocity channel! It takes some digging, but being able to observe different effects and their magnitude gives you an idea of how to approach particular issues you may experience on track, or to help find why the car behaved in a way that you did not expect.

Conclusions

Investigating yaw rate and its effect on the car’s balance is not a simple task, and its complexity also means that it may be an impractical way to quickly discern the overall balance of the car, especially in comparison with understeer angle. That said, observing this metric will lead to a very robust understanding of how the car behaves in ACC, and best of all, no need to change any values in the settings! It works best as a complement to understeer angle and I recommend using both channels to validate the other.
Given the relative complexity and difficulty of analyzing attitude velocity, I will try to provide additional examples and specific track situations to help demonstrate how attitude velocity works in future posts. As always, design your own tests and verify the driver’s observations with the available data! It is easy to get caught up in the numbers and traces, but it does not mean much if we are not methodical in our approach. Good luck!

 

Hi there! I am answering here in the comment in youtube, about what doesn't work in the spreadsheet.
I am using Google Sheets and I copied the Sections Report w/Damper v0.3.
I get the following in the Damper Histogram tab:
https://prnt.sc/Mz-lOhEVbqto
It is the same in the Zero Bin tab. The rest seem to be ok.

It is correct in your spreadsheet, so it is me, but I cannot seem to find why.

P.S. Sorry for the very late reply.

 

I will take a look at this and let you know!

 

I am going to guess as to the problem (works fine for me atm) but those tabs are sensitive to positioning of the individual data columns. For instance, if in MoTeC you added a columns AHEAD OF all the individual damper channels, the output will be incorrect (FLSB, FLSR, etc. must absolutely be in columns AN-BC for instance). Any additional columns in MoTeC should come after the damper channels, sorry to say (I would like to fix this in the future, but this is quite a difficult problem to circumvent). It can also be the case that if any of the columns referred to in the damper histogram tab are empty (such as the driver column) there will be an error. So if that is the case, simply put something under those columns in the data dump tab so that the graph can populate correctly. The reason this happens is because depending on how many driver names there are, the spreadsheet will automatically collate different data sets for EACH driver, which is pretty neat, but does require each column is filled.

Be careful not to fill down those columns, otherwise you are going to end up with a spreadsheet with millions of rows

 

Hi, there! It was the driver column that was missing. That fixed it for me, thanks a lot!

 

Installed i2 Pro

Followed steps to get ACC workbook into Motec

When I click the data folder it defaults to C:\users\GRENKE\documents\assetto corsa competizione\motek'

However there is NO USER called GRENKE so where on earth has it got this from???

So I have to have to navigate to the correct folder grrrr

Also when looking at Histogram properties how can I save the setting so I dont have to keep inputing them?

 

Save your settings by "saving" the workbook.

 

And if you wanted to know the roll and pitch gradient what would be the formulas to enter?

 

@Manic_Driver
Jesus, this is so hard core and I surprisingly finished reading.
But as new comers, I have a question:
when a car is jacking down like in stragiht line, why the front damper spend more time in rebound damping, and the rear damper spend more in bump damping?

In my head, it's like the car is nodding its head, and the front damper has more compression, thus it should spend more time

 

Congrats on reading it all! It is a very difficult topic to cover for sure, so I'll do my best to explain.

You have to remember that the histogram shows you time spent in a particular region. The damper applies force according to speed, so whereas a spring absorbs and releases energy by converting energy into distance (spring rate), the damper spreads that energy out over time instead. Dampers are uniquely designed, too, to "blow off" fast acting forces like going over a bump by letting the spring take the brunt of that movement, but at slower speeds like a transition into cornering, the damper plays an outsized role in slowing down the speed of that roll motion.

More time in rebound typically means the suspension is moving slower in the rebound vs the bump, and the suspension is moving less on extension. Imagine a bicycle pump that is easier to push down vs pull up - if you applied the exact same force in both directions, you would - on average - move closer to the down vector vs. the up vector, because there is more resistance when you pull up and it's easier to move down.

Now, this is only in the instance where there aren't other forces acting on the car. With GT3 cars, there is a significant amount of aero at play on long straights. The car is being pushed down as the aero increases, so naturally the suspension will start to move in that downward direction over time. This is a velocity, so with regards to the dampers, this effect will be represented by a bell curve that is slightly biased towards rebound or bump, depending on which end of the car you are looking at. High downforce setups push the front of the car down and creates a torquing action that lifts the rear end like a see-saw (the amount of chassis flex is minimal and not something we need to worry about too much). A 'balanced' histogram is not what we want to go for in this instance because we want to encourage that movement where the aero plays a big role, but we don't necessarily want that in low speed situations, so a balance needs to be struck. You could, if you wanted, separate the damper histogram depending on whether you are going through a low speed or high speed section to get the most out of the car, but in all honesty you will have to sacrifice one for the other, and a lower front end is typically what you want anyway. A "balanced" histogram might make more sense for somewhere like Monza, where you want to squeeze out every last bit of km/h out of the car, so you might try and setup the car so that the front and the rear compress in equal amounts. This is very tricky for us to do, though, because suspension movement ≠ ride height movement. We can only get a vague idea of what is happening to ride heights based on what the suspension is doing. Luckily, it is quite easy to test our theories by measuring top speeds.

This is all to say, the histogram is much more complicated than what has been explained previously by many others. Once again, at the end of the day, what matters most is how you the driver sense any difference, and whether it makes an appreciable difference to your lap times.

 

It's pretty close though, right? When would the ride height change without the suspension moving?