MoTeC, a Journey: Getting The Most From Data Analysis

Ciao, sto cercando di portare gli ammortizzatori della McLaren 720s, sul circuito di silverstone, alle percentuali che hai consigliato come base di partenza, 34% compressione/estensione lenta e 16% veloce, al posteriore non riesco ad arrivare a quelle percentuali, cosa sbaglio? Per arrivare a quelle percentuali devo impostare le compressioni/estensioni quasi al massimo mentre quelle veloci al minimo. Grazie a tutti per l'aiuto.

 

In questo caso, la sospensione posteriore è troppo rigida per essere gestita dall'ammortizzatore. Ammorbidite il tasso delle ruote o gli ARBs e dovreste essere in grado di ottenere risultati migliori.

Reinum asked a great question! What happens if I try to reduce dampening in the rear suspension, but my damper values do not change? In that case, the suspension is too stiff for the damper to have enough of an effect. Reducing the wheel rate or ARBs should help the suspension become more compliant and you should be able to reduce your hi-speed damper values from there.

 

Ok, grazie per l'aiuto, magari posto qualche foto delle regolazioni delle sospensioni e ammortizzatori.

 

Ciao, carico qualche immagine del setup e del motec

 

Sembra un pò troppo portare a quelle percentuali gli ammortizzatori posteriori, compressione/estenzione lenta al massimo dei clik, viceversa quelli veloci al minimo quasi a zero, lo stesso non riesco a portare l'histo a le percentuali 34% e 16% . Come mai?

 

Non devi per forza avere i grafici a spigolo. Pensaci. Se hai tutto quel carico all'anteriore, con tamponi belli spessi, il posteriore è costretto ad assorbire tutti gli urti e gli avvallamenti da solo, per dire.
Non è che non si muove abbastanza negli spostamenti di carico, semplicemente l'anteriore è incollato a terra e il posteriore assorbe.
Non è necessariamente una cosa negativa, se ti trovi.
Guarda il tuo SUS_TRAVEL al posteriore: anche in rettilineo dovresti avere il posteriore che fa su e giù tanto, ma non per questo vuol dire che lo devi cambiare.
Almeno credo.
Puoi verificare in WHEEL_SPEED se perdi velocità per questo motivo.
Gli istogrammi ti dicono semplicemente che succede.

Se li vuoi a spigolo per forza, puoi provare ad abbassarla dietro e mettere tamponi meno spessi davanti e più spessi dietro.
Ma non è detto che riesci a trovare il giusto carico in quel modo.

 

Ciao, si capisco quello che dici, ma non è che voglio portare il grafico a spigolo, nella guida di Manic_Driver scrive di partire da una base di 34% compressioni/estensioni lente e 16% compressioni/estensioni veloci. Adesso a prescindere da come và l'auto, volevo sapere come mai non riesco a portare gli histogrammi posteriori a quelle percentuali per poi da lì lavorare sul setup per riuscire a guidare bene.

 

Secondo me non e' questione di ammortizzatori. Nella guida c'e' scritto anche che in base al carico aerodinamico e come esso viene generato, cambia la dinamica. Da qualche parte questo carico dovra' pur finire. prova pressioni piu' basse dietro XD.

 

sto provando di tutto alzare abbassare altezze, bumpstop, ala a 0 ala a 8 via di mezzo... niente, la questione non è che voglio portare per forza a quei paramentri, ma a quali fare riferimento se non a quelli 34% e 16%? Si vede che la mclaren non può arrivare a quelle percentuali di lavoro al posteriore. Booo

 

le impostazioni vanno bene?

 

Tutti al 25% significa che la sospensione spende tempo equamente in tutte e quattro le fasi. Non c'è un traguardo da raggiungere. Al più serve a sapere se va tutto come uno si aspetta.
Dipende dall'applicazione.
Non esiste un valore ottimale generale, ma valori ottimali per ogni specifico utilizzo.

 

Scusami ancora, scrivo sempre delle percentuali 34% e 16% per quello che ho letto nella guida di Manic_Driver allego screenshot

 

.

 

Scusate, ho sbagliato a rispondere, ho messo un punto perché non riuscivo cancellare il messaggio.

 

In questo caso dobbiamo concludere che la sospensione posteriore è impostata in modo tale che qualsiasi tipo di smorzamento non sarà in grado di raggiungere la distribuzione normale. Questo va perfettamente bene - ogni auto ha caratteristiche di sospensione uniche che possono creare eccezioni alla teoria dello smorzamento. La McLaren ha un layout a motore centrale, quindi una certa rigidità potrebbe essere necessaria per mantenere il trasferimento del peso gestibile.

Invece di preoccuparsi della distribuzione 34/16, si consideri il monitoraggio dell'altezza del bidone zero. Prendete nota della maneggevolezza dell'auto a diverse altezze del contenitore. Se le altezze dei cassonetti non cambiano, allora molto probabilmente la rigidità posteriore dell'auto è troppo rigida perché le regolazioni degli ammortizzatori possano fare la differenza. In un post futuro guarderò la rigidità del rollio che esaminerà la rigidità di varie auto.

Reinum came across a very clear exception to the theory I presented in my post about dampers. In the case of the mclaren the rear suspension even at its softest settings refuses to react to any damper changes. The simple answer to this is the rear suspension is just much stiffer than the damping rate, so no amount of clicks will change the histogram. This does not entirely mean the dampers cannot still have an affect - you will want to track your zero bin height (-5m/s - +5m/s) and see if adjustments have any affect on its height. If even the zero bin height does not change, I would simply move on to other parts of the car to adjust the rear's behavior.

I will have to reiterate, that the histogram tuning method is a theory - you always want to take note of your on-track experience before tuning.

 

ho capito, volevo portare a 34/16 come base di partenza per poi lavorare a mio piacimento, allora dovrò regolare gli ammortizzatori con dati a mio piacere, cioè...se non riesco ad arrivare al posteriore ai 34/16 di percentuale, a quale dati fare riferimento? A mio piacimento? Mamma che casino

 

Ciao a tutti, rieccomi con qualche domanda...con l'aggiornamento 1.8 per dati motec rimane tutto invariato, Manic_Driver la tua guida resta sempre valida?
Per gli ammortizzatori McLaren hai provato come modificare?
Grazie come sempre.

 

Hey guys, I had been following this thread and came back to it, would it be possible to keep it in English so I can still keep up with it. @Manic_Driver thanks for all of this and also replying in both lingo, much appreciated.

 

Thank you and yes, I think I'm going to stick to English for now

So to address reinum's question and also a few other things: the 1.8 update has made a significant physics update in the bump stops, so that now the bump stops have some damping to them. This is going to to make our damper histogram move a little slower, and I hope to address this in a future post about roll stiffness. But it should mean that any 1.7 setups might require less damping overall, assuming they were correctly damped before the update. There's a lot more going on than I can say right now but I look forward to addressing those concerns!

As for the rear damper settings, I'm going to repost this from discord which was a similar discussion:

The Lambo/Mclaren/etc. has a mid engine setup, so how the suspension behaves is going to be a lot different. The rears, by and large, are going to be moving a lot faster because of the amount of weight the rear suspension is carrying. No two way about it, trying to get the 'ideal' distribution with the parameters I suggested will simply not be possible without some extreme compromises. Instead, a better way to look at the histogram would be to just increase your lo speed hi speed range. 50 is a good alternative. Check your average bump/rebound speed to get a sense of what might be a good range. That might seem strange, but as I mentioned earlier in the article about rally dampers, sometimes we have to adjust our histogram parameters to make it useable as a reference.

The main key is to understand why histograms are used and then suit it to your purposes. The whole point of giving the 34%/16% numbers is so people can understand what a normal/gaussian distribution is defined as, despite every damper article mentioning it but never explaining it. Some cars are already in this ideal range, like the M6. All the mid engine/rear engine cars, however, do not fit a 25mm/s lo/high range at the rear, and need a higher velocity range to fit that normal distribution. Please do not confuse this with "moving the goal posts": we are looking to meet the dampers on its terms, not ours. And as I mentioned about filtering and motion ratios, we have to be careful in accepting the data as sacrosanct because we simply do not have access to all the information necessary, so we can expect some degree of variability in the numbers. In the end, driver feel matters the most.

I hope to write an article about tracking grip levels, because that helps us get closer to finding proper damper settings. 1.8 kind of threw a spanner in the works as the data all looks different now and testing just became more time consuming, but the principles still remain the same as is how to test and track these changes.

For the time being, when in doubt, go softer.

Why, again, does increasing damping rate increase time spent in that region?

One of the big myths with dampers is that increasing damping in a region decrease time in that region. This is for the most part wrong and here's a simply analogy why.

Drop a marble into a glass of water. How long does it take to get to the bottom? Now drop a marble into a glass full of honey. How long does it take then? Is the marble spending more time going fast or slow? Now, limit your range to 5 seconds. What would a histogram look like for each glass?

In the water glass histogram, you would see a small bump in the fast range and then the zero bin shooting up because the marble has hit the bottom of the glass pretty quick and is now not moving. If you dropped it into the ocean, the histogram would just look like a super long tail until it hits a terminal velocity. This is what happens with less damping - a histogram that is wider and shallower but can also have a tall zero bin if it hits some kind of extreme (i.e. bottoming out).

The honey glass would have taller bins in the slow speed range before it hits the bottom and the zero bin shorter than the water glass' zero bin. In fact, the marble might hit a terminal speed due to drag and stop accelerating, thus causing a certain velocity bin to become the tallest in the histogram. This is what happens when you increase damping - you spend more time in the slow range and too much damping might cause it to stall on a particular velocity i.e. the dampers are damping too slowly. This is why a skew towards rebound, for instance, means the car never really is able to get back up to its original ride height and can potentially jack down if rebound damping is too high.

And yet, a lot of people still misunderstand this about dampers. I admit that it's not easy to conceptualize at first, and it doesn't help that these misconceptions have existed in real life racing for a long time as well because let's face it, most of us would rather change a few clicks on a damper than have to change out the coilovers.

But us sim racers are not allowed to use that as an excuse. We have the data, and we can do a lot of testing to see how these things work within the game.

 

Adjusting Damper Histogram Parameters: Mclaren 720s @ Silverstone Case Study

As it has been quickly pointed out for those trying to get a handle on dampers, the parameters I suggested of 34/16 split between low and high speed ranges does not always fit if we use the 25mm/s range. So let’s walk through one of these instances and let’s see what we can do to still be able to utilize the histogram in a useful way.



Many will be familiar with this problem if you have tried to adjust dampers for any of the mid/rear-engined cars. Weight, as I mentioned in my original post, will widen and flatten the dampers due to the limited suspension travel and the amount of force needed to get the rear moving. The problem is further compounded because the suspension needs to manage this weight so the suspension has to be stiff to handle this weight. Thus, we cannot just soften the rears using less aero or softening the springs. We will be making way too many compromises, and it’s impossible anyway with the changes that we are allowed to make.

As such, we need to manipulate the histogram in a way that we can still use it. In the screenshot above, I have circled the average speed of the rear dampers in bump and rebound. These numbers are what we are going to use as a reference point to start adjusting our low/high speed range. I like to keep things even, so any value close to a factor of 5 is my preference, given that it fits our bin sizes as well. So, let’s choose 50mm/s for this example (you can change the boundary in the properties menu).



Now our histogram numbers are starting to look more like the normal distribution numbers. We could if we wanted increase the windows range (-300 to 300mm/s for instance), increase bin sizes so the graph look more like a normal distribution, but this is good enough for our purposes. It’s enough to see how the damper speeds are distributed and if there’s a skew of any kind. In this case, we have a skew towards bump, most likely due to aero.

But say we want to emphasize our mechanical grip? In that case we want our histogram to look more even. I’m going to go a little softer, say -2 slow bump, and let’s see what the histogram looks like afterwards.



Now it’s looking a little more even! And the numbers are corresponding as well. We could increase both slow rebound and bump by one to get us closer to our 34% value if we really wanted to do some super fine tuning. You will notice that making this change has also affected our hi-speed percentages, both in bump and rebound. This is one of the main drawbacks of adjusting our range, as it puts us outside of the range that lets us work high and low speed ranges independently. Still, we are now able to see how the damper clicks affect our new range. It saves us from going crazy making adjustments to fit an arbitrary value like 25mm/s.

Ultimately what we are looking for is some kind of balance on both sides of the histogram, and it is easier to see and manage this balance when we have an appropriate set of windows. We could also start testing if we can narrow our speed range, say 45mm/s rather than 50mm/s? How do the damper clicks operate in the narrower range? Can we get the dampers to move a bit slower on average without significant compromises? The dampers are actually capable of reducing this average by quite a bit if you are willing to take the time and experiment. Now we have a method to try and fit our dampers to any kind of distribution we wish.

My next post will delve into tracking grip across all these situations. This will give us an idea if our damper changes are actually doing anything. Stay tuned!