Suspension tuning Lever ratio

  • Thread starter gti_sdn
  • 35 comments
  • 15,661 views
63
Australia
Australia
gti_sdn
I am trying to tune a car and want to get a base to start from. I have been looking at information on the natural frequency, which is used in-game.
I have looked at car specs and weight, including weight distribution, to try and work out a spring rate. From there, I can calculate the natural frequency.
The only issue is I am missing one important value.

So looking through the manual and it states
https://www.gran-turismo.com/us/gt7/manual/carsettings/02
The effective spring rate value changes depending on the suspension type and lever ratio.

The lever ratio is what I am missing. Depending on the value of this, my natural frequency changes. For example, for a given car and weight with a target spring rate, if the lever ratio is 1:1, the natural frequency is 2.9, but if that ratio is 1.1:1, my Natural frequency should be 2.65.

So if I were to assume the lever ratio is 1:1 and set my natural frequency to 2.9 and the lever ratio is, in fact, 1.1:1, my spring rate is way too hard!!!
I am guessing there is no way to find this information out in-game or if it changes for a given car if you upgrade to a different suspension.
 
You won't be able to find lever ratio in the game, although it is possible that someone who can access the underlying game code might have found parameters for it.

But in terms of tuning in GT7, not knowing the lever ratio isn't really a big deal... Unless you were trying to create a simulation of real life performance springs with known spring rates, in which case you would need to know the lever ratio to work out the "wheel rate" and nat freq. If that is the case, its better to forget GT7 and try to find real life data on the real life motion ratio, although this is not always easy to find. For common cars like an MX5 or Silvia there are probably guys on enthusiast forums that have measured the lever ratio on their own cars. Then you can work out a nat freq and transfer that into GT7.

In GT7 it's still possible to start with a nat freq and work out the "wheel rate" (the effective spring rate as would be measured at the wheel). i.e. how many Newtons of force needed to move the wheel up by a distance. In real life the wheel rate arises from the spring rate and the lever ratio, but it just acts like an imaginary spring is perfectly above the wheel and this can be used when comparing front to rear stiffness. You can imagine the wheel rate a being an imaginary spring acting with a lever ratio of 1 (the imaginary spring is exactly above the tyre).


Within GT7 you have car weight and F/R weight distribution, so you can calculate corner weight. There will be some unsprung weight for the wheels, which we don't know, but you should be able to get a pretty close estimate for a corner sprung weight. Then from the game's nat freq you can calculate wheel rate.

Kw = 4(pi)^2 * F^2 * M

Kw = wheel rate in N/m (you may wish to then convert to N/mm, or kgf/mm for convenience)
F = nat freq in Hz
M = corner mass "corner weight"

Likewise, going the other way to work out nat freq from a specified wheel rate

F = 1/(2(pi)) * SQRT(Kw/M).


So if you want a 60/40 spring stiffness split F/R you can imagine wheel rates of, say 3kgf/mm and 2 kgf/mm = 30N/mm and 20 N/mm = 30,000 N/m and 20,000 N/m, plug them in with your corner weights and get the frequencies.
 
Yes been trying to work on known setup information for a particular car. Similar to what you have done, I have worked out the difference between the natural frequency of the front and rear and then moved the front and rear stiffer or softer, keeping the difference between them the same.
For unsprung mass, I have used 50Kg for the driven wheels and 45Kg for the undriven wheels as a guestimate.

It seems to have worked so far.
 
This is an interesting thing I hadn't considered when using nm figures from other games for GT3 cars to work out a hz natural frequency figure... could this be why some cars still feel a little 'off' even when I've matched the suspension values?

@Nenkai I'd be interested in those figures for basically all the Gr3 cars.

@gti_sdn I worked out a system to convert the range of suspension values given in ACC for the GT3 cars in there to the equivalent cars and Natural Frequency values for GT7. This didn't include the concept of the lever ratio though, just the weight distribution and spring rate for each corner of the car. I'd be interested to know if incorporating this other variable and reaching a new hz/nm equivalent figure would make things better for those slightly off-feeling cars?
 
Last edited:
Lever Ratio is stored for each car, yes. I have no issues providing them - they're all values between 0.5 and 1.5.
That would be amazing, as I have been working on some setups, and a slight change in Motion ratio could make a huge difference in natural frequency.
It would be nice to see if the game's motion ratio is the real car's actual value.

@danardif1 Yes, motion ratio should be a variable considered. Who knows if ACC and GT7 use the same especially if GT7's Gr3 cars are not based on any real-world data?
 
That would be amazing, as I have been working on some setups, and a slight change in Motion ratio could make a huge difference in natural frequency.
It would be nice to see if the game's motion ratio is the real car's actual value.

@danardif1 Yes, motion ratio should be a variable considered. Who knows if ACC and GT7 use the same especially if GT7's Gr3 cars are not based on any real-world data?

I've had some success in recreating comparable feel and laptime using the available suspension values from ACC for the cars it shares with GT7 (so cars like the R8 LMS Evo etc). Whilst I can't say for sure whether the motion ratio is modelled in ACC though I would suspect it is, the suspension values used are sourced from the OEMs and operating teams for those cars so at least there is some realistic source to that information. So it would be very useful for my conversion process to know and better understand the motion ratio variable to help make my GT7 setups better.
 
Sorry to add to the list but

... any scaling on the anti roll bars rather than the generic 1-10 level? I can't work out if they're a completely standalone spring constant or based on a percentage of the spring nat freq (like the dampers are).
 
@gti_sdn so I'm understanding this correctly, with the example of my R8 LMS Evo and I have a front suspension NF figure of 3.83 based on 153000nm over 264.6kgs of weight on the front end, with the lever ratio of 0.75, does that mean I have to cut the NF figure by 25%?

This gives me a figure of 2.87... which then isn't a selectable figure in GT7... Could this be why the car can feel a bit off? That it's actually too stiff?
 
Last edited:
@gti_sdn I've figured it out. I'm using to work out the ratios now https://www.calculator.net/ratio-calculator.html?t1=0.75&b1=1&t2=4.08&b2=&ctype=1&x=0&y=0

The point about the unsprung weight is something I hadn't considered either. I wonder the if the physics engine factors in unsprung weight or it just uses the total mass and weight distribution in it's suspension calculations...


Okay, so as I now seem to understand the concept, it seems to me that you would apply the lever ratio to the rate of the spring you're aiming to get an 'effective' spring rate, so a spring rate of 144000nm with a lever ratio of 0.871:1 produces an effective spring rate of 165327nm, which I then put through my conversion process using https://amesweb.info/Vibration/spring-frequency-calculator.aspx to produce a Natural Frequency number. The numbers I've been getting all fit within the GT7 system so I'm going to get them all worked out for my current data and then test them out when I get home after Christmas.

They all seem to be a lot stiffer which fits in with the issues I've had with the car before, where the Audi pitches forward and backwards far more easily than it should seem given I'm matching spring rates from the same car in ACC. I'm hoping that in a few days this will prove a major breakthrough!!!
 
Last edited:
Late response, took me a bit to figure out how to format this.


Thanks for that it is appreciated.
Seems like there is a min and a max but what's the DF??
a quick look at the data and it does not appear that any car has a lever ratio that changes so will stick to using the DF.
 
@gti_sdn I've figured it out. I'm using to work out the ratios now https://www.calculator.net/ratio-calculator.html?t1=0.75&b1=1&t2=4.08&b2=&ctype=1&x=0&y=0

The point about the unsprung weight is something I hadn't considered either. I wonder the if the physics engine factors in unsprung weight or it just uses the total mass and weight distribution in it's suspension calculations...


Okay, so as I now seem to understand the concept, it seems to me that you would apply the lever ratio to the rate of the spring you're aiming to get an 'effective' spring rate, so a spring rate of 144000nm with a lever ratio of 0.871:1 produces an effective spring rate of 165327nm, which I then put through my conversion process using https://amesweb.info/Vibration/spring-frequency-calculator.aspx to produce a Natural Frequency number. The numbers I've been getting all fit within the GT7 system so I'm going to get them all worked out for my current data and then test them out when I get home after Christmas.

They all seem to be a lot stiffer which fits in with the issues I've had with the car before, where the Audi pitches forward and backwards far more easily than it should seem given I'm matching spring rates from the same car in ACC. I'm hoping that in a few days this will prove a major breakthrough!!!
Looking at the data and going from the R8 weight of 1235KG, I get the following using the default spring rate from iRacing (160n/mm front and 190n/mm rear)
2022-12-25 08_56_44-Suspension Calculator - Google Sheets.png
I had to guess the unsprung weight. The rear is heavier due to the driveshafts. Not sure of the wheelbase, so the recommended rear frequency can not be calculated for a less pitchy car.
I have not had time to jump into the game to see if these values are obtainable.

Another thing that I thought might affect the Motion Ratio for the road cars was changing the suspension from stock to racing.

Once I get this spreadsheet tidied up, I will share it for people to use.
 
Those figures seem a little low by my previous workings. For example, for the Audi R8 LMS at it's base weight of 1235kg, and a weight distribution of 42-58, for 153n/mm (one of the selectable options on ACC) on the front I get a Natural Frequency of 3.87hz, and then with the BOP weight of 1260kg 3.83hz. This is before including any lever ratios.

With the ratio of 0.75:1, what figure goes in what side of the ratio?

I've been using the initial spring rate as the left hand side of 0.75:1

153 as 0.75:1 = 204n/mm

204n/mm with a mass of 264.4kg on the front axle = 4.42hz

This is without any rough working out of unsprung weight of course, but it's much different to what you are reaching. I will say from my in-game experience of both games and the same car, that with the suspension settings matched the way I was doing it before, the GT7 version felt too soft still and there was too much pitch on the front and back compared to the ACC version, so the higher figure I've got now with my way of incorporating the lever ratio seems to make some sense to my eyes (I'm away from home so can't try on the game yet)
 
Hi all - I started by searching for a car list as I'm looking to do some time trial laps with different cars at different PPs after tuning them, and wanted to get a car list

I went into a rabbit hole of data and have found threads like this. Can I ask, for things like the lever ratio and the list of cars, where is that source data being pulled from?

I am aware of https://www.kaggle.com/datasets/prasertk/gran-turismo-7-car-list but I don't see all the info listed there which is also listed on sheets like

Is there a primary data source that has all of the info on cars?
 
The source data from this post https://www.gtplanet.net/forum/threads/suspension-tuning-lever-ratio.412121/post-13912553 was pulled directly from in-game data by @Nenkai
What other information are you looking for?
Personally, my ideal list of data in a spreadsheet for tuning would be

Make, Model, Group(roadcar/Gr1/gr2,etc), Motion ratio, wheelbase

other than that information, the rest is variable, weight, weight distribution etc

It would be nice if anti-roll bar settings had any real-world value behind them if we could get that, but I would highly doubt it, same with sprung vs unsprung weight.
 
I’m looking for a “feed” of that data, in one table, so when we get updates the data can be refreshed

the data from the Reddit spreadsheet is sufficient for my needs, and default tyre would be nice

If other data can be appended then great, but it seems there’s multiple sources… if @Nenkai can pull game data, pulling it once for all cars and having a format that can be refreshed after updates would be amazing

For my purposes I have:

  • calculated bhp/ton on stock cars
  • divided that by PP
  • looked to identify the cars with the best bhp/ton for each PP bracket (I’ve just used 50PP increments)

Then I’m driving them at Spa, seeing what they’re like, and thinking about how to tune them (ie if they’re heavy, do weight reduction, if they’re already spinning wheels then focus on controlling the power rather than adding to it)

Then if I think they feel good out the box, I’ll tune them, set a time at Spa, and stick it on the spreadsheet

I bought the “tuning calculator” from a thread on here, that makes it quick to setup the suspension for a balanced setup. What I might do for my ‘best/favourite’ cars is append those settings to the sheet… but don’t know yet

It’ll take a while but my concern is that when the data changes, what I’m basing my priority cars on will change, and I may miss a gem

And as more cars get added, I’d need to add them manually. Not the the end of the world of course, just a faff
 
It would be nice if anti-roll bar settings had any real-world value behind them if we could get that, but I would highly doubt it, same with sprung vs unsprung weight.
I think in GT6, someone managed to get physical units for these in N/mm or whatever. The generic 1-7 settings spanned a range but it was often different for the front and rear axles. Would be great if the info was there for GT7 as then we could set them vs the spring stiffness which you can derive from the nat freq.

My gut feeling is the range of spring stiffness adjustment in GT7 is more than the range of ARB, probably about double.
 
Well, if it was in GT6, then fingers crossed it is carried over. It would also be interesting to see if it is there, and if the range changes depending on the suspension installed (Stock, street, hight adjustable, fully customisable).
I tried replicating a bunch of TCR cars as close as I could using known information on actual spring rates used and some guestimation, and they came out ok.
I know anti-roll bars is more of a driver preference and a fine tuning tool but really wanted people to have a good base to start from

https://outatimesimracing.wordpress.com/tag/tcr/

Spring rates were all worked out using the supplied motion ratios, weight, weight distribution and wheelbase.
 
Hi Nenkai,

Would you be able to provide an updated list for the new cars added recently please?

Many thanks
I know there has only been a few cars added since you last updated the list @Nenkai but would it be possible to get another update?

Also, big ask, I know but are wheelbase and front and rear wheel track included in the data?

If so is it possible to get the list formatted like so:

Front Motion ratio, Rear motion Ratio, Front track, rear track, wheelbase

Appreciate what you do 👍
 
Hey Nenkai!

Would it be possible to get an updated list with the new cars added please mate?

When you get time…

Thank you!
 
Back