using logged data to adjust high speed dampers?

[samsepi0l]

Does anyone use the logged data to adjust the high speed dampers? I have tried this, but not got it to work sucessfully.

Sometimes you can see when the car hits a height transition on the track in the suspension travel values. You can see the car get light which makes it unstable on one end. I am wondering if you can look at this to determine if a corner of the car is in "bump" or "rebound" and then soften or stiffen the damper to correct for the lack of stabilization.

[samsepi0l]

guys - someone must be good at adjusting the dampers with data. I have been messing with f1perfview to copy the data into excel and I want to write a macro that will tell me which damper settings I should start with.

please share your thoughts and what works for you

[DannyDarwen]

I can't fetch the data at the minute (an issue with the new x86 mod maybe), but would you not set the fast bump / rebound dampers from how the car reacts going over kerbs etc?

With the squiggly lines, you could probably try to reduce the impact with +- bump / rebound.

Sorry, I know that doesn't answer your actual question..

[samsepi0l]

Well, I don't race with x86 gp2, I use dosbox ECE. (I might switch to this if they ever get it functional with all of the carsets/tracks and mods that I use)..

I started saving lap data in GP2, and I started using f1perfview to review the data. With this it gives you the option to make a CSV file that can be opened with excel or libreoffice.

I raced Adelaide the other day and there is one corner where I have problems with the car snapping loose. It is the "brewery bend". This goes back to something we all know about gp2. The car likes to snap loose on corners that have height transitions in them.

I can normally take this one flat out if I get the line just right, but I will get into trouble if I put too much steering input in it. Here is a graph of my suspension travel and steering as recorded an done with f1perfview:



[I'm sorry that the colors are not consistent with the normal colors gp2 uses.]

I had a chat with chatGTP AI about this, and it made the following comments about this:

The Dip in Rear Suspension Travel (Near 1890):
This indicates that both rear suspensions are approaching full compression (close to 0 remaining travel). This is likely a result of weight shifting to the rear as the car crests the hill under throttle. The sharper dip on the left rear suggests it’s hitting its limit faster, which could destabilize the car.

Difference Between Left Rear and Right Rear:
The left rear appears to compress significantly more than the right rear, creating an imbalance. This could exacerbate oversteer as the rear becomes unsettled.

Comparison to Flat Surfaces:
On flat sections, you’d expect a more even distribution of travel between the left and right rear, without such abrupt dips, further confirming the crest amplifies this imbalance.

Adjustments Based on This Interpretation

Given the data now reflects remaining travel rather than actual compression, the following tweaks should help:

Fast Bump Dampers (Rear):
Slightly stiffen both left rear and right rear, focusing more on the left rear. This will resist rapid compression and help keep the rear suspension from hitting the stops too abruptly.

Fast Rebound Dampers (Rear):
If the rear is unloading too quickly after hitting the stops, slightly soften these to allow smoother extension.

Balance the Left and Right Rear:
The imbalance suggests the left rear bump damper should be slightly stiffer than the right to prevent it from compressing excessively and destabilizing the car.

[DannyDarwen]

Theres' a lot of information I can take from this graph. Here is what I can see:

So, you are accelerating up the incline towards Brewery Bend (a high speed right-hander). As you add steering input, the left rear compression (bump) increases due to load being shifted in that direction, and the right rear increases rebound. On the graph, it is showing remaining suspension travel left before it bottoms out (at zero). This shows the yellow line (left rear) drop - indicating its under compression, and the green line (right rear) rises - indicating its under rebound.

The bigger gap between the Rear-left and Rear-right suggests that you are running a softer setup at the rear; The lines for the 'right and left rear suspension travel' separtes at a faster rate than the lines for the front suspension, which also suggest a softer setup at the rear; The way the rear rebounds after the 'dip' suggests these settings are softer/faster.

Suddenly, midway through the corner (probably just before the apex), all four corners compress. The left rear looks like it compresses slightly after, to a point where it bottoms out for a period of time. I'm guessing this unsettles the car.

You start to look to straighten the steering wheel ever so slightly, and all for corners are in rebound at the peak of the hill. As you start to straighten up completely, the load is sent towards the right hand side, especially at the rear, before the car stabilizes itself along the straight.

Thoughts:

You could try to reduce the chances of the left rear bottoming out by either stiffening the rear or increasing the ride height. The consequence of this is that it may increase the likelihood of oversteer, so you may need to adjust the front to find balance again.
It would be interesting to see what the ride height telemetry is showing with the suspension travel. I'd also like to see wheel spin data.

The answer might be that there is no solution.

I will try to replicate this, as I do like this sort of problem-solving with telemetry

[samsepi0l]

Danny - thank you for your reply, I saw it the other day but I have been busy so I didn't get a chance to write back.

I also saw the spot where the left rear was bottoming out. I was embarrassed that I missed this even when I was just reviewing the data in the game. I think the best way to fix this might be just to eliminate some of the left rear packers since I have plenty in there and there is no issue with the plank rubbing in the back.

I want to start sharing my logged data like this with AI so it can help me understand and tune the setup better. This is something that was black-magic to me when I was a kid and it would feel good to master it.

[DannyDarwen]

I studied this a couple of years ago and used AI to understand it and create macros for MoTec software (and also using F1perfview).

A couple of books I found helpful were:

Analysis Technique for Racecar Data Acquisition - Jorge Segers
Making Sense of Squiggly Lines - Christopher Brown

I haven't spent time looking at telemetry in quite a while, but I'm hoping to get back into it..

[samsepi0l]

I studied this a couple of years ago and used AI to understand it and create macros for MoTec software (and also using F1perfview).

A couple of books I found helpful were:

Analysis Technique for Racecar Data Acquisition - Jorge Segers
Making Sense of Squiggly Lines - Christopher Brown

I haven't spent time looking at telemetry in quite a while, but I'm hoping to get back into it..

Awesome that you know what MoTec is. I wanted an aftermarket ECU for my Evo IX (now sold) but could not switch due to needing the factory ECU for emissions testing.

[DannyDarwen]

Yeah, the i2 software can be used with racing sims like Assetto Corsa, ACC and GTR2.. that would be good to use with GP2 if we can export log files to .csv . If not, F1perfview is still very good