datalogging interpretation

[5axis]

Okay let's say I have one. Now how do I understand and use the info like
track mapping and segments
G Vs time
G Vs diistance
Velocity Vs time
Velocity Vs distance

and suspension inputs and throttle and RPM and wheel speed and so on.

What do I look at and how should I use this information. I do not have a team of engineers to break this down while I cool off in the team lounge trailer with my umbrella girl and not much time between practice sessions.

Any good websites on how to use this info to make decisions?

[TD in DC]

Good Luck. I started a similar thread a few days ago with the offer of real data and I haven't gotten any volunteers yet.

Any Data Gurus out there? Data from VIR

[5axis]

I wonder if everyone else who dropped a $G or two is also sitting around too embarrassed to ask. I recall there were a lot of DL-1 users on PP. Maybe they just point at the screen with a beer after the day is over and tell their buddies "I was really flying there" LOL

BTW
My use is for a bike
aprilia 917

[TD in DC]

OT: Very cool bike. Actually, a sport bike rider almost killed himself into the back of my car today . . . seriously. I was driving at about 70-75 mph in the left lane of a two lane road with a 55 mph speedlimit (GW parkway). There were cars in both lanes, but I decided to get over to the right lane because someone was blocking the left lane (but there were cars about 20 yards in front of me in the right lane as well). I put on my blinkers and luckily I saw this guy coming in the right lane at about 130 mph . . . I swerve back into the left lane (I was about 1 foot over at that point), and about 0.5 second later he passes where I was. There is no way he could have stopped, and he was threshold braking to stop for the cars in front of me . . . Of course, he was wearing nothing but a t-shirt and shorts. Moron. I used to race dirtbikes, and would love to buy a ducati but my balls have shrunk since then so I will continue to race only Porsches.

[Jack Olsen]

The people using data the most are probably the ones who don't want to give away their secrets. But I'm slow enough to share. Problem is, I don't know a whole lot.

But here's a set of graphs (speed against distance, lateral g's against distance, and 'time slip' against distance) for two sessions from last weekend. This was when I was testing a wing I'd built against a standard ducktail. In the graph, the red line is the ducktail lap and the black line is the lap with the big effin' wing.



Sorry about the width. I hope you have a big screen.

And since neither of you are southern California drivers, here's a track map:



And here's a video of a lap

These three graphs are probably the cornerstone of the DL1 software. You can see differences in braking points, differences in acquired and maintained speeds, and differences in cornering forces -- all with the time slip feature, where you can see what each difference cost, in terms of a final lap time. It's pretty self-explanitory, but I labeled some of the differences, which will also help you orient it to the map.

[TD in DC]

Thanks Jack. Sorry if I am being dense, but I so do not get the time slip graph. Why is it not constant? How is it calculated? How am I supposed to interpret it?

Thanks.

Also, what can you tell, if anything, from the shapes of the various graphs (e.g., braking)

Finally, why on earth is the lateral g graph so spiky? Is it due to dirty data, or does driving technique contribute to the smoothness of the graph or lack thereof.

[924RACR]

Jack, excellent data to work with.

I work with data acquisition in the day job, focused on vehicle dynamics, so I'll share what I focus on.

First of all, to answer TD's question - accelerometer data can look pretty spiky if it's not smoothed/filtered.

I too don't know what the "time slip" represents, so can't comment. I know my AIM SW (software - for data analysis) has a graph that will allow me to compare, over the course of a lap, where one given lap gains or loses time relative to a baseline lap. This is excellent IMO for seeing very easily where you are gaining or losing time.

Mainly I focus on speed vs. distance - how fast am I at any given point. Let's pretend, for the purposes of discussion only, that the two shown laps are identical setup. Here's my conclusions based on your speed traces:

Front straight: Substantial difference in accel from the two laps. While your corner exit speed is higher in the darker lap, your speed at the end of the straight is equally slower. If we're comparing data from different sessions/days, I might wonder if you were making substantially more power in one session vs. the darker lap. You could also have a different aero setup - notably more drag in one - or maybe you got a good tow from someone.

I'd also want to look at the previous lap, find out why you were so much faster exiting the last corner onto the front straight. Of course, you've pointed out that you were running a bigger wing, so we can clearly see that the big wing really helps you get through that corner quickly, but you pay for it down the straight - giving someone else a tow! I'd want to review the back-to-back comparison and see how much time you gain in the corners over the whole lap, vs. time lost on the straight - it may be worth it. Clearly this track isn't like Monza! Or maybe you can dial back a little on the wing angle, keep most of your downforce, but improve your front straight speed. Otherwise you might be setting yourself up for being passed into 1.

Turn 1 - again, we can see the higher cornering/apex speed, presumably due to the wing. Looking at the slope of your braking zone, we can see you're also braking harder on the lap with the wing. Is this due to the wing, or driving style? As a result of that greater decel, you actually stop braking sooner, as well as higher corner entry speed. But the shape of the curve might suggest you could trail-brake a little more into the curve; I like to see a bowl shape to my corners, rather than a flat level point in the speed trace as you have here - just as you have done in turn 9. This would seem to imply to me that you can carry just a scotch more speed into 1, trail braking in. But it's not bad by any means!

Looks like you're trailing speed decently into 2 on the lap with the wing; kinda odd what happened in the ducktail lap going into 2, like you stabbed the brakes. Maybe the car stepped out on you? You lost a lot of speed there. And yet, your apex speed seemed to be pretty much the same with either wing. Something isn't right with that; either you should be going WAY faster with the wing, or what, I don't know, maybe the wing isn't that much of an improvement? Need to see more laps through there with each wing config.

Furthermore, the triangular/sawtooth speed trace shape shows some definite uncertainty about your appropriate speed with the big wing - perhaps you need more time to get used to it? Did the car step out there? All of a sudden, you dropped down to your ducktail speed. Is there a dip in the track there (I've never been out west for racing)? You do lose lateral g's there, could be due to a dropoff. I'd want to know. Still should be able to see the bowl or U shape, should be smoothed out more, looking at the track map in particular - nice big sweeper, should be able to really hang the car on the tires there.

Turn 3 and 4 - accelerating to Turn 3 in the lap with the big wing, looks like maybe you're overdriving the corner exit or sliding the car too much, because you actually _slow down_ on corner exit (as the lateral g's trail off), whereas the lap with the ducktail shows smooth accel and pickup of speed. Interesting to still see the big wing config overtake the ducktail in speed going into 3; if you hadn't had that slide or whatever caused the slowdown (presumably you lifting due to undesireable vehicle dynamics), you might well have had even another 5 or so mph going into 3. Do most of your laps with the big wing show this?

Continuing with 3 and 4 - doesn't look like you're getting/using any benefit from the big wing in 4. You're actually just as fast with the wing as with the ducktail in 4. Now, since these are low-speed, that would make sense - losing the aero effect. But why are you SO much faster through 3 with the big wing, which is only 10mph faster than 4 when running with the ducktail? I would expect less of a speed difference, if it's only a wing change. Newer tires??? Plus, note how 3 is more bowl-shaped with the wing vs. 4 with the wing? That would also seem to indicate that you might be leaving some speed on the table in 4 either way.

Turn 5 - Dang, do I wish my car would accelerate like that! Again, while you go in later to 5 with the wing, you've got that sawtooth deal going on again with the wing, maybe overbraking still? Would be nice to see more of the bowl shape there, but that may simply not be possible, with such a high entry speed and low apex speed.

Turn 8 - agree with your comments. Certainly looks great with the wing. If the ducktail lap does not have any traffic - still looks like you're doing well there. BTW, the laterals don't seem to support your thought that you might have hit traffic - you're still pulling high laterals, which I would expect to drop off if you're off your pace.

Turn 9 - looks quite good, except for the small bobble/dip in speed right before the lap marker; not sure why it's there, but it is in both traces. Generally the speed trace should be smoother, but maybe there's a track issue there?

That's how I would analyze those two traces - hope that helps by explaining what I look for and how I draw my conclusions??

[TD in DC]

Hey, thanks Vaughan,

Would you mind giving similar notes to my data in the other thread?

Mine shows what I looked like in April in much cooler temps vs. August where the temps were very warm but where I "think" I was driving better.

with respect to the spikiness of the g logs, apart from filtering, is there a way to distinguish between 'dirty data" and some variation due to driver input and/or track anamolies?

[TD in DC]

Also, here is the type of questions I have . . . what can you tell from the "shape" of various curves. I have attached an example drawing to show what I am pondering.



Is that a valid way of judging corner entry speed . . . or is a better way to look at max lateral Gs? Then again, is it possible to carry more speed through a corner by "minimizing" lateral Gs through your driving technique (e.g., better line)?

[5axis]

Yes, thank you. I'll need to spend some time trying to see what you are seeing, but at least I have a map to make heads or tails of it.

Question 2. Have you found it usefull on a raceday or is it more a post day analysis tool. Between wrenching and umbrela girls, I do not have a lot of time.

Lastly, I hope you don't mind a bike guy asking stuff here. Club level car guys are way ahead of us "two wheel cheap skates" with this technology.

[924RACR]

Yeah, I'll look for that other thread.

Re: spikiness of g-data - no, there is no way other than experience - knowing your track. Can build up a picture, once you have many laps and a light filter.

Your example speed traces - yes, I'd agree with those discussions. My usual failure is not starting my braking crisply. I also rarely brake at the limit, to conserve my tires - hate flat-spots. Your trace flagged as your typical indicates to me that you're overbraking. I fixed that by not trying to have the perfect corner entry, but rather focus on a smooth entry to the corner that allowed me to nail the apex perfectly. Since I'm entering corners at .7-.8g decel, I've got time to get just the entry, apex, and track out speed to beat others down the corner. Of course, the downside is I lose ground going into the corners, but no-one ever passes me coming out of corners. Now I need to improve my corner entry.

[924RACR]

Oh, as for the other question - no, I don't often have the presence of mind to deal with all of this on race day. But I force myself to try, and I do look at my data shortly after each race. If nothing else, I can usually find two important things: corners where I'm giving up something relative to previous good laps, and corners where I'm doing well. It's just as important to know what you're doing well as what you need to improve - if you're already maxed out on a corner, but don't know it, and keep pushing there, you're gonna go off!!!

I try to find 2-3 corners to work on for each race, try to improve my technique and find speed. 2nd place is nice, I've got a lock on it, but it's not as fun as 1st place!!!

[TD in DC]

Originally posted by 924RACR
Yeah, I'll look for that other thread.

Re: spikiness of g-data - no, there is no way other than experience - knowing your track. Can build up a picture, once you have many laps and a light filter.

Your example speed traces - yes, I'd agree with those discussions. My usual failure is not starting my braking crisply. I also rarely brake at the limit, to conserve my tires - hate flat-spots. Your trace flagged as your typical indicates to me that you're overbraking. I fixed that by not trying to have the perfect corner entry, but rather focus on a smooth entry to the corner that allowed me to nail the apex perfectly. Since I'm entering corners at .7-.8g decel, I've got time to get just the entry, apex, and track out speed to beat others down the corner. Of course, the downside is I lose ground going into the corners, but no-one ever passes me coming out of corners. Now I need to improve my corner entry.

yes, but I frequently get 0.8-0.9 decel going into the corner and then 1.1-1.2 through the corner. I wonder what is possible in our cars? How much more lateral load could a ITS spec 944 handle? 1.4? Seems a lot to me. Yeah, the theory indicates that there might be more corner entry speed on the table, but . . . since I pull a decent lateral G, is that really true? Interesting.

[924RACR]

1.2-1.3, depending on vertical loading - function of the tires. You get less than that on decel, due to the camber you have to run for cornering.

Yes, the speed is there. Remember, going fast isn't about maximizing your G-sum!!! But even with that proviso, you're looking at trading off g's in one direction for the other... and since you're carrying brake into the corner, you'll be starting your braking later (once you get used to it and the car set up for it). Your sawtooth traces kinda look like what you'd expect out of a "brake, then turn" cornering strategy - especially not needed in a 924/944.

[924RACR]

Oh, another point of eval during a race weekend - I'll often review my lateral g's pulled through a corner. Knowing, as TD pointed out, that the car should regularly be able to pull 1.1-1.2 g's, that gives me a quick sanity check of whether or not I'm pushing the tires to their limit on every corner, consistently. Sometimes I find I'll be slacking off in some corners and only pulling 0.9-1.0g's - particularly in the real quick (short) corners, or in the fastest corner.

Of course, that has to be taken with a grain of salt and some perspective too - I can't pull much more than .9g if I've got a little bit of brake decel too!