Can someone explain how to do this in Motec?
I've mapped the longitudinal G forces on the Y axis, and mapped Lateral G Forces on the X axis but I get something that looks like this:

Surely I brake more than that? Or is the bottom supposed to represent Acceleration?

Seems like there is data missing from the top, so surely I've got it wrong.

Any ideas?

Acceleration is on the bottom and it is to be expected that it's very dense here as most of the lap is acceleration.

It looks ok to me, I'd imagine though there's only one or two heavy braking zones, possibly low speed corners after straights. You can see most of your deceleration occurs as lift-off/gentle braking whilst cornering (0 to +0.5 range in Y-axis).

That's a fairly normal scatter plot in MoTeC. The top of the graph in this case is braking so a max of 2.5 g's is typical depending upon the type of car and the bottom is acceleration, also fairly typical values.
I've never used the trend line (straight diagonal line) so I have it disabled in the "properties" screen. It might be useful, but I've never studied it to find out.

Thanks guys. The data is from a run in the rf3 F3000 cars at Northamptonshire. There isn't really a lot of hard braking, and considering these cars don't generate too much G force I suppose it could make sense that I don't see a lot of action in the upper G's.

I'll pop off the trend line - thanks for the idea Ralph.

What are you trying to analyze? What do you expect to get out of the analysis? Mapping the G forces as you have done doesn't really tell everything about a tire's "traction circle". You may not have been pushing the tires to the limit. Another mapping would be tire slide percent vs the length of the G force vector vs vehicle speed...that will tell you what your maximum G force is at the given speed. It would require a 3D graph...not sure if MoTeC can do those. And, another problem is that the velocity vectors provided through data acquisition are not even close approximations of the actual vectors...follow them as a vehicles moves in a straight line and you will see them change...even going from large positive to small negative numbers...totally bogus...they add up to the scalar velocity, but not the direction vector.

That all sounds very promising Steve. But i'll take your word for it and scrap the traction circle.

I had just hoped to get a rough estimate as to how i was doing in terms of driving at the limit, on a lap to lap basis. I've just read about the traction circle and how it can show how close to the limit you are, all the way around, an entire lap.

However, I have heard about motec not doing a great job of representing the right values for a traction circle type graph. So I guess that ideas gone.

EDIT: so is there a decent way to discover the coefficient of friction for a given tire? I know that the variables of weather/track/etc change, giving different limits, but could motec be of any help in this area at all?

EDIT AGAIN: ah, and I might have to point out that I'm no good with all of these fancy terms yet, so can we get any possible responses in plain english ? :-)

I'll put some more thought to it this weekend. If the data acquisition numbers are accurate for G forces, tire load, and, rotational velocity, and slide percentage then yes...the coefficient of friction for the tires can be calculated. What we would look for is the lateral G force at the very instant that slide percentage changed from 0 to get static friction, then while there is a slide percentage get dynamic friction...there's also rolling friction that can be calculated from rotational velocity and longitudinal G force. Divide those #s from the tire load and you will get the coefficient of friction.

However, the simulator may not be using actual physics...it really depends on the calculations that the sim is using. So, in the end...it may be better to do as Tim says and rely on the lap times to judge overall performance. Use the lap time matrix to determine where you need to improve. Then when you want to sqeeze that extra .2 seconds out of your lap times start looking at your turn-in, brake, and throttle application points; plus, steering percentage, cornering speed, slide percentage, and tire load. Analyze those #s and alter driving where you aren't maxed out on something. But, the most important data...the angle of the tires to the road...is screwed up in the sim...it would tell you if you are turning the steering wheel too much or not enough.

Steve,
Would a graph of combined g's (calculated from Lat and Long g's)be of any use for what Andrew is asking? Not a scatter plot, but a line graph. If I recall we used to use that trace along with Lat and Long g's to determine if the driver was braking too early and not carrying enough speed into the corner. I've been meaning to try it in MoTeC but your comments about the inaccuracies in the sim physics compared to RL now make me wonder about its' usefulness.
The combined g graph would show a gap or drop when the long g's dropped under braking, but the lateral g's not picking up soon enough.

As i understand it and use it, a traction circle is good for telling you if you are using the setup to it's limits.

the more circular, the more you are getting out of that setup.

i also use it as a quick visual to assess whether i'm getting more overall useful grip from a change, in specific corners, by layering before and after laps.

Ralph: Yes...if the longitudinal and lateral G data points are accurate then the length of the sum of those two vectors would be very important. And, yes...a line graph of the length (magnitude) of the total G forces vs time would be a good graph. Integrating (calculating the area under that graph) over a complete lap would be a good indicator of improvement. If your total G forces over a lap are increasing then you are pushing the car towards the limit and should continue to do whatever you are doing, but if the total area drops off then take a step back. Of course you can get inaccurate data if you are swerving around on the straights, but it would be a good tool to use.