I process measurements taken during a rig test to compile (for each set-up) a simple linear model of a vehicle. The beauty of that is that I can then assemble carpet plots of how the Performance Index (PI) might be expected to vary with changes to vehicle parameters. The results should be taken with a pinch of salt: the vehicle itself is not linear and the PI addresses vehicle control & not necessarily lap time, although there is frequently a strong correlation between the two. Nevertheless, the results can be quite instructive. Here is an example, extracted from a rig test of an open-wheeled "aero" vehicle, with D/F set for around 100 kph.
This is a plot of PI as a function of damper "strengths". The black axis shows changes in front damper strength & the red axis shows changes in rear damper strength. The heavy green lines show current settings, and each axes covers the range from 0.5 to 1.3 times the current values. On screen, the plot is "live" & the controls panel allows the plot to be manipulated (apologies for the lack of notation - I was not up to the task of creating moving legends). The "Current PI Value" box contains the current value and, importantly, estimates of the front & rear sensitivities. Finally, the "Minimum" box contains rather more dubious estimates for "best" settings (as multiplying factors) and expected minimum PI value (also indicated by the heavy blue vertical line) - dubious because they will only be correct when the linear model accurately represents the vehicle. In the present case, the dampers had already been optimized (more or less).
The next is a similar plot, but this time showing how PI would be expected to vary with changes to spring stiffness. The results are not untypical for a "aero" vehicle, suggesting that control would benefit by reducing both springs. Interestingly, however, it suggests that the rear springs should be adjusted more (0.7*front & 0.5*rear), but that a front spring change has the greater sensitivity.
This is again similar, but this time shows how PI might change with tyre rates. The results suggest that stiffer tyres would help - again this is not untypical for an "aero" vehicle. I does suggest, however, that a change in stiffness split would be helpful (higher rear, lower front). This is consistent with the spring results (probably, actually lowering the rear spring rate would improve the tyre balance). This is the same plot, but this time shown as an overhead view - a poor man's contour plot. Arguably, it shows the most efficient was of reaching the valley would be to increase rear rate by 4%, and to reduce front rate by 4%. This could be achieved (in part, perhaps) by increasing rear tyre pressure and reducing front pressure. Increasing camber also tends to reduce tyre rate.
The point of this long introduction is to present this plot. Although it looks similar in form to others, and was produced by the same model, it is a bit different. Here, the black axis shows the effect of changing c.g. position (measured aft of the front axle) and the red axis shows the effect of changing pitch radius of gyration (pitch inertia). In numbers, it suggests that the current c.g. position (at 59.5% w/b aft of the front axle) was rather too far aft. The "optimum" would be around 56.7%. The good news is that PI does not depend much on pitch inertia, so moving ballast should work. Making that change should mean that the existing spring and tyre pressure splits would be OK.
To summarize, the vehicle weight distribution, suspension settings & tyres, all work together to achieve optimal control. Deviations can be accommodated by making changes to other settings, provided these are small.
- Login or Register
No account yet? Sign up