Measured cg height

In your example, you have removed the front wheels from the car for the lifted measurements. Obviously if the front wheels were installed, they would be completely carried by the lift straps and would not affect the weight measured at the rear wheels. As only the rear wheel weights are needed for the height calculation, the removal of the front wheels does not affect the calculation.

I have made similar measurements on my own car but not used your sling/hoist method. Lifting one axle 400 mm is a very nervous activity. When the front axle is lifted vertically, the rear axle rolls forward, but must remain on the scale. Alternately the rear axle can be blocked but then the front lift needs to follow the resulting arc.

An interesting experiment would be to measure the rear weights at several different front lifts along the way to see how much the calculation changes (i.e. how sensitive the method is to raising height). Your scales presumably read in whole pounds or half kg and are probably accurate to one count in the display. It is also instructive to do the calculations at plus or minus that error to see how sensitive the method is to weight errors.

Ballasting with a driver seems dodgy! Ballast weights would need to simulate the cg height of the driver to be useful. Probably neither is really important as the vertical cg of the driver is likely very close to the vertical cg of the complete vehicle, and being a small fraction of the total mass, can only have a very small effect on the result. In any case, the effect could easily be calculated after the fact by the same methods pilots use to calculate aircraft cg. (i.e. summing masses and moments, then arm = total moment / total mass).

The effort required to do this measurement is large, and the risks involved are not at all insignificant. Also the results are very sensitive to all the potential sources of error in measurement. Hopefully most people will be discouraged from trying, but I really applaud you for providing this example.

As you have noted, an outright estimate is almost as useful as a measurement. SAE 1999-01-1336 (which can be found on line) provides an interesting empirical approach to estimating cg height of road cars, that some may find helpful for a first go at an estimated cg height.

OK, without seeing the actual proceedure, I can assume this involved lifting one side of the vehicle to check the lateral transfer of the mass, and hence loading of the outer wheels/tyres? There are numerous issues with this method, some of which you mentioned.

What I have advocated for years is to take a simple load for each wheel on weight plates, then lift the front by plating the plates on a support of known height, then checking the loads recorded. At this point it may be wise to take the effective wheelbase (saves a calc') measurement. then repeat with the rear lifted.

With these measurements, and some simple vector analysis, the load transfer due to moving the CoG can be found and, from that, it's position.

By taking both the front and rear readings one can average out the affects of fluids moving in the fuel, oil, and any other tanks.

Oh, regarding the driver's mass, as one of the chaps pointed out in another forum, there's usually a 'helpful' body that can be placed in the driver's (and co-drivers?) seat(s) if the actual driver isn't available. I wouldn't agree that an actual body isn't required as not only can the driver be a significant percentage of the total mass - >10% is certainly possible - but much of that mass will be a significant height above the ground surface - add a co-driver and it'll be even more important.

Unfortunately measuring c of g as Tim demonstrates is a tricky task at best and I don't argue with any of the points you've raised. Significant care needs to be taken to ensure that the front of the car is safely and securely supported and that the rear does not move off the scales. We found it takes a few attempts to find the 'sweet spot' for everything in order to get a workable result. The ballast approach for the driver is unlikely to be perfect however it's certainly good enough to be useful.