ROAD TESTING PROBLEMS

Firstly, different people have very different ideas about how to test fuel saving devices. These range from simply logging a vehicle's MPG record over a period of time prior to fitting the device, then doing the same, over the same period of time, after fitting the device, but otherwise paying no attention to any of the possible variables - to hiring time at the local race track, and running the vehicle, with and without the device, at the conventional steady (ish) speed of 40 MPH - to taking the vehicle to a dynamometer station, both with and without, the device being fitted.

The last method is not only the best of the three, it is seen by many reasonably knowledgeable people as being the 'very' best possible way to run such a test. Indeed, if it were possible to arrange for ALL of the variables to be controlled or scientifically adjusted for, it would certainly be a good test. However, only under the most professional and dedicated testing regime (as one might find within an engine producer's or top of the range engine tuning facility) would this be entirely satisfactory. The simple fact is that the same vehicle, taken to two different stations, on the same day, is very likely to end up with two significantly different reports.

On the other hand, an engine, run on the same 'engine' dynamometer, under the same environment conditions, with the only difference being the activity or not of the device under test, is the only safe and truly reliable way.

ROAD TESTING PROBLEMS

Varying road surface conditions and temperature.

Wind speed and direction - with respect to the vehicle.

Road speed and its constant variations.

Air temperature and pressure.

Tyre condition, temperature and pressure.

Vibration - especially of the fuel supply vessel.

Fuel measurement.*

Driver constancy.

*For a person wishing to test their newly installed fuel saving device (I have interviewed several) the problems are pretty well insurmountable. For a start, most have to rely upon assessing their fuel level meter, no easy task in itself, and far from reliably accurate anyway. Or, they rely upon the 'nearly empty' warning light for their second reading, and the fuel station pump indicating 'tank full' for their first reading. This method, according to both the meter and the fuel pump producer's published specifications, can easily produce a plus or minus 2 gallon error, and even more if the weather conditions are variable! Even with a big fuel tank, a possible 4 gallon difference makes these tests a little less that scientifically acceptable. Even if they took their vehicle to a dynamometer station, rather than just driving on the road, the results would still be most unreliable.

When we say 'the same engine' (with and without the device), we need to be sure that the device itself does not render any other aspect(s) of the engine different. For example, does it have any meaningful effect upon the fuel flow rate or fuel temperature, and if so, does this need to be adjusted for, and can it reliably be adjusted for?

So, with respect to testing an IC engine's fuel conversion efficiency, what are these variables, and how critical are they:-

The fuel: The only safe way to measure the amount of fuel consumed, is to weigh the supply before and after the test. There is no 'fuel-flow' metering device (that I am aware of) that is immune to error, and even worse, most can be critically effected by vibration induced resonance. The fuel temperature will affect its bulk, successive tests carried out even under good 'laboratory' conditions, are very likely to affect the fuel reservoir temperature. Weighing is the only safe method. Although temperature does affect a fuel's bulk, it does not affect it's weight.

If the reservoir pressure is significant to the performance, a similar starting fuel weight is essential.

Engine speed. This is one of the really difficult variables. If the device increases or decreases the engine speed, this can affect road tests and vehicle dynamometer tests to a considerable extent - but re-tuning the engine to maintain the same engine speed must not be permitted, nor must using a different throttle position. As to 'adjusting' the results - I would very much like to see a truly rational description of how this should be done, particularly as it would have to be different for every single vehicle! - Not just every make / model - each and every vehicle.

Is there any point in further pursuing this subject?

It is generally the opinion of the scientific community that the best way to test anything that might be subjected to significant variable conditions, is to completely remove all such variables as cannot be positively and accurately adjusted for, and which are not critical to the function under test, and measure and declare all remaining possible variables.

Who could ague with that? However, even this regime is not without certain amount of risk - as has been discovered on numerous occasions.

If we examine the above testing philosophy, we find that, in our case, it is saying:

Run the tests in a constant environment.

Remove all non-essential components.

As directly as is possible, convert the engine's mechanical output into a reliably constant and accurately measurable medium - preferably one that would be acceptable to all scientific personnel, irrespective of their particular discipline.

Measure all possibly variable aspects of the tests, and include these measurements in any conclusion declarations.

If there are any variations that are known or expected to be encountered in the proposed 'real' operating conditions, these must be declared, and if possible, the worst case conditions applied to the testing procedures*.

I believe that, as far as is practical, our cracker test-rig meets these conditions.

John Allen

*The cracker, operating under very low fuel pressure conditions, is prone to produce tiny bubbles of gas. This does have a significant negative effect upon its performance. However, because there are many engines that operate under such conditions, we have elected to run the tests upon which we base our claims, in this way. High pressure, direct injection engines will produce further improved MPG. The bubbles are not easily seen by the naked eye - but by running the device under conditions that allow the bubbles to rise back out of the device, and so form larger ones. That these bubbles do form, and will reduce the cracker's effects, is very easy to demonstrate.