Interesting how the consos on 5000km vary sawtooth: surely the impact of the outside temperature and the winter months ...
It's not just that. As I indicated the load of the car was changed during the months (from 1 to 3 people) but it is certain that the seasonal variations have played but more because of the state of the roads and the caps than on the outside temperature. That's just my opinion.
It is also interesting to see that the average consumption tends to decrease and not increase with the km: better fuel?
I do not think since we were driving with supermarket fuel.
Less friction?
Possible.
This is usually typical of diesel vehicles (unlike gasoline vehicles ...).
However if we look at the curve between 15 000 and 50 000 terminals we can see that the conso remains the same. By cons, later, there will be more than 2 passengers in the car and it will therefore lighten 70 kg. I think that can play on this decline in consumption.
Woodcutter,
I have not followed the post since the beginning and therefore not taken the original XL file, but cons I know at least one thing: 19 liters / 100 (if I understand correctly) in Saxomazoutte, it is TOTALLY IMPOSSIBLE so nothing that should make you think about how to organize the data ...
Absolutely! It is for this reason that calculating its consumption on full 1 (more or less well done because between the triggering of the gun and the actually full filling of the car there is a pack of liters) is completely and archi false (read the explanations I gave a few weeks ago).
The consumption of a car must be measured on a sliding average and not on 1 full of 500 kms (as it is unfortunately often the case). For example, with the case of the consumption of the AX, it is seen that consumption measurements on 5000 kilometers (7 or 8 full) give good results and no doubt make it easier to interpret improvements (Pantone, aero, weight...). We must not confine ourselves to raw data (consumption on full 1 which was only a full complement for example) but on a global consumption. For that everyone can take the raw data of this table to make its own experiments (as Christophe for example with its black curve)
EDIT: Well, apparently these are erratic data, but it would be better to compute only on full numbers.
In my opinion no. It is necessary to calculate on a sum of full after that which will completely erase the small differences due to an incomplete full example.
Elephant,
Well no, let's see: the driver gets older and the number of speed cameras and speed increases!
At the time (beginning 90) we did not have these concerns. The driver did not have a light foot ...