Hello,
Question metrology (and methodology), I do not think I have any lesson to receive, when a gain is validated on a power bench, with a measurement conso established by weighing timed, it is no longer in the approximate, but in the simple systematic error of the measuring device (2,5% for the scale, and 1 / 10 s for my thumb which stops the stopwatch).
Ben precisely Camel, the shoemaker is always the worst shod. So if you have a balance is a laboratory gas balance (I doubt it ...) you have actually 2.5% error but otherwise you are beyond 10%.
From the moment when one highlights differences greater than 15%, one is necessarily outside the margins of noise.
Theorem that remains to show ... But do not spend too much time because your statement is false metrologically.
What is right, however, is that this measure is framed by the margins of error of different measuring devices.
But that is not enough to compare TWO measures between them in a scientific way unfortunately.
This is undoubtedly the whole difference in method, precisely, between measuring on a bench, with precisely defined conditions, which make it possible to make comparisons "all other things being equal", and the "full to full" method that you use , Laurent Delaon, who is subject to variable external conditions and your way of driving ...
Ben it is more accurate repeating with imprecise material that measuring once with a material say a little more precise.
Moreover in my case I evaluated my conso without system (imprecise repeated measures) BEFORE intaller my system. So the driving factor is integrated before AND after. I can compare my results. Remind me how did you evaluate your drinks before and after? You'll do it soon it's true but you did not do it. Also before announcing that you have results presented as acquired you should have done that. Or relativize the thing.
All those who work seriously on the pantone have learned to be wary of their own way of driving, because we know that following the nervousness of driving, the conso can change considerably, therefore, even by doing, as you recommend, 5 full consecutive you will ultimately have a weighted result of a supposed phenomenon, combined with a conduct that remains a question mark, postponing this question mark on your overall result.
Not precisely not in my case (doubly because I do not have a pantone) you see. Reread explanation above.
You on the other hand you announce results as acquired and after you check them and you speak of being serious it is right?
While on a bench, a given braking torque with a given engine speed can only lead to a measurement outside the operator's field of influence (objectivity of the experiment and the measurements - subject already discussed elsewhere). .)
It is true we are out of the influence of the operator but is it really necessary? are not all cars driven by a driver? what is its influence?
Also have little freedom (see above) of this phenomenon because not everyone has a bench even if it is true it is more convenient for comparison (subject to knowing how to use and operate properly).
But the bench contributes nothing to the development of endurance and reliability of the system in real conditions. You'll see when the montage of the Mercedes Didier will leak (leaked into the pot) after 5000km.
As you say, there are always uncertainties, but there are methods that can reduce them to a quantified (and acceptable) range.
Forks are not quantified, depending on the repetitions and the variability of the measurements.
Have undergone more than have the reduies.
And 15% more, unless you take those who lead the tests for perfect morons, is not a result that can be dismissed by a hand, as you seem to want to do. under the pretext that the manipulation was not repeated X times ...
I do not rule out a single result I'm afraid it's different.
Knowing how to play with uncertainties is the basis in mechanical engineering (to mention only this subject).
He learns the mechanics so ... (or revises ...)
Mecanique stay modest huh? welds and pipes: we are more in plumbing boilermaking than the general mechanics ... (if I looked at your photos ...)
Making a prototype is a part of measurement, part of drawing, and part of realization. If you do not master a minimum of the measure (and the errors which are attached to it) you will make a buggy realization, the punishment in the matter being the observation "It does not fit!"
Note the caveman melted a bronze ax, did he draw a picture before? did he take measurements? was the ax buggy? perhaps....
It seems to me that you exaggerate a little. It's a trend at home ...
Obviously, you stopped at the doors of the pantone, probably repelled by the immensity of the task to carry out its development, and telling you that basically, if a simple injection of steam is enough, why get bored ... perfectly respectable attitude.
From there to put you above others, to rely on your results as a reference, and put in a simplistic way on the account of the error of measurement all the other results is an attitude which can not claim to be scientific.
I made a complete study of a pantone reactor and then thinking about it I opted for the current system. So I did not fall into the panel of the Pantone from my point of view, and I allow myself to point out that There is not really any focus in pushing a steel rod into a tube (well, it's really minimal).
When to my results being I think the only one to have realized them and to have proceeded in this way I am above anyone because no one has done it by myself.
The polemics.
I do not know your scientific and technical course, but I promise you for next week an avalanche of results to analyze ... we will resume this discussion after this phase ...
There is no need to have done studies to have common sense and imagination and know how to tinker.
It is a habit apparently of you to generalize quickly and to draw hasty and erroneous conlusions.
Do you think that I am incapable of understanding the results of your measurements under the pretext that you would have made studies that would allow you to do so without doubt?
I too hallucinate ...