Throttle-Stop Test, a granular look

[aerohead]

Originally Posted by AeroMcAeroFace

So "mechanical efficiency of the driveline components are a function of transmitted power" a linear one? It can certainly be approximated to linear over the 5% speed increase, so it can be incorporated into "rolling resistance"

How much does BSFC actually vary by? It is a tiny amount that is negligible over the small speed increase, and so can be ignored altogether.

1) Under the circumstances of constant-throttle, and a variable load, according to Sovran, mechanical efficiency would be non-linear. SAE may have papers dedicated to this specific topic.
2) Personally, I'm reluctant to combine variables, choosing the 'long-hand' approach, just for book-keeping purposes. There may be a particular vehicle road test which incompasses all data, from which one could run the numbers each way in order to detect whether or not something 'funny' was introduced into the quanta.
3) At Bonneville, I experienced over a 10% change in speed, over redline rpm. Technically, engine efficiency would be falling. Without a dyno, it's problematic assigning numerical values of a certain precision. We would have no 'a priori' knowledge of a velocity increase with a given modification.
4) The BSFC question is the most important of all, and again, it's an unknown quantity. An engine map from the manufacturer would be of some help. Perhaps.
5) We know for a fact that, any conclusions about a drag reduction from any particular modification could be off by 40% without gear-matching. That's the sticking point for the whole exercise. Any presumptions about quanta cannot have any degree of confidence above 60%.
-------------------------------------------------------------------------------------
Any thought exercise I come up with that might help requires a test track, away from public traffic.

[AeroMcAeroFace]

Originally Posted by aerohead

1) Under the circumstances of constant-throttle, and a variable load, according to Sovran, mechanical efficiency would be non-linear. SAE may have papers dedicated to this specific topic.
2) Personally, I'm reluctant to combine variables, choosing the 'long-hand' approach, just for book-keeping purposes. There may be a particular vehicle road test which incompasses all data, from which one could run the numbers each way in order to detect whether or not something 'funny' was introduced into the quanta.
3) At Bonneville, I experienced over a 10% change in speed, over redline rpm. Technically, engine efficiency would be falling. Without a dyno, it's problematic assigning numerical values of a certain precision. We would have no 'a priori' knowledge of a velocity increase with a given modification.
4) The BSFC question is the most important of all, and again, it's an unknown quantity. An engine map from the manufacturer would be of some help. Perhaps.
5) We know for a fact that, any conclusions about a drag reduction from any particular modification could be off by 40% without gear-matching. That's the sticking point for the whole exercise. Any presumptions about quanta cannot have any degree of confidence above 60%.
-------------------------------------------------------------------------------------
Any thought exercise I come up with that might help requires a test track, away from public traffic.

Are you saying that a 3% increase in speed will be 40% down to gearing? That is ridiculous.

Are you saying that a 3% increase in speed is down to BSFC suddenly jumping? Again that is ridiculous.

Gearing may be significant if the drag reduction is large, but it is typically not and so can be ignored.

[AeroMcAeroFace]

Originally Posted by aerohead

--------------------------------------------------------------------------------------
3) Typically, in mechanical engineering, it's common to assume a constant BSFC, with respect to predictions surrounding road load hypotheticals. As a mechanical engineer you would know that.
--------------------------------------------------------------------------------------

Glad we can agree on that now

[aerohead]

Obviously, your post-doctoral education and 40-years of research in a world-class laboratory have found complete fault with the SAE.
I apologize for wasting your precious time.
I'm sure your students and faculty members will be make the proper correction, by dismantling the stacks from your technical library, and banning any communication with the global automotive community.
Please accept my heartfelt sorrow for inflicting such a intellectual insult to your magnificent command of the subject.
It is only in this moment, that I can properly discern the infinitude of your scientific mastery.

[Cd]

Will there be an effect on the test if I hesitate to fully press the throttle by a second or so after seeing my start point ? ( Starting the test at a landmark, such as a tree by the side of the road. )

[AeroMcAeroFace]

It depends where you take the measurements, if you are using time taken to cover the distance then maybe, but I think the way Julian recommended to do it was to keep your foot in the same place on the throttle and wait for the speed to settle to a stable value and then take the measurements using an accurate speedometer/GPS/Mobile phone app.

If you are using the time taken to cover a distance, you need to be at the top speed before you get to the tree so get your foot on the throttle well before you start the measurements. Hope that helps

[aerohead]

Originally Posted by AeroMcAeroFace

Glad we can agree on that now

1) The issue is that we have an 'atypical' situation.
2) any assumption regarding a constant BSFC would necessarily be predicated upon the ECU sensing 'driver intent', which is impossible if the signal from the TPS has been disabled.
3) on the merits of 2) I'm not in agreement.
But let's all get our heads together and explore all options. The beauty of Julian's technique is it's simplicity and avoidance of the necessity of a 'large' sample, in order to arrive at useful quanta.
' Watch out where the huskies go, and don't you eat that yellow snow.' Zappa

[aerohead]

I agree that, you want to already be at 'equilibrium' for the configuration you're testing for when you begin measurement.

[Piotrsko]

Why would the acceleration / deceleration data not be smooth-able or be able to be integrated into the overall bucket? @ the rocket site we had the same issue with thermocouples not measuring instantaneous, but the whiz kid engineers from Stanford used the data anyways since they found it was actually linear.