Analyzing Gen1 Insight Acceleration to 30 MPH

[3-Wheeler]

There was a post here at EM, where a member was asking for help regarding lugging the engine, and resultant fuel economy.

I responded to the thread, but my response got me thinking about a better way to analyze my reported gas mileage findings.

The analysis at this point is quite limited, and probably will not get much better, as part of the gas measurements and resultant calculations, are based on several factors that are hard to measure without dedicated equipment.

The first graphic below shows a page from a spreadsheet that discusses my field measurements and close approximations to distance traveled during those measurements. The distances are estimates because the car dash display only starts to indicate travel every 0.1 mile, and several of my acceleration tests occurred over a shorter interval than the display can reveal. Therefore there is *uncertainty* in my measurements, and these will be discussed below.

Three acceleration mileage tests were performed:

A) Accelerate from 15 to 30 mph in second gear only, TPS of 70.

B) Accelerate from 15 in second to 18, shift to third and accelerate to 26, shift to fourth and accelerate to 30 at TPS of 70.

C) Accelerate from 15 in second, to 18, shift to third and accelerate to 26, shift to fourth and accelerate to 30 at TPS of 30.

I will discuss some of the important assumptions:

Item 6: These distances are my best estimate on distance to 30 mph. The three graphics show that being off slightly shifts the best efficiency from Option A to B or C.

Item 7: Calculated total gas usage in CC's based on estimated distance from Item 6.

Item 14: Is the total distance needed to travel to get the equivalent gas consumption of 100 mpg, after initial acceleration.



If the real moving distance to 30 mph is only 0.4 mile then accelerating to 30 mph with Option A is better.



On the other hand, if the real moving distance to 30 mph is actually 0.6 miles, then Option B or C is better.



My personal favorite is Option B, because the throttle plate open for low pumping losses, and the RPM's low for low piston friction. But is this really the best of the three options?

You can clearly see that determining the best way to accelerate a Honda Insight from 15 to 30 mph takes more than just seat of the pants gas mileage measurement. Dedicated equipment would go a long way to determining true gas consumption and distances traveled.

P.S.

After documenting my findings in this thread, it *finally* occurred to me that accelerating to a higher speed will accomplish two things:

a) cause more gas consumption over the total acceleration, making it easier to get a solid measurement.

b) because of the higher speed, get a more accurate distance measurement.

Jim.

[jamesqf]

Just a thought: most laptops these days have acceleration sensors, which can be accessed by various APIs and are sensitive enough to measure/time vehicle acceleration quite accurately.

[SentraSE-R]

It looks to me like you're extrapolating fuel consumption and distance from average mpg for guessed time at different % engine load, a very crude guesstimation of actual consumption, time, and distance.

The Scangauge II with performance upgrade allows us to log actual fuel consumption and distance every 0.4 seconds. No guesswork, no estimates. All you have to do is maintain the % engine load from speed A to speed B. See this thread.

[3-Wheeler]

Originally Posted by SentraSE-R

It looks to me like you're extrapolating fuel consumption and distance from average mpg for guessed time at different % engine load, a very crude guesstimation of actual consumption, time, and distance.

The Scangauge II with performance upgrade allows us to log actual fuel consumption and distance every 0.4 seconds. No guesswork, no estimates. All you have to do is maintain the % engine load from speed A to speed B. See this thread.

I actually used the car FCD by resetting the gauge just before the test, then toggling the gauge right after the test.

This updates immediately compared to every 0.4 seconds from the ScanGauge. I think my ScanGuage is even slower than this, as the temperature readings when climbing hills take much longer. However that could be due to the sensor itself.

You don't specify how to get the data off of the ScanGauge after each test?

Can you elaborate?

Thanks, Jim.

[SentraSE-R]

I manually transferred the SGII data to a notebook. I'm not seeing where your precision comes from. AFAIK, you are guessing distances, and stacking multiple gear shifts.

When the SGII logs its 0.4 sec. data, it logs distance down to 16' intervals,, and it provided 44 data points for my 83% load pulse.

[3-Wheeler]

Originally Posted by SentraSE-R

I manually transferred the SGII data to a notebook. .....

I understand from your earlier post that you used a laptop, but HOW did you perform this feat? Does the Scan Gauge have either a USB or serial port on the back that I don't know about?

My Scan Gauge is about five years old.

P.S. I downloaded the Version 6 Manual for SG, and see there are many options for saving this data. I see two OBDII ports (page 5) but nothing else for data transfer???

Thanks, Jim.

[SentraSE-R]

Manual transcription. My first job was copying well depth records from one sheet of paper to another. I figured if I could do that for a year, I could do anything for 30 years, and I did. I scrolled through the SGII's performance logging data (you need to upgrade the firmware to v. 4.06 or better to get the performance logging ability) and wrote the numbers onto a paper notepad.

[mikeyjd]

I love these kinds of threads. I'm very interested in the discussion you 2 are having. Darrel your thread with evidence suggesting improved efficiency with heavier (up to 9x%) engine loads has been a major factor in my getting my best tanks over my last 3 fills, all other things being equal.