BSFC (brake specific fuel consumption) explained

[jeff88]

Hypothetical situation time...

Lets assume somebody had an alternator disconnect that automatically connected and disconnected based on various parameters. Let's use the AutoSpeed BSFC graph as a reference.

So if somebody had 60% engine load and 2,000 RPM (green island) and they "turned on" the alternator which increased the load say to 90% load (but same RPM), then the BSFC would be in the red island.

Would the car get better gas mileage with the alternator on rather than off, since the BSFC is in a better range?

[Focus-Ak]

Originally Posted by jeff88

Hypothetical situation time... Lets assume somebody had an alternator disconnect that automatically connected and disconnected based on various parameters. Let's use the AutoSpeed BSFC graph as a reference.

Jeff, that's brilliant.

Also, it occurs to me that although a BSFC chart is at the foundation of the indicator lights you conceptualized a few messages ago, what we really need is an indicator for ASFC -- Acceleration Specific Fuel Consumption (new term?). The nut of the driving problem for us is how to get from one speed (zero or whatever you are moving at) to a higher speed with the least investment of consumed fuel. A logic board with inputs from an accelerometer and the OBD-II's gal/hour computation could feed an LCD screen could give the result of dividing those two instantaneous quantities. During acceleration we then use that instead of the %-Load or vacuum readout on an UltraGauge or ScanGauge. (At constant speed or coasting it doesn't matter, just during acceleration.) That would give a direct reading of what is really of interest to us.

[Focus-Ak]

I just started a new topic, "On Coasting (Two interesting papers)," and those papers have some really interesting stuff on the integration of certain hypermiling techniques with BSFC factors. See here: http://ecomodder.com/forum/showthrea...tml#post463536

[some_other_dave]

Originally Posted by jeff88

Would the car get better gas mileage with the alternator on rather than off, since the BSFC is in a better range?

No. You're making more power with the alternator on, so the total consumption goes up. The consumption per horsepower generated goes down, but the overall consumption goes up. (There may be exceptions, but I'm thinking they are very rare.)

-soD

[Focus-Ak]

Originally Posted by some_other_dave

No. You're making more power with the alternator on, so the total consumption goes up. ... -soD

Dave, what you say is true. However, what Jeff proposes is that if the alternator has to be run some of the time but not all of the time, what it is the most advantageous circumstance for having the alternator on? One answer to that is at times when using the alternator can contribute to raising the engine's efficiency to a higher level, i.e. using the BSFC plot to its best advantage. Doing so has a double benefit -- not only does it make the electrical generation more fuel efficient, it also makes the driving power delivered by the engine more fuel efficient. So, yes, turning the alternator on for that instant does raise the fuel consumption at that moment, but it is the best time to perform this necessary function. It is not an additional fuel requirement in the context of the length of a trip; it is a minimization of that necessary fuel requirement.

[some_other_dave]

Ah, that is a different question than the one I thought I was answering.

What you describe is more in the way of pulse-and-glide for the alternator. But that can easily be done by simply pulse-and-engine-off-coast. That way you not only avoid having to power the alternator, but the whole engine as well!

If you're keeping the engine on, then yes--turning the alt on while pulsing is the best. Same for A/C in the hotter months.

-soD

[teoman]

What would be needed to obtain this bsfc chart from an existing car assuming you had all of the car's canbus data?

[redpoint5]

Originally Posted by teoman

What would be needed to obtain this bsfc chart from an existing car assuming you had all of the car's canbus data?

You'd need all the data contained within those efficiency maps. Torque, RPM, and fuel consumption (HP can be derived from RPM and torque, or torque can be derived from RPM and HP).

What you would be missing is a dyno; a way to measure horsepower at a given RPM.

I'm not sure how precisely fuel consumption can be measured from the canbus either. If I understand correctly, fuel consumption is calculated based on fuel pressure, injector flow rate, and "open" time. If that's the case, you're assumptions about flow rate for a given open time would have to be fairly accurate.

Even with a dyno, you'd need to run many scenerios, such as low torque and high RPM, high torque and high RPM, and everything inbetween. And then the same for low RPM...

There has to be a way to mathematically estimate the map based on a few measurements.

[teoman]

The torque app uses the accelerometer (if i am not mistaken) to calculate the acceleration of the car and provided the weight of the vehicle that information is used to calculate the power output of the engine.

I do not need something that is accurate, but I do need something that is very precise. Some value that I can minimize or maximize.

[teoman]

What if i build a multi dimensional map with the data. Log injection ms, rpm, speed (acceleration derived from that), boost pressure, afr?

Then for 2 points in the map if the injection ms is lower then I have lower fuel consumption for that point.

Anyone see problems? It would not take the hills in to consideration so maybe an accelerometer can take care of that. Am I missing something?