A BSFC (brake specific fuel consumption) map of a SI (spark ignited) ICE (internal combustion engine) describes the engine's fuel efficiency for every combination of engine torque (usually in Newton meters) and engine RPM. For every driving condition (engine horsepower requirement) there is an optimum combination of engine torque and RPM for best fuel economy. Usually the SI engine is most efficient at about 35% of redline RPM and 75% of maximum torque. For the 2.5L engine in my 2003 Subaru Forester with a 5-speed manual, the maximum efficiency is at about 120 Nm and 2200 RPM (27.6 kW or 37 HP). City driving usually requires less than about 10 HP (easily calculated for steady speeds). For power requirements below about 30 HP, the BSFC map shows that the optimum engine speed for best fuel economy is below about 2000 RPM. I usually get into 5th gear between 25 mph (1100 RPM) and 35 mph (1500 RPM), depending on driving conditions. I consistently get 38 to 39 mpg in the summer (34 to 35 mpg in winter) and drive over 500 miles between fuel stops. Do not be afraid of using torque up to 80% or 90% of WOT (wide open throttle). The engine computer will optimize the spark timing to prevent pinging and engine damage.

Thanks - I find this helpful! I drive "blind" in my '89 car that can't take a Scangauge. I do have a tach. With any luck I'll find the BSFC map for my engine.

Thanks for the low-down. Quick search of Hobbit's Site yields what I'm after for my Prius Hat's off to Hobbit for including it in his training slides. 11011011

Hi Guys: ___Love pretty BSFC maps but I never drive in an efficient area according to most I have seen. Even the Prius is supposed to allow garbage down in the 1.2- 1.4K range and I find just the opposite? Slower speeds of course but that is my own real world including pulses. The Accord allowed 54.5 into the drive tonight as. An Accord? Impossible! The BSFC maps do not allow anything near that level of FE under any regiment! The first and second laws of thermo would have to have been broken so it must be impossible! At least that is what some keep saying ___Good Luck ___Wayne

Yeah, but Wayne, while you may not be able to break the laws of physics either, you're the guy finding all the loopholes! . _H*

Hi Al: ___You are not exactly sitting on the fence allowing mother nature to control your Priusâ€™ destiny either ___Good Luck ___Wayne

Reply to Wayne- I think my engine's maximum thermal efficiency is at about 120 Nm and 2200 RPM, corresponding to about 37 HP. The thermal efficiency is about 31%, corresponding to 260 grams of gas per kilowatt-hour of brake (flywheel) power output (1 gram of gasoline contains about 44,000 joules of energy). However, if I could drive at a steady speed of 35 mph using only 8 HP (6000 watts) at 1500 RPM (5th gear), the thermal efficiency is probably closer to 24%, corresponding to 340 grams per kWh. This corresponds to 58 grams of gas per mile, or 48 miles per gallon (using 2800 grams per gallon). Retired physicist.

Ah -- see there is the rub. We do not drive in steady state -- and we choose not to because we can enhance fuel economy substantially by making this choice. Your calculations are correct but only cover steady state driving. They do not take into account the games we play with efficiency bands for accelerating and coasting for bringing the average consumption rate over a given distance down substantially from those steady state numbers. That understanding (gained from discourse with Wayne here) was the "AHA!" moment that defined my departure from somewhat enhanced mileage numbers to "What the... ???" numbers.

So it sounds like what you're saying is that it needs to be studied as a dynamic rather than steady state optimization problem. Dynamic optimization with discontinuities (because the engine uses no fuel when it's off) would make you do all sorts of fun math to do a good job modelling what Wayne does. The BSFC map data might provide you with useful data for implementing a numerical solution. Wayne's result that pulse and glide beats steady state is interesting and appears to have been informally reproduced by a number of people including me. Has anyone formally tested the proposition?

I did a quick-n-dirty test a couple months ago. A 5.4 mile loop, with 4 right turns and rolling 50-foot ups and downs. A - 45.6 mpg - cruise control 55mph B - 49.0 mpg - no CC, steady 55mph C - 63.5 mpg - P&G 60-45mph A - 48.4 mpg - CC again Hardly anything close to formal or scientific. Not enough to prove / disprove CC against foot steady-state. But I took the P&G number to heart, and now I P&G whenever I possibly can. My last tank was 56.3 mpg with that approach.

Interesting indeed. Look at this info: http://www.rangerpowersports.com/tech/?p=25 Following retiredphysicist's numbers, for TorqueNada 35% of max rpm = 1960 rpm, and 75% max torque rpm = 1800 rpm. Looks like my SWAG shift points of 1800-2000 rpm have some basis besides the seat of my pants. I have found definitely found that Pulse & NICE-ON coasting from 55-50 is much better for FE than steady state 53: I can get 40 P&N vs 32-34 steady.

Even with standard (non-hybrid) cars, P & G (pulse and glide) is a very viable option when driving at a steady speed below about 40 mph. In my Subaru, the maximum engine efficiency (roughly 31%) is at about 37 HP. However, driving at a steady speed of 35 mph requires only roughly 8 HP (1500 RPM in 5th gear), where the engine thermal efficiency is roughly 24%. In this case, accelerating (pulsing) to 40 mph using 37 HP, and gliding down to 30 mph, would use the engine at maximum efficiency. This is conceptually similar to hybrids, except that the excess pulsed energy is stored as kinetic energy in the moving vehicle, not as electric energy in a battery (and is more efficient). In this example, the pulsed stored kinetic energy for a 1500 Kg vehicle is about 104,900 joules, or 29 watt-hours. retired physicist

I get slightly higher numbers for steady state in my vehicle than Ken's truck, but significantly better numbers with P&G -- not surprising given the aerodynamics involved. I just finished a trip over the weekend covering 366mi with significant hills (most of them I could only do about 27mpg up with really careful throttle control) -- this is on I94 between Milwaukee and Minneapolis -- and I pulled 51.2mpg over that distance. For comparison, I last drove that route about 4.5 years ago and with flatter segments after this one to finish up the tank (adding an uptick) I ended up with 36mpg. That number included coasting in neutral down the inclines when I could but otherwise fairly typical highway driving. Ignoring the flatter segments which would have been about half the tank, (guessing here as this was way before the SG) the mileage was probably closer to the 33mpg EPA highway rating over the segment in question. So at an average of 65mph I managed somewhere in the neighborhood of 55% improvement? I can often do even better than that if the terrain is flatter. I have seen segments in the lower 60's with this car and I've actually parked the car after 20+miles on the highway with 57.5mpg showing on the SG. That's what, a 74% improvement? That one was at an average speed of 62mph. I'd say that P&G is viable at quite a bit higher speeds than just 40mph. It may not be quite as effective, but I think not using it at those speeds would be a mistake.

A problem with most published BSFC maps is they only describe performance at WIDE OPEN THROTTLE! What happens at part throttle is usually a total mystery but that's where we drive!

Where do those useless plots come from? As retiredphysicist said, proper normal BSFC maps show BSFC at every combination of speed and load from zero to max. Look for the speed-torque combination that offers minimum BSFC at every power level.

The real BSFC maps (representations of 3D data) do show part-throttle. It's the BSFC curve that is always done at full throttle. I don't remember why manufacturers and sometimes engine builders will generate the WOT BSFC curve, but they do. (I think I have heard one builder say that a relatively flat curve shows the tuning of the engine is good?) Unfortunately the maps are a lot harder to find than the WOT curves are. -soD