BSFC (Brake Specific Fuel Consumption)

EddyKilowatt

Veteran Member
Joined
Mar 1, 2006
Location
Carmel Valley CA
TDI
2003 Golf GL 5M
GoFaster said:
It's better to perhaps spend a wee bit more on fuel, but get acceptable drivability and vibration properties, and not smash up turbochargers, clutches, and engine mounts.
QFT!

I keep two maps in mind as I drive... one is the BSFC map posted above by TDIMeister... the other is this cool one that I orginally saw in a post by Drivebiwire, which gives a rough idea of Wear Rate vs Load % and RPM:



My understanding was this map is based on actually running an engine under the stated conditions and measuring wear products in the engine oil...

Eddy
 

MikeMars

Veteran Member
Joined
Mar 6, 2008
Location
UK
TDI
Vento 1.9 TDi (retired), A4 1.9 TDi (rear end collision), VW Passat 1.9 TDi (retired), Audi A2 1.4 TDi
Sorry for being dense, but is red bad and green good, or vice versa?
 

wjdell

Top Post Dawg
Joined
May 17, 2006
Location
Central Florida
TDI
06 Jetta TDI DSG PKG 1 17" VV Campy White/Beige
The legend shows blue as the worst wear - Not sure if I am reading it correct but I guess its showing load vs rpm and the wear at that range. Light load and low rpm is good and heavy load and low rpm is bad. 3D like that its a bit hard to understand. My load is always low and I run 1700 to 2500 rpm. Passengers and luggage is usually under 370lb, my UOA's reflect this.
 

MikeMars

Veteran Member
Joined
Mar 6, 2008
Location
UK
TDI
Vento 1.9 TDi (retired), A4 1.9 TDi (rear end collision), VW Passat 1.9 TDi (retired), Audi A2 1.4 TDi
The reason I was confused is that the graph shows heavy load + low RPM to be good (at least below 1,300 rpm) which doesn't make sense to me.
 

McBrew

Top Post Dawg
Joined
Oct 30, 2002
Location
Annapolis, MD
TDI
2003 Golf GLS TDI, 5 speed, Silver/Grey
That confirms what´s been discussed many times, these cars are most fuel efficient at 55mph.
You mean if speeds under 55 MPH didn't exist? You will get better FE at 54, even better at 53, better yet at 52, etc.

Sorry folks, there isn't a magic RPM/load where a higher speed (in a given gear) will yield better fuel economy. And by fuel econom, I mean miles per gallon. I don't mean BSFC or thermal efficiency.

You might produce more HP per gallon at a higher RPM/load, but you will be burning more fuel per mile traveled than you would at a lower speed.

Think about it this way: A tractor trailer towing a 40,000# trailer getting 5 MPG is a more efficient use of fuel than a Ford Explorer getting 12 MPG, but the Explorer is still getting better miles per gallon than the tractor trailer. Unless the point is to haul a 40,000# trailer, the "thermal efficiency" argument doesn't make the tractor trailer use less fuel to get from one place to another.
 

McBrew

Top Post Dawg
Joined
Oct 30, 2002
Location
Annapolis, MD
TDI
2003 Golf GLS TDI, 5 speed, Silver/Grey
tnp, what you want is a ScanGauge. This shows your real-time fuel efficiency, as well as other useful information.
 

TDIMeister

Phd of TDIClub Enthusiast, Moderator at Large
Joined
May 1, 1999
Location
Canada
TDI
TDI
EddyKilowatt said:
... the other is this cool one that I orginally saw in a post by Drivebiwire, which gives a rough idea of Wear Rate vs Load % and RPM:

I really wished Drivbiwire would include more information or attribution as to the source of this chart, or at least some context like a caption, as I try to do when I post charts. For all I know, this could have been created in MS Paint. Unlabeled axes (which one can deduce what they are from the magnitudes and the semi-ambiguous chart title) is not helpful. It would also seem odd to me that wear would be less at, for example, 100% load @ 4450 RPM than 83% load at the same RPM.
 

McBrew

Top Post Dawg
Joined
Oct 30, 2002
Location
Annapolis, MD
TDI
2003 Golf GLS TDI, 5 speed, Silver/Grey
Now, if they built a car that was electric, driven off of a battery that was charged by a TDI engine/generator combo, then it would seem to be most efficient to have the engine start up when the voltage dropped below a certain point, run at a high load at peak thermal efficiency until the battery was charged back up, then shut the engine down. This would yield the most energy for the fuel burned, which would (since the battery is storing that energy for later use) yield high miles per gallon as well.
 

TDIMeister

Phd of TDIClub Enthusiast, Moderator at Large
Joined
May 1, 1999
Location
Canada
TDI
TDI
McBrew, that's the operating principle of "range extender" hybrids. However, this is not a panacea. Even if the engine could be only operated almost always at- or near the best-point BSFC, there are losses as the work from the engine (assumed to be at an efficiency of ~43% for a TDI at the best point BSFC) goes through the generator (for simplicity's sake 90%), power electronics (95%), battery charging (80%), battery discharging (80%), power electronics (95%), electric motor (90%). The "round-trip" efficiency is around 47% based on the individual efficiencies I estimated, and this is in the the ballpark, which means that the 43% efficient engine is actually only about 20% efficient once the energy from the engine comes back to the drivetrain via the electrics.

It has been my experience doing hybrid drive cycle simulations in my day job, that the biggest gains from hybridization actually come from shutting the engine off during traffic light stops and charging the batteries only regeneratively when braking. You don't need a high-degree of hybridization to achieve this. BMW's Efficient Dynamics incorporates start-stop and a mild-form of regerative braking without complicated hybrid hardware, and it's extremely effective. Another large benefit of hybridization is to be able to run on electric-only mode for low-speed, low-load conditions, when the combustion engine is very inefficient. Using the engine to charge the battery should only be used very sparingly and in very specific operating conditions, where the total "roundtrip" efficiency with this hybrid "load shifting" is greater than or equal to that of the baseline non-load-shifted condition.
 

McBrew

Top Post Dawg
Joined
Oct 30, 2002
Location
Annapolis, MD
TDI
2003 Golf GLS TDI, 5 speed, Silver/Grey
Meister... good points about the losses. I was intrigued by a system I read about that used regenerative braking... but by using hydraulic pumps/motors at the wheels to compress gas (over oil) in high pressure cylinders while braking, and then use that system in reverse to recover the energy while accelerating. This would be, of course, in conjunction with an ICE or electric drive. The efficiency, according to the article, was much greater than electric regenerative braking. I'm sure you've heard of this, but I'm mentioning for the benefit of others reading this thread.
 

bokeh

Active member
Joined
Mar 22, 2008
Location
Alicante (Spain)
TDI
Bluemotion
TDIMeister said:


Each curve represents a constant horsepower developed by the engine.
Why are your blue lines curves? If power is inversely related to RPMs why aren't they straight lines?
 

bokeh

Active member
Joined
Mar 22, 2008
Location
Alicante (Spain)
TDI
Bluemotion
I see it more like this:



Maybe I'm overlooking something but I've cross checked this against a torque and power curve map and it appears to be correct (for a 105PS TDI PD).

Was the BSFC map for a 105PS TDI PD (I don't know what motor is in that Beetle).

Moving on from this does anyone have any idea how much horsepower it takes to keep the car rolling at various speeds?
 
Last edited:

TDIMeister

Phd of TDIClub Enthusiast, Moderator at Large
Joined
May 1, 1999
Location
Canada
TDI
TDI
You have the red curve of best efficiency correct. However, plot horsepower as a function of BMEP and RPM according to the following equation (you can do this in Excel by adapting the equation below):

POWER = RPM * BMEP {bar} * DISPLACEMENT {cc} * 8.333333e-7

This is give you power in kW. To convert to horsepower, multiply by 1.341022. You will certainly get curves shaped like hyperbolas, not straight lines.
 

bokeh

Active member
Joined
Mar 22, 2008
Location
Alicante (Spain)
TDI
Bluemotion
TDIMeister said:
However, plot horsepower as a function of BMEP and RPM according to the following equation

POWER = RPM * BMEP {bar} * DISPLACEMENT {cc} * 8.333333e-7
Ok. We can forget the stuff in red because that is just constant. Now we are just left with:

POWER = RPM * BMEP * CONSTANT
POWER = RPM * TORQUE * CONSTANT
BMEP = TORQUE * CONSTANT

The scale down the left is BMEP which according to the above is synomous with torque. Seeing as the
x-axis, RPM and y-axis, torque are both linear scales the only possible way to plot from point "A" to point "B" while maintaining a constant power level is a completely straight line (according to the above formulas). How could it be any different?

Also:

BHP = (TORQUE ft/lb * RPM) / 5252

If you check any point on my straight lines using that formular you will find they are correct.
 
Last edited:

GoFaster

Moderator at Large
Joined
Jun 16, 1999
Location
Brampton, Ontario, Canada
TDI
2006 Jetta TDI
You're missing the point.

Pick a given horsepower number at a given middle-of-the-range torque and speed. To keep that same power level with (say) double the torque, the speed is half. To keep that same power level with (say) half the torque, the speed has to double.

That is not a straight line, it is a hyperbola.

A straight line graph, if extrapolated, will cross the zero axis at some point. This implies, if your presumption were true, that it would be possible to generate that power level with zero torque by using some high RPM, or with zero RPM by using some high torque. Obviously this cannot be true.

A constant-horsepower line with axes of torque in one direction and rpm in the other direction must be a hyperbola that can never cross the zero axes no matter how arbitrarily high the other variable is chosen.
 

TDIMeister

Phd of TDIClub Enthusiast, Moderator at Large
Joined
May 1, 1999
Location
Canada
TDI
TDI
Sigh. It would be a straight line if you were drawing a linear function of ONE independent variable (e.g. RPM). But you're drawing a function of TWO independent variables from the same graph (RPM and BMEP).
 

GoFaster

Moderator at Large
Joined
Jun 16, 1999
Location
Brampton, Ontario, Canada
TDI
2006 Jetta TDI
For example, note the point (on the graph which shows the correct hyperbola constant-horsepower lines) where the 50 horsepower curve crosses 2000 rpm. It crosses it at 12 bar. Now note where it crosses 4000 rpm. It does so at 6 bar, which is correct ... double the speed at half the torque to maintain constant horsepower.

On your graph it so happens that these two particular points maintain that relationship (although it's scaled wrongly), but what happens if we pick a point halfway between?

The correct result is that the BMEP should be 8 bar at 3000 rpm (1.5 x higher speed than starting point but 2/3 the pressure, 1.5 x 2/3 = 1.000000....) and this is correct on the graph with hyperbolic curves, but incorrect on the graph with straight lines.
 

shizzler

Veteran Member
Joined
Jun 25, 2005
Location
Ann Arbor MI
TDI
05 BEW Wagon
This thread hasnt moved in a while, but I just found through a link from ecomodder.com. The post I just put up over there is particularly handy for this forum too, since well, its our dear old ALH being discussed.

I really appreciate finding the BSFC map for my engine! As soon as I saw it I knew I had to start experimenting in excel. Unfortunately, not having the data meant re-creating it. Didn't take long though: Pretty close, I'd say.

from:


to:


For some background: BMEP (brake mean effective pressure) is directly comparable to torque, regardless of engine speed. I wont try to type formulas here, but you basically just relate torque to the mean pressure acting downwards on the piston times displacement and divided by 4*pi.
With units of bar (= 100 kPa), the result is torque in N-m (newton-meters).
BMEP is most often simply calculated from the brake torque as measured on an engine dynamometer. It becomes very useful when you measure the actual cylinder pressure with a transducer, calculate the indicated mean effective pressure (IMEP), and subtract the BMEP to obtain the FMEP, or friction losses of the engine due to rubbing, rotating, accessory drives, and pumping losses.

Now that we have a decently accurate BSFC map set of data, we can overlay other data for comparison. For example, the lines of constant power in blue on that version of this plot. While those are neat for reference, they aren't especially practical, since you cant just adjust your rpm while driving down the road to maximize efficiency. (of course gearing changes do this though - more on that later).

I decided to look at the power requirements due to real road loads when traveling in top gear (where I do most of my driving). I was going to calculate a bunch of stuff when I found the aerodynamic and rolling resistance HP loss calculator on ecomodder, very handy!
http://ecomodder.com/forum/tool-aero-rolling-resistance.php

My input data:
VW Jetta Mk4
weight = 3300 lbs
Cd = 0.30
Crr = 0.010
Fuel = B20 biodiesel
Air Density = 1.293 kg/m^3 (0 C)

very cool. Copied the data out and pasted into excel. Now we have the aerodynamic and rolling losses calculated from 5 to 120 mph.
Next, I calculated the engine rpm in top gear when traveling through that same speed range with my tire size and gear ratio #s. If we assume we're stuck in 5th gear, we can overlay a plot of the aero and rolling load on the vehicle vs engine rpm. The calculated HP loss is converted to torque through rpm. Vehicle speed is on the top axis for reference.



But what if we can lower our aerodynamic drag? My goal for my jetta is to get down to Cd = 0.27. Probably optimistic given how good it is to start with. We'll see (mirror delete and other tricks on the way this spring).

Re-calculating, we can compare.



So you can see that reducing our road load with the same vehicle gearing, we are actually moving the engine operation into a less efficient area. Bad thing? No way! because even though the efficiency has declined for the same vehicle speed, we are requesting less load and our overall fuel consumption has still declined. I ran a comparison calculation:

Traveling at 2500rpm = 72.5 mph
I fit 2nd order polynomials to the aero and rolling calculation lines (R^2 = 1). Back calculating real road load at this speed, converting to power @ 2500rpm, and interpolating between BSFC lines to estimate the specific fuel consumption, we generate the following data:

2500 rpm = 72.5 mph
Cd Nm kW BSFC est g/hr L/h gal/h mpg
0.30 69.21 18.11 263 4762.91 5.60 1.48 48.98
0.27 64.48 16.87 269 4538.95 5.34 1.41 51.39

Pretty darn cool, especially since the calculated fuel consumption matches what I record in driving nearly perfectly (the 49 mpg, that is). So we can now calculate directly the effect that a given aero improvement will have upon mileage! nice. Interpolating between the BSFC lines is a little bit of a guesstimation, but its the best we can do.

Next up: Gearing reductions through a swapped out 5th gear or tire size changes.
 

shizzler

Veteran Member
Joined
Jun 25, 2005
Location
Ann Arbor MI
TDI
05 BEW Wagon
Thought more folks might chime in... guess no one cares, haha. Either that or I just nerded out hardcore. Well I said I would post up my gear swap calculations, so here they are.

Available options for my TDI are 0.681 and 0.658 5th gear ratio sets. (Stock is 0.755), but you all know that. I have also been considering a set of 205/60/16 tires (3% more circumference vs the stock size of 205/55/16).
The combination of the .681 gear and the 205/60/16 tires is almost exactly equal to the effective overall gear ratio as the .658 gear set, btw.

So whereas we could place a mph scale atop the former graph since our gear ratio was unchanging, we can no longer do this. Instead, trusting that our aerodynamic and rolling resistance is constant at a given speed, despite the vehicle gearing situation, we can interpret vehicle speed as a constant power requirement. (e.g. the blue lines from before). So below I plotted 72.5 mph = 24.3 HP, and for a Cd of 0.27, 72.5 mph = 22.2 HP.



The green dot is our starting point. Changing the gear ratio with a .681 ratio 5th, and the 205/60/16 tires, we move from 2496 rpm at 72.5 mph in stock form down to 2189 rpm, along the 24.3 HP line. Finally I again looked at a coefficient of drag reduction down to 0.27, sticking with the new overall gear ratio. Estimating the BSFC for each point, these are my results:

mph rpm load(kW) BSFC est gal/h mpg
stock gear & tire 72.5 2496 18.11 263.0 1.480 48.98
0.681 gear 72.5 2251 18.11 248.5 1.399 51.84
0.658 gear 72.5 2175 18.11 243.0 1.368 53.01
205/60/16 tire only 72.5 2427 18.11 260.0 1.463 49.54
.681 and 205/60/16 72.5 2189 18.11 244.0 1.373 52.79
Cd=0.27 .681, 205/60 72.5 2189 16.54 248.0 1.275 56.87


So, I'm sold. Probably won't ever recoup the entire cost of the gear swap . But the reduction in engine noise, wear, and fuel consumption are all worth money to me.
 

SoCalC

Veteran Member
Joined
Oct 24, 2004
Nice. Thanks for nerding out hardcore. In particular Michigan's gonna need as much of that as it can get in the years to come along with the rest of the United States.

May throw that gear swap in the mix myself. Ever test for fuel economy change with the muffler job? I wouln't think it worth much for speed limit and slower stuff.
 

KILL CARB and the US EPA

Well-known member
Joined
Dec 3, 2007
Location
da Bronx
TDI
2011 Golf 2.0 tdi CR 6 spd
real diesels come equipped with indicator valves so one can measure imep and keep your plant balanced as parts wear etc. These rotary pumps are about as effective at metering as my spaghetti collander. 1900rpm seems to be this engines peak bmep hence 55 mph=good mileage, soon as i hit lotto i'm going for taller tires and taller final drive i want 75 mph@1900
 

EddyKilowatt

Veteran Member
Joined
Mar 1, 2006
Location
Carmel Valley CA
TDI
2003 Golf GL 5M
I have eyeballed the .681 swap onto that same BSFC graph a couple of times, with similar results (not to four signifcant digits though!)... thanks for posting the rigorous numbers.

Makes me wonder why better results are not experienced by folks who actually do the swap... I think I've heard of 1-2 mpg, but not 3, even among people who drive full tanks on the highway.

Eddy
 

GoFaster

Moderator at Large
Joined
Jun 16, 1999
Location
Brampton, Ontario, Canada
TDI
2006 Jetta TDI
If ALL the driving that you do is constant highway speed, then you should get the results predicted by the graph.

The problem is that most real people have a bunch of city driving mixed in (not using 5th gear = no effect) and acceleration (ditto). The other problem is that a really tall 5th requires using 4th to a higher road speed to avoid lugging, and during whatever driving you do in that speed range, you end up *worse* off.

I figure the real world benefit is about half of what theory predicts, on the overall.

should note that I drank some of the hypermiler's kool-aid a few months ago and started using 5th down to lower road speed, and ended up vibrating the bolt that retains the output-shaft gear loose inside the tranny. Fixed that, but no more 5th below 80 km/h, at all!!
 

shizzler

Veteran Member
Joined
Jun 25, 2005
Location
Ann Arbor MI
TDI
05 BEW Wagon
No problem guys. I got all nerd-excited when I realized i could calculate real fuel economy estimates based off the BSFC chart.

Just ran the numbers for the top gear USA rabbit - ALH swap car too. With a few aero improvements they should be hitting their 70mpg no problem. Today's post shows they are only up to 52, they need to learn how to drive though.

I think gofaster nailed it for the real world explanation of the gear swap results. For me, 24 out of 25 miles of my 1-way commute are highway. MI highway at that, where the slow lane goes about 75 mph. I ordered my 0.658 gears today!

Serious bummer on the low rpm 5th gear useage though, since thats kind of the exact reason for the swap. What rpm is that (80km/hr) for you gofaster?
We'll see how mine goes....
 

rockyrunner99

Veteran Member
Joined
Aug 20, 2009
Location
Grand Rapids, Mi
TDI
99.5 Golf
nokivasara said:
That confirms what´s been discussed many times, these cars are most fuel efficient at 55mph.

Edit: The BSFC is always at peak torque, isn´t it?
With stock gearing!

I agree, from my experience, BSFC is always lowest at peak torque.

Anyone else notice the BSFC curve looks like a function of the torque curve, like half the inverse of something. If some were to try to move the efficiency point to higher rpms how would that be done? If you could change absolutely anything in the engine, assume we are designing it from scratch. What could we change to put the highest efficiency at say 2500-2700rpms? I am just very curiouse about engine design, and I know there are some very smart people here.
 
Top