Jeep 4.0 Volumetric Efficiency

[czb83]

I'm using a narrowband sensor right now. Once I get it tuned nice and working without flaws I plan on adding a turbo, at which point I will add a wideband sensor. Probably an LC-1, since they can be had for like $200 new. btw, its a stock 4.0 right now.

 

[5-90]

I'm using a narrowband sensor right now. Once I get it tuned nice and working without flaws I plan on adding a turbo, at which point I will add a wideband sensor. Probably an LC-1, since they can be had for like $200 new. btw, its a stock 4.0 right now.

Good idea. Once you get away from the range of about 14.5-14.9:1AFR, the "narrowband" or "knuckle" sensor loses response - and you'll know "rich" or "lean," but not how "rich" or "lean." Very important, when you get into troubleshooting...

I honestly think that if OEMs were to use WHEGO sensors, we could reduce emissions and increase fuel efficiency and economy by eliminating all that "dancing" the ECU ends up doing with the NHEGO sensors - but that's just me. Reprogramming the fuel response routine should not be all that difficult...

 

[BBeach]

Yeah, I know what you mean.

As I've highlighted above, you need to multiply the VE (real-world percentage) with the displaced volume/revolution (theoretical quantity) to get the amount of air you're actually moving. So, it's VEx241.5ci, or whatever that works out to in cubic centimetres (~3957.5cc - found my calculator.)

For more detail, I'll probably have to crunch some numbers myself, then "show my work" for it to make any sense. However, I can definitely say you're running a little lean - 15.5:1 isn't something that the computer will usually pass for cruise RPM (ideally, it doesn't go past 15.0:1 under normal operation - and should run just a touch rich, as I'd said, to avoid detonation. The catalytic converter should handle the HC and CO emissions that will result. However, the control computer tries valiantly to run at the "ideal" 14.7:1 ratio.)

Since my wife will be in New York for the week, I might tackle some of this one evening or another, and see what I come up with. If you don't see anything by Friday, would you please PM me a reminder?

Dont forget 5-90 :wantyou: This weekend im going to see what I can do to figure things out but its tough with finals and all.

 

[BBeach]

Any updates?

 

[czb83]

I'm redoing the wiring and connectors due to some intermittent issues, but the wires are still in freight. So no MegaSquirt playing until I get it in(hopefully in a couple days)

 

[5-90]

OK - you'll have to play with the formatting on this a bit (cut-and-paste into the text editor of your choice,) but here's the air and fuel consumption at stoichiometric fuel ratio (14.7:1, theoretical.)

Best cruise is closer to 14.5:1, and best power is down around 13:1. Also, the VE comes from simulation, so offset the fact that consumption will be slightly higher than stated with the fact that laminar airflow will be mildly disturbed by the cast surfaces.

FORMULAE AND CONSTANTS:
Air weight taken as .0807 pounds/cubic foot at STP
Fuel density taken as 5.98#/gallon at STP

Air Consumption: CFM = (cid * rpm)/3456
Weight of Air Consumed: #Air = CFM * .0807
Weight of Fuel Consumed: #Fuel = #Air/14.7
Volume of Fuel Consumed: Gal = #Fuel/5.98

I can, upon request, adjust for "best cruise" and "best power" ratios - I just didn't want to crunch that many numbers at once. I didn't get to this until earlier to-day, and I've been interrupted several times, so I'll have to fix it later.

This is a "cut-and-paste" from Excel into plaintext, which explains the formatting.

----- SNIP -----

2000 3000 4000
#2686 CFM 115.3 179.9 251.6
#2686 VE (%) 82.5 85.8 90.0
#2686 Air Wt (#/min) 9.3 14.5 20.3
#2686 Fuel Wt (#/hr) (14.7) 38.0 59.3 82.9
#2686 Fuel Vol (Gal/hr) (14.7) 6.4 9.9 13.9
#7120 CFM (14.7) 116.3 180.7 253.0
#7120 VE (%) 83.2 86.3 90.5
#7120 Air Wt (#/min) 9.4 14.6 20.4
#7120 Fuel Wt (#/hr) (14.7) 38.3 59.5 83.3
#7120 Fuel Vol (Gal/hr) (14.7) 6.4 10.0 13.9
#0630 CFM 115.3 179.7 251.6
#0630 VE (%) 82.5 85.7 90.0
#0630 Air Wt (#/min) 9.3 14.5 20.3
#0630 Fuel Wt (#/hr) (14.7) 38.0 59.2 82.9
#0630 Fuel Vol (Gal/hr) (14.7) 6.4 9.9 13.9

(I don't have flow data for the #0331 head, so it's not included. I'd tend to presume it is similar to the #0630, but I can't confirm.)

 

[BBeach]

Im assuming the first number is the head # because I see the 0630. At 2000rpm (close enough to 60mph)...
#7120 Fuel Vol (Gal/hr) (14.7) = 6.4 gallons per hour
In one hour you'll travel 60 miles giving you 9.375 miles per gallon. Maybe there's a factor of two missing in there somewhere or Im misinterpreting the data. When doing out the 3000rpm it came out to maybe 9.6mpg which might be accurate considering you're going 90mph. But for the 60mph at 2000rpm using 6.4 gallons per hour there might be something missing. But thanks a bunch for the data, how'd you get that off your jeep or is it some program? Perhaps the factor of vol efficiency is missing or that air/fuel isn't 14.7.

 

[5-90]

Correct - the numbers at the beginning of the lines are cylinder head numbers. I should have said:

#2686 used from 1987 to 1990
#7120 used from 1991 to 1995
#0630 used from 1996 to 1999/2000

And, this is pure fuel consumption. However, the TE (Thermal Efficiency) for most internal combustion heat engines runs around 30% or so - don't forget to factor that into MPG (along with frontal area, drag coefficient, and drivetrain parasitic losses.) Figuring MPG from pure maths isn't as easy as it might first seem. These numbers also don't account for gear ratios and the like.

Moreover, I stopped at 4000rpm for two simple reasons:
1) Most people won't cruise up around that high anyhow.
2) That's peak VE, per simulation, all heads.

Of course, these are also airflow numbers for bare cylinder heads - they'll be different (slightly lower) with manifolds and ancillary parts in place. This is data that I don't have at the moment. So, the maths are correct - but there are a number of assumptions not made, and there are quite a lot of data missing in these calculations, since I'm only working with the engine proper, and not the entire vehicle.

Make any more sense?

 

[BBeach]

So this is something that was done on a software program? Is so which is it and is it free to public? So basically at the 2000rpm, it calculated how much fuel it would use per hour under no to little load?

 

[czb83]

So, I guess you really need a MAF to measure it accurately. I found some cheap ones on ebay( < $10.00), but I'm not sure how they will fit. It would be neat to temporarily hook one up and measure airflow directly.

 

[5-90]

So this is something that was done on a software program? Is so which is it and is it free to public? So basically at the 2000rpm, it calculated how much fuel it would use per hour under no to little load?

It's an engine simulator, and it's not free. Right now, I'm using Desktop Dyno2000, but I'd like to upgrade to something that's a bit less, er, "optomistic" in its estimates. The curves for DD2K have been right about spot on for shape, but lower them by 2-3% (to account for parasitic drag and real-world pumping losses) to get more realistic numbers.

I'm not sure if there are any freeware engine simulators out there, but they'd probably be worth whatever you end up paying for them. Mathematical modelling of engine operation isn't the easiest thing in the world to do.

If you want to figure true VE, you're going to have to stick the engine in a dyno cell and have it run - that's the only real way to get any sort of accurate number. Using a chassis dyno still gives you the variables of parasitic drag in the drivetrain and (if you have an automatic) hydraulic pumping losses in the transmission - which can be fairly significant.

And, as I recall, dyno pulls are typically "full load" vice "no load" - else, how would you know how much power/torque you're making? They're still calculated values, but if you don't load the engine, you won't have anything to use as a basis for your calculations...

 

[Jim Schweikert]

Say you have a jeep with 4.0, AW4, 3.55 gears, and 235/75R15 tires (right there may be a variable because they may have rated with AX-15 and 215/70R15 for that matter...). At 60mph with the 235's you'll be turning about 700revs/min at the wheel. Therefore its about 1864rpm at the engine. For every 2 turns of the crankshaft, every cylinder will have taken in air (3955cc or about 4.0L). <---Now is it taking in 4.0L or the V.E. x 4.0L? That's my first question. But since the V.E. is just a constant at the exact rpm, I can toss it in later if need be. With the 1864rpm/2 = 932rpm x 3955cc = 3686L/min which is the same as 3686L/mile of Air. To find the mass of air you use the formula M=D/V and find that mass to be 4423.2grams of air per minute assuming 1.2g/L density. Using an air/fuel ratio of 15.5:1, that 4423.2/15.5= 285.37grams of gas per minute. Using wiki's density of gas and the known mass, you find the volume to be 2.5834L gas/minute. But.....thats .6825gallons per minute and therefore 41gallons per hour....which is just a little bit wrong. :huh:

Anyone know where I screwed up and if theres somewhere where V.E. of say 60% is in effect, or that air is only 23% oxygen, or that combustion obviously isn't 100%? And if so, does anyone know how much its off by and why? I'm not sure why I'm doing this but this is what you do when you're bored.

Just looking at you numbers I noticed the “M=D/V”
formula, as written, was incorrect. It should have been “M=DV”. In the calculations you had it correct.
I think one of the problems lies in the gasoline density gasoline is ~6 lbs. /gallon or 2724 grams/gallon. You calculate that you use 285 grams /minute or ~0.1 gallons/ minute. 0.1x60 minutes is 6 gallons per hour or 10 mpg.

 

[BBeach]

Correct, I lost my scratchwork but I did the last part over, I didnt notice the M=D/V but it should be the D=M/V . When I did it out the right way I get around 10mpg, but due to the EPA rating, were off by almost exactly 2, so I'll try to find my scratchwork sometime or do it over completely. But perhaps you could answer my question of what the vehicle speed is travelling in 4th gear locked to get the best mpg for the setup I presented. At some point in 4th gear the engine will have a prime air/fuel ratio that'll give me best results but I dont know if its at 45 when it first starts kicking in or at 80mph because people keep saying the "sweet spot" is 2700rpm though I havent seen anyone back it up. I would assume when the peak torque is reached, but mine is 3400rpm or so on the dyno and im not gonna go 100mph nor do I think the AFR at that speed would be well because of the drag forces on the vehicle making it run richer. Any thoughts?

 

[BBeach]

http://www.convert-me.com/en/convert/flow_rate_mass
http://www.convert-me.com/en/convert/flow_rate_volume
Good easy conversions.

 

[scottbky]

Are you still working on this? I just recently got a ScanGaugeII OBD2 plug in to monitor real time ECU stuff. I can supply some empirical data, as I have a 10 mile highway drive to and from work every day. My interest is purely to reduce the money leaving my pocket for gas.

ScottB

01 XJ Limited

 

[BBeach]

What kind of information does it give you? Welcome to naxja btw.

 

[scottbky]

It plugs into the OBD2 port and reports what the ECU is seeing.
Real time readouts (~2 sec update):
speed, rpm, mpg, water temp, intake air temp, fuel consumption rate, closed/open loop, fuel pressure, ignition timing, engine load, manifold pressure, throttle position, voltage.
Haven't used them all, yet, but some stuff isn't reported by my ECU, like fuel pressure, so no data.
Also has a trip mode that reports:
max speed, max water temp, max rpm, average speed, average mpg, miles, time, fuel used
Has some setups that can show for current tankful:
fuel used, fuel remaining, distance, distance remaining, time, time remaining
Also acts as a scan/reset tool. Already used it to diagnose a bad o2 sensor. Check out the website, a pdf manual there, too:
http://www.scangauge.com/
Amazon has some resellers for a few bucks less. Worth it to me.

ScottB

 

[BBeach]

Looks pretty cool to me, decent price too. If only it was a little cheaper Id get it. What have you been seeing for best highway speed for highest mpg?

 

[scottbky]

So far, best mpg is somewhere between 60-65, at 25-26 mpg. It's been awful windy around here, so I've been getting headwinds and tailwinds depending which way I go. rpm is right at 1900 at 60 mph, so maybe that's the "sweet spot."
I'm a little surprised; I thought best mileage would be at lowest rpm in highest gear, which is 45-50 mph and around 1600 rpm, when overdrive kicks in. There's a shorter flat stretch of 2-3 miles that I tried at 45, but the SG2 reported 22-23 mpg. Maybe the ECU is programmed a little rich for power right after a gear change.

Any progress on your mileage model? The prime air/fuel ratio may get muscled out by wind resistance.

An old engineering saw is that several small, independent, random, variables will tend to average themselves out.

ScottB

 

[5-90]

So far, best mpg is somewhere between 60-65, at 25-26 mpg. It's been awful windy around here, so I've been getting headwinds and tailwinds depending which way I go. rpm is right at 1900 at 60 mph, so maybe that's the "sweet spot."
ScottB

Not so. Ask any piston-driven pilot where he gets best fuel live, and it's at the torque peak of the engine. So, you gear for "best cruise" at or near peak torque.

Since our "peak torque" is agreeably broad and flat, we have a fairly wide cruise range. I'll typically cruise right around 70mph - which gives me 2600rpm or so in fourth gear. That whole "lowest RPM" thing is a misinterpretation by pretty much everyone - I picked up about 2-3mpg once I stopped using fifth gear! Others here have borne this out through their own experimentation as well.

The 3.55 gearing typically used with the AW4 complements it well, and you get good cruise RPM at typical speeds (45-50, 65-70 being most typical.) The 3.07 gearing used with the BA-10/5 and AX-15 is nothing but a mistake - don't know what they were smoking there. To make overdrive useful, 3.73 gearing would be better, as I recall (I'd have to crunch the numbers again, it's been a while.)

Peak Torque = peak VE = optimum fuel mileage. Coincidence? I think not. Gear for cruise right around 2600-2800rpm with the AMC242, and you'll be good. Peak cruise crankshaft speed will vary from engine to engine - a small block Chevvy will cruise at a different RPM than a Chrysler Hemi, than an old Slant Six (but the Slant Six will probably cruise at a lower rpm than either of the V8s and most V6s. And nearly all inline fours...)

Also, bear in mind that "optimum" AFR will vary, depending on what you're trying to do. "Stoich" (or "ideal") AFR is 14.7:1 by mass. "Best Rich" AFR runs around 13.5:1, and "Best Cruise" is a bit lean - at around 14.9:1. The ECU should be programmed to detect what's going on, and adjust fuel trim accordingly (I'd have to look up these numbers as well - I don't use them very often, so they're a little fuzzy...)