Keith Tanner wrote:
One cool thing about modern engines - the throttle pedal isn't a throttle pedal. It's a torque request. The engine then does what needs to be done to deliver that torque. Cam timing, ignition timing, throttle plate position, assist from a hybrid system, whatever. I know the Skyactiv engines are messing around with the throttle opening a lot more than you think, manipulating things like fuel useage and cat temperature.
Being that people drive using power, it would be better if it was a power demand. Being torque- it's hard to optimize the BSFC- which is fuel per power number (aka thermal efficiency). It's also quite interesting to see the fuel consumption maps in the power domain and not the torque domain. The islands take on a very different shape.
It's an odd thing.
rslifkin wrote:
Keith Tanner wrote:
The LS engines are great on the highway, but thirsty around town. I swear they use more fuel at idle than they do at 60 mph.
There's a reason for that. Big engine does have a higher idle fuel consumption than a small engine. A lot of stock LS powered cars were also not terribly light, but have decent aerodynamics and tall highway gearing. So it takes some grunt to get it up to speed from a stop. Add that with the high idle consumption and boom, E36 M3ty fuel economy around town.
On the highway, with good aero, you're not using a ton of power and are likely in a more efficient operating range than idle, so it's probably not burning a whole lot more fuel, and thus, it ends up being fairly efficient.
Most of my experience is with them in 2400-2600 lb packages that aren't very aerodynamic 
On the mpg-meter thingy, the one in the Jeep is definitely inaccurate. If I tweak the calibration factor for it in my tune, I can get it to be consistently within 0.1 - 0.2 mpg in DD use. But it will consistently read lower than actual mpg on a long highway run. When the thing was bone stock, it tended to read a little high around town and anywhere from accurate to a little low on the highway.
Aspen
Reader
6/7/16 3:25 p.m.
kb58 wrote:
Two other side notes:
1. In this day and age of super exacting ECU measurements, it amazes me - in a bad way - how the ECU's MPG estimate vs the measured value always "happens to be" off in the direction that favors the mfg. My truck regularly claims 19.8 mpg but measured it's consistently about 10% less. 10% is a lot when ECUs are measuring things in microseconds.
2. Re best cruising speeds, it amazes me how the Ford V6 Ecoboost drivetrain will leave the engine in 6th gear all the way down to an engine speed of 1500 rpm, and it still has torque to go up slight grades. Technology is pretty amazing.
I believed this is intentionally done to reduce customer complaints about low fuel economy. If the customer is seeing 2 mpg better than actual they are less likely to complain about low economy vs. EPA stated numbers. It reduces the number of people coming in asking for warranty repairs to a car that is functioning as designed. It seems that very few people actually get out the calculator to calculate their actual consumption figures.
The one in my Volvo was 3mpg high until I fixed the fuel pressure regulator that was A: not pulling fuel pressure down and B: blowing fuel into the intake manifold. So the injector pulsewidths were artifically low. (Oddly enough it never once set a system rich code - the allowable range must be tremendous)
Now it reads accurately. The only caveat is, it loses accuracy the more idling you do, since the average is the average of miles driven versus MPG reading during that time. When you're stopped, it doesn't read any MPG because divide by zero. But if you were getting 20mpg when moving, it will show 20mpg average, even if you were idling for an hour.
That's an odd way to calculate it Knurled. Mine seems to work based on injector pulsewidth to estimate fuel use (and it definitely accounts for idle time). Issue is, I don't think it accounts for there being slightly less flow at the start and end of each pulse.
rslifkin wrote:
That's an odd way to calculate it Knurled. Mine seems to work based on injector pulsewidth to estimate fuel use (and it definitely accounts for idle time). Issue is, I don't think it accounts for there being slightly less flow at the start and end of each pulse.
But that is how it's calculated. And the pulse open and close flow changes are all factored into the injector calibration.
How difficult is the NVH problem with large displacement, few cylinder designs?
I ask because I don't remember Porsche 944s rattling my teeth out. The 300 ci straight six was pretty smooth as well. The new Mini triples as well.
944s used balance shafts to quell the shaking. Generally, you tend to find them on 4 cylinders with more than about 2L capacity. Straight sixes are inherently the best balanced engines this side of a V12 (which is just two straight sixes), so they can handle bigger cylinder capacities.
I used to know a great website with illustrations. There are a bunch of similar sites out there with animations, it's really cool stuff.
alfadriver wrote:
rslifkin wrote:
That's an odd way to calculate it Knurled. Mine seems to work based on injector pulsewidth to estimate fuel use (and it definitely accounts for idle time). Issue is, I don't think it accounts for there being slightly less flow at the start and end of each pulse.
But that is how it's calculated. And the pulse open and close flow changes are all factored into the injector calibration.
Exactly, the computer knows the injector characteristics twelve ways from Tuesday. The calculations are way involved, because the computer needs to know to the milligram (finer?) how much fuel is going into the cylinder per cycle. (Note that I didn't say how much fuel is being injected - not all of the fuel makes it in, and not all of the fuel going to the cylinder is coming from the injector... MAN I bet DI actually makes this simpler)
However, it's not really a very smart driver display.
Keith Tanner wrote:
944s used balance shafts to quell the shaking. Generally, you tend to find them on 4 cylinders with more than about 2L capacity. Straight sixes are inherently the best balanced engines this side of a V12 (which is just two straight sixes), so they can handle bigger cylinder capacities.
I used to know a great website with illustrations. There are a bunch of similar sites out there with animations, it's really cool stuff.
Fours are a "special" case because an even firing four cylinder will have all four pistons at TDC/BDC at the same time and all four pistons will be at 90xxDC at the same time. When at the DCs, friction and piston forces on the crank drop to zero, and at 90deg out, friction is at its highest. That makes for way uneven friction on the crank. Also, when at 90deg out, all of the pistons will be even with each other, but more than halfway down the stroke (forms right angle: one leg is .5 stroke, hypoteneuse is rod length, the other leg is how far down the bore the pistons are at. NO right triangle has the hypoteneuse the same length as one of the legs!)
So what you get is the center of mass of the reciprocating assembly keeps oscillating the difference between .5 stroke (average of TDC and BDC), and whatever that triangle's other leg is. The shorter the rod, or the longer the stroke, the more the oscillation distance. The heavier the pistons and rod small-ends, the more mass is moving. So big fours with heavy pistons and short rods to fit under car hoods make a big heaps vertical boing-boing-boing-boing...
This reminds me of a funny thing from USENET a while back. (Early-mid 1990s) Someone who really hated 944s said they were the worst engine because they shook so bad they needed TWO balance shafts!
Well, no, all four cylinders need two balance shafts. You need two balance shafts so they can rotate the opposite direction of the other so they cancel each other out side-side since the only shake force you need to stop is vertical. A single balance shaft would not do any good against the forces we're concerned about.
To this day I'm not sure if it was a case of someone who was bound and determined to hate something and saw everything about it as a reason why it was Bad, or someone who was banking on other peoples' ignorance to use neutral facts in a negative light. Anyway, this is how propaganda works against people not-well-versed in any particular topic.
Knurled wrote:
alfadriver wrote:
rslifkin wrote:
That's an odd way to calculate it Knurled. Mine seems to work based on injector pulsewidth to estimate fuel use (and it definitely accounts for idle time). Issue is, I don't think it accounts for there being slightly less flow at the start and end of each pulse.
But that is how it's calculated. And the pulse open and close flow changes are all factored into the injector calibration.
Exactly, the computer knows the injector characteristics twelve ways from Tuesday. The calculations are way involved, because the computer needs to know to the milligram (finer?) how much fuel is going into the cylinder per cycle. (Note that I didn't say how much fuel is *being injected* - not all of the fuel makes it in, and not all of the fuel going to the cylinder is coming from the injector... MAN I bet DI actually makes this simpler)
However, it's not really a very smart driver display.
Better in some ways, worse in others. There's still a lot of noise on the 1-2% range, but that noise isn't as bad of a problem as most may think.
DI has fuel that is lost, too. Just different.
For my job, the noise is actually welcome.
Knurled wrote:
Keith Tanner wrote:
944s used balance shafts to quell the shaking. Generally, you tend to find them on 4 cylinders with more than about 2L capacity. Straight sixes are inherently the best balanced engines this side of a V12 (which is just two straight sixes), so they can handle bigger cylinder capacities.
I used to know a great website with illustrations. There are a bunch of similar sites out there with animations, it's really cool stuff.
Fours are a "special" case because an even firing four cylinder will have all four pistons at TDC/BDC at the same time and all four pistons will be at 90xxDC at the same time. When at the DCs, friction and piston forces on the crank drop to zero, and at 90deg out, friction is at its highest. That makes for way uneven friction on the crank. Also, when at 90deg out, all of the pistons will be even with each other, but *more than halfway down* the stroke (forms right angle: one leg is .5 stroke, hypoteneuse is rod length, the other leg is how far down the bore the pistons are at. NO right triangle has the hypoteneuse the same length as one of the legs!)
So what you get is the center of mass of the reciprocating assembly keeps oscillating the difference between .5 stroke (average of TDC and BDC), and whatever that triangle's other leg is. The shorter the rod, or the longer the stroke, the more the oscillation distance. The heavier the pistons and rod small-ends, the more mass is moving. So big fours with heavy pistons and short rods to fit under car hoods make a big heaps vertical boing-boing-boing-boing...
Or you could say that in primary frequency, 4 cyl engines are balance, but in a secondary one they are not- which is why the balance shafts spin at twice engine speed, and opposite directions.
The Bosch automotive handbook has a good reference section on this. It has a handy table that shows the first and second order balance/imbalance of all the common cylinder configurations.
But, NVH isn't all about engine balance... its simple math. More firing events of a lower amplitude are going to be smoother in general than less firing events of a higher amplitude.
alfadriver wrote:
rslifkin wrote:
That's an odd way to calculate it Knurled. Mine seems to work based on injector pulsewidth to estimate fuel use (and it definitely accounts for idle time). Issue is, I don't think it accounts for there being slightly less flow at the start and end of each pulse.
But that is how it's calculated. And the pulse open and close flow changes are all factored into the injector calibration.
In the Jeep, the computer actually doesn't give a crap about that part of the injector data. The only things it has in the tune are the min pulsewidth and offset vs voltage.
Being speed density, there's no VE tables or anything, so the fuel table is just raw injector pulsewidth values that it then applies various adders and multipliers too. Because it's not calculating the injector pulsewidth from scratch, it doesn't ask for the injector slopes, etc. That's all the human's problem.
rslifkin wrote:
alfadriver wrote:
rslifkin wrote:
That's an odd way to calculate it Knurled. Mine seems to work based on injector pulsewidth to estimate fuel use (and it definitely accounts for idle time). Issue is, I don't think it accounts for there being slightly less flow at the start and end of each pulse.
But that is how it's calculated. And the pulse open and close flow changes are all factored into the injector calibration.
In the Jeep, the computer actually doesn't give a crap about that part of the injector data. The only things it has in the tune are the min pulsewidth and offset vs voltage.
Being speed density, there's no VE tables or anything, so the fuel table is just raw injector pulsewidth values that it then applies various adders and multipliers too. Because it's not calculating the injector pulsewidth from scratch, it doesn't ask for the injector slopes, etc. That's all the human's problem.
Most of the systems have all of the flow issues built into the injector calibration. Why jeep did not- dunno. We've had it all factored in since I've been doing this- which goes back to the early 90's at a bare min.
It's not as if there were better systems than a straight map- pulsewidth table back in the late 80's even.
alfadriver wrote:
Most of the systems have all of the flow issues built into the injector calibration. Why jeep did not- dunno. We've had it all factored in since I've been doing this- which goes back to the early 90's at a bare min.
It's not as if there were better systems than a straight map- pulsewidth table back in the late 80's even.
Yeah, it's the unfortunate discrepancy between "what they could/should have done" and "the cheap, E36 M3ty, lazy way they did it"
alfadriver wrote:
Knurled wrote:
Keith Tanner wrote:
944s used balance shafts to quell the shaking. Generally, you tend to find them on 4 cylinders with more than about 2L capacity. Straight sixes are inherently the best balanced engines this side of a V12 (which is just two straight sixes), so they can handle bigger cylinder capacities.
I used to know a great website with illustrations. There are a bunch of similar sites out there with animations, it's really cool stuff.
Fours are a "special" case because an even firing four cylinder will have all four pistons at TDC/BDC at the same time and all four pistons will be at 90xxDC at the same time. When at the DCs, friction and piston forces on the crank drop to zero, and at 90deg out, friction is at its highest. That makes for way uneven friction on the crank. Also, when at 90deg out, all of the pistons will be even with each other, but *more than halfway down* the stroke (forms right angle: one leg is .5 stroke, hypoteneuse is rod length, the other leg is how far down the bore the pistons are at. NO right triangle has the hypoteneuse the same length as one of the legs!)
So what you get is the center of mass of the reciprocating assembly keeps oscillating the difference between .5 stroke (average of TDC and BDC), and whatever that triangle's other leg is. The shorter the rod, or the longer the stroke, the more the oscillation distance. The heavier the pistons and rod small-ends, the more mass is moving. So big fours with heavy pistons and short rods to fit under car hoods make a big heaps vertical boing-boing-boing-boing...
Or you could say that in primary frequency, 4 cyl engines are balance, but in a secondary one they are not- which is why the balance shafts spin at twice engine speed, and opposite directions.
I was aware of the secondary imbalance problem. It just seems like the nvh problem is pretty easy to solve with counter balance shafts.
Is there some other obstacle?
DaewooOfDeath wrote:
I was aware of the secondary imbalance problem. It just seems like the nvh problem is pretty easy to solve with counter balance shafts.
Is there some other obstacle?
Well, balance shafts cost money, space, and friction. Other than that.... 
In reply to alfadriver:
Very true ...
But more than going to a 6 cylinder?
alfadriver wrote:
In reply to rslifkin:
What year is your jeep?
It's a 98. Uses a Chrysler JTEC ECU in dirt simple form (speed density, no knock sensors, etc.)
