Pumping losses

[monroe74]

gary: "With the throttle closed 'pumping losses' are greatly reduced"

I think you have it backwards. Pumping losses are greatest when the throttle is closed.

As you said, you can think of your engine as an air compressor. But every air compressor is also a vacuum pump. It's just a question of which end you look at.

When intake is restricted (that is, when you are at any setting other than WOT), there is a vacuum in the intake manifold. It's obviously not an absolute vacuum, but it's a vacuum relative to atmospheric pressure. A lighter throttle setting means more restriction and more vacuum.

Your engine has to do work to create that vacuum. This work is wasted. It's energy that is not heading toward your wheels. That's why this graph looks the way it does:

http://www.autospeed.com.au/cms/gall...0&a=110216&i=6

WOT means you are wasting little or no energy creating a vacuum. That's why WOT produces optimal BSFC (at least on that graph; I think on some engines, in certain conditions and for specific reasons, the optimal setting might be 70-90%, instead of WOT).

I think proper use of heavy throttle can be a critical element of maximizing FE. This is perversely counterintuitive, and therefore I think it's one of the most challenging concepts in the world of hypermiling. In particular, I think it's the reason why extreme P&G can be so effective. In my opinion, the essence of P&G is to maximize throttle angle and minimize throttle duration. These two go in hand-in-hand; you can't do one without the other (assuming you want to keep your speed within a reasonable range). And note that this is the exact opposite of what most of us were taught to do, and still do: use light throttle, and maintain this close to 100% of the time. So for most of us, throttle angle is low and throttle duration is high. Trouble is, that's the opposite of what the BSFC graph is trying to tell us to do.

"losses due to continuous operation of the engine necessary in normal vehicle operation"

Those other losses (like the cost of running various accessories) are not normally called pumping losses. The term is normally used specifically to refer to the work involved in drawing ("pumping") air into the engine, and getting rid of it.

[monroe74]

"A 3 inch diameter piston in a perfect vacuum (combustion chamber) would have just over 7 sqaure inches of surface area. Multiply that by 14.7 and you get 102.9 pounds of force."

I think you're making a major mistake in overlooking this essential fact: when the throttle is open, atmospheric pressure is operating equally on both the top and the bottom of the piston.

"When you close the throttle (restrict airflow) in essence you are taking the engine to a high altitude instantly."

No. You're only taking the top of the piston to a high altitude. The bottom of the piston is still at sea level. That means as the piston travels toward the ground (Subaru and Porsche owners please make the appropriate mental adjustment), at a small throttle setting, it's fighting against atmospheric pressure. It has atmospheric pressure below it, and a vacuum (relatively) above it. You are requiring the engine to create a vacuum. This wastes energy.

"Now you have almost no compression"

That makes no sense. Of course you have no compression. The piston is moving down, not up. (So I suppose you could say it's compressing the air below it.)

"and your vacuum is also reducing the suction forces as well"

I wonder what suction forces you're talking about. The vacuum itself is a suction force, and the piston is fighting it. More throttle means less vacuum (suction), and less work for the piston to do.

"When you open the throttle to test compression the engine cranking speed drops, due to increased compression even though your suction has almost disappeared."

You have it backwards. When you open the throttle to test compression, cranking speed doesn't drop. It increases. That's the whole point of opening the throttle: to achieve a higher cranking speed, and get a more realistic, accurate compression reading. And the cranking speed went up because you relieved the motor of the burden of creating a vacuum. So the piston has atmospheric pressure both at the top and the bottom, instead of just at the bottom.

"Reciprocation forces are generally not considered because the assumption is they are inevitable"

Good point. But sometimes there are exceptions, like Wankel/Mazda.

[monroe74]

holy: "Why does a more free-flowing exhaust reduce torque?"

I think it has to do with the concept of resonance. The exhaust isn't moving in a steady flow. Rather, it's a series of pulses. Gas behavior inside an exhaust pipe has some important similarities to gas behavior inside a pipe organ (or any other musical tube, like a saxophone or a flute). This same concept explains why a high-performance header gives each cylinder a pipe that is equal in length to the others, even though it's not the shortest path to follow. I think there are similar issues on the induction side.

I think road is saying the same thing, in different terminology, when he talked about how the pulses move more efficiently when they are in phase with each other.

It's not an accident that we talk about "well-tuned" headers. It's a lot like tuning a musical instrument.

[theholycow]

Originally Posted by dosco

Diesels do not have a valve in the intake, which is one component of it's superior efficiency over spark-ignition engines.

Gasoline Direct Injection to the rescue! Available now in the Cadillac CTS and soon in the new base model v6 Camaro, coming to a Chevrolet dealer near you... No throttle, just like a diesel. Almost like a diesel engine with spark plugs and low [for diesel] compression.

Originally Posted by monroe74

In my opinion, the essence of P&G is to maximize throttle angle and minimize throttle duration. These two go in hand-in-hand; you can't do one without the other (assuming you want to keep your speed within a reasonable range). And note that this is the exact opposite of what most of us were taught to do, and still do: use light throttle, and maintain this close to 100% of the time. So for most of us, throttle angle is low and throttle duration is high. Trouble is, that's the opposite of what the BSFC graph is trying to tell us to do.

This is a concept I've been using since I began hypermiling. You can even have long duration if you manage to keep your RPMs low. At low RPMs, with near-WOT, you're looking at no restriction, totally free flow of all the air it can gulp. I accelerate at near-WOT, using early shifting to limit my acceleration; by varying my shift points I can accelerate as slow or fast as the rest of the traffic, generating the same amount of power in a higher gear with less pumping loss.

[monroe74]

holy: "You can even have long duration if you manage to keep your RPMs low."

Just to be clear, I was saying that long throttle duration is generally a good thing to avoid. But as long as you're maximizing BSFC, by having WOT in a high gear (i.e., low RPM), then it's OK to stay in this mode as long as you need to. Because the only choice that's more efficient is staying home! I think you understand this, but I wanted to try to avoid confusing people.

Anyway, what you described is basically identical to the way I drive. The one thing that interests me is that you use near-WOT, instead of WOT. I can guess certain reasons why you might do this, but I'm curious about your views on this.

I've done some experimenting both ways, but I don't have enough data yet to reach a conclusion.

[R.I.D.E.]

Guys I post here to contribute information not to be insulted by assumptions about my background knowledge.

To prove my point beyond question here is a test scenario

Same road same speed same car.

Accelerate to a certain speed:

1. Coast in neutral to a lower speed, note time and distance
2. Decelerate using the exact same shift points you used in acceleration, downshifting with your foot off the gas with engine off at exactly the same points you upshifted
3. Decelerate using same shift points as 2 with accelerator at WOT, of course with engine off.
4. Remove spark plugs push car to same speed and repeat 3.

Scenario 1 demonstrates total energy losses without engine as a factor.
Scenario 2 demonstrates total energy losses with throttle closed-engine included (your maximum pumping losses although they are not moving the maximum amount of air through the engine.
Scenario 3 demonstrates total energy losses with WOT otherwise identical to 2. Now you are moving the maximum amount of air through the engine.
Scenario 4 demonstrates total losses without suction or compression losses which are all the other relevant losses inherent in IC engines.

Which scenario will give you distance travelled that is closest to neutral coast which is your base measurement?

Now we can debate the point to infinity.

This test can be done with some effort and it will demonstrate what I tried to explain in my previous post.

If your knowledge is so superior that you can state wrong or recommend reading material, maybe you can predict the results without having to actually do the experiment.

i know the results, like to guess why.

regards
gary

I find real data to be the true test. Have you ever conducted such a scientific experiment?

It may be your book has flawed data or it was not completely understood.

Since you seem to believe suction losses exceed compression losses, I can only assume you would think scenario 2 would be less distance than scenario 1 since you assume it takes more work to create vacuum than compression.
Scenario 2 has no suction losses since you have WOT. You do however have maximum compression. While scenario 3 will eliminate almost all of your compression and suction losses you will experience the losses due to reciprocation.

It's 3 separate losses, and this test is a way to separate those individual losses, which allows them to be estimated separately.

You might just learn something about losses that are relevant and the relative percentages.

[R.I.D.E.]

Monroe your conclusion about WOT operation assumes the BSFC map for a VX is identical to a map for a non lean burn engine, which is a false assumption.

You also assume that WOT is preferrable to 70% which is more than enough throttle position to fill the cylinders (check it with a vacuum guage if you wish to not believe this) at low engine speeds when WOT is not necessary. Check your BSFC map at 50 % compared to higher throttle opening positions AT LOW ENGINE SPEEDS where it is relevant to our discussion (although your reference map is not relevant).

You also choose to ignore the fact that if you maintain lean burn you are using about 60% of the fuel you would use at WOT (check AF ratios of WOT and lean burn).

This is why your conclusion is basically flawed, the map you are using as a reference is not a VX engine map.

The basic design of the VX lean burn system is to increase efficiency at throttle positions that would be less efficient in a normal 1.5 liter non lean burn engine.

Thats why I use less than 50% throttle position, because I can pulse quite a bit longer for the same amount of fuel which increases the total distance of one P&G cycle.

If you want to reply try to be somewhat civilized gentlemen. I find communication flourishes when it is conducted with civility.

I use the sil indicator to determine throttle position when I accelerate. It takes me 13 seconds in 5th gear to accelerate from 50 to 65 MPH, so you can duplicate that time and know exactly the amount of throttle I am using.

I only pulse in 4th gear when my initial speed is below 30 MPH and the light never comes on during the pulse.

Last but not least is the mileage figures I achieve, and remember I am not using engine off coasting, which would increase those figures by about 15%.
I made it clear that I don't feel comfortable using EOC in my situation. That certainly does not mean I dont recognize the additional benefit of EOC. I have used it under very controlled situations without traffic to worry about, and I feel that it would affect my situational awareness and increase the chances of getting in an accident.

regards
gary

[monroe74]

"Scenario 2 has no suction losses since you have WOT."

Earlier you said this: "3. Decelerate using same shift points as 2 with accelerator at WOT." I think you're getting your numbers mixed up. That's one reason I'm finding your example hard to follow, but it's not the only reason.

Anyway, I think your example reduces to the following much simpler comparison.

A) Take an engine that's not running, and spin the crankshaft, with the throttle closed.
B) Take an engine that's not running, and spin the crankshaft, with the throttle open.

I think you're claiming that A takes less work than B. Really?

"your conclusion about WOT operation assumes the BSFC map for a VX is identical to a map for a non lean burn engine"

No, I'm not claiming it's identical. I'm just claiming that there's enough similarity to make for a worthwhile comparison.

"the map you are using as a reference is not a VX engine map"

True. But I think it's helpful, as compared with using no map at all. If you have a map or reference that you think is superior, I hope you will give us a chance to see it.

I think it's clear that it's possible to save gas by utilizing lean burn, and it's possible to save gas by minimizing pumping losses. The interesting thing is that it's just not possible to do both of those things at the same time. Lean burn is generally associated with relatively small throttle settings. Trouble is, small throttle settings are generally associated with large pumping losses. Therefore it's necessary to make a choice. And the right choice probably depends on the immediate situation.

P&G is a technique that relies on relatively large throttle settings. And there seems to be lots of evidence that P&G works (and I think it works primarily by minimizing pumping losses). Trouble is, P&G is basically incompatible with lean burn, because in the pulse phase, you're accelerating. It's generally possible to maintain lean burn while at a steady speed, but not while accelerating. So we're back to that choice again, to either minimize pumping losses, or maximize lean burn. I have a hunch that the former strategy is more effective, when driving conditions allow for P&G (ideally in the range of 25-45 mph, I think). When driving conditions require a steady speed (especially at higher speeds, like 50+), then my hunch is that it makes sense to maximize lean burn. I'm not sure about this, and I'm open to evidence that I'm wrong. But I just haven't seen that evidence yet.

"Check your BSFC map at 50 % compared to higher throttle opening positions AT LOW ENGINE SPEEDS"

I have. It shows that WOT is more efficient than 50%. Not always by a wide margin, but nevertheless the margin is positive. Therefore it's probably worth taking. It also shows that at 1000 rpm, the margin is actually quite large. And 1000 rpm/WOT is a state I use quite frequently.

"You also assume that WOT is preferrable to 70%"

True, I do. I don't believe the margin is always large, but I believe the margin is always positive. That's what I see on the graph we're discussing. On the other hand, I think you assume that 70% is always preferable to WOT. I'm open to believing this, but so far I've seen no proof.

"You also choose to ignore the fact that if you maintain lean burn you are using about 60% of the fuel you would use at WOT"

I think the number might be even less than 60%. But what's more important than the air/fuel ratio is the work/fuel ratio. In other words, it's important to approach the analysis in terms of BSFC, not absolute fuel consumption.

"Thats why I use less than 50% throttle position, because I can pulse quite a bit longer for the same amount of fuel which increases the total distance of one P&G cycle."

If I understand correctly, you're using P&G, and you're claiming that you're in lean burn during the pulse phase. May I ask how you know? My experience using a DMM to monitor my O2 sensor leads me to believe that lean burn generally does not occur during acceleration, even mild acceleration. And this is consistent with the comments of others who have monitored lean burn in this manner. How are you monitoring lean burn?

"I use the sil indicator to determine throttle position when I accelerate. It takes me 13 seconds in 5th gear to accelerate from 50 to 65 MPH, so you can duplicate that time and know exactly the amount of throttle I am using."

I'm very confused by that paragraph, because the SIL is off 100% of the time that you're in 5th gear, regardless of vehicle speed, engine speed, or throttle position.

"I only pulse in 4th gear when my initial speed is below 30 MPH and the light never comes on during the pulse."

If you are in any gear other than 5th, and using a throttle setting large enough to keep the SIL turned off, then you are almost certainly not in lean burn.

[theholycow]

Gary, I think you might be reading more feeling in Monroe's posts than he intends. Just a guess.

Everybody, please use the built-in quote feature. It's much easier to follow, as it does a better job separating and highlighting text. It's not hard to trim quotes and respond to individual points, you just have to put in ['quote'] and ['/quote'] tags like the ones that are automatically inserted when you click "Quote".

Originally Posted by monroe74

The one thing that interests me is that you use near-WOT, instead of WOT. I can guess certain reasons why you might do this, but I'm curious about your views on this.

I'm avoiding open loop. What I've read is that when you are at WOT, the ECU goes into open loop mode, ignoring the post-cat o2 sensors (if that makes sense) and using pre-set fuel tables instead of metering fuel based on actual measured conditions. In more simple terms (for anybody reading this who is more of a layman than me), it just dumps fuel into the engine rather than trying to be efficient.

Making things more complicated for me, my VW has a Drive By Wire system, so the gas pedal is not necessarily a throttle control. I am not sure that 50% of the gas pedal == 50% throttle opening, or that 100% of the gas pedal == WOT.

Now, that said, I'd really like a way to know in real-time whether I'm in open loop or closed loop. I think I may be able to floor the pedal and still not be in open loop at low RPMs. I started using a Vag-Com cable to monitor it on a laptop, but I can't tell if it's lagging a few seconds in its readings or if it really doesn't go into open loop easily. It takes until like 3500 or 4000 rpm before it shows open loop, then stays open loop a couple seconds after I take my foot off the gas.

Since this thread discusses P&G quite a bit, I might as well mention this: I have been tried P&G (neutral, not EOC) and it looks promising. Based on the mileage and position of my gas gauge I'd guess I've gained at least 2 mpg over my previous record, but it's hard to tell (harder still because of a minor technical difficulty I'm having with my digital LCD odometer reading in km instead of mi).

[R.I.D.E.]

The source of my assumptions is what I read in this forum, that doesn't make it a fact any more than everything you read in any book makes it a fact. Its all hypothetical until real testing proves otherwise. Try the throttle closed versus open test that I suggested. If you coast down distance is lower with throttle open I am correct (more resistance at WOT). If it is lower with throttle closed you are correct (more resistance at closed throttle). The reason I think I am correct is because compression is greater at WOT, while suction losses are insignificant at WOT. A simple test that proves my point.

I have done thousands of compression tests and the conclusion I have reached is the cranking speed is lower with WOT. My reasoning to support that is simple. You are pumping more air through the engine at WOT, because the compression is lower if you don't open the throttle. Since suction losses are practically nil at WOT I can't rationalize any other conclusion.

Please let me know the results of your test.

Think about the Jake brake on big rigs, they are using trapped compression, not vacuum, obviously they can't use vacuum. The difference in compression versus vacuum is directly proportional to the compression ratio. No closed throttle will produce a perfect vacuum, but in my previous example with 185 pounds of compression you have over ten times the resistance that 14.7 negative pounds of vacuum would ever produce.

I wish I could prove when lean burn is engaged. Since that is not possible I rely on those contributors to this forum who have developed a means of proving that it is engaged when climbing a slight grade in 2nd, 3rd, 4th, and 5th gears. In fact in the links Tom O proivided there is information that justifies that assumption. The contributor was one of the hypermilers that was involved with Wayne Gerdes, who made the statement that lean burn was possible in lower gears and climbing grades.

"You can climb a slight grade and still stay in lean burn, even in lower gears".
The use of lower gears would indicate to me that steeper grades would still allow lean burn as long as you used lower gears.

Think of it this way, we know the VtecE closes one valve almost completely to create swirl and improve low speed fuel air mixture (below about 2500 RPM). Homogenous Charge Compression Ignition relies on close to perfectly even distribution of the fuel particles in the fuel air charge. Testing has shown that this creates the potential for compression ignition of gasoline with a 25% improvement in efficiency.

You can check that by googling HCCI. The problem is its presently not possible to accomplish HCCI over all the loads and ranges of normal engine speed. I takes some time for the mixture to completely emulsify which brings on the problem of having a fairly significant amount of fuel air mix in the manifold, which would be a scary thing if you had a backfire. One solution is ultra high injector pressures, but it may be that it will not be possible with fuel injection occuring soo close to the intake valves. We may actually have to go back to a form of throttle body injection with mechanical turbulence and preheating of both air and fuel to allow better fuel particle distribution and to offset the fact that when vaporized fuel tends to become cooler. In the carb era without preheating of the air in cold temps you had carb icing where the fuel would not suffeciently atomize. The result was hesitation to the point of stalling. Its the same principle as warm air intake which while reducing the mass of air going in the engine also allows for better atomization of the fuel particles. Preheating the fuel would also help, but when the fuel is heated you risk vaporizing the fuel in the lines which causes serious problems.

My position is lean burn requires better atomization in order to function at 22-1 AF ratios. While it is not true HCCI at certain speeds and loads it is at least a significant step towards HCCI.

We disagree on when lean burn is actually occuring. I am at a disadvantage with my old time pre computer education which ended in the late 60's. The shift indicator light is there to tell you when to shift for best economy. To me it makes no sense that the light would be on when lean burn was working but off when it wasn't. The light uses several imputs, but basically it is telling you your vacuum is to high for efficient operation. Thats why I believe (can't prove it other than my mileages) that lean burn occurs if you stay just above the point where the light comes on and below 2500 rpm. Of course the light does not come on in 5th gear, but you can still use the same throttle position percentage. It doesnt take much throttle to accelerate from 50-65 in 5th gear in 13 seconds. The coasting time is about 17 seconds,and it may be better to just forget P&G at those speeds. Other forum members have also stated that it is difficult to achieve the percentages of improvement in average mileage with a VX, compared to the same year vehicle without lean burn. Thats easy to understand when you look at the difference in combined mileage for both vehicles with the difference about 33%.

I enjoy civilized debate and hope to learn from our interaction. It certainly is possible that your position is correct and mine is less correct.

I don't think a "I am totally right and you are totally wrong" position would produce any learning.

regards
gary

My rationale is based on this statement by a fellow hypermiler who apparently has a means of verification.

"I can maintain lean burn up to 70 MPH"

TomO stated that lean burn is a function of several imputs one of which is the load placed on the engine.

If I use the same throttle position as I would at a constant 70 mph, I should be able to stay in lean burn. At lower speeds that means you would be accelerating since total energy losses are lower at lower speeds maintaining the same load would allow for acceleration.

If the potential for lean burn did not exist, I would only disagree with you on the throttle position. In that case I would use 75% because at WOT you are creating enrichment that doesn't exist at 75%.

If you looked at a BSFC map for the VX (even though I don't know where on is or I would provide a link) it would be two maps.

One in lean burn

One outside of lean burn

Combining the two would be confusing to say the least.

This is what I believe. Mr Honda was a pioneer in lean burn technology, going back to the stratified charge prechamber CVCC engines in the 70's. He died in 1991 the year the first VX was produced. He was not a highly educated man, but his intelligence is unquestionable. After his death the driving force behind lean burn technology seems to have been gradually lost. Combined with increasingly stringent emissions requirements lean burn fell out of favor, principally due to NOX emissions.

Lean burn was a way of compensating for the poor BSFC of normal engines when there is any significant throttle restriction ,and it was a practical solution that proved to be more economical.

Try one more test.

Hook up a vacuum guage and see how little throttle opening you can use to achieve the lowest possible vacuum reading at low engine speeds. I beleive the throttle opening percentage would be below 25%, because you are only running the engine at about 1/5 of maximum speed (give or take).

I believe the sil light uses vacuum as its principle imput (not the only imput), to show the driver when upshifting will create a greater load and lower vacuum to produce the same work, which I think we can agree is more efficient.

regards
gary