Efficiency of a modern internal combustion engine

[Ulysees]

You'll be laughing in the end when you realize how simple it really is.

 

[Ulysees]

I think the complications you get drawn to arise because you imagine the car on a dyno and that raw dyno measurements are shown on the power/torque curves. But such curves would be car-specific, not engine-specific.

Manufacturer gives power/torque from the axle, the engine output. Where that torque goes, how much goes to fight friction, how much goes into acceleration, makes little or no difference in instantaneous consumption per second, or efficiency.

 

[Mech_Engineer]

Based on your previous posts, you are even farther from understanding the dynamics of the problem that I thought, and it's pretty obvious that you have no interest in learning why you are wrong.

 

[Ulysees]

Would it help if I drew a system diagram for you?

 

[Mech_Engineer]

I think it would be a lot more useful for you to first define exactly what efficiency you are looking for, and how you plan to calculate it.

 

[TVP45]

Ulysees,
The power chart you are using requires calibration to road use before you do the kinds of calculations you are trying. Actual output will then be in the neighborhood of 8 MJ/l. This is common for lab measurements and something you have to learn to take into account. I suggest you contact the company that made the power chart and ask them about the details of calibrating that to road driving.

Carnot can sleep easy tonight.

 

[Ulysees]

I think it would be a lot more useful for you to first define exactly what efficiency you are looking for, and how you plan to calculate it.

The variability of efficiency has been addressed:

Efficiency would not be the maximum possible under these stressful conditions, it would only give some idea. But you have no way to measure the maximum efficiency without opening up the engine output part and connecting something to the axle

In other words, I'm happy with just some idea of efficiency. It is obvious that efficiency would vary somewhat, from the outset of this effort. Ideally we'd want the full range of efficiency. Ie the maximum and the minimum observed over all combinations of rpm and torque.

But we don't have the equipment to do that. Manufacturer has the option of measuring power by attaching a generator directly to the axle, but we don't. And using a dyno under the car would require additional knowledge of the transmission characteristic, ie power loss as a function of rpm (probably what tvp45 cryptically calls calibration before going into poetic mode) and perhaps something to measure the force of the car on the dyno.

So since we have no knowledge of the transmission losses, not to mention no access to a dyno, the best we can do is what we've done. To just get some idea of the efficiency. The true range of efficiency is probably on both sides, starting below and ending above the 3500rpm-full-throttle efficiency that we've measured.

 

[TVP45]

The variability of efficiency has been addressed:


In other words, I'm happy with just some idea of efficiency. It is obvious that efficiency would vary somewhat, from the outset of this effort. Ideally we'd want the full range of efficiency. Ie the maximum and the minimum observed over all combinations of rpm and torque.

But we don't have the equipment to do that. Manufacturer has the option of measuring power by attaching a generator directly to the axle, but we don't. And using a dyno under the car would require additional knowledge of the transmission characteristic, ie power loss as a function of rpm (probably what tvp45 cryptically calls calibration before going into poetic mode) and perhaps something to measure the force of the car on the dyno.

So since we have no knowledge of the transmission losses, not to mention no access to a dyno, the best we can do is what we've done. To just get some idea of the efficiency. The true range of efficiency is probably on both sides, starting below and ending above the 3500rpm-full-throttle efficiency that we've measured.

There's nothing cryptic about calibration - it's a commonly used engineering phrase for mapping one set of data (in this case a graph) onto another.

 

[Mech_Engineer]

The variability of efficiency has been addressed:

In other words, I'm happy with just some idea of efficiency. It is obvious that efficiency would vary somewhat, from the outset of this effort. Ideally we'd want the full range of efficiency. Ie the maximum and the minimum observed over all combinations of rpm and torque.

You still haven't explicitly stated what kind of efficiency it is you're looking for, but it seems to me you're trying to find an efficiency that cannot be measured by the methods you're currently utilizing...

From what I can tell, you're wanting to measure the efficiency of the engine itself (chemical energy into mechanical work out), but that can't be measured without removing the engine from the vehicle and bolting it to a specially designed engine dynamometer. It's important to note this sort of measurement is where your power output chart came from. Manufacturers measure the average power output of their engines before they go in the vehicle so that they can advertise the highest numbers available from the engine. If they advertised the power put to the ground (through the transmission, driveshafts, and axles), the numbers would be significantly lower because of power losses in each component of the power/drivetrain.

But we don't have the equipment to do that. Manufacturer has the option of measuring power by attaching a generator directly to the axle, but we don't.

That's not how the manufacturer measures the power output of their engines, they use an engine dyno that bolts directly to the engine's crankshaft before it ever goes in the vehicle. Mouting to the axle will essentially give you the same numbers that rollers on the floor do, because power is still being transmitted through the entire drivetrain, minus rolling resistance from the tires.

And using a dyno under the car would require additional knowledge of the transmission characteristic, ie power loss as a function of rpm (probably what tvp45 cryptically calls calibration before going into poetic mode) and perhaps something to measure the force of the car on the dyno.

This is a VERY common way to measure net power output on performance vehicles. By measuring net system output, you can see exactly how much mechanical work will actually go to accelerating the car, and then you don't have to worry about drivetrain efficiencies. Besides, if you take meausrements on a vehicle dyno, and compare those numbers to the manufacturer's engine output graph, you can see exactly what losses you are experiencing in the drivetrain.

...the best we can do is what we've done. To just get some idea of the efficiency. The true range of efficiency is probably on both sides, starting below and ending above the 3500rpm-full-throttle efficiency that we've measured.

You haven't done anything!

Your full-throttle calculation is wrong because you do not have an accelerometer in your car, and relying on the manufacturer's power output chart is incorrect because it does not take into account efficiency losses through all of your drivetrain components.

Your efficiency calculations are still incorrect for the reasons originally stated to you. You have to analytically or experimentally describe the loads your vehicle is fighting at 100 km/hr before you can even think about trying to calculate it's efficiency; and the only efficiency you will be able to calculate is the system efficiency of the entire vehicle from gas into power to the ground.

 

[Ulysees]

There's nothing cryptic about calibration - it's a commonly used engineering phrase for mapping one set of data (in this case a graph) onto another.

Sure, tvp45. Except without giving the underlying principle, one is left guessing. That's not the approach here, we seek understanding. I still don't know for sure if I guessed right (transmission and tyre losses?)

------

Mech_Engineer, can I suggest that you let this thread go for a while. Because emotion is beginning to cloud your judgement and you're forgetting what the title of the thread is. It's about the efficiency of a (figuratively) isolated engine. Isolated engines have an efficiency defined too.

Maybe in the industry you have a convention that efficiency is only defined specifically to a car (taking into account transmission etc, ie it's wheel power / fuel power). No problem. Just allow us common mortals to be interested in the efficiency of a figuratively isolated engine (ie axle power / fuel power), and make threads about it.

relying on the manufacturer's power output chart is incorrect because it does not take into account efficiency losses through all of your drivetrain components.

In other words, you want to take into account the drivetrain components because you're after wheel power / fuel power. But we're not after this efficiency. We want the efficiency of the figuratively isolated engine: axle power / fuel power.

We focus on the engine as a block in a system diagram, a block that the manufacturer luckily gives a lot of details of in isolation, both in the book, and live on the panel.

 

[TVP45]

Ulysees,
I must apologize; I thought you were just looking for the correct answer. If you seek understanding, I would recommend the excellent paper by H. R. Ricardo in the Proceedings of the Royal Aeronautical Society, 1920. It really is the seminal paper on this topic and is quoted by researchers such as Tizard.
Good luck on your efforts.

 

[Mech_Engineer]

Just allow us common mortals to be interested in the efficiency of a figuratively isolated engine (ie axle power / fuel power), and make threads about it.

I would like to know what you are defining as the "axle"? Are we talking the axle the wheels are attached to, or the crankshaft in the engine?

... We want the efficiency of the figuratively isolated engine: axle power / fuel power.

We focus on the engine as a block in a system diagram, a block that the manufacturer luckily gives a lot of details of in isolation, both in the book, and live on the panel.

So you want to use the fuel consuption measured on your dash, and assume the power output is the same as the dyno chart you have. This will give you a number, but it's impossible to tell if that number means anything, since the ambient conditions you are driving your engine in could be vastly different from the test case shown in the graph.

It is possible to approximate how much power your engine is actually putting out, but if you just want to use the power graph that you have (which may or may not be the power output of your specific engine) you're done.

 

[Ulysees]

I would like to know what you are defining as the "axle"? Are we talking the axle the wheels are attached to, or the crankshaft in the engine?

In the context of an isolated engine, pretend you don't know what axle refers to as much as you like, if it makes you feel better.

So you want to use the fuel consuption measured on your dash, and assume the power output is the same as the dyno chart you have.

Exactly, the chart shows what was physically measured with a generator attached to the axle of the engine, an engine dyno.

This will give you a number, but it's impossible to tell if that number means anything, since the ambient conditions you are driving your engine in could be vastly different from the test case shown in the graph.

Angular acceleration would be different alright. So would ambient temperature. So would the fuel. So would the actual engine due to manufacturing tolerances. Maybe the inlet air is wetter too if you want some ambient conditions. Will you be happier if we spend the rest of today coming up with tiny reasons why the chart is not exact, when we know from the outset that efficiency varies anyway, and that's why we're only looking for some idea what efficiency is. Different points in the area under the power graph have different efficiencies, it would make an interesting surface plot if we had an engine dyno.

 

[brewnog]

So you want to use the fuel consuption measured on your dash, and assume the power output is the same as the dyno chart you have. This will give you a number, but it's impossible to tell if that number means anything, since the ambient conditions you are driving your engine in could be vastly different from the test case shown in the graph.

Well said.

Ulysees, if you're actually keen to learn how to solve your problem, take the time to read and understand such comments, and possibly even a book about fundamental engine operating parameters (the Ricardo paper already suggested is an excellent start).

 

[Ulysees]

Then I recommend you read the previous post carefully too. Nothing personal btw.

 

[Ulysees]

See what he's doing, brewnog. He's trying to massively amplify tolerances in the chart to render all practical use of the chart incorrect, just because he made the mistake to falsely involve transmission when he thought the thread was about fuel-to-wheel-efficiency. What is one supposed to do with such ego-driven disinfo?

I mean, are we going to throw our common sense out of the window, as well as our physics and engineering background, just to please someone who just can't admit he was answering the wrong topic (fuel-to-wheel instead of fuel-to-engine-output efficiency)?

 

[Mech_Engineer]

He's trying to massively amplify tolerances in the chart to render all practical use of the chart incorrect, just because he made the mistake to falsely involve transmission when he thought the thread was about fuel-to-wheel-efficiency. What is one supposed to do with such ego-driven disinfo?

I'm sorry you feel that way, but I was under the impression you wanted to calculate a number that actually meant something. As TVP stated: I must apologize; I thought you were just looking for the correct answer.

 

[Ulysees]

I must apologize; I thought you were just looking for the correct answer.

And the correct answer is... 26% now, 24% now, 31% now, 27% now, ...

 

[Ulysees]

But the one about the wind tunnel was a cracker.

 

[hxtasy]

My friend Mech_Engineer, substantiate your statements please. This is not like bible reading where the pastor speaks and the others accept, we can't accept figures without some substantiation.

Just like topics on engineering, sciences, etc. - you should always find out for yourself as much as possible, the truth. The same with religion, trusting a pastor's word is like trusting your mechanic or doctor, just because someone is labeled a professional doesn't mean they are right. Find out for yourself.


Sorry to be off topic. This is an interesting thread though.

 

[OmCheeto]

Just like topics on engineering, sciences, etc. - you should always find out for yourself as much as possible, the truth. The same with religion, trusting a pastor's word is like trusting your mechanic or doctor, just because someone is labeled a professional doesn't mean they are right. Find out for yourself.


Sorry to be off topic. This is an interesting thread though.

I want to see all the data. I haven't had an mpg gauge in one of my cars since my computer burned up in my '84 Impulse about 12 years back

I want to see L/100 km at full throttle and steady state speed at 10 mph increments from 30 mph to top speed with the gear noted.

Given this data, I will be able to determine the range of efficiencies of the motor.

 

[russ_watters]

We're done here.