air filter vs fuel economy (tricky question?) 11

[ivymike]

Perhaps this is a tricky question...

Does having a dirty air filter hurt the fuel economy of a modern gasoline-fueled car? (Obviously the parts stores would have you believe that it does)

If so, how exactly?

If it's inlet restriction, then would you say the effect is the same as modifying the throttle actuator such that the throttle opens slightly less for any given pedal input than it would otherwise have opened? If this is equivalent, would you expect to get better or worse fuel economy as a result of such a modification?

 

[ivymike]

I don't think you've quite got it. If X1+S < X0+S', then he hasn't pushed the pedal far enough and manifold density is less (so the power output is less and fuel consumption is less). If X1+S > X0+S', then the pedal has been pushed too far, and power output is too high and fuel consumption is higher.

 

[JSteve2]

If I flow 10 lbm/min through either system, then I have the same fuel mass per unit time (based on air mass) in either engine and therefore the same power. What I cannot guarantee is that I have the same flow restriction through the inlet because the flow v. pressure curve of a valve (throttle) looks different than that of a filter. More work may be coming off the crankshaft in one case v. the other to achieve that flow rate.

 

[ivymike]

Can't guarantee it? Sure you can. At a given rpm, you can't get 10 lbm/min through both systems unless the total pressure drop (and thus density) are the same:

Atmospheric pressure P1 stays the same...

Pressure after the filter is lower by some amount, deltaP_F, depending on filter restriction.
Pressure after the throttle is lower by some amount, deltaP_T, depending on throttle opening

Pressure (and thus density) at the intake manifold inlet, P_M, is what determines mass flow into the engine for a given rpm if everything downstream is held constant.

Pressure at the intake manifold inlet (P_M again) is what will affect pumping work if everything else is held constant.

P_M = P1 - DeltaP_F - DeltaP_T
If you want to get the same power out of the engine at a given rpm, then P_M must be held constant. P1 is constant, so any case where (deltaP_F)+(deltaP_T) is changed will give a different P_M, a different mass flow rate, and a different power.

 

[JSteve2]

At the same flow rate, I agree that power is fixed. I don't agree that RPM is fixed or that "everything downstream is held constant." Those two conditions assume the answer.

If the piston pulls harder (i.e. takes more work off the crankshaft) in one case than the other, then your solution is not correct. Whether the piston pulls harder is in fact the question, so your solution just assumes the answer.

 

[patprimmer]

The question is clearly answered several times.

The only way a fuel economy decrease could result is if there is a throttle position sensor for a WOT fuel enrichment system and the driver was driving in the range where restricted filter was at WOT, but the clean filter was at not quite WOT to get the same power level. This is a very unlikely situation for any significant time as mostly if a driver is over half throttle, they are at WOT.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[carfreak01]

I'm not an expert but this is my opinion on the matter:

The mass of air that is flowing thru the air intake system is metered so that the computer can calculate the amount of fuel that has to be injected.
So if a dirty filter is used, then less mass of air will flow thru, so less mass of air will be metered and less fuel would be injected.
So fuel economy wont be affected by this, only performance.
The driver will notice that with the same pedal travel he or she used to have more power, so if X amount of pedal travel was needed to cruise at 60mph, now X+Y pedal tarvel will be required to cruise at the same speed, so the driver's economy will be affected but not the engine's.

 

[carfreak01]

I notice I wrote that the driver's economy will be afected, that's wrong, I'm sorry, only engine performance.

 

[ivymike]

jsteve, is there something other than pressure that tips off the piston to the fact that something is afoot? Does the transmission change so that rpm @ 65mph is different?

for everyone else, a grand conspiracy:

http://www.fueleconomy.gov/feg/maintain.shtml

"Check & Replace Air Filters Regularly
Replacing a clogged air filter can improve your car's gas mileage by as much as 10 percent. Your car's air filter keeps impurities from damaging the inside of your engine. Not only will replacing a dirty air filter save gas, it will protect your engine."

http://www.knfilters.com/ads/linegas.aspx

"According to the U.S. Government, replacing a clogged air filter can significantly improve gas mileage*."

http://www.aempower.com/ViewTopic.aspx?TopicID=9

"According to the U.S. Government, replacing your car's dirty air filter can net an improvement in fuel mileage of up to 10%"

http://www.carcare.org/Fuel_AirIntake/air_filter.shtml

"Driving with a dirty filter restricts the air entering the engine, and if severe, can impact fuel economy and performance."

http://www.dailyfueleconomytip.com/uncategorized/fuel-economy-tip-change-your-dirty-air-filter/

"This past week, I conducted my own, very unscientific study after I changed my air filter. My gas mileage for this past week was 33.77, up from 32.84 the previous week. "

But wait! Not everybody is onboard with this conspiracy!

http://www.consumerreports.org/cro/...t-mileage-for-your-fuel-dollars-406/index.htm

"Keep your air filter clean. According to our tests, driving with a dirty air filter in modern engines doesn't have a significant impact on fuel economy, as it did with older engines. While fuel economy didn't change, however, power output did. Both cars accelerated much more slowly with a dirty air cleaner. We drove both vehicles with their air cleaners restricted and found little difference in gas mileage with either engine. That's because modern engines use computers to precisely control the air/fuel ratio, depending on the amount of air coming in through the filter. Reducing airflow, therefore, caused the engines to automatically reduce the amount of fuel being used."

 

[IceStationZebra]

From a theoretical standpoint I'm in the "doesn't affect part throttle fuel economy, does affect max hp" camp.

From a real world standpoint it depends on the driving habits - if the driver demands near maximum power frequently fuel mileage would go down. (though it would be bad to start with) An excessively plugged filter would also cause fuel mileage to go down - though the filter restriction would have to be high enough to cause a fully open throttle at cruise.

I was thinking a little deeper, like would the piston need to do more work to draw the charge in - which is no. Take the driver out of the equation - Set the cruise control to 55mph and only vary the restriction in the air filter - within reason. The ECU will adjust the throttle to achieve the same intake manifold vacuum. Essentially you need X amount of fuel to create Y amount of HP to maintain 55mph. The ECU will let in Z amount of air to properly combust X amount of fuel by adjusting the throttle.

ISZ

 

[SomptingGuy]

Just to confuse the hand-wavers more, a clogged filter would affect economy for a diesel, because it would introduce additional pumping losses.

- Steve

 

[JSteve2]

I can pick a mass flow and filter dirtiness level, short of WOT, where the intake system has more pressure drop than a similar mass flow with just the throttle and a clean filter. Until someone tells me who pays for that extra pressure drop, I will continue to protest. Quietly after this, unless someone has something new to present.

Are air-fuel corrections by the computer enough so that moderately clogged filters in normal driving conditions possibly below the level of an unscientific consumer reports measurement? Probably.

However, it's nonsense that regardless of the means the air takes to get to the engine, it has no effect on fuel economy. Replace the filter with a tiny orifice - no effect. Put a banana (only 90% blockage, though) in the intake, no effect. Make the intake 12 miles long in a pile of spaghetti configuration - no effect. The entire intake-design industry is a scam, and you guys have exposed it.

Sorry, I'm not buying. Even under consumer reports evaluation, a dirty filter gives a severely degraded engine. Did they try to pull anything with that engine or even attempt to get similar acceleration out of it (but something less than WOT)? We don't know.

 

[ivymike]

In your opinion, how is mass flow of air determined for a particular engine operating condition? Please do just one example calculation! Don't even post it, just write it down and think about it on your own!

So far you don't seem to be grasping the fact that the user can (and does) select the throttle restriction to compensate for filter restriction up to the point where throttle restriction reaches a minimum. There is, therefore, no "extra" pressure drop whatsoever until you reach WOT. For any "before the dirt" throttle position & resultant pressure drop, there exists an "after the dirt" position to give the same pressure drop, up to the point where you reach "after the dirt" WOT. Any pair corresponding (equal pressure drop) conditions will give identical engine performance.

Make the intake 12 miles long, as you suggest, and (assuming the total restriction is similar to idle position throttle restriction) you'll be stuck driving around with the throttle at WOT to achieve what used to be idle performance. You'll get the same fuel economy that you would have gotten if you froze the banana and wedged it under the pedal to prevent pushing it down. Whether the filter or the throttle restricts the flow makes no difference to the engine.

 

[btrueblood]

"unscientific consumer reports measurement"

They are usually much more scientific than the average car-nut magazine, in that they take multiple samples from multiple cars purchased "off the lot" at dealerships...vs. single tests of mfgr-supplied cars, or worse, simple regurgitation of mfgr-supplied data...

And they are the only reliable source of reliability data for cars, or other major purchase appliances, etc.

 

[JSteve2]

ivymike,
I'm just offering my honest opinion, not trying to get anyone spun up.

The boundary conditions we've placed are that the operator applies the throttle until the mass air flow is the same in the two systems. Just because the mass air flow is the same, how can this enforce identical intake manifold pressure where I have a piston downstream capable of performing variable work on the system? Let's assume one system had a lower intake manifold pressure but the same mass flow rate - the piston pulls harder (not a violation of physics) or the volumetric efficiency is decreased (also not a violation of physics). Either of those conditions reduce the fuel economy of the engine, and I do not see how they can be excluded.

In your proposed system, the two configurations have the same pressure drop to each other at 1 lbm/min, at 5 lbm/min, all the way up to the point where the valve saturates. Given that valves/throttles and filters have different flow versus pressure drop characteristics, I find it extremely unlikely that comparing two systems where the bulk of the pressure drop is placed in the valve or the filter that they could have the same pressure drop for all flow rates until the valve saturates. I also do not see how that result is required by the physics.

 

[ivymike]

Let's assume one system had a lower intake manifold pressure but the same mass flow rate

That won't happen. You're implying that you can get the same mass of air into each cylinder regardless of density. If you could do that, then there'd be no market for devices to increase density.

 

[Lou Scannon]

JSteve, please do some scientific analysis, or better still, instrument an engine and run some tests, then report back.

 

[JSteve2]

"That won't happen. You're implying that you can get the same mass of air into each cylinder regardless of density. If you could do that, then there'd be no market for devices to increase density."
You are assuming the engine speed must be the same, and that the volumetric efficiency is fixed.

"JSteve, please do some scientific analysis, or better still, instrument an engine and run some tests, then report back."
Would that I had the time, because I think this is a really cool problem. Where does the burden of proof lie, and what has been incorrect in the current discussion?

 

[Surestick]

JSteve2 maybe you are looking at this the wrong way.

How will the engine management system operating in closed-loop tell the difference between a partially blocked filter and a throttle valve?

Assuming changes like a lower manifold pressure and the same mass flow rate assumes other changes (different engine or gearing) that invalidate the initial question.

I think the initial post was based on certain assumptions; filter isn't blocked to the point most people would have gone to see a mechanic, driver isn't a 16 year old with F1 ambitions, etc.

 

[GregLocock]

Woow this poor old dead hoss is getting a flogging.

OK, here's one that got me beaten up on another forum.

How does variable valve lift improve economy? It changes the effective timing so the scavenging will change, but that is a second order effect I suggest.

Again, how can the engine 'see' any difference in the restrictions in the intake duct? Why is an obstruction caused by a valve closing 'better' than a butterfly valve (or a dirty air filter)?

Or are the second order effects the reason?


Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[ivymike]

'spose it depends on the events in question. If you're closing an intake valve early, then lemme see...

(for ease of handwaving, I'll go with some improbable intake valve events)

let's say our hyp. engine has bore and stroke of 100mm, TDC volume of 9.5E-5m3, BDC volume of 8.8E-4m3.

let's say that 50% vol eff. is desired, and ignore pumping losses past partly-open valves (so the valves work as though the lift events were square waves, and flow extremely well when open)

(let me know if I get too far from reality for useful dicussion)

case1: restriction achieved via throttle
crazy intake valve event 1 (opening at TDC)
pressure in intake manifold 50,000 Pa
pressure in crankcase 100,000 Pa
gas up-force on piston during downstroke: 393N
gas work done on downstroke: -39.3J (pumping work)
crazy intake valve event 2 (closing at BDC)
pressure at start of upstroke: 50kPa
pressure at end of upstroke: 1129kPa (math error? no time)
compression work done (polytr): -142.8J

case2: restriction achieved via inlet valve closing
crazy intake valve event 1 (opening at TDC)
pressure in intake manifold 100,000 Pa
pressure in crankcase 100,000 Pa
gas up-force on piston during first half of downstroke: 0N
gas work on first half of downstroke = 0N
crazy intake valve event 2 (IVC @ half stroke)
polytropic expansion second half stroke, pressure at BDC = 49.8kPa
work done on downstroke = -73.1J
poly compression, pressure at TDC = 1125kPa
work done on upstroke = -142J

the crazy valve events version seems to come out behind in this case, since it seems to have greater pumping losses on the intake stroke for a given amount of air intake... so not better the way I set it up, but clearly not the same.