Vacuum In Exhaust-What would it take?

[Gollum]

Just thought of this, while sitting on the crapper no less...

 

Ok, so pretty much almost all internal combusion engines are air pumps. We use turbos and superchargers to FORCE air into them so they can pump more air, effectively making more power. Pumpings lots of air is a good thing.

 

But in nearly all situations there's POSITIVE pressure in the exhaust system, meaning you're not getting ALL the exhaust out of the chamber before the valves close.

 

Would it be possible to rig some sort of electric or gasoline powered fan/turbine to the exhaust to force air out, possible creating a mild vacuum?

 

Has it been tried? How long ago? How did they do it? What was thier success.

 

I can't imagine it being a HUGE improvement on power, as we can have open headers with not a huge power increase, but I'd imagine it could get you extra power in the lower RPM range, similar but better to the effects of exhaust scavenging.

 

But the hard part in my head seems to be "how do you design a turbine that will pull enough air to not be a restriction at full throttle"? And how on earth to power it? Using gasoline might work, but how do you rig it? You could use an electric motor, but where's the power going to come from. For some reason I have doubts even an oversized alternator would generate enough power.

 

What do you guys think. A pipe dream, or a possible idea?

[OlderThanMe]

It's like an inverse turbo.

You would have to pull on the exhaust with greater volume than the cyinder has. That means with a 2.8 liter a pump would would need to pump 2.8 liters(hot and probably still under soome pressure...) every two rotations of the engine to break even. At 2500RPMs you would have to "pull" more than 3500 liters of air every minute to come out at neutral if the air exits at ambient pressure and temperature. That is a lot of air.

[Gollum]

Yes it IS a lot of air... But what would it take?

 

If a turbo can double an engine's HP, surely a simlar device to pull air from the engine's exhaust instead of ambient air shouldn't take more power than the engine's creating, right?

 

 

EDIT: I was just punching some numbers on an online convertor, and it said 2500 liters would be 88 cubic feet... that can't be right. If that's the case it shuold only take about 300CFM for a well maxed out 2.8 liter engine. I'm expecting to need a vacuum/fan that can move more like 3000+CFM

[RTz]

But in nearly all situations there's POSITIVE pressure in the exhaust system, meaning you're not getting ALL the exhaust out of the chamber before the valves close.

 

 

If I understand you correctly, you're talking about 'blowing down the clearance volume'. Part of the purpose of valve overlap is to do just that. And it works very well on its own. However, depending on the engine, it doesn't become efficient until some RPM is generated. At low RPM, as you've implied, there is still a fair bit of exhaust in the chamber, diluting the 'useful' air.

 

Keep in mind that its far easier to push air than it is to pull air. That is the reason we get by on smaller exhaust valves even though the exhaust 'volume' is greater (thermal expansion).

[Gollum]

Keep in mind that its far easier to push air than it is to pull air. That is the reason we get by on smaller exhaust valves even though the exhaust 'volume' is greater (thermal expansion).

 

True, but aren't superchargers and turbos basically air pullers?

 

Anyone actually know how much CFM a given engine is putting out at WOT? looknig into it I don't think it would be too hard to make an array of fans pumping upwards of 10k CFM. The hardest part I think would be finding fans rated at the temperatures they'd be subjected to.

 

The geek in me wants to buy this:

 

http://www.toolexperts.com/airflow-measurement/anemometers-and-cfm-measurement/mannix-deluxe-air-flow-meter-wdcfm8951/wdcfm.html

[RTz]

True, but aren't superchargers and turbos basically air pullers?

 

They are not typically installed with significant restriction on the 'ingest' side.

 

Anyone actually know how much CFM a given engine is putting out at WOT?

 

The geek in me wants to buy this:

 

Since you've made the self proclamation of geekage, I bet you can figure it out on your own... convert 2.8 liters to cubic inches, know that a 4-stroke engine displaces half its volume every revolution, factor in time, and bingo... static CFM.

[G.I.jonas]

You should be concerned with the fact that this mechanism has to be so lightweight that the minimal gains are still enough to overcome the weight gains,not sure how much power you are planning to save but this sounds like it could get heavy.

[Gollum]

They are not typically installed with significant restriction on the 'ingest' side.

 

 

 

Since you've made the self proclamation of geekage, I bet you can figure it out on your own... convert 2.8 liters to cubic inches, know that a 4-stroke engine displaces half its volume every revolution, factor in time, and bingo... static CFM.

 

Well yea, that's easy... But me and you both know that static CFM will be vastly different from what the engine actually pumps at various throttle and RPM variations. That's what I'm not so keep on calculating, but I'm sure I can figure it out. But if someone knew off hand...

 

You should be concerned with the fact that this mechanism has to be so lightweight that the minimal gains are still enough to overcome the weight gains,not sure how much power you are planning to save but this sounds like it could get heavy.

 

Yea I've thought about that. I guess it really depends on how much gain could be seen. I'm trying to figure out how to even calculate how much gain cuold be seen at various vacuums, but have no clue how to even go about that without using software to create a simulation.

 

I don't have a whole lot of hope for this idea, but it's still fun to think about.

 

EDIT: At 8000RPM a 2.8 liter engine should only be displacing 395.5242672 cubic feet... That's not that much. Granted we have to figure in expansion of gases, I don't see it being that hard... A shop fan as used for dynos is around 7000CFM, and is about 1/4-1/2 HP, so I can't imagine it taking that much power to create an array of aluminium fans to push around 2000CFM, which should be plenty of flow to keep the exhaust at a slight vacuum, or close to ambient.

 

Time to look up gas expansions... (off to google)

[Scottie-GNZ]

But in nearly all situations there's POSITIVE pressure in the exhaust system, meaning you're not getting ALL the exhaust out of the chamber before the valves close.

 

Would it be possible to rig some sort of electric or gasoline powered fan/turbine to the exhaust to force air out, possible creating a mild vacuum?

To some degree, a well designed header system is supposed to contribute to this.

[OlderThanMe]

I assume that you are trying to create a vacuum to draw in air through the intake...and maybe increase your VE to above 100%.

 

NASCAR engines do a VERY good job at cramming as much air as possible into the chamber and even use the helmholtz effect to naturally supercharge.

 

I can not see any easy or cost effective way to increase exhaust scavanging other than any static exhust/header design.

I would rather work on where the engine really hurts(in the intake, pulling air) and get the most for the money/effort and try to get more power.

 

Sometimes calculating until your brain pukes in your skull is a good idea.

I've been doing plenty of that lately.

[Gollum]

Ok, did some quick number crunching. In order to DOUBLE the air volume via heat you'd need a heat jump from about 0 degrees F to about 500 degrees F, which I doubt many engines at street HP levels could do. At the header maybe, at the exhaust tip... not so sure.

 

So if you take say, a 5.0 liter engine, convert it to cubic feet, and at 7000 rpm it's producing 618 cubic feet of air. Now it'd be an overly cautious generalization to assume double that for heat expansion, so now that's 1232 cubic feet at redline. If there's a power adder or ram air that changes the numbers dramtically though.

 

I don't think creating 2000CFM worth of flow would be that hard though, and would be more than enough for most engines.

 

I'd just like to clarify at this point that I understand a well designed header won't restrict power too much and will give you good scavenging effects in the right RPM range, but what if for 10-25 pounds added to the car you could have ideal scavenging at ALL rpm? wouldn't that be worth it? It might even help drivability on engines with wild cams, but that's just a guess. I don't know NEARLY enough about cam dynamics to assume anything like that.

[bschiltz]

nevermind

[Dan Juday]

...

 

Keep in mind that its far easier to push air than it is to pull air...

 

This is the real key.

 

I'm not a professional mechanic and I don't design engines. But I am a HVAC contractor and I do design duct systems for buildings. I have some pretty sophisticated load calc software that I use to design these systems. I can tell you that the airflow inside a return duct is very different than the airflow in a supply duct. To get equal airflow the size, shape, and types of fitting can be very different. And the results are not that intuative. You can take the same fitting, lets say a square-to-round transition for example, send the air through in one direction and have a very low pressure drop. Send the same amount of air at the same speed and temperature the other way and the pressure drop and be three times that of the opposite direction.

 

Think about it this way: A turbo is just a fan in the intake and a fan in the exhaust connected by the same shaft. Why doesn't the force of the air entering the motor spin the shaft with enough force to pull the exhaust gases out? Why, in fact, is it actually the other way around?

[proxlamus©]

you can do this with aerodynamics =)

 

imagine a radical design to build a venturi and create a vacuum to suck the exhaust out =)

[(goldfish)]

Centrifugal supercharger in the exhaust. Problem solved.

[Flexicoker]

you could have the exhaust exit at a low pressure point on the car, like F1 did back in the day by having the exhaust exit in the diffuser tunnels under the car. The problem with that is that your downforce is engine dependent, and your super exhaust scavenging is now speed dependent.

 

A good solution is using a stepped header, but stepped in precisely the right spot(s). You can use the harmonics created by the closing exhaust valve to seriously change and augment your torque curve.

 

For example... 2 Formula SAE cars, both with bone-stock, unopened Honda CBR600's. Both cars are relatively heavy, yet both scored well in the acceleration event at competition. The F07 car, however, seems much faster, is slightly faster, and is way easier to drive than the F05 car solely due to exhaust resonence tuning we did with Ricardo Wave software.

[80LS1T]

Ok, the only question I have is how big is the fan going to have to be to suck 2000CFM I mean a shop fan that blows 7000 CFM is like 4' across isn't it? Just put a turbo on it(intake side!) and call it a day! LOL

 

Guy

[Gollum]

Centrifugal supercharger in the exhaust. Problem solved.

 

I'd thought about that, and it would be one take one wicked header to do that. Could definitely be done.

 

Ok, the only question I have is how big is the fan going to have to be to suck 2000CFM I mean a shop fan that blows 7000 CFM is like 4' across isn't it? Just put a turbo on it(intake side!) and call it a day! LOL

 

Guy

 

Well I was actually looking at radiator fans yesterday. Many of the 14 and 16inch variety flow well over 2kCFM, and some over 3k. Those aren't very big and they're light to. The downside is that almost all of them are made out of plastic and I have a feeling they'd just melt in this application. Though maybe I'm wrong. Maybe exhaust temps would go WAY down and I wouldn't have to worry about the exhaust being any hotter than say the coolant. In that case a radiator fan WOULD work.

 

Yesterday while I was browsing through fan catalogs it seems as though it would be best to go with a fan over 12 inches in diameter. Reason being is that you gain a LOT of airflow per inch the larger you go, so an array of 4 to 8 inch fans just wouldn't really get the required flow that easily. You'd need like 10+ 4inch fans on the back of the car, or just use a single 14 inch.

 

Also, by the looks of it "blower" fans, like the ones used for AC and heat systems in cars seem to generate much more flow for their size. It be harder to adapt into an exhaust system, and harder to vent out, but the size/CFM ratio might be worth it.

 

I'm seriously thinking I want to build a test rig I can adapt onto cars being dyno'ed to at least see what would happen. I could probably use larger than needed fan to exaggerate the test to see what's possible, then try to find out what's realistic.

[RTz]

I'd thought about that, and it would be one take one wicked header to do that. Could definitely be done.

 

The supercharger won't work, at least not like it does in the intake. It seems your forgetting that, while the intake valve is open and cabable of ingesting air, the exhuast valve is closed (for most of the duration anyway). You can't draw intake through a closed exhaust valve. There's 489 other things wrong with the theory but I don't have time to write them.

 

 

Maybe exhaust temps would go WAY down and I wouldn't have to worry about the exhaust being any hotter than say the coolant. In that case a radiator fan WOULD work.

 

I think not.

 

Also, those CFM ratings are in 'free air'. Introduce a restriction and the CFM will drop significanty. Axial fans suck at making static, pun intended. Juday will confirm that.

 

My honest opinion is that you're barking up the wrong tree. Its not my place to tell you that you can't do something, but I feel your efforts could be better spent elswhere.

 

I hate watching someone shoot themself in the foot... No amount of wishful thinking will change the outcome... its gonna hurt.

[BobbyZ]

I am not really sure what gains you are expecting to see here. Forced induction increases performance by forcing more air/fuel into the cylinders for a more energetic explosion.

 

By creating a vacuum in the exhaust the only increase in performance is going to come from the decrease in energy required to expel the exhuast gases from the cylinder. I think that the gain in performance over a well designed header would be minimal if any. Am I missing something here?