I was going to do a search, but I couldn't figure out how to word it.


Years ago I worked in a lab doing FAA certification testing on fuel transfer pumps for Jet engines. (Among other things.)
In case you've never seen one, it looks sort of like a portable airhorn, only instead of an air can on the bottom you have a hose leading to a fuel tank. On the backside, opposite the pump outlet, you have a line that feeds in jet fuel from a relatively low pressure pump and shoots it straight out the "horn". This creates a low pressure area in the "body" of the pump do (IIRC) to the venturi effect, and that sucks in a large volume of fuel to send it to the other tank.


So here's what I'm wondering; and I would like to address this question in not so much an "is it worth it?" mode as a "how do you make it work if you want to try it?" mode.

Lets say that you have a forced induction method, either turbo or mechanical, and it's feeding too much pressure to the engine; so it's venting the excess pressure through your blow-off valve.

So you capture the excess pressure, pipe it to your exhaust system and inject it to create a low pressure area behind the injection point.

a) Where in the system should this be injected? Personally, I'm thinking that you would inject it as far back as you can manage, and increase the exhaust tubing size after the injection point.

b) Would you need to create a "horn" to get the venuri effect, or is a diameter increase sufficient?

c) Does there need to be a 90 degree bend in the piping system so that the injection can be directly parallel to the exist stream?

d) What does the ratio of injection pipe to exhaust pipe size need to be? I suspect that it's not as important as it is in a jet fuel pump, but that there are still some optimum ballparks to be in.

e) Would there be a difference between turbos and superchargers in a system like this, given that they have different characteristics, or does it not matter, since it would be downstream of the turbo anyway?

f) What are the potential downsides, OTHER than wasting your time/money?


Please add any other thoughts that come to mind; other than the "What are you smoking?" thoughts.

 

Turbocharger systems already bypass exhaust past the turbo(s) with a wastegate to prevent overboost conditions and their charge relief or "blow off" valves open and vent to atmosphere (or back into the non-pressurized section of the intake on MAF-equipped cars) only on engine vacuum (throttle closing) to prevent compressor surge.

There is nothing you can do to improve that system by connecting the output of the blow-off valve to the exhaust. All you'd be doing is creating a heat pipe to the intake from the exhaust, and likely pulling exhaust gasses back into the engine when the valve opened, especially since it opens based on engine vacuum.

Supercharger systems may use a "pop-off" or "overboost" valve to vent excess pressure, since you can't easily bypass the supercharger or slow the belt speed in an overboost condition. However, you don't want the pop-off valve to be activated often, if ever. Pulley sizes should be selected to keep the supercharger in a safe operating zone and prevent an overboost condition. It's a fail-safe device, not a constant source of pressurized air.

A pop-off valve differs from a charge relief valve in that pressure of a specific level (typically controlled by spring pressure) opens the valve, not vacuum. In a situation where the pop-off valve would open, you'll typically get valve flutter as the valve opens, vents pressure, closes, and the pressure builds and exceeds the spring pressure of the valve again. Once again, this design doesn't lend itself to "harnessing" the output and doing something useful with it.

In other words, there's a reason why no one has tried this.

 

Its all been done. Dont try it because it was worthwhile, it would be in production. ****ty outlook for an R&D point of view, but when it comes to the internal combustion engine- I cant disagree.

Im not sure I understand what you are getting at here, but its very very very hard to inject more fluid into a pipe and get the original fluild to flow faster.

For example, you have a mass flow rate of 2lbs/hr. Inject 1 lb/hr of very high pressure fluid and now your at 2.75 lb/hr mass flow rate. It is going to be damn hard (impossible actually, if my thinking is rational) to get it above 3 lb/hr.

 

Thats exactly how a fuel transfer pump in a jet airplane works.

 

I know what you are talking about, I a helicopter pilot and we have jey pumps to evacuate our refuel transfer tube. I have seen headerst that use a collector that looks like it is trying to use this effect, Afterburners or someting like that, where 3 pipes come in together and 1 comes in later in the collector, so someone is trying. I think to do this you would need a good volume of air and a large cone shaped collector, maybe a leaf blower or something that is not engine driven would be the best.

 

What I was basically envisioning was using a centrifugal blower that's larger than it normally would be for a particular application, then using a blow-off valve to continuously vent the excess pressure.
(Blow-off valve may be the wrong term for the type of valve I'm thinking of, but I think you get the idea.)

So you have this car going down the track at WOT, the blower begins to exceed the pressure level that the valve is set for and excess pressure begins venting. Obviously, this blower would be blowing INTO a MAF sensor, so that regardless of how much pressure is vented; the computer still dispenses the proper amount of fuel.

Now, I could vent this excess pressure to the atmosphere, but then I thought... Why not inject it into the exhaust stream?

I'm thinking that a 90 radius in front of the rear tires would be a perfect injection point. The spent gasses coming down the pipe behave like a fluid, so in THEORY this setup should behave like a fuel transfer pump.

Here's the problem though...

I know that those pumps I used to test were designed very carefully and that the various proportions and shapes MATTER.
I just don't know what the general proportions should be for this exhaust setup should be.

 

I read this carefully and that can not be how a transfer pump would work. The system you describe would dump fuel out into the air at the back of the horn. That would be very wasteful and would not work.

The venturi effect is real, the center of the horn would be a low pressure area. The forward facing small tube would be a high pressure area at full stall pressure of the air flow. None of these pipes would be connected directly to the fuel supply.

So how could this system work? It works by vaccum, air pressure diffferences. You connect the small tube to a vent on the full tank, this supplys a higher air pressure on the top of the full tank. The large tube is connected to a vent on the empty tank, this supplys a low air pressure.

Another line, below the fluid level in the full tank, and connected to the empty tank will allow fuel to flow from the high pressure full tank to the low pressure empty tank at a high rate. If you see any fuel at the pump horn the tank you are filling is now over full.

Sorry to be a killjoy but you have the operating principle all wrong. No jet fuel is injected into the air stream, counter productive.

That is why it won't work on a car either. Sometimes the vent from the valve covers are connected to the exhaust to expolit the low pressure from the velocity and the pulsing of the exhaust.

You are thinking about these things and that is good, but you are on the wrong track.

Z28

 

Well then I must be doing a poor job of explaining, because that is exactly how they work.

Please understand; I used to TEST these pumps. One of the things I had to do was to measure fuel flow at different pump pressures. In other words, this isn't something I read in a book, this is something I have held in my hand and watched function.

Where my explanation fails, I do not know; but I am describing reality, not theory.

 

If the pump was powered by fuel flow not air flow and was completely enclosed in the large fuel line it could work that way.

The small high pressure stream would give the fuel velocity through the narrow part of the horn. This will cause a low pressure in the venturi of the horn. The narrow part, due to high velocity, fluids moving fast are at a lower pressure. This could be used to drag more fuel along at a low pressure and higher volume.

I thought you where talking about an air powered system. This is a fuel powered system.

This will not work very well with a compressable fluid like air. It takes a 10hp compressor to run a 1hp air motor. Since fluids are not very compressable pumping them is much more efficent.

The shape of the horn would be desgined to keep a constant mass flow rate. As the fuel slows down it would need more area to maintain the same mass flow rate. That is why the shape is horn like. Shapes would be very different with air due to lower mass ect. I would not know how to easily figure them, but they would likely work best over a narrow range.

In theory you might get it to work but using crank horsepower to drive a roughly 10% efficent system would be a bad trade, I think.

Z28

 

My explanations can't possibly be this bad.

I think you're simply jumping to conclusions about what I'm saying, before you actually finish reading what I said.

 

The explanation is not bad, just incomplete.

Any air you deliver, wasted, pumped into the engine, or pumped into the exhaust. Was made with crank horse power.

If the air is pumped into the engine you have crank horse power losses doing so, but your pay back is huge due to the amount of power produced. Air injected into the exhaust will not have this kind of gain.

I was trying to point out how bad air is at transferring power i.e. 10hp compressor to run 1hp air motor. With a piston conpressor 90% of the supplied power will be lost in flow and heat losses.

A blower will do better than this, but the effiency is very poor.

The fuel pumps have a high vicosity fluid and little compression losses due to heat. Their performance will not translate to good results with a low viscosity fluid that has compression losses, air.

The most likely result will be to, as other have said earlier, to increase restriction in the exhaust by increasing the total flow through it.

That is all I know on the topic, good luck with your project.

Z28

 

So what?

The air I'm talking about is excess pressure from the intake tract that was just going to be vented to the atmosphere anyway.

Injecting it into the exhaust in the way I am thinking of may not do anything useful, but it certainly isn't going to WASTE anything.

 

Have you considered the cooling effect on the exhaust and what that might do to exahust velocity?

-Mindgame

 

I'm completely lost with the tech intent of this question. How could adding mass and volume to the exhaust pipe do anything other than increase the pressure drop in the exhaust pipe, hurting the overall exhaust flow.

And, if I read one of the tech explanations correctly, you actually intend to oversize the compressor to make more "excess" air available. The larger you make the compressor, the more HP it takes to turn it and the more HP you waste by venting the excess flow.... doesn't matter whether its being driven by a belt off the crank, or by an exhaust turbine.... you are adding a load to the engine, to vent the artificially increased "excess" into an exhaust system.

Appears you may be talking about an "eductor", with regard to your fuel flow experience .... I really can't follow the concept any better than that?????

 

Mind -

No, I hadn't. You're right, the release of pressure as it leaves the blow-off valve (or whatever the proper valve is for continuous venting), will cause the air to cool rapidly


Injun -

I shouldn't have said that. It's not really my intention to use an oversize compressor for the purpose of creating injectable pressure.

It IS my intention to use an oversize compressor, and I do have specific reasons for doing so; but until the injection idea occurred to me I was planning to simply vent it to atmosphere.

As to pressure drop, I believe that's what I'm after. A pressure drop BEHIND the injection point. I am assuming that this will tend to "pull" the upstream exhaust out of the system, ultimately aiding scavenging.


I do know from my experience with pumps that the flow of fluid through a venturi can be used to create a pressure drop that sucks "extra" fluid along with it. That's why I think this will work.

Is there a fundamental flaw in my idea or is it just that you don't think it will work out the way I hope?