What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Why not a mechanical pump for automotive FI ?

Larry DeCamp

Well Known Member
The most concerning aspect of automotive FI referenced is the need for two electric pumps . The electrical hardware and design to achieve this is considerable and discussed at length. I recall a comment that an engine driven pump is not suitable for the high rail pressure required.

What is the design constraint causing this limitation ??

It would sure be nice to enjoy the benefits with the same reliability and simplicity of legacy FI systems.
 
A gear or gerotor pump driven off an accessory pad could work, easy to develop the 50ish psi needed for EFI.

The twin electric pumps have proven to be reliable. Never had a report of both failing on the same flight. They should be powered off separate breakers with separate switches for best reliability. You should always have a backup power source with EFI which can be isolated from the primary power.
 
The most concerning aspect of automotive FI referenced is the need for two electric pumps . The electrical hardware and design to achieve this is considerable and discussed at length. I recall a comment that an engine driven pump is not suitable for the high rail pressure required.

What is the design constraint causing this limitation ??

It would sure be nice to enjoy the benefits with the same reliability and simplicity of legacy FI systems.

I hope you get a definitive answer for this, but I don't think this is precisely correct. Maybe it is a practical limitation or availability but not in theory i.e. maybe you can not get a mechanically drive bolt on pump with the right performance. A properly designed positive displacement pump (OTS) can pump the pressure, and volume. But it would need to be placed low on the engine or suction head requirements for hot fuel would likely allow cavitation.We don't have a good location in the gear train for such a bolt-on pump. I am not sure of the limitations of a diaphragm pump.

Or you could just wish for an electric source and pump with the reliability of a mechanical unit. Likely more readily achieved and lower development and production cost. Recall that our automotive EFI systems are very complicated and work so much better in reliability and durability than the previous mechanical pump and carb. Just a matter of proper attention to the goals. Oh- and $^3:D

Ross is more succinct. Up early Ross!!
 
Another part to this is the recycle - to do it right, and be able to run higher vapor pressure fuels without worrying about vaporlock, the fuel pump(s) need to move quite a bit more fuel than the engine will consume and return the unused part back to the tank via an Andair-style dual fuel selector. That's a flowrate that most mechanical pumps cannot match, at the normal size used for the given engine.

For vapor lock consideration, separating the fuel pump from the Lycoming-brand heater is key.
 
I hope you get a definitive answer for this, but I don't think this is precisely correct. Maybe it is a practical limitation or availability but not in theory i.e. maybe you can not get a mechanically drive bolt on pump with the right performance. A properly designed positive displacement pump (OTS) can pump the pressure, and volume. But it would need to be placed low on the engine or suction head requirements for hot fuel would likely allow cavitation.We don't have a good location in the gear train for such a bolt-on pump. I am not sure of the limitations of a diaphragm pump.

Or you could just wish for an electric source and pump with the reliability of a mechanical unit. Likely more readily achieved and lower development and production cost. Recall that our automotive EFI systems are very complicated and work so much better in reliability and durability than the previous mechanical pump and carb. Just a matter of proper attention to the goals. Oh- and $^3:D

Ross is more succinct. Up early Ross!!

From my research, the diaphragm pumps can flow a lot. However they sacrifice pressure as the volume goes up (inverse relationship). The FI servos can deal with a very wide range of pressure inputs. The EFI systems, OTOH, require a steady and precise pressure, as it along with injector open time determine fuel delivery rates to the cylinders. Most EFI's are not set up to vary the open time if the pressure changes.

Larry
 
Last edited:
Believe Airguy meant lower VP.

He is correct as the pressure losses are a function of the square the flow if frictional losses of 35 versus 10 gpm are 12x higher. Throw in the extra elevation head of the engine mount location (not much, i know) and things keep compounding. Everything is good, until it's not; fuel blend issue, dirty pre-filters or strainers, etc.
 
An EFI system (which presumably also has EI) requires reliable electrical power. Once you have that, driving two pumps (fused and switched independently) isn't a big challenge. Yes, the pumps have good-sized current draw that you need to account for with your electrical system design, but that's just a matter of scale, not techincal feasibility.

Replacing an electric pump with a mechanical pump doesn't change any of that. You still need power to drive the ECU, open the injectors, and give you spark.
 
Lots of good points brought up here.

At least one of the EFI Sport Class winners has used an engine driven pump since day one. Can and has been done.

Some high hp (1500+hp) EFI race cars use mechanically driven gear/ gerotor pumps. Has been done.

We recently worked with an Italian company to develop a mechanical pump setup for our Rotax 912 EFI kits. This was mainly done because the factory generator system is only rated at 18 amps and the EFI current draw takes a good portion of this if running ignition too, leaving not a lot for other aircraft systems.

As mentioned in the post above, with EFI, you need a reliable source(s) of electrons for the ECU, Injectors, coils etc. anyway, so these same systems can run the electric pumps as well.

0500 is my usual time up. Only way I can get stuff done...
 
Last edited:
Bennair; Dunnatt

I had a rotary mechanical pump reset for 50PSI and 50GPH at one of the shops that repairs such devices along with others (vac pumps, hyd pumps etc). The new pump fit exactly where the other pump was - but the new pump had no widget to set the fuel flow - not needed in fact with the SDS setup minding things for you - TCM engine in this case.

If you have a rotary drive type aft section for your engine instead of the diaphragm type you are golden. If not, you can switch covers and have the correct drive, and you are on your way.
 
Believe Airguy meant lower VP.

Nope, I meant what I said. Higher vapor pressure = lower boiling point = more exposure to vaporlock, which is what you get with automotive fuel and E-gas when you start exposing the fuel to heat. Removing the heat input (mechanical pump is a big one) helps, as does a high flowrate and full recycle back to the tank.

100LL has a lower vapor pressure, higher boiling point.
 
Sorry about that. Read it wrong. I know you knew the applicability but assumed you brain-farted it. Turns out it was me
 
OK with a mechanical pump, how do you start the plane?

To start, the injectors need very little fuel volume, the regulator orifice is closed or mostly so. Fuel pressure easily builds up to the required 50psi or so as you crank. It's been done as I mentioned, works fine.
 
“PilotJohnS
So how do you start….”

Aux or Boost pump. Similar to what you’d find on a mechanical FI system. Carb engines too.
 
Last edited:
To start, the injectors need very little fuel volume, the regulator orifice is closed or mostly so. Fuel pressure easily builds up to the required 50psi or so as you crank. It's been done as I mentioned, works fine.

Isn't the SDS system closed on shutdown and retains the fuel at pressure after shutdown? Then to prevent solid expansion lock, it has a pressure relief? There fore on starting, a pump, the pump, has little work to get the a start?

If not retained an automotive system can heat and vent fuel so starting has to recover the pressure/volume of fuel loss resulting in a rough start, BAD for emissions.
 
Switch covers to rotary pump ?

F1 Boss says switch covers to rotary drive. Airguy says mount it low to minimize suction lift vapor lock. Where does this cover come from and does it include the hardware to convert the levered diaphragm drive components to rotary motion ?
 
Isn't the SDS system closed on shutdown and retains the fuel at pressure after shutdown? Then to prevent solid expansion lock, it has a pressure relief? There fore on starting, a pump, the pump, has little work to get the a start?

If not retained an automotive system can heat and vent fuel so starting has to recover the pressure/volume of fuel loss resulting in a rough start, BAD for emissions.

Correct, correct and correct in the case of the electric pump setup. With a gear pump, maybe not without external check valves fitted.

The flying mechanical pumps are mostly on Contis to my knowledge however the Lear rotary pumps used on some Bendix FI systems should probably work too. I believe these were fitted to some turbocharged Aztecs and possibly other turbo aircraft. I am not very knowledgeable about this aspect.
 
F1 Boss says switch covers to rotary drive. Airguy says mount it low to minimize suction lift vapor lock. Where does this cover come from and does it include the hardware to convert the levered diaphragm drive components to rotary motion ?

The rotary fuel pump drive pads on the 540's I've familiar with are very low on the case. I would have considered going with a mechanical fuel pump as the primary on my EFI system for the Rocket, but there as no way I could see to get the pump to fit even with radical firewall/footbox mods.

That said, even a "low" engine mounted fuel pump has a bunch of rise to draw through, so a low wing airplane using automotive gasoline (like mine) needs to be carefully considered.
 
That said, even a "low" engine mounted fuel pump has a bunch of rise to draw through, so a low wing airplane using automotive gasoline (like mine) needs to be carefully considered.

And is still a large heater for the fuel system - though if you're running a looped fuel rail with tank return that becomes much less of a problem. A dead-headed fuel rail should be avoided to eliminate hot-start problems.
 
We don't see any hot start issues dead heading to the injectors as the high fuel pressure on EFI generally raises the boiling point of the fuel enough so that it doesn't.

Dave Anders did some hot start tests at 118F OAT and reported pretty much instant starts using a deliberate 10 minute heat soak period after shutdown.

We do recirculate fuel from the distribution block which is either rear baffle or top crankcase mounted. This purges any hot fuel up to that point instantly as soon as the pumps are on.

Like most things in aviation, we build on what's been proven and stick with it. The twin electric pumps have never caused a forced landing to my knowledge but be sure they are separately fused, switched and powered. Don't do what one company did on their EFI equipped aircraft- tying the fuel pumps to the gear pump circuit- Popped the breaker when cycling gear, 2 forced landings to date.
 
Back
Top