What's new
Van's Air Force

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

Fuel Pressure loss on Loop upline

SMO

Well Known Member
Airplane - F1 Rocket, IO-540 D4A5, recent install of Red Cube between fuel servo and spider. Engine driven fuel pump has ~350 hours.

Pulling 3 g's at the start of a loop I am experiencing a fuel pressure loss from the normal 20+ psi to ~8 psi (engine loses power) after which my boost pump automatically kicks in (power picks back up) and brings the pressure back to normal.

I can fly around all day, pulling 2 g's in 60* banks, and no issues, fuel pressure stays 20+ psi.

Anybody experience the same? Am I correct to assume a new engine driven fuel pump is in my near future?
 
Consider the changes you made, and start there.

First, it worked fine prior to relocating the red cube, yes?

I assume you're measuring pressure at the engine pump outlet. If so, nothing you plumbed between the throttle body and the divider would make a difference.

I further assume the red cube was previously firewall mounted between the bulkhead fitting and the engine pump. If so, you changed a hose and/or hard line as part of the relocation.

Thus three possibilities; the new lines pick up more heat than the previous install, or one of the fittings has a small air leak, or you installed a right angle fitting and (combined with heat) it is forming vapor bubbles in the flow.
 
Pulling 3 g's at the start of a loop I am experiencing a fuel pressure loss from the normal 20+ psi to ~8 psi (engine loses power) after which my boost pump automatically kicks in (power picks back up) and brings the pressure back to normal.

I can fly around all day, pulling 2 g's in 60* banks, and no issues, fuel pressure stays 20+ psi.

20 PSI "normal" is too low (18 minimum), typical pump pressure should be around 24-26, I'd start there.
 
Consider the changes you made, and start there.

First, it worked fine prior to relocating the red cube, yes?

I have had 10 flights since the change where I performed some aerobatics, only the last 3 - 4 have shown these symptoms. BTW, the other change I made at the same time was replacing my Andair gascolator with and Aerolab.

I assume you're measuring pressure at the engine pump outlet.

Correct, just to clarify this is not a measurement issue, there is a corresponding loss of engine power till the boost pump cuts in.

I further assume the red cube was previously firewall mounted between the bulkhead fitting and the engine pump.

Close, it was a Floscan that was replaced, and it was located in the tunnel between the stick well and the firewall. Replacement hose is a straight run.

Thus three possibilities; the new lines pick up more heat than the previous install, or one of the fittings has a small air leak, or you installed a right angle fitting and (combined with heat) it is forming vapor bubbles in the flow.

To add the Red Cube I put a bracket on the front of the fuel servo and mounted the cube there. Not a likely source for additional heat.

Will check the fittings for a leak, although why "G" forces would exacerbate a leak I am not sure, but possible.

No right angle fittings were introduced.

EDIT: This is not fuel flow related, I have confirmed via flight testing.
 
Last edited:
20 PSI "normal" is too low (18 minimum), typical pump pressure should be around 24-26

I should have been more clear, the normal fuel pressure fluctuates between ~22 and ~25. If I look at this on a graph it has a very constant oscillation - is that oscillation normal?
 
I should have been more clear, the normal fuel pressure fluctuates between ~22 and ~25. If I look at this on a graph it has a very constant oscillation - is that oscillation normal?

Approximately what is the frequency? I've noticed this at about 2400 RPM as monitored by the G3x Touch. I believe it may have to do with the sampling frequency of the engine monitor vs pressure pulses in the mechanical pump. Unverified, but the oscillation goes away when I dial the rpm down from 2400 to about 2350. The period of the osc. I have observed is about 5 seconds, and it will vary between maybe 21 and 25.
 
I have had 10 flights since the change where I performed some aerobatics, only the last 3 - 4 have shown these symptoms. BTW, the other change I made at the same time was replacing my Andair gascolator with and Aerolab.

So check it for an air leak. Seal the inlet and pull suction on the outlet, see if it will hold a vacuum.

Will check the fittings for a leak, although why "G" forces would exacerbate a leak I am not sure, but possible.

If fuel weighs 3x more, it requires 3x the pressure drop for the same lift. More suck means a small air leak becomes a big air leak. Entrained air doesn't pump very well.

Hey, it's the long distance diagnosis game, with the usual lack of information. Maybe right, maybe not, but I usually have a logical reason for a suggestion.
 
location, location, location

Where did you mount the aerolab relative to the exhaust tips?

Or how many inches aft of the firewall? I am guessing it is somewhere in the center tunnel belly skin.

I am also guessing on take off, below 500ft, you have your boost pump on?

The exhaust pulse is a vibrational shake and bake to anything on the floor skin near the center tunnel. Might just be bubbles manifesting from the increased flow demand when you pull the nose vertical. A 2 G level turn will need less power and fuel flow than a 3 or 3.5 G pull up.
 
Last edited:
Approximately what is the frequency? I've noticed this at about 2400 RPM as monitored by the G3x Touch. I believe it may have to do with the sampling frequency of the engine monitor vs pressure pulses in the mechanical pump. Unverified, but the oscillation goes away when I dial the rpm down from 2400 to about 2350. The period of the osc. I have observed is about 5 seconds, and it will vary between maybe 21 and 25.

Frequency between the top and bottom is ~5 seconds, between tops ~ 10 seconds, 2300 rpm. Same EFIS/EIS as you, sampling rate is 1 per second.
 
Ok, I have a few things to check but will have to wait till next week now. Any further flights will stay within gliding distance of the airport till I get this sorted out.
 
Mark---can we assume some of the other flights after the changes also were up to 3g, and no incidents? So this indication just started? And boost pump on during loops?
Just asking--
Tom
 
Mark
I am not sure which servo you have but the two that I am familiar with have an internal filter that needs to be cleaned. Symptoms of a partially clogged filter, that I have experienced, is a rough running engine at high power settings and a high angle of attack.
Just another thing to check off the list.
 
Without looking/measuring my guess is 12 - 16" aft of the exhaust. It is in the stick well.

On the belly centerline? As F!R's suggests, a hot canister is a fine way to develop vapor bubbles....and since that particular gascolator is riveted into the belly skin, the entire belly is conducting exhaust heat to it.

Remember, vapor pressure rises with temperature. When vapor pressure exceeds local pressure, bubbles form. A 3G pull drops local pressure as stated previously.

So, is your canister actually hot? Can you fly with the top off the center console so you can reach it?
 
Frequency between the top and bottom is ~5 seconds, between tops ~ 10 seconds, 2300 rpm. Same EFIS/EIS as you, sampling rate is 1 per second.

I'm guessing if you change the rpm some, you'll see the oscillations go away. Regarding the sampling rate, the 1/s is probably what is saved to your flight log. I wasn't able to figure out what the GEA-24's electronics do to sample/condition the raw signals. I see the same oscillations on MAP at 2400 rpm also.

I'd bet your oscillations are measurement artifact. The drop on pulling g's is likely not. I've seen that on occasion for 17 years FWIW, but I've never had any power loss associated with it.
 
I have done some more research into my data logs and have also flown a test flight.

Test flight confirmed this is a "g" related problem, not fuel flow. I can fly 20 gph without the boost pump on and pressure stays up fine. I flew at 14 gph, pulled up into a loop and by the top of the loop fuel pressure was down to 6.8 psi and my boost pump had kicked in. Fuel flow dropped a little to 11 gph and came back up to 14.

Prior to the test flight I opened my stick boot so I could use my IR temperature gun to check the temperature of the gascolator during flight. Max temperature I saw was 100* F. My exhaust is turned down on the ends so should limit the amount of heat applied to the gascolator. Measured distance from the end of the exhaust to the gascolator is 18". Gascolator is offset from the centerline so it doesn't interfere with the stick (located in stick well).

Looking at my logs, it appears EVERY aerobatic flight since I changed out the transducer and the gascolator had at least some level of fuel pressure drop (below 18 psi) associated with increased g force, though not always enough to trigger my boost pump to kick in (11 psi). Previous flights show some corresponding drop (~5 psi) but nothing as dramatic as I am seeing after the gascolator and transducer changes.

Tom, I checked the internal screen on my RSA servo during the annual this winter and it was clean, although certainly worth checking again.

I changed out the Andair gascolator because I had to remove it in order to access the filter. The concept of the Aerolab was appealing because of the ease of access to the filter.

Besides checking for air leaks, the next test I will do is to bypass the gascolator (using an inline filter) and see if the symptoms go away. Next week's project.
 
Replace engine driven fuel pump (?)

- If you apply takeoff power with the boost pump off (as an experiment), does your fuel pressure drop, or stay up in the 20+ psi range? If pressure drops at high fuel flow, that could point to a fuel flow (volume) related pump performance issue. When mine developed that problem, replacing the engine driven pump solved the problem.

Thinking out loud...
- You can perform the loop with the boost pump on (ensuring adequate fuel pressure) without a problem.
- Thus the fuel system components forward of the fuel pressure sensor location (near the engine drive pump output, I assume) appear to be exonerated.
- In the boost pump off configuration, the engine driven pump appears to generate inadequate pressure.
- It could be the case that maneuver G loads are impacting the engine driven pump's performance for some reason.
- Replacing the pump is fairly inexpensive, and not too difficult. Consider putting that on your list of troubleshooting options. (Recall the lacing cord trick to hold the actuating plunger up during new pump installation).
 
Here is an update - I bypassed the Aerolab gascolator and went flying. Fuel pressure stayed above 20 psi. Max pull was 3.5 g's.

The Aerolab is designed such that the inlet is at the side of the bowl and the outlet at the top. I had a 90* fitting at the top and thought that maybe that was restricting the flow. So I changed it to a straight fitting and had a more gentle bend with tubing. I greased the orings in the gascolator with Molykote 55, vacuum tested the unit, and reinstalled it. Went flying today and no joy, fuel pressure dropped off to 8 psi.

My next thought is to reverse the flow through the gascolator and see if that makes any difference. My previous unit was an Andair - it had a horizontal inlet and outlet, both at the top, and worked fine. I am wondering if I make the top port the inlet, and the side port the outlet, what difference that would make. I sent an email to the mfg to get their thoughts. The impact I see would be:

  • Any debris caught by the filter would end up on the inside of the filter rather than the outside (fyi, my gascolator filter has always been clean).
  • The gascolator would hold less water before it ended up going downstream (I have never seen water in my tanks or the gascolator).
  • When pulling the filter I wouldn?t have an automatic shutoff (my fuel valve is quite close to the gascolator, I use it to stop the flow of fuel).

Anybody see any other issues?

Some might say just get rid of the gascolator all together. I agree but in Canada the law says I have to have one.
 
I think from your test of bypassing the Aerolab unit that the problem is somewhere inside it. Looking at the Owner's Manual p7/9, I see the automatic fuel cutoff valve is right beside the inlet. I wonder if this valve is being affected by your 3.5 g pullup and trying to close the inlet port. You might talk to the manufacturer to see if the gascolator can be operated with this valve removed (I know, it's there to prevent fuel draining out when the bottom is opened). I can't think of an easy way to bench test this, so a careful test flight would probably be required.

I don't think it is a good idea to plumb the gascolator to operate in the reverse direction.
 
I wonder if this valve is being affected by your 3.5 g pullup and trying to close the inlet port.

I had the same thought. I took the gascolator apart and looked at this piece in particular but it became obvious that it is not the problem. The cylinder that shuts off the fuel is on top of the rod that is captured by the bottom of the gascolator. It is firmly attached (probably threaded but I am not sure) and it would take a LOT more than a few g to move it - I believe it would have to break first.

I don't think it is a good idea to plumb the gascolator to operate in the reverse direction.

Ok, can you tell me why? BTW, I won't do this unless the mfg says it can be done without issue.
 
Vibration Bubbles

Only a hunch, but I suspect the belly skin mounting could be the culprit as it subjects the entire gascolator to the exhaust pulse .
I am guessing when you had the Andair , that it was NOT mounted to the floor skin.

Either it is the vibration or, it could simply have a lower flow capacity than the Andair. ( set them both up on a bench and time a gravity flow test with the same volume of fuel)

I doubt the pull up is flexing the spar and skin to the point where it distorts the aeoliab and allows air to enter, but it is within the realm of possible.

Perhaps try a foam spacer gasket or else place the entire unit inside of the bottom skin and test fly if you can secure it safely free of the control trunion.

Any chance a flex line in or out of the unit is able to deflect downward with G loading to the point where it restricts flow? Both sides of the gascolator are under suction so the flex lines would not need to droop much to flatten slightly and restrict flow.
Good Luck
 
Last edited:
Either it is the vibration...

I will check the floor for vibration, certainly nothing visible - I suspect my turned down exhausts mitigate that quite a bit.

.... or, it could simply have a lower flow capacity than the Andair.

Not a flow issue. I have burned over 20 gph without the boost pump on, was only burning 12 gph doing a loop today when the fuel pressure dropped.

Any chance a flex line in or out of the unit is able to deflect downward with G loading to the point where it restricts flow?

No flex fuel lines rear of the firewall.

Thanks for the thoughts, one of them will eventually turn out to be the answer. Maybe I just have a weak engine driven fuel pump, and the Aerolab needs a bit more suction than the Andair did.
 
Not a flow issue. I have burned over 20 gph without the boost pump on, was only burning 12 gph doing a loop today when the fuel pressure dropped.

Set gascolator pressure loss aside for a moment, and consider only the effect of flow and G. Pressure loss due to flow increase is a function of line friction. Pressure loss due to G is a function of head, which is elevation change x acceleration. Assume 1 foot of head in level flight, 0.375 ID smooth lines, and 3.5 G. Run the numbers, and you find the pressure drop for 12 GPH at 3.5 G is 2.9 times the pressure drop for 20 GPH is level flight. Head increase has much more effect than flow increase.

Maybe I just have a weak engine driven fuel pump, and the Aerolab needs a bit more suction than the Andair did.

If true, "...needs a bit more suction" would mean the Aerolab unit results in more pressure drop than the Andair. I say "if true", because we don't know it as measured fact. It would be easy to check with a simple experiment.

Heat is also a big factor, and we have reason to believe the Aerolab runs a bit warmer due to its mounting in the belly skin.

Don't blame the engine pump. Remember, when local pressure drops below vapor pressure, bubbles form. Restriction and head reduce local pressure. Temperature increases vapor pressure.
 
Hi Mark,
I also fly an F1 evo, AEIO540 D4A5, two aerolab filter-gascolator installed on the belly with a 90 ? fitting on top, one for each tank.
Only 20 hours flight from the maiden flight.
The fuel pressure normally ranges between 20 and 22 Psi without auxiliary pump. The maximum flow I recorded is 115 l / h, full rich, wide open, 2700Rpm, 1018mb, 22 C ? with 18-20 Psi fuel pressure.
This morning, OAT 20C?,1016, 3500', to check if also my installation is prone to pressure drop, I pulled +4.5 G, 2500 Rpm, 24 Map, 35 l right tank, 65 L left tank, rich, FF 65l/h, controlling the pressure, no variation.
I did it on right and left tank with the same result.
The filters are installed near the exit of the fuel line from the tanks, in the two small bays near the stick, where Mark F. suggests to put the battery.
I do not have any idea about your problem, the hydrostatic head between inlet and outlet of the filter appears to me very small also at 4G and, in all cases, the total head between fuel surface ( in the tank ) and the pump inlet remain the same.
I am obviously very interested in knowing how your tests are proceeding.
If you think I can do something useful I'm at your disposal.
Claudio
 
This morning, OAT 20C?,1016, 3500', to check if also my installation is prone to pressure drop, I pulled +4.5 G, 2500 Rpm, 24 Map, 35 l right tank, 65 L left tank, rich, FF 65l/h, controlling the pressure, no variation.

Thank for your response Claudio, good to hear from you. Can you confirm when you did the test quoted above, was your auxiliary pump on or off?

I found 2 new symptoms today I have never before experienced with this airplane:
  • The ambient air temp was ~80* F (27* C), the hottest I have flown in so far this year. I was 1,000' AGL and went to change tanks. I do not normally use the boost pump when I do this, but I watch the fuel pressure. I have never seen a drop in fuel pressure in the past and have done this in much hotter temps. Today the fuel pressure started dropping from the 24 psi it was holding, 21, 17, 14, and the RPM started to decay. I switched tanks back and switched the boost pump on. I then climbed to 4,500' AGL and, with the boost pump on, switched tanks. No problem.

  • When I returned to the airport I fueled the plane. Next was the hot start. My hot start procedure is full throttle and mixture at idle cutoff. Crank till it starts, pull the throttle back and advance the mixture to half. Done properly RPM never exceeds 1,000 and after 5 seconds the idle smooths out and all is fine. Today it started normally, but then did not want to stay running. I played with the throttle to try to keep it running, but it would not smooth out until I put the boost pump on. Even then it was rougher than normal but ran well enough to taxi to the hangar.
These two events are symptomatic of vapor lock, similar to the g force issue.

I think I will put the Andair gascolator back in and see if I get back to normal operations.
 
Yes, the aux fuel pump was off during the flight.
I do not have your automatic low pressure control ( may I now how it function? )
so the aux pump can't run without intentional pilots action.
My exhaust gas flows near the filters but I see that the flow ( grey smoke tracks) do not impact directly the filter caps . Two Vettermann 3in1. The cowl is not standard to accept a cold air sump.
Claudio
 
I tried to figure out how to verify the temperature-vapor lock hypothesis.
One simple way to exclude exhaust impact is to cover from outside the small area of the filter with thin aluminum foil taped to the belly.
I never had problems switching tanks ( every 20 minutes). I always fly with aux pump off.
If vapour bubbles are generated in the filter , running the aux pump with a clear outlet tube to a canister the bubbles will be visible and also bubbles if air is coming in upstream the pump.
I discovered at my expenses that it is necessary to change compressed o-rings if dismantled and reinstalled.
Claudio
 
I do not have your automatic low pressure control ( may I know how it function? )

I have the Vertical Power VP-200 electrical unit, and this is one of the many features. The VP-200 was discontinued a number of years ago.
 
Thank you, I think it's a very useful feature.
After the pump trips do you have to switch it off? or the pump runs for a wile and go off automatically if the pressure remain stable?
Claudio
 
Tanks Mark,
very rich device. Your successfull application of the FP function suggest me to develop a small circuit only for the automatic aux fuel pump function, may be with a delayed auto off and cyclic pressure test. The alarms are actually managed by the efis.
Claudio
 
Back
Top