Fuel Flow/Pressure Anomalies

[skippydiesel]

While I quite enjoy & learn from many of the  "thread drifts" that have come to typify this Forum, I am concerned about the tone of some of the respondents and the lack of any relationship to matters aircraft.

 

SO! I thought I might repost this interesting (to me) matter, that is homebuilt aircraft in its origin & intent;

 

"My Sonex/912ULS has an interesting fuel anomaly;

To give you some ides where the matter may arise from, I should tell you that my Sonex is fitted with 2 x 30L wing tanks and a 40L in fuselage "header" tank. I call it a "header" because it has no external fuel point, being "fed" from the two wing tanks and the fuel return.

I can select from any of the three tanks but prefer to use just the "header" .

The plumbing to to achieve, the above, is quite convoluted.

I have an electric fuel transfer pump and a "boost" / auxiliary. The latter being used for all higher risk situations i.e take-off/landing, low level operations.

For the most part fuel pressure is between 4-5+psi.

The anomaly;

In Cruise, my Dynon will occasionally (it seems to be quite random) give me a low fuel pressure warning - Pressure may be indicating a low as 1 psi. 

If I am high, I will just monitor the situaton and pressure will slowly recover, without any intervention on my part.

If low, I will turn on my Boost pump for an immediate increase to 5 psi - Turn off the pump, pressure remans between 4-5psi.

At no time has the engine ever hesitated.

The above is puzzling enough - what could be happening ?

BUT

There is another weird symptom - When I switch on the Boost pump, in response to low pressure, my fuel flow rate reduces by up to 5L/hr.

My Red Cube (x2)/Dynon fuel flow indication has, despite my best efforts to calibrate accurately, never been much good - usually over indicating fuel flow by a significant margin.  I think it likly, the indication, with the Boost pump on, is reflective of the actual flow rate in Cruise.

The Fuel Used/Remaining is acurate.

To rule out engine heat causing fuel vaporisation, I have double insulated the, in cowling, fuel supply to the mechanical pump. The fuel distribution lines from the pump are all coved with orange fire sleeve.

 

My best guess:

Due to the convoluted fuel supply system, low pressure is being generated somewhere, causing the fuel to "cavitate". The cavitation encourages bubbles to form. The bubbles effect the fuel pressure sensor - low pressure warning!

Why:

•    The above only seem to happen occasionally, I can not explain.
•    Would fuel flow indication, reduce with the Boost pump on?
•    Does my fuel flow, without the Boost pump, indicate high flow?

 

Anyone got any other suggestions?"

 

😈

[Gadjet]

My random thoughts FWIW:

 

As you say, you may have a low pressure in the tanks due venturi effect or climbing to altitude. Do the pressure indications change with time/altitude/airspeed. Is the fuel pump a suction or pressure pump?

Gravity: Relative positions of header tank and wing tanks.

 

Edited May 11 by Gadjet

[Thruster88]

We would need to see a diaphragm of the complete fuel system showing tanks, selectors, pumps, vents, orifice, check valves, flow meter,  pressure transducer.

 

Does this pressure drop occur  if using from the left or right tanks or only the main? 

[skippydiesel]

9 hours ago, Gadjet said:

My random thoughts FWIW:

 

As you say, you may have a low pressure in the tanks due venturi effect or climbing to altitude. Do the pressure indications change with time/altitude/airspeed. Is the fuel pump a suction or pressure pump?

Gravity: Relative positions of header tank and wing tanks.

 

Each tank has its own breather, facing into airflow, so should be between ambient and slightly elevated (+) pressure.

Correction; My speculation was low pressure (cavitation) in the reticulation system, not in the tanks.

For the most part pressures remain stable from Take-Off, through Cruise, to Landing. Not every flight involves a low pressure indication, but it is common.

As far as I understand pumps, they all exhibit suction, on the inlet side and pressure, on the delivery side. Rotax 912 use a diaphragm pump.

The wing tanks are below the "header". The header may have some "gravity" flow but it would be minimal, primarily when full. Fuel supply is ensured by an engine/mechanical pump.

I have yet to determine a consistent "condition" for the low pressure indication.

😈

 

[skippydiesel]

5 hours ago, Thruster88 said:

We would need to see a diaphragm of the complete fuel system showing tanks, selectors, pumps, vents, orifice, check valves, flow meter,  pressure transducer.

 

Does this pressure drop occur  if using from the left or right tanks or only the main? 

I can't see me developing the complex  diagram with all the features you describe (not that good of a draftsman).

 

There is the option to operate from any one of the three tanks (or all three at once) however I use the "header" tank for all operations and the wing tanks to supply the header.

 

While a picture may save many words , I will attempt the words-

• With the exception engine bay all fuel lines are aluminum.
•  All tanks have their own breather, facing into air flow.
• A gascolator is located at the bottom of the engine side of the firewall 
• There are three fuel drain valves
• In line  Boost/Auxiliary pump (with a fuel bypass circuit in case of pump failure) used for all higher risk operations - TO/Landing & Low Level
• Inflight fuel transfer pump
• A four way wing tank selector valve - Left / Right / Both / Off. This valve is only ever on for fuel transfers
• The header has an ON /Off valve. On for all flight/engines running operations
• The system has a fuel return, from the distributor to header tank. This is fitted with a #35 restrictor jet. Fuel return flow was measure (using Boost pump) at 7L/hr
• The two (30L) wing tanks are located about mid wing
• The "header (40L) is in fuselage below the glair shield. When full (37 L), this tank may provide some gravity feed
•  Red Cube fuel flow sensors x2 . One located between engine pump & fuel distribution block, the other on the return to tank line. Installed as per instructions however I may change this sometime in the future
• The pressor sensor is located as high in the engine compartment as possible/above the fuel distribution block to which it is connected via rubber hose (The seal around the sensor, has been intentionally damaged to facilitate ambient pressure entry)
• Fuel Pressure, Flow, Fuel Remain & Used, are displayed on a Dynon Classic Skyview EMS
• Fuel Used/Remain indication is acurate to about 1L in 4 hrs
• Fuel Flow indication only drops below 17L/hr when engine rpm reduced for descent. Calculated (Fuel added) Average fuel flow, for whole of trip, is below 15L/hr (usually 14.25)
• There is only one "check" valve (other than fuel pumps) which is in the Boost pump bypass circuit
• The gascolator to engine fuel pump line has been heavenly insulated, to rule out temperature related fuel vaporisation
• The over engine fuel fines, from pump, through distribution block to carburettors, are all orange fire sleeved, also to mitigate against fuel vaporisation
• The wing tank fuel selector & Boost pump system/associated plumbing, are in a  very"tight package"

😈

[T510]

Are you sure the fuel pressure sensor is accurate?

 

I spent ages chasing a boost issue on an engine only to find out the boost pressure sensor was randomly giving an incorrect reading.

 

90% of the time it worked fine, the other 10% it was borked. Only managed to find it by data logging the sensor output

[skippydiesel]

Hi T510,

 

With small aircraft one should always be sceptical about the readings from engine condition related electronic indicators.

 

The fuel pressure indication, pre engine start, run up & TO/Climb are all as I would expect - cant say that it may  be "borked" from time to time.

 

I guess I could interrogate/down load the Dynon & see if there is any tendency to bork. As a slight mature person, all this computer stuff is a tad intimidating but will give it go.

 

 Its not just pressure anomalies, its also flow. Completely separate & type of sensors so may not be related.

 

😈

 

 

 

 

 

Edited May 11 by skippydiesel

[onetrack]

Fuel flow operating conditions that produce "anomalies", are usually the result of something physical that can move around in the fuel system - or there's a physical fuel flow impediment that only raises its head under specific conditions.

 

As an example, many decades ago, I was travelling with an older brother and he was driving an old International truck that belonged to his boss, a builder. The truck started and ran fine, and went well for most of the trip until we ran into some hilly country.

Upon the need to have maximum power from the engine, it started to show signs of fuel starvation - to the point eventually, where it wouldn't climb one relatively small hill. The brother suspected a fuel supply problem, and as he was quite good mechanically despite being only a young man, he promptly removed the main fuel line from the fuel tank, found some fencing wire, and poked it up the fuel line.

To our amazement, a piece of circular thin plate, virtually shaped like a carburettor butterfly, fell out of the end of the (steel) fuel line. This tiny piece of plate appeared to be the centre remnant from a hole punching machine.

It immediately became obvious, that being a neat fit in the fuel line, when it went up the line, it was initially moving with its body parallel to the line. Then it must have reached a restriction point where it was turning at 90° to the line and effectively shutting off the fuel supply, like a tiny shutoff valve. We replaced the cleaned fuel line, and the problem was solved.

 

I'd suggest it's possible there's a manufacturing obstruction in the fuel system, that is creating intereference in the fuel flow when it reaches a specific set of circumstances. Sometimes that restriction is where fittings have been drilled during manufacture, or holes drilled in other components such as tanks. Even an irregular shaped orifice can cause restriction or cavitation under specific conditions.

 

Without seeing photos or the layout of the fuel system, the problem is difficult to diagnose exactly. Because there are electronics/electrics involved in the fuel system as well, this adds another layer of source potential for the problem.

I would start by trying to determine if the problem is physical fuel flow restriction, or an electronics/electrics problem. Fault finding is a process of gradual elimination, and can be very time-consuming.

[facthunter]

Chopping off rubber pipe often produces a scarf that would cause flow Problems. Be super careful with everything.  Ensure the flow is about 150% of Max rate required.  Nev

[PureCaboose]

When you say the fuel breathers are air flow facing, if they are on top of the wing they may be subject to low pressure and induce a vacum. I had this on a Gizelle where if you faced the breather what felt to logical way ended up causing an issue.

[facthunter]

A Poor  sealing fuel cap can do that. Nev

[skippydiesel]

19 minutes ago, PureCaboose said:

When you say the fuel breathers are air flow facing, if they are on top of the wing they may be subject to low pressure and induce a vacum. I had this on a Gizelle where if you faced the breather what felt to logical way ended up causing an issue.

 The wing tank bather, for each tank,is under the op posit outer wing😈

Edited May 12 by skippydiesel

[skippydiesel]

NOTE: At no time has the engine shown any indication (hesitation) of low/nil fuel supply😈

[T510]

2 hours ago, skippydiesel said:

Hi T510,

 

With small aircraft one should always be sceptical about the readings from engine condition related electronic indicators.

 

The fuel pressure indication, pre engine start, run up & TO/Climb are all as I would expect - cant say that it may  be "borked" from time to time.

 

I guess I could interrogate/down load the Dynon & see if there is any tendency to bork. As a slight mature person, all this computer stuff is a tad intimidating but will give it go.

 

 Its not just pressure anomalies, its also flow. Completely separate & type of sensors so may not be related.

 

😈

 

 

 

 

 

From your first post it sounds like the issues are not necessarily related. 

 

First issue being the low fuel pressure warning which triggers you switching the boost pump on which solves the low fuel pressure issue.

 

The second issue is a decrease in indicated fuel flow rate when the boost pump is activated.

 

 Do you use the boost pump in any other flight modes, ie landing or take off, or do you only use the boost pump when you get the low fuel pressure warning?

 

If you don't get the low fuel pressure warning and switch the boost pump on do you still get the reduced indicated fuel flow?

 

Does the boost pump bypass have a one way valve? If it does could this be staying partially open allowing the pump to recirculate fuel?

[skippydiesel]

T510,

 

You seem to be really getting into this little anomaly.

 

"Do you use the boost pump in any other flight modes,"

Other than what I call high risk situations (TO/Climb/Low level) I do not routinely use the Boost pump (the exception to this rule being low fuel pressure warning)- See my earlier comments

 

"If you don't get the low fuel pressure warning and switch the boost pump on do you still get the reduced indicated fuel flow? "

The Boost pump ON, reduction in flow, only happens when flow indication is above that what I would expect in Cruise. High Flow indication is almost the norm ie 17-19L/hr, when I would expect sub 15L/hr. Turn the pump on, more often than not, the flow indication will go to something like what I would expect.

 

"Does the boost pump bypass have a one way valve? "

Yes - as stated above.

 

"If it does could this be staying partially open allowing the pump to recirculate fuel?"

Interesting thought - please expand.

 

😈

 

 

[T510]

Skip,

 

I spent years in R&D, testing and manufacturing as a small engine specialist. I enjoy this type  of challenge.

 

If the one way valve on the bypass is not closing fully you could be getting reverse flow through the bypass circuit. 

 

Is your fuel pressure sensor inline or dead head style?

Have you tried cracking the fitting on the fuel pressure sensor when it is pressurised (preferably without getting a face full of fuel) to see if there is air in the line? If there is, the air being compressible, could give you false fuel pressure readings.

As the pressure sensor is the highest point in the fuel system if I understand your description correctly, any air bubbles could be expected to rise to the pressure sensor.

 

How are the internal fuel tank pickups connected to the external fuel lines?

 

 

[skippydiesel]

3 hours ago, T510 said:

Skip,

 

I spent years in R&D, testing and manufacturing as a small engine specialist. I enjoy this type  of challenge.

 

If the one way valve on the bypass is not closing fully you could be getting reverse flow through the bypass circuit. Boost pump On without mechanical gives 5psi. With both about 5.5psi

 

Is your fuel pressure sensor inline or dead head style? Dead head (only ever seen this style in Rotax insulations)

Have you tried cracking the fitting on the fuel pressure sensor when it is pressurised (preferably without getting a face full of fuel) to see if there is air in the line? If there is, the air being compressible, could give you false fuel pressure readings. There is air/gas in the line. As per insulation instructions sensor is set high, relative to fuel distribution point, so that raw fuel unlikly to contact sensor and for fuel to drain away when system turned off (not sure if latter is [possible)

As the pressure sensor is the highest point in the fuel system if I understand your description correctly, any air bubbles could be expected to rise to the pressure sensor. No -Sensor take off point ,from fuel distribution block, is below return line (vapour extraction) take off point.

3 hours ago, T510 said:

How are the internal fuel tank pickups connected to the external fuel lines? Hard/threaded fittings (no leaks)

 

 

😈

[skippydiesel]

Read installation where spell check has delivered insulation

[Thruster88]

Pressure is Pressure, it makes no difference to the sensor if it is seeing liquid or gas. Gas will give a benefit of damping any pulses from the fuel pumps.

 

The instructions re the sensor to keep raw fuel away from it is a concern, is it 100% compatible with ulp? 

 

When you have the low fuel pressure indication in cruise flight have you tried, without the boost pump, going full throttle, if it is a flow problem the Pressure should drop further or increase Pressure if you go to idle.

 

Perhaps you need a good old mechanical gauge to know what is really happening.

[skippydiesel]

Thank Thruster,

 

"The instructions re the sensor to keep raw fuel away from it is a concern, is it 100% compatible with ulp? "

 

The problems with the fuel pressure sensor are  well documented (check out Rotax Owners Forum). Various "fixes" to pressure reading anomalies have been recommended (all of which I have applied). I have no reason to think that the pressure reading is wrong, more likly effected by bubbles/vapour causing the mechanical pump to loose efficiency/pressure.

 

"When you have the low fuel pressure indication in cruise flight have you tried, without the boost pump, going full throttle.....",

No I haven't tried incensing power - will give that a go next opportunity

 

"Perhaps you need a good old mechanical gauge to know what is really happen"

 

Yep! It could only be temporary - no panel real estate left

 

😈

[T510]

This is a good one Skip, you certainly have me scratching my head.

 

Where does the return enter the header tank? if it is in the top it could be aerating the fuel, I had a fuel aeration issues on one of my race cars that was fixed by returning in to the bottom of the surge tank instead of the top.

 

If it where my aircraft the next steps I would take is to confirm the fuel pressure and flow readings are accurate.

 

I would be calibrating the fuel pressure gauge using a syringe with the required weights and using a graduated cylinder to calibrate the fuel flow sensors.

 

I would also confirm the one way valve in the pump bypass is operating as intended. 

 

[IBob]

Skippy, a short anecdote to support my much earlier suggestion:

   The Savannah puts hot air into the airbox when carb heat is selected.
   I had poor indicated carb heat, which I improved a bit with some attention to the ducting, but it was still very poor.
   So, either I had poor carb heat, or I had a faulty carb heat reading.
   If I could determine which, I could then greatly narrow down the problem.
   I bought a very cheap digital temp gauge with a 1m probe lead, tested it (iced then boiling water) and installed it in the airbox. 
   And got the expected much higher readings. So, faulty carb heat reading.
   (There is a further 'gotcha' in this, but I'll put that at the end.)

It seems to me you have a similar situation: is it the fuel pressure, or is it the fuel pressure sensing. In your situation I would be strongly inclined to temporarily install a steam gauge. You will then know which side of the question to be looking at, greatly reducing the number of possibilities and variables that you are currently chasing.
 

 

  (The gotcha: had I tested the probe supplied with the kit (iced then boiling water), I would have found nothing wrong with that. The problem wasn't the sensing, but the probe type (very short) and it's mounting (into a

   thick threaded bush in the airbox wall). So what that was measuring was not air temperature but airbox wall temperature. And the wall is constantly cooled by outside air.

   I did write to ICP about it, they thanked me politely, but I have no idea if they made any subsequent changes.)

 

Edited May 13 by IBob

[skippydiesel]

I will throw in a suggestion from Rotax Owners Forum;

 

"While I can only speculate as to your low fuel pressure readings, I can perhaps suggest what may be afflicting your fuel flow readings.

The EI FT-60 "red cube" fuel flow transducer operates using an optical sensor which senses when the light from an internal LED is interrupted by the rotor passing between the LED and the sensor. Bubbles in the fuel can interrupt (refract) the light from the LED, which the sensor cannot differentiate from interruptions caused by the rotor. Either event will cause the transducer to send a pulse to the EMS. Additionally, because vaporisation (what you've called cavitation) increases the volume of the fuel itself, the rotor would also spin faster for the same mass of fuel passing through.
Therefore when fuel vaporisation is present the fuel flow transducer will erroneously present a higher fuel flow measurement to the EMS. This can apply to either or both transducers, but probably not in such a way that the error in one exactly cancels out the error in the other. Turning the boost pump on probably suppresses the formation of vapour in the fuel line, allowing your flow transducer(s) to operate as designed while the boost pump is on.

As I touched on above, fuel vaporisation is exactly the same phenomenon as what you referred to as cavitation. From a physics point of view, it doesn't matter whether it is the fluid's temperature rising above its boiling point at a given pressure or if it is the fluid's absolute pressure dropping below its vapour pressure at a given temperature - vapour will form just the same. Excessively low pressure in the fuel line on the suction side of the pump (whether caused by flow restriction/s or a large 'lift' distance) will cause vaporisation just the same as heating the fuel by too much will."

😈