Fuel pressure mystery

[Thruster88]

On 28/12/2022 at 3:53 PM, skippydiesel said:

Aircraft; New Sonex, with new Rotax 912ULS. Dynon Skyview engine instruments.

 

Mystery;

Fuel pressure with & without boost pump, meets Rotax fuel pressure standards, while ground running, taxi run up etc. Fuel pressure drops (Dynon alarms), during climb out (when boost turned off) and must be maintained with boost pump on for duration of flight. Pressure returns on landing. Engine has never shown signs of fuel starvation (loss of power/hesitator, etc).

 

Have checked & rechecked:

• Fuel sensor security  (attached to firewall)-appears to be okay.
• Gascolator screen for contamination - all good
• Aircraft has two wings tanks - makes no difference which one is in use.

 

Wondering;

Faulty sensor reading but why only in flight?

If cowling air pressure  (through drain hose) could be influencing the pump diaphragm.

 

Ideas?

 

 

 

Based on my experience with the RV described above, the sensor is not likely to be at fault. My aircraft also has two tanks.

 

There are three possible causes. 

 

1 the mechanical engine fuel pump is faulty. (my case)

 

2 Air is able to enter the fuel system before the mechanical engine pump. (unlikely)

 

3 There is a restriction in the fuel system before the mechanical engine pump. This could take the form of a hose kink, faulty fuel selector,  internal hose collapse or just excessively strong springs in the boost pump valves. At low flow on the ground (0.1  LPM?) this will not a problem. When you give it the beans on takeoff and climb, the fuel flow rate jumps to 27lph. Any restrictions, faults, would become evident however with the boost pump on ( assume it is mounted close to the tanks) faults will be masked until switching it off.

 

Have just read your test flight report today. Suggest you switch off boost pump at 100% power, safe altitude, to ensure engine pump can supply correctly.  

 

[IBob]

1 hour ago, skippydiesel said:

Reporting in after .9 hr test flight -

 

Before TO, checked fuel pressure sensor has a small (about 1.5mm) hole in casing, to atmosphere - appeared to be clear.

 

After leveling out

Turned boost pump off, fuel pressure dropped to below 2 psi  - fuel pressure then returned to 4.5 psi & stayed there. This suggests to me that the pressure sensor may be taking time to "catch" up with change in air pressure.

 

(FYI - Rotax Forum - seems quite a few pilots have the same/similar problem and never find the exact cause - most just live with it)

 

After landing

Noted fuel pressure at 3500 rpm, about 4.5 -5psi

 

After shut down

Checked sample bottle on end of air box breather - about 15ml fuel. Seem I have a flooding/overflowing float chamber - will investigate tomorrow.

Checked sample bottle end of pump drain - bit hard to see but might be empty  - will investigate tomorrow.

Skippy, I confess I have not spent time studying my fuel pressure in flight. I do know that sometimes it is at or just above 2PSI in a climb, other times (and more usually) it is at or above 4PSI. And I have seen it shift quite quickly from one to the other, but I don't know why it should do that.
I shall watch it more carefully when next out, and see if I can add to that. I am on steam gauge.

I don't know why your sensor would be slow to catch up with air pressure. Having said that, these sensors (once you get away from the VDO stuff) are standard industrial equipment, and I suppose it is possible they could have a very tiny or filtered internal atmospheric aperture....since they are normally deployed in static applications, not required to respond to rapid or large atmospheric changes, and the manufacturers quite probably have internal filtering etc to protect the works from the ingress of moisture and dirt.
 

It's certainly interesting. If problems persist, one option would be to borrow a steam gauge and plumb it in, You would then be able to clearly establish if you have a fuel pressure problem, or a fuel pressure sensing problem.

[IBob]

PS here's a further minor complication for you: the air behaviour under your cowl quite probably alters not only with speed, but also with angle of attack. For instance I found the Naca scoop on my cowl top doesn't work well at high angles of attack, so I was drawing air from under the cowl instead of outside air. I was able to correct that by greatly reducing the gap between Naca scoop and airbox inlet.

How much such things may affect your fuel pressure sensor mounted in the engine bay.......if at all.......we'll never know, without something else to check it against. But I'm not surprised to hear that some pilots get odd results, if that's where they are putting them.

[RFguy]

 

"Seems I have a flooding/overflowing float chamber - will investigate tomorrow."
Skippy -  Be very mindful of a fire hazard .  I know you know this but worthwhile  highlighting. Am pleased to see your overflow goes somewhere instead of just into a tray under the carb....

Skip do you have a fuel tank return in the setup ? another thing to .....

[skippydiesel]

1 hour ago, RFguy said:

 

"Seems I have a flooding/overflowing float chamber - will investigate tomorrow."
Skippy -  Be very mindful of a fire hazard .  I know you know this but worthwhile  highlighting. Am pleased to see your overflow goes somewhere instead of just into a tray under the carb....

Skip do you have a fuel tank return in the setup ? another thing to .....

Yes,  to the return line - due to an unusual fuel tank(s) configuration, I am a bit concerned about the amount (too much)of fuel the return delivers. Another niggle that will be addressed in due course. The niggle list doesn't seem to get any shorter - par for the course?

[IBob]

Skippy, the Sav (amongst other aircraft) has a 6L 'receiver' tank behind the passenger seat with a drain cock underneath (this being the low point in the system) and a float switch in the top which activates an indicator on the panel unless it is full.
This arrangement gives approx 20mins of warning of fuel exhaustion. It also compensates for unporting of tanks, as may happen in long descents or momentary unbalanced flying. And some folk run their fuel return back to that tank, so avoiding the problem of overfilling a wing tank with returned fuel.
All in all, I see the receiver tank as a very worthwhile innovation. Might it be possible to fit something similar, albeit smaller, in the Sonex?

[skippydiesel]

FYI - The intermittent low fuel pressure story is now centering on the return line jet/restrictor.

 

Rotax recommend;

• a #35 slow/idle jet as the restrictor
• 2-3 L/hr return line fuel

So far the only known fact is that my system is returning just short of  7.5L/hr (Hobbs).

This flow rate will give excellent fuel vaporisation scavenging/removal and fuel cooling BUT has implications for fuel delivery at max engine  power.

 

I am told that the pump is rated at max 35L/hr. The engine may require near 30 L/hr at take off power. If the max fuel demand is a combination of engine and return line, this may be around 37L/hr - exceeding the pumps capacity - leading to a low fuel pressure situation during climb out. (Information comes from Rotax Owners Forum)

Should this be the case changing the existing restrictor for a smaller orifice may remedy the situation.

 

My restrictor was supplied by B Flood however it turns out that #35 jets (if that is what it is) can be measured differently depending on origional supplier.  Flow rates vary from 2-7.5L/h r@ 3.5psi

 

I will be attempting to measure my restrictor today.

Edited January 11, 2023 by skippydiesel

[IBob]

Hi Skippy, following this with interest.
Here is a pic of the (spare) return jet setup from my aircraft. (It as 2 identical jets, the second one goes to the fuel pressure gauge).
My measure of the hole is a best guess: I can get a .35mm strand of wire into the hole, but certainly not 2 strands.

I would guess the hole is 0.5mm

 

I should add that this is mounted where the fuel line splits to the two carbs, and works fine for me: I have had vapour lock after short shutdowns: I always use the booster pump pre-start, and with vapour it runs fast  for 5 to 15secs before settling down once the vapour is gone, at which point I get a fuel pressure reading.  (Obviously, vapour will pass through the jet much faster than liquid.)

 

 

Edited January 11, 2023 by IBob

[RFguy]

and I beleive that the fuel returns for the rotax need to go from T jets at the carburettor fuel inputs, not at the fuel pump output ... the fuel pump gets a good air blast, but it can be pretty damn hot in the fuel line running over the top of the engine which is very crowded and hot.  hence the returns at the carbs.

Also, the float bowls need heat shields. (IMO)  many aircraft have drip trays under the carbs which function well as shields.  I would always have the fuel overflow/etc pipe not drip into the tray like some aircraft I see with a drain pipe on the drain .  jesus no, needs to go out in a pipe away from the exhaust pipe.  otherwise just what you need a drip tray full of fuel !

Edited January 11, 2023 by RFguy

[skippydiesel]

I now have an orifice measurement (sort of)  for the restrictor/jet fitted to my Rotax.

Sort of, because I used oxyacetylene nozzle cleaners/wires to do the job.

I am imperial measurement challenged, so I hope you can make sense/convert the following;

Oxyacetylene Cleaner No 22 (Drill 55-54) fits. When I say fits,  the non burred (smooth) end of the cleaner fits snuggly, the burrs do not.

Cleaner No 24 (Drill 53-52) is loose and this includes the burred section.

The above, are found on the back of the cleaner kit - got no idea what they mean to a metric person😀.

[IBob]

Skippy, so #54- #55 is 54 -55 thou. (or 1.3 - 1.4mm)

Which is huge.
My jet (see above), is approx 20 thou. (or 0.5mm)

[facthunter]

Getting that bleed hole too big is not safe and the carbs should not be on the top of the engine. Injection HAS to be the go for a multitude of reasons.  Nev

Edited January 12, 2023 by facthunter

[skippydiesel]

27 minutes ago, IBob said:

Skippy, so #54- #55 is 54 -55 thou. (or 1.3 - 1.4mm)

Which is huge.
My jet (see above), is approx 20 thou. (or 0.5mm)

Hmmm! I tried to get a 1mm drill bit shank into the orifice without success. The  jet has a wider outer opening than the inner. !mm drill bit would not go into the inner opening, whereas the No 22 fits snugly - something does not add up here.

[skippydiesel]

2 minutes ago, facthunter said:

Getting that bleed hole too big is not safe and the carbs should not be on the top of the engine. Injection HAS to be the go for a multitude of reasons.  Nev

Ah! but do you know the diameter of a No 22 oxyacetylene cleaner?

[facthunter]

I have one of a range of sizes and use it on oxy tips to clean them out. You can check the plain part with a micrometer. Fuse wire is soft and good to clear small orifices. You could drill out and tap your fitment to take AMAL carburetter jets. Their numbers represent a flow RATE and I think the thread is 2 BA.. They are brass and hexagonal.. You have one in your ARIEL twin.   Nev

[rodgerc]

When I installed my 912ULS fuel return line the BPR-Rotax Installation Manual’s specified “pilot jet 35” oriface was 0.35mm / 0.014”

Edited January 12, 2023 by rodgerc
Image added

[IBob]

42 minutes ago, skippydiesel said:

Hmmm! I tried to get a 1mm drill bit shank into the orifice without success. The  jet has a wider outer opening than the inner. !mm drill bit would not go into the inner opening, whereas the No 22 fits snugly - something does not add up here.

Sounds like you need to get the V callipers/micrometer out and measure...um....what you are poking in there?

 

Based on your reported flow (which must be a constant if the pressure is) I would guess your jet is just under 1mm.

Edited January 12, 2023 by IBob

[facthunter]

Size and finish affect flow. You could easily be 7x where you  should be. Looks as though you have found your problem.   Nev

[rodgerc]

Might also be worth mentioning at this point that low wing aircraft utilise/benefit from a different model Facet pump (higher head) than that used in high wing designs. 

[skippydiesel]

At the risk of making a fool of myself this is what I have found;

Using my very old micrometer screw gauge (metric)

Cleaner 18 for drill 56, measures 0.510 mm

Cleaner 22 for drill 55-54, measures 0.503 mm

Cleaner 24 for drill 53-52, measures 0.450mm

Cleaner 26 for drill 51-50-45, measures 0.365 mm

The larger the cleaner number, the smaller the wire/orifice. My cleaners go up to 6 for drill 77-76

So, if my gauge is correct, the diameter of my jet is about .503 mm.

My 1mm drill bit, did not fit.

[IBob]

Skippy, I am confused by (part of) this:
EG: the drill size chart I am looking at has a 56 drill as .046" or 1.18mm

So clearly we are looking at different standards/charts.

BUT

 

The main thing here is that you have been able to find something that fits in the jet, and that you can then measure with your micrometer.
I can only add that, if your jet is 0.5mm, then it is approximately the same size as the one I have fitted (may be exactly the same size, but I have no exact fit to measure).
I hope that is of help.

[Yenn]

If you have a micrometer, then get a piece of wire which doesn't quite fit and file a taper on it. Push it in till it stops, then measure at the point where it stopped and you have the hole diameter. Just need to have a nice even taper on your wire.

[Blueadventures]

4 hours ago, rodgerc said:

When I installed my 912ULS fuel return line the BPR-Rotax Installation Manual’s specified “pilot jet 35” oriface was 0.35mm / 0.014”

35 is likely 0.35mm orifice.

[facthunter]

That's about ALL you need. To measure insert a needle and note how far it goes and measure THAT part of the taper. (As YENN says). with a micrometer. Nev

[skippydiesel]

1 hour ago, IBob said:

Skippy, I am confused by (part of) this:
EG: the drill size chart I am looking at has a 56 drill as .046" or 1.18mm

So clearly we are looking at different standards/charts.

BUT

 

The main thing here is that you have been able to find something that fits in the jet, and that you can then measure with your micrometer.
I can only add that, if your jet is 0.5mm, then it is approximately the same size as the one I have fitted (may be exactly the same size, but I have no exact fit to measure).
I hope that is of help.

IBob

I find almost every aspect of Imperial measurement to be confusing - in my humble opinion the very many Imperial  system(s) should have been abandoned long long ago.

 

The gauges I used to get my hole diameter are in fact oxyacetylene cleaning wires. The back of the cleaning kit (there are many wires of different diameters) has the Cleaner numbers & Drill number. The numbers mean nothing to me. To find the Cleaning number 22/ Drill 55-54 I first found which Cleaning wire fitted the hole in my jet. I then determined where inlay in the size chart. I did this by counting the number of Clear wires of diminishing size and compared that with the chart. For all I know, this is another mysterious Imperial measuring system, unique to oxyacetylene cleaning tools.

 

As for the ramifications of my 0.503 mm jet discovery -

• The only fact that I am sure of  is that my engine fuel return line is delivering 7+L/ hr(Hobbs) through the oft mention restrictor jet.
• Some have suggested that Rotax specifications call for 2-3L/hr.
• The #35 restrictor jet seems to be a Rotax standard
• Some have pointed out jet manufactures can have differing fuel flow performance criteria for a #35 jet. The variation being in the realm of about 3-7 L/hr

Reducing my jet size to achieve 2-3L/hr will possibly;

• Benefit the low fuel pressure that occurs when max  engine power (max fuel demand) is required.
• Increases the chances of fuel vaporisation and the speedy venting of the same.

As my engine/pump are new, I am also concerned by the possibility that the pumps performance may degrade slightly over its 5 year service life span. If this does occur and I don't change the jet ,then the low fuel pressure warnings I am currently getting ,may become actual engine fuel starvation, with the potential for serious incident.

 

If there is no further commentary/advise on this matter, I think I should change my fuel return line reducer/jet to one that delivers 2-3L/hr