Fuel flow riddle

See the graph below. This was taken from the log of a short flight yesterday. Total flight was a half hour, the log starts at about the 15 minute mark. Elevation was 3,000 ASL (~1,250 AGL) the entire time. Aircraft is an F1 Rocket, IO-540, constant speed prop.

Question: What caused the initial fuel flow deviation?

The initial increase from 11 gph to 20 gph (as well a changes to MP and EGT) occurred without any intervention from me, however the subsequent decrease was caused by me leaning with the mixture (till the engine stumbled) and then richening it back again. The final drop off of MP is me pulling the throttle back. There is no indication of fuel leakage, and I flew two flights today with normal fuel flows.

I am withholding one critical clue that I believe was the cause, however lets see what you all think could cause this.

?

Boost pump?

First of all, at that power setting, 11gph seems way too low, which is confirmed by the engine stumble when you leaned that far.

Since it is a rocket, it makes me wonder if it was the flight attitude that caused the low reading. The red cube will only read accurately in certain orientations.

It also could be something obstructing the flow of fuel, but that is not likely unless the engine was running very roughly and the EGT is where it should be.

I have seen some of these sensors read low because they are going bad and there is friction in the wheel, but I would expect it to continue being low if this were the case.

You should really give all the information unless you are wanting some to post a foolish remark. A riddle has all the information contained, however obscure.

If the fuel system had been opened and air entrained, it is possible that an air bubble interfered with the flow indication for a short period then cleared itself.

Did your mixture control creep? looks like normal ff, leaned out at low power cruise, followed by a sudden switch to rop conditions at much higher ff.

Perhaps the gib messing around, or a loose quadrant friction control or similar.

Help us out.

More Info

Here is what I found next the cowl after I shut down.



I have a spring loaded alternate air door on the air box, I believe this bag partially or fully blocked the air intake which resulted in the increased fuel flow.

Well - I am not sure if you are saying this was sucked in the air intake, or was sucked into the alt air door from the lower cowl. If you are running at 23 by 2300 and the alt air door is opening due to the bag covering the filter, then it works, if it was in the lower cowl and got sucked into that gap, then the alt air door spring is weak or filter pressure drop is high.

Thanks for posting, it reminds me to add filter pressure drop to my instrumentation list for Phase I.

SMO said:

I have a spring loaded alternate air door on the air box, I believe this bag partially or fully blocked the air intake which resulted in the increased fuel flow.

Click to expand...

That's a phenomenon Don measured on the AFP flow bench. This spring loaded alternate air door caused a fuel flow increase with the primary intake blocked, although not nearly as much as the 9 GPH in your graph.

OK, I'll admit that I'm not understanding how a blocked air intake (and I'm assuming a reduction in air flow) resulted in increased fuel flow. Can some one explain what is happening in this situation?

Think of a choke

Todd,
If you were an old geezer, you would remember a knob on the dash of cars until (about) 1950 or so, called the choke. It controlled a butterfly disc in the carburetor, similar in appearance to the throttle, but at the top (inlet) of the carb. When partially closed while cranking the engine, it increased the vacuum (decreased the pressure, actually) in the carb passage, thus causing more gasoline to enter the air stream. It was needed while cranking, due to the low velocity of airflow not creating enough pressure differential in the venturi.
Since the engine here is fuel injected, the effect is not as pronounced, but apparently the fuel flow is influenced to some extent by a significant drop in pressure at the point of injection. That's my thought, anyway.

My thought is that the fuel servo interprets a lower pressure within the body (relative to the external static pressure) with higher velocity airflow through the device so will add fuel. Correct?

The FI regulator (aka servo) measures intake air velocity and increases fuel flow accordingly. i.e. more power demand increases measured intake air velocity and resultant fuel dispensed by the regulator. If the intake were blocked, I would think the regulator would see very low intake velocity and hence flow reduced fuel. I must be missing something here.

Jesse said:

First of all, at that power setting, 11gph seems way too low, ....

Click to expand...

My fuel flow instrumentation (GRT) has proved to be accurate in the past. At 23 squared I can normally pull the FF down under 10 gph before it stumbles - injectors are balanced so peak egt happens on each cylinder within ~.2 gph.

Jesse said:

Since it is a rocket, it makes me wonder if it was the flight attitude that caused the low reading.

Click to expand...

Attitude was straight and level flight.

BillL said:

Well - I am not sure if you are saying this was sucked in the air intake, or was sucked into the alt air door from the lower cowl.

Click to expand...

I believe the plastic bag was blocking the air intake and the alt air door operated as it should, sucking in hot air from the lower cowl.

DanH said:

That's a phenomenon Don measured on the AFP flow bench. This spring loaded alternate air door caused a fuel flow increase with the primary intake blocked, although not nearly as much as the 9 GPH in your graph.

Click to expand...

I have a Precision Silverhawk RSA-5 fuel servo, not sure if that makes any difference. There is undoubtedly differences in the spring tension and the alternate air door size and shape. Also, would Don's measurements change based on parameters for an IO-540 (assuming his measured increase was based on a 4 cylinder)?

At the end of the day I cannot state for certain what caused this, but given its isolated nature (two more normal flights today) and the presence of the bag beside the airplane after it was shut down, its the only thing that makes sense to me. However it may have been something else and the presence of the bag was a coincidence. I may never know unless it happens again.

I had the Vans original alternate air door on my RV-7A. It was held closed by a magnet. I found out that under cruise or higher power it would open due to the pressure difference across the air filter and I got results as you describe - a very high fuel flow. I could cause it to happen by a slight change in throttle position and it baffled me until I realized that the door was opening when it should not have. This airplane used Airflow FI, but I can imagine Bendix would probably give the same results - the disturbed airflow through the servo caused mismetering of fuel. A spring loaded one could very well cause the same problems.

Removing that door and going to the sliding one Vans now sells completely cured the problem.

SMO said:

I believe the plastic bag was blocking the air intake and the alt air door operated as it should, sucking in hot air from the lower cowl.

Click to expand...

That's why I did mine as a spring loaded door rather than a manual control; no pilot action required to continue flight. And I think your bag theory is a good one. You could check it easy enough; do a quick runup on the ramp, record the MP, RPM and fuel flow, then shove the bag into the intake and do it again.

I have a Precision Silverhawk RSA-5 fuel servo, not sure if that makes any difference.

Click to expand...

The Silverhawk does sense pressure differently, using a low-gain venturi rather than the high gain venturi of the FM series fuel controls.

There is undoubtedly differences in the spring tension and the alternate air door size and shape.

Click to expand...

I sized the spring so the door starts to open at around 3" H2O pressure differential across the filter element (actually designed for 2", but it turned out to be 3" on the flow bench). The arm to which the spring is attached has less and less advantage as the door opens, so the force necessary to open the door does not rise much as it opens further. The long spring helps too; less tension rise with length increase. I think arranging the leverage to offset rising spring pressure is a key point to the design of a spring loaded alternate air door.

Also, would Don's measurements change based on parameters for an IO-540 (assuming his measured increase was based on a 4 cylinder)?

Click to expand...

The difference would merely be increased air flow demand. If all airbox dimensions were proportionally larger to match, the numbers would probably be the same

Break.

I looked up the measured numbers for the above airbox. At 1400 lbs per hr air flow, carb loss was 3.1" H2O for a bare FM200 with an inlet bell, 5.0" with the filtered airbox, and 6.2" with the airbox inlet completely blocked and the alternate air door floating open in equilibrium with air pressure. Fuel flows were 122 lbs per hr, 123 lbs per hr, and 130 lbs per hr respectively. You're reading that correctly; blocked inlet and alternate air door operation only increased fuel flow about 5%.

You win!

Looks good that you had the alt. air door to catch your chocked intake. Brings back old engine stories when we would put our hand over the intake and turn it over a few times to chock it before start up. Only one tip out of this old f_ _ _. You may wish to clean your spark-plugs a little early as a heavy chock or rich event will tend to dirty them up a lot faster. Thanks for post, I think I will go pull the air filter off the mower and not us the chock to start it, just hand-cup the intake and pull it through a couple of times. It is good to stay in practice. Yours as always. R.E.A. III #80888

Rupester said:

The FI regulator (aka servo) measures intake air velocity and increases fuel flow accordingly. i.e. more power demand increases measured intake air velocity and resultant fuel dispensed by the regulator. If the intake were blocked, I would think the regulator would see very low intake velocity and hence flow reduced fuel. I must be missing something here.

Click to expand...

One side of the air diaphragm is ported to dynamic pressure (the 4 little tubes in the throat of a Bendix or Silverhawk) and the other side is ported to venturi pressure. The higher the differential, the more the diaphragm moves against a fixed spring, further opening the fuel control ball valve. If I have my theory correct, a blocked filter increases fuel flow due to the increased drop at the venturi (suction, if you prefer) while dynamic pressure remains roughly the same, a function of RPM.

DanH said:

That's a phenomenon Don measured on the AFP flow bench. This spring loaded alternate air door caused a fuel flow increase with the primary intake blocked, although not nearly as much as the 9 GPH in your graph.

Click to expand...

Dan, where did you get your cowl scoop/air box set up? It looks like such a thing could be adapted to my smooth Hr-ii cowl that has the inlet just below the spinner.

vlittle said:

Dan, where did you get your cowl scoop/air box set up? It looks like such a thing could be adapted to my smooth Hr-ii cowl that has the inlet just below the spinner.

Click to expand...

It's all scratch-built, processes detailed in the fiberglass section of the forums.