fuel flow fluctuating

[mhrivnak]

I have an original IO-360-A3B6D, in service since 1990 with ~2250 hobbs hours in my M20J.

 

Since my recent annual, the fuel flow will fluctuate about 0.5gph several times per minute when running above about 20" manifold pressure. It's persistent, but not on a regular interval. All 4 EGTs follow the changes exactly, so I assume it's not an indication issue. It happens when I'm lean of peak or rich of peak. Turning on the boost pump has no effect, nor does switching tanks. It happens on the ground and in the air. Fuel pressure is stable. There are no visible fuel leaks anywhere.

 

This week, we hooked up a clear tube at the divider and confirmed that there are no air bubbles in the fuel line going into the divider, even as I ran it to 2200 RPM and saw the fluctuation happening on my engine monitor.

 

Here is a quick in-flight video, plus charts from engine monitor data showing FF before and after the annual.

 

Background:

 

On the last flight before my recent annual, I saw a cylinder start behaving like it was very lean, and I suspected a clogged injector. During annual we cleaned the injectors, and sure enough I found a small piece of debris that fell out of that injector. My A&P took the top off the divider to have a look and make sure there wasn't more debris in it; it was perfectly clean. I also checked the gascolator screen and servo screen as a normal part of the annual. That's all we did to the fuel system. We also sent out the magneto for IRAN, but I don't know how that could be related here.

 

Ideas?

 

Given that the diaphragm in the divider is 31 years old, and that this problem occurred immediately after messing with it during annual, it's tempting to start by replacing it. But: 1. the diaphragm isn't available for sale as far as I can tell, and 2. I can't think of how it or the valve in the divider could cause this behavior.

 

Is it just time to overhaul the servo and divider? I'd feel better if there was a reasonable theory on why this behavior is happening now, just after some work happened on the fuel system, especially given that it's a significant expense on an old engine. But based on my understanding of how the system works, the servo handles the regulating and is most likely to be at fault.

 

I thought about trying to get a look at the servo's air impact tubes. If some bit of debris sneaked by while I was changing the air filer, and partially obstructed the servo's ability to measure impact air, maybe that could cause this? Seems unlikely, but at least it would be an explanation. Maybe someone has a better theory?

[N201MKTurbo]

If they are all changing it must be the servo. Is there any looseness in the mixture cable? Its hard to imagine the divider doing anything at that MP. The plunger should be fully raised at that fuel flow. It would be interesting to see if the fuel pressure is leading or lagging the fuel flow changes. This would indicate weather the fuel pressure changes were causing the flow changes or reacting to the flow changes.

[mhrivnak]

1 minute ago, N201MKTurbo said:

If they are all changing it must be the servo. Is there any looseness in the mixture cable? Its hard to imagine the divider doing anything at that MP. The plunger should be fully raised at that fuel flow. It would be interesting to see if the fuel pressure is leading or lagging the fuel flow changes. This would indicate weather the fuel pressure changes were causing the flow changes or reacting to the flow changes.

I checked the mixture and throttle cables, and both are very secure with no slop.

I can't see any change in fuel pressure while this is happening, either on the display or in the recorded data. But I think it's only measuring in increments of 1psi, so there could be smaller fluctuations happening below that threshold.

[N201MKTurbo]

24 minutes ago, mhrivnak said:

I checked the mixture and throttle cables, and both are very secure with no slop.

I can't see any change in fuel pressure while this is happening, either on the display or in the recorded data. But I think it's only measuring in increments of 1psi, so there could be smaller fluctuations happening below that threshold.

When I watched your video, the fuel pressure was changing from 21 to 23 PSI while the fluctuations were happening. You didn't have FP in your graphs.

[mhrivnak]

4 minutes ago, N201MKTurbo said:

When I watched your video, the fuel pressure was changing from 21 to 23 PSI while the fluctuations were happening. You didn't have FP in your graphs.

Right, when I add FP to the graph, there is no apparent correlation. It's mostly stable at 22psi. I don't know why it moved during that short video, but I haven't been able to identify any correlation between its occasional changes and the constant FF changes. Here's an example.

[mhrivnak]

Update: I had the fuel servo and divider overhauled, but that did not resolve the issue. Airflow Performance did find corrosion in the servo and noted that it it was behaving well out of spec when they tested it. But again in the flight yesterday, the fluctuation happens lean or rich of peak, left or right tank, left or right mag, boost pump on or off, and at multiple power settings. If anything, the amount of variation seems to have increased after overhauling the servo and divider, now up to .8 gph and occasional .9.

Any other theories on what could cause this? Fuel pressure appears stable, though the JPI resolution seems to only be 1 psi. The mechanical pump is the original with 32 years and 2250 hours, and they're relatively cheap, so I may go ahead with replacing that, though I don't have a good explanation for how it could be the problem.

Here's a snapshot from yesterday's flight, running LOP in this segment.

[PT20J]

Since the EGTs rise when the flow increases while LOP, I would tend to believe it is real. One way to double check that it is not instrumentation would be to repeat the test ROP and see if the EGTs decrease when the fuel flow increases. That would tend to rule out instrumentation.

The engine-driven pump works on displacement rather than pressure and if it can supply enough fuel for full power at reasonable pressure it should be fine. I might also try running the boost pump to see if that makes a difference.

The servo meters fuel based on airflow. I would check the intake to see if anything might be affecting the airflow. You mentioned that this happened after the last annual inspection. Perhaps the air filter was replaced and is an issue. 

Skip

[mhrivnak]

Thanks for the ideas.

When ROP, EGTs do in fact decrease as expected. You can also see the RPM correlation, so I'm confident it's not an indication issue.

Running the boost pump makes no difference.

The intake path is clear with no obvious cracks or defects. I have the usual Brackett air filter, and I did replace the element at annual. I've double checked that it is seated well in the housing. Anything in particular you'd look at regarding the filter? Naively is seems hard to screw up, but maybe I found a way.

[Ragsf15e]

Guys, any chance this can go all the way back to an issue with the gascolator?  Does it seem the same no matter which tank you’ve selected?  I guess I’m thinking about large air  bubbles being pulled through.

[mhrivnak]

1 minute ago, Ragsf15e said:

Guys, any chance this can go all the way back to an issue with the gascolator?  Does it seem the same no matter which tank you’ve selected?  I guess I’m thinking about large air  bubbles being pulled through.

Before overhauling the servo, we hooked up a clear tube at the divider to look for bubbles and observed none. I ran it up to 2200rpm and saw the fluctuation happening, but my mechanic watching the tube saw no signs of air bubbles.

Indeed, the behavior is the same on either tank.

Anything else you would do to rule out the gascolator? It was of course open during the annual (found to be whistle-clean).

[PT20J]

I would hold the Brackett element up to the light and make sure it doesn't have an internal issue just to be sure. I might also try running the engine briefly in a non-dusty area without the filter just to see if it makes a difference. I would also check the alternate air door and check the intake tubes though usually a leak there causes idle issues that smooth out at high power.

Skip

[Marc Taylor]

Had a fluctuating fuel flow indication on my 252. Was at the avionics shop for software updates to the Garmins and mentioned it to the tech.  When I was ready to leave, the tech said he had fixed the problem-it was a connection issue-don’t recall if it was a sensor grounding or an intermittent connection.  Anyway, it was simple and quick.

[PT20J]

2 hours ago, Marc Taylor said:

Had a fluctuating fuel flow indication on my 252. Was at the avionics shop for software updates to the Garmins and mentioned it to the tech.  When I was ready to leave, the tech said he had fixed the problem-it was a connection issue-don’t recall if it was a sensor grounding or an intermittent connection.  Anyway, it was simple and quick.

Interesting. Did it affect your EGTs as well?

[larryb]

This is what caused my fluctuating fuel flow. That is the connector for the flowscan sensor feeding the JPI.

[carusoam]

See if @M20Doc has seen a fuel flow variation that appears to be very real… has confirming EGT match… (M20J)

What does the mixture knob look like?  Anything loose there?

The fuel servo has an MP sensor, doesn’t it? It may be sensing something incorrectly…

PP guessing while inviting the doc…

Best regards,

-a-