Adam,

 

Nice recovery.

 

Your engine monitoring system should have given you an alert when the fuel pressure went to zero or fuel flow dropped below some minimum.  Some injected Lycomings require a minimum of 12 psi.  Continental 6-cyl may have a higher requirement.

 

Most fuel system quantity sensors (even spam cans) are calibrated with the fuselage level and are not useful indicators in non-level pitch conditions.

 

Some slosh compartment's doors leak and continued non level pitch attitude can cause loss of fuel from the compartment (you do have a slosh compartment?).

 

During the build when it is easy, almost everyone fails to test "usable fuel" quantities in wing-main-tank systems.  That is, the wing can be tilted up and down in pitch whilst measuring how much fuel it takes to keep the pickup covered.  It seems that usable fuel in a nose down pitch is  wing capacity less 10 or more gallons in each wing.  Hard to tell what the pitch reading means - was that merely down by 5 or 6 degrees? - only for 2 minutes?  The high fuel reading meant the fuel had gone up higher on the probe - was this an uncoordinated descent?  Where did the left wing fuel hide during the pitch over or was that in a steep bank?

 

So many questions........

 

Scott Krueger

 

Yesterday I had a brief loss of engine power. I'm 26 hours into my Phase I testing and was doing fuel flow/airspeed tests at various altitudes while playing around with the leaning function on my new Dynon SkyView. The attached 'full flight log' file shows the entire 1.7 hour flight. I was on the right tank for the first 30 minutes, then switched to the left tank for the next hour. You can see this in the third graph down from the top, which shows the 'Fuel R' level dropping steadily then leveling off when I switch tanks. 

I have EI fuel probes with output a frequency proportional to fuel level. They feed Princeton frequency-to-voltage converters which in turn talk to the EFIS. The f-to-v converters have a considerable amount of damping. When combined with the filtering in the EFIS' fuel readout, we determined that it takes nearly 2 minutes for a fuel level change to register on the display. As we'll see, that's good for long flights, not so good when maneuvering. 

The second attachment, 'flight detail' zooms in on the last 15 minutes of the flight. It starts with a calibration of the Dynon AOA probe. That requires a series of pitch oscillations followed by a full stall. Once that was finished it was time to head home. I was at 10,000', 9 miles north of my home field, which is at 81' MSL. I turned towards the airport and pointed the nose downhill. With the throttle set to 10" MAP and the prop lever most of the way back I was screaming downhill at 220 KIAS and 2000 RPM. What a blast! I'll be able to enter the pattern at the correct altitude and the prop isn't driving the engine. What could possibly go wrong?

When I shallowed out the descent, the engine gave a brief increase in power. I hadn't changed any control settings, so that meant it wasn't running prior to that. I hadn't detected the power loss earlier because the engine was configured to provide essentially zero thrust. I opened the throttle to test the engine and got no response. You can see the manifold pressure briefly go to 25" at the 45 minute mark when I add throttle. That's when I noticed the fuel pressure was zero. I then tried increasing the prop RPM but all that did was add braking action. I put the prop lever back to its original position, switched tanks and turned on low boost. The engine started running again. I was now 3 miles north of the airport so I entered the pattern and landed normally. 

Once I landed and allowed the fuel levels to stabilize, I showed 10 gallons in the left tank and 16 in the right (which agreed with my totalizer). Right now I'm thinking that the left fuel pickup unported during the descent. I surmise that the fuel pickups are positioned to stay immersed during high angle climbs, but if the fuel sloshes forward all bets are off.  

You can also see that the the fuel level indications are all over the place. It appears that the reading is sensitive to pitch angle and increases dramatically during descent. That combined with the 2 minute lag reduces the usability of these gauges. 

Thoughts?

Thanks,

Adam Molny

--
For archives and unsub http://mail.lancaironline.net:81/lists/lml/List.html