Possible fuel starvation?

[Chuck]

A question for the hive-mind:

 

I was recently on a long cross-country (NH to NC) and was intending to do it without a fuel stop.  I had done the math and figured I should land with about 6 gallons still in the tanks and had been carefully watching the fuel totalizer on the Dynon and the fuel in the sight tubes.  Towards the end of the flight I was making sure to fly slightly uncoordinated at times to move a little fuel from left to right to try and keep the tanks balanced.

 

With about 30 minutes to go I felt the engine give a slight 'hiccup' - just a split second hesitation before returning to full power.  I sat up straight and looked at everything and all looked good.  I was still showing about 4-5 gallons per side according to the sight tubes, and the totalizer was telling me what I expected.  I buzzed along for about another 10 minutes and it did it again.  At that point I was right near the airport in Galax, VA, so I just landed and put 10 gallons in and gave things a look on the ground.  I didn't see anything amiss and so took off and flew the last 20 minutes to my destination.  I've put about 8 hours on the plane since then and haven't had it happen again.  

 

That's the lowest I've ever had the fuel in the plane and am just curious if it's possible that a little air got into the lines or something because the fuel was so low?  There was a little turbulence so fuel was definitely sloshing around in the tanks.  Does the fuel to the engine feed out of both tanks or just the right one?

 

Thanks in advance for any thoughts - I want to fine tune my flight planning mindset...

 

-Chuck

[Anticept]

Don't try to "balance tanks". Just fly coordinated and it will draw the fuel as needed, preferring the higher level tank if you're truely coordinated.

The other thing I find is letting BOTH tanks get below 5 gallons, you can start getting quirks (Mine doesn't like it when i get below 4 gallons). If one tank is >5 gal it's fine.

There's not enough head pressure to deal with all the serpentine routed lines if you let it get too low, and at that point the engine fuel system has to drop in pressure to pull more. Wasn't an issue with the old pierburg pumps as those things were champions, but the corona pumps don't deal with vapor very well and that's what you can expect when the pressure drops.

[Mike Koerner]

Though I’m loath to disagree with anything Corey says, in this case I feel so moved.

Most of us don’t fly perfectly coordinated. Over the course of a long flight you may find the tanks have not drained evenly. Purposely flying uncoordinated for a few minutes - “half a ball” out with the heavy wing high - is an appropriate means of correcting this situation relatively quickly.

Yes, you could just fly perfectly coordinated from then on and this would also correct the imbalance, over time. The problem is this will take so long you may forget what you’re trying to do and revert to the uncoordinated flight that created the condition in the first place.

It’s not important that the tanks be balanced as long as there is fuel showing in both sight tubes during coordinated flight. However, if one tank threatens to empty, transferring fuel to avoid that condition is appropriate.

Also, as long as there is fuel showing in both tubes, there is sufficient head to maintain proper engine operation.

[Anticept]

I've never seen the fuel move at any appreciable rate, at least in a CTLS. We parked a CT wing high a few degrees once and after an hour, only a couple gallons moved.

We have rudder trim, for the most part that's been good enough if it bothers you that a wing is low. Use that to continue to fly coordinated, make some adjustment but don't start flying sideways.

I would argue that especially if you're low on fuel and trying to balance them, you're at greater risk of unporting fuel. Inside of the tanks, there is a coarse mesh that picks up the fuel in many aircraft, and it's not at the edge of the tank. You can actually trap fuel up against the inboard side of the wing and fuel won't have a way to enter, and due to the sight gauge's construction, it too can indicate higher than what you actually have when not flying coordinated.

Given the finickyness of our CT fuel systems, one wing full and one wing empty is far more valuable than two equal tanks due to the higher head pressure to displace any air that gets into the system.

If you're reading off a dynon unit, those balls are EXTREMELY sensitive. If it's compared to a whiskey filled inclinometer, a full ball deflection on a dynon unit barely registers on the old style inclinometer! I hung an 8 inch string in the cockpit with an M8 nut it barely moved from center while the dynon unit was quite far out of the cage (testing something unrelated).

The only time I have ever seen fuel move appreciably is when the fuel cap disappeared. It drew all the fuel to that wing with the missing fuel cap. Swapped the cap to the other wing and when we landed, most had shifted back.

Anyways, outside of flying sideways, usually what I see is a CT draining more from one wing or another until there's a few gallons difference, then they draw equally. Trying to fine tune it ends up being a lot of extra work in my experience.

[Mike Koerner]

Corey,

We've had different experiences... with different aircraft.

My CT2k doesn’t have rudder or aileron trim, just elevator.

I don’t rebalance fuel often. It’s only at the end of a long flight, for example when one sight tube shows 2 or 3 gallons and the other shows 5 or 6. My personal minimum is 5 gallons total with fuel showing on both sides during coordinated flight. In a few minutes of fuel rebalancing, I can move a gallon of fuel, gaining some additional flight time and ensuring that I don’t un-port either side while maneuvering for landing.

I don’t worry about un-porting a wing during the transfer. As Ed has pointed out, as long as there is fuel against the root of the high wing, it'll should get to the carbs.

The “half ball” I’m referring to is on a mechanical turn and bank.

With our nearly flat wings the difference in head between a full fuel tank and an almost empty one is only a few inches; vs. 30 or so from the carbs up to the wing.

Fuel transfer works for me anyway, but perhaps I shouldn’t recommend it to others.

[Anticept]

None the less, it's an interesting comparison.

Besides that I think the biggest takeaway is don't run it close to empty

[Jim Meade]

2 hours ago, Anticept said:

None the less, it's an interesting comparison.

Besides that I think the biggest takeaway is don't run it close to empty

You don't give a very specific reason to not run it close to empty.  If there is gas available, the engine should run normally.  What difference does it make if there is a pint or 30 gallons?

[Anticept]

Head pressure and air/vapor locks.

The amount of bends and turns that our fuel lines take, it traps vapor. If you somehow get vapor in the line and you don't have much fuel, the head pressure may not be enough to dislodge the air. Flow will be reduced or stopped all-together, and you'll have to stand the bird on its side to get it flowing again. Not a pleasant experience if your engine is already having issues.

I can't get my CTLS to flow fuel if it's below 4 gallons in a dry system and an open gascolator. Tested with 2 other CTLS, same issue. The air traps in CTSW are even larger. Granted, a little air is a lot different than a dry system, but if you couple these issues with the fact that the corona pumps struggle a little with purging air, you should understand my concerns.

The reason I believe it to be an issue with the plumbing, is if I disconnect the line at the A-pillar, flow starts with barely anything in the tank. Yet I reiterate, nothing if the gascolator is the open point.

This is part of the reason that I will be installing a fuel pump in mine, to put an end to some of this silliness.

This video goes over all the various issues of air lock, pressure, energy loss with air in a pipe, and how it all adds up and can even stop flow all together.

 

 

[FlyingMonkey]

My recent thread on power loss during takeoff turned out to be a vapor issue.  I did three things to resolve it:

1) New fuel pump -- my previous one was almost 8 years old and probably quite weak.  Rotax likes them to be changed every 5 years.

2) Trimmed the fuel pump vent line to be an inch or two up from the bottom of the firewall, to ensure no low pressure air causing vacuum condition to the pump.

3) Changed out a fuel line behind the panel that had a rise in it (pressure trap), and made it a direct route to the fuel valve via a 120° fitting.

I had three power loss episodes in two hours of flight, all in climb out where head pressure is minimized due to the pump position relative to the fuel tanks.  After taking the above actions, there have been zero repeat incidents in about ten hours.  In my case (3) above was self-inflicted, because I changed out all my hoses with Teflon "lifetime" hoses and the routing changed behind the instrument panel compared to factory routing.  Before changing out those hoses I had no incidents of vapor lock in 900+ hours or flying...so routing matters!  

[FredG]

For many years I had a mechanical inclinometer (old school, curved glass tube and solid ball ball) mounted to the horizontal bar above the panel in my CT.  The deflection of the ball was nearly identical to that of the "ball" on the Dynon D100.  When I built my RANS S20, the only mechanical gage I installed on the panel was a standard mechanical inclinometer.  It also agrees very closely with the "ball" displayed on the Dynon Skyview PFD.  So, my experience is a bit different than what Corey described.  

Regarding head pressure, the determinants of head pressure are the height of the column of liquid and the density of that liquid.  The actual volume (or total weight) of the liquid is irrelevant.  Since the tanks are pretty flat and wide, head pressure doesn't change a whole lot as fuel is burned.  

[Anticept]

There's one important thing to also point out: When air is underneath a column in a small enough tube, it also acts as a blockage because it's very difficult for the two to trade places.

Anyways, head pressure is the only thing that I can come up with here. I get that the vertical height of the column of fluid is the only thing that matters and won't change much, but that's the only thing I've found to fit the observations:

1) the flow will not start until sufficient fuel is in the tank. As soon as it does, all of it will drain out (minus a little puddle that sits under the fuel tank outlet acting as a sump).

2) to start the flow at the gascolator with a dry system requires 4+ gallons before flow begins.

3) to start flow from the bottom of the a pillar is almost immediate with very little fuel requiring to be added. I can even disconnect the hose when before there was no fuel flow, and as soon as it is pulled away, the fuel will rapidly begin to flow down the a pillar tube and make a mess if i don't re-fit it timely.

4) I can get the aforementioned less than 4 gallons to begin flowing if I lift the wing high enough.

5) full tanks show significantly less issues with fuel pressure complaints than nearly empty ones. I found this out during some test flying while trying to locate the source of these vapor issues during hot weather.

This is observed again with no obstructions that I could find in the system.

Finally, the possibility is very real that when originally designed, flight design's engineers chose tubing sizes that allowed enough flow but did not really consider vapor issues. Since the old pierburg pump was an absolute champ and had no problem with vapor and keeping pressure while vapor is in the system, I suspect nobody ran across these issues.

When I replaced my engine (and consequently, it came with the new fuel pump), we've had issues ever since. It's even in a rotax service bulletin saying that the pump does have issues with vapor and it does eventually purge it, but it shouldn't take more than a few seconds to stabilize.

[Madhatter]

9 hours ago, FlyingMonkey said:

My recent thread on power loss during takeoff turned out to be a vapor issue.  I did three things to resolve it:

1) New fuel pump -- my previous one was almost 8 years old and probably quite weak.  Rotax likes them to be changed every 5 years.

2) Trimmed the fuel pump vent line to be an inch or two up from the bottom of the firewall, to ensure no low pressure air causing vacuum condition to the pump.

3) Changed out a fuel line behind the panel that had a rise in it (pressure trap), and made it a direct route to the fuel valve via a 120° fitting.

I had three power loss episodes in two hours of flight, all in climb out where head pressure is minimized due to the pump position relative to the fuel tanks.  After taking the above actions, there have been zero repeat incidents in about ten hours.  In my case (3) above was self-inflicted, because I changed out all my hoses with Teflon "lifetime" hoses and the routing changed behind the instrument panel compared to factory routing.  Before changing out those hoses I had no incidents of vapor lock in 900+ hours or flying...so routing matters!  

I don't believe the rise in the fuel hose is a factor. Many aircraft have a rise in fuel lines but have a fuel drain at the bottom of the low spot for water. My guess is your issue was the fuel pump or some other restriction. 

One thing I noticed when changing my hoses was the fuel valve was incorrectly rigged and not fully opening by about 1/3.

[FlyingMonkey]

1 hour ago, Madhatter said:

I don't believe the rise in the fuel hose is a factor. Many aircraft have a rise in fuel lines but have a fuel drain at the bottom of the low spot for water. My guess is your issue was the fuel pump or some other restriction. 

One thing I noticed when changing my hoses was the fuel valve was incorrectly rigged and not fully opening by about 1/3.

Thanks.  I did all the potential fixes at once before test flying, so it's hard to say which one or combination was the key. 

I was careful when I re-rigged my fuel valve to make sure the lever arm gave full travel to the valve, I had to adjust the rod end a half-turn or so.

[Jim Meade]

Sometimes it's better to say "I don't know".

[FlyingMonkey]

7 hours ago, Jim Meade said:

Sometimes it's better to say "I don't know".

I get to say that a lot!  

[Tom Baker]

Corey, there is a difference between the static drain test and when flying. When flying there is a little pressure inside the tank to help the fuel flow. The angled hoods on the CTSW fuel vents and the little NACA ducts on the winglets for the CTLS provides a little bit of pressure in the tank while in flight.

[Anticept]

I did acknowledge that, but there is another observation:

During flight with low tanks, i had many more occurrences of fuel pressure and flow warnings when I only had 5 gallons on board when going around the pattern. It was going off often.

After filling the tanks full and flying again half an hour later, no alarms except a brief one on takeoff after suspecting head pressure.

This was all during my attempts to get to the source of this irritating fuel pressure issue.

The last thing to test after conversion to experimental is to check pressure going to the inlet side of the fuel pump and see what happens with the installation of a boost pump.

[Madhatter]

When I was replacing the sight tube plate gaskets I noticed that even with the inlet screens touching the bottom of the tank the actual inlet is higher. I would guess with 2 1/2 gallons per side a lot of air would get into the fuel line. In all the years I have been flying I have seen dozens of crashes due to fuel mismanagement. When I get to 1/4 fuel load I am looking for an airport, just me.  Lots of fuel is good insurance,  too many have died needlessly.  There are enough other issues to worry about in aircraft.

[Anticept]

The Only Time an Aircraft Has Too Much Fuel On Board Is When It Is On Fire.

I guess in response to what madhatter added: I don't see a point in flying with almost dry tanks, and I don't see a point in moving fuel around especially considering the various observations I have made trying to chase down fuel system issues.

I also admit there is one other avenue I have not tried yet, and that is replacing some of the small tubing with larger stuff.

PS: there is one part of the fuel system that is ESPECIALLY bad, and that's the long vertical hose after the fuel valve and fuel flow sensor and turns down to the gascolator. That's a long vertical run, which is the worst kind when dealing with air in a system. I think I mentioned before in another thread that a tiny air line tap right there going back to the tank would make a lot of difference. But so would a boost pump.

[Madhatter]

18 minutes ago, Anticept said:

The Only Time an Aircraft Has Too Much Fuel On Board Is When It Is On Fire.

I guess in response to what madhatter added: I don't see a point in flying with almost dry tanks, and I don't see a point in moving fuel around especially considering the various observations I have made trying to chase down fuel system issues.

I also admit there is one other avenue I have not tried yet, and that is replacing some of the small tubing with larger stuff.

PS: there is one part of the fuel system that is ESPECIALLY bad, and that's the long vertical hose after the fuel valve and fuel flow sensor and turns down to the gascolator. That's a long vertical run, which is the worst kind when dealing with air in a system. I think I mentioned before in another thread that a tiny air line tap right there going back to the tank would make a lot of difference. But so would a boost pump.

All my fuel lines aft of the fuel pump are now 3/8". I was surprised on how small the fuel valve ball diameter was when I replaced it, quite a restriction especially when you have small diameter hoses which cause significant line friction for reduced flow. Fluid flow charts easily show a significant flow difference from 1/4" or 5/16" to 3/8".

[Anticept]

You know what? I am betting that valve along with the vertical drop are where the fuel flow disturbances happen.

Did you see a change in takeoff fuel pressure alarms? I don't know your aircraft configuration and I'm sure you will clarify in a moment

Another thing to add to my list of experiments... see what happens with the small valve and a few gallons of gas, and then what happens with a bigger valve and a few gallons of gas.

[Madhatter]

1 hour ago, Anticept said:

You know what? I am betting that valve along with the vertical drop are where the fuel flow disturbances happen.

Did you see a change in takeoff fuel pressure alarms? I don't know your aircraft configuration and I'm sure you will clarify in a moment

Another thing to add to my list of experiments... see what happens with the small valve and a few gallons of gas, and then what happens with a bigger valve and a few gallons of gas.

I had no change. I also removed the small filter from the fuel system. My main reason for changing the hoses is to go to AN fittings which are much better than the original setup. 

[FredG]

The difference in pressure at the inlet to a hose compared to the outlet of a hose (the pressure equivalent of voltage drop across a resistor) is proportional to the fourth power of the radius of the hose.  So, if the radius is halved, the pressure drop in that length of hose increases sixteen times.  A 20% reduction in hose diameter results in a 2.4 fold reduction in pressure (which is the same as 2.4 fold increase in resistance to flow).  Small changes in hose diameter have very large effects on the resistance to flow in that hose (and, very large decreases in flow).

[Jim Meade]

What diameter hose is necessary to flow lets say 30 liters per hour (roughly 8 gal per hour - much more than I ever see on my fuel flow gauge)?

I see a flow rate calculator but it's over my head.

[Anticept]

Hose length also matters for that question, Jim. Length of hose and diameter are enough to estimate straight runs, but it's more complicated if you're trying to calculate minimums for real world installations.