Fuel Flow Discussion, Moved from Float Mounting

[Battson]

Additional constraint, albeit the most obvious - the tanks must be independently vented and not interconnected.

[Battson]

Jon mentioned he had two engine failure in a Bearhawk while running on “both”. I see it did not have the standard Bearhawk fuel caps. I wonder about other details of this airplane. Like, were they vented caps, any other venting, fuel system details, FI or not, anything else different than standard Bearhawk. If any of this has been given, sorry I missed it.

Another interesting side note, all the planes which have reported this issue are scratch built, and may have modifications from the basic design which we aren't aware of.

[Battson]

You can't have this problem if the tank cap vents are working correctly.

[Archer39J]

If those vents are on both tanks so as to provide a bypass for unequal pressures it will. But let's not get too far in the weeds of detailed design, let's also stay away from absolute statements.

[whee]

I am hesitant to talk about this much more before hearing what Bob has to say; I don't want to cause and additional confusion.

I don't think it will hurt anything to state the constraints that must be met for this failure mode to occur and perhaps that will identify why we have only seen this failure a few times in the fleet.

1. Tanks must be independently vented with no cross tank vent line. ie, tank airspaces are independent from each other.

2. The fuel valve must be in the "BOTH" position.

3. The vent pressure in one tank must be sufficiently different than the vent pressure of the other tank.

4. The fuel quantity in the tanks must be low. One tank must be below approximately 4ish gallons. How low the fuller tank must be depends on how large the vent pressure differential is.

5. The above conditions must be met and remain that way until the fuel remaining in the fuel lines downstream of the fuel valve is consumed. Example; the fuel in the carburetor bowl must be consumed before the engine stops running.

There are probably few that run sufficiently low on fuel while somehow inducing a prolonged vent pressure differential. With Bob caps a leak thought the O-rings isn't likely so you'd have to misalign the cap vent or fly uncoordinated to cause a sufficient difference in vent pressure. Or fly really really low on fuel.

[whee]

The BH I was flying had a O360 and the fuel system was built according to plans. The fuel caps were Luscombe or Atlee Dodge Cub style (half turn cap with a forward facing snorkel vent). Construction on this plane began before the kits were available so the Bob caps were not commercially available and the builder didn't have the machinery necessary to build them himself.

[whee]

Maybe worth noting that the NTSB did not determine what caused the vent pressure imbalance in the Luscombe wreck. They said possible a plugged vent, which they did not say they found and evidence of, or uncoordinated flight.

[whee]

I've never heard my tanks pop.

Adding additional vents will not have any affect in regard to the failure mode which I am referring too.

[Archer39J]

For technical correctness, 14 CFR 23.975 has been superseded by 14 CFR 23.2430: https://www.law.cornell.edu/cfr/text/14/23.2430

A cross vent is no longer required by the language of the regulation. The requirements now address possible adverse effects of incorrectly designed fuel vent systems 23.2430 (a)(1), (a)(3), (a)(6), and (b)(3). (b)(5) may be of note to those of us with 2700 GTOW, but I doubt anyone will go out of their way to meet that "reg" (which of course don't apply to experimental AC). I personally like the move away from dictating specific designs by regulation, let folks figure out how to avoid the undesirable outcome rather than lock them into a solution that might not be right for everyone.

In any case, a cross vent is no longer required by the "regs".

[whee]

The TCM fuel injection system does return a significant amount of fuel through the "vapor return". As much as 50% of the pumped fuel could be returned based on power settings. The amount of fuel pumped is determined wholly by engine RPM. The amount of fuel that reaches the fuel spider is determined by throttle and mixture settings.

None of this had anything to do with the failure which we are discussing here.

[N942VT]

The Super Cub has several STCd fuel valves to add the both position. Each one I’m familiar with requires a cross vent. The simplest way is to tie them together at the top of the site gauge. These STCs required flow testing in all flight regimes. No both setting, no cross vent. Both setting requires a cross vent. Should tell us something. This is one of the areas that following the certified world makes good sense. FWIW I flew 942vt for 500 hrs always on both. Lots of hard slips into tight places, steep banking turns down the valleys etc. Never a problem but it is carbureted so add that to the equation.

[zkelley2]

Matt, how close are you to Campinas?

[Ed.Meyer]

Jon mentioned he had two engine failure in a Bearhawk while running on “both”. I see it did not have the standard Bearhawk fuel caps. I wonder about other details of this airplane. Like, were they vented caps, any other venting, fuel system details, FI or not, anything else different than standard Bearhawk. If any of this has been given, sorry I missed it.

[Archer39J]

Running some numbers I've reached the same conclusions Ed. Of course we're all just speculating, hopefully Matt will continue to update us as more becomes known. But I'm still unsure how you can get, with correctly vented and oriented caps, a pressure difference between the tanks.

I also recall some folks talking about how their tanks "pop" with changes in altitude, and that it's normal/expected. I would say that tanks popping in flight are a sign of inadequate venting.

I'm looking hard at, if not putting in a cross vent, at least an extra vent (per tank) with its pickup on the top-outboard-forward corner of the tank.

[Ed.Meyer]

Like several others on this thread, I am struggling some with this. Seems like there is a lot of theory given here and I can appreciate the theory stuff and in a situation like this, facts may be hard to prove. I do, like others, appreciate the discussion which I have re-read, some of it multiple times. Some stuff running through my mind:

First, I cannot see how a pressure imbalance can occur with two properly installed vented fuel caps, which Matt said was the case with his failure. With the two fuel caps only a few feet apart and moving at the same speed they would be applying equal pressure (pitot pressure) to vent both tanks. If one was installed backwards then I could see unequal pressure which would be less than one PSI. At 150 kts, pitot pressure is about 15 inches H2O or about.5 psi. The one benefit that I can see with an interconnected vent would be venting redundancy. This would come with the potential cross feed problem mentioned. I wonder if something else might have caused Matts failure. Also, I see in the photos that the airplane came to rest with the left wing low. Could some fuel have flowed through the fuel valve from the right to the left tank after the accident?

I am starting to grasp how differential tank pressure might cause fuel starvation such as the Luscombe accident Jon cited. Being a gravity fed system, the head pressure feeding the carburetor is very small with only the weight of the fuel in the system above the carburetor supplying the pressure. If one tank had low pressure, thus sucking fuel from the other, the suction might be greater than the head pressure at the carburetor causing flow into the carburetor to stop. If a single tank is selected and it is not vented it would also quit feeding fuel. It would seem that a system that includes a fuel pump would be less prone to this potential problem since the sucking from a low pressure tank would have to be greater than the gravity head pressure plus whatever suction the pump provides. Any time there is suction, especially when temperatures are warm, there is potential for vapor lock causing pump to stop being fed (which I have experienced in a low wing Cherokee 140 on a hot day with auto fuel, at least that is what I think happened. No way to prove since engine ran fine after landing.)

I will certainly keep following this to determine if venting modification might be needed on our Patrol. In 150 hours so far running on both most of the time there has not been any problem. I don't like running it very low on fuel though...