Fuel Flow Discussion, Moved from Float Mounting

[Archer39J]

Sure the engine isn't gravity fed but the pumps are, verses pumps that are sucking fuel. The pressure differential in the tanks isn't more than the pressure head at the fuel selector. There's no difference between a gravity fed pump and a gravity fed engine at that point, and one clogged vent shuts off that tank after a while. I don't know what would happen with a pump that is sucking fuel, but I think it could only be worse.

[svyolo]

Cessna fuel systems are pretty Googleable. It is amazing how they evolved over the years. They didn't get simpler. I guess over the years the fleet got so many hours they kept finding a new corner of the envelope where the system didn't quite work.

I think I do want some additional venting. Right now I think I will just go with the cross vent in the middle, maybe with a shutoff valve. Mostly because it is easy to implement. I am not sure it is the best. I really like the Cessna 182S venting, but at this point it looks too hard to do. Maybe next time.

[whee]

I think I fixed the pic.

Theses "snorkel" style caps use either a cork or rubber gasket and their function is indeed to provide positive pressure inside the tanks.

I'd offer to induce the failure in my own airplane and video record it but since I have Continental fuel injection it would be of little value to anyone and the failure would probable get attributed an incompatibility with my FI. This summer I'll be testing an LSA. If I get feeling brave maybe I'll induce the failure in it and record it. But it is scratch built...

[Battson]

The caps were ram air pressurized. Same function as the Bob caps just a different form. The fuel system was built exactly as Bob specified except Bob caps were not available so a conventional cap was used.

Picture seems to have failed?

Did they also have the o-ring seal to ensure the tank could develop positive pressure? As the air rushing over the tank could theoretically develop suction, if the vent blocked.

This is all getting quite theoretical and esoteric, it makes it hard to discuss!

[whee]

I understand your machine had some other kind of vented cap, where the vent is not pressurised by ram air. It was unwritten, but I have assumed that unpressurised fuel cap must have played some part of the cause of the issues with your last plane. Set me straight if you see fit.

I think we've established that with a gravity system conforming with Bob Barrow's design and with his design of fuel cap, the risk is negligible. As ever, it appears Bob is exactly right, let's be clear on that. We have no evidence of failure in a system entirely as Bob intended it. Bearhawk Aircraft has always been very clear that we diverge from Bob's approved design at our own risk. So here we are.

With a conventional design, gravity feed system, it appears the chances of a failure seem almost impossible when you run the numbers. The engine needs so little fuel pressure to run and one of the tanks is always under some pressure, which reduces the risk further.

In post #81, I drew the conclusion that fuel injection means you need to install this vent. I think that was jumping two steps ahead when I meant to just take one. I think all we can say for certain is, some fuel injection systems are at-risk of this failure, the high pressure ones. We actually have no evidence to suggest Bendix systems are at risk.

Similarly, I didn't say that ONLY fuel injection systems are at risk. There are a million ways to design a dangerous fuel system, and only a small number of ways to make an inherently safe system.

The caps were ram air pressurized. Same function as the Bob caps just a different form. The fuel system was built exactly as Bob specified except Bob caps were not available so a conventional cap was used. See pic of the conventional cap style cap.



U286116__92249.1368114694.1280.1280.jpeg

[Archer39J]

The check valve idea is interesting, and I am especially tempted because the cross vent would need a check valve anyway.
However, I am not sure whether I would trust a check valve in my main fuel line - seems to introduce another potential engine-stopping failure point.
In the vent line, a failure of the check valve is less risky.

Yeah I was thinking a check valve in a separate vent line that terminates inside the wing with a weep hole just in case, I described it in a comment a bit ago. That would solve the clogged cap issue but the commercially available ones I've found have opening pressures slightly too high for a reversed vent case so I'm considering making my own.

Here's the one I was looking at https://www.mcmaster.com/5492k51

ETA: I think the risk for my system is low enough for the reverse vent case though, fleet history being what it is in addition to my analysis. My main concern is icing since two clogged vents will take you out of the sky.

[Battson]

It affects both gravity feed carburetor systems and fuel injection systems. Matt’s failure was with Continental Fuel injection. The two times I experienced the failure it was in a Lycoming O360 powered plane. Bendix fuel injection might actually be more tolerant of this issue because of the types of pumps used.

I understand your machine had some other kind of vented cap, where the vent is not pressurised by ram air. It was unwritten, but I have assumed that unpressurised fuel cap must have played some part of the cause of the issues with your last plane. Set me straight if you see fit.

I think we've established that with a gravity system conforming with Bob Barrow's design and with his design of fuel cap, the risk is negligible. As ever, it appears Bob is exactly right, let's be clear on that. We have no evidence of failure in a system entirely as Bob intended it. Bearhawk Aircraft has always been very clear that we diverge from Bob's approved design at our own risk. So here we are.

With a conventional design, gravity feed system, it appears the chances of a failure seem almost impossible when you run the numbers. The engine needs so little fuel pressure to run and one of the tanks is always under some pressure, which reduces the risk further.

In post #81, I drew the conclusion that fuel injection means you need to install this vent. I think that was jumping two steps ahead when I meant to just take one. I think all we can say for certain is, some fuel injection systems are at-risk of this failure, the high pressure ones. We actually have no evidence to suggest Bendix systems are at risk.

Similarly, I didn't say that ONLY fuel injection systems are at risk. There are a million ways to design a dangerous fuel system, and only a small number of ways to make an inherently safe system.

[Battson]

My solution is a check valve on each tank, ones that open at a pressure well below the pressure head of the fuel. Air gets in if your cap vents get clogged, doesn't let fuel out, doesn't remove your ability to isolate tanks on a slope, and won't let fuel settle in a low point for those wanting to go outboard side to outboard side with their cross vent (dihedral isn't shown in any of those fuel diagrams).

Not 100% on this solution but it's where I'm leaning.
​​​​​​

The check valve idea is interesting, and I am especially tempted because the cross vent would need a check valve anyway.
However, I am not sure whether I would trust a check valve in my main fuel line - seems to introduce another potential engine-stopping failure point.
In the vent line, a failure of the check valve is less risky.

[Battson]

I understand you are not a 'gravity feed' fuel system if you need pumps to keep the engine running. You are pumped. In that case, you run the risk of losing all fuel supply to the engine, if one vent is blocked.

[Archer39J]

Correct, I only mention sucking pumps because I haven't looked into that at all, since it's not what I'm running. You can equalize the pressures by having them vent to the same open space.

[whee]

There is still confusion on this topic but I don’t know how to resolve it.

The issue has nothing to do with a pump sucking air. It affects both gravity feed carburetor systems and fuel injection systems. Matt’s failure was with Continental Fuel injection. The two times I experienced the failure it was in a Lycoming O360 powered plane. Bendix fuel injection might actually be more tolerant of this issue because of the types of pumps used.

The issue is having the higher air pressure in the airspace of one tanks vs the other. IDK how you equalize the pressure in the airspace of each tank without connecting them.

Putting a check valve in the cross vent line, like pretty much all certified high wing planes do, easily solves the parking on a slope fuel transfer issue. Worth nothing that most fuel valves connect the tanks together when in the off position so you need to put the valve in either right or left when parked on a slope to prevent fuel transfer even if you don’t have a cross vent. Or verify that you valve maintains tank isolation when in the off position. My SPRL valve connects the tanks when placed in the off position.

[Archer39J]

You mean like carrying two complete independent ignition systems, stabilizer struts as well as flying wires, tires larger than really necessary, a backup battery for electric ignition?
Sometimes you need an backup plan.

Clearly this is a real risk, we have two firsthand accounts right here. So anyone with fuel injection should be worried about differential pressures - especially high pressure systems.

If a mason bee lodging itself in a tank vent can bring the whole plane down, then I'll be installing a backup system.

What are the "simpler ways around that" - I am keen to hear what other options I have, which are simpler than a bit of tube between tanks.

The designer who made one of those decisions also said a cross vent isn't necessary. The risk is there fuel injection or not. But I'm sticking with gravity feed to the pumps so my solution would only be applicable to a otherwise gravity fed system, if you're sucking at the pump good luck and I don't see a cross vent affecting that anyway off the top of my head.

My solution is a check valve on each tank, ones that open at a pressure well below the pressure head of the fuel. Air gets in if your cap vents get clogged, doesn't let fuel out, doesn't remove your ability to isolate tanks on a slope, and won't let fuel settle in a low point for those wanting to go outboard side to outboard side with their cross vent (dihedral isn't shown in any of those fuel diagrams).

Not 100% on this solution but it's where I'm leaning.
​​​​​​

[Archer39J]

So I think the conclusion is, if you have fuel injection - you NEED to install these vents to be on the safe side, or avoid unbalanced flight if you have a low fuel tank.

I wish I knew this before the plane was assembled! Going to be a difficult mod to get right without easy access.

That said, I slip all the time with low fuel, I have no cross vent and never had a problem with the Bendix system... because I never had a blocked vent. If the vent blocks, then I run the risk.

"To be on the safe side" is a great way to add unnecessary weight to an airplane.

​​​​​​Fuel injection doesn't inherently necessitate a cross vent. If you're worried about differential pressures there are simpler ways around that.

[Bcone1381]

Here is a collection of fuel tank diagrams of certified aircraft I found on-line that I thought may help visualize this discussion.

Below is (with a 95% confidence feeling) a C-152. Lycoming O-235 engine. The C-152 merely has a shutoff valve, so its always on BOTH. One vented cap + a cross vent + a tank vent with a check valve.
Screen Shot 2020-02-06 at 11.00.26 AM.png

Below is a Cessna 174N with one vented cap. The C-172 N has a carbureted Lycoming O-320 engine.

I note two things here, the cross vented tanks and the vent line between the fuel supply line and the cross vent. I see this on all the later model Cessna's, C-172's, 180's, 182's, If parked on a left leaning slope, some fuel will drain from the right tank. If the vent line was run back to the right outboard of the right tank it would not drain out. What happens if the check valve fails closed? The R tank's vented filler cap should relieve a pressure differential.

Screen Shot 2020-02-06 at 10.47.39 AM.png


Cessna 172S. This is the latest model of C-172 that has a Lycoming IO-360 engine instead of a carburated O-320. Cessna designed the system with a header tank (fuel reservoir tank). Note how the header tank is vented. I don't see vented fuel caps on this application. What happens if the Left tanks vent's check valve fails closed? I think negative pressure would build up and the engine would fail with fuel in the tanks. I question if the schematic is accurate by showing the unvented fuel caps.

Screen Shot 2020-02-06 at 11.06.18 AM.png

Cessna 177 Cardinal with header tank. Cardinals were made with both carberated and injected Lycoming 360 CI engines. I'm not sure which one this set up was designed for. No check valve. No vented caps. Will fuel drain onto the ground if parked on a slope? I don't think so.

Screen Shot 2020-02-06 at 10.55.10 AM.png


Cessna 182S with a 230 hp Fuel Injected Lycoming IO-540 engine. Note fuel venting is all over the place. It has two vented fuel caps, a cross over vent, fuel supply lines vented, and a vent in each tank. This fuel system does NOT have a header tank.


What conclusions should we draw?

[Archer39J]

Then you have to think about how to purge fuel that gets into the low spots or you'll not be equalizing the pressure between tanks.