You omitted the "or" at the end of (1). Gravity is the "means to prevent introducing air into the [fuel] system" of an aircraft with the tanks higher than the fuel selector/EDP etc. That works, even if the fuel selector is able to select more than one of those tanks, simultaneously.

All fuel systems will eventually introduce air into the system, through exhaustion or starvation. That's not the point of 23.951.

The point of 23.951 is system design. An aircraft with the fuel tanks lower than a fuel selector that can select more than one of those tanks simultaneously is not designed in accordance with 23.951, unless there's some further component to overcome the laws of physics demonstrated in the video I posted and which 43 explained. Given the ATSB report I posted earlier, it seems to me the Commander 114 has a basic design flaw in the fuel system. I'd never fly one with the fuel selector in "BOTH".

How about the humble Chipmunk? Two tanks but (in its standard form) a simple ON/OFF selector.

I'll make a wild guess and say that the 'humble Chipmunk' wasn't designed and certified under the FARs (at least not the current FARs), but there may in any event be components in the fuel system that mitigate the risk of air being drawn into the fuel system due to one one tank emptying before the other.

What happens if a Chipmunk has one tank full and the other one empty, and is started and flown with the "simple ON/OFF selector" in the "ON" position?

The pilot manual says “if one tank empties before the other the latter will supply the engine provided the fuel cock is ON”. There are no warnings in relation to the fuel system.

Then...

It seems to me that there must be components in the fuel system on a Chipmunk that mitigate the risk of air being drawn into the fuel system due to one tank emptying before the other.

Either that or the design somehow manages to position the fuel selector and inlet to the fuel components of the engine below the outlets from the tanks.

Rather than endless, unproductive PPRuNe speculation, the answer can be given by someone with first-hand knowledge of the fuel system 'plumbing' and components on a Chipmunk.

Looking at the fuel system schematics on the Chipmunk they do seem to have solved this by having the fuel selector position lower than the tanks. So there is still a gravity feed to where the lines join and the selector, from the sump outlet, as well as built in non return valves in the tank outlets.

I think the fuel selector of the PA-28-180 is also below the tank outlets in level flight but it certainly is not at the highest sustainable pitch attitude.

I have found the discussion interesting and informative and wondered what addition guidance had been written by FAA. I found AC 23-16A which has some discussion of the subject in section "23.951 Fuel System General (Amendment 23-43)" which starts on page 43.

It includes the following -

"(Many in-service airplanes were certified under the Civil Air Regulations (CAR). For background, CAR § 3.430, which preceded § 23.951(b), required the fuel system arrangement to “permit any one fuel pump to draw fuel from only one tank at a time.” In the past, this has sometimes been interpreted (incorrectly) as “a fuel pump may only draw fuel from one tank at a time.” The CAR regulation is actually more permissive than § 23.951(b) as it did not address the possibility of introducing air into the fuel system.)"

ref https://www.faa.gov/documentLibrary/.../AC_23-16A.pdf

edit to add text of CAR 3.340 -

"§ 3.430 Fuel system arrangement. Fuel systems shall be so arranged as to permit any one fuel pump to draw fuel from only one tank at a time. Gravity feed systems shall not supply fuel to any one engine from more than one tank at a time unless the tank air spaces are interconnected in such a manner as to assure that all interconnected tanks will feed equally. (See also § 3.439.)"

FWIW, here's the pic of the fuel system for the Chipmunk: They speak of a 'flexible fuel tank'.. presumably a bladder tank which may reduce the amount of air available to get into the system?

https://cimg8.ibsrv.net/gimg/pprune....69ae081709.png

How I wish they drew detailed and clear diagrams like these in the AFM for the PA-28's.. but that's progress I suppose :(

Interesting. It may just be the perspective of the drawing, but as 43 noted it appears the fuel cock where the outlets of the tanks meet is below the level of the tank outlets. And the non-return valves may play a role.

The issue I see in the Chippie though is if a non-return valve fails and you roll/bank the aircraft, you risk losing a tank-full overboard. With no "Both", this can't happen in a PA28.

For comparison to the above, this is what you get in a typical Archer POH:

https://cimg3.ibsrv.net/gimg/pprune....46b3c2644d.png

No non-return valves fitted in the PA28 series. I've also heard some say non-return valves can get stuck and get you in trouble but not sure how true that is.. Maybe it's simply a cost thing?

If you fly in balance why would any fuel go overboard? Cessna fuel systems don’t have a non return valve and I’ve yet to lose fuel overboard even with prolonged circling on photographic flights.

Pipercameron,

Note that with your drawing the vent slot/fairing/pipe arrangement was a retrofit item introduced around 1953; so that all the original Chipmunks that came to Australian aero clubs did not have this (the RAF understandably didn't bother to install this on aircraft already ear-marked for disposal). This installation was intended to pressurize the tanks to ensure even feeding. Certainly the RVAC (at least) modified their Chipmunks without the vent to have a L/R/OFF selector.

If the fuel cock is lower than the tanks then it's only very very marginally so. On visual inspection I'd argue that it's at the same level.

And lastly, despite what the Pilot's Notes say, I know of at least one instance where with one tank empty and the other full (surely this evolving situation would have rung alarm bells to the pilot?) the engine failed.

That's some very telling (and expert) background, Dora-9.

We've therefore established that the two nominated low wing exceptions to the rule so far - Commander 114 and Chipmunk - have a basic flaw in the original fuel system design.

That's interesting.. I was wondering what all that was for. Thanks for the info! :ok:

Fair point.. but who would want to fly a fully-aerobatic aircraft like a Chipmunk in balance? Slips (especially slips to land) are all just part of the fun!! :)

I don't fly Cessnas but I have heard that, for the reason you mention, you do need to park them level to top off the tanks.. true?

The Romans figured this stuff out over 2000 years ago, it's not that hard. Fluid likes to flow downhill with gravity, if the tank (inlet) is higher than the outlet it all works fantastically, if not, you have to wait until complicated 18th century steam pumping to make it all work, and that even just involved pumping it back to a height it could flow downhill again. So you pick it up from a single sump where gravity deposits it and pump it upwards to a point it flows down again. All extra complications and weight you don't need in a simple light aircraft. Any extra holes in the system between point A and B add complications where fluid will either leak out of or air leaks in.