I suspect this is down to the more direct supply of fuel to the engine and the sensitivity to fuel pressure.
And probably also due to the fact that high pressure fuel pumps require by injected engines don't work properly when they are dry. Just like Diesel powered cars that are very difficult to get started again after being allowed to run dry. Some time ago I used to fly a C340 with six tanks. Unless you wanted unknown quantities of unused fuel in every tank you needed to run them dry before switching. When one was quick enough to switch tanks and turn on the electric boost pump at the first sign of coughing the whole process would not be noticed by the passengers. If you were slow and the engine coughed for a few seconds with the aircraft yawing, some passengers would get very scared.
It seems that carburetted engines recover very quickly, but my experience with injected systems ia about 10 seconds of rough running before the lines are flushed with fuel again.
I suspect this is down to the more direct supply of fuel to the engine and the sensitivity to fuel pressure.
In either event it is no great issue if planned for. There is little loss of RPM or altitude when I have tried it.
If you have refilled a tank that has been emptied, it's a very good idea to start and run on that tank on the ground to purge those fuel lines.
Do you really think that air stays in the pipes after the tank is filled again? I don't think so. Even if a little air stayed there, with the electrical pump turned on for take off, with the fuel used for heating the engine, there would be no air left in the circuit during take off. In my opinion that's not the reason for this rule. It simply a precautionnary thing. In France many pilots fly in aeroclubs with DR400. The 120 and 140 versions of the aircraft have but one tank. For many pilots, simply switching tank is a thrilling adventure... It would maybe have been one for me too, if I had not done it, including running the tank dry, with a instructor first.
That rule could make sense for a club or group aircraft. If you run a tank dry you put air in the fuel line from that tank. The next time that tank is selected it could give somebody a nasty surprise when the engine stumbles. This obviously depends on the fuel system in that aircraft.
I also run a tank dry during my training to get my licence with my intsructor. As said previously, the engine did not stop suddenly, but coughed once or twice. I turned the electrical pump on, then switched tanks and the engine ran smoothly again immediately. The instructor had a point in showing it was a non event. However, in all the other clubs I have been after this one, rules were not to run a tank dry and pilots never had the opportunity to do it during their training. Too bad, in my opinion.
I had dry the tanks twice. The time fully intentionally - I was to ferry flight a new microlight to Italy (I really mean new type) and wanted to get fuel gauges confirmed before setting up for several hundreds miles trip. Done above the airport at 3000 ft with Rotax 912. Engine started shaking and after switching to the other tank power was restored immediately. The second time I was flying on old 172 with Rolls-Royce 0-200. (it is a Continental you know). I was somehow not happy with the left fuel gauge reading, after visual inspection I made all take-offs and landings on the right tank. High enough I switched to left. And of course forget about chance the tank will be dry soon. When the power was lost, I didn´t even lowered the nose, just moved the fuel selector and power was restored immediately.
I've used the handy trick with the minute hand on my analoge watch. If it's pointed to the right (0-30 mins past the hour) I'll run on the right tank. If it's pointed to the left (30-60 mins past the hour) I'll run on the left tank.
As as result, I'm never more than 30 mins out of balance and it's easy to remember which tank to switch to and when.
In terms of running a tank dry, I'd occasion once where I thought this appropriate. At sufficient altitude (and no passengers) I was determined to run a tank dry. At roughly the predicted time, the effects were obvious.
The engine did not stop, but it started to die/recover. Sort of a surging. It wasn't very severe...perhaps a 200rpm at peek. The fuel flow needle showed the fuel pressure dropping a bit and then recovering with the surges. I left this for about 20 seconds. The engine didn't die, but the surges became faster/more pronounced, and I lost my nerve then and changed tanks.
But good to know that if it happened unintentionally, that I have some symptoms to warn me in advance.
This was on a Piper Arrow by the way.
Colm
Useful feedback....leading to some further questions
In the TB20 a fuel tank imbalance is noticeable via the yoke, as one would expect
Do you notice a larger effect on the yoke for the same fuel imbalance when the tanks are full versus less than half?
The fuel should not slosh around much when banking if the flight is coordinated (ball more or less in the middle)
Why not? Is that because the main force in a co-ordinated turn is directly underneath the banked aircraft rather than vertically downwards to earth?
In years past, when driving older cars and using what was then probably less sophisticated fuel, I was cautioned against running the tank dry because of the gunge/grit etc. that might be lurking at the bottom. Presumably this doesn't happen in GA aircraft, where any sediment would have been drained from the tanks pre-flight.
It seems to me there is a choice when running a tank dry between the risk of not knowing how much fuel is left on the live tank during approach/landing (or just plain running out of fuel) vs removing the redundancy of being able to switch tanks if there is any fault/problem with the tank in use.
What I sometimes find is that the fuel warning light (either tank below 8 USG) comes up during taxi
I have seen the low fuel warning light come on in our aircraft, but through this was connected to the Shadin fuel flow meter (which at that time thought we had less than 60 minutes of fuel remaining). I'd not configured it correctly after refuelling, so knew this was wrong. I'd better check it's not driven from some other sensor.
I ceased to run the tanks dry after I got a fuel totaliser fitted, as it simply wasn't necessary.
As JasonC points out, it wasn't something to do without briefing the Px; done correctly, you would sit with the fuel pump on waiting for the fuel pressure needle to waver, ready to change tanks; there would barely be a cough.
As Peter rightly points out, it was important to switch off the autopilot (or at least the Alt Hold) to prevent the thing trying to maintain height until the stall!
As I previously stated, this was only necessary on long flights without a fuel totaliser; think 4 hrs with a 5hr range, where the lack of absolute certainty as to where the fuel was located could lead to one arriving at a destination with 10l in each of the four tanks, where even a modest angle of bank on an approach procedure to land could lead to a fuel intake being uncovered.
On short flights (3hrs or less) there was never any need to run the tanks dry.
However, the fuel management of the PA32's with four tanks is one of the model's "gotcha's", the others being rate of descent power off and the consequences of loading the aircraft outside of C of G (which is very easy to do).
As ever, as long as the pilot is properly trained as to how to deal with known issues, there is no problem; issues do arise when these have never been covered.
Air-restarts are not covered in the PPL course, but then neither is leaning or the use of GPS or 1001 other useful things....a whole different topic!
There is always going to be a theoretical chance of some problem.
As Deakin correctly says, a certified SE plane is supposed to restart within X seconds (a different and longer number for a twin), but you never know what might be lurking in there. A bit of dirt maybe? For example you would not tempt fate by switching tanks in mid Channel on the way to Le Touquet... or when halfway across the Alps.
And if you are on autopilot, altitude hold, it will stall the plane. If you are not paying attention, the airspeed will drop off to the point where the engine stops windmilling and then you have to use the starter motor, which will (depending on the avionics master switch installation) interrupt power to all avionics, so you then have to wait for everything to restart, etc. In any case you have to turn off avionics power when starting the engine, due to starter motor inductive spikes. Not the sort of thing you fancy doing in a hurry, especially in IMC, when the autopilot has just stalled the plane...
Also when the engine stops, any vacuum instrument(s) will wind down.
An instructor I used to fly with used to stop the engine when doing pleasure flights (oops I meant to say trial lessons) with passengers.
