Clunk!

[onetenor]

Sounds like Spitfire ( and any Merlin powered plane) Achilles heel.They had a problem with the float bowl in the carb emptying under Neg "G".The float was pressed against the fuel cut-off either owing to the fact that gravity alone would shut off the valve or neg "G" forces would do it. Or a combination of the two. Now as we are not dealing with a carb/float system the only conclusion I can come to if that the g forces are pushing the fuel bodily away from the pick up somehow. Going by the tank shape and size I'm guessing that the fuel is being spread up the sides of the container. Something to do with the S.Gravity and fluidity of fluids under gravitational and similar forces. There was I believe a study done on it somewhere .(Aus rings a bell ).You can get some idea of the effect by swinging some coke or other coloured fluid around on a string tied to the bottle in various positions and observing what happens to the fluid under load. Not exactly scientific but! It would be fun for the kids to spin round while we observe ('til they get fed up of course). A conical shape might be good.  I've done this as a kid just for fun but saw some funny effects. One was the fluid spread thinly against the sides on such a way that a clunk would not be immersed just simply wet.  In this position the clunk would pick up air. I suggest that something like this could be happening here Try different tank shape perhaps?????  P.S try different string lengths doing the tests.

Edited By onetenor on 02/04/2018 22:07:09

Edited By onetenor on 02/04/2018 22:09:26

[Petrol Pete]

Hi Gary, I think everyone's fousing on the tank, remember these carbs have a very effective pump, so we should be focusing on the fuel in the flote chamber ( the diaphragm side of the carb) do put a extension on the vent into the fuz. Check the height of the needle arm and a good check of the diaphragm, a full kit to overhaul these on e bay is only about £6 . It must be going weak or rich for a moment if it cuts out and then will restart within a few flicks if it was a fuel line problem you would have to flick it many times to draw fUel back up. I rember the early Dle30 with the side carb used to cut out if you flicked it round after landing to taxi it back, this has got to ba a mixture problem. Pete

[Gary Manuel]

I'm not ruling anything out yet because I simply don't understand what's going on.

In my head though, I'm reasonably satisfied that it's not the fuel supply that's causing the problem. That's why I'm concentrating on the air / fuel mixture and the ignition spark.

[Gary Manuel]

I've changed my plans due to the weather and because the pressure nipples and M4 Taps arrived this morning.

I've decided that rather than flying without the cowl on, I'll do the carb diaphragm regulation mod first, see how it behaves, then tweak the mixtures and leave the ignition unit till last.

This is one of the nipples I'll be using. It's not quite what I thought I was ordering. I assumed that the gasket washer would fit between the small nipple head and the casing and that I'd discard the nut. It doesn't - it's obviously intended to be fitted with the nipple on the outside, flush up to an unthreaded hole in the tank, with the washer and nut on the inside of the tank. At this point, I thought that I would just make a smaller washer out of some suitable gasket material, but it didn't quite work out that way.

This is the diaphragm casing I'll be modifying.

Casing removed to reveal the diaphragm.

The existing vent hole blocked with solder from the inside .....

..... and the outside.

2.5mm pilot drill followed by 3.2mm. The final hole should have been 3.3mm to suit the M4 taps I'll be using, but 3.2 is the nearest I had available.

Hole tapped with an M4 tapered tap.

Continued in next post ......

 

Edited By Gary Manuel on 06/04/2018 17:48:47

[Gary Manuel]

Screwing the nipple into the threaded hole revealed a problem!

There is a small shoulder between the threaded part and the hex head, which prevents the nipple from being screwed in until the head is flush with the diaphragm casing. This is because the nipple is intended to be pushed into a plain hole rather than being screwed into a threaded hole.

My plan was then to screw the large brass nut onto the nipple, add the washer, then screw the assembly into the casing..... but there's still a problem. The brass nut does not fit flush up to the nipple hex head for the same reason that the nipple didn't screw flush into the casing - due to the shoulder on the nipple.

To get round this little problem, I used a ball shaped grinding wheel in my Dremel and cut a recess in the brass nut to accommodate the shoulder on the nipple. That's better. I assembled these two parts with a strong thread-lock.

The nipple assembly and washer were then screwed into the diaphragm casing using medium thread-lock and the solder on the outside tidied up a bit with a Dremel / grinding wheel.

Inside tidied up in a similar way. The small length of the threaded part of the nipple was also ground flush with the casing.

Modified diaphragm casing reassembled onto the carb.

Tube pushed onto the nipple and routed down the side of the engine box ......

..... into the fuselage. I decided to route it into the fuselage rather than the engine box to keep it away from any moving air associated with the carb air intake. I didn't bother with the 35mm film can because I think the air should be fairly static inside the fuselage as there are no deliberate air holes in it.

Edited By Gary Manuel on 06/04/2018 17:51:58

[Rich too]

I like you use the same method always. I use very large Mahl filters as used in garden equipment, fit and forget them. And no special install - doesn’t matter if it is not completely full.

I always run three lines. I do not run two clunks. Leave fuel in the tank, why empty it? It is not good for the plumbing. I never use T lines.

I am surprised to see that you have not looped the fuel line over the tank on the vent line? Perhaps I am not seeing it correctly.

And as Ron said, I more often than not, carry out maiden flights with no cowl.

Good luck Gary, it must be very frustrating.

[Gary Manuel]

Frustrating is one way of putting it Rich! Engine runs lovely and model flies great apart from this problem.

I'm hoping to get chance to test it tomorrow. I'll take the spare ignition unit and tools needed to change it if all else fails.

PS vent line loop is now on top of the tank rather than down the side of it.

[Gary Manuel]

I've just realised that I never updated this thread to report how I got on.

The next two flights after the above modifications went very well. I found that I was unable to force the engine to cut, even with sustained inverted spins, bunts and rolling circles - manoeuvres that would previously have ended in a dead-stick. I was feeling rather chuffed when the engine suddenly went from a nice quiet tone to sounding like "crazy frog". I landed immediately only to find that the exhaust header had broken near the manifold.

If the model hadn't felt like the best thing I'd ever flown, I would have given up at this point, feeling that the model was jinxed. Instead, I arranged for a replacement header pipe and began to think about what I'd done to resolve the problem of the engine cutting on negative G.

I was happy with the plumbing. I was not convinced about the inline filter as this was potentially adding stored air into the system. I was happy that the problem was nothing to do with the ignition module. I was reasonably convinced that the problem had been resolved by venting the carb diaphragm into the fuselage.

I decided to simplify the plumbing by getting rid of the inline filter. The system is now a very simple 2 line tank. The vent pipe double back over the TOP of the tank and out of the bottom of the fuselage. The clunk pipe goes to the engine via the straight legs of a T piece. The offset leg of the T piece goes to a fuel dot for filling / emptying.

New header arrived and fitted. Test flights were faultless. I am absolutely convinced that the problem was fixed by venting the carb diaphragm to the inside of the fuselage. There must have been a negative pressure created within the cowl when undertaking negative G manoeuvres, that caused the engine to cut.

I'm happy with the way the model is performing, but will report back if I have further issues.