RCU Forums - View Single Post - Fuel system plumbing and air trap consideration
Old 11-06-2009 | 04:31 AM
  #1  
olnico's Avatar
olnico
 
Joined: May 2007
Posts: 4,120
Likes: 0
Received 16 Likes on 10 Posts
From: Houston, Texas.
Default Fuel system plumbing and air trap consideration

Hello guys.
I am starting this thread following the discussion on the FEJ F-16 thread where several guys had flame outs with their P-200 and a normal fuel plumbing.
I think that this discussion might be of a general interest.

https://www.ultimate-jets.net/blogs/...considerations

https://www.ultimate-jets.net/blogs/...ing-cavitation

https://www.ultimate-jets.net/blogs/...considerations

I would first like to mention that I have written an article that partly covers this matter, to publish in the December/January RCJI issue.

The big engine available on the market ( 160 N and above ) have a high flow requirement. This means that the fluid velocity in a standard tygon tube is very high.
The flow Q is given by: Q=SxV. If the section S of the tube decreases, the the velocity V will increase at a constant flow ( non compressible fluid ). What happens is that passed a certain velocity, the fluid will start to cavitate: bubbles of gaseous kerosene will form in a perfectly leak free system. So any restriction in the lines could create cavitation at a certain point ( restriction= S decrease = V increase )...
Additionally the flexible UATs like the BVM or Hansen products will collapse to such an extend ( on very big engines ) that the walls of the tank will touch each other. This will restrict the flow around the air trap bag and might induce cavitation downstream it !
So in that case the air trap system becomes the bubble generator...

I have done extensive tests with all kinds of UATs and all kinds of engines.
I have a very simple rule: no flexible UAT for engines above 120N.

1) I use the biggest Tygon size for my vents ( 8 mm ) and two 6 mm brass tubes for the vent pickup ( big bore ).
2) All the cluncks are re-drilled
3) 6 mm fuel tubes everywhere ( big bore ), no filter: the air trap does this job.
4) the 6mm tube slides on the pump nipple with a small piece of 4 mm tubing and 2 turns of safety wire.
5) I use the rigid PST air trap that is fitted with 8 mm nipples and ultra large air trap foil folded about 8 times ( very large air retention area )
6) Keep the fuel lines downstream the air trap as short as possible

The PST air trap is very big and offers a significant buffer time in case of intermittent bubbles. It does not cavitate at all because of the large retention area. It is also correctly priced given its technical specifications ( 50 USD ).
The other option is to use the JMP accumulator (air trap) that is made of aluminium and is of course very rigid.

I prefer the PST version for the several reasons:
1) The transparent bottle makes you see what is happening while doing the test runs
2) It has a very high volume
3) It is much cheaper





When I have finished setting up the system, I do a stress test in real conditions without the engine running. I just loop the fuel line that is supposed to feed the engine back to the fuel vent ( possibly with a Tee in a 3 tank configuration ). This way the system runs on a closed loop.
I then enter the engine test menu and rum the pump at whatever voltage is needed ( 6v in the P-200 SX case ). I can then have a good look at what the system is doing at a simulated full load.
While doing this test, I move the plane in all the possible positions ( ie: nose up, nose down, both sides, inverted ). I always keep a good look at the air trap behaviour and any air bubble entering/exiting it, as well as at the engine fuel feed ( which is looping back to the tank and not connected to the engine on this test ).
I also do the same test while feeding the fuel back to the my Jersey modeller field tank ( open loop ). This way I can see the behaviour of the fuel clunks when the tanks empties.

I have not had any problems with this setup on my two P-200 for many test and flight hours...

Last edited by olnico; 06-07-2017 at 01:10 AM.