Integrated aerobatic fuel tank with exotic engine

Discussion in 'Firewall Forward / Props / Fuel system' started by Scheny, Mar 6, 2019.

Help Support HomeBuiltAirplanes Forum by donating:

  1. Mar 6, 2019 #1

    Scheny

    Scheny

    Scheny

    Active Member

    Joined:
    Feb 26, 2019
    Messages:
    39
    Likes Received:
    9
    Location:
    Vienna, Austria
    Hi,
    for my project I would like to build integrated wing tanks located between double spars (like Cirrus and Diamond but without the aluminum). The wing is is built from carbon over honeycomb and it will use Jet A1. Short durations of inverted flight should be possible (5min). I have knowledge of how this is normally done with gasoline engines, but unfortunately my new project is quite exotic.

    The engine is using a built-in electrical fuel pump (which is mainly used as sort of a throttle and not for suction) and is located in the vertical CoG position approximately one foot above the tanks. The fuel pump is assuming the fuel to be coming at pressures >=0.

    My plan would be to have a fuel pump (or redundant one) getting the fuel into a really small aerobatic center tank (directly next to the engine) under slight pressure and relieving any trapped air back into the main tanks (this way no problem if they are steadily kept on, then fuel just flows back instead or I have a fuel detector switch to turn pump off). The engine pump could then get fuel from the aerobatic center tank. The center tank would have a fuel strainer for checking/draining.



    • Is this design OK or flawed?
    • Can I have a Left/Both/Right fuel switch or do I need Left/Right for a low wing (why?)?
    • When sticking to Jet A1, is CFK/GFK over honeycomb OK, or should I go for Jeffco/Rhino (to be paranoid about future changes in Jet A1 composition)?
    • Does it make sense to segment the left/right tank into main/AUX or is it enough to have baffle walls?
    • Does it make sense to go for redundant pumps while engine pump (and ECU) is not redundant?

    Thanks very much in advance for all your inputs!
     
  2. Mar 6, 2019 #2

    Himat

    Himat

    Himat

    Well-Known Member

    Joined:
    May 5, 2011
    Messages:
    2,752
    Likes Received:
    631
    Location:
    Norway
    First one question, do the engine flame out if it for a moment get air and no fuel?
     
  3. Mar 6, 2019 #3

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,531
    Likes Received:
    2,785
    Location:
    Saline Michigan
    Lots of recent discussion on most of your topics, these just on the first two pages of the Firewall Forward/ Fuel System category. Most topics have been talked about sometime in the past...

    https://www.homebuiltairplanes.com/forums/showthread.php?t=30820
    https://www.homebuiltairplanes.com/forums/showthread.php?t=30876
    https://www.homebuiltairplanes.com/forums/showthread.php?t=30050
    https://www.homebuiltairplanes.com/forums/showthread.php?t=31145
    https://www.homebuiltairplanes.com/forums/showthread.php?t=30838

    Those pretty well cover the system design except inverted fuel, which could be interesting. A flop tube in your header tank will work, but I must believe that little slugs of air will make it into the line during some attitude transitions, and your engine had better tolerate that...

    As for graphite/carbon fiber tolerating Jet A, they do fine with it. Some resins do not play well with alcohols, but Jet A seems to work well with most resins. Big deal is to select resins that are known to be good with Jet A.

    Billski
     
    BoKu likes this.
  4. Mar 6, 2019 #4

    Toobuilder

    Toobuilder

    Toobuilder

    Well-Known Member

    Joined:
    Jan 20, 2010
    Messages:
    4,244
    Likes Received:
    2,972
    Location:
    Mojave, Ca
    Really not clear on your architecture here but it sounds like this was solved on the L-39. The main boost pump pressurizes a small tank with a bladder diaphragm inside. With loss of boost pressure, the diaphragm forces fuel to the engine pump until depleted.

    Timely question, as I had to deal with this particular tank leaking on an L-39 op just yesterday.
     
    rv6ejguy likes this.
  5. Mar 7, 2019 #5

    Scheny

    Scheny

    Scheny

    Active Member

    Joined:
    Feb 26, 2019
    Messages:
    39
    Likes Received:
    9
    Location:
    Vienna, Austria
    Thanks for your inputs guys.

    My design is most comparable to a BD-5 (see Avatar), but it has the wet wing between two spars instead of the aluminum leading edge tanks. As for the BD-5, size is the biggest issue shortly followed by weight.

    The manufacturer of the turbine uses an internal pump for fuel control and suggests a center tank with a boost pump and a solenoid valve as shutoff. As said before, fuel has to be delivered at positive pressure to the engine as the internal pump has no pressure rise unfortunately.

    So based on the links and comments you provided, I see two possibilities:

    1.) Filters and drains in the wing, then an Andair left/right/both followed by the boost pump delivering the header tank
    2.) Filters and drains in the wing and a fuel pump inside each tank, delivering directly the header tank

    For solution 2 I would need a solenoid valve behind the header tank (fuel pumps act as some sort of tank selector as the DA42 AUX tanks have it). I do not like this solution, as there is only one real shutoff.

    I prefer solution 1, as I rather have a parallel redundant pump after the (reliable mechanical) selector, than two pumps in the wings. Anyway, the L39 solution will not work for me, as the header would be rather small (about 1-2 gallon).

    My biggest concern at the moment is to get rid of trapped air in the header. What are your suggestions there? Air would mainly get in while I fly inverted, but I don't know how to get it out if not by flushing it back into the tanks.

    Fly safe, Andreas
     
  6. Mar 7, 2019 #6

    BJC

    BJC

    BJC

    Well-Known Member

    Joined:
    Oct 7, 2013
    Messages:
    8,672
    Likes Received:
    5,569
    Location:
    97FL, Florida, USA
    Do not use a "Both" feeding a single boost pump.


    BJC
     
    pictsidhe and BoKu like this.
  7. Mar 8, 2019 #7

    Monty

    Monty

    Monty

    Well-Known Member

    Joined:
    Jul 15, 2010
    Messages:
    1,283
    Likes Received:
    478
    Location:
    Fayetteville, AR / USA
    Hi Andreas,

    Interesting project you've got there. There are some fabric membranes which will allow fuel to pass, but not air. Have a vent that allows air to escape, and use an inlet to the pump that only lets fuel pass. You can get commercially available:

    "Air Traps"

    Monty
     
  8. Mar 8, 2019 #8

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,531
    Likes Received:
    2,785
    Location:
    Saline Michigan
    Just to check: The engine driven pump merely sets the flow rate to the injectors? And pressure to the injectors is provided solely by the airframe mounted pumps?

    Next question: There is no fuel return line from the engine? If a return line is used, that presents some more issues...

    I see issues - been doing Failure Modes and Effects Analysis for a long time... With all fuel flow pressure coming from the airframe pumps, I have a several reliability concerns:

    With the first one, should the airframe pump fail, the engine quits. Should a filter be blocked, you only have the fuel remaining in the other tank. I think a selector valve that opens to access to one tank and shuts off access to the other tank is essential, as you must keep pressure up in your header to have pressure at the engine driven pump and injectors. I suggest you look into paired boost pumps with internal check valves for this application. They are made and SDSefi.com sells such paired pumps with internal check valves that may be suitable;

    With the second one, both tank/filter/pump systems are connected to the header that you must keep pressurized for the engine to run. When you select one tank electrically, the other tank must be positively blocked, or you will just pump fuel from one tank to the other. If a solenoid shutoff mentioned is used on each wing system, that will block that system when everything is working correctly, but remember to go through the faults the solenoid valve can have and if that will let the header back drain to the non-selected tank, shutting off the engine when you select the other tank. I suggest you study using low loss check valves on each side to isolate the system not making pressure from the header tank instead of being dependant upon extra circuits and solenoids for fuel pressure;

    I concur with the BJC - You do not want a BOTH setting on your valve for a pumped system. A both setting will result in the pump drawing air from the first tank to empty. L_R_OFF is good.

    I hope that this helps...

    Billski
     
  9. Mar 8, 2019 #9

    Tiger Tim

    Tiger Tim

    Tiger Tim

    Well-Known Member

    Joined:
    Apr 26, 2013
    Messages:
    2,733
    Likes Received:
    1,513
    Location:
    Thunder Bay
    I love eastern bloc engineering, that's also how I get well water into my house.
     
  10. Mar 10, 2019 #10

    Mad MAC

    Mad MAC

    Mad MAC

    Well-Known Member

    Joined:
    Dec 9, 2004
    Messages:
    381
    Likes Received:
    150
    Location:
    Hamilton New Zealand
    So how long will it run for on each tank? and what sort of landing speed are you looking at.

    Given its a small airframe with a small turbine, so high fuel flow, probably very low endurance and an engine that probably won't restart quickly (likely high cockpit work load as well), I would be VERY weary of installing any fuel system with a tank selector as a primary fuel management system, better to develop it to the point of only an on/off selector. While its a common fuel system config for piston GA, this airframe config just makes the usual death by engine failure caused by selecting the wrong tank too probable.

    Guess you are really tight for fuel, did you considered using tip tanks with bladders in them as primary fuel source unlikely to suck air (ugly, painful for roll but unlikely to actually add much more weight).
     
  11. Mar 10, 2019 #11

    Toobuilder

    Toobuilder

    Toobuilder

    Well-Known Member

    Joined:
    Jan 20, 2010
    Messages:
    4,244
    Likes Received:
    2,972
    Location:
    Mojave, Ca
    Didn't realize this thread was a comprehensive design of your fuel system. My offering of the L- 39 solution was not in that context - rather - it was to suggest a component (an accumulator) between the ship and the engine.

    How you implement that proven solution is up to you.
     
  12. Mar 11, 2019 #12

    Scheny

    Scheny

    Scheny

    Active Member

    Joined:
    Feb 26, 2019
    Messages:
    39
    Likes Received:
    9
    Location:
    Vienna, Austria
    The engine is a PBS TJ-100. Consumption is between 40l/h idle and 120l/h at take-off. We could have chosen a centertank like the JSX-2, but we rather do not. With wet tanks the CoG will change less (very important for such a small plane, one of the biggest concerns for BD-5 etc.) and we will have less bending moment on the wing root. Although engine limits us to 6G (9G ultimate), it is nice to have the wing designed for 9G and 14G ultimate.

    As for the safety, I am a certified safety manager myself, so the whole project was about how can we make a safe version of the BD-5. Unfortunately it is impossible to have the standards I normally would like to apply, but we do our best to excel in any place we can. The turbine only has non-redundant pump and ECU (redundancy only offered for turboprop afaik), so based on FMEA a singular pump will not be my biggest concern ;-)

    Most small turbines I know have a solenoid shutoff after the tank and a redundant pump to the engine. In this case, I have said problem with dual tanks so that standard does not apply. It seems good to have flop tubes in both main tanks in any case and for the header there should be valves to relief air available (like air-separators for oil?). If not, it seems that RC jets do well without them?

    What I didn't understand is why a "BOTH" switch is bad. If I detect imbalance I will switch to the fuller tank and the same for when one is empty and sucks air. For most of the flight a both setting would reduce workload. In "my" DA40 I would switch between L/R every 10 minutes. I prefer the C172 system.

    I was not aware that I would get this much of an echo with my question but I am thanking you for your many replies.
     
  13. Mar 11, 2019 #13

    Toobuilder

    Toobuilder

    Toobuilder

    Well-Known Member

    Joined:
    Jan 20, 2010
    Messages:
    4,244
    Likes Received:
    2,972
    Location:
    Mojave, Ca
    Why is BOTH bad on a low wing?

    Put two straws in your mouth, place one straw in a full glass and the other in a nearly empty glass. Start drawing from both and see what happens when the lower glass sucks air
     
    wsimpso1 likes this.
  14. Mar 12, 2019 #14

    Scheny

    Scheny

    Scheny

    Active Member

    Joined:
    Feb 26, 2019
    Messages:
    39
    Likes Received:
    9
    Location:
    Vienna, Austria
    I get the point, but this only applies after one tank is (almost) empty and I still have have the possibility to switch to either one. From the technical perspective I understand you, but if I have only L/R I would run in to the same situation that one tank will run dry and suck air at some point.

    Nonetheless, I will project a LEFT/RIGHT/OFF now. Keeps only the question open how the header should look like and where draining/filtering should be possible. So I recapitulate: Filter and drain in each wing, go to selector, go through (potentially redundant) pump, go through header, go to the engine. I will ask the engine manufacturer about air bubble tolerance and foresee flop-tubes in each wing.
     
  15. Mar 12, 2019 #15

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,531
    Likes Received:
    2,785
    Location:
    Saline Michigan
    Duplex valve from selected wing to header, header has a flop tube pickup, and a return line out the top that returns fuel to the tank selected via a duplex selector valve. Then you have options - either:

    Unrestricted return line plus redundent pressure pump from the header or:

    Reurn line through a pressure regulator that keeps the header at nominal feed pressure for the engine.

    Nominal operation overfills the header, venting any air to wings, inverted ops keeps working, and you might want an inverted timer. You can get more inverted time by running flop tubes in the wings.

    Billski
     
  16. Mar 12, 2019 #16

    Monty

    Monty

    Monty

    Well-Known Member

    Joined:
    Jul 15, 2010
    Messages:
    1,283
    Likes Received:
    478
    Location:
    Fayetteville, AR / USA
    I've flown a lot of aircraft with wing tanks. I don't think a single one of them has drained the tanks at the same rate with the fuel valve in the "both" position. This means eventually, maybe with one tank still half full, the engine will suck air. I always wind up switching tanks every 30 min to keep a wing from getting heavy. I would imagine the situation would be even worse with turbine fuel flow rates. Recovering from a flame-out with a small turbine in a small plane with a high landing speed is not something I would enjoy. The engine won't just come back to life the instant fuel is returned like a piston.

    That said, if the header tank was of sufficient volume and the wing tanks feed the header, you could probably get away with a both position if you pay very close attention to the venting arrangements. It does introduce a possible failure though.

    Nice to see someone doing something with the TJ-100. Looking forward to seeing your progress.

    Monty
     
  17. Mar 12, 2019 #17

    Monty

    Monty

    Monty

    Well-Known Member

    Joined:
    Jul 15, 2010
    Messages:
    1,283
    Likes Received:
    478
    Location:
    Fayetteville, AR / USA
    One other thing I forgot to mention. I would carefully consider whether or not to use flop tubes in the wings. Flop tubes can cause all kinds of problems. They can hang up, get stiff with age, flop back on themselves, un-port in turbulence, collapse etc. The place for the flop tube is in the header tank. You only need a few min of inverted fuel.

    Monty
     

Share This Page

arrow_white