Transfer Pumps and Fuel System Configuration

Discussion in 'Firewall Forward / Props / Fuel system' started by wsimpso1, Nov 28, 2018.

Help Support HomeBuiltAirplanes Forum by donating:

  1. Nov 28, 2018 #1

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,627
    Likes Received:
    2,866
    Location:
    Saline Michigan
    I have a couple questions for folks.

    My fuel system has a header tank on centerline to feed the engine and along slender wet wing tank on each side to supply to the header. I am planning on using a paired selector valve to both pick the tank to draw from and to send excess fuel back. The idea is that a full header tank just overflows back to the tank fuel is coming from. I do plan to put the fuel into the header tank up high to prevent siphoning when pumps are off. Since it is a low wing airplane, I will have to pump to the header tank. I was figuring on a couple low pressure continuous duty electric pumps, but then the exact configuration and which pumps become the question.

    Configuration - So, am I better served to put the transfer pumps in parallel or in series between the selector and the header tanks? Why would I prefer series or parallel? I imagine if the transfer pumps are in parallel, each will need either an internal or external check valve to prevent the system with only one pump turned on from driving fuel around in a circle instead of keeping the header tank full. And if the pumps are better in series, will I need any check valves at all?

    Pumps - Facet makes their Solid State, Dura-Lift, Gold-Flo, and Posi-Flo pumps. I have replaced a couple of the Gold-Flo over the years in the Archer, which is not confidence inspiring. Are any of these particularly bad or good? Then there are various models for bus voltage, fuel connections, flow and head, and check valves or not. Which ones should I use and WHY? Then there are many variety of race car pumps too. Any of those to particularly stay away from or gravitate towards?

    Then we move to the high pressure pump that takes header tank fuel and sends it to the fuel injectors. Ross over at Simple Digital Systems sells a pair of Walbro pumps in a yoke that he really likes. Does anyone have reason to disagree? Ross just sent me a note explaining that each Walbro has a check valve and they normally run them one at a time, so that is cleared up.

    Onto the bigger system - Where should I place the filters and gascolator? Default is like the Cherokee that has been so good to me - after the selector. This means I only need one filter and gascolator, but it also means that the selector and pump can see debris that the filter/gascolator might have caught. To put the filter and/or gascolator upstream of the transfer pumps means that I will have to pull fuel through the finger screen, four feet of 3/8” fuel line, through a filter and/or gascolator, then through the selector valve before finally getting to the suction side of the fuel pump. I suspect that this may be an unwise idea – can anyone conclusively say yes or no on that topic​? If Yes, I am putting the components lower than the tank, but how much can the pump do?

    For discussion, there are systems out there where they put a pump at the wing tanks, pushing fuel through the filters and gascolator, then the selector. I am reluctant to do this way for a couple reasons. The first reason is that I not only will have to select the tank electrically (turning on its pump), but I also have to work a way for overflow fuel to return to the selected tank. Pump plus selector valve means an opportunity to screw up, using solenoid valves and such give more opportunities for fuel starvation. Next reason is that if I have a pump failure, all of the fuel on that side of the airplane is inaccessible. Last is that I then have to find places for twice as many gascolators and filters. I suppose I could install a panic switch that turns on both wing pumps and a standby pump too, but then I would have to add the system I am currently planning on in the first place on top of the other system components... see my problem?

    Last thoughts – there is a rather long thread on herehttps://www.homebuiltairplanes.com/forums/showthread.php?t=30050&page=9, post 128 where a manufacturer uses jet pumps to move fuel from one place to another. Anyone know where to get these jet pumps and how much flow they need to work? That would remove the need for transfer pumps by simply using the excess flow from the high pressure pump to move fuel from selected wing tank to the header tank.

    Billski
     
    Last edited: Nov 28, 2018
  2. Nov 28, 2018 #2

    Marc Zeitlin

    Marc Zeitlin

    Marc Zeitlin

    Well-Known Member

    Joined:
    Dec 11, 2015
    Messages:
    424
    Likes Received:
    571
    Location:
    Tehachapi, CA
    Like the Andair dual valve? And why the header tank? Worried about unporting wing tanks during slips?

    I'll get to pump recommendations, but you don't mention venting the header tank anywhere. Let's say you have two pumps, but, assuming that you have two pumps for redundancy purposes [and I'd recommend a SSA/FMEA to determine the need for two pumps, given that you have a header tank that has what - a 5 gallon capacity?]) both fail - if you use identical pumps, there are common mode failures as well as independent ones. Once you cannot flow fuel to the header tank, you need to vent it so that fuel can be pumped to the engine from it. You can't have a separate vent to the outside world, or else a continuous transfer pump would possibly just pump fuel out onto the ground through the vent when on. I would HIGHLY recommend using the return line from the header to the valve/feed tank as the vent, ensuring that it is always returning above the fuel level in the wing tank, wherever your wing tank vent enters the tank. This way, there's always either fuel returning to the wing tank when the header is full, or air moving from the wing tank to the header when the transfer pumps are off.

    Put them in parallel. If they're in series, a "closed" failure of one of them eliminates all fuel flow.

    Canards have used the Facet pumps:

    https://www.aircraftspruce.com/catalog/eppages/facetpumps.php

    for many years with good success. If you want checkvalves, they're available. If you want high lift (36" suction, to deal with one of your later questions), they're available. All in flow rates that are more than adequate.

    Do what Ross says.

    You don't want/need a gascolator, assuming you're installing fuel injection. You want filters. You will want a 100 micron filter such as:

    https://www.summitracing.com/parts/sum-230124b/overview/

    between the wing fuel tank and the valve/selector, and then whatever Ross recommends for a fuel filter between the header tank and his fuel pumps. If you don't have a separate filter for each tank, then crud in ONE tank could cause a complete fuel system clog - if you clog one of two wing tank filters, you just switch to the other wing tank.

    [/QUOTE]To put the filter and/or gascolator upstream of the transfer pumps means that I will have to pull fuel through the finger screen, four feet of 3/8” fuel line, through a filter and/or gascolator, then through the selector valve before finally getting to the suction side of the fuel pump. I suspect that this may be an unwise idea – can anyone conclusively say yes or no on that topic​? If Yes, I am putting the components lower than the tank, but how much can the pump do? [/QUOTE]This is NOT a problem. The electric pumps indicated above can lift 36" of fuel when starting dry, and you're not starting dry. Assuming that you put the pump at the level of the wing tank or close to it, then there is no issue at all.

    If you can put an automotive pump IN each wing tank, activated by a microswitch on the fuel valve, this is the safest solution. Bag filters on automotive pumps NEVER get clogged. The fuel valve will return fuel from the header to the selected tank - why is this a problem? The fuel doesn't know if the pump is in the tank, between the tank and the selector valve, between the selector and the header, or on Jupiter. There's a feed to the header and a return to the feeding tank - that's all it needs to know.

    If you control which pump is running with microswitches (turn on the pump system - both pumps - from a switch on the panel, and then only allow current to flow to the selected tank with the microswitches at the selector valve) then there's no manual labor involved after startup - just pick the tank you want and off you go. Fixes the issues you posit below.

    See above - none of what you posit is necessary.

    These are not the droids you seek. See:

    http://www.eaton.com/ecm/groups/public/@pub/@eaton/@aero/documents/content/ct_207260.pdf

    for an example of why. Also, not very efficient, energy wise.

    Hope this helps. Depending upon your mission, you may be complicating things too much - keep your system as simple as possible. With a header tank holding 5 - 10 gallons, even if you can't get fuel from the wing tanks, if there's a vent in the header you've got 1/2 - 2 hours of flight time to get somewhere safe. Unless you're planning long overwater flights or going to northern Canada, total redundancy may not be necessary. But even if you DO want total redundancy, two wing to valve filters, two check-valved simple, cheap electric pumps in parallel, and one dual high pressure pump and associated low micron filter per SDS recommendations will do it.
     
    wsimpso1 likes this.
  3. Nov 28, 2018 #3

    rv6ejguy

    rv6ejguy

    rv6ejguy

    Well-Known Member

    Joined:
    Jun 26, 2012
    Messages:
    3,546
    Likes Received:
    2,572
    Location:
    Calgary, Alberta, Canada
    I'd try to put the header tank (if you use one) low down on the floor and gravity feed it from each tank with minimum -6 hose. This assumes you can have the tank feed fitting in the lower corner near the spar. Have a shutoff valve for each at the root. You can return fuel to the header as long as the main tanks are vented and you have some dihedral.

    Alternately, dispense with the header and the transfer pumps altogether. Mount the EFI pumps on the floor after the Andair Duplex selector as Marc suggested. Use a large SS element filter like we offer pre-pump and another post-pump. Return fuel comes back from the regulator through the selector to the selected tank. This is how most low wing EFI aircraft are set up. It's well proven.

    The problem with pumps in the wing roots is that if they fail, all the fuel is stuck in that tank. That could be a bad surprise leaving few options if you ran the other tank almost dry. With high pressure pumps there, you need to open the valve as the pump is energized as well which increases complexity.

    The Walbro pumps are very reliable if you don't run them dry and don't feed junk through them.

    I'm with Marc on the gascolator- don't install one with EFI. Use filters and have quick drains on the main tanks at the lowest point. Water elsewhere is a minor concern with EFI which mixes it all up anyway due to the very high return rates.
     
    wsimpso1 likes this.
  4. Nov 29, 2018 #4

    Toobuilder

    Toobuilder

    Toobuilder

    Well-Known Member

    Joined:
    Jan 20, 2010
    Messages:
    4,286
    Likes Received:
    3,029
    Location:
    Mojave, Ca
    Some random thoughts: Please don’t take them as design advice without verification

    Most of the jets I’ve supported over the last few decades have a single engine mounted in the fuselage fed by a sump (header) tank system. The wing tanks feed the sump; the sump feeds the engine. In each case, the sump is a relatively small capacity (maybe 5%) of the total fuel volume, but the thing to keep in mind is that the sump is the “brains” of the fuel system. In contrast, the wing tanks are relatively dumb – only serving to supply the sump when commanded. This is usually done with simple float valves to maintain a relatively constant level. In your case, the system can be pretty simple. I’ve considered such a system myself. Consider a 60 gallon total capacity for the ship including a 5 gallon sump (for example). The sump will contain dual submerged pressure pumps (I'm assuming EFI) and a float valve that would either electrically activate the inexpensive Faccet transfer pumps at the wing roots or open a mechanical valve to let the fuel in (or both). In such a system the engine’s fuel return simply goes back into the sump, and the “fuel selector” is a toggle to route the float switch signal to L or R wing root pumps. Such a configuration significantly reduces fuel plumbing, eliminates the need for a duplex valve, and even removes the selector valve entirely. With the availability of float switches, valves and pressure switches it would be pretty simple to have a nearly automatic fuel system with plenty of warning outputs for pilot SA. All that said, it works on a simple system diagram, but venting and other considerations are going to have to be well thought out.

    As for the jet pumps, they make sense in turbine applications because there is a lot of fuel being returned due to a mechanically driven pump that moves a LOT of fuel to support the engine at TO, but needs just a fraction of that volume at cruise altitude (one airplane I maintain has essentially the same fuel flow at high altitude with WOT as it does on the ground at idle). With all that otherwise wasted effort of pumping mass quantities of fuel that's not being burned, might as well put it to work running jet pumps. For a gasoline application, it might make sense to use the return from an EFI system, but I’ll bet you would have to scale down the jet pumps to make it work. In the end, the cheap little cube clicker pumps would probably be lighter and more reliable.

    Also agree with dumping the gascolator. They don’t really do much good in a low wing design with a carbureted engine, and nothing for a FI engine. They ARE good at absorbing heat from the engine and causing vapor lock issues though. In my case, the builder installed one on my Rocket which was quickly relocated from the firewall to the inlet of the fuel pump module, serving now as a pre filter for the pumps. My main filtration is a Delco high pressure automotive disposable unit installed just upstream of the injectors. Inexpensive, easy to service, and well proven with billions of hours on a generation of GM FI vehicles.
     
    Last edited: Nov 29, 2018
    MadRocketScientist and rv6ejguy like this.
  5. Nov 29, 2018 #5

    Mad MAC

    Mad MAC

    Mad MAC

    Well-Known Member

    Joined:
    Dec 9, 2004
    Messages:
    388
    Likes Received:
    151
    Location:
    Hamilton New Zealand
    Odd thoughts
    Check Valves can have non independent failure modes depending on system layout (contaminated fuel resulting in slowly leaking check valves).

    Ensure the fuel pump can self prime (have encountered aircraft where it won't but was fine till you have to put additional tankage).

    Have you had a read through at FAR 27, its fuel system section is much stricter than FAR 23 (before they broke it) with some interesting practical testing (most of it is not applicable but still thought provoking).
     
  6. Nov 29, 2018 #6

    narfi

    narfi

    narfi

    Well-Known Member

    Joined:
    Feb 23, 2016
    Messages:
    660
    Likes Received:
    295
    Location:
    Alaska
    You should look at production planes of similar configuration and power requirements and see what you like, then ask to look at the manuals and copy them. Why reinvent the wheel when someone has put a lot of money into doing the research for you?

    For example, the bonanza, Low wing, IO520 or turbo normalized IO550,

    fuel.jpg

    What engine are you planning to use? I feel some of the answers you received assume specific engines, did you post your plans for that somewhere else?

    For example if your engine normalcy just runs with the engine driven fuel pump, then I wouldn't see the need for two electric fuel pumps for redundancy, the first one is already redundant. (none of the cessna/beech singles I work on have more than one electric pump*) But if your are using an engine that does require an electric pump on at all times, then I could understand the desire for more.

    Same goes for the gascolator, all those planes I am familiar with use one, so again it would depend on the engine being used and what requirements it has.

    I have a lot to learn, and the spinney parts and what make them spin are one of my weaker areas, so I may be way off course.


    *Not entirely true... the ones with aftermarket long range tip tanks use electric transfer pumps into the main tanks, but are not really part of the ships fuel system.
     
  7. Nov 29, 2018 #7

    Toobuilder

    Toobuilder

    Toobuilder

    Well-Known Member

    Joined:
    Jan 20, 2010
    Messages:
    4,286
    Likes Received:
    3,029
    Location:
    Mojave, Ca
    Bill- what is your expected fuel delivery method (carb, mechanical FI, or EFI), and what's the HP?
     
  8. Nov 29, 2018 #8

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,627
    Likes Received:
    2,866
    Location:
    Saline Michigan
    Thanks for the speedy response. I can do structures, but fuel and wiring are a little new to me. This was a big help in confirming I am on a pretty good path.

    Yes, one path the fuel feed, one path a vent line back to the tank being drawn from.

    Yes. Low wing, 4 degrees of dihedral, AR of 9. Long skinny tanks.

    Actually I did, but you probably did not recognize my use of "overflow" as "vent". My apologies there. That is why I am going for a double tank selector valve - The one for fuel coming from the wing, one for air or fuel going back to the wing, and serving as vent when the transfer pump is not running. I get ya!

    I have been doing FMEA since 1980. Are you implying that the extra plumbing, switch, and circuit plus common failure modes to run a second pump may lead to a net loss in reliability? Hmm... The Dual Battery Dual Alternator with cross feed and an Essential Bus is also in the plan with one of each transfer pump and one of each high pressure pump on each main bus.

    I get ya! Fly in the ointment - I only carried the return line about 2/3 of the tank span into the wing tank, so it could be covered with fuel during the fullest part of the tank draw. I have done the thought experiment - if the vent line in the wing tank is covered with fuel and the transfer pumps are blocked, we would need to lift the fuel about 18" (level attitude, top of the vent line in the wing to top of the header tank) to go from wing to header tank. That means we would need a combined deltaP of about 0.5 psi to keep fuel moving from the wing to the header. A positive vent pressure plus vacuum from the Walbro exceeding 0.5 psi would still result in back flow up the vent line... 80 knot glide speed is 135 ft/s is about 21.8 lb/ft^2 is about 0.15 psi. We would have be able to stand 0.35 psi vacuum in the tank and get it from the Walbro pump. I found multiple pumps in the Walbro site that have 48” of dry lift, so we ought to be OK with that. Just have to design the header tank to stand that much vacuum.


    Understood. I wondered how low the probability of that was versus the additional issues with check valves on both transfer pumps...

    Been to their site, like what they have available.

    Oh, I am planning on that.

    That is why I listen when you speak. Lots of news in that paragraph.

    “To put the filter and/or gascolator upstream of the transfer pumps means that I will have to pull fuel through the finger screen, four feet of 3/8” fuel line, through a filter and/or gascolator, then through the selector valve before finally getting to the suction side of the fuel pump. I suspect that this may be an unwise idea – can anyone conclusively say yes or no on that topic​? If Yes, I am putting the components lower than the tank, but how much can the pump do? “
    All good news. I shall have to look up the draw of Ross' Walbro pumps. Saw pumps rated for 48” of dry lift, I think this will work even when the end of the return line is submerged in the wing tanks.

    Wings are built and closed up, so no automotive pump in the wing tank.

    I can put a Facet pump on each wing very close to the tank and work a switch mount on the tank selector, but if I lose that one pump, I lose that fuel. I suspect that would drive a need for a second pump on each side. I think that a filter on each wing, selector, and then two transfer pumps in parallel with check valves on the floor will work nicely...

    These are not the droids we seek. The smallest ones I found were sized for the KingAir... Design and build my own? I probably could, but I am not going to.

    Oh, this was very helpful to me. A 5 gallon header tank will give me a little less than an hour best range speed, which does not feel like enough out west. Re-thinking 10 to 12 gallons. My remaining concern is how much lift Ross’ Walbro’s can make – I might need 18” and Walbro has pumps with 48”, so I might be OK here.
     
  9. Nov 29, 2018 #9

    rv6ejguy

    rv6ejguy

    rv6ejguy

    Well-Known Member

    Joined:
    Jun 26, 2012
    Messages:
    3,546
    Likes Received:
    2,572
    Location:
    Calgary, Alberta, Canada
    In aircraft EFI fuel systems we don't really want to lift much due to "vapor lock" concerns especially if mogas will be the fuel. We ideally want to flood the pump inlet so it has little to no work to do on the suction side. Walbros will lift about 36" in my tests to prime themselves which some Bosch designs won't do. This is a major reason for using them over the Bosch plus they are lighter, more compact and less expensive.
     
  10. Nov 29, 2018 #10

    Toobuilder

    Toobuilder

    Toobuilder

    Well-Known Member

    Joined:
    Jan 20, 2010
    Messages:
    4,286
    Likes Received:
    3,029
    Location:
    Mojave, Ca
    Which is why I like submerged pumps in the header. Cooler running, no lift concerns
     
  11. Nov 29, 2018 #11

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,627
    Likes Received:
    2,866
    Location:
    Saline Michigan
    The fuel port on my wing tank is 3" above the floor of my cockpit, and it goes away completely at 5 degrees of deck angle, so gravity feed just will not cut it. This bird is engine-in-front, low wing two seat side-by-side. Header tank top is about 18" above the high spot on the return line in the wing tank.

    That was my going in design, with a half gallon chamber protected by a surge door, but the more I looked at it, the more I became convinced that my usable fuel would be pretty limited if I were to do a slip down final or any aerobatics. My tanks are about 4" deep at their deepest, 60" span and 4 degrees of dehedral. The header tank does look like the right way to do all of this.

    Agreed.

    Ergo the filters and warning switch in each tank for low fuel plus a warning switch in the header for less than full.

    That one was big news for me. Thanks for the pointers and all the help.

    Billski
     
  12. Nov 29, 2018 #12

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,627
    Likes Received:
    2,866
    Location:
    Saline Michigan
    Thanks for all the inputs. The wife and I are off for the eye doc (followup from vitrectomy last week, return to flight in a month or so) then off to Minnesota by car. I have only had time for the first two notes so far, I will get through them all in the next couple days.

    Billski
     
  13. Nov 29, 2018 #13

    rv6ejguy

    rv6ejguy

    rv6ejguy

    Well-Known Member

    Joined:
    Jun 26, 2012
    Messages:
    3,546
    Likes Received:
    2,572
    Location:
    Calgary, Alberta, Canada
    I don't understand when you say at 5 degrees deck angle, gravity feed would disappear. Are your tanks aft of the spar?

    My RV6A has a similar dihedral to your design as do hundreds of other RVs, Lancairs etc. fitted with EFI, although my wing is thicker. There are no issues with fuel feed in banked attitudes. I too thought like you and initially had Facet pumps pulling from the main tanks, filling a high mounted header with the EFI pumps mounted below it. I had reservations about eliminating the header and Facet pumps. My flight testing using the new system with 5 gallons of fuel and bank angles up to 30 degrees in both directions, even without the ball in the center showed no fuel pressure loss. I've never had any issue in the last 9 years.

    There is nothing wrong with using your proposed system. It will work as that's what I had for the first 6 years in my RV but it's heavier and more complicated. You should use what you're most comfortable with however.
     
  14. Nov 29, 2018 #14

    Toobuilder

    Toobuilder

    Toobuilder

    Well-Known Member

    Joined:
    Jan 20, 2010
    Messages:
    4,286
    Likes Received:
    3,029
    Location:
    Mojave, Ca
    Indeed. My tanks on the Rocket are going to be 70 inches long when I put the bigger ones in and I don't anticipate a need for a header. Appropriately baffled tanks will handle min fuel OK as long as the pilot takes normal precautions.

    In a slight, but important revision to my initial post, I liked the header tank in an EFI scenario which by defenition is electrically dependand and also set up to "push" fuel where it needs to be. In such a case you can sometimes get away without mechanical selector valves. If there is a pump required to pull however, then a selector valve is needed upstream to positively close off an empty tank. I also like header tanks as a way to increase fuel capacity if the configuration works out, and also as previously mentioned a great way to submerge the primary fuel pumps and simplify the return. All that said, I'm not sure they are appropriate as a means to protect from unporting a primary tank. As Ross points out, there are just too many similar designs flying with long, skinny tanks mounted low that have otherwise completely conventional "L R OFF" fuel selectors. A header in this case is just adding complexity and weight.
     
    wsimpso1 likes this.
  15. Nov 29, 2018 #15

    TFF

    TFF

    TFF

    Well-Known Member

    Joined:
    Apr 28, 2010
    Messages:
    11,128
    Likes Received:
    3,012
    Location:
    Memphis, TN
    Jet pumps or motive flow is popular on airliners. I don't know if anyone has done a GA plane. You are one of the few who could come up with a home design. Except for some inverted flying I can't see the reason for a header for EFI. Airliners put the jet pumps in a box on the floor of the fuel tank. Open on the bottom. If the plane, heaven forbid, goes up side down, the fuel pickup is in a that box already, and can draw the fuel from that small box that just became the tank. I don't know how EFI would work with a flop tube, but the header can't take any air if you are relying on one fixed pickup if you go inverted.
     
  16. Nov 29, 2018 #16

    Marc Zeitlin

    Marc Zeitlin

    Marc Zeitlin

    Well-Known Member

    Joined:
    Dec 11, 2015
    Messages:
    424
    Likes Received:
    571
    Location:
    Tehachapi, CA
    Until you do the SSA/Fault Tree, you don't know whether you've made things better from some baseline or worse. I generally find that good engineers can find/fix about 80% of failure modes without an SSA/FT, but finding the last 20% (standard 80/20 rule, I guess) takes the analysis. Don't ask what the #'s are for bad (or even average) engineers...

    Good. I'm in the process of creating a new wiring diagram and IP for my plane (mission creep from an ADS-B upgrade plan), and will be using the same Nuckolls Z-14 diagram with some minor modifications and ground power input, and will be doing the same - one pump on each bus, pumps in parallel (assuming I ever upgrade to an EFII system - I want the infrastructure in place). I don't have transfer pumps or a header tank, though.

    If I understand you correctly, you'd only need to "suck" the fuel this high if the siphon was broken in the return line from the header tank to the wing tank and you needed to "suck" the fuel back up from the wing to the header through the return line. If you fill the return line, then the total suction pressure needed would only be the difference between the HP pump position and the top of the fuel in the wing tank - the up and down to the header is a wash. Of course, if there's less than 2/3 fuel in the wing tank, you're not getting any fuel from it through the return line, so this isn't a particularly useful failure mitigation in any case.

    A failure mode to be concerned with, however, is the valve sticking in the "off" position as you transition from "LEFT" to "RIGHT". Unlikely, but possible. You think - well, BFD, I have 5 gallons in the header tank. But with the valve "OFF", you can't get air into the header, and as you suck 40 gph (or whatever the Walbro pumps do in normal use) out of the header, you'll quickly collapse the header tank due to vacuum - at 7 gph usage, you'll either starve the pump or collapse the tank in a couple of minutes.

    Now, this isn't necessarily a failure mode that you should worry about - if MY fuel valve sticks closed, my mechanical FI engine stops in about 4 - 7 seconds at cruise power - with a float bowl in a carb, maybe you get 15 - 20 seconds. We live with it, because it's a very rare failure mode, and I switch tanks only when I'm over an airport. But I don't destroy a fuel tank and spray fuel all over the inside of my cabin should this happen, as would occur if your header tank collapsed.

    Yes - for the siphon case, that's easy, assuming you've got a 2 core 2 tank with a few baffles inside - 1 psi or so is not a problem. HOWEVER, see above - the issue is what happens when the vent AND feed lines are blocked - you can get a LOT of vacuum (the max of the Walbro suction). If the Walbro can do 48", that's about 2 psi, so you might want to design the header for 3 - 4 psi suction, to be safe - then, the worst case is that you just run out of fuel if the fuel valve sticks "OFF", just like every other plane on the planet.

    Or maybe have a dedicated vent on the header with a check valve so that it can only allow air IN, not pump fuel OUT. This would fix the "collapse" issue, and also ensure that in the case of a valve failure, you can still draw ALL the fuel out of the header. But it would keep you from drawing fuel from the wing tanks in the "pump stuck closed" failure mode you originally posited above. I don't see that as a problem. This would NOT hurt the standard transfer pump usage, as the transfer pump is just pressurizing the header, not de-pressurizing it.

    I hope that you've also got an "ON/OFF" fuel shutoff between the header and the engine. Both for maintenance reasons and fire reasons. You may not use it often, but you have to have it (and I don't remember a mention of it).

    I'm waving my hands around, because I have no idea what the relative failure rates of any of these components is - you'd have to look in a reliability database and then do the analysis. All I know is that when doing the WK2/SS2/RM2 SSA/Fault Trees, we did run across situations where things were non-intuitive. Not often, but sometimes. You'd certainly expect that the simpler the device, the more reliable. There were a few things we needed to redundify after the design was complete, and other things where we found that the redundancy that had been designed in wasn't helping any, due to other failure modes that dominated.

    Yes. But you've got a whole other tank and the header tank - it's not an emergency, it's just a "land somewhere within the next 1/2 hour or so" MAJOR issue - it's not Catastrophic or even Hazardous, unless you're in the middle of the ocean.

    I agree - then either pump can be used for either tank. More flexible, for sure, and there's no reason that the pump has to be upstream of the valve or very near the tank.

    So I live in Tehachapi and fly out west all the time. While there are many places that look like the surface of the moon when you're flying over them, there's almost no-place where you're not within 15 minutes of a paved airport at a 160 KTAS cruise speed (which I what I'm imagining your airplane will be able to cruise at easily). If you stay somewhere near a road during your flight, then there's never a time that you can't land safely within 1/2 hour. There's certainly nothing wrong with a larger header tank, if you have the space for it, but I wouldn't consider it necessary. For you to get into the situation where it's the last resort, BOTH transfer pumps and/or the fuel valve would have had to fail. Even with a 1/1000 hour failure rate pump, that's 1/1M hour failure probability. In an airplane that I fly 100 hours/year, I don't worry about 1/1M hour failure rates.

    I'd be very surprised if the pumps couldn't do that, but even if they can't, see the discussion above - you're no worse off than in any other plane in which the fuel flow completely fails. The only single point of failure, with the system you're planning, is the fuel valve. And NO-ONE puts in redundant fuel valves :).
     
    wsimpso1 and rv6ejguy like this.
  17. Nov 30, 2018 #17

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,627
    Likes Received:
    2,866
    Location:
    Saline Michigan
    Thanks for the inputs.

    I have thought through the electrical selection system, and I find myself looking at the level sensors in the pump circuit, adding more failure modes in both sensors and in wiring, wondering if I am really making a net improvement in reliability. After that, I figure that a pump failure deprives me of the fuel on that side of the airplane. So, we end up with EITHER a third transfer pump, check valves, and a selector valve anyway OR redundant pumps in each wing. Not happy, FMEA to get a better feel.

    Agreed, but the topic is still interesting. The pumps everyone is recommending do not like to be run dry, while a jet pump does not care. If I run a tank dry with a Facet pump, I may not figure it out until the pump is wrecked. Nearly as I can figure, the Walbro FI pumps are 35 gph while the most the emerging will need is somewhere around 25 gph, giving me a minimum of 10 gph to run a suitably designed jet pump. I did find awebsite with all of the design equations and practical range of coefficients, so I could do a jet pump for each wing instead of transfer pumps. It would be big design/build/test effort, but it might take some wiring and switches and failure modes right out of the airplane while allowing me to full drain one wing tank instead of having to switch to another tank when I still have 20 minutes fuel in the thing... which is probably the route I will take anyway.

    Agreed.

    Billski
     
  18. Nov 30, 2018 #18

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,627
    Likes Received:
    2,866
    Location:
    Saline Michigan
    Well, actually, a slowly back flowing check valve in my system is a pretty benign thing in the system I was proposing. The check valves are NOT keeping anything full. The header tank fill and return connections are all up high in the header tank, so they do not need check valves - they can only back flow a tiny amount before the lines are uncovered. I am only seeing need for check valves on the pump outlets to keep them from running fuel in a circle - right way through the one that is on, wrong way through the one that is off.

    Pumps under consideration have plenty of dry lift. No issues...

    I shall review Parts 23 and 27.

    Thanks for the inputs...

    Billski
     
    Last edited: Dec 1, 2018
  19. Nov 30, 2018 #19

    wsimpso1

    wsimpso1

    wsimpso1

    Well-Known Member

    Joined:
    Oct 19, 2003
    Messages:
    5,627
    Likes Received:
    2,866
    Location:
    Saline Michigan
    Aside from curiousity, what difference would that make?

    SDS EFII. Engine type is up in the air, with O-360 having been the default and being tempted by Ross on an EJ257 and AeroMomentum on their new engine. Planning on 150 hp in cruise. The remaining option is Gemini keeps teasing us with a 200 hp. The airframe based fuel system would be the same for all of them, with the exception that we should run a air in and an air out line for the diesel. I installed both when I built my wings.

    Billski
     
    Last edited: Dec 1, 2018
  20. Nov 30, 2018 #20

    dcstrng

    dcstrng

    dcstrng

    Well-Known Member

    Joined:
    Oct 17, 2010
    Messages:
    912
    Likes Received:
    321
    Location:
    VA or NoDak
    Am not an engineer, A&P or anything of the type (nor do I am I related to any), but have been following a related thread over on the Zenith forum; and this cautions strongly against in-line pumps without a bypass loop…

    [Posted by Bob Jones on November 17, 2018] “I was asked to tear down a plane that had crashed, and find the problem , the FAA and Rotax had inspected the plane and decided it was a pilot problem, the fuel bowls were empty and they were able to bypass all systems and start the motor , I found the FACET fuel pump that was inline to have failed and totally shut off the fuel, one person died because of the crash, DO NOT PUT A PUMP INLINE < PUT IT IN A BYPASS LOOP or leave it out you do not need a boost pump on most aircraft engines , also fly the plane no matter what happens… something came apart inside the pump , I actually thought I had made that clear , all pumps allow gravity flow in one direction , this pump is totally blocked both ways, and the filter which should always be installed before pumps was cut open and found spotless, I know and have worked on lots of set ups that have boost or emergency pumps but they Mostly have been installed in by-pass loops there - by giving double safety and back up , from the exit of the switching valve a "Y" should be installed with the ELECT. in one line and the MECH.in the other then rejoin before the carbs with a second "Y" both ELECT. and MECH. pumps have reverse flow checks , so can independently supply the carbs ... p.s. does any one know if Facet or and other aftermarket elect. fuel pump supplier sell a AIRCRAFT rated pump or are they all automotive ?? I will try to call Facet and ask if they approve their pumps to be installed in airplanes…”

    Sorry if this is thread-creep… seemed germane but am not sure… {eliminate if needs be...}
     
    wsimpso1 likes this.

Share This Page

arrow_white