REFLECTOR: fuel feed - my solutions

Brett Ferrell bferrell at 123mail.net
Mon Feb 28 17:58:02 CST 2005


  ----- Original Message ----- 
  From: Jim Sower 
  To: Velocity Aircraft Owners and Builders list 
  Sent: Monday, February 28, 2005 1:59 PM
  Subject: Re: REFLECTOR: fuel feed - my solutions


  Brett Ferrell wrote: 
    I actually don't have anything riding on it, because if I need the check valve to open, my primary line is already plugged, and the factory-recommended supplemental vent check valve can only *help*, not hurt the situation.  

  True.  Provided that it operates as hoped.  My (and Al's?) whole point is that it may very well not operate as hoped, and that a check valve with a 0.5 psi cracking pressure almost certainly will NOT operate as hoped. 

  As I've explained previously, you'll have much more that .5 psi to open the valve.  You'll have whatever the current cabin air pressure is (a few psia) versus vacuum.
    I haven't done any calculations on it, but I do have an engineering degree and have done process safety studies, and I don't ever want my tanks to be unvented, if possible.  The mass of fuel leaving the tanks needs to be replaced with an mass of air of like volumne, or a vacuum will begin to be drawn on the tanks 
  Exactly.  Remember that "begin" is the key word here.  But you have three vents.  Two of them are A) vented outside and subject to blockage by ice, critters, etc.  B) backed up by a check valve with suspect cracking pressure (which is the whole point of this discussion).

    and potentially cause thier failure or cavitation at the pump, 
  Failure is a non-issue since cavitation (or fuel exhaustion, which is the issue here).

    neither of which is desirable in my view.  I just wanted folks to hear a voice saying that they believe the factory design is a sound one, in addition to all of the talk of modifications.  
  Agreed again.  I'm talking about what I regard as the most important failure mode factory design as it stands now - without modifications. 
  Follow this scenario:
  1.  Wasp or whatever takes up residence in vent line under the belly of the airplane.
  2.  With the blockage undetected, you take off for Easter Vacation
  3.  As you burn off fuel, the strakes will attempt to gravity feed into the sump.  As the fuel level drops and absent any air from the blocked vent to replace the missing fuel, there will be a pressure drop in the air bubble above the fuel in the strakes (a vacuum will begin to form).
  4.  The head pressure of fuel (as measured between the surface of the strake fuel and the surface of the sump fuel) will cause the fuel to continue to transfer for a while, gradually decreasing as the underepressure increases toward the head pressure.  The more fuel you have in your strakes (smaller air bubble) the faster the pressure drop will increase (removing a pint of fuel from a tank with a small "air bubble" will cause MUCH more "vacuum" than removing the same pint of fuel from a tank that is half empty and has a very large "bubble").
  5.  Soon (again, much sooner if you started with full fuel) the underpressure in air above the fuel in your main tanks will equal the head pressure of the fuel trying to gravity feed to the sump.  At that point the fuel flow from the main tanks (which has been gradually decreasing as the underepressure increased) will stop altogether.  Al and I are arguing that this stoppage will occur at something like .4 psi or less - not enough to unseat your check valve - so your .5 psi check valve will amount to a second blockage.
  Agree up to here, where you believe that a pump with a fully submerged inlet will not pump fuel, with admittedly very low head.  If you got the pump curves they would tell you exactly what happens here, which is not good to be sure, and might in fact harm the pump over a prolonged period of time, but the pump will continue to move some fluid, the pressure will continue to drop in the tank and pull a (relative, not 0 psia, but significant) vacuum on the tank.  I've designed systems to work like this.   Here's an article for you with an example pump curve, note the NPSH is only "to prevent cavitation", not to allow fluid to be pumped.
  http://www.mcnallyinstitute.com/02-html/2-03.html
  http://www.coleparmer.com/techinfo/techinfo.asp?htmlfile=ReadPumpCurve.htm

  Besides the fact that the check valve doesn't have 18" of fuel sitting on it, so it "sees" a significantly lower pressure than the pump sees at any given moment.  Here is also a concrete, real world example, that an airplane fuel pump can generate significant, damaging, vacuum on a fuel tank.  I don't think the lightweight spring in a 3/8" check valve is going to hold back the kind of vacuum that implodes a metal fuel cell.  I'm not sure what Cessna's tank can do, but in my experience they can usually take at least 1-2 psi of underpressure, which would have already opened said check valve before damage.  In fact, in my process safety work, I've seen examples where I don't even need a pump, just opening the drain will draw enough vacuum to crush the tank, which is why people make and sell vacuum breakers.
  http://www.aopa.org/asf/asfarticles/2003/sp0305.html


  6.  All the while, the engine is consuming upwards of 15 gph of fuel.  Since the sump is vented and the sump vent is open, it will happily deliver fuel to the engine which will be very easily replaced by air flowing quite freely thank you very much, through the unrestricted sump vent.  As the fuel consumption of the engine surpasses what the mains are able to deliver (having been slowed to a relative trickle by the increasing "vacuum") the fuel level in the sump will begin to drop.

  I'm not sure what you're saying here, about the vent being open, because you've stated above that it has been plugged, and all three tanks share a common vent manifold.

  7.  At some point (perhaps a gallon below full) the low sump fuel will light.  Let's suppose you don't notice it right away and / or take a little time analyzing the situation and figuring out what's happening - let's say a few minutes.  Now, you're down to three gallons of usable fuel and fifty gallons of unusable fuel.  

  If you plug the sump vent line (which I wouldn't do), you don't really have a sump tank at all, you have a wide piece of pipe.  Mostly that will probably work out for you, but could cause issues.  If you get air entrained in the fuel (say from it sloshing around), it could come out of solution due to the change in cross-section of the pipe (when it enters the sump), and then it would not be able to vent and you'd have a trapped air bubble sloshing around in there (potentially setting off your low level sensor).  I'm not saying that you would ever see that, but I don't see any value in it.  Also, if your design is meant to allow the pumps to draw a vacuum on the sump to pull fuel out of the mains in the event their vents get plugged, I hope you don't draw so much vacuum on the sump that you crack a seam.

  8.  I'll accept your estimate of how far you are from home at this juncture, but I think we'll agree that you've got a very serious fuel situation.  Let me further suggest that if this hasn't happened to you yet and you've largely forgotten this thread, you will be experiencing some strong psychological pressures: a)  to focus on the 50 gal in the mains and try really hard to recover that fuel (false hope)
  b)  to ignore the low sump fuel indication in light of all the fuel in the wings (denial)
  c)  to put the best face possible on the (unknown at this point) amount of fuel left in the sump.
  These pressures serve only to aggravate your already dire situation.

  If you were to make a case (which is within the realm of possibility) that the head pressure full tanks is over the cracking pressure of your check valve (say 0.6 psi) you still haven't solved anything much.  At 0.1 psi over the cracking pressure of the check valve, it will do just that - crack - definitely not open up far enough to provide more than a trickle of fuel.  You will only delay the onset of  conditions 5 and 6 above, not prevented them.  You will be farther from home when the sump light goes on and the shit hits the fan.

  If we ignore all other factors (like the ones that caused my and other's assymmetric transfer, the scenario above can happen at various points along your trip depending on the cracking pressure of your check valve.  Since the checking mechanism of a check valve must weigh something, by definition there must be a cracking pressure.  I strongly suspect that the factory design pretty much presupposed a "perfect" check valve, that is one with zero cracking pressure.  But of course only the designer can speak to that with any authority.  For my own part, I have what I believe to be a rubber reed type check valve which I orient upside down so as to have the weight of the check mechanism trying to open the valve and rely on properly operating vent overpressure to keep it closed.  I've no way of knowing how well that scheme is working.

  My point in this whole discussion is and has been to neutralize what I believe to be a design defect in  the factory vent system.  I propose to accomplish this by interrupting the scenario above at Step-6.  If the sump vent is capped (the sump not allowed to vent) then the pumps will create an overpressure in the now closed fuel storage system.  Instead of the "vacuum" caused by the pumps being vented by the sump vent, it is transferred to the main tanks.  Now, instead of only head pressure trying to crack your check valve,  you have the added suction capability of the pumps.  Your 0.5 psi check valve will now tend to work better.  You will still have at least a .5 psi underpressure in the tanks.  Since the vapor pressure of Avgas is over 7 psi and that of Mogas maybe 10 psi or perhaps more) it will work less and less well as the atmospheric pressure declines with altitude.  All of this will be further degraded by the fact that the fuel pump will not draw anything remotely like a perfect (14.7 psi at sea level, 7 psi at something below 20k ft altitude) vacuum.  As you climb, the vapor pressure of your fuel remains the same, the atmospheric pressure decays and the difference is the only pressure you have to open your check valve.


  So now the pump can pull a vacuum on the sump, and in course, the mains, but in the previous scenario it couldn't?

  The vapor pressure only comes into play if the fuel can offgas as fast as your pump is draining fuel from the tank, which is highly dubious (do you have a figure for that?), and if it could, then you would never need a vent, since the tank would never been exposed to vacuum, and you could happily allow the vapor-pressure push the fuel out of the tank.  Vapor pressue is the pressure that the would result in a closed container if it was allowed to reach equilibrium, which will never happen in this dynamic pumping scenario, what we really need to know if we thought it were important, is the evaporatation rate.  http://hyperphysics.phy-astr.gsu.edu/hbase/kinetic/vappre.html

  Further, once the valve opens, it will not reseat until the flow has subsided, so descent will not be an issue.

  Your engineering degree and certainly your process safety experience more than qualifies you to examine what I've been discussing.  The physics is high school level, as is the math.  All you have to do is examine and quantify some of the very general terms like vacuum and check valve that tend to get waaaay oversimplified in a superficial discussion of the issue.


    I believe it is, and that's what I'm installing.  I'll leave the math to those changing that design.
  I don't believe it is.  Your understandable faith in the design presupposes that the designer examined these issues in detail and designed around them.  I am not at all convinced that they did.  I strongly believe that you and the designer ignore an in depth and detailed examination of these issues at your peril.

  My education and experience lead me to understand the soundness of the design.  It is simple, fault-tolerant, and the basis of the majority of pumping systems in the world - it's not sexy, it just works.  It also has the advantage of having a significant number of flight hours under it's belt.  I don't see any reasonable basis to investigate it further.  Those who do, can.  I just don't want anybody to go out to fix a problem that I don't believe exists without hearing both sides, and the potential risks.  I can now leave the conversation for others with clear conscience.

  Brett

  But you're both going to make your own decisions.  I wish you well ... Jim S.

  PS  If you want to check my math, the numbers I used are roughly as follows:

  1 atmosphere = 14.7 psi
  1 atmosphere = 32+ ft H2O
  therefore 14.7 psi = 32 ft H2O = 384 in H2O
  and 1 in H2O =  14.7/384 =0.03828 psi
  1 in Avgas = 1 in H2O * 6.0 / 7.4 = .03104 psi 
  1 psi = 384/14.7 =~ 26 in H2O = ~ 32 in Avgas


    Brett
      ----- Original Message ----- 
      From: Jim Sower 
      To: Velocity Aircraft Owners and Builders list 
      Sent: Sunday, February 27, 2005 10:29 PM
      Subject: Re: REFLECTOR: fuel feed - my solutions


      Brett Ferrell wrote:

        I don't avodcate sucking the fuel, but if you're drawing 15 gph, I don't believe the fuel is going to off-gas fast enough to fill the void, warm or not, so I don't believe that partial pressure will keep the check valve from opening and venting the tank. 

      Do you have anything like say, physics or math to support that postulation?
      I hope so.  You've got a lot riding on it ... Jim S.


        Brett 
        ----- Original Message ----- From: "Al Gietzen" <ALVentures at cox.net> 
        To: "'Velocity Aircraft Owners and Builders list'" <reflector at tvbf.org> 
        Sent: Sunday, February 27, 2005 6:29 PM 
        Subject: RE: REFLECTOR: fuel feed - my solutions 



          Subject: Re: REFLECTOR: fuel feed - my solutions 

          As soon as the pump pulls some fuel out of the sump with a blocked vent 
          line, the check valve will open and vent some air into the tanks.  1 
          atmosphere versus vacuum will open nearly any check valve 

          Of course you don't really have that option since warm avgas has a vapor 
          pressure of around 8-9 psi, and warm mogas 10-12.   Which means at _______ 
          ft, you have no "suction". 
          That's why you don't want to depend on sucking fuel anywhere. 

          Al 

          _______________________________________________ 
          To change your email address, visit http://www.tvbf.org/mailman/listinfo/reflector 

          Visit the gallery!  www.tvbf.org/gallery 
          user:pw = tvbf:jamaicangoose 
          Check new archives: www.tvbf.org/pipermail 
          Check old archives: http://www.tvbf.org/archives/velocity/maillist.html 




        _______________________________________________ 
        To change your email address, visit http://www.tvbf.org/mailman/listinfo/reflector 

        Visit the gallery!  www.tvbf.org/gallery 
        user:pw = tvbf:jamaicangoose 
        Check new archives: www.tvbf.org/pipermail 
        Check old archives: http://www.tvbf.org/archives/velocity/maillist.html 





--------------------------------------------------------------------------
      _______________________________________________
      To change your email address, visit http://www.tvbf.org/mailman/listinfo/reflector

      Visit the gallery!  www.tvbf.org/gallery
      user:pw = tvbf:jamaicangoose
      Check new archives: www.tvbf.org/pipermail
      Check old archives: http://www.tvbf.org/archives/velocity/maillist.html
----------------------------------------------------------------------------
_______________________________________________
To change your email address, visit http://www.tvbf.org/mailman/listinfo/reflector

Visit the gallery!  www.tvbf.org/gallery
user:pw = tvbf:jamaicangoose
Check new archives: www.tvbf.org/pipermail
Check old archives: http://www.tvbf.org/archives/velocity/maillist.html

------------------------------------------------------------------------------


  _______________________________________________
  To change your email address, visit http://www.tvbf.org/mailman/listinfo/reflector

  Visit the gallery!  www.tvbf.org/gallery
  user:pw = tvbf:jamaicangoose
  Check new archives: www.tvbf.org/pipermail
  Check old archives: http://www.tvbf.org/archives/velocity/maillist.html
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.tvbf.org/mailman/private/reflector/attachments/20050228/a371cdbd/attachment.htm


More information about the Reflector mailing list