Received-SPF: pass
 receiver=logan.com; client-ip=64.136.30.75; envelope-from=alwick@juno.com
To: flyrotary@lancaironline.net
Date: Wed, 7 Dec 2005 14:18:06 -0800
Subject: Re: [FlyRotary] Re: fuel system issues
Message-ID: <20051207.141825.2636.4.alwick@juno.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
From: al p Wick <alwick@juno.com>

This is so significant. It's clear you have marginal fuel system. I've
heard descriptions of guys who experienced vapor lock. If I understand
the physics correctly, your design is darn close to that condition. You
say "what? you idiot, I've got 300 hours on it!"  Let me try to explain. 

Let's say that it takes 100 points to cause vapor lock. We know a number
of the causes. Lets add up the causes. If the total =100, then you WILL
get vapor lock.


Fuel temp > 100F= 40 points.
Altitude > 10k ft =5 points
Fuel inlet filter=10 points
Engine at idle=10 points
Low main tank fuel level=20 points
Auto fuel= 30 points
Hot fuel pump=30 points

So basically the factors are heat, head pressure, and type of fuel.

If you add up all  of these factors, they total more than 100 points. So 
if you operate your plane under those conditions, you are sunk. Doesn't
matter how many fuel pumps you have. They can't pump vapor. (you can do
test to verify that). 

But each of those are independent probabilities. It doesn't often happen
that you have your plane soaking on hot ramp, filled with auto fuel, at
high altitude. Every once in a while you get close to that condition.
Then one day, wham, no fuel. Engine stops.

So the main lesson here is that the various causes are additive. Some
pilots are smart enough to  avoid inlet filter, so a little less risk
than others. Others may use only aviation fuel, some may live in colder
climate.

Do you realize that Vans brake design has the same causes? He too has
marginal design. It works  great 95% of the time, but every once in a
while, the independent odds stack up to cause a failure. I was appalled
to see the brakes don't have a heat shield. Shield is highly effective at
delaying heat transfer to tubing, legs. Fibrefax also is highly
effective. Very marginal design.

How would the Japanese design this system? They would actually arrive at
a number which represents the real risk. So instead of building fuel sys,
firing up engine, then saying "Seems to work ok", they would say "How
close is this to failure?" They would measure it, arrive at a number. 
So anything we design, if we use attributes to make decisions, we are
likely to have sys close to failure. I've seen this time and again. Can't
emphasize the concept enough. We have to find ways to measure how close
we are to failure. There are tons of simple ways. We can do it all by
bench  testing. This is the biggest oversight in design.

To prove you are at low risk of vapor lock, just measure the pressure at
pump inlet. There is direct correlation of that pressure to your risk.
Then heat the fuel, pull a vacuum on the "fuel  tank", add inlet
restriction.

Or you can place your pump inside main tank and do what millions of
vehicle designers have already proven.


-al wick
Artificial intelligence in cockpit, Cozy IV powered by stock Subaru 2.5
N9032U 200+ hours on engine/airframe from Portland, Oregon
Prop construct, Subaru install, Risk assessment, Glass panel design info:
http://www.maddyhome.com/canardpages/pages/alwick/index.html


On Tue, 6 Dec 2005 21:29:31 -0500 "Ed Anderson"
<eanderson@carolina.rr.com> writes:
> Always more  than willing to hear another view pointsone fuel systems 
> - 
> can't know too much about how they function and possible problem 
> areas in my 
> opinon- given there somewhat critical function{:>)
> 
>  Its always interesting to view our beliefs and assumptions in the 
> context 
> of someone else's experience or knowledge.  I do fly (300+ hours)  
> with an 
> unvented, small header tank that is approx 1 1/4 pint (36 cubic 
> inches)capacity (or if I want to make it sound large - it has a 
> capacity of 
> 590ccs {:>)).
> 
> All my fuel from the engine is returned there via the pressure 
> regulator 
> discharge.  My theory was that the smaller volume of fuel gets used 
> very 
> quickly and replenished (especially at high power settings) with 
> cooler fuel 
> from the tanks.  Indeed this does seem to be the case with my setup. 
>  I view 
> my setup perhaps functionally similar (but not identical) to the 
> No-return 
> fuel systems being seen more frequently in fuel injected 
> automobiles.
> 
> I did find early on when using auto gas that on a hot day,  and with 
> low 
> engine power settings (returning more hot fuel and using less 
> thereby 
> causing less cool fuel to be pulled into the tank),the fuel pressure 
> could 
> start to fluctuate.  Turning on the boost pump (approx 6 psi) 
> pressurizing 
> the header tank would cause the fluctuations to end.   I have since 
> place a 
> fiberglass box with aluminum foil around my header tank and fuel 
> pumps and 
> have a blast tube routing air to the box.  I am also using 100LL 
> rather than 
> auto gas.  In any case, I have had not further fluctuations in the 
> pressure 
> or any other indication of a fuel flow problem even on hot days down 
> in 
> Louisiana this summer.  .
> 
> I suspect that as in many cases, its the details that makes one 
> system 
> function well vs a similar - but not identical installation which may 
> have 
> problems.  So "copying" a "good" installation should be done with 
> utmost 
> care, small - seemingly insignificant - changes to the design may 
> have large 
> unintended consequences.  The one  change I would make to my own 
> system 
> would be to move this header tank out from under the hot cowling 
> area say to 
> a wing root or anyplace protected and cooler.
> 
>  I considered EFI pumps in the wing tanks but dislike the long lines 
> of 
> highly pressurized fuel running through the cockpit area.  Certainly 
> many 
> different design approaches have been made to work - but none are 
> without 
> some risks in some area.
> 
> 
> FWIW
> 
> Ed A
> 
> .
> > It could be an issue, however, if the fuel is
> >> returned
> >> to a smaller header tank depending on how the header tank is
> >> constructed and
> >> vented.
> >
> > I suspect you want it vented. Unvented opens you to new problems. 
> Like
> > fuel sys that won't self prime.
> >
> >
> >>
> >> If the fuel is recirculated long enough, and if the bubbles are
> >> really air;
> >> the dissolved air should be eventually nearly eliminated from 
> the
> >> system,
> >> but I didn't run the pump long enough to see if this is true.  I
> >> haven't
> >> tried the tests with avgas, either.
> >>
> >> My apologies to the list if this is common knowledge.
> >
> > Don't apologize. This is valuable info that deals with one of the 
> biggest
> > risk items our conversion have. Thanks for providing some 
> science.
> >
> >
> > -al wick
> > Artificial intelligence in cockpit, Cozy IV powered by stock 
> Subaru 2.5
> > N9032U 200+ hours on engine/airframe from Portland, Oregon
> > Prop construct, Subaru install, Risk assessment, Glass panel 
> design info:
> > http://www.maddyhome.com/canardpages/pages/alwick/index.html
> >
> > --
> > Homepage:  http://www.flyrotary.com/
> > Archive and UnSub:   
> http://mail.lancaironline.net/lists/flyrotary/
> > 
> 
> 
> --
> Homepage:  http://www.flyrotary.com/
> Archive and UnSub:   http://mail.lancaironline.net/lists/flyrotary/
> 
> 


-al wick
Artificial intelligence in cockpit, Cozy IV powered by stock Subaru 2.5
N9032U 200+ hours on engine/airframe from Portland, Oregon
Prop construct, Subaru install, Risk assessment, Glass panel design info:
http://www.maddyhome.com/canardpages/pages/alwick/index.html