Scott Gettings wrote:

This should give a higher-than-expected temperature rise.
Factors making the temp lower than in real operation include: no
incoming heat from the fuel rail return, and no heat from the backup
pump running during takeoff and landing.

I'm not certain I agree with the above statement. I believe you are
ignoring the heat gain the fuel receives simply by passing through the
engine compartment. Most engine compartments are darned toasty,
especially so during the Summer.  I would not be surprised if the fuel
rail temp were 150 degF.  If you take 2.5 gal of 90 degF fuel and
recirculate it through the 150 degF fuel rail, in a few passes your
fuelwill be darned close to 150 degF.
So, I don't mean to poo-poo your test; the test is good as it
quantifiesthe pump's contribution to the fuel heating issue. But, I
think we need
some more data before anyone can claim to have data proving that fuel
heating isn't a problem with your proposed setup.
What's the worst case temperature for the fuel loop? I'd guess that it
would be extended climb-out in AZ in August at 15:00 (local) after the
aircraft has sat on the ramp all day. The whole engine compartment has
been heat soaked, the airflow is low, the engine is making best power
(and throwing gobs of waste heat), and this is about the worst time
forthe engine to quit.  Or maybe touch-n-goes in the same environment,
since the fuel flow is quickly reduced when you level off, but the
system is still hot as blazes.
We need to get a handle on how much heat the fuel can acquire.  I'm
going to assume that all of the fuel lines (both flexible and
rigid) in
the engine compartment are wrapped in fire sleeve, which is a good
thermal barrier.  So most of the heat gain from the engine compartment
is going to happen while the fuel is in the fuel rail and the pressure
regulator (assuming it's on the end of the fuel rail).  We need to
measure the fuel rail's surface temperature to get an idea of how much
heat could be transferred. Anyone got a plane we can use to measure
this?  The measurement could be as simple as using the temperature
indicating paints or crayons, or maybe one of those IR-based
temperaturesensing guns.
Russell Kent

Perhaps I didn't communicate this clearly.  I was trying to say that the
temperature rise I measured was probably LOWER than it would be in real
operation because there was no heated fuel coming back from the fuel
rail return (regardless of its cause).  We all know this is going to be
hot.  And only 1 pump, not 2, was running.
So heating might actually be greater than measured for these reasons.
Incoming fuel would lower the temps, but by how much?
Also note that this was only a 15-minute test, not a steady-state
temperature plateau.
Scott Gettings

Hello Scott,

We amphibian builders have a few more things to consider since our runways
sometimes move.

I am going through a similar thought processes right now with the mid wing
SeaWind I'm building.  The manufacturer cautions to take care about design
alteration concerning the  use of the  BOTH tank setting on the Andair valve
or the interconnection between tanks because of  water operations.

For example lets say you pull up to near a dock, sort of, and get out and go
have lunch.  When you come back you may find one of your airplane wings
underwater and the other one in the air. The reason is clear I'm sure.

So the solution they advised were check valves between the tanks and header
tank, which turned out to be not always reliable.  I am guessing either
seepage to dirt in line here.  Anyway  they eventually  went with electric
solonoid valves. Not exactly KISS but seems to work.

A couple of builders wanting a gravity drain system, use the Andair valve
and do pull off both tanks through header tank like you are suggesting. One
guy  rigged a buzzer that sounded if the canopy was opened and the valve was
left in the BOTH position.  This scenario of course requires at least two
switches.  Another guy have the an cover door over the valve selector lever
that would not close unless it was in the OFF position.  Presumably he would
notice the open door between the seats.
Another fellow had two header tanks of the size you are recommending, one
for each wing and fed the fuel back to the header tank from whence it came.
Complicated and doesn't really solve the fuel heating problem in my opinion.

This is the first time I ever heard anyone mention a fuel radiator.  That is
an interesting idea for hot days, but perhaps not for winter. Food for
thought though.

Still the safest solution I've heard of yet is the 'return to same tank'
idea.

The best logical solution however would be to have a system that eliminated
the need for fuel to be returned to the tank at all.  Non return pulse width
modulated fuel pumps are new for the airborne rotary group and I'm not
familiar with the failure modes yet.  It sounds like they would have to be
interfaced to an electronic pressure gauge and have some type of
proportional response circuit in the controller to prevent pressure
fluctuations.  Basically a microprocessor.

As an aside, with the Glass Goose, do you have to transfer fuel between
sides to balance the plane properly for water takeoff?  In the SeaWind you
want one wing heavy to compensate for the torque, so a transfer pump and
additional fuel lines is also included in the kit.

Keith Moore and Ian Beadle have a interesting system you might be interested
in if you like 3 way solonoids.  Do you have their post from 5-25-2005?  If
not I can send it  to you in a Word document.


Cheers

DougFir

No need. Here it is. Don't forget the Andair detents are weak. PL

------------------------
> Paul;
>
> You have focused on a critical problem; i.e. EFI fuel system with two fuel
> tanks .  At some risk to my reputation, may I offer an EFI Fuel System
> Design for consideration!
>
> The attached diagram features a 3 way NO (normally open) solenoid valve (5/8
> dia).  A switch in the cockpit selects the fuel tank and powers the pump for
> that tank.
>
> When the Right tank is selected, the Right pump is ON and the 3-way NO valve
> returns fuel to the Right tank.  When the Left tank is selected, the Left
> pump is ON and the solenoid valve closes to direct return fuel to the Left
> Tank.
>
> Panel LED's show which tank is selected. Fuel Level gages show the tank fuel
> levels.  A "flashing / bell" alarm signals a low tank.  A pressure gauge /
> alarm reports the status of the fuel feed to the engine.
>
> Note that:
>    - Its a simple, fail safe design - intuitive for all  pilots
>    - There are no fuel lines in the cockpit!
>    - The expensive multiple pipe line selector valve is not required!
>    - All active components may be mounted on the fire wall, and/or under the
> fuel tanks.
>    - There are no surge tanks - all flow is "From - To" the selected fuel
> tank
>    -  The fuel level in each tank can be easily managed by switching tanks as
> required.
>     - Dual pump redundancy is maintained - perhaps improved
>
> As further comment,
> 1.)  Each tank vent line shall first route from the top of the tank to the
> bottom of the fuselage and then to the top of the fuselage - so if the plane
> flips inverted, gas will not dribble out - don't ask me how I know!
>
> 2.) All fuel lines shall include a ground wire to a common point, including
> the fuel cap on the tank (particularly for fiber glass tanks).
>
> 3.)  Fuel lines in the engine compartment should be stainless steel, and
> fire sleeved.  Note the flex lines - as required.
>
> 4.) The Long EZ fuel tank has  built in baffles to minimize the movement of
> the fuel, such as when in a banked turn.  The design also includes a sump
> tank with screen.
>
> 5.) Fuel flow rate is computed by summing the EFI injector pulse width.
>
> 6.) Observe that the EFI pump flow rate is typically about 10 GPM.  This may
> quickly empty a surge tank, even with a large connecting line.    I have
> always worried about additional / separate fuel sump tank.
>
> Regards,
>
> Keith (Black Hat) Moore
> LEZ


Keith
You have described excactly the fuel system that I had in my RV6A. It worked
perfectly. I will be installing the same system in my new RV7.

Ian Beadle

In studying this LEZ system a little more, it seems that:
1. If one fuel pump failed, the engine would quit, so you must switch
to the other pump to restart the engine, AND that tank must have fuel.
This means you can't ever run either tank near empty, because you might
need the pump on that side.

2. If the solenoid stopped working, it is also a single point failure
that could either block fuel from coming back from the rail, or dump
fuel back to the wrong tank.

Am I missing something?

Scott Gettings
BTW, I measured



The Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site. http://www.rotaryeng.net
Copyright 1998-2006 All world wide rights reserved.