[Jsbsim-devel] [ jsbsim-Feature Requests-963659 ] Detailed Fuel Tank design options

[SourceForge.net <no...@so...>]

Feature Requests item #963659, was opened at 2004-05-31 11:09
Message generated for change (Tracker Item Submitted) made by Item Submitter
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=369399&aid=963659&group_id=19399

Category: Architecture
Group: None
Status: Open
Resolution: None
Priority: 5
Submitted By: Jon S. Berndt (jberndt)
Assigned to: Jon S. Berndt (jberndt)
Summary: Detailed Fuel Tank design options

Initial Comment:
Attributes:

1) Capacity
2) Contents (actual)
3) Empty weight
4) Selected
5) TransferRate
  a) dependent on whether gravity feed or,
  b) pump assisted,
  c) acceleration vector affects rate, too (inverted, 
zero "g", etc.)
6) TransferToTank (Tank0, Tank1, ... Tankn)
7) Pressure

Methods:

A) Consume
B) Transfer (double rate)
C) Pressurize
D) Vent
E) Fill
F) Drain

It could be that what is needed is an expansion of the 
fuel system parts of FGPropulsion containing a list of 
FGTank objects and refueling valves as inputs, and lists 
of FGTank, FGEngine and dump valves as outputs.  On 
the other hand, this may be more than we need, as it 
could get awfully complicated without much benefit in 
return.

One benefit of a fuel system is that the tanks don't have 
to know where they're sending the fuel, and likewise the 
engines don't have to know from which tank they're 
getting fuel.  This will make it easier to model fuel feeds, 
fuel transfers, refueling and dumping.

To keep things manageable we can assume that each 
tank has one fuel boost pump, and the pump is always 
working.  We can assume all the fuel is usable, leaving it 
up to the user to initially set the tank capacity to usable 
capacity.  Inverted flight can get messy to configure, so 
maybe we could have one configuration item per tank, 
INVERTED, which takes a "double" value indicating how 
long the tank will supply pressure if inverted.

Somebody (I can't remember who) offered a fuel 
temperature algorithm.  What to do with this 
temperature is another issue.  It should be available for 
instrument panel display, but the effect of too-hot or 
too-cold fuel should vary enough from engine to engine 
(and between fuel types) that we probably shouldn't 
hard-code anything here. 

Currently the FGEngine object consumes fuel by asking 
FGPropulsion which tanks are assigned to that engine, 
then removing fuel equally from all assigned tanks.  In 
Boeings, and some other airplanes, you can assign more 
than one tank to an engine, and the engine will remove 
fuel from one tank until it is empty, then remove fuel 
from the next tank (i.e. sequentially rather than in 
parallel).  This could be easily modeled by assigning each 
tank an output pressure.  The engine will then remove 
fuel from the tank(s) with the highest pressure first.

I agree that the gravity feed thing is important.  For 
example, high-wing Cessna singles usually have a "both" 
fuel-selector position, because the fuel flows down via 
gravity, while low-wing Piper singles do not, because the 
fuel is pumped (imagine sucking through two straws, one 
in a full glass and the other in an empty glass).

That makes fuel management an important issue in a 
Piper -- owners tend to do it instinctively after a while (I 
set a timer on the yoke just to make sure), but renters 
who are used to Cessna’s sometimes get into trouble: 
one Ottawa pilot ditched (successfully) in Toronto 
harbor a 
few weeks back after running one tank dry.

You might consider 'unusable fuel' also.  That tends to 
depend on pitch, perhaps a 'handle' on dynamically 
cutting fuel flow would be appropriate.

Large jets often have a fuel heater to reduce 
coagulation below –40 C.  I think the 727 has a heater 
ahead of the fuel filters to get rid of ice crystals that 
can cut fuel flow.

Another possibility is to be able to set the 2'nd moments 
for large tanks (typically wing tanks) about their 
own 'centers'.  That could be added to  m*d^2, where d 
is the distance to the 'weight center' of a tank.  The 
first moment, g*m*d is what affects balance.  
For 'space vehicles' g clearly changes.

The problem is that long wing tanks have higher Roll 
MoI's than the distance to the center of mass accounts 
for.  One can't set both Balance and MoI with one 
distance to the tank's center of fuel mass.

If one tank could be broken into two or more tanks the 
difference in Radius of Gyration and 'Balance Distance' 
becomes much less significant. IOW, break one wing 
tank into 1A and 1B so the moments come out better 
while the user sees 1A and 1B as just '1'.

In practice, I've set FS wing tanks a bit further out than 
the center of mass so the MoI changes more 
appropriately wrt fuel level. Wing tanks are normally 
balanced, so the effect of the First Moment isn't 
generally a problem.  But, it would change more than it 
should if MoI is favored in setting the distance(s).

Wing tanks may also tend to be tapered, which would 
reduce both the first and second moments from what a 
top view would suggest.

Properly set, the effect of changing fuel in wing tanks in 
jet transports is quite noticeable on roll dynamics.

One could also account for difference in volume capacity 
and mass capacity.  I expect large AC have higher 
volume than can be used in cooler temperatures since 
they have to handle the fuel expansion on hot days. 
They are always filled with 'lbs' <sic>, not gallons.



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

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=369399&aid=963659&group_id=19399