Sign in to follow this  
Guest Ron Freimuth

Questions about Jet-A fuel?

Recommended Posts

Hi,I'm working on a DC-9-41 project and I'm in the process of modeling the fuel system.could anyone answer these questions....1) What is the freezing point of Jet-A fuel?2) Once freezing point has been met, how fast will ice accumulate and at what point would this seriously clog the fuel filters? (Assuming the fuel heater were not turned on.)3) Is there a basic formula for calculating the time it would take for fuel to cool from the takeoff temp to the cruise temp? A full fuel load would not immedeiately cool to the outside temp, but would do so over a period of time. (Hope my wording makes sense)4) what else can you add to this line of questions? Please share your experiances.Thanks,John(P.S. these are follow-up questions to the long "fuel system tutorial" thread a little while back. Things are geting a little more complicated now trying to model a realistic code sequence.)

Share this post


Link to post
Share on other sites
Help AVSIM continue to serve you!
Please donate today!

>Hi,>>I'm working on a DC-9-41 project and I'm in the process of>modeling the fuel system.>>1) What is the freezing point of Jet-A fuel? A goggle search should find something. >>2) Once freezing point has been met, how fast will ice>accumulate and at what point would this seriously clog the>fuel filters? (Assuming the fuel heater were not turned on.) Generally, the fuel gels, gets too viscous to flow. I doubt it occurs suddenly. Possibly around -40C. The heaters may only be on the fuel tank outlets. >3) Is there a basic formula for calculating the time it would>take for fuel to cool from the takeoff temp to the cruise>temp? A full fuel load would not immedeiately cool to the>outside temp, but would do so over a period of time. (Hope my>wording makes sense) Ha, depends on physics of system. Specific Heats, thermal resistances, air flow over wing, sunlight, etc. Didn't you mention two hours to cool was typical? Anyway, here is my Time Constant algorithm in XML. Works very well. The first line calculates the values of the two 'weighting factors', wfa0 and wfb0. Note they add to 1.00. This is based on 18 ticks/second, so I start with '5', take the reciprocol, negate, exponentiate, then calculate wfa0 and wfb0. More recently I added a little bit so I could enter "TC" in seconds. You need something like 60*60*2 *18 (/18ticks/sec) seconds.5 s0 (>L:TCnor1,number) 1 l0 / /-/ exp s0 (>L:wfb1,number) 1.0 l0 - (>L:wfa1,number) (** ~TC=5/18.25 sec **) TCnor1 means 'per tick', would be in seconds if iterated once per second. wfb1 = e^(-1/5), wfa1 = 1.0- wfb1 The above is calculated only once, so could be done manually. Do Not recalculate every dos tick!I used G:Var1 here, but later changed to L:Variables. G:Var1 stores the last 'average' of the variable. First, check if it is zero, if true, set it to current 'input value'. This was written to average EGT_C.(G:Var1) 0 == if{ (L:EGT_C,number) (>G:var1) }(G:Var1) (L:wfb1,number) * (L:EGT_C,number) (L:wfa1,number) * + s0 (>G:Var1) l0 (>L:EGTavg,number) In algebraic notation: avg_value = avg_value * wfb1 + input_var * wfa1 EGTavg = avg_value Note if wfa1 is very small (0.01), each iteration takes 0.99 of old value and adds 0.01 of new value. This results in an exponential change. After 'one TC', the value has changed 63% of the way to its final value (assuming input is a step from zero to x). This works very well, takes only one saved variable, and has minimial computational overhead. One could even change the 'Time Constant' during accumulation. Or, cascade two or more 'time constants'. In fact, I have about seven of these 'exponential averaging' blocks in my Test Gauge to 'filter' some rates I calculate. For those familar with a bit of electronics, it is equivalent to an Resistor-Capacitor low pass filter (series R, shunt C), which has a TC = RC. In your case, rather than 'average and delay' EGT, you need to input a function of OAT. In fact, I'd suggest a weighted average of OAT and TAT, maybe 80% OAT and 20% TAT. One might then rapidly toggle the fuel on/off as the fuel reaches the gel temperature. Might have to have 'ignition' On or the first fuel cut might stop the turbine. Seems thurst would start dropping as fuel was starved, pilot would advance throttle, eventually run out of excess thrust. Ron

Share this post


Link to post
Share on other sites

>>2) Once freezing point has been met, how fast will ice>>accumulate and at what point would this seriously clog the>>fuel filters? (Assuming the fuel heater were not turned on.)>Generally, the fuel gels, gets too viscous to flow. I doubt it >occurs suddenly. Possibly around -40C. The heaters may only be on >the fuel tank outlets.Until I fully digest you reply... let me just add this to my problem statement.On the Dc-9 there are no fuel heaters in or near the tanks. (that I know of) Rather, there are fuel heaters in the Fuel Control Unit (FCU) to address the problem of "ice crystals" clogging the fuel filter as the result of water in the fuel. The heaters melt the accumulated ice restoring proper fuel flow.What I was thinking of doing was using a variables (L:FUEL_ICE_PERCENT1 & 2) to keep a running count of accumulated ice crystals. An annunciator light will come on when a pressure differental occurs at the fuel filter of approx. 3psi. (This is a warning to the pilot to turn on the fuel heaters and melt that ice.) I was also thinking of modeling in a Fuel PSI wariables for this. (L:FUEL_PSI1 & 2)When the fuel pressure drops due to icing of water that is in the fuel, how do I effect the engines in FS?What variables should I change (independant of throttle controls) that would force the pilot to advance the throttle to "maintain" speed?I'm using two jet engines (JT8D-11's) I want to see the ice crystals have an effect on the engines before thay actually cut out ...should the fuel heaters not be used.Thanks,John

Share this post


Link to post
Share on other sites

Roman,I've been going over your last reply ...lots to digest.Please start me off with the simple questions:>The above is calculated only once, so could be done manually.>Do Not recalculate every dos tick!How is this done? How do you prevent a calculation from being processed? How do you control the number of times a gauge is processed per second? Is there a way?Also,Your time constant, please explain it's purpose for me again and what exactly are wfb0 and wfb0? Thanks,John

Share this post


Link to post
Share on other sites

.........."Also,Your time constant, please explain it's purpose for me again and what exactly are wfb0 and wfb0? Thanks,John I'll get back to you on this. Or, email me. I orginally developed the 'filtering' algorithm in Fortran, it should be easy in any language. Some of the basic algorithms are in my XML Jet Test gauge, but I'm just getting it working in FS9. My test gauges also generate strobes that set how often different code blocks are calculated. Such strobes are available to all XML gauges used in the same panel. Ron

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this