PFPX Wind Summary

[Hardy Heinlin]

This small difference between totalizer (tank sensors) and calculated fuel (FF summation) is normal and realistic; the former lags behind because the tank sensor system incorporates a great, artificial "inertia" so that changes of the aircraft attitude won't instantly change the totalizer display. (You can reset this lag on the Service page for sim setups).

In PSX, the tank fuel can only disappear through engines, APU, jettison, or activated fuel leak malfunctions. There are no other hidden possibilities.


Regards,

|-|ardy

[MFarhadi]

Tried resetting the lag, still didn't help with the missing fuel.
Also, an instance of the Instructor's Panel is attached to the image below after pressing the "Reset Lag" button. I had tried the button before during the flight, and the only times it did have an effect was on the ground, after (and during) refuelling (to stabilize the calculated amount, just as you mentioned due to the inertia).


Thus, I loaded a Cold and Dark situation file and just powered up the aircraft using EXT PWR (Leaving engines and APU off). I put 13,287 Kg of fuel in tanks 1 and 4, 18,000 Kg in tanks 2 and 3 and 4,017 Kg in Reserve tanks no. 2 and 3 (tallying up to 70,608 Kg so far).
Went to the upper overhead panel and put the guarded RESERVE 2 & 3 XFER switch to OPEN and waited around 58 minutes.
After the transfer had been completed, a total of 102 Kg of fuel was missing from the instructor's panel "Service" page (also tried Resetting the lag, but I believe the lag reset option is only meant to be used for the FQIS and EICAS Fuel Synoptics page).
If the experiment is done using time acceleration, the missing amount is reduced by a considerable margin. Don't know if RES 2 & 3 XFER takes longer or shorter while in the air (and how different the fuel difference might be), but I can run a "x 0 Fuel Flow" test in the air to see how much fuel goes missing while flying.


Before the Transfer:



Transfer Completed:

[Hardy Heinlin]

to stabilize the calculated amount

The reset button doesn't stabilize the calculated amount; it removes the present lag of the totalizer value. The calculated value is always stable as long as no fuel leak malfunction exist.

I just retested my Reset button in flight a few times, and it always worked.

As for the RES transfer with or without time acceleration: There may be tiny integer rounding errors in the long run. 50 kg tolerance in one RES transfer is really small; I would say this is certainly not a problem for flight planning :-)

The fuel network synchronization in PSX has a resolution of 0.1 pound. That is, 0.05 or 0.14999999 is rounded to 0.1 per fuel update. Updates are done at circa 0.3 Hz. In my opinion, this slight "imperfection" makes the simulation even more realistic. The real-world machine probably has much greater tolerances in its physical structures and electronics.


|-|ardy


P.S.: Just to be sure: When you shutdown the APU, it will keep running for a minute to cool down. That also uses some fuel.

[MFarhadi]

In my opinion, this slight "imperfection" makes the simulation even more realistic. The real-world machine probably has much greater tolerances in its physical structures and electronics.

You're entirely right on this one. Many times (many...) I've witnessed fuel hiding where it goes missing from a tank (by up to a ton or so) and then magically comes back after a while or even after landing. It adds to the complexion of the simulator, and it's beautiful.
But my point is that I suspected that might've been the cause of the disparity between FMC-predicted fuel estimates and what the plane actually had on board when reaching that point, especially the sudden drop after holding true for quite some time (if the fuel load is heavy enough).

Unfortunately, I haven't come up with any better explanations than the TOTALIZER/CALCULATED/FUEL USED figures in the PROGRESS page 2/4 regarding the matter to aid in expressing my point about the seemingly un-random (... if that's actually a word or something that I invented on the fly) continuously reducing EFOB and TOTALIZER falling behind the CALCULATED value.

However, I admire and appreciate your time and effort in helping me out with the problem.

[Hardy Heinlin]

You're welcome.

Hmmm ... I just revisited my code that runs the RES transfer, and I see it's not a rounding function but truncate function, i.e. fraction-round-down-only. I wrote that 15 years ago; I guess for the 0.1 pound resolution in the transfer I thought the truncate was better than rounding because when a disagreement occurs, some loss on the way would be more plausible than some additional pounds from nowhere :-)

Maybe I'll test it again with a rounding function ... (whether that is a good decision -- I don't know yet) ...


|-|


Edit:

OK, the rounding works, and I think I'll put that into PSX version 10.163. But I had to include a safety function that prefers the old truncate method over the rounding method when the tank update difference is so small that the tank's updated quantity is rounded up to its existing quantity while the actual FF fraction is still greater than zero, especially when the time acceleration is below 1 during a scavenge transfer from the center tank.

[simonijs]

Hi MR,

You may want to read this (https://www.infinidim.org/calculatedtotalizer-fuel-pre-fuelling), written by a Virgin Australia 777 pilot.*
The article was later updated by this one (https://www.infinidim.org/totalizer-vs-calculated-fuel).

Regards,
Simon

* Note the small error in this text: 0,5% of 8000 kg equals 40 kg, not 400 kg.

[Hardy Heinlin]

The networked fuel tank quantity rounding errors are now drastically reduced in PSX update 10.163:

https://aerowinx.com/board/index.php/topic,4191.0.html


Regards,

|-|ardy

[MFarhadi]

The networked fuel tank quantity rounding errors are now drastically reduced in PSX update 10.163

Thanks Hardy.

You may want to read this (https://www.infinidim.org/calculatedtotalizer-fuel-pre-fuelling), written by a Virgin Australia 777 pilot.*
The article was later updated by this one (https://www.infinidim.org/totalizer-vs-calculated-fuel).

Thanks for sharing the article, Simon.
Not only was it a good read, I already like the guy! Went ahead and followed him on Youtube.