Leaderboard

Got the CRJ working!

RFDS - Newcastle to Sydney CBD Appears through the haze... Lined Up Runway 25 YSSY

Hi everyone, this thread has been quiet all too long, and it is mostly our fault for not providing more frequent updates. This time we don´t have a new video for you, but things keep progressing nonetheless. We are still targeting a release in 2015 - again this is not a promise, but our very urgent desire. So where are we at? As you all know, the airplane is almost complete, save for some minor 3D items, like the outboard landing lights, movable cabin doors, a moving ouflowvalve vane, and a few more. The bulk of the documentation is also there, but it needs to be formatted into a nice document. The biggest chunk is still the FMS. We have the PERFormance part working, maintaining weights, takeoff page, approach page, etc. The RTE page and entry of a lateral route already works (as you could see in the latest videos). We are still missing the VNAV calculation, and the PROGgress calculation (estimating times, fuel usage). Two weeks ago, Tom Kyler, our lead FMS programmer, spent a week at my place in Germany, so we could crunch down on the VNAV part. We had done a similiar session a few months ago, where we worked on the VNAV climb part, now it was time to do the VNAV PTH and SPD descent. So what is VNAV? In a nutshell, VNAV is the vertical path that the plane takes along the lateral route. The climb, cruise, descent and missed approach phases. VNAV is pretty complicated, as the computations try to predict the vertical profile (a non-linear climb), and the descent (mostly an idle-power descent), provides altitude and speed predictions for the waypoints (therefore changing the lateral route again, due to changing turn radii) and heeding several different types of restrictions. The climb is "fairly" straightforward: The aircraft will climb with maximum climb power to the cruise altitude. It will, however, change airspeed at least three times, and possibly come up against altitude restrictions that delay the climb. All of this can be changed and modified by the pilot, so we need to take that into account as well. The descent is almost a climb in reverse, but the goal is now to target a three-dimensional spot in space, again changing speeds and descent regimes, ducking under and levelling off for altitude restrictions, taking into account all sorts of influencing factors (wind, weight, etc.). We also have to provide steering cues for the autopilot to do the right thing - climb while in the climb phase, level off for altitude restrictions and the cruise altitude, follow a three-dimensional path during the VNAV PTH descent, hitting deceleration segments, etc. It is absolutely mind-boggling - but we are on a good way to get it right! Tom has actually written the code to identify the E/D (end of descent) point correctly and reliably. It is "the last point before the end of route or route discontinuity with an AT altitude restriction". Try to put that into code. Phew. We hope to have VNAV in testing by early October. The CRZ page and PROG page are mere matters of coding crunch, adding up mileage, divide by groundspeed, multiply with fuel-flow... that kind of thing. Not trivial, but a piece of cake compared to the LNAV and VNAV computations. I really hope that many of you will exert the effort to understand and use the VNAV functionality of this aircraft - unlike most real-life pilots, who consider the VNAV button a "magic button" where you "never quite know what the airplane will do". That´s it for now - you can expect to see more frequent updates here from now on until we relase. Cheers, Jan