nikit

Interesante lo que comenta uno de los creadores del FF A320.

Perdón por fusilarlo diréctamente del otro foro,quizás huviera sido mejor pegar el enlace simplemente.

20 hours ago, skitapa said:

can give out API access to this massive system

It's already presented and open. As i writed before, everything is exported to build in javascript engine that allows to make any kind of scripts. And you can share everything outside, using network sockets, for example. Each variable in each object already have a readable description that is possible to find using build in developer javascript console.

20 hours ago, skitapa said:

is not based on air resistance

Internal flight model, works like XP flight model (integrating forces of different parts), but with other math to properly simulate some things (actually is a some kind of VLM) and remove XP bugs. This is not a tables, it is wing aerodynamic profile and geometry based general solution with some polynomials for transonic region... exeact parameters are fitted using a huge matlab script to FCOM tables and other data sources.

20 hours ago, skitapa said:

plugins that handle weather

Can't say anything at this point, first task is to use XP weather model right.

20 hours ago, skitapa said:

native shared cockpit possible

This is an integral part of development. State save\load system, then full replay system, then network synchronization. May be we have only save\load at the first release point, but then we will expand...

On 15/5/2017 at 5:21 AM, cassis said:

is th flaps handle animagtion will be updated to the real one (lift and drag) ?  thanks

Animation is already updated (and it gets ~5 minutes of my time), but because of alot of such questions... 

This A320 is a completely different project from other FF models, and even on this market. With other level and priorities. We can say, that this is an experiment, of how real life technology can be embed in the sim. So, priorities are shifted from earning money and making happy of the most users, to identity of the simulation of the real aircraft. Not a fake identity of simple things, that it is possible to achieve by simple imitation to make something that “looks like an Airbus”, but fundamental simulation of the whole aircraft with all it's machines, computers, mechanics, electrics and the processes with in. Of course this is a very huge and complex work, consuming a lot of time, but we are in the end now (or at least nearly).

Because of system complexity and scale, there are a lot of minor bugs, malfunctions and problems (mainly animation, indication and minor logic). It is not possible to control everything during active development. In other hand, this product was continuously tested by the pilots, and some peoples that is very experienced in A320 simulations. And it is not contains any significant problems that can't be resolved fast without any significant modifications. So, this is a complete base, with most of very complex things (flight controls, autopilot, flight management, other systems) are implemented and working in, with the same technologies and logic as in real aircraft.

For example, it uses own aerodynamics because a lot of aircraft systems requires access to clean and isolated flight model or it's derivatives, and it is not possible to make things right without it. FACs are continuously computing aircraft CG and GW from ADIRS and engines data. There are not using direct, known values from the sim. It computes everything by itself, then using this info, computes reference speeds and other values for various aircraft systems. So... it is possible to completely reproduce, for example "XL Airways Germany Flight 888T" only by freezing two AoA sensors with the same conditions.

Now, having this base, we are tuning it to fit real aircraft as much as possible. Some things can work bad, even badly than in usual models, but this means only one thing. This thing is waiting for a tuning or some info to match completely. Some things are already works better than in real life training systems for pilots, because real traning systems, simulates some things in a fake manner (some transition effects and minor things). So this base allow us to make an exceptionaly good and identical aircraft. And this is a focus of our work. Of course, as a part of FF team, we are trying to make a user friendly product for most users, but this is not a MAIN priority. If you are looking for best of the best visual model, interchangeable engines, winglets, longer or shorter fuselage, sounds... then you need to wait for a while. Until other things with higher priority is finished. We do not want to make fakes, and this is a key priority. Aircraft systems, it's logic, then other things...

On 15/5/2017 at 10:02 AM, Gael Mebiame said:

I'm interested to see how much of the FMGC will actually be modeled. Will it be 100% 90% or 70% modeled?

First of all, there are different models of Flight Management systems in the real life. At this point, we are implementing Thales FM rev1 exacly like this http://www.737ng.co.uk/airbusfms.pdf (do not be confused about link name, just download). Expect a 100% availability of main functions. More over, alot of sweet things, are not described in this doc (like T/C symbol dynamic prediction in V/S mode base on selected V/S speed) but implemented. In other hand, some things have low priority, for example, predicitve GPS (alot of work to make things right for nearly nothing) or some alaternate functions, because of Navigraph database supply. Other things: SID/STAR legs management, lateral and vertical profile predictions, direct to, holds, steps, phase logics, constraints as it really should work, phase logics, dynamic nav interceptions, etc mostly finished and working, or exacly near the finish line and will be finished at the release point. Descend profile computation and guidance, as the most painfull thing for all Airbuses, works very good. Flight Guidance part, that is actually an autopilot, is implemented in 100% with all it's modes, control laws and works exacly as in real aircraft.

16 hours ago, skiselkov said:

Also, this wasn't a CAT III approach - do Airbusi even allow autoland if both A/Ps aren't engaged?

A320 supports two types of approaches. R-NAV (that is on the video), using on board navigation, down to minimum where the AP must be disengaged (and disengages by itself), and the ILS approaches, using ILS facality... to use autoland function, you need a CAT III C (no DH) landing facality. It is possible to use autoland with only one AP (so, you are engaging both at the start of approach, then one AP fails, for example, in 10FT above runway, but landing can be finished with rest), but your autority level will be degraded from CAT III fail operative (DUAL) to CAT III fail papssive (SINGLE). Will make a video with autoland soon.