Hey,
Yesterday I was trying to do an RNP Y arrival for LHBP 31R. For those who can`t check the chart it is ODVAS (IAF, 180KT, 3000FT) - BP328 (IF, 180KT, 2300) - BP331 (165KT 1800) and RW31R.
Normally in Airbus logic the procedure here would be is to set the FCU ALT to 3000 feet (coincidentally this equals the missed app climb, so it is convenient) and BEFORE passing the IAF (ODVAS in this case) set APPR mode. FMA should be F-LOC and F-GS blue (as there is no FINAL APP mode in the 350) and from the IAF (ODVAS) it should follow the lateral and vertical profile. LS should be ON.
What happened is, that the lateral profile was followed correctly, but it kept levelled off at 3000 until i set 1300ft (MDA) and set DES mode again. It only activated F-LOC and F-GS when it war turning the final track, despite it should have been in that mode passing the IAF. Also, the APP phase DECEL point was set after BP331 (where you should be way below green dot) what is clearly wrong, as this is a 4.2 nm final, so if you are not in full config before 331 you should have a hard time as this point is 500ft above MDA.
Second attempt. I set 1300 on the FCU (i know it is wrong see below) and set APPR before ODVAS. It kept the 3000 constraint nicely, then after ODVAS dropped like brick to 1800 (disregarding the 2300 constraint) without even thinking of doing something like a CDA (CDFA), so I flew levelled off at 1800ft for a while until it crossed the 3 degree FLS G/S and started descending. The whole behaviour is similar to the Fenix, where you have FLS ON in the EFB and you turn on the LS it will never go to FINAL APP mode, instead it wil be in F-LOC F-APP so only the final track would work. But that is by design, as the A320ceo was not (yet) equipped with the upgraded Honeywell FMS)
Third and fourth attempt:
Did the same as first and second, but with LS off (in the A350 LS should be on according to FCOM, but I guessed it might be the same issue as the Fenix
Airbus FCOM clearly states that MDA should NEVER be set in the FCU to coordinate descent, so this can not be intentional. I guess it is the logic where the system identifies the start of the final approach phase and it thinks like it is an ILS where of course the final app is on the final track, but for RNP procedures (especially with turn to radius RNP and some RNP (AR) approaches with less then 1 nm final it is not the case. And the real thing can fly those.
Some reference: Airbus bulletin
Or, it is also possible, that I am doing something wrong here.
Can anyone point me into the right direction?