XXXXA altitude constraints and ILS

[threexgreen 3,135]

Just flew the ILS into 12R at KMSP coming in via the KRUGG IAF. At G/S alive the A/P sent the aircraft into a steep dive in order to capture the G/S and I had to switch it off and fly a missed approach. Now I'm conducting my own NTSB investigation...

Between the KRUGG IAF and the FAF are two waypoints, EFEXX and ZESTY. The approach chart shows no altitude for EFEXX but 4000 for ZESTY. The FMC showed 6000A for EFEXX and 4000A for ZESTY, the latter being exactly what the chart said (underlined = do not go below?). I'm thinking I was too high because the aircraft crossed the waypoints "above" instead of exactly at the respective altitude, thus being too high to properly capture the G/S and sending the aircraft into a dive.

This has happened before although very seldom. Even when I get an FMC message informing me there's no descent path after XXXXX it usually works just fine. Now I'm wondering: is it correct practice to put in the exact altitudes into the FMC or leave the XXXXA altitudes be?

Thank you

 

chart http://155.178.201.160/d-tpp/1709/00264IL12R.PDF

[Olympic260 415]

I am usually putting  the exact altitudes, in that case I would put 4000 at ZESTY and 3000 at KINNS.  Most probably I would level off at 4000 and est the g/s there

[downscc 1,327]

I differ from Chris in that I'd leave it 4000A and might change the FAF to a hard 3000.  Like Chris, I would plan to arrive at 4000, capture LOC and GS all at ZESTY so the ats and aboves are a moot point. 

I just looked at the latest arrival on KMSP 12R and he was cleared to descend to 3000 until established on final so that is also a real world option.  ATC has the ability to clear you in lower than the charted route segments.

[threexgreen 3,135]

1 hour ago, downscc said:

I differ from Chris in that I'd leave it 4000A and might change the FAF to a hard 3000.  Like Chris, I would plan to arrive at 4000, capture LOC and GS all at ZESTY so the ats and aboves are a moot point. 

I just looked at the latest arrival on KMSP 12R and he was cleared to descend to 3000 until established on final so that is also a real world option.  ATC has the ability to clear you in lower than the charted route segments.

But if you leave ZESTY at 4000A you're not going to arrive at 4000, are you? I think I was higher than 4000 and above the G/S at that point so the A/P made the aircraft plunge to try and capture G/S.

On the second approach I was at 3000 and the aircraft plunged again when G/S became active...

 

[downscc 1,327]

22 minutes ago, threegreen said:

But if you leave ZESTY at 4000A you're not going to arrive at 4000, are you?

There is no reason why not.  The "A" means literally AT OR ABOVE so arriving AT might be the best vertical path the FMS can calculate.

[skelsey 765]

4000 at ZESTY to 3000 at the FAF in 3.1NM is almost exactly a 3 degree path, so it should be possible to establish at ZESTY as already mentioned.

Personally if I were flying this I'd be inclined to take it out of VNAV and instead fly it down in V/S for a nice continuous descent until established on the glideslope; VNAV with lots of constraints (which you should respect even if flying in V/S, incidentally: such restrictions are often in place to provide separation from traffic or terrain below, or to keep you in controlled airspace and out of 'bandit country') has a nasty tendency to screw you over by levelling off just when you don't really want to...

[voske 34]

11 hours ago, threegreen said:

Between the KRUGG IAF and the FAF are two waypoints, EFEXX and ZESTY. The approach chart shows no altitude for EFEXX but 4000 for ZESTY. The FMC showed 6000A for EFEXX and 4000A for ZESTY 

The segment between KRUGG and EFEXX has a minimum altitude of 6000 according to the chart. That's the reason why the FMC shows 6000A at EFEXX.

[downscc 1,327]

14 hours ago, skelsey said:

Personally if I were flying this I'd be inclined to take it out of VNAV and instead fly it down in V/S for a nice continuous descent until established on the glideslope;

Simon, I used to use V/S a lot but RSR chastised me and others including line pilots have pointed out the problems so I have switched to using LVL CHG (or FLCH) mode and use thrust to modulate my descent rate and MCP SPD to set pitch.  It takes a little practice but once you know about how much N1 to crank in for 800 fpm clean and dirty it gets second nature and you always have speed protection.

Otherwise, I agree with you on this.  About the only time I use VNAV all the way to FAF is on a RNAV/GNSS arrival with smooth transition to approach.  As for the MEA of 6000 mentioned by Marc..., watch the traffic and you'll see approach clearing flights to 3000 on the vector to intercept final course so there are many ways to simulate this.  One is to fly the full approach (a request that ATC will rarely hear but will grant if it's not busy) and the other with knowledge of how the traffic flow looks from a source such as Flightaware and use that.

[skelsey 765]

On 09/09/2017 at 4:38 PM, downscc said:

Simon, I used to use V/S a lot but RSR chastised me and others including line pilots have pointed out the problems so I have switched to using LVL CHG (or FLCH) mode and use thrust to modulate my descent rate and MCP SPD to set pitch.  It takes a little practice but once you know about how much N1 to crank in for 800 fpm clean and dirty it gets second nature and you always have speed protection.

Interesting. I have used FLCH with thrust to fine-tune the descent (mainly in the 'big sim' where we have the benefit of motorised thrust levers) but whilst I agree that V/S in the climb is a dangerous thing I've never heard of anybody discouraging its use in the descent. If anything I would have thought you have more protection in V/S in the descent compared to FLCH, not less; in V/S the A/T will always be in SPD so will always give you thrust on levelling off, whilst the 'FLCH trap' may leave you levelling off with idle thrust and no speed protection in  FLCH if you are not careful (ref Asiana).

[downscc 1,327]

18 minutes ago, skelsey said:

Interesting.

Yeah, it was and it took me awhile to get used to it.  I've read, maybe from RSR, that the VS use was prevalent until VNAV became the norm.  I still find the easiest way to arrive in the 777 is using FPA and I might not change from that but I've all but dropped the use of VS.

[threexgreen 3,135]

On 9/9/2017 at 0:14 AM, downscc said:

There is no reason why not.  The "A" means literally AT OR ABOVE so arriving AT might be the best vertical path the FMS can calculate.

I've observed on other approaches that VNAV used to cross points with an "A" at about 500ft above. As I couldn't remember what altitude I was at when the aircraft plunged I assumed VNAV crossed at something like 500 above instead of the exact altitude and therefore was a bit too high and tried to grab the G/S in a pretty wild way. Just a theory though. I still have no idea why the plunge.

This occasion makes it desirable to have a good, realistic ATC...