Narrative:

Deja vu all over again. On a previous leg this day; this crew; flying this arrival; in this aircraft; encountered the same problem. This time we thought we were better prepared. Enroute to lga on the haarp one arrival; we crossed bdl at 16;000 due to turbulence above and request by boston center due to traffic. Approximately 20 NM south of bdl; we were instructed to cross basye at 9;000. The wind was from the north; varying in direction from 340 to 360 true from 5 to 7 KTS; causing continuous light chop. The perf init page had been programmed for the standard descent speed of 290 KTS/.76M and a more conservative 2.5 degree angle. During the cruise and descent portions of the flight; full automation was utilized: the autopilot and auto throttles were engaged with FMS speed; LNAV and VNAV flight guidance modes. The 9;000 ft crossing restriction had previously been entered into the mcdu flight plan page and an additional restriction to cross 2 NM prior to basye at 9;000 (pseudo waypoint PBD01); and the altitude select was set to 9;000. Standard altitude awareness procedures were followed; including verifying 'big (font) pink (mcdu display) and TOD (nav display and prog page 2)'. In addition; the 9;000 restriction was displayed on the vertical profile on both navigation displays. I stated an estimated top of descent which the first officer confirmed. We then began to brief the approach and enter the landing data for the expressway visual approach into lga. During the briefing; the pink FMS vertical deviation indicator appeared followed by an aural alert. At the depicted TOD; the aircraft began its descent to 9;000. We resumed the briefing items and the checklist. We monitored the aircraft very closely to avoid a repeat of the previous issue. Up until 7 miles from valre intersection; all indications were normal; and we fully expected to meet altitude and speed restrictions. We had started the descent approximately 10 NM earlier than the previous leg. The FMS showed no deviation from the lateral or vertical path. Approximately 7 NM north of valre; the aircraft began a left turn to intercept the dpk 338 radial for the 158 course to basye and PBD01...a 103 degree course change. The distance from valre to basye is only 7 NM. We had added a fix 2 NM closer; leaving a segment of 5 NM; and the aircraft was flying an arc; cutting the corner. Now; it started the turn sooner than the last time! The FMS vertical deviation began dive; indicating we were too high. Midway through the turn; the lateral deviation indicated 2.6L (more than the previous leg). This coincided with my progress indication of 2.8 NM (also more than the previous leg) to valre. Until the aircraft had turned halfway and cycled the valre waypoint; followed immediately by PBD01. The EFIS and mcdu provided no indication that we would not cross basye at 9;000. Once valre was cycled; the aircraft was in a position approximately 3 miles north of bayse at 10;000 ft and 270 KTS. Even with full speed brakes; we were too high; crossing at 9;500 and 250 KTS. Aside from deploying the speed brake and overriding the throttles (which seemed slow to retard); we did not intervene. On the prior leg into lga; we had the exact same issue. This time; we programmed a 2.5 degree descent angle instead of the standard 3.0 degree angle and moved the restriction to 2 miles prior to basye; providing plenty of tolerance for an acceptable crossing. The alterations to the flight plan did nothing to change the outcome. Changing the descent angle only changed the TOD; thus only marginally improving our energy management. Placing a 9;000 crossing restriction at PBD01 did move our bod and TOD 2 NM; but it also forced the FMS to start its turn 2 NM earlier too! In short; our inputs were negated by the FMS logic to turn early and remain within 2 NM of the centerline. In the enroute mode; the FMS evidently calculates a descent based upon a linear model; in this case; approximately 8 miles were eliminated bythe leading turn. Full automation is simply not capable of flying the procedure as charted. Aside from modifying the haarp one arrival; or recoding the FMS software; the simple solution is to make a procedural change to amend the flight plan instructing the crew to reduce the level of automation: 1) calculate the TOD based on crossing valre at 9;000. 2) avoid VNAV descents on the arrival by using fpa or vertical speed. Perhaps reducing the rnp to .3 NM would provide an early alert; though probably not enough of a warning to prevent a clearance deviation. An FMS overfly at valre or prior to basye or moving the crossing restriction to valre would have helped (although not entirely eliminated the problem.

Related reports:

Google
 

Original NASA ASRS Text

Title: EMB190 Captain discovers that his aircraft's FMC cannot properly fly the HAARP 1 arrival into LGA from over BDL. The FMC begins a gentle turn well before VALRE and does not make the crossing restriction at BASYE.

Narrative: Deja Vu all over again. On a previous leg this day; this crew; flying this arrival; in this aircraft; encountered the same problem. This time we thought we were better prepared. Enroute to LGA on the HAARP ONE arrival; we crossed BDL at 16;000 due to turbulence above and request by Boston Center due to traffic. Approximately 20 NM south of BDL; we were instructed to cross BASYE at 9;000. The wind was from the north; varying in direction from 340 to 360 true from 5 to 7 KTS; causing continuous light chop. The PERF INIT page had been programmed for the standard descent speed of 290 KTS/.76M and a more conservative 2.5 degree angle. During the cruise and descent portions of the flight; full automation was utilized: the autopilot and auto throttles were engaged with FMS Speed; LNAV and VNAV flight guidance modes. The 9;000 FT crossing restriction had previously been entered into the MCDU flight plan page AND an additional restriction to cross 2 NM prior to BASYE at 9;000 (Pseudo waypoint PBD01); and the Altitude Select was set to 9;000. Standard altitude awareness procedures were followed; including verifying 'Big (font) Pink (MCDU display) and TOD (Nav display and Prog page 2)'. In addition; the 9;000 restriction was displayed on the Vertical Profile on BOTH navigation displays. I stated an estimated top of descent which the First Officer confirmed. We then began to brief the approach and enter the landing data for the Expressway Visual approach into LGA. During the briefing; the pink FMS vertical Deviation indicator appeared followed by an aural alert. At the depicted TOD; the aircraft began its descent to 9;000. We resumed the briefing items and the checklist. We monitored the aircraft very closely to avoid a repeat of the previous issue. Up until 7 miles from VALRE Intersection; all indications were normal; and we fully expected to meet altitude and speed restrictions. We had started the descent approximately 10 NM earlier than the previous leg. The FMS showed no deviation from the lateral or vertical path. Approximately 7 NM north of VALRE; the aircraft began a left turn to intercept the DPK 338 radial for the 158 course to BASYE and PBD01...a 103 degree course change. The distance from VALRE to BASYE is only 7 NM. We had added a fix 2 NM closer; leaving a segment of 5 NM; and the aircraft was flying an arc; cutting the corner. Now; it started the turn SOONER than the last time! The FMS Vertical Deviation began dive; indicating we were too high. Midway through the turn; the lateral deviation indicated 2.6L (more than the previous leg). This coincided with my PROGRESS indication of 2.8 NM (also more than the previous leg) to VALRE. Until the aircraft had turned halfway and cycled the VALRE waypoint; followed immediately by PBD01. The EFIS and MCDU provided no indication that we would not cross BASYE at 9;000. Once VALRE was cycled; the aircraft was in a position approximately 3 miles north of BAYSE at 10;000 FT and 270 KTS. Even with full speed brakes; we were too high; crossing at 9;500 and 250 KTS. Aside from deploying the speed brake and overriding the throttles (which seemed slow to retard); we did not intervene. On the prior leg into LGA; we had the exact same issue. This time; we programmed a 2.5 degree descent angle instead of the standard 3.0 degree angle and moved the restriction to 2 miles prior to BASYE; providing plenty of tolerance for an acceptable crossing. The alterations to the flight plan did nothing to change the outcome. Changing the descent angle only changed the TOD; thus only marginally improving our energy management. Placing a 9;000 crossing restriction at PBD01 did move our BOD and TOD 2 NM; but it also forced the FMS to start its turn 2 NM earlier too! In short; our inputs were negated by the FMS logic to turn early and remain within 2 NM of the centerline. In the enroute mode; the FMS evidently calculates a descent based upon a linear model; in this case; approximately 8 miles were eliminated bythe leading turn. FULL automation is simply not capable of flying the procedure as charted. Aside from modifying the HAARP ONE arrival; or recoding the FMS software; the simple solution is to make a procedural change to amend the flight plan instructing the crew to reduce the level of automation: 1) Calculate the TOD based on crossing VALRE at 9;000. 2) Avoid VNAV descents on the arrival by using FPA or vertical Speed. Perhaps reducing the RNP to .3 NM would provide an early alert; though probably not enough of a warning to prevent a clearance deviation. An FMS overfly at VALRE or prior to BASYE or moving the crossing restriction to VALRE would have helped (although not entirely eliminated the problem.

Data retrieved from NASA's ASRS site as of April 2012 and automatically converted to unabbreviated mixed upper/lowercase text. This report is for informational purposes with no guarantee of accuracy. See NASA's ASRS site for official report.