View Single Post
Old 5th April 2019, 12:14 PM
Philip Argy's Avatar
Philip Argy Philip Argy is offline
Senior Member
Join Date: May 2008
Location: Drummoyne
Posts: 1,387
Angry Bad software

As I understand it, MCAS was limited to 2.5 degrees of horizontal stabiliser deflection, which in theory could be overcome by manual nose-up input from the pilots. However, Boeing's SOP, which recommended deactivating the MCAS, did not explain that such a procedure just reset it rather than disabled it. If the same faulty AOA input recurred MCAS would then deflect the stabiliser up to 2.5 degrees. But here's the fatal flaw: it had no "memory" of previous deflections after each reset, and it had no provision for checking the current deflection, so each activation of MCAS ADDED ANOTHER 2.5 DEGREES of nose-down!!

That explains why both Lion Air and EA appeared to porpoise - manually regaining climb after the first deactivation but increasingly unable to do so as the jackscrew added each incremental 2.5 degrees of MCAS-commanded nose-down to its existing setting until the horizontal stabiliser was beyond the pilots' ability to counter with elevator nose-up commands. What is shocking is both the flaw in the software and Boeing's deliberate decision to withhold full information about MCAS so that they could market MAX as a no-conversion option for existing B738 pilots. As is now apparent, no B738 simulator incorporated MCAS and no pilot was aware how to fully disable MCAS to avoid it continually re-activating, assuming they would even appreciate the need to do so given the limits of the SOP Boeing had provided.

If my theory is confirmed by the investigation team Boeing's civil liability may be the least of its problems ...
Between 16R and 16L glide slopes (33.842252 S 151.153295 E) - wave as you fly over!

Last edited by Philip Argy; 5th April 2019 at 12:21 PM. Reason: Correct typos
Reply With Quote