According to the NTSB report on the crash of United Airlines Flight 585, the ability of the 737’s lateral controls (ailerons and flight spoilers) to counteract the rolling forces produced by (say) a rudder hardover is considerably worse at lower flap settings, such that, in some cases, a hardover occurring with the airplane’s flaps fully extended might produce only minor control difficulties, whereas the same malfunction occurring with the flaps at (say) the ten-degree setting would result in a loss of directional control (pages 60-61 of the report; pages 78-79 of the PDF file of the report):
...Flap position and airspeed are important when determining controllability during a rudder hardover condition. With the rudder at about 25 degrees airplane nose right (ANR), the following conditions would exist at 150 to 160 knots calibrated airspeed (KCAS). Bank angles are noted as left or right wing down (LWD, RWD) and provide constant heading trim solution (no turns), except for the last case.
Rudder Angle …… Flaps …… Side Slip Angle …… Wheel Angle …… Bank Angle
25 ANR …………… 40 ……… 14 ANR ……………… 35 LWD ……… 18 LWD
25 ANR …………… 30 ……… 15 ANR ……………… 44 LWD ……… 17 LWD
25 ANR …………… 25 ……… 15 ANR ……………… 68 LWD ……… 16 LWD
23 ANR* ………… 15 ……… 17 ANR ……………… 107 LWD ……… 23 LWD
21 ANR* ………… 10 ……… 16 ANR ……………… 107 LWD ……… 19 LWD
25 ANR** ………… 10 ……… 13 ANR ……………… 107 LWD ……… 40 RWD
* Less than full rudder allowed to maintain directional control.
** Loss of directional control.
At 10 and 15 degrees of flap setting, heading cannot be maintained with full rudder deflection. If full right rudder is achieved with a 10-degree flap setting, for example, heading control is lost and, according to Boeing, a steady 40-degree right-wing-down trim solution is attained that results in turning flight to the right even with full left wheel deflection...
Why, for a given airspeed, do the 737’s lateral controls become increasingly ineffective as the flaps are retracted?