When I first saw the column force charts on FDR data, one thing that impressed me was the force applied was very ragged toward the end of the flight.
Control Column Force Applied
Then there was this comment by birdspeed which prompted an examination of the breakout mechanism which separates the two control columns if one side becomes jammed.
What perhaps happens is a breakout of the control yokes just before the dive. FDR shows a difference in the applied control force on the yokes. I believe when elevator breakout occurs there is sudden loss of elevator effectiveness. This would cause a sudden bunt over, with a wtf moment and the pilots stopped any sustained opposite trim.
There was then an interesting NTSB Safety Recommendation A-11-7 through -11 dated February 2010 which discussed several control jam incidents and indicated that the 737NG (and MAX I assume) had some additional features not found on the earlier 737-300 through 737-500 aircraft. and were
improved by the addition of several mechanical override mechanisms
. One form of these mechanical override mechanisms is called a POGO. See this link for an explanation
Pogo Load Limiter; Pogos are spring loaded linkages that act as a rigid link until a sufficiently high level of force is reached. Then the spring begins to allow motion and effectively disconnects mechanisms downstream of the POGO (which would normally have to be jammed for sufficient force to be generated.)
When you pull back on the control column inflight, you are not actually pulling back against the air loads on the elevator. That surface is being moved by an irreversible hydraulic cylinder. What you actually are working against is the elevator feel and centering unit. that applies force (torque) on the Elevator PCU input torque tube and which is actuating two elevator Power Control Unit hydraulic cylinders. One for the right elevator and the other for the left elevator. Both elevators are synchronized by the elevator torque tube. There is a pogo unit going to each elevator PCU control valve. Normally, the input valve force needed is rather light but should the PCU control valve jam, then its POGO will allow the rest of the input mechanism to move so that you can still actuate the opposite PCU control valve.
One thing I have not yet been able to establish is whether there are additional POGO-like devices installed between the individual control columns and the PCU input torque tube. If each column was connected directly to the PCU input torque tube, then activation of the breakout mechanism would seem to have minimal ability to free up a jam on one side of the control system forward of the elevator PCU input torquel tube since the control cables would back drive the opposite control column. I speculate that there are indeed POGO-like override mechanisms between the control columns and that they are set to a rather high force differential that would not be normally encountered. (More on this in a minute)
The elevator feel and centering unit applies its force profile on the PCU iinput torque tube by means of a linkage. As we can see from the force applied by the JT610 crew, they applied in excess of 100 lb force to the column during the final seconds. Suppose this level of force was sufficient to override the POGO unit on that side of the control pathway. leading to the elevator PCU input torque tube. If that happened, pulling further back on the control column would do nothing more! Remember, the trim had just reached full nose down. Could that be what the JT610 crew was trying to figure out by moving the controls, trying to find a spot that would "grab" and increase g? It would have felt like a control disconnect. Not the kind of thing that they would have had time to figure out once the nose began to drop.
For links to two more detailed posts on this subject in Tech Log see:
Cockpit End Discussion and
Elevator System Discussion