I have a geometrically nonlinear simulation of a beam, modeled with shell elements. The beam is loaded with *CONNECTOR LOADs applied to CONN3D2 elements. There are two connector elements in the model, both of type *AXIAL. One represents an actuator that loads the beam with vertical/lateral load (mostly vertical) and the the other represents an actuator that loads the beam with axial/vertical load (mostly axial). The reason I'm using connector elements/connector forces (instead of concentrated forces) is because I want the to force consistently point at a fixed anchor point, as the model deflects. See the figure below for a description of the model. The deflected shape of the model and the reaction forces appear reasonable. 

My question arises when I examine the symbol plots of connector force (CTF) for the two elements. The resultant CTF appears correct in magnitude and direction, but when I examine CTF2/CTF3, there are non-zero (and non-negligible) values. How is it possible for the AXIAL connector type to carry load in any direction other than its local X direction? It is noted that the same result exists regardless of if , DIRECTIONS=YES exists for the CTF field output request. 

EDIT: The figure below inadvertently states that the two couplings are distributed couplings but should say kinematic couplings (in the model they are kinematic couplings).