Hi Steve Apologies for the lack of clarity! I am adding some additional information below, so I hope this helps make things more clear.
The main interface I am working with looks something like the cross-section representation below:
Where zones 1, 2, and 3 are three separate fluid parts, and the yellow lines indicate the areas that I am applying system coupling between the fluid and solid models. The Piston moves down and up in one cycle. The fluid zone 1 extends from the top of the piston to the bottom of the cylinder, but the piston extends beyond the bottom of the cylinder.
The issue arises at the coupled interface between zone 1 and the Piston. It appears that ANSYS FSI is designed such that the fluid at the interface wants to follow the motion of the corresponding solid face. In this case, the fluid elements in zone 1 and at the yellow line above will follow the piston as it moves outside of the cylinder, resulting in the following profile:
As you may be able to see from the red rectangle at the bottom right of the cylinder, the fluid at the outlet and on the piston face extends beyond the bottom of the cylinder, resulting in elements with a similar face shape as the rhombus on the left. In the expanded image of the red rectangle on the right, you can see the dashed black line representing the bottom of the cylinder, and the blue-yellow spectrum representing the fluid. The fluid should stop at the bottom of the cylinder, as that is where the outlet is located.
This motion is not desired, and results in solver failure via floating point error in the fluid model. This is likely due to the elements along the piston stretching and thus having too large of an aspect ratio.
I also tried a simulation where the outlet was set to be stationary. This results in the profile below:
This solution diverges even sooner than the initial one, likely due to the fact that the fluid elements are still moving with the piston, but are limited in range to the bottom of the cylinder, and thus have a profile similar to the triangle to the left of the piston-cylinder interface.
So, after all of this, I ask:
Is there a way in ANSYS to have the FSI between the piston and fluid, such that the radial displacements of the piston are transferred to the fluid, but not the axial displacement of the piston?

Sometime after I asked my initial question, I viewed the Ansys Fluent Fluid Structure Interaction with Ansys Mechanical Learning Hub Training Module, and in the Lecture 07: System Coupling Workflows slide deck, it mentions Coupling with Deformable Interfaces can happen with Node Projection. This is enabled through the following TUI command: /define/mesh-interfaces/transfer motion across interfaces yes project nodes. It also comes with this image for reference:
As well as this 2D model as an example:
This approach is something I do not believe I have seen anywhere outside of these lecture slides. Since the material in this lecture was only briefly covered, and the slides themselves do not have much additional information, I thought it may be useful to ask the following questions here as well:
As the example model is in 2D, will the same workflow work for a 3D model as well? Or has this not yet been implemented?
It appears that there is only 1 fluid element in the radial direction of the thin fluid gap between the plunger and barrel. Can this be modified to include a greater number of elements in the radial direction, or is this limited to just 1? From my experience, you need at least 3 elements to fully form the fluid profile in this thin fluid space.
When should the TUI command be entered? Before or after initialization? How often does it need to be re-entered? Before every simulation, only when modifications are made to the simulation, or something in between?
Is there anywhere on the Ansys Help database that goes into more detail about this? As of now, I have never encountered this option outside of this slide deck.
Would it be possible for me to receive the example model to study? Some details are missing in the slides, such as the fluid mesh and the purpose of several of the dynamic mesh zones. Having the model available to play with would be very helpful!
To follow up after those 5 questions, is this the only/easiest method to allow for the FSI between the piston and fluid to not consider the axial motion of the piston?

I hope this helped clear up my initial question, and that you are able to answer at least some of the questions I have!
Henry