I'm trying to understand how weighted links work and could use some help. I should mention that I am attempting to use a weighted link to distribute a reaction force to a surface. I create a weighted link attached to a surface (for example the bottom of my part) and then I assigned the DOF to be the direction normal to the surface (say y). I then created another point offset from it along the y direction, connected the points with a spring and fixed the non link point. I thought the link would assign the average y displacement of the surface to the y displacement of the dependent point. This would create change in spring length which would translate to a spring force acting on the point, which would then be uniformly distributed back to the surface.
It appears to "kind of" work this way. In a test case I ran, I created a cylinder of diameter .25", applied 1000 lbs to one end and the other end I used a weighted link between the bottom surface and a point on the axis, offset .050" from the surface. I then used a1E6 lbf/in spring to a fixed point on the axis. Everything seemed to work as expected, the cylinder saw a uniform 20 ksi stress, the dependent point of the link moved .001" and the the reaction force as the fixed point was 1000 lbf.
I then moved the two points off axis by an inch, and re-ran the results to see if things would work the same (I thought they should be). My measures of dependent point displacement and reaction load were identical, .001" and 1000 lbf, but the stress fields and fringe plot of displacement make no sense (100s of ksi stress and 30" displacement).
Can anyone explain how the weighted link takes the displacement of the independent side and determines the dependent side, and then how it propagates forces? Is it a uniform distribution or is it non-uniform across the independent surface? Why does the location of the dependant point matter?
Thanks