@C_R After adding the force sensor, the reading in the direction of gravity is 10 N.

It seems that the reason is considering the weight of the cable. Are there any other reasons? What are the other precautions when using the cable? What are the solutions for slow calculation? Will it still be slow when calling the calculation after exporting it as an FMU model?

Three-dimensional forces can be measured using force and moment sensors in the multibody. However, if the cable is not orthogonal to the axes, a force synthesis is needed, and it would be nice if the cable force measurement could be implemented using this summary_Tension.