Most probably the linearization is zero because you are computing it at the operating point where the orifice is closed. The tool "perturbs" the model a bit to calculate the difference between the operating point and the perturbed state. If the valve stays closed, you get zero linearization. One solution that often works in those cases is to run a model to a time where nothing is fully closed or fully open, and linearize at that time. For example, if you know that at 5 seconds into the simualtion, valves are partially open, in the Linear Analysis tool specify that you want to linearzie the model at simulation snapshot of 5 seconds, as discussed in the doc here.
Hope this helps. Arkadiy
