You could try this: set number of steps = 2, and set the end time of step 1 to be immediately after your chirp excitation decays to zero. Set the end time of step 2 = 1.5e-3 s (the end of your transient):
There will then be a "split" in the chart after the chirp signal:
You can RMB on the part of the graph to the right of the split. A drop down should appear. Choose "Deactivate". If you do this, then if your excitation is an imposed displacment, it will be deleted after the split (the displacement will no longer be enforced).
I should say however, that if it is your intention to mimic piezo actuation, a force seems more realistic to me. If you haven't already done so, maybe you need to define damping appropriately so that the structure does not continue to "ring" after the applied force decays to zero.
Of course, using a Coupled Field Transient analysis system, you could model the piezoelectric bodies in detail, including the piezoelectric matrices that achieve coupling between stress and electric field. Your chirp excitation could then be applied as a voltage to the electrode surfaces of the piezoelectric bodies:
This would be the highest fidelity representation.