Valve closure for water-hammer effect with interpolation function

Hello,
Comsol has provided a water hammer benchmark problem for a 20 m long piping. This is solved using Water-hammer module. This module gives a pre-defined boundary condition 'Closed' and selects the 'node' at 'valve' location. I do not want to bring the velocity/flow rate zero at valve location at t=0 i.e. no instantaneous closure. I defined a linearly reducing flow rate from 1m/s to 0 m/s from 0 s to 0.03 s or 1 s or 10 s or 100 s. I create one more boundary condition named ' velocity' and assign the 'valve' node to it. Now the 'closed' boundary condition shows that the valve node is 'override'. After solving the problem and plotting pressure at valve location I have one doubt. The pressure plot starts at time t=0. If I provide a function that the velocity is 1 m/s for 0<t<10 s and 0 m/s for 10<t<20 then the pressure at valve location should not change for 10 s and the pressure rise should start from 10 s on wards. But the pressure plot at valve node shows pressure rise from t=0 s.

interpolation function

t f(t)
0 1
10 1
20 0


Interpolation: Liner
Extrapolation: Constant

Units:

Arguments: s
Function: m/s

Boundary condition:

Velocity: uin = u0*f(t[1/s]) m/s

where u0 is initial velocity u0=1m/s