Modeling a Process - Filling a Tank
The tank above is filled at a flow rate of Qin m3/sec which is the
input to the system. The output is the discharge flowrate,
Qout m3/sec. If Qin = Qout, the level, h, remains constant. If Qin >
Qout, the level, h, rises. If Qin < Qout, the level, h, falls. This much
is obvious but what exactly is the relationship between the flow
in, the flow out and the level.
The following equation is a mass balance that can be applied to
any system:
In - Out = Accumulation
In this case, the accumulation manifests itself as an increase or a
decrease in volume. Accumulation is the change in volume with
time.
Qin - Qout = V/t
Volume, V = area x height = A x h. The diameter, d, is a constant
because the walls of the tank are vertical and parallel. Therefore,
h is the only variable which means that V = h x A. Writing the
equation in differential form, we have:
Qin - Qout = A dh/dt
Next, consider the output flowrate, Qout. The driving force for the
discharge flow is the head of water in the tank which is given
by gh. The restriction to the discharge flow is the presence of
the valve (and to a lesser extent the pipe) and this can be
represented by a resistance, R, i.e.
Qout = gh/R
R has units of Newton seconds/metre5
Therefore, the above equation can be rewritten as follows:
Examination of this equation reveals that it has the following
form:
i.e. it has the first order characteristic where the output, o, is
equivalent to height, h; the input, i, is equivalent to the flow in,
Qin, and the time constant and gain are as follows:
The value of the resistance, R, depends upon the position of the valve. If
the valve is almost closed, R is large and if the valve is fully open, R is
small. R can be determined if the area available for discharge is known.
The above equation describes the relationship between the input and the
output of the system in the time domain. Converting this equation to the
frequency domain using a Laplace transform allows us to determine the
transfer function for this system. Once we know the transfer function we
can include this process as one block in a block diagram model of a
control loop.
The other benefit of determining the Laplace transform of the equation is
that we can choose a particular input (e.g. a step change in the input) and
multiply this by the transfer function to give the output in the frequency
domain. The equation is then converted back to the time domain and hey
presto we now have an equation that describes the response of the output
to a step change in the input.
Before continuing, understand the following:
What is a Laplace transform?
Transform a first order differential equation to the frequency
domain
Determine the transfer function of the process in the frequency
domain
Evaluate the Laplace transform of a unit step signal
Multiply the transfer function by a step input and transform the
result back to the time domain
Applying the Laplace Transform to the Tank Filling System
The first order differential equation derived above that describes how the
level in a tank changes with time for a given input is as follows:
where
Converting this equation to the frequency domain and rearranging in
terms of the output over the input gives the following:
Think of a step change in the input, let i(s) = a (=a/s in the frequency
domain). Substituting for i(s) and rearranging we get:
Applying an inverse Laplace transform, we get that:
where the time constant and steady state gain are the same as above.
