2021‐01‐07

5.4 RESPONSE OF SECOND-ORDER PROCESSES

As noted in Chapter 4, a second-order transfer
function can arise physically whenever two first-order
processes are connected in series. For example, two
stirred-tank blending processes, each with a first-
order transfer function relating inlet to outlet mass
fraction, might be physically connected so that the
outflow stream of the first tank is used as the inflow
stream of the second tank. Figure 5.6 illustrates the
signal flow relation for such a process. Here

where 𝐾 𝐾1𝐾2. Alternatively, a second-order process transfer function will

arise upon transforming either a second-order differential equation process model
such as the two coupled first-order differential equations, for the CSTR (Eqs. 4-88
and 4-89). In this chapter

we consider second-order transfer functions in the standard form

1
2021‐01‐07

2
 2021‐01‐07

The underdamped form of Eq. 5-38 can arise from

some mechanical systems, from flow or other
processes such as a pneumatic (air) instrument line
with too little line capacity, or from a mercury
manometer, where inertial effects are important.
For process control problems the underdamped
form is frequently encountered in investigating the
properties of processes under feedback control.
Next we develop the relations for the step responses
of all three classes of second-order processes.

3
 2021‐01‐07

5.4.1 Step Response

For the step input (U(s) = M/s) to a process described by Eq. 5-38,

After inverting to the time domain, the responses can be categorized into three classes:

Overdamped (ζ> 1)
If the denominator of Eq. 5-46 can be factored using Eqs. 5-44 and 5-45, then the
response can be written

Critically Damped (ζ= 1)

Underdamped (0 ≤ ζ < 1)

4
 2021‐01‐07

Figure 5.9 illustrates the characteristics of the step

response of a second-order underdamped process.
The following terms are used to describe the
dynamics of underdamped processes:
1. Rise Time. 𝒕𝒓 is the time the process output takes
to first reach the new steady-state value.
2. Time to First Peak. tp is the time required for the
output to reach its first maximum value.
3. Settling Time. ts is the time required for the
process output to reach and remain inside a band
whose width is equal to ±5% of the total change in y
(±1% is also used for some applications).
4. Overshoot. OS = a/b (% overshoot is 100 a/b).
5. Decay Ratio. DR = c/a (where c is the height of the
second peak).
6. Period of Oscillation. P is the time between two
successive peaks or two successive valleys of the
response.

5
2021‐01‐07

6
 2021‐01‐07

EXAMPLE 5.5
A stirred-tank reactor has an internal cooling coil to remove heat liberated in the
reaction. A proportional controller is used to regulate coolant flow rate so as to
keep the reactor temperature essentially constant. The controller has been
designed so that the controlled reactor exhibits typical underdamped second-
order temperature response characteristics when it is disturbed, either by feed
flow rate or by coolant temperature changes.
(a) The feed flow rate to the reactor changes suddenly from 0.4 to 0.5 kg/s, and
the temperature of the reactor contents, initially at 100 C, changes eventually to
102 C. What is the gain of the transfer function (under feedback control) that
relates changes in reactor temperature to changes in feed flow rate? (Be sure to
specify the units.)
(b) The operator notes that the resulting response is slightly oscillatory with
maxima estimated to be 102.5 and 102.0 C occurring at times 1000 and 3060 s
after the change is initiated. What is the complete process transfer function?
(c) The operator failed to note the rise time. Predict 𝑡𝑟 based on the results in (a)
and (b).t

SOLUTION
(a) The gain is obtained by dividing the steady-state change in temperature by the
feed flow rate (disturbance) change (cf. Eq. 4-41):

(b) The oscillatory characteristics of the response can be used to find the dynamic
elements in the transfer function relating temperature to feed flow rate. Assuming the
step response is due to an underdamped second-order process, Figs. 5.7 and 5.10 can
be used to obtain estimates of ζ and τ. Alternatively, analytical expressions can be
used, which is the approach taken here. Either Eq. 5-51 or 5-52 can be employed to
find ζ independently of τ. Because the second peak value of temperature (102.5 C) is
essentially the final value (102 C), the calculated value of peak height c will be
subject to appreciable measurement error. Instead, we use the relation for overshoot
(rather than decay ratio) to take advantage of the greater precision of the first peak
measurement. Rearranging Eq. 5-51 gives

7
 2021‐01‐07

(c) The rise time tr can be calculated from Eq. 5-49. When t = 𝑡 , y(t) is equal to its
final steady-state value, KM. In other words, the bracketed quantity is identically
zero at t = tr:

8
 2021‐01‐07

5.4.2 Sinusoidal Response

When a linear second-order system is forced by a

sinusoidal input A sin ωt, the output for large values of
time (after exponential terms have become negligible) is
also a sinusoidal signal given by

9
 2021‐01‐07

EXAMPLE 5.6
An engineer uses a temperature sensor mounted in a thermowell to measure the
temperature in a CSTR. The temperature sensor/transmitter combination operates
approximately as a first-order system with a time constant of 3 s. The thermowell
behaves like a first-order system with a time constant of 10 s. The engineer notes
that the measured reactor temperature has been cycling approximately sinusoidally
between 180 ∘C and 183 ∘C with a period of 30 s for at least several minutes.
What can be concluded concerning the actual temperature in the reactor?

SOLUTION
First, note that the sensor/transmitter and the transmission line act as two first-
order processes in series (Eq. 5-37) with an overall gain K = 1, and with the
approximate transfer function

10
 2021‐01‐07

From the reported results, we conclude that some disturbance has caused the actual
reactor temperature (and its deviation) to vary sinusoid ally, which, in turn, has caused
he recorded output to oscillate. The cycling has continued for a period of time that is
much longer than the time constants of the process—that is, the instrumentation
system. Hence, the transients have died out and we can infer the conditions in the
reactor from the measured results, using Eq. 5-61 for the sinusoidal response of a
second-order system. From Eq. 5-65, τ = 3 s and τ = 10 s; τ and ζ are calculated
from Eqs. 5-40 and 5-41:

The frequency of the perturbing sinusoidal signal (reactor temperature) is calculated

from the observed period of 30 s:

The amplitude of the output perturbation also is obtained from observed results as

Equation 5-61 now can be rearranged to calculate the amplitude of the actual reactor
temperature
183 180
181.5 ∘C
2

from which A = 4.12 ∘C. Thus, the actual reactor temperature is varying between
(181.5−4.12) = 177.38 ∘C and (181.5+4.12) = 185.62 ∘C, nearly three times the
variation indicated by the measured value. Because the second-order process in this
example is overdamped (ζ = 1.19), we expect that sinusoidal perturbations in the
reactor temperature always will be attenuated (reduced in amplitude) in the
measurement system regardless of the frequency of the perturbation. Further
discussion of sinusoidal forcing is contained in Chapter 14 on frequency response
techniques.

11