Physical examples of first-order systems (Part I)

1
Chapter 5

Lecture 7
  In the first part of this chapter ,Chapter 5, we shall consider several
physical systems that can be represented by a first-order transfer
2
function.
 In the second part, a method for approximating the dynamic
response of a nonlinear system by a linear response will be
presented. This approximation is called linearization.
 Examples of first-order systems :
A. Liquid level :
 Consider the system shown in
Fig. 5.1, which consists of a tank of
uniform cross-sectional area 𝑨
to which is attached a flow
resistance 𝑹 such as a valve , a pipe, or a weir.
  𝒒𝒐 (volumetric flow rate) is related to the head (𝒉) by the linear
3 relationship
𝒉(𝒕)
 𝒒𝒐 (𝒕) = (𝟓. 𝟏)
𝑹
 A resistance that has this linear
relationship between flow and
head is referred to as a linear resistance.
 Note :
 A pipe is a linear resistance if the flow is in the laminar range.
 At unsteady-state, the mass balance becomes :
 𝐦𝐚𝐬𝐬 𝐟𝐥𝐨𝐰 𝐢𝐧 − 𝐦𝐚𝐬𝐬 𝐟𝐥𝐨𝐰 𝐨𝐮𝐭 = 𝐫𝐚𝐭𝐞 𝐨𝐟 𝐚𝐜𝐜𝐮𝐦𝐮𝐥𝐚𝐭𝐢𝐨𝐧 𝐨𝐟 𝐦𝐚𝐬𝐬 𝐢𝐧 𝐭𝐡𝐞 𝐭𝐚𝐧𝐤
𝒅 𝝆𝑨𝒉
 𝝆𝒒 𝒕 − 𝝆𝒒𝒐 𝒕 =
𝒅𝒕
𝒅𝒉
 𝒒 𝒕 − 𝒒𝒐 𝒕 = 𝑨 (𝟓. 𝟐),Combining 5.1 and 5.2 to eliminate 𝒒𝒐 (𝒕) giving
𝒅𝒕
:
 𝒉(𝒕) 𝒅𝒉
 𝒒(𝒕) − = 𝑨 (𝟓. 𝟑)
4 𝑹 𝒅𝒕
 Eq. 5.3 is a linear differential equation.
𝒅𝒉
 Initially, the process is operating at st-st, which means that =𝟎
𝒅𝒕
𝒉𝒔
 𝒒𝒔 − =𝟎 (𝟓. 𝟒)
𝑹
 Subtracting 5.4 from Eq. 5.3 :
𝟏 𝒅 𝒉−𝒉𝒔
 𝒒 − 𝒒𝒔 − 𝒉 − 𝒉𝒔 = 𝑨 (𝟓. 𝟓)
𝑹 𝒅𝒕
 𝑸 = 𝒒 − 𝒒𝒔
 𝑯 = 𝒉 − 𝒉𝒔
𝟏 𝒅𝑯
 𝑸 − 𝑯=𝑨 (𝟓. 𝟔)
𝑹 𝒅𝒕
𝑯 𝒅𝑯
 𝑸 = +𝑨
𝑹 𝒅𝒕
 Taking the Laplace transform of E.g. 5.6 :
5 𝟏
 𝑸 𝒔 = 𝑯 𝒔 + 𝑨 𝒔𝑯 𝒔 − 𝑯 𝟎 , 𝑯 𝟎 =𝟎
𝑹
𝟏
 𝑸 𝒔 = 𝑯 𝒔 + 𝑨𝒔𝑯 𝒔 (𝟓. 𝟕)
𝑹
𝑯 𝒔 𝑹
 = (𝟓. 𝟖)
𝑸 𝒔 𝝉𝒔+𝟏

 𝝉 = 𝑨𝑹 = 𝐭𝐢𝐦𝐞 𝐜𝐨𝐧𝐬𝐭𝐚𝐧𝐭
 In comparing the transfer function of the tank given by Eq. 5.8
with the transfer function for the thermometer (4.7)
𝒀 𝒔 𝟏
 [ = ]
𝑿 𝒔 𝝉𝒔+𝟏
 The term 𝑹 is simply the conversion factor that relates 𝒉(𝒕) to
𝒒(𝒕) when the system is at steady state.
 For unit-step change :
6  𝑸 𝒕 =𝒖 𝒕
𝟏
 𝑸 𝒔 =
𝒔
 By combining this forcing function with Eq. 5.8 :
𝟏 𝑹
 𝑯 𝒔 = ∗
𝒔 𝝉𝒔+𝟏
 Applying the final-value theorem :
𝐑
 𝐇 𝐭 = ∞ = 𝐥𝐢𝐦 𝐬𝐇 𝐬 = 𝐥𝐢𝐦 =𝐑
𝐬→𝟎 𝐬→𝟎 𝛕𝐬+𝟏
 This shows that the ultimate change in 𝑯(𝒕) for a unit change in 𝑸(𝒕) is simply 𝑹
at steady state.
 If the transfer function relating the inlet flow 𝒒(𝒕) to the outlet flow 𝒒𝟎 (𝒕) is
desired, note that we have from Eq. (5.1)
𝒉𝒔
 𝒒𝒐 𝒔 = (𝟓. 𝟗)
𝑹
 𝒉 𝒕
 Subtracting 5.9 from 5.1(𝒒𝒐 (𝒕) = (𝟓. 𝟏)) and using deviation variable
𝑹
7  𝒒𝒐 (𝒕) − 𝒒𝒐𝒔 =
𝒉(𝒕)
−
𝒉𝒔
=
𝒉(𝒕)−𝒉𝒔
𝑹 𝑹 𝑹
 𝑸𝒐 = 𝒒𝒐 (𝒕) − 𝒒𝒐𝒔 , 𝑯 = 𝒉 𝒕 − 𝒉𝒔
𝑯
 𝑸𝒐 = (𝟓. 𝟏𝟎)
𝑹
 Taking the transform of equation (5.10):
𝑯 𝒔
 𝑸𝒐 𝒔 = , 𝑯 𝒔 = 𝑹 𝑸𝟎 𝒔 (𝟓. 𝟏𝟏)
𝑹
 Combining 5.11 and 5.8 to eliminate H(s) gives :
𝑯 𝒔 𝑹
 = (𝟓. 𝟖)
𝑸 𝒔 𝝉𝒔+𝟏
𝑹 𝑸𝟎 𝒔 𝑹 𝑸𝟎 𝒔 𝟏
 = , =
𝑸(𝒔) 𝝉𝒔+𝟏 𝑸(𝒔) 𝝉𝒔+𝟏
 Recall that the unit−impulse function is defined as pulse of unit area as the duration of the
pulse approaches zero, and the impulse function can be approximated by suddenly increasing
the flow to a large value for a very short time; that is, we may pour very quickly a volume of
liquid into the tank.
 The nature of the impulse response for a liquid-level system will be described by the following
example.
  Example 5.1. A tank having a time constant of 𝟏 min and a
𝟏 𝒇𝒕.𝒎𝒊𝒏
8 resistance of 𝟑 is operating at steady state with an inlet flow
𝟗 𝒇𝒕
𝒇𝒕𝟑 𝒇𝒕𝟑
of 𝟏𝟎 . At time 𝒕 = 𝟎, the flow is suddenly increased to 𝟏𝟎𝟎
𝒎𝒊𝒏 𝒎𝒊𝒏
for 𝟎. 𝟏 min by adding an additional 𝟗 𝒇𝒕𝟑 of water to the tank
uniformly over a period of 𝟎. 𝟏 min . (Fig. 5.1 for this input
disturbance).
 Plot the response in tank level and compare with the impulse
response.
 Solution :
 𝝉 = 𝟏 𝒎𝒊𝒏
𝟏 𝒇𝒕.𝒎𝒊𝒏
 𝑹=
𝟗 𝒇𝒕𝟑
𝑯 𝒔 𝑹
 =
𝑸 𝒔 𝝉𝒔+𝟏
 The input may be expressed as the difference in step functions.
 𝑸 𝒕 = 𝟗𝟎𝒖 𝒕 − 𝟗𝟎𝒖 𝒕 − 𝟎. 𝟏 = 𝟗𝟎[𝒖 𝒕 − 𝒖(𝒕 − 𝟎. 𝟏)] pulse function
  The transfer function of this is :
𝟗𝟎 𝟗𝟎 𝟗𝟎
9  𝑸 𝒔 =
𝒔
− 𝒆−𝟎.𝟏𝒔
𝒔
=
𝒔
(𝟏 − 𝒆−𝟎.𝟏𝒔 )
𝟏 𝟏 𝟗𝟎
 𝑯 𝒔 = ∗ ∗ (𝟏 − 𝒆−𝟎.𝟏𝒔 )
𝟗 𝒔+𝟏 𝒔
𝟏 𝒆−𝟎.𝟏𝒔
 𝑯 𝒔 = 𝟏𝟎( − ) (5.13)
𝒔 𝒔+𝟏 𝒔(𝒔+𝟏)

 According to the theorem on translation of functions:

 𝑳−𝟏 𝒆−𝒔𝒕𝟎 𝑭 𝒔 = 𝒇 𝒕 − 𝒕𝟎
𝒆−𝟎.𝟏𝒔
 𝑳−𝟏 =𝟎 𝒕 < 𝟎. 𝟏
𝒔 𝒔+𝟏

= 𝟏𝟎 𝟏 − 𝒆− 𝒕−𝟎.𝟏
𝒕 > 𝟎. 𝟏
 𝑯 𝒕 = 𝟏𝟎 𝟏 − 𝒆−𝒕 − 𝟎, 𝒕 < 𝟎. 𝟏
 𝑯 𝒕 = 𝟏𝟎 𝟏 − 𝒆−𝒕 − 𝟏 − 𝒆− 𝒕−𝟎.𝟏 , 𝒕 > 𝟎. 𝟏 (𝟓. 𝟏𝟒)
10
  The response of the system to an impulse of magnitude 𝟗𝟎 ∗ 𝟎. 𝟏 = 𝟗
11  𝑸 𝒕 = 𝟗𝜹 𝒕 ,
 𝑸 𝒔 =𝟗
𝑯 𝒔 𝑹 𝟏 𝟏
 = = ∗
𝑸 𝒔 𝝉𝒔+𝟏 𝟗 𝒔+𝟏
𝟏 𝟏 𝟏
 𝑯 𝒔 = ∗𝟗 =
𝟗 𝒔+𝟏 𝒔+𝟏
−𝒕
 𝑯 𝒕 =𝒆

 The responses to step and sinusoidal forcing functions are the same for the
liquid-level system as for the mercury thermometer.
 This is the advantage of characterizing all first-order systems by the same
transfer function.
 B. Liquid-Level process with constant-flow outlet :
12
 The resistance shown in Fig.5.3 is replaced by a constant-flow
pump.
 The same assumptions of constant cross-sectional area and
constant density that were used before also apply here.
  For this system, Eq. (5.2) still applies, but 𝒒𝟎 is now a constant;
thus
13
𝒅 𝝆𝑨𝒉
 𝝆𝒒 𝒕 − 𝝆𝒒𝒐 = (𝟓. 𝟐)
𝒅𝒕
𝒅𝒉
 𝒒 𝒕 − 𝒒𝒐 = 𝑨 (𝟓. 𝟏𝟓)
𝒅𝒕
 At st-st :
 𝒒𝒔 = 𝒒𝒐
 𝒒𝒔 − 𝒒𝒐 = 𝟎 (𝟓. 𝟏𝟔)
 Subtracting 5.16 from 5.15 and introducing deviation variables :
𝒅 𝒉−𝒉𝒔
 𝒒 𝒕 − 𝒒𝟎 − 𝒒𝒔 + 𝒒𝟎 = 𝒒 𝒕 − 𝒒𝒔 = 𝑨
𝒅𝒕
 𝑸 = 𝒒 𝒕 − 𝒒𝒔 , 𝑯 = 𝒉 𝒕 − 𝒉𝒔
𝒅𝑯
 𝑸= 𝑨 (𝟓. 𝟏𝟕)
𝒅𝒕
 Taking the Laplace transform for both sides :
14  𝑸 𝒔 = 𝑨 𝒔𝑯 𝒔 − 𝑯 𝟎 ,𝑯 𝟎 = 𝟎
𝑯 𝒔 𝟏
 = (𝟓. 𝟏𝟖)
𝑸 𝒔 𝑨𝒔
𝒕 𝑭 𝒔
 Note : 𝑳 𝟎
𝒇 𝒕 𝒅𝒕 =
𝒔
𝟏𝑸 𝒔
 𝑯 𝒔 =
𝑨 𝒔
𝟏 𝒕
 𝑯 𝒕 = 𝑸 𝒕 𝒅𝒕
𝑨 𝟎
𝟏 𝒕
 𝒉 𝒕 − 𝒉𝒔 = 𝑸 𝒕 𝒅𝒕 𝟓. 𝟏𝟗
𝑨 𝟎
 If step-change 𝑸 𝒕 = 𝒖 𝒕 = 𝟏
𝟏 𝒕 𝒕
 𝒉 𝒕 − 𝒉𝒔 = 𝟎 𝒖(𝒕)𝒅𝒕 =
𝑨 𝑨
𝒕
 𝒊. 𝒆 𝑯 𝒕 = (𝟓. 𝟐𝟎) 𝐫𝐚𝐦𝐩 𝐟𝐮𝐧𝐜𝐭𝐢𝐨𝐧
𝑨
  The step response given by Eq.(5.20) is a ramp function
𝒕
15 that grows without limit. 𝑯 𝒕 = 𝟓. 𝟐𝟎
𝑨
 The transfer function for the liquid-level system with constant
𝑯 𝒔 𝟏
outlet flow given by Eq.(5.18) ( = ) can be considered as a
𝑸 𝒔 𝑨𝒔
𝑯 𝒔 𝑹
special case of Eq. (5.8), ( = ) as 𝑹 → ∞.
𝑸 𝒔 𝝉𝒔+𝟏
𝑹 𝑹 𝟏
𝑯 𝒔 𝑹 𝟏
 = = 𝝉
𝟏 = 𝑨𝑹
𝟏 = 𝑨
𝟏 = 𝟓. 𝟏𝟖
𝑸 𝒔 𝝉𝒔+𝟏 𝒔+ 𝒔+ 𝒔+ 𝑨𝒔
𝝉 𝑨𝑹 ∞
 C.Mixing process :
16
 Consider the mixing process shown in Fig.5.4 in which a stream of
solution containing dissolved salt flows at a constant volumetric
flow rate 𝒒 into a tank of constant holdup volume 𝑽.
 The concentration of the salt in the entering stream, 𝒙 (mass of salt/volume),
varies with time.
 It is desired to determine the transfer function relating the outlet concentration
𝒚 to the inlet concentration 𝒙.
  Assuming the solution density is constant,
17  by making balance on the salt :
 (Flow rate of salt in) – (flow rate of salt out) = rate of accumulation
of salt in the tank
𝒅 𝑽𝒚
 𝒒𝒙 − 𝒒𝒚 = (𝟓. 𝟐𝟏)
𝒅𝒕
 We shall again introduce the deviation variables
 At st-st : 𝒒𝒙𝒔 − 𝒒𝒚𝒔 = 𝟎 (𝟓. 𝟐𝟐)
 Subtracting Eq.5.22 from 5.21 and introducing the deviation variables :
 𝑿 = 𝒙 − 𝒙𝒔 , 𝒀 = 𝒚 − 𝒚𝒔
𝒅𝒀
 𝒒𝑿 − 𝒒𝒀 = 𝑽 𝒅𝒕

 Taking the Laplace transform :

 𝒒𝑿 𝒔 − 𝒒𝒀 𝒔 = 𝑽 𝒔𝒀 𝒔 − 𝒀 𝟎 , 𝒀 𝟎 =𝟎
𝑽
 𝝉=𝒒 𝑿 𝒔 − 𝒀 𝒔 = 𝝉𝒔𝒀(𝒔)
  𝑿 𝒔 = 𝒀 𝒔 𝝉𝒔 + 𝟏
18 𝒀 𝒔 𝟏
 = (𝟓. 𝟐𝟑)
𝑿 𝒔 𝝉𝒔+𝟏
 This mixing process is another first-order process.
 Ex: A stirred tank is used for the dilution of a concentrated salt solution
𝒌𝒈
containing 𝟒𝟎% 𝐛. 𝐰 salt at a rate of 𝟏𝟎𝟑 𝟑 to a solution containing 𝟐𝟎%
𝒎
𝐬𝐚𝐥𝐭 𝐛. 𝐰.
𝒎𝟑
 Pure water is added for this purpose at a rate of . If the
𝟏
𝒉
concentration of the inlet salt stream is suddenly increased to 𝟓𝟎% 𝐛. 𝐰.
a. What will be the new st.st concentration of the outlet stream ?
b. How long will it take for the outlet concentration to reach 𝟗𝟓% of its new
st.st value ?
 Assume that:
 Constant 𝝆 and equal to that of 𝐇𝟐 𝐎 (𝝆 = 𝟏𝟎𝟎𝟎 𝒌𝒈/𝒎𝟑 )
 Constant hold up in tank (𝐯 = 𝐜𝐨𝐧𝐬𝐭.)
 Solution :
19  Total material balance :
 𝒒 = 𝟏𝟎𝟎𝟎 + 𝟏𝟎𝟎𝟎 = 𝟐𝟎𝟎𝟎 𝒌𝒈/𝒉
 Salt balance under steady state
 𝟏𝟎𝟎𝟎𝒙𝒔 − 𝟐𝟎𝟎𝟎𝒚𝒔 = 𝟎 ,
 𝒙𝒔 − 𝟐𝒚𝒔 = 𝟎 (𝟏)
 Salt balance under un-steady state
 Input-output = rate of Acc.
𝒅 𝝆𝑽𝒚
 𝟏𝟎𝟎𝟎𝒙 − 𝟐𝟎𝟎𝟎𝒚 =
𝒅𝒕
𝒅𝒚 𝒅𝒚 𝒅𝒚
 𝟏𝟎𝟎𝟎𝒙 − 𝟐𝟎𝟎𝟎𝒚 = 𝝆𝑽 = 𝟏𝟎𝟎𝟎 ∗ 𝟓 = 𝟓𝟎𝟎𝟎
𝒅𝒕 𝒅𝒕 𝒅𝒕
𝒅𝒚
 𝒙 − 𝟐𝒚 = 𝟓 (𝟐)
𝒅𝒕
  Subtract (2) from (1)
20 𝒅(𝒚−𝒚𝒔 )
 𝒙 − 𝒙𝒔 − 𝟐 𝒚 − 𝒚𝒔 = 𝟓
𝒅𝒕

 Deviation variables :
 𝑿 = 𝒙 − 𝒙𝒔 ,
 𝒀 = 𝒚 − 𝒚𝒔
𝒅 𝒀
 𝑿 − 𝟐𝒀 = 𝟓
𝒅𝒕
 𝑿 𝒔 − 𝟐𝒀 𝒔 = 𝟓 𝒔𝒀 𝒔 − 𝒀 𝟎 , 𝒀 𝟎 =𝟎
 𝐗(𝐬) = 𝐘 𝐬 𝟐 + 𝟓𝐬
𝒀 𝒔 𝟏 𝟎.𝟓
 = =
𝑿 𝒔 𝟓𝒔+𝟐 𝟐.𝟓𝒔+𝟏
  𝐒𝐭𝐞𝐩 − 𝐜𝐡𝐚𝐧𝐠𝐞 𝐨𝐟 𝐦𝐚𝐠𝐧𝐢𝐭𝐮𝐝𝐞 = 𝟎. 𝟓 − 𝟎. 𝟒 = 𝟎. 𝟏
21  𝑿 𝒕 = 𝟎. 𝟏𝒖(𝒕)
𝟎.𝟏
 𝑿 𝒔 =
𝒔
𝟎.𝟓∗𝟎.𝟏 𝟎.𝟎𝟓
 𝒀 𝒔 = =
𝒔(𝟐.𝟓𝒔+𝟏) 𝒔(𝟐.𝟓𝒔+𝟏)

 𝒀 𝒕 = 𝟎. 𝟎𝟓 𝟏 − 𝒆−𝒕/𝟐.𝟓 = 𝟎. 𝟎𝟓(𝟏 − 𝒆−𝟎.𝟒𝒕 )

 The new st.st condition is :
 𝒂𝒕 𝒕 → ∞ 𝒀 𝒕 = 𝟎. 𝟎𝟓 = 𝒚𝒔𝒏𝒆𝒘 − 𝒚𝒔𝒐𝒍𝒅 = 𝒚𝒔𝒏𝒆𝒘 − 𝟎. 𝟐
 𝒚𝒔𝒏𝒆𝒘 = 𝟎. 𝟐𝟓, 𝒙𝒔𝒏𝒆𝒘 = 𝟎. 𝟓
 𝒚 = 𝟎. 𝟗𝟓𝒚𝒔𝒏𝒆𝒘 = 𝟎. 𝟗𝟓 ∗ 𝟎. 𝟐𝟓 = 𝟎. 𝟐𝟑𝟕𝟓
 𝒀 𝒕 = 𝟎. 𝟐𝟑𝟕𝟓 − 𝟎. 𝟐 = 𝟎. 𝟎𝟑𝟕𝟓
 𝒀 𝒕 = 𝟎. 𝟎𝟓(𝟏 − 𝒆−𝟎.𝟒𝒕 )
 𝟎. 𝟎𝟑𝟕𝟓 = 𝟎. 𝟎𝟓(𝟏 − 𝒆−𝟎.𝟒𝒕 )
 𝒆−𝟎.𝟒𝒕 = 𝟎. 𝟐𝟓 , −𝟎. 𝟒𝒕 = −𝟏. 𝟑𝟖𝟔, 𝒕 = 𝟑. 𝟒𝟔 ℎ
 D. Heating Process
22  Consider the heating process shown in Fig. 5–5 . A stream at temperature 𝑻𝒊 is fed
to the tank. Heat is added to the tank by means of an electric heater. The tank is
well mixed, and the temperature of the exiting stream is 𝑻. The flow rate to the tank
is constant at 𝒘 , 𝑰𝒃𝒎 𝒉.
 A transient energy balance on the tank yields
 {𝐑𝐚𝐭𝐞 𝐨𝐟 𝐞𝐧𝐞𝐫𝐠𝐲 𝐟𝐥𝐨𝐰 𝐢𝐧𝐭𝐨 𝐭𝐚𝐧𝐤} − {𝐑𝐚𝐭𝐞 𝐨𝐟 𝐞𝐧𝐞𝐫𝐠𝐲 𝐟𝐥𝐨𝐰 𝐨𝐮𝐭 𝐨𝐟 𝐭𝐚𝐧𝐤}
+ 𝐑𝐚𝐭𝐞 𝐨𝐟 𝐞𝐧𝐞𝐫𝐠𝐲 𝐟𝐥𝐨𝐰 𝐢𝐧 𝐟𝐫𝐨𝐦 𝐡𝐞𝐚𝐭𝐞𝐫} = {𝐑𝐚𝐭𝐞 𝐨𝐟 𝐚𝐜𝐜𝐮𝐦𝐮𝐥𝐚𝐭𝐢𝐨𝐧 𝐨𝐟 𝐞𝐧𝐞𝐫𝐠𝐲 𝐢𝐧 𝐭𝐚𝐧𝐤}
 Converting this energy balance to symbols results in
𝒅(𝑻−𝑻𝒓𝒆𝒇 ) 𝒅𝑻
 𝒘𝑪 𝑻𝒊 − 𝑻𝒓𝒆𝒇 − 𝒘𝑪 𝑻 − 𝑻𝒓𝒆𝒇 + 𝒒 = 𝝆𝑽𝑪 𝒅𝒕
= 𝝆𝑽𝑪 𝒅𝒕 (5.24)
 Where 𝑻𝒓𝒆𝒇 is the reference temperature and 𝑪 is
the heat capacity of the fluid.
𝒅𝑻
 At steady state, is zero, and Eq. (5.24) can be written
𝒅𝒕
 𝒘𝑪 𝑻𝒊𝒔 − 𝑻𝒔 + 𝒒𝒔 = 𝟎 (5.25)
 Where the subscript 𝒔 has been used to indicate steady state.
 Subtracting Eq. (5.25) from Eq. (5.24) gives
𝒅 𝑻−𝑻𝒔
 𝒘𝑪 𝑻𝒊 − 𝑻𝒊𝒔 − 𝒘𝑪 𝑻 − 𝑻𝒔 + 𝒒 − 𝒒𝒔 = 𝝆𝑽𝑪 (𝟓. 𝟐𝟔)
𝒅𝒕
  If we assume that 𝑻𝒊 is constant (𝑻𝒊 = 𝑻𝒊𝒔 ) and make disturbance in the
23 outlet temperature(𝑻) and introduce the deviation variables in 𝑻
 𝑻′ = 𝑻 − 𝑻𝒔
 𝑸 = 𝒒 − 𝒒𝒔
 Eq. (5.26) becomes
′ 𝒅𝑻′
 −𝒘𝑪𝑻 + 𝑸 = 𝝆𝑽𝑪 (𝟓. 𝟐𝟕)
𝒅𝒕
 Taking Laplace transforms of Eq. (5.27) gives
 −𝒘𝑪𝑻′ 𝒔 + 𝑸 𝒔 = 𝝆𝑽𝑪𝑻′ (𝒔) (𝟓. 𝟐𝟖)
 Rearranging Eq. (5.28) produces the following first-order transfer function relating 𝑻′ 𝒔
and 𝑸(𝒔):
𝑻′ (𝒔) 𝟏 𝑲
 = 𝝆𝑽
𝒘𝑪
= (5.29)
𝑸(𝒔) 𝒘 𝒔+𝟏 𝝉𝒔+𝟏

 Thus, this process exhibits first-order dynamics as the tank temperature 𝑻 responds to
changes in the heat input to the tank.
  Example 5.2. Consider the mixed tank heater shown in Fig. 5–6 . Develop a transfer
function relating the tank outlet temperature to changes in the inlet temperature.
24
Determine the response of the outlet temperature of the tank to a step change in the inlet
temperature from 𝟔𝟎 to 𝟕𝟎𝟎 𝑪 . Before we proceed, intuitively what would we expect to
happen? If the inlet temperature rises by 𝟏𝟎𝟎 𝑪, we expect the outlet temperature to
eventually rise by 𝟏𝟎𝟎 𝑪 if nothing else changes. Let’s see what modeling the process will
tell us.
 From Eq. (5.26) we can write the following simplified balance, realizing that
 𝒒 = 𝒒𝒔
𝒅 𝑻−𝑻𝒔
 𝒘𝑪 𝑻𝒊 − 𝑻𝒊𝒔 − 𝒘𝑪 𝑻 − 𝑻𝒔 + 𝒒 − 𝒒𝒔 = 𝝆𝑽𝑪
𝒅𝒕
𝒅 𝑻−𝑻𝒔
 𝒘𝑪 𝑻𝒊 − 𝑻𝒊𝒔 − 𝒘𝑪 𝑻 − 𝑻𝒔 = 𝝆𝑽𝑪
𝒅𝒕

 In terms of deviation variables, this becomes

′ 𝒅𝑻′
 𝒘𝑪𝑻𝒊 − 𝒘𝑪𝑻′ = 𝝆𝑽𝑪
𝒅𝒕

 𝒘𝑪𝑻𝒊 ′ (𝒔) − 𝒘𝑪𝑻′ 𝒔 = 𝝆𝑽𝑪[𝒔𝑻′ 𝒔 − 𝑻′ 𝟎 ] (𝟓. 𝟐𝟖)

 𝒘𝑪𝑻𝒊 ′ (𝒔) − 𝒘𝑪𝑻′ 𝒔 = 𝝆𝑽𝑪𝒔𝑻′ 𝒔
  Finally, after rearranging,
25 𝑻′ (𝒔) 𝟏 𝟏
 = =
𝑻𝒊 ′ (𝒔) 𝝆𝑽
𝒔+𝟏 𝝉𝒔+𝟏
𝒘

 Substituting in numerical values for the variables, we obtain the

actual transfer function for this mixed tank heater.
𝛒𝐕 𝐕 𝐕 𝐭𝐚𝐧𝐤 𝐯𝐨𝐥𝐮𝐦𝐞 𝟏𝟎𝟎𝟎 𝐋
 𝛕= = = = = = 𝟓 𝐦𝐢𝐧
𝐰 𝐰/𝛒 𝐯 𝐯𝐨𝐥𝐮𝐦𝐞𝐭𝐫𝐢𝐜 𝐟𝐥𝐨𝐰 𝐫𝐚𝐭𝐞 𝟐𝟎𝟎𝐋/𝐦𝐢𝐧
𝑻′ (𝒔) 𝟏 𝟏
 = =
𝑻𝒊 ′ (𝒔) 𝝆𝑽
𝒔+𝟏 𝟓𝒔+𝟏
𝒘

 If the inlet temperature is stepped from 𝟔𝟎 − 𝟕𝟎𝟎 𝑪

 𝑻𝒊 ′ = 𝟕𝟎 − 𝟔𝟎 = 𝟏𝟎
′ 𝟏𝟎
 𝑻𝒊 𝒔 =
𝒔
𝟏𝟎 𝟏
 𝑻𝒊 ′ 𝒔 = ( )
𝒔 𝟓𝒔+𝟏
  Inverting to the time domain, we obtain the expression for 𝑻′ (𝒕)
26 𝒕
−𝟓
 𝑻′ 𝒕 = 𝟏𝟎(𝟏 − 𝒆 )
 Finally
𝒕
−
 𝑻′ 𝒕 = 𝑻 𝒕 − 𝑻𝒔 = 𝟏𝟎(𝟏 − 𝒆 ) 𝟓
𝒕
−𝟓
 𝑻′ 𝒕 = 𝑻 𝒕 − 𝟖𝟎 = 𝟏𝟎(𝟏 − 𝒆 )
𝒕
−𝟓
 𝑻 𝒕 = 𝟖𝟎 + 𝟏𝟎(𝟏 − 𝒆 )
 A plot of the outlet temperature
(in deviation variables) is shown in
the Fig. 5–7 a. The actual outlet
temperature is shown in Fig. 5–7 b.
 Note that for the uncontrolled mixing tank,
a step change of 10 0C in the inlet temperate ultimately produces a 10 0C change in the
outlet temperature, just as we predicted intuitively before we began our modeling
27
  In summarizing the previous examples of first-order systems,
28 the time constant for each has been expressed in terms of
system parameters; thus
𝒎𝑪
 𝝉= for thermometer
𝒉𝑨
 𝝉 = 𝑨𝑹 for liquid level process
𝑽
 𝝉= for mixing process
𝒒
𝝆𝑽
 𝝉= for heating process
𝒘