Access full Solution Manual only here

https://www.book4me.xyz/solution-manual-automatic-process-control-smith-corripio/

%TO := % %CO := %
Smith & Corripio, 3rd edition
Problem 6-1. Second-order loop with proportional controller.

D(s)
G2(s)
R(s) + E(s) M(s) + C(s)
Gc(s) G1(s)
+
-

K
G1 ( s) = Gc( s) = Kc
(τ 1⋅ s + 1)⋅ (τ 2⋅ s + 1)
%TO
Problem parameters: K := 0.10 τ 1 := 1min τ 2 := 0.8min
%CO
(a) Closed loop transfer function and characteristic equation of the loop.
K
Kc⋅
C( s)
=
(τ 1⋅ s + 1) ⋅ (τ 2⋅ s + 1) =
Kc⋅ K
R( s)
1 + Kc⋅
K (τ 1⋅ s + 1)⋅ (τ 2⋅ s + 1) + Kc⋅ K
(τ 1⋅ s + 1) ⋅ (τ 2⋅ s + 1)
Characteristic equation:
2
( )
τ 1 ⋅ τ 2 ⋅ s + τ 1 + τ 2 s + 1 + Kc⋅ K = 0

C( s) 0.1Kc
Closed-loop transfer function: =
R( s) 2
0.8s + 1.8s + 1 + 0.1Kc

2
Characteristic equation: 0.8s + 1.8s + 1 + 0.1Kc = 0

(b) Values of the controller gain for which the response is over-damped, critically
damped, and under-damped

Roots of the characteristic equation:

( )
2 2
−1.8 + 1.8 − 4 ⋅ 0.8⋅ 1 + 0.1Kc 1 + 0.1Kc
−1.8
+  1.8  −
r1 = =  1.6
2 ⋅ 0.8 1.6   0.8
https://www.book4me.xyz/solution-manual-automatic-process-control-smith-corripio/

The response is critically damped when the term in the radical is zero: 2 1 + 0.1Kc
 1.8  −
 1.6 =0
  0.8
1   1.8   2
%CO
Critically damped:
Kccd := 0.8  − 1 Kccd = 0.125
0.1   1.6   %TO

%CO %CO
Over-damped (real roots): Kc < 0.125 Under-damped: Kc > 0.125
%TO %TO

The loop cannot be unstable for positive gain because,

• for real roots the radical cannot be greater than the negative term, so both roots are negative
• for complex conjugate roots the real part is always negative, -1.8/1.6, or -(τ1 +τ2 )/2τ1 τ2
This is true for all positive values of the time constants and the product K. cK.

(c) Equivalent time constants for different values of the gain:

%CO
Kc := 0.1 (over-damped, two equivalent time constans)
%TO
−1 2⋅ τ 1⋅ τ 2
τ e1 = τ e1 := τ e1 = 0.935 min
r1
(τ 1 + τ 2) − (τ 1 + τ 2) 2
(
− 4 ⋅ τ 1 ⋅ τ 2 ⋅ 1 + Kc⋅ K)
−1 2⋅ τ 1⋅ τ 2
τ e2 = τ e2 := τ e2 = 0.847 min
r2
(τ 1 + τ 2) + (τ 1 + τ 2)
2
(
− 4 ⋅ τ 1 ⋅ τ 2 ⋅ 1 + Kc⋅ K)
%CO
Kc := 0.125 (critically damped, two equal real time constants)
%TO
−1 2⋅ τ 1⋅ τ 2
τ e1 = τ e1 := τ e1 = 0.889 min
r1
(τ 1 + τ 2) − (τ 1 + τ 2)
2
(
− 4 ⋅ τ 1 ⋅ τ 2 ⋅ 1 + Kc⋅ K)
−1 2⋅ τ 1⋅ τ 2
τ e2 = τ e2 := τ e2 = 0.889 min
r2
(τ 1 + τ 2) + (τ 1 + τ 2) 2
(
− 4 ⋅ τ 1 ⋅ τ 2 ⋅ 1 + Kc⋅ K)
%CO
Kc := 0.2 (under-damped, time constant and damping ratio)
%TO
2 2 τ 1⋅ τ 2 2
τ1 + τ2
τ s + 2ζ ⋅ τ ⋅ s + 1 = s + s+1
1 + Kc⋅ K 1 + Kc⋅ K

τ 1⋅ τ 2 τ1 + τ2
Match coefficients: τ := ζ := τ = 0.886 min ζ = 0.996
1 + Kc⋅ K (
2 ⋅ τ ⋅ 1 + Kc⋅ K )
(d) Steady-state offset for a unit step change in set point.
1
Final value theorem: lim Y( t ) = lim s⋅ Y( s) R( s) = (Table 2-1.1)
t→∞ s→0 s
 %CO Kc⋅ K 1 -3
Kc := 0.1 lim s⋅ → 9.9009900990099009901 ⋅ 10
%TO s→0 2
( )
τ 1 ⋅ τ 2 ⋅ s + τ 1 + τ 2 ⋅ s + 1 + Kc⋅ K
s

offset := ( 1 − 0.0099)%TO offset = 0.99 %TO

%CO Kc⋅ K 1 -2
Kc := 0.125 lim s⋅ → 1.2345679012345679012 ⋅ 10
%TO s → 0 2
( )
τ 1 ⋅ τ 2 ⋅ s + τ 1 + τ 2 ⋅ s + 1 + Kc⋅ K
s

offset := ( 1 − 0.01235 )%TO offset = 0.988 %TO

%CO Kc⋅ K 1 -2
Kc := 0.2 lim s⋅ → 1.9607843137254901961 ⋅ 10
%TO s→0 2
( )
τ 1 ⋅ τ 2 ⋅ s + τ 1 + τ 2 ⋅ s + 1 + Kc⋅ K
s

offset := ( 1 − 0.01961 )%TO offset = 0.98 %TO

These are very large offsets because the loop gains are so small.
Excerpts from this work may be reproduced by instructors for distribution on a not-for-profit basis for testing or instructional purposes
only to students enrolled in courses for which the textbook has been adopted. Any other reproduction or translation of this work
beyond that permitted by Sections 107 or 108 of the 1976 United States Copyright Act without the permission of the copyright owner
is unlawful.

https://www.book4me.xyz/solution-manual-automatic-process-control-smith-corripio/
Smith & Corripio, 3rd edition
Problem 6-2. Inverse-response second-order system with proportional
controller.

D(s)
G2(s)
R(s) + E(s) M(s) + C(s)
Gc(s) G1(s)
+
-

6 ( 1 − s) %TO %CO
G1 ( s) = Gc( s) = Kc⋅
( s + 1 ) ⋅ ( 0.5⋅ s + 1 ) %CO %TO

(a) Closed-loop transfer function and characteristic equation of the loop.

C( s) Kc⋅ 6 ( 1 − s)
Closed-loop transfer function: =
R( s) ( s + 1 ) ⋅ ( 0.5s + 1 ) + Kc⋅ 6 ( 1 − s)

( )
2
Characteristic equation: 0.5⋅ s + 1.5 − 6Kc s + 1 + 6Kc = 0

(b) Values of the gain for which the response is over-, critically, and under-damped
Roots:

( ) (1.5 − 6Kc) ( ) = −1.5 + 6K

2
− 1.5 − 6Kc + − 4 ⋅ 0.5⋅ 1 + 6Kc 2
r1 =
2 ⋅ 0.5 c+ 0.25 − 30Kc + 36Kc

The response is critically damped when the term in the radical is zero:
2
2 30 + 30 − 4 ⋅ 0.25⋅ 36 %CO
0.25 − 30Kc + 36Kc = 0 Kc := Kc = 0.82491
2 ⋅ 36 %TO
2
30 − 30 − 4 ⋅ 0.25⋅ 36 %CO
Kc := Kc = 0.00842
2 ⋅ 36 %TO
%CO %CO
Over-damped (two real roots): Kc < 0.00842 and Kc > 0.825
%TO %TO

%CO %CO
Under-damped (complex conjugate roots): 0.00842 < Kc < 0.825
%TO %TO
 %CO
The response is unstable when Kc > 0.25 (one real root is positive or the real part of the
%TO complex roots i positive)
(c) Effective time constants or time constant and damping ratio for various values o
the gain:

0.5 2 1.5 − 6Kc 2 2

s + s + 1 = τ s + 2ζ ⋅ τ ⋅ s + 1
1 + 6Kc 1 + 6Kc

%CO 0.5min
2 (1.5 − 6Kc)min
Kc := 0.1 τ := ζ := τ = 0.559 min ζ = 0.503
%TO 1 + 6Kc 2 ⋅ τ ⋅ ( 1 + 6Kc)

%CO 0.5min
2
(1.5 − 6Kc)min
Kc := 0.125 τ := ζ := τ = 0.535 min ζ = 0.401
%TO 1 + 6Kc 2 ⋅ τ ⋅ ( 1 + 6Kc)

Kc := 0.2
%CO
0.5min
2 (1.5 − 6Kc)min
%TO τ := ζ := τ = 0.477 min ζ = 0.143
1 + 6Kc 2 ⋅ τ ⋅ ( 1 + 6Kc)

Kc := 0.3
%CO
0.5min
2
(1.5 − 6Kc)min
%TO τ := ζ := τ = 0.423 min ζ = −0.127
1 + 6Kc 2 ⋅ τ ⋅ ( 1 + 6Kc)
(unstable)
Try values that result in equivalent time constants:
%CO 1min
Kc := 0.005 τ e1 := τ e1 = 0.868 min
%TO 2
1.5 − 6 ⋅ Kc − 0.25 − 30Kc + 36Kc

1min
τ e2 := τ e2 = 0.559 min
2
1.5 − 6 ⋅ Kc + 0.25 − 30Kc + 36Kc

%CO 1min
Kc := 1 τ e1 := τ e1 = −0.143 min
%TO 2
1.5 − 6 ⋅ Kc − 0.25 − 30Kc + 36Kc
(unstable)
1min
τ e2 := τ e2 = −0.5 min
2
1.5 − 6 ⋅ Kc + 0.25 − 30Kc + 36Kc

(d) Offset for various values of the gain and a unit step change in set point.
%CO Kc⋅ 6 ⋅ ( 1 − s) 1
Kc := 0.10 lim s⋅ → .37500000000000000000
%TO s
( )
s→0 2
0.5s + 1.5 − 6 ⋅ Kc s + 1 + 6Kc

%CO
offset := 1 − 0.375 offset = 0.625
%TO
https://www.book4me.xyz/solution-manual-automatic-process-control-smith-corripio/

%CO Kc⋅ 6 ⋅ ( 1 − s) 1
Kc := 0.125 lim s⋅ → .42857142857142857143
%TO s
s→0 2
( )
0.5s + 1.5 − 6 ⋅ Kc s + 1 + 6Kc

%CO
offset := 1 − 0.429 offset = 0.571
%TO

%CO Kc⋅ 6 ⋅ ( 1 − s) 1
Kc := 0.20 lim s⋅ → .54545454545454545455
%TO s
s→0 2
( )
0.5s + 1.5 − 6 ⋅ Kc s + 1 + 6Kc

%CO
offset := 1 − 0.545 offset = 0.455
%TO

The offsets are high because the gains are small. Of course, since for gains greater than
0.25%CO/%TO the loop is unstable, offsets can only be high with a proportional controller.
Excerpts from this work may be reproduced by instructors for distribution on a not-for-profit basis for testing or instructional purposes
only to students enrolled in courses for which the textbook has been adopted. Any other reproduction or translation of this work
beyond that permitted by Sections 107 or 108 of the 1976 United States Copyright Act without the permission of the copyright owner
is unlawful.
K := 1.8R
Smith & Corripio, 3rd edition
Problem 6-3. First-order process and proportional-integral controller.

D(s)
G2(s)
R(s) + E(s) M(s) + C(s)
Gc(s) G1(s)
+
-

Gc( s) = Kc⋅  1 + 
K 1
G1 ( s) =
τ ⋅s + 1 τ I⋅ s
 

To work in dimensionless units, t/τ, set: τ := 1

(a) Closed-loop transfer function and characteristic equation of the loop. Offset.

Closed-loop transfer functon:

C( s)
=
(
KKc⋅ τ I⋅ s + 1 )
R( s) τ I⋅ s⋅ ( τ ⋅ s + 1 ) + KKc⋅ ( τ I⋅ s + 1 )

( )
2
Characteristic equation: τ I⋅ τ ⋅ s + 1 + KKc τ I⋅ s + KKc = 0

Offset: the steady state gain is: lim

(
KKc⋅ τ I⋅ s + 1 ) =
KKc
=1
KKc
I (
s → 0 τ ⋅ τ ⋅ s2 + 1 + KK τ ⋅ s + KK
c I c )
(no offset)

(b) Is there an ultimate gain for this loop?

2
Substitute s = iω: (
−τ I⋅ τ ⋅ ω u + KKcu + i 1 + KKcu ω u = 0 ) ω u := 0 KKcu := 0

No, there is no ultimate gain. This result just means that a negative loop gain will make the loop
unstable. Another way to show it is to determine the roots of the characteristic equation:

2
( ) (1 + KKc)
2
− 1 + KKc ⋅ τ I + ⋅ τ I − 4 ⋅ τ I⋅ τ ⋅ KKc
r1 =
2τ I⋅ τ

The real root cannot be negative for any positive value of the loop gain KK c because the radical is
always smaller than the negative term. Also, for complex conjugate roots, the real part is always
https://www.book4me.xyz/solution-manual-automatic-process-control-smith-corripio/

negative:
(
− 1 + KKc )
Real = <0
2⋅ τ
(c) Response of the loop to a step change in set point for τI = τ as the gain varies
from 0 to infinity.
R( s) =
1
(Table 2-1.1) KKc⋅ ( τ ⋅ s + 1 ) 1 KKc 1
s Y( s) = =
τ ⋅ s⋅ ( τ ⋅ s + 1 ) + KKc⋅ ( τ ⋅ s + 1 ) s τ ⋅ s + KKc s

τ 1 1 1 1
Let τc = Y( s) = = −
KKc τ c⋅ s + 1 s s 1 u ( t) := 0 if t < 0
s+
τc 1 if t ≥ 0

τ c := 1
−t R( t) := u ( t)
Invert using Table 2-1.1: τc
Y( t) := u ( t) − e

1
As KKc increases τc decreases and the response
is faster. Y ( t)

R( t) 0.5

0
0 2 4
t

Excerpts from this work may be reproduced by instructors for distribution on a not-for-profit basis for testing or instructional purposes
only to students enrolled in courses for which the textbook has been adopted. Any other reproduction or translation of this work
beyond that permitted by Sections 107 or 108 of the 1976 United States Copyright Act without the permission of the copyright owner
is unlawful.

K
R :=
1.8
https://www.book4me.xyz/solution-manual-automatic-process-control-smith-corripio/

Smith & Corripio, 3rd edition

Problem 6-4. Second-order process with pure integral controller.

D(s)
G2(s)
R(s) + E(s) M(s) + C(s)
Gc(s) G1(s)
+
-

K KI
G1 ( s) = Gc( s) =
(τ 1⋅ s + 1)⋅ (τ 2⋅ s + 1) s

(a) Ultimate gain and period with the parameters of Problem 6-1:
%CO
K := 0.1 τ 1 := 1min τ 2 := 0.8min
%TO
K⋅ KI
( )
Characteristic equation: 2
τ 1⋅ τ 2⋅ s + τ 1 + τ 2 s + 1 + =0
s
3
( )2
τ 1 ⋅ τ 2 s + τ 1 + τ 2 s + s + KKI = 0

( )
3 2
Substitute s = iωu : −τ 1 ⋅ τ 2 ⋅ i ⋅ ω u − τ 1 + τ 2 ω u + i⋅ ω u + KKIu = 0 + 0i

( )
2 3
− τ 1 + τ 2 ω u + KKIu = 0 −τ 1 ⋅ τ 2 ω u + ω u = 0

ω u :=
1
KIu :=
(τ 1 + τ 2)⋅ ω u2 ω u = 1.118 min
−1
KIu = 22.5
%CO
τ 1⋅ τ 2
K %TO⋅ min
2π
Tu :=
ωu Tu = 5.62 min

(b) Ultimate gain and period for other values of the smaller time constant:

τ 2 := 0.1min
ω u :=
1
KIu :=
(τ 1 + τ 2)⋅ ω u2 ω u = 3.162 min
−1
KIu = 110
%CO
τ 1⋅ τ 2
K %TO⋅ min
π
Tu := 2
ωu Tu = 1.987 min
https://www.book4me.xyz/solution-manual-automatic-process-control-smith-corripio/

τ 2 := 2min
ω u :=
1
KIu :=
(τ 1 + τ 2)⋅ ω u2 ω u = 0.707 min
−1
KIu = 15
%CO
τ 1⋅ τ 2
K %TO⋅ min
2π
Tu :=
ωu Tu = 8.886 min

Reducing the non-dominat time constant increases the ultimate gain and reduces the ultimate
period, as expected. When τ2 is increased to 2 min, it becomes the dominant time constant and
the ultmate gain should be higher than for part (a). However, in this case K I has units of rate and,
since the loop is slower, it results in a smaller ultimate gain.

Excerpts from this work may be reproduced by instructors for distribution on a not-for-profit basis for testing or instructional purposes
only to students enrolled in courses for which the textbook has been adopted. Any other reproduction or translation of this work
beyond that permitted by Sections 107 or 108 of the 1976 United States Copyright Act without the permission of the copyright owner
is unlawful.