4 NUMERICAL TECHNIQUE

SIX-PLATE ABSORPTION COLUMN – PROBLEM 1

 It is desired to develop the steady-state tray compositions for a six-plate absorption column. It can be assumed that
a linear equilibrium relation holds between liquid (Xm) and vapor (Ym) on each plate:
𝑌𝑚 = 𝑎𝑋𝑚 + 𝑏

Solution
 The inlet composition to the column x0 and y7 are specified along with the
liquid (L) and gas (G) phase flow rates (moles/time). The system is shown
schematically in Figure .
 To solve the problem, a material balance is written on a representative tray,
n, shown in Figure
 Applying the macroscopic mass balance yields
Rate of mass in = Rate of mass out
Lxn−1 + Gyn+l = Lxn + Gyn
 Two key component(A+B)  Three component
 Inert gas B (Ex:air)
 One feed (normally)  Inert liquid C (Water)
 Two pure product.  Solute A( CO2 )
 Two sections Goes from gas to liquid
 Two feeds
 Heating & cooling  One pure product
 Boiling mixtures  One section
 Non-boiling mixture
 P1 – SOLUTION (CONTINUED)

 Using the linear equilibrium relation in the mass balance gives

Lxn−1 + G axn+1 + b = Lxn + G axn + b
OR
Lxn−1 − L + Ga x n + Gaxn+1 = 0

 The entire set of equations for the six-plate column is

• n=1: −(L + Ga)x1 + Gax2 = −Lx0 ( where x0 is specified)
• n=2: Lx 1 − (L + Ga)x2 + Gax3 = 0
• n=3: Lx 2 − (L + Ga)x3 + Gax4 = 0
• n=4: Lx3 − (L + Ga)x4 + Gax5 = 0
• n=5: Lx4 − (L + Ga)x5 + Gax6 = 0
y7 −b
• n=6: Lx5 − L + Ga x6 = −Ga
a
(where y7 is specified)
 SOLUTION PROCEDURE

 The simultaneous solution can be expressed in matrix notation as

 It is in the form of A X = B or X = B A-1

 Since it is a tridiagonal matrix, with upper and lower triangle containing Zeros, The above matrix can be solved by
Using Thomas Algorithm.
 SOLUTION OF NONLINEAR EQUATIONS (1/2)

Solving a Single Nonlinear Equation in One

Unknown

Wegstein Method
 An initial guess is used (x1) to generate a
second estimate x2 = g(x1).
 Note in the figure that x2 is diverging from
the solution.
 At this point we project a straight line through
the points (Xl, g(xl)) and (x2, g(x2)) to the
diagonal line, that is
𝑦 − 𝑔 𝑥1 𝑔 𝑥1 − 𝑔 𝑥2
=
𝑥 − 𝑥1 𝑥1 − 𝑥2
 SOLUTION OF NONLINEAR EQUATIONS (2/2)

at the point y = x = x3; then

𝑥3 − 𝑔 𝑥1 𝑔 𝑥1 − 𝑔 𝑥2
=
𝑥3 − 𝑥1 𝑥1 − 𝑥2
 Rearranging the above equation in term of X3
x1 g x2 − x2 g x1
x3 = −−− −(1)
x1 − x2 − g x1 + g x2
Note
1) Given function/ Equation should be in explicit form
2) Since x2= g(x1) for first iteration only then x3 (above eqn.) is given by

x1 g x2 − x2 2
x3 = −−− −(2)
x1 − 2x2 + g x2
3) For second iteration onwards use equation (1) to calculate X 3
4) For second iteration, x2 replaces xl and x3 replaces x2 and in equation (1)
𝑥2 → 𝑥1
𝑥3 → 𝑥2
 PROBLEM 1

 Solve the following equation using Wegstein method

15.04
𝐶𝑃 = 0.716 − 4.257 × 10−6 𝑇 −
𝑇
where the units of Cp are (kJ/kg K) and the unit of temperature T is K. Use Two iteration and use T1=600 K as
starting value.
Solution
15.04
𝑔 𝑇 = 0.716 − 4.257 × 10−6 T − =0
T
Rearranging the above equation in explicit form

0.716 − 15.04ൗ
𝑇
𝑇=𝑔 𝑇 = −−− −(3)
4.257 × 10−6
 P1 – SOLUTION (1/3)

Iteration 1
T1= 600 K; g(T1) = T2
g(600) = 23959 (using equation (3))
g(T2) = g(23959) = 145368.6
 Now sub. in

𝑇1 𝑔 𝑇2 − 𝑇2 2
𝑇3 =
𝑇1 − 2𝑇2 + 𝑔 𝑇2
600 × 145368.6 − (23959)2
𝑇3 =
600 − 2 × 23959 + 145368.6
T3 = -4964.9 K
 Negative temperature is physically inadmissible) => can’t proceed further
 Reformulate the above equation
2
15.04
𝑇= −−− −(4)
0.716 − 4.257 × 10−6 𝑇
 P1 – SOLUTION (2/3)

Iteration 1
T1=600 K;
g(T1) = T2 => g(600) = 444.40 ---- using eqn.(4)
g(444.40) = 443.57

T1 g T2 − T2 2
𝑇3 =
T1 − 2T2 + g T2
600 × 443.57 − 444.42
𝑇3 =
600 − 2 × 444.40 + 443.57
T3 = 443.587 K

 Iteration 2
T2 → T1 i.e. T1 = 444.40 K
T3 → T2 i.e. T2 = 443.587 K
g(T2) = g(443.587) = 443.57 ---- using eqn.(4)
g(T1) = g(444.40) = 443.57
 P1 – SOLUTION (3/3)

𝑇1 𝑔 𝑇2 − 𝑇2 𝑔 𝑇1
𝑇3 =
𝑇1 − 𝑇2 − 𝑔 𝑇1 + 𝑔 𝑇2
444.4 × 443.57 − 443.587 × 443.57
𝑇3 =
444.4 − 443.587 − 443.57 + 443.57
T3 = 443.57 K

 Since the value of T3 of first iteration and T3 of the second iteration are nearly same, third iteration is not required.

Iteration T1 g(T)= T2 T3

1 600 444.40 443.587

2 444.2 443.587 443.57

 PROBLEM 2

 Find the volume of the following equation using Wegstein method. Perform Two iterations.
𝑅𝑇 − 𝑎
𝑃= 𝑒 𝑉𝑅𝑇
𝑉−𝑏
𝑎𝑡𝑚 𝑐𝑚3 𝑐𝑚3
Given 𝑎 = 4620 ; b= 46.3 ; P=100 atm; T=50oC ; n=1 gmol.
𝑔𝑚𝑜𝑙 𝑔𝑚𝑜𝑙

Solution
Starting Guess value of volume is found from Ideal Gas Law
𝑛𝑅𝑇
𝑉1 =
𝑃
𝑙𝑡. 𝑎𝑡𝑚 3
𝑐𝑚3 . 𝑎𝑡𝑚
𝑅 = 0.08206 = 0.08206 × 10
𝑚𝑜𝑙. 𝐾 𝑚𝑜𝑙. 𝐾
1 × 0.08206 × 103 × 323
𝑉= = 265𝑐𝑚3 /𝑔𝑚𝑜𝑙
100
Rearranging the above equation
𝑎
−
𝑅𝑇 𝑒 𝑉𝑅𝑇
𝑉 =𝑏+ −−− −(1)
𝑃
 P2 – SOLUTION

Iteration 1 Iteration 2
V2 = g(V1) = 311.196 cm3 V2 → V1 i.e. V1 = 311.196
g(V2) = 311.2 cm3 V3 → V2 i.e. V2 = 311.2

V1 g V2 − 𝑉2 2 g(V2) = g(311.2) ---- using eqn.(1)

𝑉3 =
V1 − 2V2 + g V2 V1 g V2 − V2 g V1
V3 =
300 × 311.2 − 311.196 2 V1 − V2 − g V1 + g V2
𝑉3 =
300 − 2 × 311.196 + 311.2
V3= 311.20 cm3/gmol

Iteration V1 g(V)= V2 V3

1 265 311.96 311.20

2 311.96 311.20 311.20

 PROBLEM 3

 Solve the following equation for f using Wegstein method

Ɛ
1 2.51
= −0.86 𝑙𝑛 𝐷 +
𝑓 3.7 𝑅𝑒 𝑓

where Ɛ/D = 10-4 and Re = 105 . Use starting f value as 0.01 and perform two iterations

Solution:
Rearranging the given equation in explicit form
2
1
𝑓=𝑔 𝑓 =
10−4 2.51
−0.86 𝑙𝑛 + 5
3.7 10 𝑓
2
1
f=g f =
2.51 × 10−5
−0.86 ln 2.702 × 10−5 +
105 f
 P3 - SOLUTION

Iteration 1
𝑓1 𝑔 𝑓2 − 𝑓2 2
𝑓3 =
𝑓1 − 2𝑓2 + 𝑔 𝑓2
Iteration 2
𝑓1 g 𝑓2 − f2 g f1
𝑓3 =
f1 − f2 − g f1 + g f2

Iteration f1 g(f1)= f2 g(f2) f3

1 0.01 0.020168 0.01872 0.0189

2 0.0201068 0.0189 0.018845 0.01885

 MULTI-VARIABLE NEWTON-RAPHSON METHOD

 Consider Fi(x) = 0, where i = 0,1,2, …….. N

 Writing the expression for Fi(x) at x(k+1) i.e Fi(x(k+1) ) using multi-variable Taylor series expansion around xk
𝑁
𝜕𝐹𝑖
𝐹𝑖 𝑥 𝑘+1 = 𝐹𝑖 𝑥 𝑘 + ෍ 𝑥𝑗 𝑘+1 − 𝑥𝑗 𝑘 + 𝐻𝑖𝑔ℎ𝑒𝑟 𝑜𝑟𝑑𝑒𝑟 𝑡𝑒𝑟𝑚𝑠 −−− −(1)
𝜕𝑥𝑗
𝑗=1 𝑥𝑘

 Forcing LHS in eqn.(1) and putting all the N eqn. together

𝜕𝐹1 𝜕𝐹1 𝜕𝐹1
…………….
𝜕𝑥1 𝜕𝑥2 𝜕𝑥𝑁 𝑥1 𝑘+1 − 𝑥1 𝑘
𝑘
𝐹1 𝑥 𝜕𝐹2 𝜕𝐹2 𝜕𝐹2
𝑘 … … … … … . 𝑥2 𝑘+1 − 𝑥2 𝑘
𝐹2 𝑥 𝜕𝑥1 𝜕𝑥2 𝜕𝑥𝑁
.. .
− , = . ×
.. .
.
𝐹𝑁 𝑥 𝑘 .
𝜕𝐹𝑁 𝜕𝐹𝑁 𝜕𝐹𝑁 𝑥𝑁 𝑘+1 − 𝑥𝑁 𝑘 𝑥𝑘
…………….
𝜕𝑥1 𝜕𝑥2 𝜕𝑥𝑁 𝑘
𝑥

 For brevity sake

− 𝐹 𝑥 𝑘 = 𝐴 𝑥 𝑘 × 𝑥 𝑘+1 − 𝑥 𝑘 𝑤ℎ𝑒𝑟𝑒 𝐴 𝑥 𝑘 = 𝐽𝑎𝑐𝑜𝑏𝑖𝑎𝑛 𝑚𝑎𝑡𝑟𝑖𝑥
𝑥 𝑘+1 = 𝑥 𝑘 − 𝐴 𝑥 𝑘 −1
𝐹 𝑥𝑘 −−− −(2)
 PROBLEM 1

 Solve the following using N-R method for two iterations

F1(y) = 4 - 8y2 + 4y3 – 2 y23 = 0
F2(y) = 1 - 4y2 + 3y3 + y32 = 0
1 1 1 𝑇
Starting value 𝑦 = 𝑦2 𝑦3 = 0.5 0.5

Solution
 Using Newton-Raphson multivariable method
𝑦 𝑘+1 = 𝑦 𝑘 − 𝐴 𝑦 𝑘 −1
𝐹 𝑦𝑘

 Expanding for two equation

F1(y) = 4 - 8y2 + 4y3 – 2 y23 = 0
F2(y) = 1 - 4y2 + 3y3 + y32 = 0
 P1 – SOLUTION (1/2)

1st Iteration
−1
𝜕𝐹1 𝜕𝐹1
𝑦2 𝑘+1 𝑦2 𝑘 𝜕𝑦2 𝜕𝑦3 4 − 8𝑦2 + 4𝑦3 − 2𝑦2 3
= −
𝑦3 𝑘+1 𝑦3 𝑘 𝜕𝐹2 𝜕𝐹2 1 − 4𝑦2 + 3𝑦3 + 𝑦3 2 [Link]
𝜕𝑦2 𝜕𝑦3 0.5,0,5

−1
𝑦2 2 0.5 −8 − 6𝑦2 2 4 4 − 8𝑦2 + 4𝑦3 − 2𝑦2 3
= −
𝑦3 2 0.5 −4 3 + 2𝑦3 0.5,0.5 1 − 4𝑦2 + 3𝑦3 + 𝑦3 2 [Link]

𝑦2 2 0.5 −9.5 4 −1
1.75
= −
𝑦3 2 0.5 −4 4 0.75

𝑦2 2 0.5 −0.182
2 = 0.5 − 5.5 × 10−3
𝑦3

𝑦2 2 0.682
2 = 0.4945
𝑦3
 P1 – SOLUTION (2/2)

2nd Iteration
−1
𝑦2 3 0.682 −8 − 6𝑦2 2 4 4 − 8𝑦2 + 4𝑦3 − 2𝑦2 3
= −
𝑦3 3 0.4945 −4 3 + 2𝑦3 0.682, 0.4945 1 − 4𝑦2 + 3𝑦3 + 𝑦3 2 0.682, 0.4945

𝑦2 3 0.6654
3 = 0.4779
𝑦3
 PROBLEM 2

 Find the roots of the following equations

3x2 – 2y2 =1
x2 -2x +y2 +2y =8

Assume x0 = -1 and y0 = 1, for two iteration

Solution
𝐹1 𝑥, 𝑦 = 3𝑥 2 − 2𝑦 2 − 1
𝐹2 𝑥, 𝑦 = 𝑥 2 − 2𝑥 + 𝑦 2 + 2𝑦 − 8

−1
𝜕𝐹1 𝜕𝐹1
𝑥 𝑘+1 𝑥𝑘 𝜕𝑥 𝜕𝑦 𝐹1 𝑥, 𝑦
= 𝑘 −
𝑦 𝑘+1 𝑦 𝜕𝐹2 𝜕𝐹2 𝐹2 𝑥, 𝑦 𝑥 𝑘 ,𝑦𝑘
𝜕𝑥 𝜕𝑦 𝑥 𝑘 ,𝑦𝑘

−1
𝑥1 −1 6𝑥 − 4𝑦 3𝑥 2 − 2𝑦 2 − 1
= −
𝑦1 1 2𝑥 − 2 2𝑦 + 2 −1,1 𝑥 2 − 2𝑥 + 𝑦 2 + 2𝑦 − 8 −1,1
 P2 - SOLUTION
−1
𝑥1 −1 −6 −4 0
1 = −
𝑦 1 −4 4 2
𝑥1 −1 0.2
1 = −
𝑦 1 −0.3
𝑥1 −1.2
1 =
𝑦 1.3
2nd Iteration
−1
x2 −1.2 6x − 4y 3x 2 − 2y 2 − 1
= −
y2 1.3 2x − 2 2y + 2 −1.2,1.3 x 2 − 2x + y 2 + 2y − 8 −1.2,1.3

x2 −1.2 −0.0071
2 = −
y 1.3 0.00214
x2 −1.19
2 =
y 1.2786