Scribd
Upload
0 views

Tutorial 1 solute boundary layer

Matlab tutorial for solute boundary layer

Uploaded by

fraudyfraudy1
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
0 views

Tutorial 1 solute boundary layer

Matlab tutorial for solute boundary layer

Uploaded by

fraudyfraudy1
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

University of Sheffield Hector Basoalto

MAT6334/MAT3334/AER444/MAT3444
Tutorial 1 Numerical solution for the solute boundary layer
Hector Basoalto
University of Sheffield

PROBLEM STATEMENT

In this tutorial you will learn to solve for the solute boundary layer ahead of a moving S/L
interface and determine the solute boundary layer thickness 𝛿𝑐 . Under steady state
conditions the diffusion equation has the following form

𝜕 2𝑐 𝜕𝑐
𝐷 2+𝑉 =0 (1)
𝜕𝑥 𝜕𝑥
where 𝐷 is the diffusivity and 𝑉 is the S/L interface velocity. Depending on diffusivity and
process conditions (i.e., the interface velocity), solute boundary layer in a binary alloy has
a thickness of the order of 𝛿𝑐 ∼ 10 − 100𝜇m.

To solve for the steady state concentration field, we need to specify boundary conditions.
For a binary alloy, the steady state solution will satisfy the following boundary conditions
𝑐0
𝑐(𝑥 = 0) =
𝑘
(2)
𝑐(𝑥 → ∞) = 𝑐0

will consider a binary alloy with composition 𝑐0 and partition coefficient 𝑘.

TASKS

i) Numerically determine solute boundary layer by implementing an iterative finite


difference algorithm. The S/L interface is located at 𝑥 = 0 and assume the computational
domain size to be 5µm. Typical parameters for a binary alloy are listed in Table 1.

ii) Numerically estimate the solute boundary layer thickness 𝛿𝑐 by locating where
𝑐(𝑥 = 𝛿𝑐 ) = 0.99 𝑐0 . How does this numerical value compare with the theoretical
approximation 𝛿𝑐 = 2𝐷/𝑉.

iii) Derive a plot of 𝛿𝑐 - 𝑉 for interface velocities ranging between 1-100 𝜇m s −1 .

Table 1 Model parameters

𝑐0 𝑘 𝐷 [𝜇m2 s −1 ] 𝑉 [𝜇m s −1 ]
0.2 0.28 10 10
University of Sheffield Hector Basoalto

FINITE DIFFERENCE SCHEME

Application of the finite difference scheme (FDS), the derivatives need to be discretised.
The space is discretised into equally spaced nodes 𝑥1 , 𝑥2 , ⋯ , 𝑥𝑛 , where the size of the
element is Δ𝑥 = 𝑥𝑖+1 − 𝑥𝑖 . The concentration field is also discretised, i.e., 𝑐(𝑥𝑖 ) = 𝑐𝑖 . In
the most popular approach the second spatial derivatives are approximated by the
second-order central formula and first order spatial derivatives by the centered
difference formula:

𝜕 2 𝑐 𝑐𝑖+1 − 2 𝑐𝑖 + 𝑐𝑖−1
≈ (3)
𝜕𝑥 2 Δ𝑥 2

𝜕𝑐 𝑐𝑖+1 − 𝑐𝑖−1
≈
𝜕𝑥 2 Δ𝑥 (4)
The finite difference formula for PDE is obtained by inserting Equations (3) and (4) into
Equation (1):
[𝑐𝑖+1 − 2 𝑐𝑖 + 𝑐𝑖−1 ] 𝑉 [𝑐𝑖+1 − 𝑐𝑖−1 ]
+ =0
Δ𝑥 2 𝐷 2 Δ𝑥

1 1
𝑐𝑖 = (1 + 𝐹0 )𝑐𝑖+1 + (1 − 𝐹0 )𝑐𝑖−1
2 2

Δ𝑥 𝑉 (5)
𝐹0 =
2 𝐷

(6)
An iterative scheme will be used. This means that the solution is found by iterating
Equation (5) until the error between successive interactions falls below a specified
tolerance. The 𝑛 + 1 iteration is given by the 𝑛 interaction via the following relation

(𝑛+1) 1 (𝑛) 1 (𝑛)


𝑐𝑖 = (1 + 𝐹0 )𝑐𝑖+1 + (1 − 𝐹0 )𝑐𝑖−1
2 2 (7)
The relative error is defined as
(𝑛+1) (𝑛)
𝑐𝑖 − 𝑐𝑖
𝜖 = max ( (𝑛+1)
)
𝑐𝑖

The iteration is carried out until tol > 𝜖, where tol is a user specified tolerance.

Boundary conditions imply (see Equation (2)):


𝑐0
𝑐1 = 𝑐𝑛 = 𝑐0
𝑘
Table 2 numerical parameters.
𝐿 [𝜇m] n tol
5 500 0.001
University of Sheffield Hector Basoalto

MATLAB IMPLEMENTATION
An example of a numerical implementation of the scheme is presented in Algorithm 1.
The numerical parameter used in this example are listed in Table 2.

Algorithm 1 Iterative finite difference scheme


.

• Define process and material


parameters.

• Discretisation of space.

• Parameters required to control


while loop.

• Iterative scheme. Repeat


while-loop until condition is no
longer true.

• Finite difference

• Relative error estimate

• Update solution

You might also like