02.
Roots of Equation
Computational Mathematics 1
Roots in Engineering
• Given
𝑔𝑚 𝑔𝑐𝑑
𝑣 𝑡 = tanh 𝑡
𝑐𝑑 𝑚
• Subtracting 𝑣 𝑡
𝑔𝑚 𝑔𝑐𝑑
𝑓 𝑚 = tanh 𝑡 − 𝑣(𝑡)
𝑐𝑑 𝑚
• “roots” problem, value of 𝑚 makes 𝑓 𝑚 = 0
Graphical method (example 2.1)
• In Matlab, determine the root of the equation for
the mass of the bungee jumper with a drag
coefficient of 0.25 kg/m to have a velocity of 36
m/s after 4s of free fall.
>> cd = 0.25; g = 9.81; v = 36; t = 4;
>> mass = linspace(0,200);
>> fmass = sqrt(g*mass/cd).*tanh(sqrt(g*cd./mass)*t)-v;
>> plot(mass,fmass),grid
Incremental Search
• 𝑓(𝑥) changed sign on opposite sides of the root
𝑓 𝑥𝑙 𝑓 𝑥𝑢 < 0
• locating an interval where the function changes
sign
• too small length, very time consuming
• too large length, very closely spaced roots might
be missed
Implementing Incremental Search
• 𝑓𝑢𝑛𝑐 within the range from 𝑥𝑚𝑖𝑛 to 𝑥𝑚𝑎𝑥
• 𝑛𝑠 number of intervals within the range
• 𝑓𝑜𝑟 loop step through each interval
• upper and lower bounds are stored in an
array 𝑥𝑏 (when sign change occurs)
M-file
(example 2.2)
• Use the M-file “𝑖𝑛𝑐𝑠𝑒𝑎𝑟𝑐ℎ” to identify
brackets within the interval [3, 6] for the
function:
𝑓 𝑥 = sin 10𝑥 + cos 3𝑥
>> incsearch(@(x) sin(10*x)+cos(3*x),3,6)
Bisection
• interval is always divided in half
• location of the root is determined within
the subinterval where the sign change
occurs
• subinterval then becomes the interval for
the next iteration
• process is repeated until the root
Using ex. 2.1
• function changes sign between values of 50 and 200
• Estimated root (𝑥𝑟 ) lies at midpoint
50 + 200
𝑥𝑟 = = 125
2
• Relative error (𝜖𝑡 )
142.74 − 125
𝜖𝑡 = × 100% = 12.24%
142.74
• While
𝑓 50 𝑓 125 = 1.87
∴ 𝑓 𝑥𝑙 𝑓 𝑥𝑢 < 0, root “must be” located between 125 and 200
• Process is repeated until acquiring accurate result
• Since the root is unknown, the actual error (𝜖𝑎 )
𝑥𝑟_𝑛𝑒𝑤 − 𝑥𝑟_𝑜𝑙𝑑
𝜖𝑎 = × 100%
𝑥𝑟_𝑛𝑒𝑤
• Once reached stopping criterion 𝜖𝑠 , the process is terminated
(example 2.3.a)
• Using Example 2.1, determine the roots until
𝜖𝑠 = 0.5% or below. Graph 𝜖𝑡 and 𝜖𝑎 vs iteration.
(example 2.3.b)
• Implement bisection using MATLAB.
False Position
• Interpolation method, locates the root by joining 𝑓(𝑥𝑙 ) and
𝑓(𝑥𝑢 ) with a straight line
• the intersection of the straight line
𝑓 𝑥𝑢 (𝑥𝑙 − 𝑥𝑢 )
𝑥𝑟 = 𝑥𝑢 −
𝑓 𝑥𝑙 − 𝑓(𝑥𝑢 )
• From Example 2.1
𝑥𝑙 = 50; 𝑓 𝑥𝑙 = −4.579
𝑥𝑢 = 200; 𝑓 𝑥𝑢 = 0.860
𝑥𝑟 = 176.277
𝜖𝑡 = 23.5%
• Next iteration 𝑓 𝑥𝑙 𝑓 𝑥𝑟 = −2.593
𝑥𝑙 = 50
𝑥𝑢 = 176.277
Example 4
• Use bisection and false position to locate
the root of
𝑓 𝑥 = 𝑥 10 − 1
• Homework: verify your answer using
MATLAB
Thank you for your attention
