Department of Mechanical Engineering

College of Engineering, University of the Philippines

PS1 and MP1

ME 91: Numerical Methods for Mechanical Engineering

Prepared by: Asst.Prof. Ryan M. Cabrera

Email: rmcabrera.updme@gmail.com

Instructions:
• For the PS part, you are encouraged to use Matlab R
but you may use other computing
tools as appropriate. For the PS1 Report, you are required to typeset your work in MS-
Word, or optionally in LATEX, and then submit a printed copy. For those who will be using
Matlab R
, include your script (in Courier New font) in the solution. For those solving
them in MS-Excel, you don’t need to include the results in tabular forms; instead, submit
the MS-Excel files through my email on the due date. Note: Aside from the printed
copies, you are also required to send the PDF file of PS1 Report through my email.

• For the MP part, you are required to use Matlab R

, to develop your script and function
files (M-files). Then you are also required to submit a Report with description of the
numerical methods being implemented, test run results, and corresponding explanations,
in printed copies; in addition, the code (M-files) and MP1 Report (PDF) must be submitted
through my email as well.

• NOTE: Always observe the Honor Code: ”You shall not take unfair advantage over
your peers. You are strictly prohibited from DELIBERATELY copying the work of your
classmates. HOWEVER, you may collaborate with your partner ONLY for both the PS
and MP.”

• Due Dates: The due date for PS1 Report (hard copies) will be on October 28, 2019
(before your Midterm Exam) while the PS1 Report (PDF) can be submitted on the same
date but not later 11:59PM; for MP1, the due date will be on November 4, 2019 (M-files
and PDF of MP1 Report through my email) and November 5, 2019 (hard copies of MP1
Report).

1
PS1 Problems:
Problems are based on topics discussed in Module 1 as well as the use of Matlab
R
, in general.

R
If not specified, other computing tools can be used; otherwise, use Matlab as default.

1. A quarterback throws a pass to his wide receiver running a route. The quarterback releases
the ball at a height of hQ . The wide receiver is supposed to catch the ball straight down
the field 60 [ft] away at a height of hR .

Figure 1: Figure for Problem 1: A quarterback

throws the football to the receiver.

Recall the kinematic coordinates of projectile motion, (x, y), are defined as follows:

x = xo + vo cos θ t (1a)
1
y = yo + vo sin θ t − gt2 (1b)
2
(a) Show that, if xo = 0 and yo = hQ , the vertical position of the football’s motion is
given by:

x2 g
y = hQ + x tan θ − (2)
2vo2 cos2 θ
(b) At x = 60 [ft] and y = hR , find the angle θ such that the receiver will catch the
football. Note the following given data: hQ = 6.5 [ft], hR = 6.85 [ft], while the initial
velocity of the throw is vo = 6.5 [fps]. Use the Bisection Method to find the solution
and use ε = 0.01◦ . Note that the acceleration due to gravity is g = 32.2 [ft/s2 ]. [Hint:
Set up first the ”f (θ) = 0” format of the problem, based on Eq. (2).]

2. The force F acting between a particle with a charge q and a round disk of radius R and
charge Q is given by the equation:
 
Qqz z
F = 1− √ (3)
2εo z 2 + R2
where εo = 0.885 × 10−12 [C2 /(Nm2 )] is called the permittivity constant. Suppose that the
disk has charge Q = 9.5 × 10−6 [C] and radius R = 0.1 [m] while the particle has charge

2
 Q = 2.75 × 10−5 [C], find the distance z from the disk (see Fig. 2 for reference) such
that the electrostatic force is F = 0.5 [N]. Use the Newton-Raphson Method in finding the
solution within accuracy of ε = 1 × 10−5 [m].

Figure 2: Figure for Problem 2: Electrostatic force

between a disk and a particle.

3. A trough of length L has a cross section in the shape of a semicircle with radius R.
Initially, it is fully-filled with water at volume Vo [m3 ]; after some time, the volume of
water discharged through the hole at the bottom of the trough is Vd [m3 ].

(a) Derive the equation of volume of the remaining water as function of the height, h
(from the top of the trough to the surface of remaining water passing along the radial
direction), i.e., derive the equation for V = V (h).
(b) Suppose that L = 10 [m], R = 1.5 [m], and Vd is 40% of initial volume, find the depth
of remaining water, within accuracy of ε = 0.001 [m]. Use the Regula Falsi Method
in finding the solution.

4. Given the following function:

x2 − 1
f (x) = (4)
ex + 1
where x ∈ [1, 10]. Suppose that Vn (x) is the Vandermonde polynomial and Ln (x) is the
Lagrange polynomial, where n is the degree of the polynomial interpolating (n+1) number
of points, do the following:

(a) For n = 4, plot f (x), V4 (x), and L4 (x) versus x ∈ [1, 10]; put a legend in the graph.
(b) For n = 8, plot f (x), V8 (x), and L8 (x) versus x ∈ [1, 10]; put a legend in the graph.
What can you say about the increase in degree of polynomial, for both Vandermonde
and Lagrange?

5. The torsion stress factor for rectangular cross-section prismatic bars made of homogeneous
and isotropic materials (i.e., conventional materials such as metals) is given by Danao and
Cabrera (2007) as follows:

B
K= (5)
A
3
 where

  h i 
∞ tanh (2m+1)πc
1  192  πc  X 2

A= 1 − 5 tanh +  (6a)
3 π c 2 m=1
(2m + 1)5 
 
∞
8 1 X 1
B =1− 2  πc

+ h i (6b)
π cosh 2 2
m=1 (2m + 1) cosh
(2m+1)πc
2

It should be noted that since the functions A = A(c) and B = B(c) are both functions of
the aspect ratio c (i.e., c = b/a, where a and b are the half-length dimensions of the narrow
and wide sides of the rectangle, respectively), then the torsion stress factor K should also
be a function of c, i.e., K = K(c). For given values of {cj } = {1.0, 1.5, 2.0, 3.0, 5.0},
find the Lagrange polynomial Ln (c) to replace the complicated-looking function of torsion
stress factor K(c). Also, plot K(c) and Ln (c) versus c ∈ [1.0, 5.0] and put a legend in the
graph.
6. A fin is an extended surface used to transfer heat from a base material (at x = 0) to an
ambient space. Heat flows from the base material through the base of the fin, through
its outer surface, and through the tip (see Fig. 3 for reference). Measurement of the
temperature distribution along a fin gives the following data (see attached text file):
ME91_PS1_Problem6_Data_HeatTransferThruAFin

Figure 3: Figure for Problem 6: Heat transfer

through a fin.

The fin as length of L = 10 [cm]. The heat flux per unit area (of the fin’s cross-section),
qx , is given by:

dT
qx = −κ (7)
dx
where k is the fin’s thermal conductivity. Compute the heat flux qx (xj ) at each node xj
(for j = 1, 2, 3, ..., 11) over the entire length of the fin. Also, plot the discrete heat flux
qx (xj ) versus xj ∈ [0, L], for each j = 1, 2, 3, ..., 11. Note that κ = 240.5 [(W/m)/K] is the
fin’s thermal conductivity. [Hint: Approximate the first-order derivative in Eq. (7) using
the three-point forward difference at x = 0 and the three-point backward at x = L; as for
the interior nodes, use the three-point midpoint difference scheme.]

4
7. A 30 [ft]-long uniform beam is simply supported at the left end (x = 0) and clamped at
the right end (x = L). The beam is subjected to the triangular load shown (see Fig. 4 for
reference). The deflection of the beam is given by the differential equation:

d2 y M
2
= (8)
dx EI

Figure 4: Figure for Problem 7: Simply-supported

beam at left end and clamped at right end.

where M = M (x) is the bending moment which is a function of x- coordinate (lengthwise

direction of the beam), while E = 29 × 106 [psi] is the elastic modulus and I = 730 [in4 ]
is the beam’s cross-sectional moment of inertia. It should be noted that the deflection
y = y(x).

(a) Develop a script file that calculates and plots the bending moment M (x) versus
x ∈ [0, L]. Use the data given in the attached text file:
ME91_PS1_Problem7_Data_BeamSimplySupportedAndClamped
(b) Plot M (x) versus x ∈ [0, L].

8. The central span of the Golden Gate bridge is 4200 [ft] long and each of the towers has
height from the roadway of 500 [ft] (see Fig. 5 for reference). The shape of the main
suspension cables can be approximately modeled by the equation:

ex/C + e−x/C
 
f (x) = C − 1 , for a≤x≤b (9)
2

where C = 4491 and [a, b] = [−2100, 2100] (as measured in [ft]). Suppose that the length
of the suspension cables, L, is calculated using the following:
Z bp
L= 1 + [f 0 (x)]2 dx (10)
a

Calculate the length L of the Golden Gate’s suspension cable by approximating Eq. (10)
using Composite Simpson’s Rule.

5
 Figure 5: Figure for Problem 8: Calculating the
length of suspension cables in the Golden Gate
bridge.

MP1 Problems:
This is purely Matlab R
implementation of the numerical methods discussed in Module 1;
hence, you are not allowed to use any other computing tool, for the purpose of consistency. You
may work on this during the Laboratory Sessions or at home or anywhere you find convenient.

1. Implement the following numerical methods of solving ”f (x) = 0” problems:

(a) Bisection Method

(b) Secant Method
(c) Newton-Raphson Method
(d) Regula-Falsi Method

make sure to incorporate the IVT checker whenever applicable. The user must have a
choice of which method to be used. The input will usually be the following: the function
f (x), the interval [a, b], and tolerance ε. The output will be the solution (or root) and plot
of f (x) versus x ∈ [a, b].

2. Implement the following interpolation methods:

(a) Vandermonde interpolation

(b) Lagrange interpolation

The user must choose between the following type of input:

• The values of x in the nodes (i.e., xj ) and f (x), for which yj ≡ f (xj ); hence, the
coordinates (xj , yj ) can be specified.
• The coordinates (xj , yj ) are already specified.

In addition, the input must also include the order (or degree) n of the approximating
polynomial and the interval [a, b]. The output will be the equation of the polynomial and
its plot versus x ∈ [a, b].