# ME 301

Kinematics & Dynamics of Machines

Class Notes

Dr. Bob

Mechanical Engineering

Ohio University

williar4@ohio.edu

http://www.ent.ohiou.edu/~bobw/
2

INTRODUCTION............................................................................................ 3
MATRIX-VECTOR INTRODUCTION........................................................ 11
MATLAB INTRODUCTION........................................................................ 25
MOBILITY..................................................................................................... 29
FOUR-BAR MECHANISM POSITION ANALYSIS................................... 34
GRASHOF’S LAW........................................................................................ 49
SLIDER-CRANK MECHANISM POSITION ANALYSIS.......................... 53
VELOCITY ANALYSIS INTRODUCTION ................................................ 62
FOUR-BAR MECHANISM VELOCITY ANALYSIS................................. 67
SLIDER-CRANK MECHANISM VELOCITY ANALYSIS........................ 78
ACCELERATION ANALYSIS INTRODUCTION...................................... 85
FOUR-BAR MECHANISM ACCELERATION ANALYSIS ...................... 91
SLIDER-CRANK MECHANISM ACCELERATION ANALYSIS ........... 100
INPUT MOTION SPECIFICATION........................................................... 106
DYNAMICS INTRODUCTION.................................................................. 110
MASS, CENTER OF GRAVITY, MASS MOMENT OF INERTIA.......... 115
SINGLE ROTATING LINK INVERSE DYNAMICS................................ 124
FOUR-BAR MECHANISM INVERSE DYNAMICS................................. 133
SLIDER-CRANK MECHANISM INVERSE DYNAMICS........................ 147
CAM INTRODUCTION.............................................................................. 156
ANALYTICAL CAM SYNTHESIS............................................................ 168
GEAR INTRODUCTION............................................................................ 175
GEAR TRAINS AND GEAR STANDARDIZATION................................ 185

3
ME 301 Kinematics & Dynamics of Machines

Introduction

Kinematics:

Kinema - Greek for motion

Dynamics:

Rigid Body Mechanics Diagram:

Required Math: Geometry, trigonometry, vectors, matrices, calculus

Mechanisms: linkages, cams, gears, gear trains

Analysis vs. Synthesis
• Analysis – determination of position, velocity, acceleration, etc.
for a given mechanism
• Synthesis – design of mechanism to do a specific job

4
Mobility - number of degrees-of-freedom (dof):

• Structure – static, no motion
• Mechanism – 1 dof device with rigid links connected with joints
• Machine – collection of mechanisms to transmit force (input /
output)
• Robot – an electromechanical device having greater than 1 dof,
programmable for a variety of tasks.

Motion – Translation and Rotation

Planar – all motion is 2D (projected onto a common plane)

Helical - rotation about fixed axis and translation along axis - screw

Spherical - 3D motion; all points in a body moves about a fixed point

Spatial - 3 independent translations and rotations

5
Joints – Pairing elements
Lower – surface contact
Revolute – pin joint, turning pair

Prismatic – sliding pair

Higher – point or line contact
ball bearing

gears

cam and follower

Kinematic chain – number of links connected by joints
open – serial robot
closed – mechanism, parallel robot

Kinematic Inversion – change which link is fixed – same relative
motion, different absolute motion.

Examples – in class; also see following Atlas
6
A Brief Atlas of Structures, Mechanisms, and Robots
Dr. Bob

Statically Determinate Structure Statically Indeterminate Structure

4-Bar Mechanism Offset Slider-Crank Mechanism

Inverted Slider-Crank Mechanism Scotch-Yoke Mechanism
7

Geared 5-Bar Mechanism Stephenson I 6-Bar Mechanism

Stephenson II 6-Bar Mechanism Stephenson III 6-Bar Mechanism

Watt I 6-Bar Mechanism Watt II 6-Bar Mechanism
8

Spur Gear Mechanism Cam-and-Follower Mechanism

Geneva Wheel Mechanism Planar 3-dof Robot

Adept 4-dof SCARA Robot Mitsubishi 5-dof Robot
9

PUMA 6-dof Robot NASA 8-dof ARMII

2-dof 5-Bar Parallel Robot 3-dof 3-RRR Parallel Robot

3-dof 3-RPR Parallel Robot 3-dof Carpal Wrist

10
Connection to Machine Design

In ME 301 we focus on kinematics & dynamics analysis, not
synthesis (design).

However, the skills gained in this course support general
(electro)mechanical design.

Before one can design a machine, the required motion must be
satisfied. All design candidates must be analyzed regarding the
motion each would provide (position, velocity, and acceleration, both
translational and rotational). This requires kinematics analysis.

Before one can size the links, joints, bearings, gear box, and actuators
condition(s) must be known, for statics and dynamics. This requires
dynamics analysis.

Engineering design is iterative by nature: each candidate design must
be thoroughly analyzed to determine its performance relative to the
design specifications and relative to other design candidates.

This kinematics & dynamics analysis is facilitated using a computer.
Without the computer, it is difficult to determine the worst-case
inefficiently applied.

The goal of ME 301 is to give the student general skills in general
matrix/vector-based kinematics and dynamics analysis which may be
applied in later classes and later careers.
11
Matrix-Vector Introduction
Vectors

Arrow in the plane with magnitude and direction. Used to
represent position, velocity, acceleration, force. Also, arrow normal
to the plane to represent angular velocity, angular acceleration, and
torque (moment) vectors (see later in notes).

Cartesian representation:

Polar representation: Magnitude at angle: θ @ P

(atan2 - quadrant-specific inverse tangent function)

12

components in same coordinate frame.

13
Vector Dot Product

Dot product is projection of one vector onto another. Scalar
result.

14
Vector Cross Product

Cross product of two vectors gives a third vector mutually
perpendicular to the original two vectors. Vector result.

Direction via right-hand-rule: Put right hand fingers along first
vector
1
P , rotate into second vector
2
P ; right thumb is direction of
2 1
P P × .
15
k
ˆ
Vectors

In planar kinematics, angular velocity, angular acceleration, and
torque (moment) vectors are arrows along about the k
ˆ
axis (the unit
direction for the Z axis, perpendicular to the plane). Still has
magnitude and direction, but simplifies to a single component with ±
sign. We will often represent these k
ˆ
vectors by curled arrows in the
XY plane.

Example:

k
ˆ
ω ω ± = ;

+ ccw (curling in the direction of the right hand fingers)
– cw (curling in the opposite direction of the right hand fingers)
16
Vector Examples

)

¹
¹
´
¦
=
2
1
1
P
)

¹
¹
´
¦
=
2
3
2
P

2 1
P P

Dot Product: = •
2 1
P P

Cross Product: = ×
2 1
P P
17
Matrices

Matrix: m x n array of numbers, where m is the number of rows
and n in the number of columns.

| |
11 12 1
21 22 2
1 2
n
n
m m mn
a a a
a a a
A
a a a

=

"
"
# # % #
"

Used to simplify and standardize the solution of n linear equations in
n unknowns (where m=n). Used in velocity, acceleration, and
dynamics analysis linear equations (not used in position which is a
non-linear solution).

Special Matrices

Square (m=n=3)
| |
11 12 13
21 22 23
31 32 33
a a a
A a a a
a a a

=

Diagonal
| |
11
22
33
0 0
0 0
0 0
a
A a
a

=

Identity
| |
1 0 0
0 1 0
0 0 1
I

=

18
Transpose
| |
11 21 31
12 22 32
13 23 33
T
a a a
A a a a
a a a

=

Symmetric
| | | |
11 12 13
12 22 23
13 23 33
T
a a a
A A a a a
a a a

= =

Column Vector (3x1 matrix)
{ }
1
2
3
x
X x
x
¦ ¹
¦ ¦
=
´ 
¦ ¦
¹ )

Row Vector (1x3 matrix)
{ } { }
1 2 3
T
X x x x =

a b e f a e b f
c d g h c g d h
+ +

+ =

+ +

Matrix Multiplication with Scalar Just multiply each term

a b ka kb
k
c d kc kd

=

19
Matrix Multiplication
| | | || | | || |
C A B B A = ≠

Row, Column indices have to line up as follows:

| | | || |
( ) ( )( )
C A B
mxn mxp pxn
=

That is, the number of columns in the left-hand matrix must equal the
number of rows in the right-hand matrix; if not, the multiplication is
undefined and cannot be done! Multiplication proceeds by
multiplying and adding terms along the rows of the left-hand matrix
and down the columns of the right-hand matrix: (use your index
fingers from the left and right hands):

Example:
| |
( ) ( )( )
2 1 2 3 3 1
g
a b c ag bh ci
C h
d e f dg eh fi
i
x x x
¦ ¹
+ +
¦ ¹
¦ ¦
= =
´  ´ 

+ +
¹ )
¦ ¦
¹ )

note the inner indices (p=3) must match, as stated above and the
dimension of the result is the outer indices, i.e. 2x1.

20
Matrix Multiplication Examples

| |
1 2 3
4 5 6
A

=

| |
7 8
9 8
7 6
B

=

| | | || |
7 8
1 2 3
9 8
4 5 6
7 6
7 18 21 8 16 18 46 42
28 45 42 32 40 36 115 108
C A B =

=

+ + + +

= =

+ + + +

( ) ( )( )
2 2 2 3 3 2 x x x ≡

| | | || |
7 8
1 2 3
9 8
4 5 6
7 6
7 32 14 40 21 48 39 54 69
9 32 18 40 27 48 41 58 75
7 24 14 30 21 36 31 44 57
D B A =

=

+ + +

= + + + =

+ + +

( ) ( )( )
3 3 3 2 2 3 x x x ≡
21
Matrix Inversion

Matrix “division”: given
| | | || |
C A B = , solve for [B]

| | | || |
C A B = ⇒
| | | | | | | || |
| || |
| |
1 1
A C A A B
I B
B
− −
=
=
=

| | | | | |
1
B A C

⇒ =

Matrix [A] must be square to invert.

| || | | | | | | |
1 1
A A A A I
− −
= =

where [I] is the identity matrix, the matrix “1”. To calculate the
matrix inverse:

| |
( )
1
A
A

=

where: A Determinant of [A]

( ) ( )
T
A A =

Cofactor(A)
( )
1
i j
ij ij
a M
+
= −

Minor
ij
M is the determinant of the submatrix with row i
and column j removed.
22
System of Linear Equations

We can solve n linear equations in n unknowns with the help of
a matrix. For n=3:

11 1 12 2 13 3 1
21 1 22 2 23 3 2
31 1 32 2 33 3 3
a x a x a x b
a x a x a x b
a x a x a x b
+ + =
+ + =
+ + =

Using matrix multiplication (backwards), this is written as:

| |{ } { }
A x b =

where:

| |
11 12 13
21 22 23
31 32 33
a a a
A a a a
a a a

=

(known coefficients)

{ }
1
2
3
x
x x
x
¦ ¹
¦ ¦
=
´
¦ ¦
¹ )
(unknowns to be solved)

{ }
1
2
3
b
b b
b
¦ ¹
¦ ¦
=
´ 
¦ ¦
¹ )
(known right-hand sides)

Unique solution { } | | { }
1
x A b

= only if [A] has full rank. If not,
0 A = and the inverse of matrix [A] is undefined (dividing by zero).

23
Matrix Example

Solution of simultaneous linear equations.

1 2
1 2
2 5
6 4 14
x x
x x
+ =
+ =

)

¹
¹
´
¦
=
)

¹
¹
´
¦

14
5
4 6
2 1
2
1
x
x

| |

=
4 6
2 1
A { }
)

¹
¹
´
¦
=
2
1
x
x
x { }
)

¹
¹
´
¦
=
14
5
b

{ } | | { } b A x
1 −
=

( ) ( ) 8 6 2 4 1 − = − = A Determinant non-zero; unique solution!

| |

=

=

8 / 1 4 / 3
4 / 1 2 / 1
1 6
2 4
1
1
A
A

check: | || | | | | | | |

= = =
− −
1 0
0 1
2
1 1
I A A A A

)

¹
¹
´
¦
=
)

¹
¹
´
¦

=
)

¹
¹
´
¦
2
1
14
5
8 / 1 4 / 3
4 / 1 2 / 1
2
1
x
x

check: Plug answer into original equations and compare to the
{b} we need to get.
24
Vector and Matrix Matlab Examples

P1 = [1;2;0]; % Define two vectors
P2 = [3;2;0];
sum1 = P1+P2; % Vector addition
sum2 = P2+P1;
dot1 = dot(P1,P2); % Vector dot product
dot2 = dot(P2,P1);
cross1 = cross(P1,P2); % Vector cross product
cross2 = cross(P2,P1);

A = [1 2;6 4]; % Define a matrix and vector
b = [5;14];
dA = det(A); % Calculate determinant of A
invA = inv(A); % Calculate the inverse of A
x = invA*b; % Solve linear equations
x1 = x(1); % Extract answers
x2 = x(2);
A*x % Check answer – should be b
25
Matlab Introduction

Matrix laboratory

Control systems simulation and design software. Very
widespread in other fields. Introduction to basics, programming,
plots, animation, matrices, vectors. Based on C language,
programming is vaguely C-like, but much simpler to use. Sold by
Mathworks (http://www.mathworks.com).

of one textbook (can use it for many classes!). ENT college has a
Matlab license; it is installed in most computer labs.

Double-click on Matlab icon to get started. Type

>>demo

to get a comprehensive overview of Matlab including built-in
functions. Try all the categories under Matlab first; you can ignore
Toolboxes, Simulink, and Stateflow for now. (Exception: there is
Symbolic Math under Toolboxes for the adventurous student!).

Type in commands (such as the Vector/Matrix examples given
earlier) at the Matlab prompt >>. Press <Enter> to see result or ;
<Enter> to suppress result.

Recommended operation mode: m-files. Put your sequence of
Matlab statements in an ASCII file name.m (create a file with the
beautiful Matlab Editor/Debugger - this is color-coordinated, tab-
friendly, with parentheses alignment help and debugging
26
capabilities). A % indicates a comment. One basic way to run your
program is to hit the ‘save and run’ button on the editor toolbar.

Alternative: at the >> prompt type the M-File name name, without
the .m, assuming your file is in the search path. Matlab language is
interpretive and executes line-by-line. Use the ; at the end of
statements to suppress intermediate results. If you use this
suppression, the variable name still holds the resulting value(s) – just
type the variable name at the prompt after the program runs to see the
value(s). If there is a syntax or programming logic error, it will give
a message at the bad line and then quit. Type:

>>who

to show you what variables you have defined;

>>whos

will show the variables, plus their matrix dimensions (scalar, vector
array, or matrix), very useful for debugging. Plus, after running a
file, place the cursor over different variables in the M-File inside the
Editor/Debugger to see the values! On-line help is generally great:

>>help

Example m-files (given on the following two pages)

1) MatEx1.m: Input, programming, plots, animation.

2) MatEx2.m: Matrix and vector definition, multiplication, transpose,
and solution of linear equations.
27
%---------------------------------------------------------------
% Matlab Example Code 1: MatEx1.m
% Matrix, Vector examples
% Dr. Bob, ME 301
%---------------------------------------------------------------

clc; clear; % Clear the cursor and clear any previously defined variables

%
% Matrix and Vector definition, multiplication, and transpose
%

A1 = [1 2 3; ... % Define 2x3 matrix [A1] (... is continuation line)
1 -1 1];
x1 = [1;2;3]; % Define 3x1 vector {x1}

v = A1*x1; % 2x1 vector {v} is the product of [A1] times {x1}

A1T = A1'; % Transpose of matrix [A1]
vT = v'; % Transpose of vector {v}

%
% Solution of linear equations Ax=b
%

A2 = [1 2 3; ... % Define matrix [A2] to be a 3x3 coefficient matrix
1 -1 1; ...
8 2 10];

b = [3;2;1]; % Define right-hand side vector of knowns {b}

detA2 = det(A2); % First check to see if det(A) is near zero

x2 = inv(A2)*b; % Calculate {x2} to be the solution of Ax=b by inversion

check = A2*x2; % Check results;

z = b - check; % Better be zero!

%
% Display the user-created variables (who), with dimensions (whos)
%

who
whos

%
% Display some of the results
%

v
x2
z
28
%---------------------------------------------------------------
% Matlab Example Code 2: MatEx2.m
% Menu, Input, FOR loop, IF logic, Animation, and Plotting
% Dr. Bob, ME 301
%---------------------------------------------------------------

clc; clear; % Clear the cursor and clear any previously defined variables

r = 1; L = 2; DR = pi/180; % Constants

%
% Input
%
the = input('Enter [th0, dth, thf] (deg): ') % User types input
th0 = the(1)*DR; dth = the(2)*DR; thf = the(3)*DR; % Initial, delta, final thetas
th = [th0:dth:thf]; % Assign theta array
N = (thf-th0)/dth + 1; % Number of iterations for loop

%
%
if anim == 1 % Animate if user wants to
figure; % Give a blank graphics window
for i = 1:N; % For loop to animate
x2 = [0 L*cos(th(i))]; % Single link coordinates
y2 = [0 L*sin(th(i))];
plot(x2,y2); grid; % Animate to screen
set(gca,'FontSize',18);
xlabel('\itX (\itm)'); ylabel('\itY (\itm)');
axis('square'); axis([-2 2 -2 2]); % Define square plot limits
pause(1/4); % Pause to see animation
if i==1 % Pause to maximize window
pause; % User hits Enter to continue
end
end
end

%
% Calculate circle coordinates and cosine function
%
xc = r*cos(th); % Circle coordinates
yc = r*sin(th);
f1 = cos(th); % Cosine function of theta
f2 = sin(th); % Sine function of theta

%
% plots
%
figure; % Co-plot cosine and sine functions
plot(th/DR,f1,'r',th/DR,f2,'g'); grid; set(gca,'FontSize',18);
legend('Cosine','Sine');
axis([0 360 -1 1]); title('Functions of \it\theta');
xlabel('\it\theta (\itdeg)'); ylabel('Functions of \it\theta');

figure; % Plot circle
plot(xc,yc,'b'); grid; set(gca,'FontSize',18);
axis(['square']); axis([-1.5 1.5 -1.5 1.5]); title('Circle');
xlabel('\itX (\itm)'); ylabel('\itY (\itm)');
29
Mobility

Mobility:

Degrees-of-freedom (dof):

How many dofs does an unconstrained planar link have?

What is the effect of constraining that link with a revolute joint?

Grubler's Criterion: Planar Jointed Devices

Where: M is the mobility
N is the total # of links, including ground
J
1
is the number of one-degree-of-freedom joints
J
2
is the number of two-degree-of-freedom joints
30
One-degree-of-freedom joints:

Revolute

Prismatic

Two-degree-of-freedom joints (all have rolling and sliding):

Cam joint

Gear joint

Slotted-pin joint

Caution: m links joining at one revolute location, must count m-1
joints!

Caution: must count ground link (its freedom is subtracted in formula
with n-1.

Planar mechanical device classification:

1 > M

1 = M

0 = M

0 < M

31
Planar Mobility Examples:

4) Scotch Yoke mechanism

5) Cam and follower

32
6) Gear pair

9) Statically-determinate structure

10) Statically-indeterminate structure

33

14) 3-RRR parallel robot
34
Four-Bar Mechanism Position Analysis

Position (Displacement) Analysis: determination of relative
orientation/ position of links in a mechanism. Required for testing
motion of a synthesized mechanism. Also required for further
analysis: velocity, acceleration, dynamics, forces.

Generic Mechanism Position Analysis Statement: Given the
mechanism and one dof of position input, calculate the position
unknowns.

Four-bar Mechanism Position Analysis

Step 1. Draw the Kinematic Diagram:

r
1
1
r
2
2
– input angle
r
3
3
– coupler angle
r
4
4
– output angle
All angles measured in right-hand sense from horizontal to link.
35
Step 2. State the problem:

Step 3. Draw the Vector Diagram. Define all angles in positive
sense, measured from the right horizontal to the link vector (tail-to-
head). Don’t try to force acute angles; the relationships we can see so
)

¹
¹
´
¦
=
θ
θ
sin
cos
L
L
P ; good for all θ .
Vector Diagram:

Step 4. Derive the Vector-Loop-Closure Equation. Start at one
equation by starting and ending at same points, but choosing a
different path.

36
Step 5. Write XY Components for Vector-Loop-Closure Equation.
Break one vector equation into its two scalar components (X and Y):

Step 6. Solve for the Unknowns from the XY Equations. Two
coupled nonlinear equations in the two unknowns
4 3
,θ θ . Isolate and
eliminate
3
θ and solve for
4
θ . Then go back to find
3
θ .

37
This equation has the form:

Solve using the tangent half angle substitution (Text Equation 4.9):

|
.
|

\
|
=
2
tan
4
θ
t
2
2
4
1
1
cos
t
t
+

= θ
2
4
1
2
sin
t
t
+
= θ

We converted a complicated coupled transcendental set of
equations into a quadratic polynomial. Much easier to solve (but we
doubled the order of the equation!).

Two solutions for
4
θ :

38
With factor two, no need to use the atan2 function.

Why two solutions? (Graphically demonstrate the two branches.)

What if 0
2 2 2
< − + G F E ? Imaginary solution, physically
means the mechanism cannot assemble for that input angle. See
section on Grashof's Law.

Go back to find
3
θ , one for each solution branch. Go back to original
two XY scalar equations.

Use ratio of Y to X equations:

Show graphical interpretation:
|
|
.
|

\
|

=

X X
Y Y
A B
A B
1
3
tan θ

39
The basic four-bar mechanism position analysis problem is now
solved. Now that we know the angular unknowns, we can find the
translational position of any point on the mechanism, e.g. coupler
point C:

Four-bar mechanism transmission angle: Transmission angle u:
relative angle between coupler and output links. Measure of
o
is ideal; 0,180
o
zero
transmission; as a rule of thumb, the absolute value of u should
remain in the range
D D
140 40 < < u for good transmission in a
mechanism. By geometry:

40
Four-Bar Mechanism Position Analysis: Term Example 1

Given
7
8
3
18 . 11
4
3
2
1
=
=
=
=
r
r
r
r
in
1
2
3
4
0.284
0.076
0.203
0.178
r
r
r
r
=
=
=
=
m

and
D
3 . 10
1
= θ (Ground link is 11" over and 2" up). Also given
5
/
=
A C
R (in) and
D
9 . 36
3
= δ for the coupler link point of interest.

Snapshot Analysis (one input angle)

Given this mechanism and
D
30
2
= θ , calculate u θ θ , ,
4 3
, and
C
P for
both branches. Results:

0.076
0.005
0.036
E
F
G
=
=
=

Branch t
3
θ
4
θ
u
C
P
Open 1.79
D
8 . 53
D
7 . 121
D
9 . 67
0.06, 0.16
Crossed -1.57
D
0 . 47 −
D
9 . 114 −
D
9 . 67
0.19, 0.02

These two branch solutions are demonstrated in the figures on the
following page. We use the SI system (m). Note u is identical for
both branches due to the conventions presented earlier.
41
-0.1 0 0.1 0.2 0.3
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.25
X (m)
Y

(
m
)

4-bar Example Snapshot, Open Branch
-0.1 0 0.1 0.2 0.3
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.25
X (m)
Y

(
m
)

4-bar Example Snapshot, Crossed Branch
42
Graphical Solution: The 4-bar position analysis may be solved
graphically, merely by drawing the mechanism and determining the
mechanism closure. This is an excellent method to validate your
computer results at a given snapshot.

• Draw the known ground link (points O
2
and O
4
).
• Draw the given input link 2 length at the given angle (point A).
• Draw a circle of radius r
3
, centered at point A.
• Draw a circle of radius r
4
, centered at point O
4
.
• These circles intersect in general in two places.
• Connect the two branches and measure the unknown values.

Graphical Solution Figure:

43
4-bar Snapshot Matlab code:
This program solves the 4-bar position analysis problem for both
branches given a single
2
θ . The results are drawn to the screen.

%-------------------------------------------------------------
% 4-bar linkage snapshot position analysis - both branches
% Fbarplec.m, with graphical output, Dr. Bob, ME 301
%-------------------------------------------------------------
clc; clear; % Clear cursor and clear previously defined variables

% Inputs
DR = pi/180;
R = input('Enter [r1, r2, r3, r4, rca, th1, th2, del3] (m and deg): ');
r1 = R(1); r2 = R(2); r3 = R(3); r4 = R(4); rca = R(5);
th1 = R(6)*DR; th2 = R(7)*DR; del3 = R(8)*DR; % Change degrees to radians
r1x = r1*cos(th1); r1y = r1*sin(th1);

% Position analysis: theta4
E = 2*r4*(r1*cos(th1) - r2*cos(th2));
F = 2*r4*(r1*sin(th1) - r2*sin(th2));
G = r1^2 + r2^2 - r3^2 + r4^2 - 2*r1*r2*cos(th1-th2);
t(1) = (-F + sqrt(E^2 + F^2 - G^2)) / (G-E); % Crossed Branch
t(2) = (-F - sqrt(E^2 + F^2 - G^2)) / (G-E); % Open Branch
th4(1) = 2*atan(t(1));
th4(2) = 2*atan(t(2));

% th3, coupler point, transmission angle; calculate for both branches
for i = 1:2,
ax = r2*cos(th2); % theta3
ay = r2*sin(th2);
bx = r4*cos(th4(i)) + r1x;
by = r4*sin(th4(i)) + r1y;
th3(i) = atan2(by-ay,bx-ax);
bet = th3(i) + del3; % coupler point
pcx(i) = r2*cos(th2) + rca*cos(bet);
pcy(i) = r2*sin(th2) + rca*sin(bet);
mu(i) = abs(th4(i)-th3(i)); % transmission angle
end

% Plot 4-bar position results
for i = 1:2,
x2 = [0 r2*cos(th2)]; % Coords of link 2
y2 = [0 r2*sin(th2)];
x3 = [r2*cos(th2) r1x+r4*cos(th4(i)) pcx(i)]; % Coords of link 3
y3 = [r2*sin(th2) r1y+r4*sin(th4(i)) pcy(i)];
x4 = [r1x r1x+r4*cos(th4(i))]; % Coords of link 4
y4 = [r1y r1y+r4*sin(th4(i))];
figure;
plot(x2,y2,'r',x4,y4,'r'); patch(x3,y3,'r');
axis('square'); set(gca,'FontSize',18);
xlabel('\itX (\itm)'); ylabel('\itY (\itm)');
axis([-0.1 0.3 -0.15 0.25]); grid;
end
44
Full-Range-Of-Motion (F.R.O.M.) Analysis

A more meaningful result from position analysis is to report the
position analysis unknowns for the entire range of mechanism
motion. The first plot gives
3
θ (red),
4
θ (green), and u (blue), all
deg, for all
D D
360 0
2
≤ ≤θ , for Term Example 1, open branch only.
The second plot gives the coupler point location for this branch,
plotting
CY
P vs.
CX
P .

0 50 100 150 200 250 300 350
20
40
60
80
100
120
140
160
180
θ
2
(deg)
A
n
g
l
e

(
d
e
g
)
Thetas 3 (red) and 4 (green), Mu (blue)

45

-0.1 0 0.1 0.2 0.3
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.25
X (m)
Y

(
m
)

4-bar Example Snapshot, Open Branch Coupler Curve
46
Trigonometric Uncertainty

3
θ solution: XY scalar equations:

2 2 4 4 1 1 3 3
2 2 4 4 1 1 3 3
s r s r s r s r
c r c r c r c r
− + =
− + =

Since
4
θ has been solved, why not calculate
3
θ using Y equation?:

e.g. ( ) 5 . 0 sin
1
3

= θ ; figure:

Problem: inverse sine function is double-valued; for each
4
θ there
are two possible solutions, only one of which is correct! Why not
calculate
3
θ using X equation? Inverse cosine has a similar problem;

e.g.
|
.
|

\
|
=

2
3
cos
1
3
θ ; figure:

Problem: inverse cosine function is double-valued; for each
4
θ there
are two possible solutions, only one of which is correct!
47
So we must use information from both sine and cosine (i.e. both X
and Y equations) - this suggests using the tangent (as we did earlier in
the
3
θ solution):

|
|
.
|

\
|
− +
− +
=

2 2 4 4 1 1
2 2 4 4 1 1
1
3
tan
c r c r c r
s r s r s r
θ

e.g.
|
.
|

\
|
=

3
1
tan
1
3
θ ; figure:

Problem: the plain atan inverse tangent function is still double-
valued!; for each
4
θ there are two possible solutions, only one of
which is correct! Solution: use the quadrant-specific inverse
tangent function atan2. Input to this function is both a numerator and
denominator; the function has built-in logic to determine the correct
quadrant for the angle answer, given the signs ± of the numerator
and denominator. The plain atan function takes a single quotient
unknown. No uncertainty with atan2:
e.g. =
|
|
.
|

\
|
+ + =
2
3
,
2
1
2 tan
3
a θ
=
|
|
.
|

\
|
− − =
2
3
,
2
1
2 tan
3
a θ

( )
2 2 4 4 1 1 2 2 4 4 1 1 3
, 2 tan c r c r c r s r s r s r A − + − + = θ
48
Now, having just cleared up this Trigonometric Uncertainty, we
already have an exception in the
4
θ tangent half-angle solution:

( ) t
1
4
tan 2

= θ

(there are two branches, one for each t value; only showing one here.)

With the 2 multiplying the inverse tangent result, it doesn't matter
whether we use atan or atan2 since the final answer will come to the
same angle. Example:

For
|
.
|

\
|
=

3
1
tan
2
1
4
θ
, from before, we don't know if the solution is

D
30
2
4
=
θ
or

D
210
2
4
=
θ

However, the multiple 2 takes care of this uncertainty:

D
60
4
= θ or

D D
60 420
4
= = θ

Now, for next time consider the following: Do the solutions for
4
θ
always exist? What if 0
2 2 2
= − + G F E ? What if
0
2 2 2
< − + G F E ? Stay tuned . . .
49
Grashof’s Law

Grashof was a German Engineer in the late 1800s. Grashof's
Law is used to determine the relative rotatability of the input and
output links in a 4-bar mechanism:
Crank - full rotation, no limits
Rocker - not full rotation, rotates back-and-forth between limits

Mechanism types (input / output links):

Identify longest, shortest, intermediate 2 links: L, S, P, Q

1) If L + S < P + Q Then we call this a Grashof Mechanism and
there are four different mechanisms and rotation conditions:

Diagrams:

a)

b)

c)
50
2) If L + S > P + Q Then we call this a Non-Grashof Mechanism
and the are four different mechanism inversions yield only one
rotation condition:

3) If L + S = P + Q Then we call this a Special Grashof
Mechanism and the four different mechanism inversions yield the
identical rotation conditions from 1) Grashof Mechanism. However,
there is the additional interesting and troublesome feature that the
mechanism may jump branches! Centerlines of links can become
collinear.

Examples

1) L = 10, S = 4, P = 8, Q = 7 – demonstrate the 4 possibilities

2) L = 10, S = 6, P = 8, Q = 7 – all Double Rockers

3) L = 10, S = 5, P = 8, Q = 7 – demonstrate branch jumping

Another interesting example: L = P = 10, S = Q = 4
parallel, locomotive linkage – subject to branch jumping unless
constrained. Also, very easy analysis:

u θ θ = =
4 2
0
3
= θ for all motion!
51
4-Bar Joint Limits

If Grashof's Law predicts the input link is a rocker, there will be
rotation limits on the input link. These joint limits occur when links
3 and 4 are aligned. As shown in the figure, there will be two joint

To calculate the joint limits, we use the law of cosines:

( )
L
r r r r r r
2 2 1
2
2
2
1
2
4 3
cos 2 θ − + = +

( )

+ − +
± =

2 1
2
4 3
2
2
2
1 1
2
2
cos
r r
r r r r
L
θ

1
r
52
Example 1: Given 7 , 8 , 6 , 10
4 3 2 1
= = = = r r r r
L + S > P + Q (10 + 6 > 8 + 7)
so we predict only double rockers from this mechanism.

( )
( )( )
| |
D
9 . 137 742 . 0 cos
6 10 2
7 8 6 10
cos
1
2 2 2
1
2
± = − ± =

+ − +
± =
− −
L
θ

Example 2: Given 7 , 8 , 4 , 10
4 3 2 1
= = = = r r r r
L + S < P + Q (10 + 4 < 8 + 7),
so we predict this mechanism is a crank-rocker. Therefore, there are
no joint limits!

( )
( )( )
| | 3625 . 1 cos
4 10 2
7 8 4 10
cos
1
2 2 2
1
2
− ± =

+ − +
± =
− −
L
θ
which is undefined!

Caution: the figure on the previous page does not apply in all joint
limit cases. For certain mechanisms, the limiting conditions occur
when links 3 and 4 fold upon each other instead of stretching straight
out. The previous method can also be used to find angular limits on
link 4 when it is a rocker; here links 2 and 3 either stretch out in a
line or fold upon each other.

Example 3: (Term Example Four-bar)
Given 7 , 8 , 3 , 18 . 11
4 3 2 1
= = = = r r r r (in) and
D
3 . 10
1
= θ , limits are:
D
1 . 120
4
=
L
θ (links 2 and 3 stretched in a line)
D
5 . 172
4
=
L
θ (links 2 and 3 folded upon each other in a line)
There are no limits on
2
θ since it is a crank.
53
Slider-Crank Mechanism Position Analysis

Converts linear motion to rotary or rotary motion to linear via
connecting rod. Internal Combustion Engine – explosion drives
piston (input), output is rotation of drive shaft. Air Compressor –
electric motor drives crank (input), piston (output) compresses air.
Two dead points where piston is at limits. Use flywheel on crank to
avoid locking. Unlike the four-bar mechanism, the four kinematic
inversions of the slider-crank mechanism yield radically different
types of motion. In class we will solve the Air Compressor case
where the crank is the input and the slider is the output.

Step 1. Draw the Kinematic Diagram:

r
2
2
– input angle
r
3
3
– coupler angle
h

– slider offset x – output displacement

Link 1 is the fixed ground link. All angles measured in right-hand
sense from horizontal to link. x is measured horizontally from the
origin to the slider/coupler revolute joint location.
54
Step 2. State the problem:

Step 3. Draw the Vector Diagram. Define all angles in positive
sense, measured from the right horizontal to the link vector (tail-to-

Vector Diagram:

Step 4. Derive the Vector-Loop-Closure Equation. Start at one
equation by starting and ending at same points, but choosing a
different path.

55
Step 5. Write XY Components for Vector-Loop-Closure Equation.
Break one vector equation into its two scalar components (X and Y):

Step 6. Solve for the Unknowns from the XY Equations. Two
coupled nonlinear equations in the two unknowns
3
,θ x . We could
isolate on unknown, square & add, and solve as in the four-bar
approach. However, notice that the two XY equations are coupled
only in
3
θ but not in x. There a simpler method - solve
3
θ using the
Y equation only and then solve x from the X equation:

What about trigonometric uncertainty? The inverse sine function is
double-valued and so there are two valid solution branches.
Graphically demonstrate the two branches.

56
Full-rotation condition

For solution to exist for entire motion range (r
2
is a crank),
absolute value of the inverse sine argument must be less than or equal
1:

1
3
2 2

r
s r h

2 2 3
s r h r − ≥

which must hold for all motion. The worst case is
D
90
2
− = θ , which
yields

2 3
r h r + ≥

This condition was derived assuming positive h; allowing negative h:

2 3
r h r + ≥ .

57
Slider-Crank Mechanism Position Analysis: Term Example 2

Given:

3
8
4
3
2
=
=
=
h
r
r
in
2
3
0.102
0.203
0.076
r
r
h
=
=
=
m

Snapshot Analysis (one input angle)

Given this mechanism and
D
30
2
= θ , calculate x and
3
θ for both
branches. Results:

Branch x (m)
3
θ
Open 0.290
D
2 . 7
Crossed -0.114
D
8 . 172

These two branch solutions are demonstrated in the figures on the
following page. We use the SI system (m).
58
-0.1 0 0.1 0.2 0.3
-0.2
-0.1
0
0.1
0.2
X (m)
Y

(
m
)

Slider-Crank Example Snapshot, Open Branch

-0.1 0 0.1 0.2 0.3
-0.2
-0.1
0
0.1
0.2
X (m)
Y

(
m
)

Slider-Crank Example Snapshot, Crossed Branch
59
Graphical Solution: The Slider-Crank position analysis may be
solved graphically, merely by drawing the mechanism and
determining the mechanism closure. This is an excellent method to
validate your computer results at a given snapshot.

• Place the grounded revolute for the crank at the origin.
• Draw the line of the slider, offset vertically from the origin by h.
• Draw the given input link 2 length at the given angle (point A).
• Draw a circle of radius r
3
, centered at point A.
• This circle intersects the slider line in general in two places.
• Connect the two branches and measure the unknown values.

Graphical Solution Figure:

Slider Limits
The crank will rotate fully if the previously-derived condition is
met. The slider reaches its maximum displacement when links 2 and
3 are aligned straight out and its maximum displacement when link 2
if folded onto link 3. We can draw two right triangles representing
these conditions and easily calculate the x limits to be
2951 . 0 0671 . 0 ≤ ≤ x , as seen in the full motion x plot, next page.
60
Full-Range-Of-Motion (F.R.O.M.) Analysis

A more meaningful result from position analysis is to report the
position analysis unknowns for the entire range of mechanism
motion. The first plot gives x (m), for all
D D
360 0
2
≤ ≤θ , for Term
Example 2, right branch only. The second plot gives
3
θ (deg), for all
D D
360 0
2
≤ ≤θ , for the right branch only.

0 50 100 150 200 250 300 350
0.05
0.1
0.15
0.2
0.25
0.3
θ
2
(deg)
x

(
m
)
Slider Displacement

61

0 50 100 150 200 250 300 350
-10
0
10
20
30
40
50
60
θ
2
(deg)
θ
3

(
d
e
g
)
Coupler Angle

62
Velocity Analysis Introduction

Velocity analysis is important for kinematic motion analysis.
Some tasks have timing, rates. Position analysis must be completed
first. Velocity analysis is also required for dynamics: position,
velocity, acceleration, dynamics, forces, machine design. Velocity
analysis is solution of coupled linear equations. Velocity is the first
time derivative of the position. Vector quantity:

Magnitude of velocity is speed; direction also crucial. Analytical
velocity analysis: write position vectors, take first time derivatives,
solve for unknowns. Units (translational and rotational):

Basic Velocity Derivation Figure:

Most general planar case: Translating and rotating rigid rod with a
slider on it. Find the total velocity of point P on the slider. Express
the position vector in Cartesian coordinates:

= + = L P P
O P

63
The angle is changing with angular velocity:

Only the planar case is this simple; the spatial rotation case is more
complicated. The length of the rod is changing with sliding velocity:

Product and Chain Rules of Differentiation
We’ll need to use the product and chain rules over and over in
velocity and acceleration analysis derivations.

Product rule:
( )
dt
dy
x y
dt
dx
xy
dt
d
+ = x, y both functions of time.

Chain rule:
( ) ( ) ( )
dt
dx
dx
df
t x f
dt
d
= f is a function of x, which is a function of t.

Example:
( ) ? cos = θ L
dt
d

64
Back to basic velocity derivation

First time derivative of position vector:

= =
dt
P d
V
P
P

65
We have just derived the Three-Part Velocity Equation:

L V V V
O P
× + + = ω

The terms for the Three-Part Velocity Equation can be
expressed in various ways, summarized below:

Vector

O
V

V

L × ω

Name

Point O
Velocity

Sliding
Velocity

Tangential
Velocity

XY
Components

Magnitude /
Direction

66
Three-Part Velocity Equation Example:
Given (instantaneously) L = 2 m,
D
30 = θ , ω = 1 rad/s, 3 V L = =

m/s
(outward), { }
3 2
T
O
V = m/s, calculate
P
V .

( )
( )
cos sin 3 3cos30 2 1 sin30
sin cos 2 3sin30 2 1 cos30
3 2.598 1 4.598
2 1.5 1.732 5.232
OX
P
OY
P
V V L
V
V V L
m
V
s
θ ω θ
θ ω θ
+ − ¦ ¹ + −
¦ ¹
= =
´  ´ 
+ + + +
¹ )
¹ )
+ −
¦ ¹ ¦ ¹
= =
´  ´ 
+ +
¹ ) ¹ )
D D
D D

or, 6.965@48.7
P
V =
D
m/s

Show magnitude and direction of each velocity component:

Vector

O
V

V

L × ω

Name

Point O
Velocity
Sliding
Velocity
Tangential
Velocity

XY
Components

Magnitude /
Direction

67
Four-Bar Mechanism Velocity Analysis

Velocity Analysis: determination of angular and linear velocities of
links in a mechanism. Required for complete motion analysis. Also
required for further analysis: acceleration, dynamics, forces, machine
design. Linear equations result from first time differentiation of
position equations. Unique solution for each mechanism branch.
Position analysis must be complete first. 1-dof mechanism, so one
velocity input must be given.

Generic Mechanism Velocity Analysis Statement: Given the
mechanism, complete position analysis, and one dof of velocity input,
calculate the velocity unknowns.

Four-bar Mechanism Velocity Analysis

Step 1. Position Analysis must first be complete.

Step 2. Draw the Velocity Diagram:

where
i
ω , i = 2,3,4, is the absolute angular velocity of link i. 0
1
= ω
since the ground link is fixed.
68
Step 3. State the problem:

Step 4. Derive the velocity equations. Take the first time derivative
of the vector loop closure equations from position analysis, in XY
component form.

Four-bar mechanism position equations:

2 3 1 4
r r r r + = +

2 2 3 3 1 1 4 4
2 2 3 3 1 1 4 4
r c r c rc r c
r s r s r s r s
+ = +
+ = +

First time derivative for velocity equations: (use chain rule several
times) Chain rule:

( )
cos
cos
sin
sin
i i
i
i
i i
i i
d d d
dt d dt
θ θ
θ
θ
θθ
θ ω
=
= −
= −

( )
sin
sin
cos
cos
i i
i
i
i i
i i
d d d
dt d dt
θ θ
θ
θ
θθ
θ ω
=
=
=

Don’t have to use product rule because 0 =
i

69
The first time derivative of the position equations is:

Gathering unknowns on the LHS:

Substituting simpler terms:

Written in matrix form:

70
Step 5. Solve the velocity equations for the unknowns
4 3
, ω ω .

Algebra solution:

71
Alternate matrix solution (yields same solution):

72
Four-Bar mechanism singularity condition:

When does the solution fail? This is a mechanism singularity,
when the determinant of the coefficient matrix goes to zero. The
result is dividing by zero, for infinite answers
4 3
, ω ω . Let’s see what
this means physically.

Physically, this happens when links 3 and 4 are straight out or folded
on top of each other (what does this correspond to?):
73
The basic four-bar mechanism velocity analysis problem is now
solved. Now that we know the angular unknowns, we can find the
translational velocity of any point on the mechanism, e.g. coupler
point C:

74
Four-bar mechanism velocity example:

Given r
1
= 0.284, r
2
= 0.076, r
3
= 0.203, r
4
= 0.178, R
C/A
= 0.127
m, and
1
10.3 θ =
D
,
2
30 θ =
D
,
3
53.8 θ =
D
,
4
121.7 θ =
D
,
3
36.9 δ =
D
. This is
the open branch of the four-bar mechanism position example (Term
Example 1).

Snapshot Analysis (one input angle)

Given this mechanism position analysis plus
2
ω π = rad/s (+, so
ccw), calculate
4 3
,ω ω , and
C
V for this instant (snapshot).

3
4
0.164 0.151 0.120
0.120 0.093 0.207
ω
ω
− −
¦ ¹ ¦ ¹
=
´  ´ 

− −
¹ ) ¹ )

3
4
1.271
0.587
ω
ω

¦ ¹ ¦ ¹
=
´  ´ 

¹ ) ¹ )

Both are negative, so cw direction. These results are the absolute
angular velocities of links 3 and 4 with respect to the ground link.

Coupler point translational velocity:
0.042
0.209
C
V
¦ ¹
=
´
¹ )
(m/s)
75
Full-Range-Of-Motion (F.R.O.M.) Analysis

A more meaningful result from velocity analysis is to report the
velocity analysis unknowns for the entire range of mechanism
motion. The plot below gives
3
ω (red) and
4
all
D D
360 0
2
≤ ≤θ , for Term Example 1, open branch only. Since
2
ω is constant, we can plot the velocity results vs.
2
θ (since it is
related to time t via t
2 2
ω θ = ).

0 50 100 150 200 250 300 350
-1.5
-1
-0.5
0
0.5
1
1.5
2
θ
2
(deg)
ω
3

a
n
d

ω
4

(
r
a
d
/
s
)
Omegas 3 (red) and 4 (green)

76
The plot below gives the translational coupler point velocity for all
D D
360 0
2
≤ ≤θ , for Term Example 1, open branch only.

0 50 100 150 200 250 300 350
-0.2
-0.1
0
0.1
0.2
0.3
θ
2
(deg)
V
C
X

a
n
d

V
C
Y

(
m
/
s
)
Coupler Point Velocities X (red) and Y (green)

77
Derivative/Integral Relationships
When one variable is the derivative of another, what are the
relationships? For example:
dt
d
3
3
θ
ω =

+ = dt
3 30 3
ω θ θ

0 50 100 150 200 250 300 350
0.4
0.6
0.8
1
1.2
θ
3

(
r
a
d
)
0 50 100 150 200 250 300 350
-1
0
1
θ
2
(deg)
ω
3

(
r
a
d
/
s
)

The value of
3
ω at any point is the slope of the
3
θ curve at that point.
The value of
3
θ at any point is the integral of the
3
ω curve up to that
point (the value of
3
θ at any point is the area under the
3
ω curve up
to that point).
This graph is vs.
2
θ , but the same type of relationships hold as
for time t since
2
ω is constant. This is the Term Example 1 result,
but we changed
3
θ from deg to rad for better comparison.
78
Slider-Crank Mechanism Velocity Analysis

Again, we will solve the Air Compressor case where the crank is
the input and the slider is the output. The Internal Combustion
Engine case (slider input/crank output) is equally interesting.

Step 1. Position Analysis must first be complete.

Step 2. Draw the Velocity Diagram:

where
i
ω , i = 2,3 is the absolute angular velocity of link i. x is the
variable slider velocity. 0
4
= ω since the slider cannot rotate.

Step 3. State the problem:

79
Step 4. Derive the velocity equations. Take the first time derivative
of the vector loop closure equations from position analysis, in XY
component form.

Slider-crank mechanism position equations:

2 3
r r x h + = +
2 2 3 3
2 2 3 3
r c r c x
r s r s h
+ =
+ =

First time derivative for velocity equations:

Gathering unknowns on the LHS:

Written in matrix form:

80
Step 5. Solve the velocity equations for the unknowns x ,
3
ω .

Actually, these equations are decoupled so we don't need a
matrix solution. First, solve
3
ω from Y equation:

Then solve x from the X equation using the
3
ω result:

The alternate matrix solution:

will yield identical results.
81
Slider-crank mechanism singularity condition:

When does the solution fail? This is a slider-crank mechanism
singularity, when the determinant of the coefficient matrix goes to
zero. The result is dividing by zero, resulting in infinite answers
x ,
3
ω .

3 3
0 A r c = =

0 = A when 0 cos
3
= θ , or "
D D
, 270 , 90
3
= θ

Physically, this happens when link 3 is straight up or down
(
3
90 θ = ±
D
). Doesn’t happen for nominal full-rotation slider-crank
mechanisms, even with offsets.

Of course r
3
cannot go to zero, otherwise we have a degenerate
slider-crank mechanism.

82
Slider-crank mechanism velocity example:

Given r
2
= 0.102, r
3
= 0.203, h = 0.076 m, and
2
30 θ =
D
,
3
7.2 θ =
D
, x = 0.290 m. This is the right branch of the slider-crank
position example (Term Example 2).

Snapshot Analysis (one input angle)

Given this mechanism position analysis plus
2
/ 2 ω π = rad/s (+,
so ccw), calculate
3
,ω x for this instant (snapshot).

3
1 0.025 0.080
0 0.202 0.138
x
ω

¦ ¹ ¦ ¹
=
´  ´

¹ ) ¹ )

3
0.062
0.686
x
ω

¦ ¹ ¦ ¹
=
´  ´

¹ ) ¹ )

Both are negative, so the slider is currently traveling to the left and
the coupler link is currently rotating in the cw direction. These
results are the absolute linear and angular velocities of links 4 and 3
with respect to the fixed ground link.
83
Full-Range-Of-Motion (F.R.O.M.) Analysis

A more meaningful result from velocity analysis is to report the
velocity analysis unknowns for the entire range of mechanism
motion. The plot below gives x (red, m/s) and
3
all
D D
360 0
2
≤ ≤θ , for Term Example 2, right branch only. Since
2
ω is constant, we can plot the velocity results vs.
2
θ (since it is
related to time t via t
2 2
ω θ = ).

0 50 100 150 200 250 300 350
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
θ
2
(deg)
x
d

(
m
/
s
)

a
n
d

ω
3

(
r
a
d
/
s
)
Xdot (red) and Omega3 (green)

84
Derivative/Integral Relationships

When one variable is the derivative of another, what are the
relationships? For example:
dt
dx
x =

+ = dt x x x
0

0 50 100 150 200 250 300 350
0.1
0.2
0.3
x

(
m
)
0 50 100 150 200 250 300 350
-0.2
0
0.2
θ
2
(deg)
x
d

(
m
/
s
)

The value of x at any point is the slope of the x curve at that point.
The value of x at any point is the integral of the x curve up to that
point (the value of x at any point is the area under the x curve up to
that point).
This graph is vs.
2
θ , but the same type of relationships hold as
for time t since
2
ω is constant. This is the Term Example 2 result.
85
Acceleration Analysis Introduction
Acceleration analysis is required for dynamics: position,
velocity, acceleration, dynamics, forces, machine design. Important
for kinematic motion analysis. Position and velocity analyses must be
completed first. Acceleration analysis is solution of linear equations.
Acceleration is the first time derivative of the velocity and second
time derivative of the position. Vector quantity:

Analytical acceleration analysis: write position vectors, take first
two time derivatives, solve for unknowns. Units (translational and
rotational):

Basic Acceleration Derivation Figure:

Rotating rigid rod with a slider on it. Find the total acceleration of
point P on the slider.
86
Recall the 2-part position and 3-part velocity results:

)

¹
¹
´
¦
+
+
= + =
θ
θ
sin
cos
L P
L P
L P P
OY
OX
O P

)

¹
¹
´
¦
+ +
− +
= × + + =
θ ω θ
θ ω θ
ω
cos sin
sin cos
L V V
L V V
L V V V
OY
OX
O P

The angle is changing with angular velocity and acceleration:

Only planar case is this simple; the spatial rotation case is more
complicated. The length of the rod is changing with sliding velocity
and acceleration:

87
Product and Chain Rules of Differentiation

Again, we’ll need to use the product and chain rules over and
over in acceleration analysis derivations.

Product rule:
( )
dt
dy
x y
dt
dx
xy
dt
d
+ = x, y both functions of time.

Chain rule:
( ) ( ) ( )
dt
dx
dx
df
t x f
dt
d
= f is a function of x, which is a function of t.

Example:

( )
2
2
cos ?
d
L
dt
θ =
88
Back to basic acceleration derivation

First time derivative of velocity vector (Second time derivative of
position vector):

= = =
2
2
dt
P d
dt
V d
A
P P
P

89
We have just derived the Five-Part Acceleration Equation:

( ) L L V A A A
O P
× × + × + × + + = ω ω α ω 2

These terms can be expressed in various ways, summarized below:

Vector

O
A

A

V × ω 2

L × α

( ) L × × ω ω

Name

Point O
Acceleration

Sliding
Acceleration

Coriolis
Acceleration

Tangential
Acceleration

Centripetal
Acceleration

XY
Components

Magnitude /
Direction

90
Five-Part Acceleration Equation Example:
Continuation of 3-part velocity example.

Given (instantaneously) L=2 m,
D
30 = θ , 1 ω = rad/s, 2 α = rad/s
2
,
3 V L = =

m/s (outward),
{ }
3 2
T
O
V = , 4 A L = =

m/s
2
(outward),
{ }
1 2
T
O
A = , calculate
P
A .

2
2
2
cos 2 sin sin cos
sin 2 cos cos sin
1 3.464 3 2 1.732 2.268
2 2 5.196 3.464 1 11.660
OX
P
OY
A A V L L
A
A A V L L
m
s
θ ω θ α θ ω θ
θ ω θ α θ ω θ
¦ ¹ + − − −
=
´ 
+ + + −
¹ )
+ − − − −
¦ ¹ ¦ ¹
= =
´  ´ 
+ + + −
¹ ) ¹ )

or, 11.879@101.0
P
A =
D
m/s
2

Show magnitude and direction of each Acceleration component:

Vector
O
A

A

V × ω 2

L × α
( ) L × × ω ω

Name Point O
Acceleration
Sliding
Acceleration
Coriolis
Acceleration
Tangential
Acceleration
Centripetal
Acceleration

XY
Components

Magnitude /
Direction

91
Four-Bar Mechanism Acceleration Analysis

Acceleration Analysis - determination of angular and linear
accelerations of links in a mechanism. Required for complete motion
analysis. Also required for further analysis: dynamics, forces,
machine design. Linear equations result from second time
differentiation of position equations. Unique solution for each
mechanism branch. Position and velocity analyses must be complete
first. 1-dof mechanism, so one acceleration input must be given.

Generic Mechanism Acceleration Analysis Statement: Given the
mechanism, complete position and velocity analyses, and one dof of
acceleration input, calculate the acceleration unknowns.

Four-bar Mechanism Acceleration Analysis

Step 1. Position and Velocity Analyses must first be complete.

Step 2. Draw the Acceleration Diagram:

where
i
α , i = 2,3,4 is the absolute angular acceleration of link i.
0
1
= α since the ground link is fixed.
92
Step 3. State the problem:

Step 4. Derive the acceleration equations. Take the first time
derivative of the four-bar mechanism velocity equations from
velocity analysis, in XY component form.

Four-bar mechanism velocity equations:

2 2 2 3 3 3 4 4 4
2 2 2 3 3 3 4 4 4
r s r s r s
r c r c r c
ω ω ω
ω ω ω
− − = −
+ =

The first time derivative of the velocity equations is:

Gathering unknowns on the LHS:

93
Substituting simpler terms:

Written in matrix form:

94
Step 5. Solve the acceleration equations for the unknowns
4 3
,α α .

Matrix solution (Algebra solution yields the same results):

95
Four-Bar mechanism singularity condition:

Same coefficient matrix A as velocity case, so singularity
condition is identical:

"
D D
, 180 , 0
3 4
= −θ θ

This condition is the same problem for position, velocity, and
acceleration. At this singularity, there is zero transmission angle u
and Link 2 is at a joint limit!

The basic four-bar mechanism acceleration analysis problem is now
solved. Now that we know the angular unknowns, we can find the
translational acceleration of any point on the mechanism, e.g.
coupler point C:

96
Four-bar mechanism acceleration example:

Given r
1
= 0.284, r
2
= 0.076, r
3
= 0.203, r
4
= 0.178, R
C/A
= 0.127
m, and
1
10.3 θ =
D
,
2
30 θ =
D
,
3
53.8 θ =
D
,
4
121.7 θ =
D
,
3
36.9 δ =
D
;
2
ω π = ,
3
1.271 ω = − ,
4
0.587 ω = − rad/s. This is the open branch of the
position and velocity example (Term Example 1).

Snapshot Analysis (one input angle)

Given this mechanism position and velocity analysis, plus
0
2
2
, calculate
4 3
,α α for this instant (snapshot).

3
4
0.164 0.151 0.877
0.120 0.093 0.589
α
α
− −
¦ ¹ ¦ ¹
=
´  ´ 

− − −
¹ ) ¹ )

3
4
0.213
6.030
α
α
¦ ¹ ¦ ¹
=
´  ´ 
¹ ) ¹ )

Both are positive, so ccw direction. These results are the absolute
angular accelerations of links 3 and 4 with respect to the ground link.

Coupler point translational acceleration:
0.676
0.582
C
A

¦ ¹
=
´

¹ )
m/s
2

97
Full-Range-Of-Motion (F.R.O.M.) Analysis

A more meaningful result from acceleration analysis is to report
the acceleration analysis unknowns for the entire range of mechanism
motion. The plot below gives
3
α (red) and
4
2
), for
all
D D
360 0
2
≤ ≤θ , for Term Example 1, open branch only. Since
2
ω is constant, we can plot the acceleration results vs.
2
θ (since it is
related to time t via t
2 2
ω θ = ).

0 50 100 150 200 250 300 350
-6
-4
-2
0
2
4
6
θ
2
(deg)
α
3

a
n
d

α
4

(
r
a
d
/
s
2
)
Alphas 3 (red) and 4 (green)

98
The plot below gives the translational coupler point acceleration for
all
D D
360 0
2
≤ ≤θ , for Term Example 1, open branch only.

0 50 100 150 200 250 300 350
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
θ
2
(deg)
A
C
X

a
n
d

A
C
Y

(
m
/
s
2
)
Coupler Point Accelerations X (red) and Y (green)

99
Derivative/Integral Relationships

When one variable is the derivative of another, recall the
relationships from calculus. For example:

dt
d
4
4
θ
ω =

+ = dt
4 40 4
ω θ θ

dt
d
4
4
ω
α =

+ = dt
4 40 4
α ω ω

0 50 100 150 200 250 300 350
2
2.5
3
θ
4

(
r
a
d
)
0 50 100 150 200 250 300 350
-1
0
1
ω
4

(
r
a
d
/
s
)
0 50 100 150 200 250 300 350
-5
0
5
α
4

(
r
a
d
/
s
2
)
θ
2
(deg)

100
Slider-Crank Mechanism Acceleration Analysis

Again, we will solve the Air Compressor case where the crank is
the input and the slider is the output.

Step 1. Position and Velocity Analyses must first be complete.

Step 2. Draw the Acceleration Diagram:

where 3 , 2 ; = i
i
α is the absolute angular acceleration of link i. 0
4
= α
since the slider cannot rotate.

Step 3. State the problem:

101
Step 4. Derive the acceleration equations. Take the first time
derivative of the velocity equations from velocity analysis, in XY
component form.

Slider-crank mechanism velocity equations:

2 2 2 3 3 3
2 2 2 3 3 3
0
r s r s x
r c r c
ω ω
ω ω
− − =
+ =

The first time derivative of the velocity equations is:

Gathering unknowns on the LHS:

Written in matrix form:

102
Step 5. Solve the acceleration equations for the unknowns x ,
3
α .

Actually, these equations are decoupled so we don't need a matrix
solution. First, solve
3
α from Y equation:

Then solve x from the X equation using the
3
α result:

103
Slider-crank mechanism singularity condition:
Same coefficient matrix as velocity case, so singularity
condition is identical (see the singularity discussion in the slider-
crank velocity section).

Slider-crank mechanism acceleration example:

Given r
2
= 0.102, r
3
= 0.203, h = 0.076 m, and
2
30 θ =
D
,
3
7.2 θ =
D
, x = 0.290 m; and
2
/ 2 ω π = ,
3
0.686 ω = − rad/s, 0.062 x = −
m/s. This is the right branch of the position and velocity example
(Term Example 2).

Snapshot Analysis (one input angle)
Given this mechanism position and velocity analysis plus,
0
2
2
, calculate
3
,α x for this instant (snapshot).

3
1 0.025 0.312
0 0.202 0.137
x
α

¦ ¹ ¦ ¹
=
´  ´

− −
¹ ) ¹ )

3
0.329
0.681
x
α

¦ ¹ ¦ ¹
=
´  ´
¹ ) ¹ )

These results are the absolute linear and angular accelerations of links
4 and 3 with respect to the fixed ground link.
104
Full-Range-Of-Motion (F.R.O.M.) Analysis

A more meaningful result from acceleration analysis is to report
the acceleration analysis unknowns for the entire range of mechanism
motion. The plot below gives x (red, m/s
2
) and
3
2
),
for all
D D
360 0
2
≤ ≤θ , for Term Example 2, right branch only. Since
2
ω is constant, we can plot the velocity results vs.
2
θ (since it is
related to time t via t
2 2
ω θ = ).

0 50 100 150 200 250 300 350
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
θ
2
(deg)
x
d
d

(
m
/
s
2
)

a
n
d

α
3

(
r
a
d
/
s
2
)
Xdotdot (red) and Alpha3 (green)

105
Derivative/Integral Relationships

When one variable is the derivative of another, recall the
relationships from calculus. For example:

dt
dx
x =

+ = dt x x x
0

dt
x d
x

=

+ = dt x x x
0

0 50 100 150 200 250 300 350
0.1
0.2
0.3
x

(
m
)
0 50 100 150 200 250 300 350
-0.2
0
0.2
x
d

(
m
/
s
)
0 50 100 150 200 250 300 350
-0.5
0
0.5
θ
2
(deg)
x
d
d

(
m
/
s
2
)

106
Input Motion Specification

Up to this point, for full range of motion we have assumed that
the input link rotates fully with a given constant input angular
velocity. Our input constraints have thus been
D D
360 0
2
≤ ≤θ , ω
2

constant, and α
2
= 0. This input motion specification is plotted like
this:

Note that we have been plotting calculated results vs. θ
2
. Since ω
2
is
constant, we have
2 2
t θ ω = , so we could just as well plot all results
vs. time t, since both θ
2
107
This constant ω
2
input specification is fine for mechanisms whose
input rotates fully and considering steady-state motion only. Many
useful mechanisms have input links that do not rotate fully but travel
between joint limits, starting and stopping at zero angular velocity.
Why is the previous page’s plots unacceptable in this case?

Simplest change – linear angular velocity starting and stopping at
zero:

We cannot plot vs. θ
2
since it is not increasing linearly – plot vs. t.

What is the weakness of this approach? (Discontinuous acceleration
function yields infinite jerk at start, middle, and finish.)
108
We can fix this with a trapezoidal input acceleration profile:

This input motion specification should be fine (trapezoidal input
torque is often used for industrial robots), but there are many
different zones to handle – what acceleration profile is similar but
with a single function?
109
Full-cycloidal function input angle specification

( ) ( )
( )
( )
( )
( )
( )
( )
2 20 2 20
2 20
2
2 20
2
2
2
2 20
2
3
1 2
sin
2
2
1 cos
2
2
sin
4
2
cos
F
F F
F
F F
F
F F
F
F F
t t
t
t t
t
t
t t
t
t
t t
t
t
t t
π
θ θ θ θ
π
θ θ
π
ω
π θ θ
π
α
π θ θ
π
β
| |
= + − −
|
\ .
− | |
= −
|
\ .
− | |
=
|
\ .
− | |
=
|
\ .

0 0.5 1 1.5 2 2.5 3
60
80
100
θ
2

(
d
e
g
)
0 0.5 1 1.5 2 2.5 3
0
2
4
ω
2

(
r
a
d
/
s
)
0 0.5 1 1.5 2 2.5 3
-1
0
1
α
2

(
r
a
d
/
s
2
)
0 0.5 1 1.5 2 2.5 3
-2
0
2
time (sec)
β
2

(
r
a
d
/
s
3
)

Example with
20
60 θ =
D
,
2
120
F
θ =
D
, and t
F
= 3 sec.
110
Dynamics Introduction

Chart:

Kinematics:

translational

rotational

Kinetics:

translational Newton’s 2
nd
Law:

rotational Euler’s equation:

Dynamics of a single rigid body in the plane
Rigid body acted on by a system of forces and moments to
produce planar motion. What is the first step in analysis? Draw . . .
111
Free Body Diagram (FBD)

Isolate each rigid body and show the forces and moments acting.
This contains all the info needed to write Newton’s 2
nd
Law and
Euler’s equation.

FBD Simplified FBD

112
Internal and External Forces and Moments

All internal and external forces and moments must be included
on the FBD.

External forces/moments:

Internal forces/moments:

Write dynamics equations

Newton’s 2
nd
Law:

Euler’s equation:

G
A is the linear acceleration of center of gravity – same direction as
R. Different points in rigid body have different linear accelerations.
α angular acceleration of rigid body. The entire rigid body
experiences the same α .
113
D’Alembert’s Principle

Turn dynamics problem into a statics problem by the inclusion
of a fictitious “inertial force”
G
A m F − =
0
and a fictitious “inertial
moment” α
G
I M − =
0
. “Centrifugal force”
2
ω mr − is an example of
an inertial force; it’s not really a force but an effect of acceleration
and inertia. Subtract RHS of equations, then sum to zero as in statics.
We won’t use this method, just wanted you to know in case you ran
into it somewhere.

0
0
G
O
R mA
R F
− =
+ =

0
0
G
O
T r R I
T r R M
α + × − =
+ × + =

114
Two Types of Dynamics Problems

Forward Dynamics:
Given the mechanism, external forces and moments, and the
applied driving force (or torque), find the resulting mechanism
motion and internal joint forces.

Inverse Dynamics:
Given the mechanism, external forces and moments, and the
desired mechanism motion, find the required driving force (or torque)
and internal joint forces.

Given
2
τ and
EXT EXT
M F , , find the motion
4 3 2
, , θ θ θ ,
4 3 2
, , ω ω ω ,
4 3 2
, , α α α and internal forces
ij
F .

Given the motion
4 3 2
, , θ θ θ ,
4 3 2
, , ω ω ω ,
4 3 2
, , α α α , and
EXT EXT
M F , , find
2
τ and internal forces
ij
F .

Next lecture: Newton's 2
nd
Law and Euler's equation require:

translational: mass center of gravity
rotational: center of gravity mass moment of inertia

m P
G
I
G
115
Mass, Center of Gravity, Mass Moment of Inertia

G
A m F =

α
G
G
I M =

Translational: mass center of gravity
Rotational: center of gravity mass moment of inertia

Mass

In Newton’s 2
nd
Law a m F =

, mass m is the proportionality
constant. Mass is measure of translational inertia – resistance to
change in motion, Newton’s 1
st
Law. Mass is also measure of storage
of translational kinetic energy
2
2
1
mv KE
T
= .
116
Examples for m, CG, I
G
:

System of particles General rigid body

Rectangular rigid body

Mass calculation:

System of particles:

General rigid body:

Rectangular rigid body:

117
Center of Gravity (CG, G)

Also called center of mass, mass center, centroid

CG calculation:

System of particles:

General rigid body:

118
Rectangular rigid body:

Using an XY coordinate frame centered at the geometric center.

/ 2
/ 2
/ 2
/ 2
/ 2
2
/ 2
2 2
2
0
2 4 4
x
x
b
b
b
b
b
b
xdm
X
dm
xdV
m
xthdx
m
th
xdx
m
th x
m
th b b
m
ρ
ρ
ρ
ρ
ρ

=
=
=
=
=
| |
= − =
|
\ .

/ 2
/ 2
/ 2
/ 2
/ 2
2
/ 2
2 2
2
0
2 4 4
y
y
h
h
h
h
h
h
ydm
Y
dm
ydV
m
ytbdy
m
tb
ydy
m
tb y
m
tb h h
m
ρ
ρ
ρ
ρ
ρ

=
=
=
=
=
| |
= − =
|
\ .

For a homogeneous, regular geometric body, the CG is the geometric
center.

119
Mass Moment of Inertia (I
G
) is not the same as Area moment of
inertia (I
G
) for beam bending:

2
Ax
y
I y dA =

2
Ay
x
I x dA =

Units:
4
A
I m ≡

Mass Moment of Inertia (I
G
)

In Euler’s equation
GZ
G
M I α =

, I is the proportionality constant. I
is measure of rotational inertia – resistance to change in motion,
Newton’s 1
st
Law. Also, it is a measure of how hard it is to
accelerate in rotation about certain axes. I is also measure of storage
of rotational kinetic energy
2
1
2
R G
KE I ω = .

Units:
2
G
I kgm ≡ .

120
Mass Moment of Inertia I
G
calculation:

System of particles:

where r
i
is the scalar perpendicular distance from the axis to the i
th

particle. With squaring, all terms will be positive, no there can be no
canceling like for CG. If first moment is balanced, second moment
will be doubled about the CG.

General rigid body:

What is the only term that matters for XY planar motion?

In the example shown above:
XXG YYG ZZG
I I I > > also
ZZG ZZ
I I >

121
Rectangular rigid body:
Using an XY coordinate frame centered at the CG.

( ) ( )
∫ ∫ ∫
− −
+ = + =
2 /
2 /
2 /
2 /
2 2 2 2
b
b
h
h
body
ZZG
tdxdy y x dm y x I ρ

|
|
.
|

\
|
|
|
.
|

\
|

− +
|
.
|

\
|

− =
|
|
|
.
|

\
|
+ =
2 /
2 /
3 3
2
2 /
2 /
2 /
2 /
3
2
8 8 3
1
2 2
3
b
b
b
b
h
h
ZZG
dx
h h h h
x t
dx
y
y x t I
ρ
ρ

|
|
|
.
|

\
|
+ =
|
|
.
|

\
|
+ =

2 /
2 /
3 3
2 /
2 /
3
2
12 3 12
b
b
b
b
ZZG
x h hx
t dx
h
hx t I ρ ρ

( )
2 2
3 3
3 3 3
12 12 12
2 2 12 8 8 3
h b
tbh bh h b
t
b b h b b h
t I
ZZG
+ =
|
|
.
|

\
|
+ =
|
|
.
|

\
|
|
.
|

\
|

− +
|
|
.
|

\
|

− =
ρ
ρ
ρ

( )
2 2
12
h b
m
I
ZZG
+ = (because tbh V m ρ ρ = = )

122

Units: mass times distance squared,
2
kgm .

Checks with result given in the textbook.

How do we find mass moments of inertia in the real-world?

• look up in tables
• CAD package such as SolidEdge
123
Parallel Axis Theorem
The mass moment of inertia through the CG is related to mass
moments of inertia of parallel axes through different points as
follows:

where d is the scalar distance separating the axis of interest from the
axis through the CG. Notice
ZZG
I is a small as it can get; any
ZZ
I
must be greater, due to the term
2
md , which is always positive.

Parallel axis theorem example:

Rectangular rigid body:

( )
( )
2 2
2 2
2 2 2 2
2 2
2 2
12 4 4
12 4 12 4
3 3
3
ZZ
m b h
I b h m
b b h h
m
b h
m
m
b h
| |
= + + +
|
\ .
| |
= + + +
|
\ .
| |
= +
|
\ .
= +

Agrees with result given in dynamics textbooks.
124

Generic Mechanism Inverse Dynamics Analysis Statement:
Given the mechanism, external forces and moments, and the
desired mechanism motion, find the required driving force (or torque)
and internal joint forces.

Single Rotating Link Inverse Dynamics Analysis

Step 1. Position, Velocity, and Acceleration Analyses must first be
complete.

Step 2. Draw the Diagrams:

Physical Dynamics Diagram:

Free Body Diagram (FBD):

125
Step 3. State the problem:

Step 4. Derive the Newton-Euler Dynamics Equations.

Newton's 2
nd
Law:

Euler's Equation:

Count # of unknowns and # of equations:

126
Step 5. Derive XYZ scalar equations from the vector equations.

Written in matrix form:

127
Step 6. Solve for the unknowns
Actually, we don’t need matrix solution; the first two equations
are decoupled and the solution is straight-forward:

Step 7. Calculate Shaking Force and Moment
After the inverse dynamics problem is solved, we can calculate
the vector shaking force and moment, which is the force/moment
reaction on the ground link due to the mechanism, motion, and

128
Terms for the inverse dynamics equations

The inverse dynamics problem has been solved analytically for
the single rotating link. Now, how do we calculate the various terms
that appear in the dynamics equations? These all must be derived
from given information.

=
)

¹
¹
´
¦
=
GY
GX
G
A
A
A

=
)

¹
¹
´
¦
=
EY
EX
E
F
F
F

=
)

¹
¹
´
¦
=
Y
X
r
r
r
12
12
12

=
)

¹
¹
´
¦
=
EY
EX
E
r
r
r

=
GZ
I

129
Single rotating link inverse dynamics example:

Given: L = 1 m, h = 0.1 m, m = 2 kg, ω = 100 rad/s, α = 0,
F
E
=150 N, 0 =
E
φ (constant relative to horizontal), M
E
=0 Nm.

Calculated terms: 5 . 0
12
= =
E
r r m
2
17 . 0 kgm I
GZ
=

2500
4330
− =
=
Gy
Gx
A
A
2
s
m

Snapshot Analysis (one input angle)
At
D
150 = θ , given this link, motion, and external force,
calculate τ , ,
12 12 Y X
F F and
S S
M F , for this instant (snapshot).

¦
)
¦

¹
¦
¹
¦
´
¦
− =
¦
)
¦

¹
¦
¹
¦
´
¦

5 . 37
4980
8510
1 433 . 0 250 . 0
0 1 0
0 0 1
12
12
τ
Y
X
F
F

¦
)
¦

¹
¦
¹
¦
´
¦
− =
¦
)
¦

¹
¦
¹
¦
´
¦
5 . 66
4980
8510
12
12
τ
Y
X
F
F
N, Nm

)

¹
¹
´
¦

= − = =
4980
8510
12 21
F F F
S
N

k M
S
ˆ
5 . 66 − = − = τ Nm

130
Full-Range-Of-Motion (F.R.O.M.) Analysis

A more meaningful result from inverse dynamics analysis is to
report the unknowns for the entire range of mechanism motion. The
plot below gives the required driving torque τ (Nm, red) for all
D D
360 0 ≤ ≤θ , assuming the given ω is constant, for the same
example from the previous page. This shows the torque that must be
supplied by an external DC servomotor to cause the specified motion.
Also plotted is the average torque (green) τ
AVG
= 0 and the root-
mean-square (RMS) torque value (blue) τ
RMS
= 106.1 Nm.

0 50 100 150 200 250 300 350
-150
-100
-50
0
50
100
150
θ
(deg)
τ

(
N
m
)
Tau (red) with average (green) and rms (blue)

131
The plots below give the Shaking Force
S
F and CG translational
acceleration results, respectively, for all
D D
360 0 ≤ ≤θ . In both plots,
the X components are red and the Y green.

0 50 100 150 200 250 300 350
-1
-0.5
0
0.5
1
x 10
4
θ
(deg)
F
S
X

a
n
d

F
S
Y

(
N
)
Shaking Force, X (red) and Y (green)

132

0 50 100 150 200 250 300 350
-5000
0
5000
θ
(deg)
A
G
X

a
n
d

A
G
Y

(
m
/
s
2
)
CG Acceleration, X (red) and Y (green)

The Shaking Moment
S
M is merely the negative of the driving
torque τ plot shown previously and hence is not shown separately. Is
133
Four-Bar Mechanism Inverse Dynamics

Generic Mechanism Inverse Dynamics Analysis Statement:
Given the mechanism, external forces and moments, and the
desired mechanism motion, find the required driving force (or torque)
and internal joint forces.

Four-Bar Mechanism Inverse Dynamics Analysis

single rotating link? Count # of unknowns and # of equations:

Step 1. Position, Velocity, and Acceleration Analyses must first be
complete.

Step 2. Draw the Diagrams:
Physical Dynamics Diagram:

134
Free Body Diagrams (FBDs):

ij
F :

ij
r :

Step 3. State the problem:

135
Step 4. Derive the Newton-Euler Dynamics Equations.

Newton's 2
nd
Law:

Euler's Equation:

Count # of unknowns and # of equations:
136
Step 5. Derive XYZ scalar equations from the vector equations.

137
Write these equations in matrix/vector form:

21
21
12 12 32 32 32
32
43
23 23 43 43 43
14
14
34 34 14 14 2
1 0 1 0 0 0 0 0 0
0 1 0 1 0 0 0 0 0
0 0 0 0 1
0 0 1 0 1 0 0 0 0
0 0 0 1 0 1 0 0 0
0 0 0 0 0
0 0 0 0 1 0 1 0 0
0 0 0 0 0 1 0 1 0
0 0 0 0 0
X
Y
Y X Y X X
Y
X
Y X Y X Y
X
Y
Y X Y X
F
F
r r r r F
F
F
r r r r F
F
F
r r r r τ
− ¦
¦

¦
¦ − −
¦

¦
¦

´

− −

− −
¹
( )
( )
( )
2 2
2 2
2 2
3 3 3
3 3 3
3 3 3 3 3 3 3
4 4 4
4 4 4
4 4 4 4 4 4 4
G X
G Y
G Z
G X E x
G Y E y
G Z E x E y E y E x E
G X E x
G Y E y
G Z E x E y E y E x E
m A
m A g
I
m A F
m A g F
I r F r F M
m A F
m A g F
I r F r F M
α
α
α
¦ ¹ ¹
¦ ¦ ¦
+
¦ ¦ ¦
¦ ¦ ¦
¦ ¦ ¦

¦ ¦ ¦
¦ ¦ ¦
+ − =
´ 
¦ ¦ ¦ ¦
− + −
¦ ¦ ¦ ¦

¦ ¦ ¦ ¦
¦ ¦ ¦ ¦
+ −
¦ ¦ ¦ ¦
¦ ¦ ¦ ¦ − + −
) ¹ )

| |{ } { } b v A =

Coefficient matrix [A] dependent on geometry (kinematics solutions).
RHS {b} dependent on inertial terms, gravity, and given external
forces and moments.
138
Step 6. Solve for the unknowns

Simultaneous matrix solution: { } | | { } b A v
1 −
=

Actually, using Gaussian elimination is more efficient and robust.

Solution to internal forces and input torque are contained in the
components of {v}.

Step 7. Calculate Shaking Force and Moment

After the basic inverse dynamics problem is solved, we can
calculate the vector shaking force and moment, which is the
force/moment reaction on the ground link due to the motion:

139
Details for the general four-bar mechanism model

The inverse dynamics problem has been derived analytically for
the four-bar mechanism. Now, how do we calculate the various
terms that appear in the dynamics equations? These all must be
derived from given information. See Fig. P11-2. Let us do link 3
terms (next page). Here is the general link 3 diagram for these
derivations:

140

=
)

¹
¹
´
¦
=
Y
X
r
r
r
23
23
23

=
)

¹
¹
´
¦
=
Y
X
r
r
r
43
43
43

=
)

¹
¹
´
¦
=
Y G
X G
G
A
A
A
3
3
3

=
)

¹
¹
´
¦
=
Y
X
P
P
P
F
F
F
3
3
3

=
)

¹
¹
´
¦
=
Y
X
P
P
P
R
R
R
3
3
3

3 E
M = given
141
Figure for Term Example 1 Inverse Dynamics Example starting on
the next page:

The coupler link 3 is a rectangle of dimensions 8” x 6” x 0.5”.
The triangle tip we have been using all along in Term Example 1 is
actually the CG; of the actual rectangular link for inverse dynamics.

-0.1 0 0.1 0.2 0.3
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
X (m)
Y

(
m
)

142
Four-bar mechanism inverse dynamics example:
This is the mechanism from Term Example 1 (open branch),
with one crucial difference: the input angular velocity was too low
for interesting dynamics, so I changed it from
2
ω π = to
2
20 ω =

Given r
1
= 0.284, r
2
= 0.076, r
3
= 0.203, r
4
= 0.178, R
G3
= 0.127
m, and
1
10.3 θ =
D
,
2
30 θ =
D
,
3
53.8 θ =
D
,
4
121.7 θ =
D
,
3
36.9 δ =
D
;
2
20 ω = ,
3
8.09 ω = − ,
4
2
0 α = ,
3
8.65 α = ,
4
244.4 α =
2
. This is the open branch of the position, velocity, and
acceleration example (Term Example 1).

All moving links are wood, with mass density 0.03 ρ = lb
m
/in
3
.
Links 2 and 4 have rectangular dimensions 0.75 by 0.50 by r
i
(in);
link 3 has rectangular dimensions 8 by 6 by 0.5 (in), as shown on the
previous page. The calculated inertia parameters are
2
0.015 m = ,
3
0.327 m = ,
4
0.036 m = kg and
6
2
7.9 10
G Z
I

= × ,
3
3
1.8 10
G Z
I

= × ,
5
4
9.5 10
G Z
I

= × kgm
2
. All external forces and moments are zero but
gravity is included.

Snapshot Analysis (one input angle)

At
2
30 θ =
D
, given this mechanism and motion, calculate the
four vector internal joint forces, the driving torque
2
τ , and the
shaking force and moment ,
S S
F M for this instant (snapshot).

143
¦
¦
¦
¦
¦
¦
)
¦
¦
¦
¦
¦
¦

¹
¦
¦
¦
¦
¦
¦
¹
¦
¦
¦
¦
¦
¦
´
¦

=
¦
¦
¦
¦
¦
¦
)
¦
¦
¦
¦
¦
¦

¹
¦
¦
¦
¦
¦
¦
¹
¦
¦
¦
¦
¦
¦
´
¦

− − −

− −

0233 . 0
095 . 0
638 . 0
015 . 0
497 . 4
955 . 8
0
034 . 0
202 . 0
0 047 . 0 076 . 0 047 . 0 076 . 0 0 0 0 0
0 1 0 1 0 0 0 0 0
0 0 1 0 1 0 0 0 0
0 0 0 122 . 0 037 . 0 002 . 0 127 . 0 0 0
0 0 0 1 0 1 0 0 0
0 0 0 0 1 0 1 0 0
1 0 0 0 0 033 . 0 019 . 0 033 . 0 019 . 0
0 0 0 0 0 1 0 1 0
0 0 0 0 0 0 1 0 1
2
14
14
43
43
32
32
21
21
τ
y
x
y
x
y
x
y
x
F
F
F
F
F
F
F
F

Solution by Gaussian elimination or: { } | | { } b A v
1 −
=

{ }
¦
¦
¦
¦
¦
¦
)
¦
¦
¦
¦
¦
¦

¹
¦
¦
¦
¦
¦
¦
¹
¦
¦
¦
¦
¦
¦
´
¦

− =
¦
¦
¦
¦
¦
¦
)
¦
¦
¦
¦
¦
¦

¹
¦
¦
¦
¦
¦
¦
¹
¦
¦
¦
¦
¦
¦
´
¦
=
43 . 0
52 . 5
60 . 3
61 . 5
96 . 2
11 . 10
99 . 5
08 . 10
20 . 6
2
14
14
43
43
32
32
21
21
τ
y
x
y
x
y
x
y
x
F
F
F
F
F
F
F
F
v N, Nm

)

¹
¹
´
¦
=
56 . 4
80 . 9
S
F N k M
S
ˆ
68 . 1 − = Nm
144
Full-Range-Of-Motion (F.R.O.M.) Analysis

A more meaningful result from inverse dynamics analysis is to
report the unknowns for the entire range of mechanism motion. The
plot below gives the required driving torque
2
τ (Nm) for all
D D
360 0
2
≤ ≤θ , for the Term Example 1 mechanism, assuming the
given 20
2
= ω rad/s is constant (Remember: this has been changed
from the π ω =
2
rad/s in the kinematics examples!). This plot shows
the torque (red) that must be supplied in all configurations by an
external DC servomotor to cause the specified motion. Also plotted
is the average torque (green) τ
2AVG
= 0 and the root-mean-square
torque value (blue) τ
2RMS
= 0.36 Nm.

0 50 100 150 200 250 300 350
-0.4
-0.2
0
0.2
0.4
0.6
0.8
θ
2
(deg)
τ
2

(
N
m
)
Tau2 (red) with average (green) and RMS (blue)

145
The plots below give the shaking force
S
F (N) and shaking
moment
S
M (Nm) results, respectively, for all
D D
360 0
2
≤ ≤θ . In
the force plot, the X component is red and the Y green.

0 50 100 150 200 250 300 350
-10
-5
0
5
10
θ
2
(deg)
F
S
X

a
n
d

F
S
Y

(
N
)
Shaking Force, X (red) and Y (green)
146

In the shaking moment plot, there is only the Z component:

0 50 100 150 200 250 300 350
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
θ
2
(deg)
M
S

(
N
m
)
Shaking Moment

147
Slider-Crank Mechanism Inverse Dynamics

This problem is very similar to the four-bar mechanism inverse
dynamics problem. In fact, links 2 and 3 are handled identically!

Step 1. Position, Velocity, and Acceleration Analyses must first be
complete.

Step 2. Draw the Diagrams:
Physical Dynamics Diagram:

Free Body Diagrams (FBDs):

ij
ij
r : moment arm pointing to link i from the CG of link j
148
Step 3. State the problem:

Step 4. Derive the Newton-Euler Dynamics Equations.
Again, links 2 and 3 are identical so let us focus on link 4, the
slider. There are two kinematic constraints on the slider:

Newton's 2
nd
Law:

Euler's Equation:

Count # of unknowns and # of equations: We need an additional
equation:
149
Step 5. Derive XYZ scalar equations from the vector equations and
beam these equations into matrix/vector form. Substitute the friction
constraint to eliminate one unknown (F
14X
); also eliminate one
equation (
4
4 4
G
G Z
M I α =

).

( )
2 2 21
2 2 21
2 32 12 12 32 32
32
43
43 23 23 43 43
14
2
1 0 1 0 0 0 0 0
0 1 0 1 0 0 0 0
0 0 0 1
0 0 1 0 1 0 0 0
0 0 0 1 0 1 0 0
0 0 0 0
0 0 0 0 1 0 0
0 0 0 0 0 1 1 0
G X X
G Y Y
G Z X Y X Y X
Y
X
Y Y X Y X
Y
m A F
m A g F
I F r r r r
F
F
F r r r r
F u
τ
− ¦ ¹
¦ ¦
+ −
¦ ¦
¦ ¦ − −
¦ ¦

¦ ¦

=
´

¦ ¦

¦ ¦
− −

¦ ¦

− ±
¦ ¦

¦ ¦

¹ )
( )
2
3 3 3
3 3 3
3 3 3 3 3 3 3
4 4 4
4 4
G X E x
G Y E y
G Z E x E y E y E x E
G X E x
E y
m A F
m A g F
I r F r F M
m A F
m g F
α
α
¦ ¹
¦ ¦
¦ ¦
¦ ¦
¦ ¦

¦ ¦
´ 
+ −
¦ ¦
¦ ¦
− + −
¦ ¦

¦ ¦
¦ ¦

¹ )

| |{ } { } b v A =

Coefficient matrix [A] dependent on geometry (kinematics solutions).
Always choose proper sign of u to be opposite to the current x
direction. RHS {b} dependent on inertial and statics terms.

Step 6. Solve for the unknowns
Simultaneous matrix solution:

{ } | | { } b A v
1 −
=

Actually, using Gaussian elimination is more efficient and robust.
Solution to internal forces and input torque contained in the
components of {v}.

150
Step 7. Calculate Shaking Force and Moment
After the basic inverse dynamics problem is solved, we can
calculate the vector shaking force and moment, which is the
force/moment reaction on the ground link due to the motion.

Figure for example starting on the next page: The slider-crank
mechanism is shown at the starting (and ending) position, with zero
input angle θ
2
.
-0.1 0 0.1 0.2 0.3
-0.25
-0.2
-0.15
-0.1
-0.05
0
0.05
0.1
0.15
0.2
X (m)
Y

(
m
)

151
Slider-crank mechanism inverse dynamics example:

This is the mechanism from Term Example 2 (right branch
only), in this case keeping the low input angular velocity
2
2 ω π =
rad/s so the previous snapshots and full-range-of-motion results still
apply.

Given r
2
= 0.102, r
3
= 0.203, h = 0.076 m, and
2
30 θ =
D
,
3
7.2 θ =
D
, x = 0.290 m; and
2
/ 2 ω π = ,
3
0.686 ω = − rad/s, 0.062 x = −
m/s;
2
0 α = ,
3
2
, 0.329 x = − m/s
2
. This is the right
branch of the position, velocity, and acceleration example (Term
Example 2).

All moving links are wood, with mass density 03 . 0 = ρ (lb
m
/in
3
).
Links 2 and 3 have rectangular dimensions 0.75 by 0.50 by r
i
(in);
link 4 has rectangular dimensions 0.75 by 0.50 by 3 (in). The
calculated inertia parameters are m
2
= 0.020, m
3
= 0.041, m
4
= 0.015,
(kg) and I
G2Z
= 1.819e-005, I
G3Z
= 1.418e-004 (kgm
2
). There is a
constant external force of 1 N acting at the center of the piston,
directed horizontally to the left; gravity is included but all other
external forces and moments are zero. We assume u = 0.2
(coefficient of friction between piston and wall);

Snapshot Analysis (one input angle)

At
2
30 θ =
D
, given this mechanism and motion, calculate the
four vector internal joint forces, the driving torque
2
τ , and the
shaking force and moment ,
S S
F M for this instant (snapshot).

152
21
21
32
32
43
43
14
2
1 0 1 0 0 0 0 0 -0.002
0 1 0 1 0 0 0 0 0.1
-0.025 0.044 -0.025 0.044 0 0 0 1
0 0 1 0 1 0 0 0
0 0 0 1 0 1 0 0
0 0 -0.013 0.101 -0.013 0.101 0 0
0 0 0 0 1 0 0.2 0
0 0 0 0 0 1 1 0
X
Y
X
Y
X
Y
Y
F
F
F
F
F
F
F
τ
− ¦ ¹
¦ ¦

¦ ¦
¦ ¦
¦ ¦

¦ ¦

=
´

¦ ¦

¦ ¦

¦ ¦

¦ ¦

¦ ¦

¹ )
99
0
-0.011
0.398
0.0001
0.995
0.150
¦ ¹
¦ ¦
¦ ¦
¦ ¦
¦ ¦
¦ ¦
´ 
¦ ¦
¦ ¦
¦ ¦
¦ ¦
¦ ¦
¹ )

Solution by Gaussian elimination or: { } | | { } b A v
1 −
=

{ }
21
21
32
32
43
43
14
2
-0.935
-0.517
-0.938
-0.318
-0.949
0.081
0.231
-0.011
X
Y
X
Y
X
Y
Y
F
F
F
F
v
F
F
F
τ
¦ ¹ ¦ ¹
¦ ¦ ¦ ¦
¦ ¦ ¦ ¦
¦ ¦ ¦ ¦
¦ ¦ ¦ ¦
¦ ¦ ¦ ¦
= =
´  ´ 
¦ ¦ ¦ ¦
¦ ¦ ¦ ¦
¦ ¦ ¦ ¦
¦ ¦ ¦ ¦
¦ ¦ ¦ ¦
¹ ) ¹ )
(N, Nm)

0.982
0.748
S
F
− ¦ ¹
=
´

¹ )
(N)
ˆ
0.053
S
M k = − (Nm)
153
Full-Range-Of-Motion (F.R.O.M.) Analysis

A more meaningful result from inverse dynamics analysis is to
report the unknowns for the entire range of mechanism motion. The
plot below gives the required driving torque
2
τ (Nm) for all
D D
360 0
2
≤ ≤θ , for the Term Example 2 slider-crank mechanism,
right branch only, assuming the given
2
2 ω π = rad/s is constant.
This plot shows the torque (red) that must be supplied in all
configurations by an external DC servomotor to cause the specified
motion. Also plotted is the average torque (green) τ
2AVG
= -0.002 and
the root-mean-square torque value (blue) τ
2RMS
= 0.086 Nm.

0 50 100 150 200 250 300 350
-0.1
-0.05
0
0.05
0.1
0.15
0.2
θ
2
(deg)
τ
2

(
N
m
)
Actuator Torque

154
The plots below give the shaking force
S
F (N) and shaking
moment
S
M (Nm) results, respectively, for all
D D
360 0
2
≤ ≤θ . In
the force plot, the X component is red and the Y green.

0 50 100 150 200 250 300 350
-1.05
-1
-0.95
-0.9
-0.85
-0.8
-0.75
-0.7
θ
2
(deg)
F
S
X

a
n
d

F
S
Y

(
N
)
Shaking Force, X (red) and Y (green)

155
In the shaking moment plot, there is only the Z component:

0 50 100 150 200 250 300 350
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
θ
2
(deg)
M
S

(
N
m
)
Shaking Moment

156
Cam Introduction
Chapter 8
Applications

Compared to linkages, easier to design desired motion with cams, but
much more expensive and difficult to produce.

Cam Classification: Disk cams with followers

157
Degrees of Freedom Recall a cam joint has two-dof; allows both
rolling and sliding.

Function Generation
The output parameter is a continuous function of the input
parameter. With linkages, we can only satisfy a function exactly at a
finite number of points: 3, 4, or 5, usually. For example, a 4-bar
( )
2 4
θ θ f =

With a cam and follower mechanism, we can satisfy function
generation at infinite points.

( ) θ f S = ( ) θ φ f =

Cam input angle is θ , output is S for reciprocating (translating) and
φ (rotating) for oscillating follower.
158
Cam Motion Profiles

Up to this point, we have been mostly concerned with
mechanism analysis: given a mechanism design and its input
parameters, determine the position, velocity, acceleration, and
dynamics behavior. With cams we must consider mechanism
synthesis for the first time: given the motion requirements (follower
motion and timing with input angle), design the cam. The first step is
to determine a “smooth” cam follower motion profile. Classification:

When the motion transitions between different motion functions, we
must ensure “smooth” motion.

159
Fundamental Law of Cam Design:

Which means:

If the Fundamental law of Cam Design is satisfied, the resulting
dynamic performance will be acceptable for high-speed cam/follower
operation. If not, there will be performance degradation due to
noise, vibrations, high wear, etc. Cyclical impulse hammering when
acceleration is not continuous.
160
S V A J Diagrams

In synthesis, we are only given total motion range and perhaps
some timing requirements. It is the engineer’s job to determine the
position curves and to match the velocity and acceleration across
junctions. Position is automatically matched by shifting axes. Draw
S V A J diagrams vs. time to graphically see if the Fundamental Law
of Cam Design is satisfied for candidate curves. We can plot vs. time
or vs. input cam angle θ (assuming constant angular velocity,
t ω θ = ).

Check out Examples 8-1 (terrible)
8-3 (acceptable)

Slope of a function is the value of its derivative at a point. Therefore,
for continuous velocity and acceleration curves, the slopes of the
position and velocity curves must match across all junctions. The
slope of the acceleration can be discontinuous (leading to finite jumps
in jerk), but the acceleration itself must be continuous.

161
Generic Cam Follower Motion Profile figure:

Define each separate function so the value is zero at the initial angle,
which is zero. Then to put the whole thing together, just shift the θ
and S axes.

Match S: easy, just do it - shift S axes.

Match V: slope of S must match across junctions.

Match A: slope of V must match across junctions.

162
Cam Follower Motion Profile Examples Example 1
rise – dwell portion. Specify Parabolic (constant acceleration) to
Straight Line (constant velocity) rise, followed by a dwell.
S: ( )
2
1 0 1 1
2
1
θ θ A f = ( )
2 0 2 2
θ θ V f = ( ) 0
3 3
= θ f

V:

A:

J:

Match S at junction B: Just shift axis up.

Match V at junction B:

Try to match A at junction B:

Plot on next page.
163
Example 1 Plots

164
Cam Follower Motion Profile Examples Example 2

Fix rise portion only. Specify Half-Cycloidal function
(sinusoidal in cam angle) to Straight Line (constant velocity) rise.

S: ( )
|
|
.
|

\
|
− =
1
1
1
1
1 1 1
sin
1
β
πθ
π β
θ
θ L f ( )
2 0 2 2
θ θ V f =

V:

A:

J:

Match S at junction B: Just shift axis up.

Match V at junction B:

Match A at junction B:

Plot on next page.

165

0 20 40 60 80
0
1
2
S

(
m
)
Half Cycloid
0 20 40 60 80
0
0.05
V

(
m
/
d
e
g
)
0 20 40 60 80
0
5
x 10
-4
A

(
m
/
d
e
g
2
)
0 20 40 60 80
-2
0
2
x 10
-5
θ (deg)
J

(
m
/
d
e
g
3
)

166
Cam Follower Motion Profile Examples Example 3

Specify Full-Cycloidal function (sinusoidal in cam angle). This
will rise all the way to meet a dwell smoothly; it satisfies the
Fundamental Law of Cam Design.

S:
( )
1 1
1 1 1
1 1
1 2
sin
2
f L
θ πθ
θ
β π β
| |
= −
|
\ .

( )
2 2
0 f θ =

V:
( )
1 1
1 1
1 1
2
1 cos
L
v
πθ
θ
β β
| |
= −
|
\ .

( )
2 2
0 v θ =

A:
( )
1 1
1 1
2
1 1
2 2
sin
L
a
π πθ
θ
β β
| |
=
|
\ .

( )
2 2
0 a θ =

J:
( )
2
1 1
1 1
3
1 1
4 2
cos
L
j
π πθ
θ
β β
| |
=
|
\ .

( )
2 2
0 j θ =

Plot on next page.
167
0 20 40 60 80
0
1
2
S

(
m
)
Full Cycloid
0 20 40 60 80
0
0.05
V

(
m
/
d
e
g
)
0 20 40 60 80
-2
0
2
x 10
-3
A

(
m
/
d
e
g
2
)
0 20 40 60 80
-2
0
2
x 10
-4
θ (deg)
J

(
m
/
d
e
g
3
)
168
Analytical Cam Synthesis

Disk Cam with Radial Flat-Faced Follower

Assume a valid cam motion profile has been designed according
to the Fundamental Law of Cam Design; i.e. we now have continuous
S, V, A curves. Given the motion profile, determine the cam contour.

Is it as simple as polar plotting of ( ) θ f S = vs. cam angle θ ?

We will use kinematic inversion to simplify the synthesis.

DCRFFF Figure:

169
As seen in the figure, the radius R out to the flat-faced follower (not
to the point of contact (x,y)) is:

where C is the minimum cam radius, a design variable, and ( ) θ f S =
is the given motion profile. The radius R and the flat-face length L
can be related to the contact point x,y and the cam angle through
geometry:

Notice that:

To calculate the follower flat-face length, double the maximum of L
from above. Doubled because by symmetry the contact point will
change to the other side at
D
180 = θ .
170
To summarize thus far:

This is sufficient to manufacture the cam; it is machined with θ, R, L
coordinates. If we want to know the cam contour in Cartesian
coordinates, we must solve the relationships for x, y. In matrix form:

The coefficient matrix [A] is orthonormal, which means | | | |
T
A A =
−1
.
The solution is:

171
Minimum Cam radius to Avoid Cusps
A cusp is a point in the cam, or actually undercut; this is to be
avoided for good motion. The condition is that for a finite θ ∆ , there
is no change in x, y:
0 = =
θ θ d
dy
d
dx
will cause a cusp.

( ) ( )
( ) ( ) θ
θ
θ
θ
θ
θ
θ θ
θ
θ
θ
θ
θ
θ
θ
θ θ
θ
cos sin sin cos
sin cos cos sin
2
2
2
2
d
f d
d
df
d
df
f C
d
dy
d
f d
d
df
d
df
f C
d
dx
+ − + + =
− − + + − =

( )
( ) θ
θ
θ
θ
θ
θ
θ
θ
cos
sin
2
2
2
2
|
|
.
|

\
|
+ + =
|
|
.
|

\
|
+ + − =
d
f d
f C
d
dy
d
f d
f C
d
dx

0 = =
θ θ d
dy
d
dx
simultaneously only when:
( ) 0
2
2
= + +
θ
θ
d
f d
f C

Therefore, to avoid cusps on the entire cam contour,
( ) 0
2
2
> + +
θ
θ
d
f d
f C

Note C is always positive and ( ) θ f starts and ends at zero and never
goes negative.
172
Disk Cam with Radial Flat-Faced Follower Design Example

Specify a full-cycloidal rise (total lift 50 mm), followed by a
high dwell, a full-cycloidal return (total fall 50 mm), and then a low
dwell. Each of these four motion steps occurs for 90 deg of cam
shaft rotation.

The cam motion profile associated with this specification is
shown below. Clearly, this satisfies the Fundamental Law of Cam
Design because the position, velocity, and acceleration curves are
continuous. The jerk is not continuous, but it remains finite over all
cam angles.

0 50 100 150 200 250 300 350
0
50
S
0 50 100 150 200 250 300 350
-50
0
50
V
0 50 100 150 200 250 300 350
-100
0
100
A
0 50 100 150 200 250 300 350
-500
0
500
θ
J

173

Choosing a minimum cam radius of C = 100 mm), the resulting
cam contour is shown below.

-150 -100 -50 0 50 100
-100
-50
0
50
100
150
X (mm)
Y

(
m
m
)

174
Let us check the cusp avoidance plot. To avoid cusps in this cam,
we require that:

( ) ( ) ( ) 0
2
2
> + + = + +
θ
θ θ θ
d
f d
f C A S C

As seen in the plot below, this inequality is satisfied for the entire
range of motion, so this cam design is acceptable.

0 50 100 150 200 250 300 350
0
50
100
150
200
250
θ
C
+
S
+
A

175
Gear Introduction

Transfer motion between rotating shafts in machinery, vehicles,
toys, etc. Gears used in electromechanical systems. Change in
angular velocity, torque, direction. Can openers to Aircraft carriers.
Related mechanisms - belt and chain drives.

Applications:

Gear Classification

External Spur Gears Internal Spur Gears
176

Rack & Pinion Bevel gears

Helical (Parallel Shaft) Helical (Crossed Shaft)

Herringbone Gears Gear Train

177

Automotive Differential Planetary Gear Train

Worm and Gear Harmonic Gearing

178
Harmonic Gearing

Taken from: http://www.roymech.co.uk/

“The harmonic gear allows high reduction ratios with concentric
shafts and with very low backlash and vibration. It is based on a very
simple construction utilising metals elasto-mechanical property.”

“Harmonic drive transmissions are noted for their ability to
reduce backlash in a motion control system. How they work is
through the use of a thin-walled flexible cup with external splines on
it lip, placed inside a circular thick-walled rigid ring machined with
internal splines. The external flexible spline has two fewer teeth
than the internal circular spline. An elliptical cam enclosed in an
antifriction ball bearing assembly is mounted inside the flexible cup
and forces the flexible cup splines to push deeply into the rigid ring at
two opposite points while rotating. The two contact points rotate at a
speed governed be the difference in the number of teeth on the two
splines This method basically preloads the teeth, which reduces
backlash.”

179
Gear Ratio

Common electric motors have high speed but low torque. A
robot joint needs lower rotation speed but high torque. A gear train
can accomplish both objectives – reduce speed and increase torque.
Gear ratio is a measure of the degree of reduction and increase.

Simple spur gear pair:

DOF:

A gear joint is like a cam joint; two-dof, teeth in contact allow rolling
and sliding.

Gear 1 is input, gear 2 is output. Like two cylinders rolling without
sliding. Arc lengths are equal.

180
Define gear ratio n:

Radii inversely proportional to angular motion. For standard spur
gears, the radii are directly proportional to the number of teeth:

For relating angular velocities, tangential velocities are equal.

Most gear applications have constant angular velocities, for
accelerating up to (or down from) constant angular velocities:

181
For relating shaft torques, interface forces are equal.

Radii directly proportional to shaft torques

Summary:

The ratio of the number of teeth is directly proportional to the radii,
diameter, and shaft torques. The ratio of the number of teeth is
inverse proportional to the shaft angles, angular velocities, and
angular accelerations.
182
If 1 > n :
1 2
ω ω < Output has reduced speed

1 2
τ τ > Output has increased torque

This is the electric motor / robot joint case.

If 1 < n :
1 2
ω ω > Output has increased speed

1 2
τ τ < Output has reduced torque

If 1 = n :
1 2
ω ω = Output speed and torque unchanged

1 2
τ τ = direction reverses (external spur gears)

183
Gear ratio examples 1. Gear toy

2. Mountain Bike Transmission

Gear Ratios:
F
R
R
F
F
R
IN
OUT
N
N
N
N
n
τ
τ
ω
ω
= = = =

Schwinn Front
Sierra 48 38 28
14 0.29 0.37 0.50
16 0.33 0.42 0.57
Rear 18 0.38 0.47 0.64
22 0.46 0.58 0.78
26 0.54 0.68 0.93
30 0.62 0.79 1.07

Unlike electric motor example, mountain bike gearing generally:
• increases angular velocity
• decreases torque
184

Cannondale
Front
M400 44 32 22
11 0.25 0.34 0.50
12 0.27 0.38 0.54
Rear 14 0.32 0.44 0.64
16 0.36 0.50 0.73
18 0.41 0.56 0.82
21 0.48 0.66 0.95
24 0.54 0.75 1.09
28 0.64 0.88 1.27
32 0.73 1.00 1.45

BikeE
Front 34 Rear hub
recumbent 1.2913:1 1:1 0.7:1
11 0.25 0.32 0.46
13 0.30 0.38 0.55
Rear 15 0.34 0.44 0.63
18 0.41 0.53 0.76
21 0.48 0.62 0.88
24 0.55 0.70 1.01
28 0.64 0.82 1.18

But considering the difference in wheel sizes (26” Cannondale, 20”
BikeE), the effective BikeE high and low gear ratios are:

Original front was 46 teeth – changed for more granny gear.

185
Gear Trains and Gear Standardization

Simple Gear Trains

Mesh any number of spur gears. Leftmost is driving gear.
Rightmost is the output gear. All intermediate gears are first the
driven gear and then the driving gear as we proceed from left to right.
Let us calculate the overall gear ratio.

OUT
IN
GT
n
ω
ω
= Example:

We can find the overall gear ratio by canceling neighboring
intermediate angular velocities:

Each term in the above product may be replaced by its known
number of teeth ratio:

186
All intermediate ratios cancel, so:

We could have done the same with pitch radii instead of number of
teeth because they are in direct proportion:

So, the intermediate gears are idlers. Number of teeth effect cancels
out, but do change direction! We should have included sign:

So, for external spur gear trains:
Odd # of gears: Output same direction as input
Even # of gears: Output opposite direction as input

187
Different case:
Mesh any number of spur gears, where the driving and driven
gears are distinct, because each pair is rigidly attached to the same
shaft. See the figure. Again, let us calculate the overall gear ratio.

OUT
IN
GT
n
ω
ω
= Example:

Again, we use the equation:

But now the gears rigidly attached to the same shaft have the same
angular velocity ratio, so:

188
General formula:

Again, we must consider direction:

So, for external spur gear trains:
Odd # of pairs: Output opposite direction as input
Even # of pairs: Output same direction as input

189
Involute Spur Gear Details and Standardization

Rolling Cylinders
Mating spur gears are based on two pitch circles rolling without
slip. These are fictitious circles; you cannot look on a gear to see
them. The actual gear teeth both roll and slide (two-dof joint).

Fundamental Law of Gearing:

From our study of linkage velocity, we know this is no easy feat.
Velocity ratios in a linkage vary wildly over the range of motion.

Velocity Ratio

The author’s velocity ratio is the inverse of our gear ratio
definition and his torque ratio is the same as our gear ratio.

190
Involute Function

Standard spur gears have an involute tooth shape. If the gears’
center distance is not perfect (tolerances, thermal expansion, wear,
etc.), the angular velocity ratio will still be constant to satisfy the
Fundamental Law of Gearing. The involute is a curve generated by
unwrapping a taut sting from a circle:

191

Base Circle: Involute starts from this circle

Pitch Circle: Fictitious circle, pure rolling in contact

Pitch Point: Contact point between the two pitch circles

Pressure Angle: Angle between the common normal (also
called axis of transmission) of the two meshing teeth and the velocity
of the pitch point (tangent to both pitch circles). Point of contact
slides along this line. Similar angle is defined for cams and
followers.

Base circle, pitch circle, pressure angle relationship:

192

Length of contact along axis of transmission. Beginning of
contact is when tip of driven gear tooth intersects the axis of
transmission. End of contact is when tip of driving gear (pinion) tooth
intersects the axis of transmission. Only one or two teeth are in
contact at any one time. For harmonic gearing, many teeth are in
contact at any one time (higher gear ratio in a smaller package).

193

Increasing center distance increases the pressure angle, increases
the pitch circle radii, but doesn’t change the base circles (of course).
Thanks to the involute tooth shape, does not affect angular velocity
ratio.

How is this possible? Relationship from last page:

194
Backlash

Clearance. Distance between mating teeth measured along the
pitch circle circumference. All real gears must have some backlash
due to tolerances, thermal expansion, wear, etc. However, must
minimize backlash for smooth operation. Example: robot joints
which must be driven both directions. Changing direction, nothing
happens until the backlash is moved, and then impact - bad for
dynamics. Non-linear effect in robots. On earth gravity tends to load
the backlash for predictable effects. In space however, the backlash
is less predictable! Figure:

195
Gear Standardization

To allow interchangeability in manufacturing and to allow
meshing of different size gears (radii and number of teeth) to achieve
desired gear ratios. For two spur gears to mesh, must have: 1) the
same pressure angle; 2) same diametral pitch; and 3) be made with
standard tooth proportions (Table 9-1, p. 441).

Diametral Pitch:

Module:

Module is the metric version of diametral pitch. Not
interchangeable with US gears because different tooth proportion
standards!

Circular Pitch:

196

Standard involute tooth proportions, see Table 9-1, p. 441.
Addendum is radial distance from pitch circle to Top Land of tooth.
Dedendum is radial distance from pitch circle to Bottom Land of
tooth (not to base circle). Clearance is radial distance from Bottom
Land to mating gear Top Land (kinda like radial backlash). Face
width is thickness of tooth and gear (mating widths needn’t be the
same). Tooth thickness is the circumferential length of each tooth..
Related to the circular pitch and backlash by:

2

INTRODUCTION ............................................................................................ 3 MATRIX-VECTOR INTRODUCTION ........................................................ 11 MATLAB INTRODUCTION ........................................................................ 25 MOBILITY..................................................................................................... 29 FOUR-BAR MECHANISM POSITION ANALYSIS................................... 34 GRASHOF’S LAW........................................................................................ 49 SLIDER-CRANK MECHANISM POSITION ANALYSIS.......................... 53 VELOCITY ANALYSIS INTRODUCTION ................................................ 62 FOUR-BAR MECHANISM VELOCITY ANALYSIS ................................. 67 SLIDER-CRANK MECHANISM VELOCITY ANALYSIS ........................ 78 ACCELERATION ANALYSIS INTRODUCTION...................................... 85 FOUR-BAR MECHANISM ACCELERATION ANALYSIS ...................... 91 SLIDER-CRANK MECHANISM ACCELERATION ANALYSIS ........... 100 INPUT MOTION SPECIFICATION ........................................................... 106 DYNAMICS INTRODUCTION.................................................................. 110 MASS, CENTER OF GRAVITY, MASS MOMENT OF INERTIA .......... 115 SINGLE ROTATING LINK INVERSE DYNAMICS ................................ 124 FOUR-BAR MECHANISM INVERSE DYNAMICS................................. 133 SLIDER-CRANK MECHANISM INVERSE DYNAMICS........................ 147 CAM INTRODUCTION .............................................................................. 156 ANALYTICAL CAM SYNTHESIS ............................................................ 168 GEAR INTRODUCTION ............................................................................ 175 GEAR TRAINS AND GEAR STANDARDIZATION................................ 185

3

ME 301 Kinematics & Dynamics of Machines Introduction Kinematics:

Kinema - Greek for motion Dynamics: Rigid Body Mechanics Diagram:

Required Math: Geometry, trigonometry, vectors, matrices, calculus Mechanisms: linkages, cams, gears, gear trains Analysis vs. Synthesis • Analysis – determination of position, velocity, acceleration, etc. for a given mechanism • Synthesis – design of mechanism to do a specific job

4 Mobility .3 independent translations and rotations . no motion • Mechanism – 1 dof device with rigid links connected with joints • Machine – collection of mechanisms to transmit force (input / output) • Robot – an electromechanical device having greater than 1 dof.number of degrees-of-freedom (dof): • Structure – static. Motion – Translation and Rotation Planar – all motion is 2D (projected onto a common plane) Helical .screw Spherical .3D motion.rotation about fixed axis and translation along axis . all points in a body moves about a fixed point Spatial . programmable for a variety of tasks.

turning pair Prismatic – sliding pair Higher – point or line contact ball bearing gears cam and follower Link – rigid body Kinematic chain – number of links connected by joints open – serial robot closed – mechanism. parallel robot Kinematic Inversion – change which link is fixed – same relative motion. Examples – in class. different absolute motion.5 Joints – Pairing elements Lower – surface contact Revolute – pin joint. also see following Atlas .

and Robots Dr.6 A Brief Atlas of Structures. Mechanisms. Bob Statically Determinate Structure Statically Indeterminate Structure 4-Bar Mechanism Offset Slider-Crank Mechanism Inverted Slider-Crank Mechanism Scotch-Yoke Mechanism .

7 Geared 5-Bar Mechanism Stephenson I 6-Bar Mechanism Stephenson II 6-Bar Mechanism Stephenson III 6-Bar Mechanism Watt I 6-Bar Mechanism Watt II 6-Bar Mechanism .

8 Spur Gear Mechanism Cam-and-Follower Mechanism Geneva Wheel Mechanism Planar 3-dof Robot Adept 4-dof SCARA Robot Mitsubishi 5-dof Robot .

9 PUMA 6-dof Robot NASA 8-dof ARMII 2-dof 5-Bar Parallel Robot 3-dof 3-RRR Parallel Robot 3-dof 3-RPR Parallel Robot 3-dof Carpal Wrist .

. Before one can design a machine. the worst-case force and moment loading condition(s) must be known. Engineering design is iterative by nature: each candidate design must be thoroughly analyzed to determine its performance relative to the design specifications and relative to other design candidates. The goal of ME 301 is to give the student general skills in general matrix/vector-based kinematics and dynamics analysis which may be applied in later classes and later careers. However.10 Connection to Machine Design In ME 301 we focus on kinematics & dynamics analysis. This requires dynamics analysis. the skills gained in this course support general (electro)mechanical design. Without the computer. joints. This requires kinematics analysis. the required motion must be satisfied. both translational and rotational). and actuators (motors) in a machine. gear box. it is difficult to determine the worst-case loading cases. not synthesis (design). velocity. bearings. Before one can size the links. This kinematics & dynamics analysis is facilitated using a computer. and acceleration. for statics and dynamics. and over-designed factors of safety may be inefficiently applied. All design candidates must be analyzed regarding the motion each would provide (position.

Cartesian representation: Polar representation: Magnitude at angle: P @ θ (atan2 .quadrant-specific inverse tangent function) . acceleration. arrow normal to the plane to represent angular velocity.11 Matrix-Vector Introduction Vectors Arrow in the plane with magnitude and direction. Used to represent position. force. velocity. Also. and torque (moment) vectors (see later in notes). angular acceleration.

. Scalar result.13 Vector Dot Product Dot product is projection of one vector onto another.

. Vector result. right thumb is direction of P1 × P 2 . rotate into second vector P 2 .14 Vector Cross Product Cross product of two vectors gives a third vector mutually perpendicular to the original two vectors. Direction via right-hand-rule: Put right hand fingers along first vector P1 .

Example: ˆ ω = ±ωk . Still has magnitude and direction. but simplifies to a single component with ± ˆ sign. angular acceleration.15 ˆ k Vectors In planar kinematics. + ccw (curling in the direction of the right hand fingers) – cw (curling in the opposite direction of the right hand fingers) . and ˆ torque (moment) vectors are arrows along about the k axis (the unit direction for the Z axis. We will often represent these k vectors by curled arrows in the XY plane. angular velocity. perpendicular to the plane).

16 Vector Examples 1  P =  1 2  3  P2 =   2  Addition: P + P2 = 1 Dot Product: P • P2 = 1 Cross Product: P × P2 = 1 .

acceleration. where m is the number of rows and n in the number of columns. Used in velocity. and dynamics analysis linear equations (not used in position which is a non-linear solution).  a11 a12 a a22 A] =  21 [    am1 am 2 a1n  a2 n     amn  Used to simplify and standardize the solution of n linear equations in n unknowns (where m=n).17 Matrices Matrix: m x n array of numbers. Special Matrices  a11 a12 [ A] =  a21 a22   a31 a32   a11 0 [ A] =  0 a22  0 0  1 0 0  [ I ] = 0 1 0    0 0 1    a13  a23   a33   0 0  a33   Square (m=n=3) Diagonal Identity .

18 Transpose  a11 T [ A] =  a12   a13  a21 a22 a23 a31  a32   a33   a12 a22 a23 a13  a23   a33   Symmetric  a11 T [ A] = [ A] =  a12   a13   x1    Column Vector (3x1 matrix) { X } =  x2     x3  Row Vector (1x3 matrix) {X } T = { x1 x2 x3 } Matrix Addition Just add up like terms f  a + e b + f   = c + g d + h  h   Just multiply each term a b   e c d  +  g    Matrix Multiplication with Scalar  a b   ka kb  k = c d   kc kd      .

. if not. the number of columns in the left-hand matrix must equal the number of rows in the right-hand matrix.e. as stated above and the dimension of the result is the outer indices. i. 2x1.Matrix Multiplication [C ] = [ A][ B ] ≠ [ B ][ A] 19 Row. the multiplication is undefined and cannot be done! Multiplication proceeds by multiplying and adding terms along the rows of the left-hand matrix and down the columns of the right-hand matrix: (use your index fingers from the left and right hands): Example: [C ] =  a b d e ( 2 x1) ≡ ( 2 x3)( 3x1) g  c    ag + bh + ci  h =   f    dg + eh + fi   i  note the inner indices (p=3) must match. Column indices have to line up as follows: [C ] = [ A][ B ] ( mxn ) ≡ ( mxp )( pxn ) That is.

20 Matrix Multiplication Examples 1 2 3 A] =  [   4 5 6 7 8  [ B ] = 9 8    7 6    [C ] = [ A][ B ] 7 1 2 3  = 9 4 5 6    7   7 + 18 + 21 =  28 + 45 + 42 8 8  6  8 + 16 + 18   46 42   = 115 108 32 + 40 + 36    ( 2 x 2 ) ≡ ( 2 x3)( 3x 2 ) [ D ] = [ B ][ A] 7 8  9 8  1 2 3 =   4 5 6   7 6    7 + 32 14 + 40 21 + 48  39 54 69  =  9 + 32 18 + 40 27 + 48 =  41 58 75      7 + 24 14 + 30 21 + 36   31 44 57      ( 3x3) ≡ ( 3x 2 )( 2 x3) .

[ A][ A]−1 = [ A]−1 [ A] = [ I ] where [I] is the identity matrix. To calculate the matrix inverse: [ A] where: A −1  Adjoint ( A )   = A Determinant of [A] T  Adjoint ( A )  = Cofactor ( A )      Cofactor(A) aij = ( −1) i+ j M ij Minor M ij is the determinant of the submatrix with row i and column j removed. .21 Matrix Inversion Matrix “division”: given [C ] = [ A][ B ] . solve for [B] [C ] = [ A][ B ] ⇒ [ A] [C ] = [ A] [ A][ B ] = [ I ][ B ] = [ B] −1 −1 ⇒ [ B ] = [ A] −1 [C ] Matrix [A] must be square to invert. the matrix “1”.

If not. A = 0 and the inverse of matrix [A] is undefined (dividing by zero).22 System of Linear Equations We can solve n linear equations in n unknowns with the help of a matrix. For n=3: a11 x1 + a12 x2 + a13 x3 = b1 a21 x1 + a22 x2 + a23 x3 = b2 a31 x1 + a32 x2 + a33 x3 = b3 Using matrix multiplication (backwards). this is written as: [ A]{ x} = {b} where:  a11 a12 [ A] =  a21 a22   a31 a32  a13  a23   a33   (known coefficients)  x1    { x} =  x2  x   3  b1    {b} = b2  b   3 −1 (unknowns to be solved) (known right-hand sides) Unique solution { x} = [ A] {b} only if [A] has full rank. .

 x1  − 1 / 2 1 / 4   5  1   =  14 = 2  x 2   3 / 4 − 1 / 8     check: Plug answer into original equations and compare to the {b} we need to get. x1 + 2 x2 = 5 6 x1 + 4 x2 = 14 ⇒ 1 2  x1   5  6 4  x  = 14   2     x1    x2  [ A] =   1 2  6 4  {x} =  {b} =   5  14 {x} = [ A]−1{b} A = 1(4 ) − 2(6 ) = −8 Determinant non-zero.23 Matrix Example Solution of simultaneous linear equations. . unique solution! [ A]−1 = check: 1  4 − 2  − 1 / 2 1 / 4  = A − 6 1   3 / 4 − 1 / 8     [ A][ A]−1 = [ A]−1[ A] = [I 2 ] =   1 0  0 1  Answer.

dot1 = dot(P1.2. % cross2 = cross(P2. % sum2 = P2+P1. A = [1 2. dA = det(A).P2). invA = inv(A).0]. b = [5. x = invA*b.24 Vector and Matrix Matlab Examples P1 = [1. % dot2 = dot(P2.6 4]. % P2 = [3. x1 = x(1). cross1 = cross(P1.P2). x2 = x(2).0]. A*x Define two vectors Vector addition Vector dot product Vector cross product % Define a matrix and vector % % % % Calculate determinant of A Calculate the inverse of A Solve linear equations Extract answers % Check answer – should be b .14]. sum1 = P1+P2.2.P1).P1).

programming. Double-click on Matlab icon to get started.25 Matlab Introduction Matrix laboratory Control systems simulation and design software. <Enter> to suppress result. programming is vaguely C-like. Simulink. Very widespread in other fields. Recommended operation mode: m-files. tabfriendly.m (create a file with the beautiful Matlab Editor/Debugger . Try all the categories under Matlab first.this is color-coordinated. ENT college has a Matlab license.com). and Stateflow for now. Press <Enter> to see result or . vectors. animation. Introduction to basics. Can buy student version software and manual for about the price of one textbook (can use it for many classes!). Put your sequence of Matlab statements in an ASCII file name. but much simpler to use.mathworks. (Exception: there is Symbolic Math under Toolboxes for the adventurous student!). Type >>demo to get a comprehensive overview of Matlab including built-in functions. you can ignore Toolboxes. matrices. plots. Based on C language. it is installed in most computer labs. with parentheses alignment help and debugging . Sold by Mathworks (http://www. Type in commands (such as the Vector/Matrix examples given earlier) at the Matlab prompt >>.

Plus. plots. place the cursor over different variables in the M-File inside the Editor/Debugger to see the values! On-line help is generally great: >>help Example m-files (given on the following two pages) 1) MatEx1. vector array. If you use this suppression. animation. . Use the . multiplication.26 capabilities). Matlab language is interpretive and executes line-by-line. 2) MatEx2. A % indicates a comment. assuming your file is in the search path. transpose.m: Input. and solution of linear equations. after running a file. or matrix). programming. it will give a message at the bad line and then quit. Alternative: at the >> prompt type the M-File name name. One basic way to run your program is to hit the ‘save and run’ button on the editor toolbar. Type: >>who to show you what variables you have defined. very useful for debugging. without the . the variable name still holds the resulting value(s) – just type the variable name at the prompt after the program runs to see the value(s). >>whos will show the variables.m: Matrix and vector definition. plus their matrix dimensions (scalar.m. If there is a syntax or programming logic error. at the end of statements to suppress intermediate results.

multiplication. and transpose = [1 2 3. = [1. ..2. = inv(A2)*b. % % % % % % Define matrix [A2] to be a 3x3 coefficient matrix b detA2 x2 Define right-hand side vector of knowns {b} First check to see if det(A) is near zero Calculate {x2} to be the solution of Ax=b by inversion Check results.. z % % % = b . 8 2 10]..check. = [3. . Better be zero! check = A2*x2..2.m % Matrix.27 %--------------------------------------------------------------% Matlab Example Code 1: MatEx1. vT = v'. % % % A1 x1 v % Clear the cursor and clear any previously defined variables Matrix and Vector definition... = A1*x1. is continuation line) Define 3x1 vector {x1} 2x1 vector {v} is the product of [A1] times {x1} Transpose of matrix [A1] Transpose of vector {v} A1T = A1'. 1 -1 1. ME 301 %--------------------------------------------------------------clc. clear. Display the user-created variables (who).. % % % A2 Solution of linear equations Ax=b = [1 2 3. Vector examples % Dr.1]. = det(A2). .. 1 -1 1].3]. Bob. with dimensions (whos) who whos % % % v x2 z Display some of the results . % % % % % Define 2x3 matrix [A1] (.

and Plotting % Dr. . axis('square'). FOR loop. Animation.m % Menu. % % % Animate if user wants to Give a blank graphics window For loop to animate Single link coordinates Animate to screen Define square plot limits Pause to see animation Pause to maximize window User hits Enter to continue Circle coordinates Cosine function of theta Sine function of theta % % plots % figure.5 -1.th/DR. cos(th). xlabel('\itX (\itm)'). % Constants = 1. ylabel('\itY (\itm)'). dth. % y2 = [0 L*sin(th(i))]. axis(['square']). thf] (deg): ') th0 = the(1)*DR. N = (thf-th0)/dth + 1. % pause(1/4).'g'). sin(th). grid. thf = the(3)*DR. % for i = 1:N. Input. % if i==1 % pause. % set(gca. % % % % % Menu to screen User types input Initial.f1.18).'r'.28 %--------------------------------------------------------------% Matlab Example Code 2: MatEx2.'No') the = input('Enter [th0. xlabel('\itX (\itm)'). IF logic.'b'). set(gca. axis([-2 2 -2 2]).5]). Bob. set(gca. dth = the(2)*DR. r*sin(th).5 1. figure. xlabel('\it\theta (\itdeg)').'FontSize'. title('Functions of \it\theta'). plot(x2. % x2 = [0 L*cos(th(i))].f2. = 2. th = [th0:dth:thf]. ME 301 %--------------------------------------------------------------clc. final thetas Assign theta array Number of iterations for loop % % Animate single link % % if anim == 1 figure. r clear. % end end end % % % xc yc f1 f2 Calculate circle coordinates and cosine function = = = = r*cos(th).'FontSize'. legend('Cosine'. title('Circle').5 1.'FontSize'. axis([-1.'Sine'). ylabel('\itY (\itm)').'Yes'.18). grid. L % Clear the cursor and clear any previously defined variables DR = pi/180. grid.18).y2). % Plot circle plot(xc. % Co-plot cosine and sine functions plot(th/DR. ylabel('Functions of \it\theta'). axis([0 360 -1 1]). delta.yc. % % Input % anim = menu('Animate Single Link?'.

29 Mobility Mobility: Degrees-of-freedom (dof): How many dofs does an unconstrained planar link have? What is the effect of constraining that link with a revolute joint? Grubler's Criterion: Planar Jointed Devices Where: M is the mobility N is the total # of links. including ground J1 is the number of one-degree-of-freedom joints J2 is the number of two-degree-of-freedom joints .

must count m-1 joints! Caution: must count ground link (its freedom is subtracted in formula with n-1. Planar mechanical device classification: M >1 M =1 M =0 M <0 .30 One-degree-of-freedom joints: Revolute Prismatic Two-degree-of-freedom joints (all have rolling and sliding): Cam joint Gear joint Slotted-pin joint Caution: m links joining at one revolute location.

31 Planar Mobility Examples: 1) 3-link serial robot 2) 4-bar linkage 3) Slider-crank linkage 4) Scotch Yoke mechanism 5) Cam and follower .

32

6)

Gear pair

7)

8)

9)

Statically-determinate structure

10) Statically-indeterminate structure

33

14) 3-RRR parallel robot

34

Four-Bar Mechanism Position Analysis Position (Displacement) Analysis: determination of relative orientation/ position of links in a mechanism. Required for testing motion of a synthesized mechanism. Also required for further analysis: velocity, acceleration, dynamics, forces. Generic Mechanism Position Analysis Statement: Given the mechanism and one dof of position input, calculate the position unknowns. Four-bar Mechanism Position Analysis

Step 1. Draw the Kinematic Diagram:

θ1 – ground link angle r1 – fixed ground link r2 – input link θ2 – input angle θ3 – coupler angle r3 – coupler link r4 – output link θ4 – output angle All angles measured in right-hand sense from horizontal to link.

Draw the Vector Diagram. good for all θ . but choosing a different path. measured from the right horizontal to the link vector (tail-tohead). Write equation by starting and ending at same points. Don’t try to force acute angles. Define all angles in positive sense. add vectors tail-to-head until reach a second point.  L sin θ  Vector Diagram: Step 4. the relationships we can see so easily in the first quadrant hold for all four quadrants:  L cosθ  P=  . Start at one point. . Derive the Vector-Loop-Closure Equation. State the problem: Step 3.35 Step 2.

Square and add: . Then go back to find θ 3 . Write XY Components for Vector-Loop-Closure Equation. Break one vector equation into its two scalar components (X and Y): Step 6. Isolate and eliminate θ 3 and solve for θ 4 . Solve for the Unknowns from the XY Equations. Two coupled nonlinear equations in the two unknowns θ 3 .36 Step 5.θ 4 .

37 This equation has the form: Solve using the tangent half angle substitution (Text Equation 4. Two solutions for θ 4 : .9): θ  t = tan 4   2 cos θ 4 = 1 − t2 1 + t2 sin θ 4 = 2t 1 + t2 We converted a complicated coupled transcendental set of equations into a quadratic polynomial. Much easier to solve (but we doubled the order of the equation!).

physically means the mechanism cannot assemble for that input angle. one for each solution branch. no need to use the atan2 function. Use ratio of Y to X equations: Show graphical interpretation: θ 3 = tan −1    BY − AY     B X − AX  . Why two solutions? (Graphically demonstrate the two branches. See section on Grashof's Law.) What if E 2 + F 2 − G 2 < 0 ? Imaginary solution. Go back to find θ 3 .38 With factor two. Go back to original two XY scalar equations.

coupler point C: Four-bar mechanism transmission angle: Transmission angle µ : relative angle between coupler and output links.39 The basic four-bar mechanism position analysis problem is now solved.g. 0. 90o is ideal. Now that we know the angular unknowns. as a rule of thumb. e. we can find the translational position of any point on the mechanism.180o zero transmission. By geometry: for good transmission in a . Measure of mechanical advantage of mechanism. the absolute value of µ should remain in the range 40 < µ < 140 mechanism.

8 − 47.06.203 r4 = 0. We use the SI system (m). and P C for both branches. Also given RC / A = 5 (in) and δ 3 = 36.79 -1. 0.9 for the coupler link point of interest.036 Branch Open Crossed t 1.284 in r2 = 0.02 These two branch solutions are demonstrated in the figures on the following page.076 r3 = 0.θ 4 .9 θ4 µ 67.9 PC 0. Snapshot Analysis (one input angle) Given this mechanism and θ 2 = 30 .57 53.005 G = 0.9 67.178 m r2 = 3 r3 = 8 r4 = 7 and θ1 = 10.076 F = 0. . 0.19.40 Four-Bar Mechanism Position Analysis: Term Example 1 r1 = 11.0 θ3 121.16 0.7 − 114. calculate θ 3 . Note µ is identical for both branches due to the conventions presented earlier. Results: E = 0.18 Given r1 = 0.3 (Ground link is 11" over and 2" up). µ .

1 -0.05 -0.2 0.2 0.1 0 0.1 -0.41 0.1 X (m) 0. Crossed Branch .1 Y (m) 0.15 0.3 4-bar Example Snapshot.05 -0.1 X (m) 0.15 0.2 0.3 4-bar Example Snapshot.1 0 0.1 Y (m) 0.05 0 -0. Open Branch 0.25 0.25 0.05 0 -0.2 0.

Draw a circle of radius r3. Draw the given input link 2 length at the given angle (point A).42 Graphical Solution: The 4-bar position analysis may be solved graphically. This is an excellent method to validate your computer results at a given snapshot. centered at point A. These circles intersect in general in two places. Graphical Solution Figure: . centered at point O4. Connect the two branches and measure the unknown values. merely by drawing the mechanism and determining the mechanism closure. • • • • • • Draw the known ground link (points O2 and O4). Draw a circle of radius r4.

43

4-bar Snapshot Matlab code: This program solves the 4-bar position analysis problem for both branches given a single θ 2 . The results are drawn to the screen.
%------------------------------------------------------------% 4-bar linkage snapshot position analysis - both branches % Fbarplec.m, with graphical output, Dr. Bob, ME 301 %------------------------------------------------------------clc; clear; % Clear cursor and clear previously defined variables % Inputs DR = pi/180; R = input('Enter [r1, r2, r3, r4, rca, th1, th2, del3] (m and deg): '); r1 = R(1); r2 = R(2); r3 = R(3); r4 = R(4); rca = R(5); th1 = R(6)*DR; th2 = R(7)*DR; del3 = R(8)*DR; % Change degrees to radians r1x = r1*cos(th1); r1y = r1*sin(th1); % Position analysis: theta4 E = 2*r4*(r1*cos(th1) - r2*cos(th2)); F = 2*r4*(r1*sin(th1) - r2*sin(th2)); G = r1^2 + r2^2 - r3^2 + r4^2 - 2*r1*r2*cos(th1-th2); t(1) = (-F + sqrt(E^2 + F^2 - G^2)) / (G-E); % Crossed Branch t(2) = (-F - sqrt(E^2 + F^2 - G^2)) / (G-E); % Open Branch th4(1) = 2*atan(t(1)); th4(2) = 2*atan(t(2)); % th3, coupler point, transmission angle; calculate for both branches for i = 1:2, % theta3 ax = r2*cos(th2); ay = r2*sin(th2); bx = r4*cos(th4(i)) + r1x; by = r4*sin(th4(i)) + r1y; th3(i) = atan2(by-ay,bx-ax); bet = th3(i) + del3; % coupler point pcx(i) = r2*cos(th2) + rca*cos(bet); pcy(i) = r2*sin(th2) + rca*sin(bet); mu(i) = abs(th4(i)-th3(i)); % transmission angle end % Plot 4-bar position results for i = 1:2, x2 = [0 r2*cos(th2)]; y2 = [0 r2*sin(th2)]; x3 = [r2*cos(th2) r1x+r4*cos(th4(i)) pcx(i)]; y3 = [r2*sin(th2) r1y+r4*sin(th4(i)) pcy(i)]; x4 = [r1x r1x+r4*cos(th4(i))]; y4 = [r1y r1y+r4*sin(th4(i))]; figure; plot(x2,y2,'r',x4,y4,'r'); patch(x3,y3,'r'); axis('square'); set(gca,'FontSize',18); xlabel('\itX (\itm)'); ylabel('\itY (\itm)'); axis([-0.1 0.3 -0.15 0.25]); grid; end

% % %

44

Full-Range-Of-Motion (F.R.O.M.) Analysis

A more meaningful result from position analysis is to report the position analysis unknowns for the entire range of mechanism motion. The first plot gives θ 3 (red), θ 4 (green), and µ (blue), all

deg, for all 0 ≤ θ 2 ≤ 360 , for Term Example 1, open branch only. The second plot gives the coupler point location for this branch, plotting PCY vs. PCX .
Thetas 3 (red) and 4 (green), Mu (blue) 180 160 140 Angle (deg) 120 100 80 60 40 20 0 50 100 150 200 θ 2 (deg) 250 300 350

45

0.25 0.2 0.15 0.1 Y (m) 0.05 0 -0.05 -0.1 -0.1 0 0.1 X (m) 0.2 0.3

4-bar Example Snapshot, Open Branch Coupler Curve

5). figure: Problem: inverse sine function is double-valued. e. figure:  2   Problem: inverse cosine function is double-valued. for each θ 4 there are two possible solutions. θ 3 = cos −1  3  . only one of which is correct! Why not calculate θ 3 using X equation? Inverse cosine has a similar problem. why not calculate θ 3 using Y equation?: e.g. θ 3 = sin −1 (0. for each θ 4 there are two possible solutions. only one of which is correct! .46 Trigonometric Uncertainty Return to θ 3 solution: XY scalar equations: r3c3 = r1c1 + r4 c4 − r2 c2 r3 s3 = r1s1 + r4 s 4 − r2 s 2 Since θ 4 has been solved.g.

+  2 2    θ 3 = a tan 2 − . for each θ 4 there are two possible solutions. The plain atan function takes a single quotient input.g. both X and Y equations) .−   1  2 3 = 2   θ 3 = A tan 2(r1s1 + r4 s 4 − r2 s 2 .this suggests using the tangent (as we did earlier in the θ 3 solution): θ 3 = tan −1   e. θ 3 = tan −1  1  . only one of which is correct! Solution: use the quadrant-specific inverse tangent function atan2. r1c1 + r4 c4 − r2 c2 ) .e. given the signs ± of the numerator and denominator. No uncertainty with atan2:  1 3 = e. hence this sign information is lost and the true quadrant is unknown.47 So we must use information from both sine and cosine (i. figure:   3   r1s1 + r4 s 4 − r2 s 2     r1c1 + r4 c4 − r2 c2  Problem: the plain atan inverse tangent function is still doublevalued!. θ 3 = a tan 2 + .g. the function has built-in logic to determine the correct quadrant for the angle answer. Input to this function is both a numerator and denominator.

. we don't know if the solution is   3  2 θ4 θ4 2 2 = 30 or = 210 However.) With the 2 multiplying the inverse tangent result. from before. Example: θ For 4 = tan −1  1  . having just cleared up this Trigonometric Uncertainty.48 Now. only showing one here. for next time consider the following: Do the solutions for θ 4 always exist? What if E 2 + F 2 − G 2 = 0 ? What if E 2 + F 2 − G 2 < 0 ? Stay tuned . . one for each t value. the multiple 2 takes care of this uncertainty: θ 4 = 60 or θ 4 = 420 = 60 Now. it doesn't matter whether we use atan or atan2 since the final answer will come to the same angle. we already have an exception in the θ 4 tangent half-angle solution: θ 4 = 2 tan −1 (t ) (there are two branches. .

intermediate 2 links: L. no limits Rocker . Q 1) If L + S < P + Q Then we call this a Grashof Mechanism and there are four different mechanisms and rotation conditions: Diagrams: a) b) c) . P.not full rotation. Grashof's Law is used to determine the relative rotatability of the input and output links in a 4-bar mechanism: Crank . rotates back-and-forth between limits Mechanism types (input / output links): Identify longest. shortest. S.full rotation.49 Grashof’s Law Grashof was a German Engineer in the late 1800s.

Q = 7 – demonstrate branch jumping Another interesting example: L = P = 10. S = 6. Also.50 2) If L + S > P + Q Then we call this a Non-Grashof Mechanism and the are four different mechanism inversions yield only one rotation condition: 3) If L + S = P + Q Then we call this a Special Grashof Mechanism and the four different mechanism inversions yield the identical rotation conditions from 1) Grashof Mechanism. Examples 1) L = 10. Q = 7 – all Double Rockers 3) L = 10. S = Q = 4 parallel. However. P = 8. S = 5. locomotive linkage – subject to branch jumping unless constrained. Q = 7 – demonstrate the 4 possibilities 2) L = 10. very easy analysis: θ2 =θ4 = µ θ 3 = 0 for all motion! . there is the additional interesting and troublesome feature that the mechanism may jump branches! Centerlines of links can become collinear. S = 4. P = 8. P = 8.

symmetric about the ground link. To calculate the joint limits. These joint limits occur when links 3 and 4 are aligned.51 4-Bar Joint Limits If Grashof's Law predicts the input link is a rocker. there will be rotation limits on the input link. there will be two joint limits. we use the law of cosines: (r3 + r4 )2 2 = r12 + r2 − 2r1r2 cos θ 2 L θ 2 L = ± cos     2 −1 r1 2 + r2 − (r3 + r4 )2   2r1r2   ± by symmetry about r1 . As shown in the figure.

5 (links 2 and 3 folded upon each other in a line) There are no limits on θ 2 since it is a crank.1 (links 2 and 3 stretched in a line) θ 4 L = 172. For certain mechanisms.9   Example 2: Given r1 = 10. r4 = 7 L + S > P + Q (10 + 6 > 8 + 7) so we predict only double rockers from this mechanism. r2 = 4. r3 = 8.742] = ±137. the limiting conditions occur when links 3 and 4 fold upon each other instead of stretching straight out. r3 = 8. r2 = 3. so we predict this mechanism is a crank-rocker. r2 = 6.52 Example 1: Given r1 = 10.18. limits are: θ 4 L = 120. . r4 = 7 L + S < P + Q (10 + 4 < 8 + 7). r3 = 8. θ 2 L = ± cos    2 2  2 −1 10 + 6 − (8 + 7 ) 2(10 )(6 ) −1  = ± cos [− 0. there are no joint limits! θ 2 L = ± cos    2 2  2 −1 10 + 4 − (8 + 7 )  = ± cos 2(10 )(4 )   which is undefined! −1 [− 1. Therefore. r4 = 7 (in) and θ1 = 10. Example 3: (Term Example Four-bar) Given r1 = 11.3 .3625] Caution: the figure on the previous page does not apply in all joint limit cases. The previous method can also be used to find angular limits on link 4 when it is a rocker. here links 2 and 3 either stretch out in a line or fold upon each other.

the four kinematic inversions of the slider-crank mechanism yield radically different types of motion. In class we will solve the Air Compressor case where the crank is the input and the slider is the output. output is rotation of drive shaft. x is measured horizontally from the origin to the slider/coupler revolute joint location. Step 1. piston (output) compresses air. . Internal Combustion Engine – explosion drives piston (input). Draw the Kinematic Diagram: r2 – input link length r3 – coupler link length h – slider offset θ2 – input angle θ3 – coupler angle x – output displacement Link 1 is the fixed ground link. Unlike the four-bar mechanism. Air Compressor – electric motor drives crank (input).53 Slider-Crank Mechanism Position Analysis Converts linear motion to rotary or rotary motion to linear via connecting rod. All angles measured in right-hand sense from horizontal to link. Two dead points where piston is at limits. Use flywheel on crank to avoid locking.

Draw the Vector Diagram. but choosing a different path. Start at one point. Vector Diagram: Step 4. . State the problem: Step 3. Write equation by starting and ending at same points. add vectors tail-to-head until reach a second point. measured from the right horizontal to the link vector (tail-tohead). Derive the Vector-Loop-Closure Equation.54 Step 2. Define all angles in positive sense.

θ 3 . notice that the two XY equations are coupled only in θ 3 but not in x.solve θ 3 using the Y equation only and then solve x from the X equation: What about trigonometric uncertainty? The inverse sine function is double-valued and so there are two valid solution branches. We could isolate on unknown. Write XY Components for Vector-Loop-Closure Equation. . Two coupled nonlinear equations in the two unknowns x. and solve as in the four-bar approach.55 Step 5. Graphically demonstrate the two branches. Solve for the Unknowns from the XY Equations. Break one vector equation into its two scalar components (X and Y): Step 6. square & add. However. There a simpler method .

The worst case is θ 2 = −90 . which yields r3 ≥ h + r2 This condition was derived assuming positive h.56 Full-rotation condition For solution to exist for entire motion range (r2 is a crank). allowing negative h: r3 ≥ h + r2 . absolute value of the inverse sine argument must be less than or equal 1: h − r2 s2 ≤1 r3 r3 ≥ h − r2 s2 which must hold for all motion. .

8 These two branch solutions are demonstrated in the figures on the following page.114 θ3 7. calculate x and θ 3 for both branches.203 m h = 0.102 r3 = 0. We use the SI system (m).290 -0.076 Snapshot Analysis (one input angle) Given this mechanism and θ 2 = 30 .57 Slider-Crank Mechanism Position Analysis: Term Example 2 Given: r2 = 4 r3 = 8 in h=3 r2 = 0.2 172. . Results: Branch Open Crossed x (m) 0.

Open Branch 0.1 X (m) 0.1 0 0.2 0.1 Y (m) 0 -0.2 0.58 0. Crossed Branch .1 -0.3 Slider-Crank Example Snapshot.2 -0.1 X (m) 0.2 0.2 0.1 -0.2 -0.1 Y (m) 0 -0.1 0 0.3 Slider-Crank Example Snapshot.

We can draw two right triangles representing these conditions and easily calculate the x limits to be 0. This is an excellent method to validate your computer results at a given snapshot. centered at point A. This circle intersects the slider line in general in two places.0671 ≤ x ≤ 0. as seen in the full motion x plot. Draw the given input link 2 length at the given angle (point A). • • • • • • Place the grounded revolute for the crank at the origin. Draw a circle of radius r3.59 Graphical Solution: The Slider-Crank position analysis may be solved graphically. Connect the two branches and measure the unknown values. merely by drawing the mechanism and determining the mechanism closure. .2951. next page. Draw the line of the slider. The slider reaches its maximum displacement when links 2 and 3 are aligned straight out and its maximum displacement when link 2 if folded onto link 3. offset vertically from the origin by h. Graphical Solution Figure: Slider Limits The crank will rotate fully if the previously-derived condition is met.

05 0 50 100 150 200 θ 2 (deg) 250 300 350 .) Analysis A more meaningful result from position analysis is to report the position analysis unknowns for the entire range of mechanism motion.3 0. for all 0 ≤ θ 2 ≤ 360 . The second plot gives θ 3 (deg).O.1 0.15 0. Slider Displacement 0. The first plot gives x (m). right branch only.25 0.R. for Term Example 2.M. for all 0 ≤ θ 2 ≤ 360 . for the right branch only.60 Full-Range-Of-Motion (F.2 x (m) 0.

61 Coupler Angle 60 50 40 (deg) 30 20 10 0 -10 0 50 100 150 200 θ 2 (deg) 250 300 350 θ 3 .

Analytical velocity analysis: write position vectors. velocity. take first time derivatives. direction also crucial. Velocity analysis is solution of coupled linear equations. solve for unknowns. Units (translational and rotational): Basic Velocity Derivation Figure: Most general planar case: Translating and rotating rigid rod with a slider on it. machine design. acceleration. Express the position vector in Cartesian coordinates: P P = PO + L = . Velocity analysis is also required for dynamics: position. Find the total velocity of point P on the slider. forces. Some tasks have timing. rates. dynamics. Velocity is the first time derivative of the position. Position analysis must be completed first.62 Velocity Analysis Introduction Velocity analysis is important for kinematic motion analysis. Vector quantity: Magnitude of velocity is speed.

dt dx dt Example: d (L cos θ ) = ? dt . Chain rule: d df dx ( f ( x(t ))) = f is a function of x. Product rule: d dx dy ( xy ) = y + x dt dt dt x. y both functions of time. The length of the rod is changing with sliding velocity: Product and Chain Rules of Differentiation We’ll need to use the product and chain rules over and over in velocity and acceleration analysis derivations. the spatial rotation case is more complicated. which is a function of t.63 The angle is changing with angular velocity: Only the planar case is this simple.

64 Back to basic velocity derivation First time derivative of position vector: VP = d PP = dt .

summarized below: Vector Name ω×L VO V Point O Velocity Sliding Velocity Tangential Velocity XY Components Magnitude / Direction .65 We have just derived the Three-Part Velocity Equation: V P = V O +V +ω × L The terms for the Three-Part Velocity Equation can be expressed in various ways.

θ = 30 .66 Three-Part Velocity Equation Example: Given (instantaneously) L = 2 m.598 m VP = =  2 + 1. T V + V cosθ − Lω sin θ   3 + 3cos30 − 2 (1) sin 30  V P =  OX =  VOY + V sin θ + Lω cosθ  2 + 3sin 30 + 2 (1) cos30    3 + 2.7 m/s Show magnitude and direction of each velocity component: Vector Name VO V ω×L Point O Velocity Sliding Velocity Tangential Velocity XY Components Magnitude / Direction . V = L = 3 m/s (outward). V P = 6.598 − 1  4. ω = 1 rad/s. V O = {3 2} m/s.732  5.965@ 48.5 + 1.232  s  or. calculate V P .

forces. Draw the Velocity Diagram: where ω i . Also required for further analysis: acceleration. Required for complete motion analysis. dynamics.67 Four-Bar Mechanism Velocity Analysis Velocity Analysis: determination of angular and linear velocities of links in a mechanism. Linear equations result from first time differentiation of position equations.3. ω 1 = 0 since the ground link is fixed. 1-dof mechanism. Position Analysis must first be complete. Position analysis must be complete first. is the absolute angular velocity of link i. Generic Mechanism Velocity Analysis Statement: Given the mechanism. . Step 2. Unique solution for each mechanism branch. complete position analysis.4. calculate the velocity unknowns. machine design. Four-bar Mechanism Velocity Analysis Step 1. and one dof of velocity input. so one velocity input must be given. i = 2.

68 Step 3. State the problem: Step 4. Take the first time derivative of the vector loop closure equations from position analysis. Derive the velocity equations. in XY component form. Four-bar mechanism position equations: r 2 + r 3 = r1 + r 4 r2c2 + r3c3 = r1c1 + r4c4 r2 s2 + r3 s3 = r1s1 + r4 s4 First time derivative for velocity equations: (use chain rule several times) Chain rule: d cosθi dθi d ( cosθi ) = dt dθi dt = − sin θiθ i = − sin θiωi d sin θi dθi d ( sin θi ) = dt dθi dt = cosθiθi = cosθiωi Don’t have to use product rule because ri = 0 (rigid links). .

69 The first time derivative of the position equations is: Gathering unknowns on the LHS: Substituting simpler terms: Written in matrix form: .

Algebra solution: . Solve the velocity equations for the unknowns ω 3 .70 Step 5. ω 4 .

71 Alternate matrix solution (yields same solution): .

72 Four-Bar mechanism singularity condition: When does the solution fail? This is a mechanism singularity. this happens when links 3 and 4 are straight out or folded on top of each other (what does this correspond to?): . Physically. when the determinant of the coefficient matrix goes to zero. ω 4 . Let’s see what this means physically. The result is dividing by zero. for infinite answers ω 3 .

e.g. Now that we know the angular unknowns. we can find the translational velocity of any point on the mechanism. coupler point C: .73 The basic four-bar mechanism velocity analysis problem is now solved.

0. This is the open branch of the four-bar mechanism position example (Term Example 1). and V C for this instant (snapshot). θ3 = 53.178.093 ω  =  0.271  =  ω4  −0. θ 4 = 121. Snapshot Analysis (one input angle) Given this mechanism position analysis plus ω2 = π rad/s (+.207    4   ω3   −1.076.209   . so ccw).120 −0.  0. calculate ω 3 .9 .042  Coupler point translational velocity: V C =   (m/s) 0. θ 2 = 30 . δ 3 = 36.151 ω3   −0. These results are the absolute angular velocities of links 3 and 4 with respect to the ground link.8 .164 −0.7 . r3 = 0.203.127 m. ω 4 .120   −0.284. and θ1 = 10. RC/A = 0.587  Both are negative. so cw direction. r2 = 0. r4 = 0.74 Four-bar mechanism velocity example: Given r1 = 0.3 .

5 (rad/s) and ω 1 0.5 0 50 100 150 200 θ 2 (deg) 250 300 350 . θ 2 (since it is related to time t via θ 2 = ω 2 t ).75 Full-Range-Of-Motion (F. The plot below gives ω 3 (red) and ω 4 (green).) Analysis A more meaningful result from velocity analysis is to report the velocity analysis unknowns for the entire range of mechanism motion.O.5 0 ω 3 4 -0. (rad/s).M.5 -1 -1.R. open branch only. Since ω 2 is constant. for Term Example 1. Omegas 3 (red) and 4 (green) 2 1. we can plot the velocity results vs. for all 0 ≤ θ 2 ≤ 360 .

76

The plot below gives the translational coupler point velocity for all 0 ≤ θ 2 ≤ 360 , for Term Example 1, open branch only.
Coupler Point Velocities X (red) and Y (green) 0.3 0.2 VCX and VCY (m/s) 0.1 0

-0.1 -0.2 0 50 100 150 200 θ 2 (deg) 250 300 350

77

Derivative/Integral Relationships When one variable is the derivative of another, what are the relationships? For example: dθ ω3 = 3 θ 3 = θ 30 + ∫ ω3dt dt
1.2 (rad) 1 0.8 0.6 0.4 0 1 0 50 100 150 200 250 300 350

ω

θ

3

-1 0 50 100 150 200 θ 2 (deg) 250 300 350

The value of ω3 at any point is the slope of the θ 3 curve at that point. The value of θ 3 at any point is the integral of the ω3 curve up to that point (the value of θ 3 at any point is the area under the ω3 curve up to that point). This graph is vs. θ 2 , but the same type of relationships hold as for time t since ω 2 is constant. This is the Term Example 1 result, but we changed θ 3 from deg to rad for better comparison.

78

Slider-Crank Mechanism Velocity Analysis

Again, we will solve the Air Compressor case where the crank is the input and the slider is the output. The Internal Combustion Engine case (slider input/crank output) is equally interesting. Step 1. Position Analysis must first be complete. Step 2. Draw the Velocity Diagram:

where ω i , i = 2,3 is the absolute angular velocity of link i. x is the variable slider velocity. ω 4 = 0 since the slider cannot rotate. Step 3. State the problem:

Take the first time derivative of the vector loop closure equations from position analysis. Derive the velocity equations. in XY component form. Slider-crank mechanism position equations: r2 + r3 = x + h First time derivative for velocity equations: r2c2 + r3c3 = x r2 s2 + r3 s3 = h Gathering unknowns on the LHS: Written in matrix form: .79 Step 4.

x . these equations are decoupled so we don't need a matrix solution. solve ω3 from Y equation: Then solve x from the X equation using the ω3 result: The alternate matrix solution: will yield identical results. Actually. Solve the velocity equations for the unknowns ω3 . First.80 Step 5. .

81 Slider-crank mechanism singularity condition: When does the solution fail? This is a slider-crank mechanism singularity. Physically. resulting in infinite answers ω3 . . A = r3c3 = 0 A = 0 when cos θ 3 = 0 . The result is dividing by zero.270 . Of course r3 cannot go to zero. otherwise we have a degenerate slider-crank mechanism. even with offsets. or θ 3 = 90 . Doesn’t happen for nominal full-rotation slider-crank mechanisms. x . this happens when link 3 is straight up or down (θ3 = ±90 ). when the determinant of the coefficient matrix goes to zero.

025   x  −0.202  ω  =  0.203.290 m. and θ 2 = 30 .062   =  ω3  −0. calculate x.ω3 for this instant (snapshot). This is the right branch of the slider-crank position example (Term Example 2).102. 1 0. These results are the absolute linear and angular velocities of links 4 and 3 with respect to the fixed ground link.2 . r3 = 0.686   Both are negative. h = 0.138    3    x  −0. θ3 = 7. so the slider is currently traveling to the left and the coupler link is currently rotating in the cw direction.076 m. Snapshot Analysis (one input angle) Given this mechanism position analysis plus ω2 = π / 2 rad/s (+. .82 Slider-crank mechanism velocity example: Given r2 = 0. so ccw). x = 0.080   0 −0.

Xdot (red) and Omega3 (green) 0. for Term Example 2. Since ω 2 is constant.M. θ 2 (since it is related to time t via θ 2 = ω 2 t ). we can plot the velocity results vs. m/s) and ω 3 (green.6 -0.83 Full-Range-Of-Motion (F.O.8 0 50 100 150 200 θ 2 (deg) 250 300 350 .6 (rad/s) xd (m/s) and ω 0.4 -0. for all 0 ≤ θ 2 ≤ 360 . The plot below gives x (red.2 -0. rad/s).) Analysis A more meaningful result from velocity analysis is to report the velocity analysis unknowns for the entire range of mechanism motion.4 0.2 0 3 -0.R. right branch only.8 0.

3 x (m) 0. but the same type of relationships hold as for time t since ω 2 is constant. . θ 2 . This graph is vs.1 0 50 100 150 200 250 300 350 xd (m/s) 0. The value of x at any point is the integral of the x curve up to that point (the value of x at any point is the area under the x curve up to that point).84 Derivative/Integral Relationships When one variable is the derivative of another.2 0 -0.2 0. This is the Term Example 2 result. what are the relationships? For example: dx x = x 0 + ∫ xdt x= dt 0.2 0 50 100 150 200 θ 2 (deg) 250 300 350 The value of x at any point is the slope of the x curve at that point.

Acceleration is the first time derivative of the velocity and second time derivative of the position. Acceleration analysis is solution of linear equations. Vector quantity: Analytical acceleration analysis: write position vectors.85 Acceleration Analysis Introduction Acceleration analysis is required for dynamics: position. Important for kinematic motion analysis. take first two time derivatives. machine design. velocity. . acceleration. Position and velocity analyses must be completed first. forces. Find the total acceleration of point P on the slider. dynamics. Units (translational and rotational): Basic Acceleration Derivation Figure: Rotating rigid rod with a slider on it. solve for unknowns.

The length of the rod is changing with sliding velocity and acceleration: .86 Recall the 2-part position and 3-part velocity results:  POX + L cos θ  P P = PO + L =   POY + L sin θ   VOX + V cos θ − Lω sin θ  V P = V O +V +ω × L =   VOY + V sin θ + Lω cos θ   The angle is changing with angular velocity and acceleration: Only planar case is this simple. the spatial rotation case is more complicated.

which is a function of t. we’ll need to use the product and chain rules over and over in acceleration analysis derivations. Product rule: d dx dy ( xy ) = y + x dt dt dt x. y both functions of time.87 Product and Chain Rules of Differentiation Again. Chain rule: d df dx ( f ( x(t ))) = f is a function of x. dt dx dt Example: d2 ( L cosθ ) = ? dt 2 .

88 Back to basic acceleration derivation First time derivative of velocity vector (Second time derivative of position vector): dV P d 2 P P = = AP = 2 dt dt .

89 We have just derived the Five-Part Acceleration Equation: A P = AO + A + 2ω × V + α × L + ω × (ω × L ) These terms can be expressed in various ways. summarized below: Vector Name AO A 2ω × V α×L Tangential Acceleration ω × (ω × L ) Centripetal Acceleration Point O Acceleration Sliding Acceleration Coriolis Acceleration XY Components Magnitude / Direction .

464 − 1 11. ω = 1 rad/s. AO = {1 2} .0 m/s2 Show magnitude and direction of each Acceleration component: Vector Name AO Point O Acceleration A Sliding Acceleration 2ω × V Coriolis Acceleration α×L Tangential Acceleration ω × (ω × L ) Centripetal Acceleration XY Components Magnitude / Direction . calculate A P . θ = 30 .268 m = =  2 + 2 + 5. A = L = 4 m/s2 (outward).660  s 2  or. V O = {3 2} . T V = L = 3 m/s (outward).732  −2. T  AOX + A cosθ − 2V ω sin θ − Lα sin θ − Lω 2 cosθ  AP =   AOY + A sin θ + 2V ω cosθ + Lα cosθ − Lω 2 sin θ   1 + 3. α = 2 rad/s2.464 − 3 − 2 − 1. Given (instantaneously) L=2 m.196 + 3.90 Five-Part Acceleration Equation Example: Continuation of 3-part velocity example. AP = 11.879@101.

91 Four-Bar Mechanism Acceleration Analysis Acceleration Analysis . Four-bar Mechanism Acceleration Analysis Step 1.3. complete position and velocity analyses. Linear equations result from second time differentiation of position equations. Position and velocity analyses must be complete first. Unique solution for each mechanism branch. Step 2. Generic Mechanism Acceleration Analysis Statement: Given the mechanism. Draw the Acceleration Diagram: where α i . Also required for further analysis: dynamics. and one dof of acceleration input. 1-dof mechanism. Required for complete motion analysis.4 is the absolute angular acceleration of link i. i = 2. so one acceleration input must be given. Position and Velocity Analyses must first be complete. . calculate the acceleration unknowns.determination of angular and linear accelerations of links in a mechanism. forces. machine design. α 1 = 0 since the ground link is fixed.

in XY component form. State the problem: Step 4. Take the first time derivative of the four-bar mechanism velocity equations from velocity analysis. Derive the acceleration equations. Four-bar mechanism velocity equations: − r2ω2 s2 − r3ω3 s3 = − r4ω4 s4 r2ω2c2 + r3ω3c3 = r4ω4c4 The first time derivative of the velocity equations is: Gathering unknowns on the LHS: .92 Step 3.

93 Substituting simpler terms: Written in matrix form: .

Solve the acceleration equations for the unknowns α 3 . Matrix solution (Algebra solution yields the same results): .α 4 .94 Step 5.

and acceleration.g.95 Four-Bar mechanism singularity condition: Same coefficient matrix A as velocity case. there is zero transmission angle µ and Link 2 is at a joint limit! The basic four-bar mechanism acceleration analysis problem is now solved. Now that we know the angular unknowns. we can find the translational acceleration of any point on the mechanism.180 . so singularity condition is identical: θ 4 − θ 3 = 0 . coupler point C: . e. This condition is the same problem for position. velocity. At this singularity.

676  2 Coupler point translational acceleration: AC =   m/s −0. θ3 = 53. and θ1 = 10.178. ω3 = −1. θ 2 = 30 . θ 4 = 121. r4 = 0. so ccw direction.96 Four-bar mechanism acceleration example: Given r1 = 0.877   −0.284. r2 = 0. RC/A = 0. r3 = 0.093 α  =  −0.203.587 rad/s.  0.164 −0.151 α 3  −0. calculate α 3 . These results are the absolute angular accelerations of links 3 and 4 with respect to the ground link. ω2 = π .3 .9 .030   Both are positive.589    4   α 3  0. ω4 = −0. δ 3 = 36.7 .271.213  =  α 4  6.α 4 for this instant (snapshot).582  .076. plus α 2 = 0 rad/s2.8 .120 −0. −0.127 m. This is the open branch of the position and velocity example (Term Example 1). Snapshot Analysis (one input angle) Given this mechanism position and velocity analysis.

R. Since ω 2 is constant.) Analysis A more meaningful result from acceleration analysis is to report the acceleration analysis unknowns for the entire range of mechanism motion. we can plot the acceleration results vs. for all 0 ≤ θ 2 ≤ 360 .M. θ 2 (since it is related to time t via θ 2 = ω 2 t ).97 Full-Range-Of-Motion (F. The plot below gives α 3 (red) and α 4 (green).O. Alphas 3 (red) and 4 (green) 6 4 2 0 α and α 3 (rad/s ) 4 2 -2 -4 -6 0 50 100 150 200 θ 2 (deg) 250 300 350 . for Term Example 1. (rad/s2). open branch only.

8 0.2 -0.4 0.8 0 50 100 150 200 θ 2 (deg) 250 300 350 .98 The plot below gives the translational coupler point acceleration for all 0 ≤ θ 2 ≤ 360 . for Term Example 1.6 ACX and ACY (m/s ) 2 0.6 -0. open branch only. Coupler Point Accelerations X (red) and Y (green) 0.2 0 -0.4 -0.

5 2 0 50 100 150 200 250 300 350 ω (rad/s) 4 θ 1 0 0 (rad/s ) 4 2 4 -1 50 100 150 200 250 300 350 5 0 50 100 150 200 θ 2 (deg) 250 300 350 -5 0 α . recall the relationships from calculus. For example: ω4 = dθ 4 dt θ 4 = θ 40 + ∫ ω 4 dt ω 4 = ω 40 + ∫ α 4 dt α4 = dω 4 dt (rad) 3 2.99 Derivative/Integral Relationships When one variable is the derivative of another.

Position and Velocity Analyses must first be complete. α 4 = 0 since the slider cannot rotate.3 is the absolute angular acceleration of link i.100 Slider-Crank Mechanism Acceleration Analysis Again. Draw the Acceleration Diagram: where α i . Step 3. State the problem: . Step 1. Step 2. we will solve the Air Compressor case where the crank is the input and the slider is the output. i = 2.

Slider-crank mechanism velocity equations: − r2ω2 s2 − r3ω3 s3 = x r2ω2c2 + r3ω3c3 = 0 The first time derivative of the velocity equations is: Gathering unknowns on the LHS: Written in matrix form: . Take the first time derivative of the velocity equations from velocity analysis. in XY component form.101 Step 4. Derive the acceleration equations.

x . Solve the acceleration equations for the unknowns α 3 . Actually. solve α 3 from Y equation: Then solve x from the X equation using the α 3 result: .102 Step 5. these equations are decoupled so we don't need a matrix solution. First.

290 m. h = 0. Snapshot Analysis (one input angle) Given this mechanism position and velocity analysis plus.062 m/s.103 Slider-crank mechanism singularity condition: Same coefficient matrix as velocity case. This is the right branch of the position and velocity example (Term Example 2). so singularity condition is identical (see the singularity discussion in the slidercrank velocity section). calculate x.686 rad/s.681   These results are the absolute linear and angular accelerations of links 4 and 3 with respect to the fixed ground link.329   =  α 3   0. x = 0. ω3 = −0. r3 = 0. and ω2 = π / 2 .203.102. and θ 2 = 30 . α 2 = 0 rad/s2.076 m.025   x   −0.α 3 for this instant (snapshot).2 . x = −0.312   0 −0. θ3 = 7.137    3    x  −0. Slider-crank mechanism acceleration example: Given r2 = 0.202  α  = −0. 1 0. .

Xdotdot (red) and Alpha3 (green) 1 (rad/s ) 3 0. for Term Example 2. we can plot the velocity results vs.5 0 50 100 150 200 θ 2 (deg) 250 300 350 .M.5 -2 -2.104 Full-Range-Of-Motion (F.) Analysis A more meaningful result from acceleration analysis is to report the acceleration analysis unknowns for the entire range of mechanism motion. Since ω 2 is constant. rad/s2).5 -1 -1. right branch only.5 0 2 2 xdd (m/s ) and α -0.O. The plot below gives x (red. θ 2 (since it is related to time t via θ 2 = ω 2 t ). for all 0 ≤ θ 2 ≤ 360 . m/s2) and α 3 (green.R.

recall the relationships from calculus.105 Derivative/Integral Relationships When one variable is the derivative of another.3 x (m) 0.2 0. For example: x= x= dx dt dx dt x = x 0 + ∫ xdt x = x 0 + ∫ xdt 0.2 0 0.5 0 50 100 150 200 θ 2 (deg) 250 300 350 50 100 150 200 250 300 350 50 100 150 200 250 300 350 xdd (m/s ) 2 .1 0 xd (m/s) 0.2 0 -0.5 0 -0.

for full range of motion we have assumed that the input link rotates fully with a given constant input angular velocity. Our input constraints have thus been 0 ≤ θ 2 ≤ 360 . since both θ2 and t increase steadily (linearly). time t. we have θ 2 = ω2t . . and α2 = 0. Since ω2 is constant. so we could just as well plot all results vs. This input motion specification is plotted like this: Note that we have been plotting calculated results vs. ω2 constant.106 Input Motion Specification Up to this point. θ2.

Why is the previous page’s plots unacceptable in this case? Simplest change – linear angular velocity starting and stopping at zero: We cannot plot vs. Many useful mechanisms have input links that do not rotate fully but travel between joint limits. middle. t. and finish.107 This constant ω2 input specification is fine for mechanisms whose input rotates fully and considering steady-state motion only. What is the weakness of this approach? (Discontinuous acceleration function yields infinite jerk at start. θ2 since it is not increasing linearly – plot vs. starting and stopping at zero angular velocity.) .

108 We can fix this with a trapezoidal input acceleration profile: This input motion specification should be fine (trapezoidal input torque is often used for industrial robots). but there are many different zones to handle – what acceleration profile is similar but with a single function? .

5 3 Example with θ 20 = 60 .5 2 2. θ 2 F = 120 .5 2 2.5 1 1.5 3 -1 0 2 0 0. .5 time (sec) 2 2.5 1 1.5 2 2.5 1 1.5 3 0. and tF = 3 sec.109 Full-cycloidal function input angle specification θ 2 ( t ) = θ 20 + (θ 2 F − θ 20 )  ω2 ( t ) = tF    t 1 2π t  sin − t F 2π tF    tF   2π (θ 2 F − θ 20 )  2π t  α2 (t ) =  sin t  2 tF  F  4π 2 (θ 2 F − θ 20 )  2π t  β2 (t ) = cos  3 tF tF    (θ 2 F − θ 20 ) 1 − cos 2π t  θ2 (deg) β 2 (rad/s3) α2 (rad/s2) ω (rad/s) 2 100 80 60 0 4 2 0 0 1 0 0.5 3 -2 0 0.5 1 1.

What is the first step in analysis? Draw .110 Dynamics Introduction Chart: Kinematics: translational rotational Kinetics: translational Newton’s 2nd Law: rotational Euler’s equation: Dynamics of a single rigid body in the plane Rigid body acted on by a system of forces and moments to produce planar motion. . . .

FBD Simplified FBD MAD (mass-acceleration diagram) .111 Free Body Diagram (FBD) Isolate each rigid body and show the forces and moments acting. This contains all the info needed to write Newton’s 2nd Law and Euler’s equation.

The entire rigid body experiences the same α . Different points in rigid body have different linear accelerations. External forces/moments: Internal forces/moments: Write dynamics equations Newton’s 2nd Law: Euler’s equation: AG is the linear acceleration of center of gravity – same direction as R . .112 Internal and External Forces and Moments All internal and external forces and moments must be included on the FBD. α angular acceleration of rigid body.

R − m AG = 0 R + FO = 0 T + r × R − IG α = 0 T +r×R+MO =0 . “Centrifugal force” − mrω 2 is an example of an inertial force.113 D’Alembert’s Principle Turn dynamics problem into a statics problem by the inclusion of a fictitious “inertial force” F 0 = − m AG and a fictitious “inertial moment” M 0 = − IG α . Subtract RHS of equations. We won’t use this method. it’s not really a force but an effect of acceleration and inertia. just wanted you to know in case you ran into it somewhere. then sum to zero as in statics.

M EXT . and the applied driving force (or torque). external forces and moments. find the resulting mechanism motion and internal joint forces. α 3 . 4-Bar Linkage Forward Dynamics: Given τ 2 and F EXT . α 2 . find the required driving force (or torque) and internal joint forces. F EXT . ω 2 . ω 2 . θ 3 . ω 4 . external forces and moments. find the motion θ 2 . find τ 2 and internal forces F ij . and the desired mechanism motion. ω 3 . θ 4 . M EXT .114 Two Types of Dynamics Problems Forward Dynamics: Given the mechanism. Inverse Dynamics: Given the mechanism. α 4 . ω 3 . and Next lecture: Newton's 2nd Law and Euler's equation require: translational: mass rotational: center of gravity center of gravity mass moment of inertia m PG IG . θ 3 . α 2 . α 3 . θ 4 . ω 4 . 4-Bar Linkage Inverse Dynamics: Given the motion θ 2 . α 4 and internal forces F ij .

mass m is the proportionality constant. Newton’s 1st Law. Mass is measure of translational inertia – resistance to change in motion. 2 . Mass is also measure of storage 1 of translational kinetic energy KET = mv 2 .115 Mass. Mass Moment of Inertia ∑ F = m AG Translational: mass Rotational: ∑MG = IGα center of gravity center of gravity mass moment of inertia Mass In Newton’s 2nd Law ∑ F = ma . Center of Gravity.

IG: System of particles General rigid body Rectangular rigid body Mass calculation: System of particles: General rigid body: Rectangular rigid body: . CG.116 Examples for m.

G) Also called center of mass.117 Center of Gravity (CG. mass center. centroid CG calculation: System of particles: General rigid body: .

the CG is the geometric center. .118 Rectangular rigid body: Using an XY coordinate frame centered at the geometric center. X= = = ∫ xdm x ∫ dm ∫ xdV b/2 Y= = = ∫ ydm y ∫ dm m∫ y ρ ρ ρ mx ydV h/2 m ∫− b / 2 ρ th b / 2 = ∫−b / 2 xdx m = xthdx ytbdy m ∫− h / 2 ρ tb h / 2 ydy = m ∫− h / 2 = ρ ρ th x 2 m 2 b/2 ρ tb y 2 h/2 −b / 2 m 2 −h / 2 b2  =  4 − 4 =0 2m   ρ th  b 2 h2  = − =0 2m  4 4    ρ tb  h 2 For a homogeneous. regular geometric body.

Also. .119 Mass Moment of Inertia (IG) is not the same as Area moment of inertia (IG) for beam bending: I Ax = ∫ y 2 dA y I Ay = ∫ x 2 dA x Units: I A ≡ m 4 Mass Moment of Inertia (IG) In Euler’s equation ∑ M G = I GZ α . 2 Units: IG ≡ kgm2 . I is measure of rotational inertia – resistance to change in motion. I is also measure of storage 1 of rotational kinetic energy KER = I Gω 2 . Newton’s 1st Law. I is the proportionality constant. it is a measure of how hard it is to accelerate in rotation about certain axes.

General rigid body: What is the only term that matters for XY planar motion? In the example shown above: I ZZG > I YYG > I XXG also I ZZ > I ZZG .120 Mass Moment of Inertia IG calculation: System of particles: where ri is the scalar perpendicular distance from the axis to the ith particle. second moment will be doubled about the CG. With squaring. all terms will be positive. no there can be no canceling like for CG. If first moment is balanced.

121 Rectangular rigid body: Using an XY coordinate frame centered at the CG. I ZZG = body (x 2 + y 2 )dm = ∫−bb/ /22 ∫−hh/ /22 (x 2 + y 2 )ρtdxdy ∫  3 h/2  y b/2  2  I ZZG = ρt ∫ x y+ dx −b / 2  3   −h / 2   3 − h3   b / 2  2 h − h  1  h x  −  dx = ρt ∫ +  − −b / 2   2 2  3 8 8     b/2   3 h3  h3 x  hx  + dx = ρt  +  12  12  3   −b / 2   I ZZG = ρt ∫ b/2  2  hx −b / 2    h  b3 − b3  h3  b − b   +  − I ZZG = ρt   −   3 8 8  12  2 2        b3h bh3  ρtbh 2 = = ρt  + b + h2  12  12  12  m 2 b + h2 12 ( ) I ZZG = ( ) (because m = ρV = ρtbh ) .

How do we find mass moments of inertia in the real-world? • look up in tables • CAD package such as SolidEdge .122 Units: mass times distance squared. Checks with result given in the textbook. kgm 2 .

. due to the term md 2 .123 Parallel Axis Theorem The mass moment of inertia through the CG is related to mass moments of inertia of parallel axes through different points as follows: where d is the scalar distance separating the axis of interest from the axis through the CG. any I ZZ must be greater. which is always positive. Notice I ZZG is a small as it can get. Parallel axis theorem example: Rectangular rigid body: I ZZ  b2 h2  m 2 2 = (b + h ) + m  +  12 4 4  b2 b2 h2 h2  = m + + +   12 4 12 4   b2 h2  = m +  3 3 m = (b2 + h2 ) 3 Agrees with result given in dynamics textbooks.

external forces and moments. and Acceleration Analyses must first be complete. Single Rotating Link Inverse Dynamics Analysis Step 1. and the desired mechanism motion. Position. Draw the Diagrams: Physical Dynamics Diagram: Free Body Diagram (FBD): . Step 2.124 Single Rotating Link Inverse Dynamics Generic Mechanism Inverse Dynamics Analysis Statement: Given the mechanism. Velocity. find the required driving force (or torque) and internal joint forces.

State the problem: Step 4. Newton's 2nd Law: Euler's Equation: Count # of unknowns and # of equations: . Derive the Newton-Euler Dynamics Equations.125 Step 3.

126 Step 5. Written in matrix form: . Derive XYZ scalar equations from the vector equations.

which is the force/moment reaction on the ground link due to the mechanism. Calculate Shaking Force and Moment After the inverse dynamics problem is solved. motion. and external loads: . we can calculate the vector shaking force and moment. the first two equations are decoupled and the solution is straight-forward: Step 7. we don’t need matrix solution. Solve for the unknowns Actually.127 Step 6.

 AGX  AG =  =  AGY   FEX  FE =  = FEY   r  r12 =  12 X  =  r12Y  r  r E =  EX  =  rEY  I GZ = . Now. how do we calculate the various terms that appear in the dynamics equations? These all must be derived from given information.128 Terms for the inverse dynamics equations The inverse dynamics problem has been solved analytically for the single rotating link.

5 m I GZ = 0. calculate F12 X . ω = 100 rad/s. M S for this instant (snapshot). h = 0.250 0. FE=150 N.5      − 8510 F S = F 21 = − F 12 =  N  4980  ˆ M S = −τ = −66.1 m. Nm  τ   66. α = 0. m = 2 kg. τ and F S . given this link.5    F12 X   8510       F12Y  = − 4980 N.129 Single rotating link inverse dynamics example: Given: L = 1 m. F12Y . 0 0  F12 X   8510   1      0 1 0  F12Y  = − 4980        0. ME=0 Nm.5k Nm . Calculated terms: r12 = rE = 0.433 1  τ   37. motion. φ E = 0 (constant relative to horizontal).17 kgm 2 AGx = 4330 m AGy = −2500 s 2 Snapshot Analysis (one input angle) At θ = 150 . and external force.

for the same example from the previous page.O.) Analysis A more meaningful result from inverse dynamics analysis is to report the unknowns for the entire range of mechanism motion. This shows the torque that must be supplied by an external DC servomotor to cause the specified motion. assuming the given ω is constant.R. red) for all 0 ≤ θ ≤ 360 . Tau (red) with average (green) and rms (blue) 150 100 50 τ (Nm) 0 -50 -100 -150 0 50 100 150 200 θ (deg) 250 300 350 . Also plotted is the average torque (green) τAVG = 0 and the rootmean-square (RMS) torque value (blue) τRMS = 106.130 Full-Range-Of-Motion (F.1 Nm.M. The plot below gives the required driving torque τ (Nm.

131

The plots below give the Shaking Force F S and CG translational acceleration results, respectively, for all 0 ≤ θ ≤ 360 . In both plots, the X components are red and the Y green.

1

x 10

4

Shaking Force, X (red) and Y (green)

0.5 FSX and FSY (N)

0

-0.5

-1 0

50

100

150

200 θ (deg)

250

300

350

132

CG Acceleration, X (red) and Y (green) 5000

AGX and AGY (m/s )

2

0

-5000 0

50

100

150

200 θ (deg)

250

300

350

The Shaking Moment M S is merely the negative of the driving torque τ plot shown previously and hence is not shown separately. Is the static loading (mg) significant?

133

Four-Bar Mechanism Inverse Dynamics Generic Mechanism Inverse Dynamics Analysis Statement: Given the mechanism, external forces and moments, and the desired mechanism motion, find the required driving force (or torque) and internal joint forces. Four-Bar Mechanism Inverse Dynamics Analysis

First, can we simplify and solve the problem link-by-link, like the single rotating link? Count # of unknowns and # of equations:

Step 1. Position, Velocity, and Acceleration Analyses must first be complete. Step 2. Draw the Diagrams: Physical Dynamics Diagram:

State the problem: .134 Free Body Diagrams (FBDs): F ij : r ij : Step 3.

135 Step 4. Derive the Newton-Euler Dynamics Equations. Newton's 2nd Law: Euler's Equation: Count # of unknowns and # of equations: .

Derive XYZ scalar equations from the vector equations.136 Step 5. .

. and given external forces and moments.137 Write these equations in matrix/vector form:  −1  0   r12Y   0  0   0  0   0  0  0 −1 − r12 X 0 0 0 0 0 0 1 0 − r32Y −1 0 r23Y 0 0 0 0 1 r32 X 0 −1 − r23 X 0 0 0 0 0 0 1 0 − r43Y −1 0 r34Y 0 0 0 0 1 r43 X 0 −1 − r34 X 0 0 0 0 0 0 1 0 −r14Y 0 0 0 0 0 0 0 1 r14 X m2 AG 2 X 0   F21 X     F   m2 ( AG 2Y + g ) 0   21Y     IG 2 Zα 2   F32 X   1     m3 AG 3 X − FE 3 x 0   F32Y         F43 X  =  m3 ( AG 3Y + g ) − FE 3 y 0     I α −r F +r F −M  0  F43Y E3    G 3Z 3 E 3 x E 3 y E 3 y E 3 x   F   m4 AG 4 X − FE 4 x 0  14 X     m4 ( AG 4Y + g ) − FE 4 y 0   F14Y      τ  I α − r F + r F − M  0  2   G 4 Z 4 E 4 x E 4 y E 4 y E 4 x E4  [ A]{v} = {b} Coefficient matrix [A] dependent on geometry (kinematics solutions). gravity. RHS {b} dependent on inertial terms.

using Gaussian elimination is more efficient and robust.138 Step 6. which is the force/moment reaction on the ground link due to the motion: . we can calculate the vector shaking force and moment. Solution to internal forces and input torque are contained in the components of {v}. Solve for the unknowns Simultaneous matrix solution: {v} = [ A]−1{b} Actually. Calculate Shaking Force and Moment After the basic inverse dynamics problem is solved. Step 7.

how do we calculate the various terms that appear in the dynamics equations? These all must be derived from given information. P11-2.139 Details for the general four-bar mechanism model The inverse dynamics problem has been derived analytically for the four-bar mechanism. Now. Let us do link 3 terms (next page). Here is the general link 3 diagram for these derivations: . See Fig.

140 Link 3 details: r23 X  r 23 =  = r23Y   r  r 43 =  43 X  =  r43Y   AG 3 X  A G3 =  = AG 3Y    FP3 X  = F P3 =  FP3Y     R P3 X  = R P3 =  R P3Y    M E 3 = given .

25 0.2 0. 0.05 -0. of the actual rectangular link for inverse dynamics.1 0 0.05 0 -0.15 Y (m) 0.5”. The triangle tip we have been using all along in Term Example 1 is actually the CG.3 .2 0.141 Figure for Term Example 1 Inverse Dynamics Example starting on the next page: The coupler link 3 is a rectangle of dimensions 8” x 6” x 0.1 -0.3 0.1 0.1 X (m) 0.

as shown on the previous page. ω3 = −8. the driving torque τ 2 . m4 = 0. with mass density ρ = 0. RG3 = 0. This is the open branch of the position.9 . link 3 has rectangular dimensions 8 by 6 by 0. I G 3 Z = 1.203. r4 = 0. Snapshot Analysis (one input angle) At θ 2 = 30 . with one crucial difference: the input angular velocity was too low for interesting dynamics.73 rad/s. given this mechanism and motion. θ 2 = 30 . and θ1 = 10.284. α 3 = 8.4 rad/s2.178.142 Four-bar mechanism inverse dynamics example: This is the mechanism from Term Example 1 (open branch). calculate the four vector internal joint forces.50 by ri (in).036 kg and I G 2 Z = 7. ω4 = −3. ω2 = 20 .5 (in). θ 4 = 121. I G 4 Z = 9. α 4 = 244.09 .8 .03 lbm/in3.076.5 × 10−5 kgm2.015 .75 by 0. θ3 = 53.9 × 10−6 .8 × 10−3 . Links 2 and 4 have rectangular dimensions 0.3 .327 . Given r1 = 0. r2 = 0. and acceleration example (Term Example 1).7 . .127 m. All moving links are wood. All external forces and moments are zero but gravity is included. m3 = 0. and the shaking force and moment F S . δ 3 = 36. so I changed it from ω2 = π to ω2 = 20 rad/s. α 2 = 0 .65 . velocity. M S for this instant (snapshot). r3 = 0. The calculated inertia parameters are m2 = 0.

60       F14 y   5.047 0.56 ˆ M S = −1.08   21 y     F32 x   5.52   τ 2   − 0.047  0  F21x  − 0.80  FS =   N 4.202  0  F21 y   0.037 0.99       F32 y   10.638     0  F14 y   − 0.143 0 1 0 0 0 0 0  −1  0 −1 0 1 0 0 0 0  − 0.015      0  F14 x   − 0.033 − 0.076 0.0233      Solution by Gaussian elimination or: {v} = [ A]−1{b} Snapshot Answer:  F21x   6.96 N.43     9.076 0.002 − 0.61    43 y    F14 x  − 3.11  {v} =  F43 x  = − 2.122 0 0 0  0  0 −1 0 0 0 0 1 0  0 0 0 0 −1 0 1  0  0 0 0 0 0.019 0.033 0 0 0 0  0 −1 0 1 0 0 0  0  0 0 0 −1 0 1 0 0  − 0.095 0  τ 2   0.019 0.955      0  F43 x  = − 4. Nm      F   5.127 − 0.034      1  F32 x   0      0  F32 y   − 8.497   0  F43 y   0.20   F   10.68k Nm .

M.4 (Nm) 0.2 0 τ 2 -0. Also plotted is the average torque (green) τ2AVG = 0 and the root-mean-square torque value (blue) τ2RMS = 0. This plot shows the torque (red) that must be supplied in all configurations by an external DC servomotor to cause the specified motion.) Analysis A more meaningful result from inverse dynamics analysis is to report the unknowns for the entire range of mechanism motion. for the Term Example 1 mechanism.2 -0.R. The plot below gives the required driving torque τ 2 (Nm) for all 0 ≤ θ 2 ≤ 360 .O.8 0.6 0. assuming the given ω 2 = 20 rad/s is constant (Remember: this has been changed from the ω 2 = π rad/s in the kinematics examples!).36 Nm.144 Full-Range-Of-Motion (F. Tau2 (red) with average (green) and RMS (blue) 0.4 0 50 100 150 200 θ 2 (deg) 250 300 350 .

for all 0 ≤ θ 2 ≤ 360 . X (red) and Y (green) 10 5 FSX and FSY (N) 0 -5 -10 0 50 100 150 200 θ 2 (deg) 250 300 350 . the X component is red and the Y green. Shaking Force. respectively. In the force plot.145 The plots below give the shaking force F S (N) and shaking moment M S (Nm) results.

5 1 MS (Nm) 0. there is only the Z component: Shaking Moment 2 1.5 -2 0 50 100 150 200 θ 2 (deg) 250 300 350 .5 0 -0.5 -1 -1.146 In the shaking moment plot.

links 2 and 3 are handled identically! Step 1. In fact. Step 2. Velocity. Position.147 Slider-Crank Mechanism Inverse Dynamics This problem is very similar to the four-bar mechanism inverse dynamics problem. Draw the Diagrams: Physical Dynamics Diagram: Free Body Diagrams (FBDs): F ij : internal force of link i acting on link j r ij : moment arm pointing to link i from the CG of link j . and Acceleration Analyses must first be complete.

links 2 and 3 are identical so let us focus on link 4.148 Step 3. Again. Derive the Newton-Euler Dynamics Equations. There are two kinematic constraints on the slider: Newton's 2nd Law: Euler's Equation: Count # of unknowns and # of equations: We need an additional equation: . the slider. State the problem: Step 4.

Derive XYZ scalar equations from the vector equations and beam these equations into matrix/vector form.149 Step 5.  −1  0   r12Y   0  0   0  0   0  0 −1 − r12 X 0 0 0 0 0 1 0 − r32Y −1 0 r23Y 0 0 0 1 r32 X 0 −1 −r23 X 0 0 0 0 0 1 0 − r43Y −1 0 0 0 0 0 1 r43 X 0 −1 m2 AG 2 X  0   F21 X    F   m2 ( AG 2Y + g ) 0 0   21Y     IG 2 Zα 2 0 1   F32 X       m3 AG 3 X − FE 3 x 0 0   F32Y    =    F43 X   m3 ( AG 3Y + g ) − FE 3 y 0 0      0 0   F43Y   I G 3 Z α 3 − rE 3 x FE 3 y + rE 3 y FE 3 x − M E 3      m4 AG 4 X − FE 4 x ± µ 0   F14Y       m4 g − FE 4 y 1 0  τ 2      0 [ A]{v} = {b} Coefficient matrix [A] dependent on geometry (kinematics solutions). also eliminate one equation ( ∑ M G 4 = I G 4 Zα 4 ). Solution to internal forces and input torque contained in the components of {v}. Step 6. Solve for the unknowns Simultaneous matrix solution: {v} = [ A]−1{b} Actually. . Substitute the friction constraint to eliminate one unknown (F14X). using Gaussian elimination is more efficient and robust. RHS {b} dependent on inertial and statics terms. Always choose proper sign of µ to be opposite to the current x direction.

05 Y (m) 0 -0.1 X (m) 0.1 -0.15 -0.2 -0. Calculate Shaking Force and Moment After the basic inverse dynamics problem is solved.2 0.150 Step 7.3 . 0. we can calculate the vector shaking force and moment.15 0.25 -0.1 0. with zero input angle θ2.05 -0.1 0 0.2 0. Figure for example starting on the next page: The slider-crank mechanism is shown at the starting (and ending) position. which is the force/moment reaction on the ground link due to the motion.

θ3 = 7.03 (lbm/in3). and ω2 = π / 2 .75 by 0. and acceleration example (Term Example 2).076 m.418e-004 (kgm2).102. link 4 has rectangular dimensions 0. Snapshot Analysis (one input angle) At θ 2 = 30 . m4 = 0. Links 2 and 3 have rectangular dimensions 0. α 2 = 0 . . M S for this instant (snapshot). There is a constant external force of 1 N acting at the center of the piston.686 rad/s.2 . ω3 = −0. All moving links are wood.151 Slider-crank mechanism inverse dynamics example: This is the mechanism from Term Example 2 (right branch only).50 by 3 (in). x = 0.020. velocity.50 by ri (in). The calculated inertia parameters are m2 = 0. We assume µ = 0. (kg) and IG2Z = 1.2 (coefficient of friction between piston and wall). IG3Z = 1. and θ 2 = 30 . m3 = 0. h = 0. calculate the four vector internal joint forces. α 3 = 0.290 m.062 m/s. in this case keeping the low input angular velocity ω2 = π 2 rad/s so the previous snapshots and full-range-of-motion results still apply. directed horizontally to the left. given this mechanism and motion.75 by 0. r3 = 0. with mass density ρ = 0.203. the driving torque τ 2 .329 m/s2.015. gravity is included but all other external forces and moments are zero.681 rad/s2. This is the right branch of the position.041. Given r2 = 0.819e-005. x = −0. x = −0. and the shaking force and moment F S .

0001  0      0 0 0 0 0 0.101 -0.150  −1  0     Solution by Gaussian elimination or: {v} = [ A]−1{b} Snapshot Answer:  F21 X  -0.044 0 0 0 1   F32 X   0       0 −1 0 1 0 0 0   F32Y   -0.025 0.002   −1  0 −1 0 1 0 0 0 0   F21Y   0.318  {v} =   =   (N.101 0 0   F43Y  0.011   0   =  0 0 0 −1 0 1 0 0   F43 X   0.748 ˆ M S = −0.081       F14Y   0.053k (Nm) .2 0   F14Y   0.152 0 1 0 0 0 0 0   F21 X   -0.517   21Y     F32 X  -0.044 -0.013 0.025 0. Nm) F43 X  -0.011 −0.982  FS =   (N)  −0.231       τ 2   -0.199        -0.938       F32Y  -0.935   F  -0.013 0.398    0 -0.949    F43Y   0.995  −1     0 0 0 0 1 0   τ 2   0.

2 0.O.05 -0. right branch only.) Analysis A more meaningful result from inverse dynamics analysis is to report the unknowns for the entire range of mechanism motion.1 0 50 100 150 200 θ2 (deg) 250 300 350 .153 Full-Range-Of-Motion (F.15 0. for the Term Example 2 slider-crank mechanism. The plot below gives the required driving torque τ 2 (Nm) for all 0 ≤ θ 2 ≤ 360 .086 Nm.002 and the root-mean-square torque value (blue) τ2RMS = 0.M. Also plotted is the average torque (green) τ2AVG = -0.05 0 -0.R.1 τ2 (Nm) 0. Actuator Torque 0. assuming the given ω2 = π 2 rad/s is constant. This plot shows the torque (red) that must be supplied in all configurations by an external DC servomotor to cause the specified motion.

In the force plot.95 -1 -1.7 -0.75 -0. Shaking Force. respectively.8 FSX and FSY (N) -0. for all 0 ≤ θ 2 ≤ 360 . X (red) and Y (green) -0. the X component is red and the Y green.154 The plots below give the shaking force F S (N) and shaking moment M S (Nm) results.9 -0.85 -0.05 0 50 100 150 200 θ2 (deg) 250 300 350 .

there is only the Z component: Shaking Moment 0.06 0 50 100 150 200 θ2 (deg) 250 300 350 .04 -0.02 0 -0.08 0.04 MS (Nm) 0.155 In the shaking moment plot.06 0.02 -0.

Cam Classification: Disk cams with followers .156 Cam Introduction Chapter 8 Applications Compared to linkages. but much more expensive and difficult to produce. easier to design desired motion with cams.

usually. . Recall a cam joint has two-dof. 4.157 Degrees of Freedom rolling and sliding. For example. S = f (θ ) φ = f (θ ) Cam input angle is θ . allows both Function Generation The output parameter is a continuous function of the input parameter. we can satisfy function generation at infinite points. or 5. a 4-bar linkage: θ 4 = f (θ 2 ) With a cam and follower mechanism. output is S for reciprocating (translating) and φ (rotating) for oscillating follower. With linkages. we can only satisfy a function exactly at a finite number of points: 3.

Classification: When the motion transitions between different motion functions. determine the position. and dynamics behavior. design the cam. The first step is to determine a “smooth” cam follower motion profile. we have been mostly concerned with mechanism analysis: given a mechanism design and its input parameters.158 Cam Motion Profiles Up to this point. With cams we must consider mechanism synthesis for the first time: given the motion requirements (follower motion and timing with input angle). . acceleration. velocity. we must ensure “smooth” motion.

high wear. etc. the resulting dynamic performance will be acceptable for high-speed cam/follower operation. Cyclical impulse hammering when acceleration is not continuous. If not.159 Fundamental Law of Cam Design: Which means: If the Fundamental law of Cam Design is satisfied. there will be performance degradation due to noise. vibrations. .

. Therefore. time or vs. but the acceleration itself must be continuous. the slopes of the position and velocity curves must match across all junctions. Check out Examples 8-1 (terrible) 8-2 (bad) 8-3 (acceptable) Slope of a function is the value of its derivative at a point. time to graphically see if the Fundamental Law of Cam Design is satisfied for candidate curves. We can plot vs.160 S V A J Diagrams In synthesis. we are only given total motion range and perhaps some timing requirements. Position is automatically matched by shifting axes. Draw S V A J diagrams vs. input cam angle θ (assuming constant angular velocity. The slope of the acceleration can be discontinuous (leading to finite jumps in jerk). θ = ωt ). for continuous velocity and acceleration curves. It is the engineer’s job to determine the position curves and to match the velocity and acceleration across junctions.

. slope of S must match across junctions. Then to put the whole thing together.161 Generic Cam Follower Motion Profile figure: Define each separate function so the value is zero at the initial angle. which is zero. just do it . just shift the θ and S axes. Match A: slope of V must match across junctions.shift S axes. Match S: Match V: easy.

Match V at junction B: Try to match A at junction B: Plot on next page. . 1 2 S: f1 (θ1 ) = A0θ1 f 2 (θ 2 ) = V0θ 2 f 3 (θ 3 ) = 0 2 V: A: J: Match S at junction B: Just shift axis up. Specify Parabolic (constant acceleration) to Straight Line (constant velocity) rise.162 Cam Follower Motion Profile Examples Example 1 rise – dwell portion. followed by a dwell.

163 Example 1 Plots .

Specify Half-Cycloidal function (sinusoidal in cam angle) to Straight Line (constant velocity) rise. S: θ πθ  1 f1 (θ1 ) = L1  1 − sin 1  f 2 (θ 2 ) = V0θ 2  β1   β1 π  V: A: J: Match S at junction B: Just shift axis up. .164 Cam Follower Motion Profile Examples Example 2 Fix rise portion only. Match V at junction B: Match A at junction B: Plot on next page.

165 Half Cycloid 2 S (m) 1 20 40 60 80 V (m/deg) 0 0 0.05 J (m/deg ) A (m/deg2) 0 -4 x 0 10 5 20 40 60 80 0 -5 x 0 10 2 0 -2 0 20 40 60 80 3 20 40 θ (deg) 60 80 .

This will rise all the way to meet a dwell smoothly. θ 1 2πθ1  f1 (θ1 ) = L1  1 − sin β1   β1 2π  L1  2πθ1  1 − cos β1  β1    S: f 2 (θ 2 ) = 0 V: v1 (θ1 ) = v2 (θ 2 ) = 0 A: a1 (θ1 ) = 2π L1  2πθ1  sin 2  β1  β1   a2 (θ 2 ) = 0 J: 4π 2 L1  2πθ1  j1 (θ1 ) = cos β13  β1    j2 (θ 2 ) = 0 Plot on next page. . it satisfies the Fundamental Law of Cam Design.166 Cam Follower Motion Profile Examples Example 3 Specify Full-Cycloidal function (sinusoidal in cam angle).

167 Full Cycloid 2 S (m) 1 20 40 60 80 V (m/deg) 0 0 0.05 J (m/deg ) A (m/deg ) 0 -3 x 0 10 2 0 -2 -4 x 0 10 2 0 -2 0 2 20 40 60 80 20 40 60 80 3 20 40 θ (deg) 60 80 .

V. DCRFFF Figure: . Is it as simple as polar plotting of S = f (θ ) vs.e. Given the motion profile. cam angle θ ? We will use kinematic inversion to simplify the synthesis. determine the cam contour. A curves.168 Analytical Cam Synthesis Disk Cam with Radial Flat-Faced Follower Assume a valid cam motion profile has been designed according to the Fundamental Law of Cam Design. we now have continuous S. i.

double the maximum of L from above. and S = f (θ ) is the given motion profile. a design variable.y and the cam angle through geometry: Notice that: To calculate the follower flat-face length. The radius R and the flat-face length L can be related to the contact point x. the radius R out to the flat-faced follower (not to the point of contact (x. Doubled because by symmetry the contact point will change to the other side at θ = 180 . .y)) is: where C is the minimum cam radius.169 As seen in the figure.

R. which means [ A]−1 = [ A]T . it is machined with θ. L coordinates. y. The solution is: . In matrix form: The coefficient matrix [A] is orthonormal. we must solve the relationships for x.170 To summarize thus far: This is sufficient to manufacture the cam. If we want to know the cam contour in Cartesian coordinates.

The condition is that for a finite ∆θ . to avoid cusps on the entire cam contour. y: dx dy = =0 will cause a cusp.171 Minimum Cam radius to Avoid Cusps A cusp is a point in the cam. this is to be avoided for good motion. dθ dθ dx df df d2 f = −(C + f (θ )) sin θ + cos θ − cos θ − sin θ 2 dθ dθ dθ dθ dy df df d2 f = (C + f (θ )) cos θ + sin θ − sin θ + cos θ 2 dθ dθ dθ dθ  dx d2 f   sin θ = − C + f (θ ) +  dθ dθ 2     d2 f  dy   cos θ = C + f (θ ) + dθ  dθ 2    dx dy = = 0 simultaneously only when: dθ dθ d2 f =0 C + f (θ ) + 2 dθ Therefore. . d2 f >0 C + f (θ ) + 2 dθ Note C is always positive and f (θ ) starts and ends at zero and never goes negative. there is no change in x. or actually undercut.

and acceleration curves are continuous. 50 S 0 0 50 0 V 50 100 150 200 250 300 350 -50 0 100 0 -100 0 500 0 -500 0 J A 50 100 150 200 250 300 350 50 100 150 200 250 300 350 50 100 150 θ 200 250 300 350 . The cam motion profile associated with this specification is shown below. Clearly.172 Disk Cam with Radial Flat-Faced Follower Design Example Specify a full-cycloidal rise (total lift 50 mm). Each of these four motion steps occurs for 90 deg of cam shaft rotation. this satisfies the Fundamental Law of Cam Design because the position. and then a low dwell. but it remains finite over all cam angles. followed by a high dwell. The jerk is not continuous. a full-cycloidal return (total fall 50 mm). velocity.

150 100 Y (mm) 50 0 -50 -100 -150 -100 -50 0 X (mm) 50 100 . the resulting cam contour is shown below.173 Choosing a minimum cam radius of C = 100 mm).

250 200 C+S+A 150 100 50 0 0 50 100 150 θ 200 250 300 350 .174 Let us check the cusp avoidance plot. so this cam design is acceptable. this inequality is satisfied for the entire range of motion. we require that: C + S (θ ) + A(θ ) = C + f (θ ) + d2 f dθ 2 >0 As seen in the plot below. To avoid cusps in this cam.

Can openers to Aircraft carriers. Related mechanisms . torque. toys. direction. etc. Change in angular velocity.belt and chain drives. Applications: Gear Classification External Spur Gears Internal Spur Gears . Gears used in electromechanical systems. vehicles.175 Gear Introduction Transfer motion between rotating shafts in machinery.

176 Rack & Pinion Bevel gears Helical (Parallel Shaft) Helical (Crossed Shaft) Herringbone Gears Gear Train .

177 Automotive Differential Planetary Gear Train Worm and Gear Harmonic Gearing .

co.uk/ “The harmonic gear allows high reduction ratios with concentric shafts and with very low backlash and vibration. The external flexible spline has two fewer teeth than the internal circular spline.” .178 Harmonic Gearing Taken from: http://www. which reduces backlash. It is based on a very simple construction utilising metals elasto-mechanical property. placed inside a circular thick-walled rigid ring machined with internal splines.” “Harmonic drive transmissions are noted for their ability to reduce backlash in a motion control system. The two contact points rotate at a speed governed be the difference in the number of teeth on the two splines This method basically preloads the teeth.roymech. An elliptical cam enclosed in an antifriction ball bearing assembly is mounted inside the flexible cup and forces the flexible cup splines to push deeply into the rigid ring at two opposite points while rotating. How they work is through the use of a thin-walled flexible cup with external splines on it lip.

Gear ratio is a measure of the degree of reduction and increase. Arc lengths are equal. Simple spur gear pair: DOF: A gear joint is like a cam joint. Like two cylinders rolling without sliding. A robot joint needs lower rotation speed but high torque. Gear 1 is input. A gear train can accomplish both objectives – reduce speed and increase torque. two-dof.179 Gear Ratio Common electric motors have high speed but low torque. teeth in contact allow rolling and sliding. . gear 2 is output.

180 Define gear ratio n: Radii inversely proportional to angular motion. tangential velocities are equal. For standard spur gears. for accelerating up to (or down from) constant angular velocities: . Most gear applications have constant angular velocities. the radii are directly proportional to the number of teeth: For relating angular velocities.

angular velocities. interface forces are equal. and shaft torques. . diameter. Radii directly proportional to shaft torques Summary: The ratio of the number of teeth is directly proportional to the radii. and angular accelerations.181 For relating shaft torques. The ratio of the number of teeth is inverse proportional to the shaft angles.

182 If n > 1: ω 2 < ω1 τ 2 > τ1 Output has reduced speed Output has increased torque This is the electric motor / robot joint case. If n < 1 : ω 2 > ω1 τ 2 < τ1 Output has increased speed Output has reduced torque If n = 1: ω 2 = ω1 τ 2 = τ1 Output speed and torque unchanged direction reverses (external spur gears) .

mountain bike gearing generally: • increases angular velocity • decreases torque .58 0.54 0.68 0.64 0.62 Front 38 0.79 28 0.29 0.57 0.93 1.50 0.46 0.07 Unlike electric motor example.33 0. Mountain Bike Transmission N OUT N R ω F τ R = = = N IN NF ωR τ F Gear Ratios: n= Schwinn Sierra Rear 14 16 18 22 26 30 48 0.38 0. Gear toy 2.37 0.42 0.183 Gear ratio examples 1.78 0.47 0.

the effective BikeE high and low gear ratios are: Stiff: 0.50 0.44 0.184 Cannondale M400 Rear 11 12 14 16 18 21 24 28 32 44 0.32 0.53 instead of 1.70 0.38 0.54 0.30 0.63 0.34 0.73 Front 32 0.34 0.25 0.55 0.54 0.88 1.66 0.41 0.25 Granny 1.64 recumbent Rear Rear hub 1:1 0.01 1.82 0.48 0.64 0.38 0.27 1.62 0.32 0.41 0.53 0.00 22 0.56 0.55 0.2913:1 0.95 1.82 0.76 0.32 instead of 0.48 0.27 0.09 1.45 BikeE Front 34 11 13 15 18 21 24 28 1.75 0.7:1 0.46 0. .64 0.25 0.50 0.18 But considering the difference in wheel sizes (26” Cannondale. 20” BikeE).88 1.18 Original front was 46 teeth – changed for more granny gear.44 0.73 0.36 0.

nGT ω IN = ω OUT Example: We can find the overall gear ratio by canceling neighboring intermediate angular velocities: Each term in the above product may be replaced by its known number of teeth ratio: . All intermediate gears are first the driven gear and then the driving gear as we proceed from left to right. Leftmost is driving gear.185 Gear Trains and Gear Standardization Simple Gear Trains Mesh any number of spur gears. Rightmost is the output gear. Let us calculate the overall gear ratio.

186 All intermediate ratios cancel. but do change direction! We should have included sign: So. Number of teeth effect cancels out. the intermediate gears are idlers. for external spur gear trains: Odd # of gears: Output same direction as input Even # of gears: Output opposite direction as input . so: We could have done the same with pitch radii instead of number of teeth because they are in direct proportion: So.

we use the equation: But now the gears rigidly attached to the same shaft have the same angular velocity ratio. Again. where the driving and driven gears are distinct. let us calculate the overall gear ratio. so: . See the figure.187 Different case: Mesh any number of spur gears. nGT = ω IN ω OUT Example: Again. because each pair is rigidly attached to the same shaft.

we must consider direction: So. for external spur gear trains: Odd # of pairs: Output opposite direction as input Even # of pairs: Output same direction as input .188 General formula: Again.

189 Involute Spur Gear Details and Standardization Rolling Cylinders Mating spur gears are based on two pitch circles rolling without slip. Fundamental Law of Gearing: From our study of linkage velocity. Velocity ratios in a linkage vary wildly over the range of motion. we know this is no easy feat. Velocity Ratio Torque Ratio (Mechanical Advantage) The author’s velocity ratio is the inverse of our gear ratio definition and his torque ratio is the same as our gear ratio. These are fictitious circles. . The actual gear teeth both roll and slide (two-dof joint). you cannot look on a gear to see them.

etc.190 Involute Function Standard spur gears have an involute tooth shape. thermal expansion.). wear. The involute is a curve generated by unwrapping a taut sting from a circle: . the angular velocity ratio will still be constant to satisfy the Fundamental Law of Gearing. If the gears’ center distance is not perfect (tolerances.

pitch circle. Point of contact slides along this line. Base circle. Similar angle is defined for cams and followers. pure rolling in contact Contact point between the two pitch circles Pressure Angle: Angle between the common normal (also called axis of transmission) of the two meshing teeth and the velocity of the pitch point (tangent to both pitch circles). pressure angle relationship: .191 Base Circle: Pitch Circle: Pitch Point: Involute starts from this circle Fictitious circle.

Only one or two teeth are in contact at any one time. . many teeth are in contact at any one time (higher gear ratio in a smaller package). End of contact is when tip of driving gear (pinion) tooth intersects the axis of transmission.192 Length of contact along axis of transmission. For harmonic gearing. Beginning of contact is when tip of driven gear tooth intersects the axis of transmission.

193 Increasing center distance increases the pressure angle. but doesn’t change the base circles (of course). How is this possible? Relationship from last page: . does not affect angular velocity ratio. Thanks to the involute tooth shape. increases the pitch circle radii.

194 Backlash Clearance. Non-linear effect in robots. Example: robot joints which must be driven both directions. However. etc. nothing happens until the backlash is moved. the backlash is less predictable! Figure: . Changing direction.bad for dynamics. In space however. and then impact . On earth gravity tends to load the backlash for predictable effects. All real gears must have some backlash due to tolerances. must minimize backlash for smooth operation. Distance between mating teeth measured along the pitch circle circumference. thermal expansion. wear.

Diametral Pitch: Module: Module is the metric version of diametral pitch. For two spur gears to mesh. 2) same diametral pitch. must have: 1) the same pressure angle.195 Gear Standardization To allow interchangeability in manufacturing and to allow meshing of different size gears (radii and number of teeth) to achieve desired gear ratios. Not interchangeable with US gears because different tooth proportion standards! Circular Pitch: . and 3) be made with standard tooth proportions (Table 9-1. 441). p.

441. Face width is thickness of tooth and gear (mating widths needn’t be the same). Clearance is radial distance from Bottom Land to mating gear Top Land (kinda like radial backlash).196 Standard involute tooth proportions. Tooth thickness is the circumferential length of each tooth. Related to the circular pitch and backlash by: . see Table 9-1.. Dedendum is radial distance from pitch circle to Bottom Land of tooth (not to base circle). p. Addendum is radial distance from pitch circle to Top Land of tooth.