AME 352 STATIC FORCE ANALYSIS

12 STATIC FORCE ANALYSIS

This chapter applies the fundamental concepts of force analysis that was discussed in Chapter 11 to perform
static force analysis of mechanisms. The discussion reviews both graphical and analytical methods of force
analysis on stationary systems. Force analysis of mechanisms in motion, known as dynamic force analysis,
will be discussed in Chapter 13.

12.1 STATIC EQUILIBRIUM EQUATIONS

A body is said to be in static equilibrium if under a set of applied forces and torques its translational (linear)
and rotational accelerations are zeros (a body could be stationary or in motion with a constant linear
velocity).
Planar static equilibrium equations for a single body that is acted
upon by forces and torques are expressed as C FB
FC

⎧⎪ ∑ Fi( x ) = 0 T
B
∑ i
F = 0 ⇒ ⎨ (12.1) (Force equation)
⎪⎩∑ Fi( y ) = 0
∑ Tj + ∑ Mi = 0 (12.2) (Moment equation)
A
FA

Equation (12.1) represents the sum of all the forces acting on the link, and Eq. (12.2) represents the sum of
all the torques, T j , and the moments, M i , about any reference point caused by all the forces that act on the
body.

Note: Since in planar systems the moments of

forces are along the z axis, we use directional
magnitudes of the moments in the moment
equation. For instance, the moment of FP about
point O in the moment equation is expressed as


(R PO × FP ) z or R PO • FP .

Example 12.1

Assume three forces and a torque act on a body. Two of

the forces, FA and FB , are known as shown. Determine
FC and the applied torque T in order for the body to stay
in static equilibrium. Consider FA = 1.5 , θ FA = 325 o ,
FB = 1.9 Units, and θ FB = 40 o .

Sum of forces
Equation (12.1) can be solved either graphically or analytically.

Graphical: Equation (12.1) states that the sum of forces that act on the FC
body must form a closed loop; i.e., FA + FB + FC = 0 . We construct this
vector sum graphically in order to determine FC . Direct measurement from FA FB
the figure yields FC = 2.7 and θ FC = 187 o .

P. Nikravesh, M. Poursina 12-1

AME 352 STATIC FORCE ANALYSIS

Analytical: Equation (12.1) is projected onto the x- and y-axes to obtain two algebraic equations:
3

∑ F cosθ
i =1
i Fi = 0 ⇒ FA cosθ FA + FB cosθ FB + FC cosθ FC = 0
3

∑ F sin θ
i =1
i Fi = 0 ⇒ FA sin θ FA + FB sin θ FB + FC sin θ FC = 0

To solve for FC , we write the equations as

FC cos θ F = − FA cos θ F − FB cos θ F
C A B

FC sin θ F = − FA sin θ F − FB sin θ F

C A B

We square both sides and add the equations to get

( )
1/2
FC = (− FA cos θ F − FB cos θ F )2 + (− FA sin θ F − FB sin θ F )2
A B A B

Then sine and cosine of θ FC can be computed as

cos θ F = −(FA cos θ F + FB cos θ F ) / FC , sin θ F = −(FA sin θ F + FB sin θ F ) / FC
C A B C A B

These expressions provide the value of θ FC in its correct quadrant.

Numerically, the two known vectors are expressed in component form as
⎧ 1.23 ⎫ ⎧1.46 ⎫
FA = ⎨ ⎬ , FB = ⎨ ⎬
⎩−0.86 ⎭ ⎩1.22 ⎭
Then vector FC can be computed as
⎧ 1.23 ⎫ ⎧1.46 ⎫ ⎧ −2.68 ⎫
FC = −FA − FB = − ⎨ ⎬− ⎨ ⎬= ⎨ ⎬
⎩−0.86 ⎭ ⎩1.22 ⎭ ⎩−0.36 ⎭
The magnitude of this vector is FC = (−2.68)2 + (−0.36)2 = 2.7 . The vector can also be expressed as
⎧ −2.68 / 2.7 ⎫ ⎧−0.99 ⎫
FC = 2.7 ⎨ ⎬ = 2.7 ⎨ ⎬
⎩−0.36 / 2.7 ⎭ ⎩ −0.13⎭
For the angle of this vector we have cos θ FC = −0.99 and sin θ FC = −0.13 . These values yield θ FC = 187 o .

Note: If we use MATLAB to compute the angle, we can take advantage of the function atan2. In this
problem the function can be used as theta_F_C = atan2(-0.13, -0.99). The function provides the
angle in its correct quadrant.

Sum of moments
For Eq. (12.2), the sum of the torque and the moments of all three forces about any arbitrary point must
be equal to zero. We take the sum of moments about A; i.e.,
T + (R BA × FB ) z + (R CA × FC ) z = 0
or



T + R BA • FB + R CA • FC = 0
Note that the moment arm for FA is a zero vector. This
equation is rewritten as



T = − R BA • FB − R CA • FC
We can also write the moment equation with respect
to point B, C or any other point in the same way. For
example, if we pick an arbitrary point, such as O, the
moment equation becomes:
T + (R AO × FA ) z + (R BO × FB ) z + (R CO × FC ) z = 0
or




T + R AO • FA + R BO • FB + R CO • FC = 0
This equation yields the unknown torque as:

P. Nikravesh, M. Poursina 12-2

AME 352 STATIC FORCE ANALYSIS




T = − R AO • FA − R BO • FB − R CO • FC
Any of the above moment equations can be solved either graphically or analytically. In this example, we
take the moment equation about O.

Graphical: We first need to determine the angle between

every R vector and its corresponding F vector. For the FB
purpose of clarity, each pair of R and F vectors is shown
separately with the corresponding angle. The angles are R AO
θB
measured (from R to F CCW) and the moment equation is FA
evaluated as θA R BO
T + RAO FA sin θ A + RBO FB sin θ B + RCO FC sin θ C = 0
From the figure we measure the following
magnitudes: RAO = 1.1 , RBO = 1.8 and RCO = 1.5 . The θC
angles between each force vector and its corresponding R CO
position vector are measured to be: θ A = 128° , θ B = 58°
and θ C = 143° . With these values the applied torque is FC
determined to be
(
T = − (1.1)(1.5)sin128 + (1.8)(1.9)sin58 + (1.5)(2.7)sin143 = −6.64 )
Analytical: An x-y frame is positioned at O. The x and y components of all the vectors are determined and
denoted as
⎧ FA ⎫ ⎧ FB ⎫ ⎧ FC ⎫
⎪ x⎪ ⎪ x⎪ ⎪ x⎪
FA = ⎨ ⎬ , FB = ⎨ ⎬ , FC = ⎨ ⎬
F
⎪⎩ Ay ⎪⎭ F
⎪⎩ By ⎪⎭ ⎪⎩ FC y ⎪⎭
⎧ RAO ⎫ ⎧ RBO ⎫ ⎧ RCO ⎫
⎪ x⎪ ⎪ x⎪ ⎪ x⎪
R AO = ⎨ ⎬ , R BO
= ⎨ ⎬ , R CO
= ⎨ ⎬
⎪⎩ RAO y ⎪⎭ ⎪⎩ RBO y ⎪⎭ ⎪⎩ RCO y ⎪⎭

The moment equation for this body is written as

T − RAO FA + RAO FA − RBO FB + RBO FB − RCO FC + RCO FC = 0

y x x y y x x y y x x y

The angle of position vectors are measured to be θ RAO = 208 , θ RBO = 344 and θ RCO = 43o . Therefore,
o o

the position vectors can be expressed in component form as

⎧−0.97 ⎫ ⎧ 1.73 ⎫ ⎧1.10 ⎫
R AO = ⎨ ⎬ , R BO = ⎨ ⎬ , R CO = ⎨ ⎬
⎩ −0.52 ⎭ ⎩−0.50 ⎭ ⎩1.02 ⎭
The applied torque can now be computed as
T = (−0.52)(1.23) − (−0.97)(−0.86) + (−0.50)(1.46) − (1.73)(1.22)
+ (1.02)(−2.68) − (1.10)(−0.36) = −6.66

Special Cases

If only two or three forces act on a body (and nothing else), and the body is in static equilibrium, the forces
are balanced in such a way that they exhibit certain characteristics. We could take advantage of such cases
when solving for unknown forces.

Two-force member

If only two forces act on a body that is
in static equilibrium, the two forces are
along the line connecting the points of
effect of the forces, equal in magnitude,
and opposite in direction.

P. Nikravesh, M. Poursina 12-3

AME 352 STATIC FORCE ANALYSIS

Three-force member C
If only three forces act on a body that is FC
in static equilibrium, their axes intersect at
a single point. This knowledge can help us
simplify the solution process in some
FB
problems. For example, if the axes of two FA
of the forces are known, the intersection of
those two axes can assist us in determining A B
the axis of the third force.

A special case of the three-force

member is when the forces meet at a pin
joint that is connected between three links. F2
When the system is in static equilibrium, F1
the sum of the three forces must be equal to F1
zero.
F3
F2 F3

12.2 STATIC EQUILIBRIUM ANALYSIS OF MECHANISMS

A one degree-of-freedom planar mechanism may contain N moving links. If the mechanism is in static
equilibrium, then 3 equilibrium equations can be written for each link, resulting into 3N equations. In these
equations, there are normally 3N - 1 unknown reaction force/torque components and 1 unknown applied
force or torque. In a static force analysis, the main objective is to solve the equilibrium equations for the
unknown applied force/torque. Depending on the method of solution, the reaction forces/torques may also be
found in the process. In these notes, two methods for solving the static equilibrium equations are presented:
the free-body-diagram (FBD) method, and the power-formula (PF) method.

12.2.1 Free-Body Diagram Method

In this method the static equilibrium equations are constructed based on the free-body diagrams (FBD) of
each link. In each FBD, all of the reaction and applied forces/torques that act on that link are considered.
For a mechanism of N moving links, the 3N equilibrium equations form a set of linear algebraic equations in
3N unknowns. Although the equations are linear, due to their large number, a solution by hand may not be
practical or may not be the best choice. In such cases, a numerical solution through the use of a computer
program is recommended. However, if certain simplifications are made and the number of equations is
reduced, then solving the reduced set of equations by hand, either analytically or graphically, may be
practical.

Example 12.2

In the shown configuration, this slider-crank A

mechanism is in static equilibrium. A known
force F acts on the slider block in the direction
shown. An unknown torque acts on the crank.
Our objective is to determine the magnitude
and the direction of this torque in order to T
B F
keep the system in static equilibrium.
We construct the free body diagrams for O2
each link. The reaction forces at the pin joints

P. Nikravesh, M. Poursina 12-4

AME 352 STATIC FORCE ANALYSIS

are unknown. Each reaction force is described in terms of its x and y components, where the direction of
each component is assigned arbitrarily. For notational simplification, simple numbered indices are used for
all the components. For each link, we construct three equilibrium equations:

Link 2: Link 3: Link 4:

F6
F4
A F3 F5 B F
b
a
T T8
F7
F2
F1
O2
F1 + F3 = 0 − F3 + F5 = 0 − F5 + F = 0
F2 + F4 = 0 − F4 + F6 = 0 − F6 + F7 = 0
−aF3 + bF4 + T = 0 cF5 + dF6 = 0 T8 = 0
(Sum of moments about O2 ) (Sum of moments about A) (Sum of moments about B)

The moment arms are measured directly from the

⎧ ⎫
figure. Note that F7 and T8 are the reaction force ⎡ 1 0 1 0 0 0 0 0 0 ⎤ ⎪ F1 ⎪ ⎧ 0 ⎫
and torque due to the sliding joint. ⎢ ⎥ ⎪ ⎪
⎢ 0 1 0 1 0 0 0 0 0 ⎥ ⎪ F2 ⎪ ⎪ 0 ⎪
⎢0 0 −a b 0 0 0 0 1 ⎥ ⎪ F3 ⎪ ⎪ 0 ⎪
These 9 equations can be put into a matrix
form. The unknowns and their coefficients are ⎢ ⎥⎪ ⎪ ⎪ ⎪
kept on the left-hand side and the only known ⎢0 0 −1 0 1 0 0 0 0 ⎥ ⎪ F4 ⎪ ⎪ 0 ⎪
quantity, the known applied force F is moved to ⎢0 0 0 −1 0 1 0 0 0 ⎥ ⎪⎨ F ⎪⎬ = ⎪⎨ 0 ⎪⎬
the right-hand side. ⎢ ⎥⎪ 5⎪ ⎪ ⎪
⎢0 0 0 0 c d 0 0 0 ⎥ ⎪ F6 ⎪ ⎪ 0 ⎪
This set of 9 equations in 9 unknowns can be ⎢0 0 0 0 −1 0 0 0 0 ⎥ ⎪ ⎪ ⎪− F ⎪
solved by any preferred numerical method. If the ⎢ ⎥ F7
arbitrarily assigned direction to a force component ⎢0 0 0 0 0 −1 1 0 0 ⎥ ⎪ T ⎪ ⎪ 0 ⎪
⎢ ⎥⎪ 8 ⎪ ⎪ ⎪
⎣0 0 0 0 0 0 0 1 0 ⎦ ⎪ T ⎪ ⎪⎩ 0 ⎪⎭
or a torque is not correct, the obtained solution
will be a negative quantity. ⎩ ⎭
Numerical values for the link lengths are L2 = 2.0 and L3 = 4.0 (assuming a consistent unit system).
From the figures we extract the following measurements: a = 1.8 , b = 1.0 , c = 2.0 , d = 3.6 . Assume the
applied force is given to be F = 10 units. These values are substituted in the equilibrium equation. The
solution to the unknowns is determined to be:
F1 = −10, F2 = 5.56, F3 = 10, F4 = −5.56, F5 = 10 , F6 = −5.56, F7 = −5.56, T8 = 0, T = 23.55
We note that the applied torque on the crank is T = 23.55 , and as a byproduct of this process, all the
reaction forces/torques are also found.
As it is observed, F7 = −5.56 is negative. In this
application, the slider moves inside a chamber (or a cylinder). F7
The negative normal force from the chamber to the slider
means that the upper side of the slider is in contact with the
chamber.

Simplified FBD method

In some mechanisms, we can identify the two- or three-force members. The corresponding free-body
diagrams provide us with useful information about the axis of the force vectors. This may help us avoid
writing the static equilibrium equations for every single link in the system to find the unknown applied
force/torque.

P. Nikravesh, M. Poursina 12-5

AME 352 STATIC FORCE ANALYSIS

The connecting rod of this mechanism is a two-force

A
member. The reaction forces at A and B must be on F3 F3
the line connecting A to B, equal in magnitude but in F4
opposite directions. These reaction forces are named 27 o
F3 and given arbitrary directions. F
A
The FBD of links 2 and 4 can now be F3 B
constructed. For link 4 we note that the sum of three F3
forces that go through point B must be zero. T
e
Essentially link 3 is a three-force member. This
yields F3 = F / cos 27 = 11.22 , which acts along the F1 F3 B F
axis of link 3. We can also determine the magnitude
O2
and direction of F4 . Furthermore, the direction of
F2 T5
the F4 in the free-body diagram should be adjusted F4
in the negative y-direction.
For link 2 we write the moment equation about O2 as −eF3 + T = 0 , where the moment arm can be
found from measurement to be e = 2.0 . This yields the unknown applied torque as T = eF3 = 23.57 .

Example 12.3

For the slider-crank mechanism in Example 12.2, we are asked to include gravitational forces on the links.
The mass center for each link is positioned at the geometric center. The free body diagrams for the three
links are constructed.

Link 2: Link 3: Link 4:

F1 + F3 = 0 − F3 + F5 = 0 − F5 + F = 0
F2 + F4 − W2 = 0 − F4 + F6 − W3 = 0 − F6 + F7 − W4 = 0
b d T8 = 0
−aF3 + bF4 − W2 + T = 0 cF5 + dF6 + W3 = 0
2 2
(Sum of moments about O2 ) (Sum of moments about A) (Sum of moments about B)

P. Nikravesh, M. Poursina 12-6

AME 352 STATIC FORCE ANALYSIS

⎧F ⎫
⎡1 0 1 0 0 0 0 0 0⎤ ⎪ 1 ⎪ ⎧ 0 ⎫
⎥ ⎪F ⎪ ⎪ W ⎪
These 9 equations are expressed
⎢
in matrix form and solved for the 9 ⎢0 1 0 1 0 0 0 0 0⎥ 2 ⎪ 2 ⎪
unknowns. We note that the left- ⎢0 ⎥ ⎪ F ⎪ ⎪ bW / 2 ⎪
0 −a b 0 0 0 0 1 ⎪ 3⎪
hand side of this matrix equation is ⎢ ⎥ ⎪ 2
⎪
identical to that in Example 1. The ⎢0 0 −1 0 1 0 0 0 0 ⎥ ⎪ F4 ⎪ ⎪ 0 ⎪
⎢0 ⎥ ⎪ ⎪ ⎪ ⎪
difference is in the right-hand sides. 0 0 −1 0 1 0 0 0 ⎨ F5 ⎬ = ⎨ W3 ⎬
⎢ ⎥
⎢0 0 0 0 c d 0 0 0 ⎥ ⎪ F6 ⎪ ⎪−dW3 / 2 ⎪
⎢0 ⎪ ⎪ ⎪ ⎪
⎢ 0 0 0 −1 0 0 0 0 ⎥⎥ ⎪ F ⎪ ⎪ − F ⎪
⎢0 ⎪ 7⎪ ⎪ ⎪
0 0 0 0 −1 1 0 0⎥ ⎪T ⎪ ⎪ W ⎪
⎢ ⎥ 8 4

⎣0 0 0 0 0 0 0 1 0⎦ ⎪ T ⎪ ⎪ 0 ⎪
⎩ ⎭ ⎩ ⎭

Note: Since three forces act on the connecting rod,

A
F3
this link is no longer a two-force member. We
cannot assume that the reaction forces at A and B
are equal, in opposite directions, and along the axis
of the link. Therefore, the simplified method is not
applicable to this problem. This is the case for w2 B
most examples when we include gravitational F3
forces.

12.3 COULOMB FRICTION

Coulomb friction can be included between two contacting surfaces in a static force analysis. Given the static
coefficient of friction, μ (s) , the friction force can be described as the product of the coefficient of friction
and the reaction force normal to the contacting surfaces. The friction force must act in the opposite direction
of the tendency of any relative motion. Since the assumption is that the system is stationary, the tendency of
motion must be considered for two cases. The process provides a range of values for the applied load while
the system remains in equilibrium.

Example 12.4

We repeat Example 12.2 with the assumption that dry friction exists between the slider block and the
ground. We solve this problem with the FBD method. We may write the complete set of equilibrium
equations or take advantage of a 2-force member in the system. Here we construct the complete set of
equations. Since the FBD for the crank and the connecting rod are the same as before, we only show the
FBD for the slider block.
First we assume that the block is about to move to the left. Therefore, the friction force must be directed
to the right. Then, we assume that the slider block is about to move to the right. Therefore, the friction force
must be directed to the left:

− F5 + F + μ (s) F7 = 0 F6
− F6 + F7 = 0 (The block tends to move to the left)
F5 B F
T8 + eμ F7 = 0
(s)
e

μ (s ) F7 T8
F7

P. Nikravesh, M. Poursina 12-7

AME 352 STATIC FORCE ANALYSIS

− F5 + F − μ (s) F7 = 0 F6
− F6 + F7 = 0 (The block tends to move to the right)
F5 B F
T8 − eμ F7 = 0
(s)
e

μ (s ) F7 T8
F7
⎧F ⎫
We construct two complete sets of equations
⎡1 0 1 0 0 0 0 0 0⎤ ⎪ 1 ⎪ ⎧ 0 ⎫
⎢ ⎥ ⎪ ⎪
for these two cases as shown: one with the ⎢0 1 0 1 0 0 0 0 0 ⎥ ⎪ F2 ⎪ ⎪ 0 ⎪
⎢0 ⎪ ⎪
positive sign for μ (s) and one with the negative 0 −a b 0 0 0 0 1 ⎥ ⎪ F3 ⎪ ⎪ 0 ⎪
⎢ ⎥ ⎪ ⎪
sign. Each solution yields a value for the ⎢0 0 −1 0 1 0 0 0 0 ⎥ ⎪ F4 ⎪ ⎪ 0 ⎪
⎢0 ⎪ ⎪ ⎪ ⎪
unknown torque, Tleft and Tright . As long as the 0 0 −1 0 1 0 0 0 ⎥ ⎨ F5 ⎬ = ⎨ 0 ⎬
⎢ ⎥⎪ ⎪ ⎪ ⎪
applied torque stays in the range of Tleft to Tright , ⎢0 0 0 0 c d 0 0 0 ⎥ F6
⎪ ⎪ ⎪ ⎪
0
⎢0 0 0 0 −1 0 ± μ (s) ⎥
0 0 ⎥ ⎪ F ⎪ ⎪− F ⎪
the system remains in static equilibrium. ⎢ ⎪ 7⎪ ⎪ ⎪
Assuming μ (s) = 0.25 , e = 0.5 , and using the ⎢0 0 0 0 0 −1 1 0 0⎥ ⎪T ⎪ ⎪ 0 ⎪
⎢ ⎥ 8
numerical values from Example 12.2, we solve ⎣0 0 0 0 0 0 ±eμ (s) 1 0 ⎦ ⎪ T ⎪ ⎪⎩ 0 ⎪⎭
the system of 9 equations twice. ⎩ ⎭

For the plus sign in front of the terms containing μ (s) , the solution yields T = 20.68 , where the minus
sign in front of those two terms yields T = 27.35 . Therefore, as long as the applied torque is in the range
20.68 ≤ T ≤ 27.35 , the mechanism remains in static equilibrium.

Example 12.5

In this example, a known force acts at P

point P and an unknown torque acts on the
crank. It is assumed that dry friction exists
FP A
at the sliding joint. Find the magnitude and
direction of T torque in order to keep the y
system in static equilibrium.
To solve this problem, we construct T
FBD’s for the links. Although it is not x
necessary, in order to simplify the process
of projecting the reaction forces onto the x-y O2 O4
axes, the x-y axes are rotated in such a way
that the x-axis is along the axis of link (4).
We construct 3 equilibrium equations for
each link as shown:

Link 2: Link 3: Link 4:

FP( x )
F4 P
T6
F3 T6 F4
FP( y) F5
A b A
μ F5
(s )
F3 μ (s ) F5
a
T F5 F8
O2
d F7
F2 F1 O4

P. Nikravesh, M. Poursina 12-8

AME 352 STATIC FORCE ANALYSIS

F1 + F3 = 0 − F3 ± μ (s) F5 = 0 F7 ∓ μ (s) F5 + FP( x ) = 0

F2 + F4 = 0 − F4 − F6 = 0 F5 + F8 − FP( y ) = 0
−aF3 − bF4 − T = 0 T6 = 0 −T6 − cF5 + dFP = 0
⎧F ⎫
⎡1 0 1 0 0 0 0 0 0 ⎤⎪ 1⎪ ⎧ 0 ⎫
These equations can be expressed ⎢ ⎥ ⎪ ⎪
in matrix form and solved twice to ⎢0 1 0 1 0 0 0 0 0 ⎥ ⎪ F2 ⎪ ⎪ 0 ⎪
⎪ ⎪
find the range of values for the ⎢0 0 −a −b 0 0 0 0 −1⎥ ⎪ F3 ⎪ ⎪ 0 ⎪
⎢ ⎥ ⎪ ⎪
applied torque T.
⎢0 0 −1 0 ± μ (s) 0 0 0 0 ⎥ ⎪ F4 ⎪ ⎪ 0 ⎪
⎢0 ⎪ ⎪ ⎪ ⎪
0 0 −1 0 −1 0 0 0 ⎥ ⎨ F5 ⎬ = ⎨ 0 ⎬
⎢ ⎥⎪ ⎪ ⎪
⎢0 0 0 0 0 0 1 0 0 ⎥ T6 0 ⎪
⎢0 ⎪ ⎪ ⎪ ⎪
⎢ 0 0 0 ∓ μ (s) 0 1 0 0 ⎥⎥ ⎪ F ⎪ ⎪− FP( x ) ⎪
⎢0 ⎪ 7⎪ ⎪ ⎪
0 0 0 1 0 0 1 0 ⎥ ⎪ F ⎪ ⎪ FP( y ) ⎪
⎢ ⎥ 8
⎣0 0 0 0 −c −1 0 0 0 ⎦ ⎪ T ⎪ ⎪ −dF ⎪
⎩ ⎭ ⎩ P ⎭

12.4 POWER FORMULA METHOD

The power formula method constructs one equation in one unknown for a one degree-of-freedom
mechanism regardless of its number of links and joints. Only applied forces and torques, including the
unknown applied force or toque, appear in the equation. This means that the reaction forces and torques
have been eliminated from the power formula.
The mechanical power is defined as the scalar product of a force F and the velocity V of the application
point of the force; i.e.,
P = F• V (12.3)
The power can also be defined as the product of a torque and the angular velocity of the link that the torque
is applied to; i.e.
P = Tω (12.4)
To use the power formula, although the system is in static equilibrium, we first assume that the system
is in motion. We assign an arbitrary (or virtual) velocity (angular or linear) to one of the links; e.g., for the
rotation of the crank, and determine the velocities that are needed in the power formula. These are the
velocities of the points where the applied forces act on, and the angular velocities of the links that the
torques are applied to.
The equation for the power formula is: FC (6)
C
∑ P P ∑
T5
F i V + T ω = 0 (12.5)
In this equation FP is a typical applied force acting on a link at (5)
point P, VP is the velocity (absolute) of point P, T is a typical
B
applied torque acting on a link, and ω is the angular velocity of that FB
link. If the torque and the angular velocity are in the same direction,
(3)
the product T ω is positive, otherwise the product is negative.
As an example consider the six-bar mechanism shown. Three T2 (2) A
applied forces and three applied torques act on the system. We need
to find the virtual velocities of points A, B, and C, and the imaginary F4
O2
angular velocities of links 2, 4, and 5. We can find these velocities (4)
by assigning an arbitrary value to one of the linear or angular
velocities, for instance, the angular velocity of link 2. We may apply T4
any preferred method to solve for the other velocities. The value of
this arbitrarily assigned velocity or its direction will not change the O4
result obtained from the power formula.

This method is also referred to as the energy method or the virtual work method. Since we have not
discussed the concepts of virtual work and virtual displacements, the method is presented using velocities
although the system is in static equilibrium.

P. Nikravesh, M. Poursina 12-9

AME 352 STATIC FORCE ANALYSIS

Note: Since reaction forces and torques do not appear in the power formula, this method is not applicable
when friction is present.

Example 12.6

We apply the power formula to the slider-crank A

mechanism of Example 12.2. One external
torque and one external force act on this system.
We need to determine the imaginary angular
velocity of link 2 and the velocity of the slider
block. T
B F
We assume link 2 rotates with an angular
velocity ω 2 = 1 rad/sec, CCW. We construct the O2
velocity polygon and determine the velocity of
the slider block (point B). Considering the
length of the crank as L2 = 2.0 , the polygon
yields VB = 2.4 .
For this problem the power formula
becomes F i VB + T ω 2 = 0 . Since F and VB are along the same axis but in opposite directions, the formula
simplifies to − FVB + T ω 2 = 0 , or T = FVB / ω 2 = (10)(2.4) / 1.0 = 24.0 .

Example 12.7

We apply the power formula to the problem of

Example 12.3. We assume an imaginary ω 2 = 1
rad/sec, CCW, and construct the velocity (as in
the previous example). The velocities of the
points where the applied forces act, including
the gravitational forces, are determined. The PF
is then expressed as:
F i VB + W2 i VG + W3 i VG + W4 i VG + T ω 2 = 0
2 3 4

The dot product between a velocity vector and a force vector

can be determined either analytically or graphically (the angle
between the two vectors can be measured directly from the figure).
The PF yields the value for the unknown applied torque.

12.5 MECHANICAL ADVANTAGE AND TRANSMISSION ANGLE

In a four-bar mechanism if link (2) is the input link, the angle between link (3) (the coupler) and the output
link (4) is called the transmission angle, μ , where μ ≤ 90° . The optimal transmission angle is μ = 90° ; i.e.,
when the largest amount of power is transmitted from the input link to the output link. However, the
transmission angle changes as the mechanism moves.
P. Nikravesh, M. Poursina 12-10
AME 352 STATIC FORCE ANALYSIS

(3)
(3) μ
(4)
(4)
(2)
(1) (1)
(2)

The transmission of power in a mechanical system can be expressed as

Pin = Pout + Plosses (12.6)
where Pin is the input power, Pout is the output power, and Plosses is the lost power most likely due to the
presence of friction. In most mechanisms, in particular in four-bar mechanisms, the lost power is very small
(less that 10% [1]). Therefore Plosses can be assumed zero and hence Eq. (12.6) can be simplified to:
Pin = Pout (12.7)
In a four-bar mechanism if T2 is the torque applied to the input link and T4 is the torque that the output
link can apply on a load, and if ω 2 and ω 4 are the angular velocities of the two links, then Eq. (12.7) can be
expressed as
T2 ω 2 = T4 ω 4 (12.8)
The Mechanical advantage (MA) for a four-bar is defined as the ratio of the torque generated by the output
link, T4 , to the torque applied on the input link, T2 :
MA = T4 / T2 = ω 2 / ω 4 (12.9)
where Eq. (12.8) has also been used. This equation shows that the ratio of the two torques is inversely
proportional to the ratio of the corresponding links’ angular velocities. Since Eq. (12.9) contains the ratio of
angular velocities, we can further express the equation in the following form using the ratio of instant
centers of velocities, Eq. (8.2):
MA = T4 / T2 = ω 2 / ω 4 = RI I / RI I (12.10)
1,4 2 ,4 1,2 2 ,4

Another useful form of the expression for

mechanical advantage is to transform Eq.
(12.10) to the following form using the μ
transmission angle, μ , and the angle between
the input link and the coupler, β : (4)
L sin μ L4
MA = 4 (12.11) L4 sin μ
L2 sin β L 2 sin β L2 β
Equation (12.11) clearly shows that to maximize (2)
the mechanical advantage, the transmission
angle should be close to μ = 90° and the angle I 2,4 I1,2 I1,4
β should be close to β = 0° or β = 180° .
As an example, consider the four-bar mechanism shown in Figure (a), which is used to crush an object
such as can or rock. In this mechanism T2 is the input torque to the system. The output torque T4 is related
to the force applied to the object as
T4 = Fc RCO4 (12.12)
where T4 and Fc are shown in Figure (b). Equation (12.12) shows the direct relationship between the output
torque T4 and Fc . In order to generate a large force, T4 must be large, and consequently MA must be large
according to Eq. (12.10). Based on Eq. (12.11), this is achieved when the transmission angle is close to
μ = 90° and angle β is close to 0° or 180° as shown in Figure (a). It should be noted that the four-bar
should be designed to satisfy these conditions during the time period that link 4 would be crushing the
object, not necessarily for the entire complete revolution of link 2.

P. Nikravesh, M. Poursina 12-11

AME 352 STATIC FORCE ANALYSIS

T4
(4)
μ
(4) object FC C
T2 (2)
β R CO4
O4

(a) (b)

Figure: (a) Four-bar mechanism used to crush an object, (b) Fc is applied to the object by link 4

P. Nikravesh, M. Poursina 12-12

AME 352 DYNAMIC FORCE ANALYSIS

13 DYNAMIC FORCE ANALYSIS

This chapter discusses some of the methodologies used to perform dynamic force analysis on mechanisms. It
will be shown that a dynamic force analysis is very similar to a static force analysis if we considered a
concept known as D’Alembert’s principle. The dynamic force analysis discussion is followed by a section
on power/motor selection and flywheel.

13.1 DYNAMIC EQUILIBRIUM EQUATIONS

The dynamic equilibrium equations for a single body in planar FC

motion are the revised version of the static equilibrium equations, C FB
T
expressed as
⎧⎪ ∑ Fi( x ) = mAG ( x )
∑ Fi = m AG ⇒ ⎨ F = mA
B
(13.1) I G

⎩⎪∑ i( y )
G
G( y)
mA G
∑T + ∑ M
j i(G )
= IG α (13.2)
A
FA

The force equilibrium of Eq. (13.1) states that the sum of forces that act on a body must be equal to the
mass of the body, m, times the acceleration of the mass center, A G . This equation is derived from Newton’s
second Law of motion. The moment equilibrium of Eq. (13.2) states that the sum of all the applied toques
and the moments of all forces that act on a body with respect to its mass center must be equal to the body’s
moment of inertia, I G , times the angular acceleration of the body, α . The moment of inertia is defined with
respect to an axis passing through the mass center and perpendicular to the plane. Equations (13.1) and
(13.2) are also known as the translational and rotational equations of motion, respectively.

13.1.1 D’Alembert’s Principle

The D’Alembert’s principle considers a rearranged form of the

dynamic equilibrium equations as FC
C T FB
∑F − mA = 0
i G
(13.3)

∑T + ∑ M − I
j i(G ) G
α =0 (13.4)

I G
G B

mA G
Although this rearrangement may appear trivial, it has profound
implications. One interpretation of this principle is that if −m A G A
FA
and − I G α are viewed as a force and a torque respectively, then
the dynamic equilibrium equations become identical to the static
equilibrium equations. The terms −m A G and − I G α are called
inertial force and inertial torque.

13.1.2 Revised Rotational Equation

D’Alembert’s principle provides us a simple set of steps to revise Eq. (13.2) if the sum of moments is taken
about a point that is not the mass center. If we consider −m A G and − I G α as additional force and torque
that act on the body, and consider point O as the reference point instead of the mass center, the rotational
static equilibrium of Eq. (12.2) yields

P. Nikravesh and M. Poursina 13-1

AME 352 DYNAMIC FORCE ANALYSIS

FC
∑T + ∑ M
j i (O ) − IG α − m (RGO × AG ) z = 0 C T FB
or,
R CO
∑T + ∑ M
j i (O ) = IG α + m (RGO × AG ) z (13.5)
G
R BO B

I G
R GO
This equation clearly shows that if the origin is not the
R AO
mA G
mass center, then we must consider the moment caused by O
the inertia force in the moment equation. We can also A
express the directional magnitude of the moment of the FA
inertial force about O as a dot product:


∑T + ∑ M
j i(O )
− I G α − m R GO • A G = 0

13.2 DYNAMIC EQUILIBRIUM ANALYSIS OF MECHANISMS

The process of dynamic force analysis of mechanisms is only slightly different from that of the static force
analysis, especially if we look at the dynamic equilibrium equations from the perspective of D’Alembert’s
principle. The main difference is that in a dynamic force analysis we need to include the linear and angular
accelerations of each link in the process of solving the dynamic equilibrium equations.
For the dynamic force analysis, as in the static force analysis, we consider the free-body diagram (FBD)
method and the power formula (PF) method.

13.2.1 Free-Body Diagram Method

The dynamic force analysis requires incorporating the applied and reaction forces and torques, and the
inertial force and torque, in the FBD of each link. Inclusion of the applied and reaction forces and torques in
this process is identical to that of the static force analysis. To include the inertial forces and torques in the
process, we need to know the linear and angular accelerations for each link.

Example 13.1

The mass and moment of inertia for the links

of this slider-crank are given. A known force
F acts on the slider block, and an unknown T
torque T acts on the crank. In the shown
2
configuration, the angular velocity and
2 B F
acceleration of the crank are given. The
objective is to find the unknown torque O2
(magnitude and direction). It is assumed that
the mass centers are located at the geometric A V
A
centers of the links. G3 G2
VG3 VG2
Based on the given angular velocity and VBA
OV
acceleration of the crank, polygons are
B
constructed and the angular velocity and VB = VG4
acceleration of all the links are found ( ω 2 ,
A A = AG4
α 2 and α 3 are CCW, and ω 3 is CW). B OA
A G3
G3
A G2
G2
A AA

In a process similar to that of the static equilibrium (refer to Example 12.2), the FBD for each link is
constructed as shown.

P. Nikravesh and M. Poursina 13-2

AME 352 DYNAMIC FORCE ANALYSIS

F1 + F3 = m2 AG ( x )
2

F2 + F4 = m2 AG ( y )
2

a b a b
F1 − F2 − F3 + F4 + T = I G2 α 2
2 2 2 2

− F3 + F5 = m3 AG ( x )
3

− F4 + F6 = m3 AG ( y )
3

c d c d
F3 + F4 + F5 + F6 = I G3 α 3
2 2 2 2

− F5 + F = m4 AG
4

− F6 + F7 = 0
T8 = 0

Note that AG ( x ) , AG ( y ) , AG ( x ) , ⎡1 0 1 0 0 0 0 0 0⎤
⎢0 ⎧ m2 AG2 ( x ) ⎫
0 0 0 ⎥⎥ ⎧ F1 ⎫ ⎪
2 2 3

AG ( y ) and AG are negative, ⎢ 1 0 1 0 0 ⎪

⎪ F ⎪ ⎪ m2 AG2 ( y ) ⎪
⎢a ⎥ 2
3 4

0 0 1 ⎥ ⎪⎪ ⎪⎪ ⎪ I α ⎪
b a b
where α 2 and α 3 are positive. ⎢ − − 0 0
⎥⎪ 3⎪ ⎪ ⎪
F
⎢2 2 2 2 G2 2
These equations are put into ⎪ ⎪
⎢0 0 −1 0 1 0 0 0 0 ⎥ ⎪ F4 ⎪ m A
⎪ ⎪ ⎪⎪ ⎪⎪
3 G3 ( x )
matrix form. This set of 9
⎢0 0 0 −1 0 1 ⎥
0 0 0 ⎨ F5 ⎬ = ⎨ m3 AG3 ( y ) ⎬
equations in 9 unknowns can be ⎢ ⎥
solved by any preferred numerical ⎢ ⎥ ⎪F ⎪ ⎪ ⎪
0 0 0 ⎥ ⎪ 6 ⎪ ⎪ I G3 α 3 ⎪
c d c d
method. ⎢0 0
⎢ 2 2 2 2 ⎥ ⎪ F7 ⎪ ⎪m A − F ⎪
⎢0 0 0 0 −1 0 0 0 0 ⎥ ⎪ T ⎪ ⎪ 4 G4 ⎪
⎢0 ⎥ ⎪ 8⎪ ⎪ 0 ⎪
0 0 0 0 −1 1 0 0 ⎪T ⎪ ⎪ ⎪
⎢ ⎥⎩ ⎭ ⎪
⎢⎣ 0 0 0 0 0 0 0 1 0 ⎥⎦ ⎩ 0 ⎭⎪

13.3 COULOMB FRICTION

Coulomb friction can be included between two contacting surfaces in a dynamic force analysis. Given the
dynamic (kinetic) coefficient of friction, μ ( k ) , the friction force can be described as the product of the
coefficient of friction and the reaction force normal to the contacting surfaces. The friction force must act in
the opposite direction of the relative motion of the contacting surfaces. The process is illustrated through a
simple example.

P. Nikravesh and M. Poursina 13-3

AME 352 DYNAMIC FORCE ANALYSIS

Example 13.2

We consider the slider-crank mechanism of Example 13.1 with the assumption that dry friction exists
between the slider block and the ground. We construct the complete set of equations for this problem. Since
the FBD for the crank and the connecting rod are the same as in Example 13.1, we only show the FBD for
the slider block.
According to the velocity polygon the block is moving
to the left, and therefore the friction force must be directed
to the right. The equilibrium equations for the block are:
− F5 + F + μ ( k ) F7 = m4 AG
4

− F6 + F7 = 0
T8 + eμ ( k ) F7 = 0
⎧ m2 AG ( x ) ⎫
⎧F ⎫
The complete set of equations ⎡1 0 0 0⎤ ⎪ 1 ⎪ ⎪ ⎪
2
1 0 0 0 0
can now be presented in ⎢ ⎥ ⎪F ⎪ ⎪ m 2 G2 ( y ) ⎪
A
matrix form. The solution to ⎢0 1 0 1 0 0 0 0 0⎥ 2 ⎪ ⎪
⎢ a −b −a b 0 0 ⎥ ⎪ F ⎪ ⎪ IG α 2 ⎪
this set of equations yields all 0 0 1 ⎪ 3⎪
⎪ ⎪
2
⎢ ⎥
the unknowns. ⎢ 0 0 −1 0 1 0 0 0 0⎥ 4 ⎪ F ⎪ ⎪ 3 G3 ( x ) ⎪
m A
⎢0 0 ⎪ ⎪ ⎪ ⎪
0 −1 0 1 0 0 0 ⎥ ⎨ F5 ⎬ = ⎨ m3 AG ( y ) ⎬
⎢ ⎥
0 0 ⎥ ⎪ F6 ⎪ ⎪ I α ⎪
3

⎢0 0 c d c d 0
⎪ ⎪ ⎪ ⎪
⎢0 0 0 0 −1 0 μ ( k ) 0 0 ⎥⎥ ⎪ F ⎪ ⎪
G3 3
⎪
⎢ ⎪ 7
⎪ ⎪ m4 AG − F ⎪
⎢0 0 0 0 0 −1 1 0 0⎥ ⎪T ⎪ ⎪ ⎪
4

⎢ ⎥ 8
⎣0 0 0 0 0 0 eμ ( k ) 1 0⎦ ⎪ T ⎪ ⎪ 0 ⎪
⎩ ⎭ ⎪ ⎪⎭
⎩ 0

13.4 POWER FORMULA METHOD

The power formula (PF) for dynamic force analysis is a revised form of the formula from the static force
analysis. If we consider the inertial forces and torques as any other applied forces and torques, the revised
formula can be expressed as
∑ FP i VP + ∑ T ω = ∑ mi AG i VG + ∑ IG α i ω i
i i i
(13.6)
In this formula, the velocities and accelerations are actual and not imaginary. If the torque and the angular
velocity are in the same direction, the product T ω is positive, otherwise the product is negative. If the
angular velocity and acceleration are in the same direction, the product α ω is positive, otherwise the
product is negative.

Example 13.3

We apply the PF method to the slider-crank of Example 13.2. The formula for this system can be expressed
as
F i VB + T ω 2 = m2 A G i VG + m3A G i VG + m4 A G i VG + I G α 2ω 2 + I G α 3ω 3 + I G α 4ω 4
2 2 3 3 4 4 2 3 4

In addition to the angular velocities and accelerations, from the velocity and acceleration polygons, the
velocity and acceleration of the mass centers are determined. All the known quantities are substituted in the
formula to find the unknown applied torque T.

P. Nikravesh and M. Poursina 13-4

AME 352 DYNAMIC FORCE ANALYSIS

α3 (CCW)
ω3 (CW)

13.5 SHAKING FORCE AND SHAKING TORQUE/MOMENT

Dynamic forces that act on a mechanism cause the foundation (the ground) to shake. The shaking force is
defined as the sum of reaction forces the links of a mechanism apply on the ground link. It essentially tends
to move the ground (the frame) up and down, and left and right.
For a four-bar mechanism, since there are two links that are directly pinned to the ground at O2 and O4 ,
the sum of reaction forces acting on the ground at these two points is the shaking force denoted as
⎧ F1 + F7 ⎫
Fs = − ⎨ ⎬ (13.7)
⎩ F2 + F8 ⎭
B
A
where F1 , F2 , F7 and F8 could be positive or negative.
The torque that is applied to the ground is called the
shaking torque (moment). It essentially tends to rock the
frame about the vertical axis to the plane.
Different references provide different O2 O4
interpretations of the shaking torque.
For a four-bar mechanism, when link 2 is the input
link (the crank), a motor must rotate it about the axis of
O2 . The motor applies a torque T on link 2 and an
opposite torque on the ground. Based on this B
interpretation, Reference [2] computes the shaking A
moment as
Ts = −T (13.8)
In Reference [1], the shaking torque is computed as the F1 T F7
entire moment about O2 . This moment essentially
O2 O4
contains the reaction torque of the motor as well as the
moment of the forces of the pins about O2 . Based on F2 F8
this interpretation, the shaking moment can be F2 F8
computed as T
F1 F7
Ts = −T − RO4O2 F8 (13.9) O2 O4
Finally, Reference [3] defines the shaking moment as
Ts = − RO4O2 F8 (13.10)
Due to the existence of these different definitions, we must be clear which definition is used when we
compute the shaking moment for a four-bar mechanism.

P. Nikravesh and M. Poursina 13-5

AME 352 DYNAMIC FORCE ANALYSIS

13.6 A MATLAB PROGRAM FOR THE DYNAMIC ANALYSIS OF A FOUR-BAR

A MATLAB program (fourbar_force.m) for static and dynamic force analysis of a four-bar mechanism
is provided. The four-bar is considered in its most general form as shown. The program allows one known
force and one known torque to be applied on each link. The gravitational force may also be considered to
act in the negative y-direction on the system. The program requires only one unknown torque, which could
be considered on only one of the links.
For the purpose of generality, the mass center of
a link, Gi , is positioned from the reference point of
the link by an angle γ i and a length LGi . The force
application point, Pi , is positioned by an angle βi
and a length LPi . Note that the reference point for
link 2 is O2 , for link 3 is A, and for link 4 is O4 .
This program retrieves the data from a file
named fourbar_force_data.m. The user is
required to provide the following data in this file:

• Constant values for the link lengths

( L1 , L2 , L3 , L4 )

• Constant angle of the ground link ( θ1 ) if an offset four-bar

• Angle of the crank ( θ 2 ) Pi
• Estimates for the angles of the coupler and the follower ( θ 3 , θ 4 )
LPi
• Angular velocity and acceleration of the crank ( ω 2 , α 2 ) LGi Gi
• Constant values for the position of P points (i)
γi
( LP2O2 , LP3A , LP4O4 , β 2 , β 3 , β 4 )
βi Li
• Constant values for the position of G points θi
( LG2O2 , LG3A , LG4O4 , γ 2 , γ 3 , γ 4 )
• Known applied forces ( FP2 , FP3 , FP4 )
• Known applied torques ( T2 , T3 , T4 )
• Masses and moments of inertia ( m2 , m3 , m4 , I 2 , I 3 , I 4 )
• Gravitational constant (g) (the value of g specifies the system of units for the analysis)

The program prompts the user for the following information:

• Should the gravitational force be included? o Answer y for yes or n for no

• The unknown torque is applied to which link? o Answer 2, 3 or 4
• Static or dynamic force analysis? o Answer s for static or d for dynamic

The program reports the results for the B

reaction forces at O2 and A on link 2, at B F6
(3)
and O4 on link 4, and the unknown applied A
F4 B F5
torque. The user could easily determine
reaction forces acting at A and B on link 3. A F3
The program also reports the shaking force
(4)
and the shaking torque/moment according (2)
to the definition of Eq. (13.8). F7
F1
O2 O4
F2 F8

P. Nikravesh and M. Poursina 13-6

AME 352 DYNAMIC FORCE ANALYSIS

13.7 INPUT POWER REQUIREMENT (MOTOR)

In this section, we discuss issues associated with the power requirement from a motor to operate a
mechanism. A typical dynamic force analysis on a mechanism to determine the required torque during a
complete revolution of the crank normally yields large variations in the torque, which a typical motor is
unable to sustain. The large variations in the required torque can be due to large variations in the
accelerations and in the applied or resistive loads on the system.
As an example consider the four-bar web-cutter mechanism shown in Figure 1. As the crank rotates, the
two blades at C and D cut a moving web, one cut in each revolution. The duration of the cut lasts only for a
short fraction of one revolution of the crank, during which large resistive forces act on the blades. The motor
applies a torque on the crank, and the gravity acts on all the links. We also consider friction to exist at the
pin joints.

If we require the crank to rotate with a desired

constant angular velocity, for example 2π rad/sec C
(1 rev/sec), a dynamic force analysis for one B 5 mm
revolution of the crank would yield a torque versus D
crank-angle curve as shown in Figure 2. The plot
shows that the motor must provide a large torque
(3)
for a very short period when the blades cut through (4) gravity
the web. During the rest of the rotation, the
required torque varies between positive and
negative values. A negative torque means that the (2) A
motor must act as a brake in order to keep the
system from speeding up (due to gravity and O2 O4
inertia). Providing such variations in the torque is
not possible for a typical motor.
Figure 1: A web-cutter four-bar mechanism.

Figure 2: Torque versus crank angle for a constant angular velocity of the crank.

Most mechanisms are operated by electric motors. Motors are rated to provide certain amount of power
(or torque) at a specified speed. For electric motors, the torque-speed (angular) characteristics could be
either linear or nonlinear. Permanent-magnet (PM) motors have linear characteristics but compound-wound
electric motors have nonlinear characteristics as shown. The rated torque and rated speed are expressed at
100%. Electric motors normally provide the largest torque when the speed is zero [1]. Our challenge is to
select a motor, with known characteristics, to provide the desired requirements for our mechanism as closely
as possible.

P. Nikravesh and M. Poursina 13-7

AME 352 DYNAMIC FORCE ANALYSIS

Figure 3: Torque-speed characteristics for a PM motor and a compound wound motor.

To determine the size of the motor for a system, we need to look at the average required torque in one
revolution of the crank. For our web-cutter, the average required torque is 6.8 N.m as shown in Figure 2 in
dashed line. Assume that we have found a PM motor with the same characteristics as shown in Figure 3, and
it is rated to provide 6.8 N.m torque at 6.28 rad/sec (100%). This means that the motor is capable of
providing a torque of 5 × 6.8 N.m at zero angular velocity. For this motor the torque-speed linear
characteristics can be expressed as
T = Tr (5 − 4ω / ω r )
where Tr = 6.8 and ω r = 2π rad/sec.
We perform a forward dynamic analysis of the web-cutter with this motor, assuming that the system is
initially at rest. A forward dynamic analysis requires numerical integration of the differential equations of
motion, which has not been discussed in this course. In a forward dynamic analysis, we determine the
motion of a system based on the applied loads, including the torque of the PM motor.
The response from the forward dynamic analysis of the web-cutter is shown in Figure 4 for 10 seconds
of simulation. Figure 4(a) shows the angular velocity of the crank (or the motor) versus time indicating that
the angular velocity fluctuates between 3.5 and 8.8 rad/sec (the desired average value of 6.28 rad/sec is
shown as a dashed line). We note that at the initial time the angular velocity is zero. The generated torque by
the PM motor versus time is shown in Figure 4(b). We note that at the initial time when the system is at rest
the motor provides the largest amount of torque possible at around 31 N.m, and then it fluctuates around 4.5
N.m.

Figure 4: Angular velocity of the crank versus time for the web-cutter with a PM motor.

This example shows that a motor cannot provide the torque leading to the desired constant angular
velocity of the crank. However, the resultant angular velocity would fluctuate around the desired value.
Therefore, the question to ask is: can we keep the average angular velocity of the crank as close to the
desired value by keeping the fluctuation in the response as small as possible? The answer is: yes, if we add a
flywheel!

P. Nikravesh and M. Poursina 13-8

AME 352 DYNAMIC FORCE ANALYSIS

13.8 FLYWHEEL

Large fluctuations in velocity can be reduced if we increase the moment of inertia of the rotating input
link—this is called adding a flywheel to the system. A flywheel stores kinetic energy when the load on the
system is light and returns the energy to the system when the load demands larger input torque. To
demonstrate this, we increase the moment of inertia of the crank/motor significantly; e.g., 50 times, as
illustrated in Figure 5(a) by the addition of a flywheel to the crank/motor. We perform a forward dynamic
simulation with the motor selected previously with the torque-speed characteristic T = 6.8(5 − 4ω / (2π )) .
The result is shown in Figure 5(b) against that obtained without increasing the moment of inertia. The
response in dashed line represents the response of the web-cutter where the moment of inertia of the
crank/motor is 0.1 kg.m2 (same as in Figure 4(a)). The response in solid line is for the case where the
moment of inertia of the crank is increased to 5.0 kg.m2. We note that the fluctuation in the response is
reduced. We also note that with a larger moment of inertia for the crank/motor, it takes a longer time for the
system to gain the necessary speed.

(a) (b)

Figure 5: Angular velocity of the crank versus time for two different values of moment of inertia of the
crank.
A very large flywheel (a very large moment of inertia) will make the response much smoother but the
drawbacks are: (a) the system becomes very large in size and much heavier; (b) to start the system moving
from rest, a secondary motor may be needed for the speed-up phase. Therefore, we need to compromise
between the acceptable amount of fluctuation in the response and the size of the flywheel.
For the web-cutter mechanism, we consider the torque-angle response from Figure 2. The response provides
two areas below the average required torque, as shown in Figure 6. These two areas represent the portions of
the crank positions where the external forces; e.g., the gravity, add energy into the system. We consider the
larger of the two areas. We determine this area to represent E = 23.6 Jules. (The smaller area in yellow
represents 8.4 Jules.) If we perform a forward dynamic analysis, where the constant angular velocity of the
crank is not artificially enforced as a condition, the crank reaches its minimum angular velocity at the start
of the larger area, denoted as point S, and its maximum angular velocity at the end of the larger area, shown
as point E.

50
Required torque (N.m)

40

30

20 8.4 Jules 23.6 Jules

10 S E Average
torque
0

-10
0 2 4 6
Angle of the crank (rad)

P. Nikravesh and M. Poursina 13-9

AME 352 DYNAMIC FORCE ANALYSIS

Figure 6: Determining areas where kinetic energy is added to the web-cutter system.

Now we need to decide how much fluctuation is acceptable to us. For example assume that ±5%
fluctuation in the angular velocity of the crank is acceptable. This provides a coefficient of fluctuation as
k = 10% = 0.1 . The formula to determine the moment of inertia of the flywheel is [1]
E
I=
k ω avg
2

For our web-cutter example, we want the average angular velocity to be 6.28 rad/sec. This yields a moment
of inertia I = 23.6 / (0.1× 6.282 ) = 6.0 kg.m2. Note that this is the moment of inertia for the combined crank,
motor, and the flywheel.

P. Nikravesh and M. Poursina 13-10

AME 352 BALANCING

14 BALANCING
As we observed in the previous chapters, mechanisms or mechanical systems in general may contain
rotating components. The inertia of any rotating member can contribute to undesirable cyclic forces and
moments applied to the bearings and the foundation. Due to their direct relationship with centrifugal effects,
these forces and moments drastically increase as the rotational speed (RPM) of the element increases.
Therefore, if they are not completely or partially eliminated or minimized, the repeated loads result in cyclic
stress in the components of the system, and may cause them to fail by fatigue. To avoid such undesired
circumstances, rotating components should be balanced. If due to the manufacturing processes, any rotating
component is unbalanced, we can add/remove masses to/from the component to make it balanced.
The following two figures are examples of rotating objects that require balancing. Figure 14.1 shows
how material has been added or removed to balance camshaft and crankshaft in an engine. In another
example, as shown in Figure 14.2, small metallic masses can be added to a tire rim for balancing purposes.

(a) (b)

Figure 14.1: (a) Balancing of a camshaft [3], and (b) balancing of a crankshaft [4].

Figure 14.2: Balancing of a tire by adding small weights [5, 6]

14.1 STATIC AND DYNAMIC UNBALANCE

A rotating component can be unbalanced statically or dynamically. The associated unbalance is mainly due
to eccentricities. As shown in Figure 14.3, the unbalance that exists in this rotating component causes the
center of mass to move to a point that is not on the axis of rotation. This is called static unbalance, which
occurs when the center of mass of a rotating element is not located on the axis of rotation.

P. Nikravesh, M. Poursina 14-1

AME 352 BALANCING



G 










  

Figure 14.3: A statically unbalanced system [7]

As shown in the Figure 14.4, the unbalance that exists in this rotating component has moved the
center of mass to a point that is not on the axis of rotation. This angular misalignment between the principal
inertia axis and the axis of rotation causes a dynamic unbalance. The unbalanced dynamics cannot be
detected unless the component rotates.


G



  









Figure 14.4: A dynamically unbalanced rotating system [8]

14.2 SINGLE-PLANE (STATIC) BALANCING

Static balancing is the process of moving the mass center of a rotating object to be on the axis of rotation.
Consider a rotating shaft with three exaggerated extruded masses as shown in Figure 14.5. Assume the shaft
rotates about its axis by a constant angular velocity ω . The objective of the single-plane balance, also
known as static balance, is to make the shaking force to be zero, by adding another point-mass to the system
if necessary.

(a) (b)
Figure 14.5: Schematic representation of a rotating shaft with three extruded masses:
(a) side view; (b) front view.
For the static balancing we consider the front view (the view along the axis of rotation) of the
system. As shown in Figure 14.6(a), the orientation of each mass-point mi with respect to the x-axis, at a
given time, can be described by an angle θ i and a position vector R i . Each mass-point will have a normal
component of acceleration equal to Riω 2 pointing toward the shaft axis. The inertia force associated with
each mass is mi Riω 2 in the opposite direction of its corresponding acceleration. For static equilibrium, the
sum of these inertia forces must be zero in order to have zero shaking force. However, due to static
unbalance if we add these three inertia forces, the sum will not be zero:
−m1R1ω 2 − m2 R 2ω 2 − m3R 3ω 2 ≠ 0
To correct this unbalance, we add another mass point, for example mA as shown in Figure 14.6(b), in order
to have the following vector sum:
−m1R1ω 2 − m2 R 2ω 2 − m3R 3ω 2 − mA R Aω 2 = 0
We note that the coefficient ω 2 can be dropped from this equation to obtain
−m1R1 − m2 R 2 − m3R 3 − mA R A = 0 (14.1)

P. Nikravesh, M. Poursina 14-2

AME 352 BALANCING

(a) (b) (c)

Figure 14.6 : Front view of the rotating system (a) before balancing, (b) after balancing, and (c) sum of
inertia forces after balancing.

Since the force-equilibrium equation becomes independent of the angular velocity of the shaft, this process
is referred to as the static balancing although the shaft is in rotation. Equation (14.1) can be solved to
determine the balance mass for its magnitude and position either graphically or analytically.

Graphical solution: The graphical process is demonstrated in Figure 14.6(c). In this figure the inertia force
vectors associated with the mass points 1, 2, and 3 are constructed first knowing their magnitudes and
angles, then the solution vector mA R A that closes the vector loop provides the magnitude of mA RA and the
angle θ A . We note that there are infinite possibilities for mA and RA - we can choose a reasonable value for
one and then determine the other one.

Analytical solution: We can write Eq. (14.1) in a more general form as

n n
− ∑ mi R i − mA R A = 0 ⇒ mA R A = − ∑ mi R i (14.2)
i=1 i=1

where n = 3. This equation can be projected onto the x- and the y-axes to obtain two algebraic equations:
n n
mA RA( x ) = − ∑ mi Ri( x ) , mA RA( y) = − ∑ mi Ri( y)
i=1 i=1
These two equations result into
2 2
⎛ n ⎞ ⎛ n ⎞
mA RA = ⎜ ∑ mi Ri( x ) ⎟ + ⎜ ∑ mi Ri( y) ⎟ (14.3-a)
⎝ i=1 ⎠ ⎝ i=1 ⎠
n
− ∑ mi Ri( y)
θ A = arctan i=1
n
(14.3-b)
− ∑ mi Ri( x )
i=1

Example 14.1

In the system shown, find the mass-radius, and the angular m1

location of the added mass such that static balance is achieved.
R1
m1 = 1.0, m2 = 1.2, m3 = 1.0 kg,
m2 O θ1
R1 = 0.3, R2 = 0.5, R3 = 0.25 m, R2
θ1 = 30°, θ 2 = 120°, θ 3 = 235°
We implement Eqs. (14.3-a, b) in the following MATLAB R3
program: m3

m = [1.0,1.2,1.0];
R = [0.3,0.5,0.25];
theta = [30,120,235];

P. Nikravesh, M. Poursina 14-3

AME 352 BALANCING

% perform single-plane_balance
[mR_A, theta_A] = single_plane_balance(m, R, theta)

function [mR_A, theta_A] = single_plane_balance(m,R,theta)

% Single-plane (static) balance
mR = m.*R;
mRx = mR.*cosd(theta); mRy = mR.*sind(theta);
sigma_mRx = sum(mRx); sigma_mRy = sum(mRy);
mR_A = sqrt(sigma_mRx^2 + sigma_mRy^2);
theta_A = (atan2(-sigma_mRy, -sigma_mRx)*180/pi);
end

Executing the program provides the following results:

mA RA = 0.4998 kg.m, θ A = −68.4481°

14.3 TWO-PLANE (DYNAMIC) BALANCING

The objective of the dynamic force balance is to eliminate the shaking force and the shaking moment. For
this purpose, we need to add two balancing masses. First we add one mass to eliminate the shaking moment,
and then we add a second mass to eliminate the shaking force. The addition of the second mass is equivalent
to the process of performing a static balance.
To eliminate the shaking moment, we consider the side-view of the rotating component. We decide
on two correction planes, A and B, in convenient locations as shown in Figure 14.7(a). The plane of each
mass-point has a distance Li from the correction plane A. The distance of plane B from plane A is LB . The
inertia force of each mass, mi Riω 2 , causes a moment with respect to point O on the correction plane A as
mi Ri Liω 2 . This moment is perpendicular to vector R i and the rotational axis as shown in Figure 14.7(b) for
one of the masses. Therefore the correspond If we add the moments associated with the three masses, most
likely the sum is not zero, meaning that the system is dynamically unbalanced; i.e.,

m1R1 L1 2
m2 R 2 L2 2
m3R 3 L3 2  0

(a) (b)
Figure 14.7: (a) two correction plane are defined on the side-view of a rotating components; (b) the
corresponding inertia forces and their moments with respect to the correction plane A.

In order for the shaking moment to be zero, we add mass point to the correction plane B, where its
front view is shown in Figure 14.8(a). Now for these four mass points, the sum of the four moments must be
zero:
−m1R1L1ω 2 − m2 R2 L2ω 2 − m3R3 L3ω 2 − mB RB LBω 2 = 0
or,
−m1R1 L1 − m2 R 2 L2 − m3 R 3 L3 − mB R B LB = 0 (14.4)

This equation can be solved for mB R B LB either graphically or analytically.

P. Nikravesh, M. Poursina 14-4

AME 352 BALANCING

(a) (b)

Figure 14.8: (a) added mass in the B plane and (b) the corresponding sum of moments.

Graphical solution: The process of solving Eq. (14.4) is already demonstrated in Figure 14.8(b). In that
process every moment vector was rotated 90 o with respect to its position vector, in order to solve for the
solution vector mB R B LB . Then we have to rotate the solution by −90 o to obtain vector mB R B . We can
simplify this process by not rotating any of the vectors; i.e., we can draw the moment vectors along their
corresponding position vectors, as shown in Figure 14.9(a). Then we can construct the vector polygon
shown in Figure 14.9(b) that shows the solution vector mB R B LB along the position vector R B . We can
determine the magnitude mB RB LB by direct measurement from the polygon, and since LB is known, we can
determine the magnitude of mB RB (infinite possibilities for mB and the angle θ B can also be determined by
measurement from the figure. The result for the first balancing mass is shown in Figure 14.10.

Figure 14.9: Graphical solution for dynamic correction without rotating the moment vectors by 90° .

Note: If the correction plane A is placed such that the point masses are on both sides of the plane, then Li ’s
must be assigned positive and negative signs. This will affect the direction of a moment whether to be
considered along R i or in the opposite direction.

P. Nikravesh, M. Poursina 14-5

AME 352 BALANCING

Figure 14.10: The correction mass in plane B to eliminate the shaking moment.

Analytical solution: We can write Eq. (14.4) in a more general form as

n n
−∑ mi R i Li − mB R B LB = 0 ⇒ mB R B LB = −∑ mi R i Li (14.5)
i =1 i =1
This equation can be projected onto the x- and the y-axes to obtain two algebraic equations:
n n
mB RB( x ) LB = − ∑ mi Ri( x ) Li , mB RB( y ) LB = − ∑ mi Ri( y ) Li
i=1 i=1
These two equations result into
2 2
⎛ n ⎞ ⎛ n ⎞
⎜ ∑ i i ( x ) i ⎟ + ⎜ ∑ mi Ri ( y ) Li ⎟
m R L
⎝ i =1 ⎠ ⎝ i =1 ⎠
mB RB = (14.6-a)
LB
n
−∑ mi Ri ( y ) Li
θ B = arctan i =1
n
(14.6-b)
−∑ mi Ri ( x ) Li
i =1

Following the process of eliminating the shaking moment, we need to eliminate the shaking force.
To eliminate the shaking force, a second balancing mass is added in plane A according to the process of
single-plane balance. In this process the added mass at B must be included in Eq. (14.1) as any other
extruded mass as
m1R1 + m2 R 2 + m3R3 + mB R B + mAR A = 0
The graphical solution for our example is shown in Figure 14.11.
Correction
plane A

mA RA
mA RA
m2 R2
m3 R3

A
m1 R1
mB RB
mA RA

(a) (b) (c)

Figure 14.11: Adding a mass to the correction plane A to complete the process of two-plane balance.
P. Nikravesh, M. Poursina 14-6
AME 352 BALANCING

Example 14.2

For the following system, find the mass- radius, and the angular location of the added masses such that
dynamic balance is achieved.
m1 = 1.2, m2 = 1.8, m3 = 2.4 kg
R1 = 1.135, R2 = 0.822, R3 = 1.04 m
θ1 = 113.4°, θ 2 = 48.8°, θ3 = 251.4°
L1 = 0.85, L2 = 1.7, L3 = 2.4, LB = 3.0 m

The following MATLAB program can perform the analysis:

% Chapter 14: Example 14.2

m = [1.2, 1.8, 2.4];
R = [1.135, 0.822, 1.04];
theta = [113.4, 48.8, 251.4];
L = [0.85, 1.7, 2.4];
LB = 3.0;
% Two-plane balance
[mR_A_B, theta_A_B] = two_plane_balance(m, R, theta, L, LB)

function [mR_AB, theta_AB] = two_plane_balance(m, R, theta, L, LB)

% Two-plane (dynamic) balance
mRL = m.*R.*L;
mRLx = mRL.*cosd(theta); mRLy = mRL.*sind(theta);
sigma_mRLx = sum(mRLx); sigma_mRLy = sum(mRLy);
mR_B = sqrt(sigma_mRLx^2 + sigma_mRLy^2)/LB;
theta_B = (atan2(-sigma_mRLy, -sigma_mRLx)*180/pi);
[mR_A, theta_A] = single_plane_balance([m mR_B], [R 1],[theta theta_B]);
% Perform single-plane balance
mR_AB = [mR_A mR_B]; theta_AB = [theta_A theta_B];
end

Executing the program provides the following results:

mA RA = 0.9136 kg.m, θ A = −82.1646°
mB RB = 0.9381 kg.m, θ B = 75.3112°

Question: When we add a second mass to balance the system statically, wouldn’t that cause an unbalance in
the first step process that eliminated the shaking moment?

Answer: No, the second mass is added to the system in the correction plane A, which has a moment arm
n
LA = 0 , meaning that no new moment will be added to the computation of
m R L .
i i i
i=1

A Helpful Table

A table like the one shown can be helpful in organizing the data for hand calculation.

Mass Point m R θ L mR mR L
1
2
3
4


B
A 0 0

P. Nikravesh, M. Poursina 14-7

AME 352

REFERENCES

1. Norton, R.L., Design of Machinery, 4th ed., 2008, McGraw-Hill.

2. Norton, R.L., Design of Machinery, 1st ed., 199_, McGraw-Hill.

3. Uicker, J. J., Pennock, G. R., Shigley, J.E., Theory of Machines and Mechanisms, 5th ed., Oxford
University Press.

4. http://www.eatonbalancing.com/2012/12/01/camshaft-balancing/

5. https://www.howacarworks.com/crankshaft

6. http://www.aa1car.com/library/wheel_balancing.htm

7. https://www.indiamart.com/proddetail/steel-wheel-balancing-weights-2947523897.html

8. http://johnmaherracing.com/2013/01/project-2110-part-4-dynamic-balancing/

P. Nikravesh, M. Poursina