Ain Shams University (ASU)

Faculty of Engineering, Mechatronics Department

Industrial Robotics
(MCT-344)

Lect-08: Dynamics

Omar M. Shehata, PhD

Assistant Professor
Mechatronics Engineering Department, Faculty of Engineering,
Ain Shams University (ASU)

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 1
Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Outlines
• The Dynamics Problem
• Lagrangian Formulation
• RP Robot Dynamics
• Dynamics-based Control
• Next time .. ☺

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 2
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

-The Dynamics
Last time(s) ☺ 𝒁𝑮
Problem • Forward Position Kinematics using
the DH-Convention approach.
• Inverse Position Kinematics solution
using trigonometric and Numerical
solution for robots. 𝑬𝒏𝒅 𝒆𝒇𝒇𝒆𝒄𝒕𝒐𝒓

• Forward / Inverse Velocity Kinematics

and the Jacobian Matrix evaluation and 𝒀𝑮
importance in robotics study.
• Acceleration Kinematics
• Trajectory Planning task/Joint Space.
• Low/High-level Control (PID/FLC). 𝑿𝑮 Dynamics
Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 3
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

-The Dynamics
The Robot Forces:
Problem • For any robotic manipulator, we finalized the study of kinematics
(position, velocity and acceleration) and planning the trajectories and
controlling the robot motion.
• But, still we can ask some important questions such:
– If the robot material was (plastic, aluminum, steel),
would the 3 cases be the same?
– If the robot was carrying an object at its end-effector,
how can this be reflected in the equations?
– If other external forces acted upon the robot,
what would be their effect? Always resistive?

This is the Dynamics

Study of robotics
Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 4
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

-The Dynamics
Dynamics Map:
Problem • Thus, for any robotic manipulator, the Dynamics Study is a relation
between the joints space and the task space. But it is on the
Forces/Torques level. This map can be visualized as:

Robot’s
Weight
ሷ
𝝉 = 𝒇𝒄𝒏(𝑴, 𝒒) +
𝝉𝟏
𝝉𝟐 End-
𝝉𝟑 Effector’s
.. 𝑻𝒂𝒔𝒌 𝑺𝒑𝒂𝒄𝒆 load
.. 𝑱𝒐𝒊𝒏𝒕 𝑺𝒑𝒂𝒄𝒆 +
Joints’ Torques External
Forces
Dynamics Study
Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 5
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

-The Dynamics
Dynamics Study Approaches:
Problem • To study the Dynamics Problem of any robotic manipulator, we have
two main approaches that are commonly adopted:

Dynamics
Study

Newtonian Lagrangian Focus of this

Approach Approach Lecture
• This approach relies upon the • This approach utilizes the concepts of
Newtonian and Euler Equations of Lagrange of the energy analysis of the
motion for studying body dynamics. robot throughout its motion.
• It depends on the Analysis of each • It depends on the Calculation of the
body’s forces and moments. Energies of the robot.

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 6
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

-The Dynamics
Newtonian Dynamics: [Disqualified]
Problem • If we wanted to utilize the concepts of Newton in the study of
dynamics for any manipulator such as: 𝒀𝑮
(𝒙, 𝒚)
• Then to apply the Newtonian Approach
on the 1st link for example, we will have to
draw the Free-Body-Diagram (FBD): 𝑭𝑬.𝑬

𝑭𝟎𝟏,𝒙 𝑭𝟏𝟐,𝒙
𝒎𝟐 𝒈
𝒒𝟏
𝑭𝟏𝟐,𝒚
𝒎𝟏 𝒈 𝑿𝑮
𝑭𝟎𝟏,𝒚 𝒎𝟏 𝒈

Internal Forces 𝚺𝐅 = 𝐦𝐚 Accelerations

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 7
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

The Lagrange:
• Adopting the Lagrangian Approach for the dynamics study of robotics
is one of the most commonly approaches worldwide.
-Lagrangian
Formulation
• This approach focuses on the study of the Energy of the System
overall throughout its motion. And it relies upon the concept of
energy shift from one form to another.

• In an energy conservation system, we can define

the Lagrange (𝑳) as the difference between:
– Total Kinetic Energy (𝑻)
– Total Potential Energy (𝑼)

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 8
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

The Lagrange:
• Then, we can formulate the Lagrange as follows:
-Lagrangian 𝑳= 𝑻−𝑼
Formulation
• Where:
– 𝑻 is the Total Kinetic Energy of all the links in the Robot
𝑵
𝑻=෍ 𝑻𝒊 𝒇𝒐𝒓 𝒊 = 𝟏, 𝟐, . . , 𝑵
𝒊=𝟏
– 𝑼 is the Total Potential Energy of all the links in the Robot
𝑵
𝑼=෍ 𝑼𝒊 𝒇𝒐𝒓 𝒊 = 𝟏, 𝟐, . . , 𝑵
𝒊=𝟏

How can we Calculate the

Energies of each link?
Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 9
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

The Kinetic Energy: (Translational)

• For a body that is conducting pure translational motion, we learnt
earlier that we can calculate the Kinetic Energy as:
-Lagrangian
Formulation 𝟏
𝑻 = 𝒎𝒗𝟐
• Where: 𝟐
– (𝒎) is the body mass and (𝒗) is the body’s velocity. 𝒀𝑮

• But for robotics, the velocity is not scalar,

instead it is a vector (3X1),
• For a point:
𝑷 𝒙 𝒗 𝒙
𝑷𝒑𝟏 = 𝑷𝒚 𝒗𝒑𝟏 = 𝒗𝒚
𝑷𝒛 𝒗𝒛
𝑿𝑮
Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 10
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

The Kinetic Energy: (Translational)

• To achieve the same effect of velocity square (𝒗𝟐 ):
𝒗𝒙
-Lagrangian
Formulation 𝒗𝟐 = 𝒗𝑻 𝒗 = 𝒗𝒙 𝒗𝒚 𝒗𝒛 𝒗𝒚
𝒗𝒛
𝒗𝟐𝒙 + 𝒗𝟐𝒚 + 𝒗𝟐𝒛 = 𝒗𝟐

• For any link (𝒊), its weight can be summarized as a mass (𝒎𝒊 )
concentrated at its Center of Mass (C.O.M) in its center. This is done
based on the assumption of homogeneous mass distribution.

𝟏
𝑻𝒊,𝒕𝒓𝒂𝒏𝒔 = 𝒎𝒊 𝒗𝑻𝑪𝑴 𝒊 𝒗𝑪𝑴𝒊
𝟐

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 11
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

The Kinetic Energy: (Rotational)

• For a body that is conducting pure rotational motion, the Kinetic
Energy can be calculated through the equation:
-Lagrangian
Formulation 𝟏 𝟐
𝑻𝒓𝒐𝒕 = 𝑰𝝎
• Where: 𝟐
– (𝑰) is the body’s moment of inertia to rotate around
the axis of rotation and (𝝎) is the body’s angular velocity.
𝒀𝑮
• But for robotics, the angular velocity is not
scalar, instead it is a vector (3X1),
• For the 1st link: 𝟎
𝝎𝟏 = 𝟎
𝒒𝟏ሶ 𝒒𝟏 𝑿𝑮
Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 12
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

The Kinetic Energy: (Rotational)

• For the case of rotation, the moment of inertia differs based upon
which axis is the body rotating about. So, we define the Inertia Tensor
-Lagrangian
Formulation
Matrix as follows:
𝑰𝒙𝒊 𝑰𝒙𝒚 𝑰𝒙𝒛 𝒊
𝒊
𝑰𝒊 = 𝑰𝒚𝒙 𝒊 𝑰𝒚
𝒊
𝑰𝒚𝒛
𝒊
𝑰𝒛𝒙 𝒊 𝑰𝒛𝒚 𝑰𝒛 𝒊
𝒊
• So, putting all this together, we can reach out that the Rotational
Kinetic Energy for link (𝒊) can be calculated through the following
equation as shown:
𝟏 𝑻
𝑻𝒊,𝒓𝒐𝒕 = 𝝎
𝟐 𝒊𝟏𝑿𝟑
𝑰𝒊 𝟑𝑿𝟑 𝝎𝒊𝟑𝑿𝟏

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 13
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

The Potential Energy:

• The potential Energy is the energy due to gravitational force effects
on the masses when they are at a certain height from the frame of
-Lagrangian
Formulation
reference as follows:
• For link (𝒊):
ഥ 𝑷𝑪𝑴𝒊
𝑼𝒊 = 𝒎𝒊 𝒈
• Where:
– 𝒈) is the gravitational acceleration vector, with components of the gravitational
(ഥ
acceleration along the axes of the Global Frame.
For example, if gravity is pointing in the –ve direction of Z-axis:
ഥ = 𝟎 𝟎 −𝒈
𝒈 𝒈 = 𝟗. 𝟖𝟏 𝒎/𝒔𝟐
– (𝑷𝑪𝑴𝒊 ) is the position vector of the C.O.M. of link (𝒊) w.r.t the Global Fixed
frame, and (𝒎𝒊 ) is the mass of link (𝒊).

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 14
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Put the Lagrange to Use:

• After calculating the Kinetic Energy and Potential Energy for each link,
we can calculate the Lagrange, based on the Total K.E. and P.E.
-Lagrangian 𝑵 𝑵
Formulation • Where: 𝑳= 𝑻−𝑼 𝑻 = ෍ 𝑻𝒊 𝑼 = ෍ 𝑼𝒊
𝒊=𝟏 𝒊=𝟏
𝟏
𝑻𝒊,𝒕𝒓𝒂𝒏𝒔 = 𝒎𝒊 𝒗𝑻𝑪𝑴 𝒊 𝒗𝑪𝑴𝒊
𝟐
𝟏
𝑻𝒊,𝒓𝒐𝒕 = 𝝎𝑻𝒊𝟏𝑿𝟑 𝑰𝒊 𝟑𝑿𝟑 𝝎𝒊𝟑𝑿𝟏
𝟐
ഥ 𝑷𝑪𝑴𝒊
𝑼𝒊 = 𝒎𝒊 𝒈
• Then the required Torque required from each Joint (𝒊) can be
calculated using the formula:
𝒅 𝝏𝑳 𝝏𝑳
𝝉𝒊 = −
𝒅𝒕 𝝏𝒒ሶ 𝒊 𝝏𝒒𝒊
Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 15
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Case-Study
• In order to apply the Dynamics Study
using the Lagrangian Approach on a 𝒀𝑮
robotic manipulator, Let’s use the
following Revolute-Prismatic (RP)
-RP Robot robotic manipulator.
Dynamics
(𝒙, 𝒚)
• The RP configuration is presented
as shown next.
𝑿𝑮
• Let’s Study it from front view.

𝒁𝑮
Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 16
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

The Robot Parameters:

• After checking the previous configuration,
the plane view of this RP robot is as shown. 𝒀𝑮

-RP Robot • Now, we add two new parameters to the (𝒙, 𝒚)

Dynamics robotic manipulator that are essential for 𝒒𝟐
the Dynamics Study: 𝒎𝟏 𝒎𝟐
– Center of Mass (CM)
– Inertia Tensor (I)
𝒒𝟏 𝒍𝟏
𝑿𝑮
• These two parameters are defined for each body (link) in the robotic
manipulator and will be used in our study.

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 17
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Dynamics: [Solution]
• Apply the Lagrangian Approach for Dynamics study:
𝑳= 𝑻−𝑼 𝒀𝑮
• Where, for such robotic manipulator we
-RP Robot have two links (bodies) so we can say that: (𝒙, 𝒚)
Dynamics 𝑻 = 𝑻𝟏 + 𝑻𝟐 𝑼 = 𝑼𝟏 + 𝑼 𝟐 𝒒𝟐
• For the shown configuration, we can 𝒎𝟏 𝒎𝟐
conclude the position vector of both center
of masses as follows: 𝒒𝟏 𝒍𝟏
𝒍𝟏 𝒒𝟐 𝑿𝑮
𝑪 (𝒍𝟏 + ) 𝑪𝒒𝟏
𝟐 𝒒𝟏 𝟐
𝑷𝑪𝑴𝟏 = 𝒍𝟏 𝑷𝑪𝑴𝟐 = 𝒒𝟐
𝑺 (𝒍𝟏 + ) 𝑺𝒒𝟏
𝟐 𝒒𝟏 𝟐
𝟎 𝟎
Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 18
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Dynamics: [Solution]
• Starting of with the 1st link:
𝟏 𝑻 𝟏 𝑻 𝒀𝑮
𝑻𝟏 = 𝒎𝟏 𝒗𝑪𝑴𝟏 𝒗𝑪𝑴𝟏 + 𝝎𝟏 𝑰𝟏 𝝎𝟏
𝟐 𝟐
-RP Robot (𝒙, 𝒚)
Dynamics
• Let’s start with the translational term, 𝒒𝟐
the velocity of the C.O.M. is:
𝒍𝟏 𝒎𝟏 𝒎𝟐
𝑪𝒒𝟏 𝒍𝟏
− 𝑺𝒒𝟏 𝒒ሶ 𝟏
𝑷𝑪𝑴𝟏
𝟐
= 𝒍𝟏
𝟐 𝒒𝟏 𝒍𝟏
𝒗𝑪𝑴𝟏 = 𝒍𝟏
𝑺 𝑪 𝒒ሶ
𝟐 𝒒𝟏 𝟐 𝒒𝟏 𝟏 𝑿𝑮
𝟎 𝟎

𝟏
• So now, we can calculate the first term: 𝑻𝟏,𝒕𝒓𝒂𝒏𝒔 = 𝒎𝟏 𝒗𝑻𝑪𝑴𝟏 𝒗𝑪𝑴𝟏
𝟐

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 19
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Dynamics: [Solution]
• The translational K.E. of 1st link is:
𝟏 𝒀𝑮
𝑻𝟏,𝒕𝒓𝒂𝒏𝒔 = 𝒎𝟏 𝒗𝑻𝑪𝑴𝟏 𝒗𝑪𝑴𝟏
𝟐 𝒍𝟏
− 𝑺 𝒒ሶ
𝟏 𝟐 𝒒𝟏 𝟏
-RP Robot 𝑻𝟏,𝒕𝒓𝒂𝒏𝒔 =
𝒍
𝒎𝟏 − 𝟏 𝑺𝒒𝟏 𝒒ሶ 𝟏
𝒍𝟏
𝑪 𝒒ሶ 𝟎 𝒍𝟏 (𝒙, 𝒚)
Dynamics 𝟐 𝟐 𝟐 𝒒𝟏 𝟏 𝑪 𝒒ሶ 𝒒𝟐
𝟐 𝒒𝟏 𝟏
𝟎
𝟏 𝒍𝟐𝟏 𝟐 𝟐 𝒍𝟐𝟏 𝟐 𝟐 𝒎𝟏
𝑻𝟏,𝒕𝒓𝒂𝒏𝒔 = 𝒎𝟏 [ 𝑺𝒒𝟏 𝒒ሶ 𝟏 + 𝑪𝒒𝟏 𝒒ሶ 𝟏 + 𝟎] 𝒎𝟐
𝟐 𝟒 𝟒
𝒒𝟏 𝒍𝟏
• But we know that: 𝑺𝟐𝜽 + 𝑪𝟐𝜽 = 𝟏
• So, we can conclude that: 𝑿𝑮
𝟏 𝒍𝟐𝟏 𝟐 𝒎𝟏 𝒍𝟐𝟏 𝟐
𝑻𝟏,𝒕𝒓𝒂𝒏𝒔 = 𝒎𝟏 𝒒ሶ = 𝒒ሶ 𝟏
𝟐 𝟒 𝟏 𝟖

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 20
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Dynamics: [Solution]
• For the rotational K.E. of 1st link is:
𝟏 𝑻 𝒀𝑮
𝑻𝟏,𝒓𝒐𝒕 = 𝝎 𝑰 𝝎
𝟐 𝟏 𝟏 𝟏 𝟎
• But we know that for the 1st link: 𝝎𝟏 = 𝟎
-RP Robot 𝒒ሶ 𝟏 (𝒙, 𝒚)
• Similarly, the inertia tensor for the 1st
Dynamics 𝒒𝟐
link as a rigid body is: 𝑰𝒙 𝑰𝒙𝒚 𝑰𝒙𝒛 𝟏 𝟏 𝟏
𝑰𝟏 = 𝑰𝒚𝒙 𝟏 𝑰𝒚 𝑰𝒚𝒛 𝒎𝟏 𝒎𝟐
𝟏 𝟏
𝑰𝒛𝒙 𝟏 𝑰𝒛𝒚
𝟏
𝑰𝒛 𝟏 𝒒𝟏 𝒍𝟏
• Thus, we can write it as: 𝑿𝑮
𝑰𝒙𝟏 𝑰𝒙𝒚 𝑰𝒙𝒛 𝟏 𝟎
𝟏
𝟏 𝑻 𝟏
𝑻𝟏,𝒓𝒐𝒕 = 𝝎𝟏 𝑰𝟏 𝝎𝟏 = 𝟎 𝟎 𝒒ሶ 𝟏 𝑰𝒚𝒙 𝑰𝒚 𝑰𝒚𝒛 𝟎
𝟐 𝟐 𝟏 𝟏 𝟏
𝑰𝒛𝒙 𝟏 𝑰𝒛𝒚 𝑰𝒛 𝟏 𝒒ሶ 𝟏
𝟏

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 21
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Dynamics: [Solution]
• This can be further simplified as:
𝑰𝒙𝟏 𝑰𝒙𝒚
𝟏
𝑰𝒙𝒛 𝟏 𝟎 𝒀𝑮
𝟏
𝑻𝟏,𝒓𝒐𝒕 = 𝟎 𝟎 𝒒ሶ 𝟏 𝑰𝒚𝒙 𝑰𝒚 𝑰𝒚𝒛 𝟎
𝟐 𝟏 𝟏 𝟏
𝑰𝒛𝒙 𝟏 𝑰𝒛𝒚 𝑰𝒛 𝟏 𝒒ሶ 𝟏
-RP Robot
𝟏 (𝒙, 𝒚)
𝑰𝒙𝒛 𝟏 𝒒ሶ 𝟏
Dynamics 𝟏 𝒒𝟐
𝑻𝟏,𝒓𝒐𝒕 = 𝟎 𝟎 𝒒ሶ 𝟏 𝑰𝒚𝒛 𝒒ሶ 𝟏
𝟐 𝟏
𝑰𝒛 𝟏 𝒒ሶ 𝟏 𝒎𝟏 𝒎𝟐
𝟏
= 𝑰𝒛 𝟏 𝒒ሶ 𝟐𝟏
𝒒𝟏 𝒍𝟏
• Thus, we can reach: 𝑻𝟏,𝒓𝒐𝒕 𝟐
• Combining both translational and rotational K.E.s for the 1st link: 𝑿𝑮
𝟏 𝑻 𝟏 𝑻 𝒎𝟏 𝒍𝟐𝟏 𝟏
𝑻𝟏 = 𝒎𝟏 𝒗𝑪𝑴𝟏 𝒗𝑪𝑴𝟏 + 𝝎𝟏 𝑰𝟏 𝝎𝟏 𝑻𝟏 = [ + 𝑰𝒛 𝟏 ]𝒒ሶ 𝟐𝟏
𝟐 𝟐 𝟖 𝟐

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 22
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Dynamics: [Solution]
• Next we move on to the Potential Energy
of the 1st link as well: 𝒀𝑮
ഥ 𝑷𝑪𝑴𝟏
𝑼𝟏 = 𝒎𝟏 𝒈
-RP Robot • Where, we have: 𝒈ഥ = 𝟎 −𝒈 𝟎 (𝒙, 𝒚)
Dynamics and: 𝒈 = 𝟗. 𝟖𝟏 𝒎/𝒔𝟐 𝒒𝟐
𝒎𝟏 𝒎𝟐
𝒍𝟏
• Thus, we can find that: 𝑪
𝟐 𝒒𝟏 𝒒𝟏 𝒍𝟏
𝑼𝟏 = 𝒎𝟏 𝟎 −𝒈 𝟎 𝒍𝟏
𝑺
𝟐 𝒒𝟏 𝑿𝑮
𝟎
𝒍𝟏 𝒎 𝟏 𝒍𝟏 𝒈
• So, we can reach that: 𝑼𝟏 = −𝒎𝟏 𝒈 𝑺𝒒𝟏
𝟐
𝑼𝟏 = −
𝟐
𝑺𝒒𝟏

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 23
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Dynamics: [Solution]
• Next we move on to the 2nd link:
𝟏 𝑻 𝟏 𝑻 𝒀𝑮
𝑻𝟐 = 𝒎𝟐 𝒗𝑪𝑴𝟐 𝒗𝑪𝑴𝟐 + 𝝎𝟐 𝑰𝟐 𝝎𝟐
𝟐 𝟐
-RP Robot (𝒙, 𝒚)
Dynamics
• Let’s start with the translational term, 𝒒𝟐
the velocity of the C.O.M. is:
𝒒𝟐 𝒒𝟐 𝟏 𝒎𝟏 𝒎𝟐
(𝒍𝟏 + ) 𝑪𝒒𝟏 − 𝒍𝟏 + 𝑺 𝒒ሶ + 𝑪 𝒒ሶ
𝟐 𝒒𝟏 𝟏 𝟐 𝒒𝟐 𝟐
𝒒𝟏 𝒍𝟏
𝟐
𝑷𝑪𝑴𝟐 = 𝒒𝟐 𝒗𝑪𝑴𝟐 = 𝒒𝟐 𝟏
(𝒍𝟏 + ) 𝑺𝒒𝟏 𝒍𝟏 + 𝑪𝒒𝟏 𝒒𝟏 + 𝑺𝒒𝟐 𝒒ሶ 𝟐
ሶ
𝟐 𝟐 𝟐
𝟎 𝟎 𝑿𝑮

𝟏
• So now, we can calculate the first term: 𝑻𝟐,𝒕𝒓𝒂𝒏𝒔 = 𝒎𝟐 𝒗𝑻𝑪𝑴𝟐 𝒗𝑪𝑴𝟐
𝟐

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 24
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Dynamics: [Solution]
• The translational K.E. of 2nd link is:
𝟏
𝑻𝟐,𝒕𝒓𝒂𝒏𝒔 = 𝒎𝟐 𝒗𝑻𝑪𝑴𝟐 𝒗𝑪𝑴𝟐
𝟐 𝒒𝟐 𝟏
− 𝒍𝟏 + 𝑺𝒒𝟏 𝒒ሶ 𝟏 + 𝑪𝒒𝟐 𝒒ሶ 𝟐
𝒎𝟐 𝒒𝟐 𝟏 𝒒𝟐 𝟏 𝟐 𝟐
-RP Robot = − 𝒍𝟏 + 𝑺 𝒒ሶ + 𝑪 𝒒ሶ 𝒍𝟏 + 𝑪𝒒𝟏 𝒒ሶ 𝟏 + 𝑺𝒒𝟐 𝒒ሶ 𝟐 𝟎 𝒒𝟐 𝟏
Dynamics 𝟐 𝟐 𝒒𝟏 𝟏 𝟐 𝒒𝟐 𝟐 𝟐 𝟐 𝒍𝟏 + 𝑪𝒒𝟏 𝒒ሶ 𝟏 + 𝑺𝒒𝟐 𝒒ሶ 𝟐
𝟐 𝟐
𝟐
𝟎
𝟏 𝒒𝟐 𝟏 𝟏 𝒒𝟐
𝑻𝟐,𝒕𝒓𝒂𝒏𝒔 = 𝒎𝟐 [ 𝒍𝟏 + 𝑺𝟐𝒒𝟏 𝒒ሶ 𝟐𝟏 + 𝑪𝟐𝒒𝟏 𝒒ሶ 𝟐𝟐 − 𝟐 ∗ 𝒍𝟏 + 𝑺𝒒𝟏 𝑪𝒒𝟏 𝒒ሶ 𝟏 𝒒ሶ 𝟐 +
𝟐 𝟐 𝟒 𝟐 𝟐
𝒒𝟐 𝟐 𝟏 𝟏 𝒒𝟐
𝒍𝟏 + 𝑪𝟐𝒒𝟏 𝒒ሶ 𝟐𝟏 + 𝑺𝟐𝒒𝟏 𝒒ሶ 𝟐𝟐 + 𝟐 ∗ 𝒍𝟏 + 𝑺 𝑪 𝒒ሶ 𝒒ሶ ]
𝟐 𝟒 𝟐 𝟐 𝒒𝟏 𝒒𝟏 𝟏 𝟐
• So, we can conclude that: 𝑺𝟐𝜽 + 𝑪𝟐𝜽 = 𝟏
𝟏 𝒒𝟐 𝟐 𝟏 𝟐
𝑻𝟐,𝒕𝒓𝒂𝒏𝒔 = 𝒎𝟐 𝒍𝟏 + 𝒒ሶ 𝟐𝟏 + 𝒒ሶ 𝟐
𝟐 𝟐 𝟒

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 25
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Dynamics: [Solution]
• For the rotational K.E. of 2nd link is:
𝟏 𝑻 𝒀𝑮
𝑻𝟐,𝒓𝒐𝒕 = 𝝎 𝑰 𝝎
𝟐 𝟐 𝟐 𝟐
• We know that the inertia tensor for the 2nd
-RP Robot link as a rigid body is: (𝒙, 𝒚)
𝑰 𝑰 𝑰
Dynamics 𝒙𝟐 𝒙𝒚 𝟐 𝒙𝒛 𝟐
𝒒𝟐
𝑰𝟐 = 𝑰𝒚𝒙 𝟐 𝑰𝒚
𝟐
𝑰𝒚𝒛
𝟐
𝑰𝒛𝒙 𝟐 𝑰𝒛𝒚
𝟐
𝑰𝒛 𝟐 𝒎𝟏 𝒎𝟐
• But for the rotational speed of the C.O.M. 𝒒𝟏 𝒍𝟏
of the 2nd link, we must consider all the links
before it that might contribute to its rotation, through: 𝑿𝑮

𝝎𝟐 = 𝟎𝝎𝟐 = 𝟎
𝝎𝟏 + 𝟎𝑹𝟏 𝟏𝝎𝟐

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 26
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Dynamics: [Solution]
Ang. Vel of link 2 w.r.t 1st link
• So, we can calculate this as follows:
𝝎𝟐 = 𝟎𝝎𝟐 = 𝟎
𝝎𝟏 + 𝟎𝑹𝟏 𝟏𝝎𝟐 𝒀𝑮
1st link Ang. Vel

-RP Robot Rot. Matrix of Frame (1) w.r.t. Frame (0) (𝒙, 𝒚)
Dynamics 𝟎 −𝑺𝒒𝟏 𝟎 𝑪𝒒! 𝒒𝟐
𝟎
• Where: 𝝎𝟏 = 𝟎 𝑹𝟏 = 𝑪𝒒𝟏 𝟎 𝑺𝒒𝟏
𝒒ሶ 𝟏 𝟎 𝟏 𝟎 𝒎𝟏 𝒎𝟐
𝟏
𝟎
𝒒𝟏 𝒍𝟏
• And since 2nd Joint is Prismatic: 𝝎𝟐 = 𝟎
𝟎
• Therefore: 𝑿𝑮
𝟎 −𝑺𝒒𝟏 𝟎 𝑪𝒒! 𝟎 𝟎
𝝎𝟐 = 𝟎𝝎𝟐 = 𝟎 + 𝑪𝒒𝟏
𝒒ሶ 𝟏
𝟎 𝑺𝒒𝟏 𝟎 = 𝟎
𝟎 𝒒ሶ 𝟏
It is Rotating!!
𝟎 𝟏 𝟎

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 27
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Dynamics: [Solution]
• So the rotational K.E. of 2nd link is:
𝟏 𝑻 𝒀𝑮
𝑻𝟐,𝒓𝒐𝒕 = 𝝎 𝑰 𝝎
𝟐 𝟐 𝟐 𝟐 𝟎
• We found out that for the 2nd link 𝝎𝟐 = 𝟎
-RP Robot 𝒒ሶ 𝟏 (𝒙, 𝒚)
• Similarly, the inertia tensor for the 2nd
Dynamics 𝒒𝟐
link as a rigid body is: 𝑰𝒙𝟐 𝑰𝒙𝒚 𝑰𝒙𝒛 𝟐
𝟐
𝑰𝟐 = 𝑰𝒚𝒙 𝟐 𝑰𝒚 𝑰𝒚𝒛 𝒎𝟏 𝒎𝟐
𝟐 𝟐
𝑰𝒛𝒙 𝟐 𝑰𝒛𝒚
𝟐
𝑰𝒛 𝟐
𝒒𝟏 𝒍𝟏
• Thus, we can write it as: 𝑿𝑮
𝑰 𝒙𝟐 𝑰𝒙𝒚 𝑰𝒙𝒛 𝟐 𝟎
𝟐
𝟏 𝟏 𝟏
𝑻𝟐,𝒓𝒐𝒕 = 𝝎𝑻𝟐 𝑰𝟐 𝝎𝟐 = 𝟎 𝟎 𝒒ሶ 𝟏 𝑰𝒚𝒙 𝑰𝒚 𝑰𝒚𝒛 𝟎 = 𝑰𝒛 𝟐 𝒒ሶ 𝟐𝟏
𝟐 𝟐 𝟐 𝟐 𝟐 𝟐
𝑰𝒛𝒙 𝟐 𝑰𝒛𝒚 𝑰𝒛 𝟐 𝒒ሶ 𝟏
𝟐

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 28
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Dynamics: [Solution]
• So, for the 2nd link we evaluated that the
translational K.E. is: 𝒀𝑮
𝟏 𝒒𝟐 𝟐 𝟏 𝟐
𝑻𝟐,𝒕𝒓𝒂𝒏𝒔 = 𝒎𝟐 𝒍𝟏 + 𝒒ሶ 𝟐𝟏 + 𝒒ሶ 𝟐
𝟐 𝟐 𝟒
-RP Robot (𝒙, 𝒚)
Dynamics 𝒒𝟐
• While the rotational K.E. is:
𝟏
𝑻𝟐,𝒓𝒐𝒕 = 𝑰𝒛 𝟐 𝒒ሶ 𝟐𝟏 𝒎𝟏 𝒎𝟐
𝟐
• Thus, we can reach by combining both 𝒒𝟏 𝒍𝟏
translational and rotational K.E.s for
the 2nd link: 𝑿𝑮
𝟏 𝒒𝟐 𝟐 𝟏 𝟐 𝟏
𝑻𝟐 = 𝒎 𝟐 𝒍𝟏 + 𝒒ሶ 𝟐𝟏 + 𝒒ሶ 𝟐 + 𝑰𝒛 𝟐 𝒒ሶ 𝟐𝟏
𝟐 𝟐 𝟒 𝟐

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 29
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Dynamics: [Solution]
• Next we move on to the Potential Energy
of the 2nd link as well: 𝒀𝑮
ഥ 𝑷𝑪𝑴𝟐
𝑼𝟐 = 𝒎𝟐 𝒈
-RP Robot • Where, we have: 𝒈ഥ = 𝟎 −𝒈 𝟎 (𝒙, 𝒚)
Dynamics and: 𝒈 = 𝟗. 𝟖𝟏 𝒎/𝒔𝟐 𝒒𝟐
𝒎𝟏 𝒎𝟐
𝒒𝟐
• Thus, we can find that: (𝒍𝟏 + ) 𝑪𝒒𝟏
𝟐 𝒒𝟏 𝒍𝟏
𝑼𝟐 = 𝒎𝟐 𝟎 −𝒈 𝟎 𝒒𝟐
(𝒍𝟏 + ) 𝑺𝒒𝟏
𝟐 𝑿𝑮
𝟎

• So, we can reach that: 𝒒𝟐

𝑼𝟐 = −𝒎𝟐 𝒈(𝒍𝟏 + ) 𝑺𝒒𝟏
𝟐

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 30
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Dynamics: [Solution]
• After evaluating all the terms, we move on to write down the
Lagrange Equation:
𝑳= 𝑻−𝑼
-RP Robot 𝑻 = 𝑻𝟏 + 𝑻𝟐 𝑼 = 𝑼𝟏 + 𝑼𝟐
Dynamics
• Therefore:
𝟐
𝒎 𝟏 𝒍𝟏 𝟏 𝟏 𝒒𝟐 𝟐 𝟏 𝟏
𝑳= +
𝟏
𝟐
𝑰𝒛 𝒒ሶ 𝟏 + 𝒎𝟐 𝒍𝟏 +
𝟐
𝒒ሶ 𝟏 +
𝟐
𝒒ሶ 𝟐
𝟐
+ 𝑰𝒛 𝒒ሶ 𝟏
𝟖 𝟐 𝟐 𝟐 𝟒 𝟐 𝟐
𝒎 𝟏 𝒍𝟏 𝒈 𝒒𝟐
+ 𝑺𝒒𝟏 + 𝒎𝟐 𝒈(𝒍𝟏 + ) 𝑺𝒒𝟏
𝟐 𝟐

• Now, to find the torques, we will use: 𝒅 𝝏𝑳 𝝏𝑳

𝝉𝒊 = −
ሶ
𝒅𝒕 𝝏𝒒𝒊 𝝏𝒒𝒊

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 31
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Dynamics: [Solution] 𝝉𝒊 =
𝒅 𝝏𝑳
−
𝝏𝑳
• Let’s start with 1st Joint: 𝒅𝒕 𝝏𝒒ሶ 𝒊 𝝏𝒒𝒊
𝟐
𝒎 𝟏 𝒍𝟏 𝟏 𝟏 𝒒𝟐 𝟐 𝟏 𝟏 𝒎 𝟏 𝒍𝟏 𝒈 𝒒𝟐
𝑳= +
𝟏
𝟐
𝑰𝒛 𝒒ሶ 𝟏 + 𝒎𝟐 𝒍𝟏 +
𝟐
𝒒ሶ 𝟏 +
𝟐
𝒒ሶ 𝟐 +
𝟐
𝑰𝒛 𝒒ሶ 𝟏
𝟐
+ 𝑺𝒒 + 𝒎𝟐 𝒈(𝒍𝟏 +
𝟏
) 𝑺𝒒
𝟏
𝟖 𝟐 𝟐 𝟐 𝟒 𝟐 𝟐 𝟐
-RP Robot
Dynamics
• Thus: 𝝏𝑳 𝒎𝟏 𝒍𝟐𝟏 𝑰𝒛 𝟏 𝒎𝟐 𝒒𝟐 𝟐 𝟏
= 𝟐∗ + + 𝒍𝟏 + + 𝑰𝒛 𝟐 𝒒ሶ 𝟏
𝝏𝒒ሶ 𝟏 𝟖 𝟐 𝟐 𝟐 𝟐

• Then: 𝒅 𝝏𝑳 𝒅 𝒎𝟏 𝒍𝟐𝟏 𝒒𝟐 𝟐
= + 𝑰𝒛 𝟏 + 𝒎 𝟐 𝒍𝟏 + + 𝑰𝒛𝟐 𝒒ሶ 𝟏
𝒅𝒕 𝝏𝒒ሶ 𝟏 𝒅𝒕 𝟒 𝟐
𝒎𝟏 𝒍𝟐𝟏 𝒒𝟐 𝟐 𝒒𝟐
= + 𝑰𝒛 𝟏 + 𝒎 𝟐 𝒍𝟏 + + 𝑰𝒛𝟐 𝒒ሷ 𝟏 + 𝟐𝒎𝟐 𝒍𝟏 + 𝒒ሶ 𝟏 𝒒ሶ 𝟐
𝟒 𝟐 𝟐
• Finally: 𝝏𝑳 𝒎 𝟏 𝒍𝟏 𝒈 𝒒𝟐
= + 𝒎𝟐 𝒈(𝒍𝟏 + ) 𝑪𝒒𝟏
𝝏𝒒𝟏 𝟐 𝟐

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 32
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Dynamics: [Solution] 𝝉𝒊 =
𝒅 𝝏𝑳
−
𝝏𝑳
• Follow the same procedure for the 2nd Joint: 𝒅𝒕 𝝏𝒒ሶ 𝒊 𝝏𝒒𝒊
𝟐
𝒎 𝟏 𝒍𝟏 𝟏 𝟏 𝒒𝟐 𝟐 𝟏 𝟏 𝒎 𝟏 𝒍𝟏 𝒈 𝒒𝟐
𝑳= +
𝟏
𝟐
𝑰𝒛 𝒒ሶ 𝟏 + 𝒎𝟐 𝒍𝟏 +
𝟐
𝒒ሶ 𝟏 +
𝟐
𝒒ሶ 𝟐 +
𝟐
𝑰𝒛 𝒒ሶ 𝟏
𝟐
+ 𝑺𝒒 + 𝒎𝟐 𝒈(𝒍𝟏 +
𝟏
) 𝑺𝒒
𝟏
𝟖 𝟐 𝟐 𝟐 𝟒 𝟐 𝟐 𝟐
-RP Robot
Dynamics
• Thus: 𝝏𝑳 𝒎𝟐
=𝟐∗ 𝒒ሶ 𝟐
𝝏𝒒ሶ 𝟐 𝟖

• Then:
𝒅 𝝏𝑳 𝒅 𝒎𝟐 𝒎𝟐
= 𝒒ሶ 𝟐 = 𝒒ሷ 𝟐
𝒅𝒕 𝝏𝒒ሶ 𝟐 𝒅𝒕 𝟒 𝟒

• Finally:
𝝏𝑳 𝒎𝟐 𝒒ሶ 𝟐𝟏 𝒒𝟐 𝟏 𝟏
= ∗ 𝟐 𝒍𝟏 + ( ) + 𝒎𝟐 𝒈( ) 𝑺𝒒𝟏
𝝏𝒒𝟐 𝟐 𝟐 𝟐 𝟐

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 33
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Dynamics: [Solution] 𝝉𝒊 =
𝒅 𝝏𝑳
−
𝝏𝑳
• From the last two slides, we can conclude: 𝒅𝒕 𝝏𝒒ሶ 𝒊 𝝏𝒒𝒊

• For the 1st Joint:

𝒎𝟏 𝒍𝟐𝟏 𝒒𝟐 𝟐 𝒒𝟐
𝝉𝟏 = + 𝑰𝒛 𝟏 + 𝒎 𝟐 𝒍𝟏 + + 𝑰𝒛𝟐 𝒒ሷ 𝟏 + 𝟐𝒎𝟐 𝒍𝟏 + 𝒒ሶ 𝟏 𝒒ሶ 𝟐
-RP Robot 𝟒 𝟐 𝟐
Dynamics 𝒎 𝟏 𝒍𝟏 𝒈 𝒒𝟐
− + 𝒎𝟐 𝒈(𝒍𝟏 + ) 𝑪𝒒𝟏
𝟐 𝟐
• For the 2nd Joint:
𝒎𝟐 𝒎𝟐 𝒒ሶ 𝟐𝟏 𝒒𝟐 𝒎𝟐 𝒈
𝝉𝟐 = 𝒒ሷ 𝟐 − 𝒍𝟏 + − 𝑺𝒒𝟏
𝟒 𝟐 𝟐 𝟐

Now, Let’s put it in Matrix Form ☺

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 34
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Dynamics: [Solution]
• Let’s put it in Matrix Form:
𝒀𝑮
𝒎𝟏 𝒍𝟐𝟏 𝒒𝟐 𝟐
𝝉𝟏 𝟒
+ 𝑰𝒛 𝟏 + 𝒎 𝟐 𝒍𝟏 +
𝟐
+ 𝑰𝒛𝟐 𝟎 𝒒ሷ 𝟏
-RP Robot 𝝉𝟐 = 𝒎𝟐
𝒒ሷ 𝟐
+ (𝒙, 𝒚)
𝟎
Dynamics 𝟒 𝒒𝟐
𝟎 𝟎 𝒒ሶ 𝟐𝟏
𝒎𝟐 𝒒𝟐 + 𝒎𝟏 𝒎𝟐
− 𝒍 +
𝟐 𝟏 𝟐
𝟎
𝒒ሶ 𝟐𝟐
𝒒𝟐 𝒒𝟏 𝒍𝟏
𝟐𝒎𝟐 𝒍𝟏 +
𝟐 𝒒ሶ 𝟏 𝒒ሶ 𝟐 +
𝟎 𝑿𝑮
𝒎 𝟏 𝒍𝟏 𝒈 𝒒𝟐
− + 𝒎𝟐 𝒈(𝒍𝟏 + ) 𝑪𝒒𝟏
𝟐 𝟐
𝒎𝟐 𝒈
𝑺𝒒𝟏
𝟐
Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 35
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Dynamics: [Solution]
• Which could be written in a simplified form
as the following form: 𝒀𝑮
𝝉 = 𝑴 𝒒 𝒒ሷ + 𝑪 𝒒, 𝒒ሶ 𝒒ሶ + 𝑮(𝒒)
-RP Robot (𝒙, 𝒚)
Dynamics 𝒒𝟐
Inertia/Mass Non-Linear Forces Gravitational
Matrix (Centrifugal + Coriolis) Forces 𝒎𝟏 𝒎𝟐
• Similar to: 𝑭 = 𝒎𝒙ሷ + 𝒃𝒙ሶ + 𝒌𝒙 𝒒𝟏 𝒍𝟏
• Which is the same form as: 𝑿𝑮
𝚺𝐅 = 𝐦𝐚
Let’s Simulate this ☺
Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 36
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Dynamics Simulation
• Now, what if we want to drive a simulated robot’s end-effector to a
desired position (x,y,z). Then we can visualize this as follows:

Desired
E.E. Pose 𝑿𝒆𝒓𝒓 𝝉 𝒒ሷ 𝒒ሶ 𝒒𝒂𝒄𝒕
+ + 𝟏 𝟏
-Dynamics-based Cntrl. 𝑴−𝟏
(FLC) − 𝒔 𝒔
Control
𝑿𝒅𝒆𝒔 − − Actual
ሶ
𝑪(𝒒, 𝒒) Joints’
Angles
Actual
𝑿𝒂𝒄𝒕 E.E. Pose
𝑮(𝒒)

𝑿 Fwd. Pos. Kin 𝒒

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 37
 Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

Next time:
• In the next Lecture we will cover:
– Effect of External Forces
– Null-Space Control
-Next time
• In the next tutorial we will cover:
– Dynamics Study Problems

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 38
Ain Shams University (ASU)
Faculty of Engineering, Mechatronics Department

See you Next time … ☺

Spring '22 Industrial Robotics (MCT-344) - Spring '22 Dr. Omar M. Shehata (c) 39