ME 240

Dynamic and Control of Mechanical Systems

Lecture-1

G R Jayanth, [email protected]
 Motion of a rigid body
• A “rigid body” is an idealization of a solid body wherein it is assumed that the body does not deform under the influence of
mechanical loads, viz., forces and moments.
• Like all idealizations, a rigid body is just a model of a solid body and does not fully capture the behavior of a solid object Y
under mechanical load : all solid materials deform under mechanical load (most often elastically first, plastically later, before
finally fracturing). However, in some scenarios, such as when the forces are small, and the body is not fully constrained, the
deformation of the body can be neglected while understanding its mechanics. In such a case, the rigid body model captures
the behavior of the solid well enough for it to be useful for engineering purposes.
QP
Q P
• Mathematically, if B is a rigid object, and Q and P are two points fixed to the object, with position vectors Q and P
respectively, then the length of the vector QP= P-Q does not change when the body moves under the influence of
mechanical loads. Let Q= 𝑄𝑥 ෝ 𝒙 + 𝑄𝑦 𝒚ෝ + 𝑄𝑧 𝒛ො , and 𝑷 = 𝑃𝑥 ෝ ෝ + 𝑃𝑧 𝒛ො , in the conventional unit-vector notation for
𝒙 + 𝑃𝑦 𝒚 B
representing a vector. Then, the implication that the length does not change is equivalent to saying that
2 2
|𝑸𝑷| = 𝑃𝑥 − 𝑄𝑥 + 𝑃𝑦 − 𝑄𝑦 + 𝑃𝑧 − 𝑄𝑧 2 =Constant X
• We note that |𝑸𝑷|𝟐 = 𝑸𝑷. 𝑸𝑷, i.e., the dot product of 𝑸𝑷 with itself. Since |𝑸𝑷| is a constant, 𝑑|𝑸𝑷|𝟐 Τ𝑑𝑡 =
2𝑸𝑷. 𝑸𝑷ሶ =0, i.e.,𝑸𝑷. 𝑸𝑷=0.
ሶ ሶ = 𝒗𝑷 − 𝒗𝑸 cannot have a component
This implies that the relative velocity vector 𝒗𝑷|𝑸 = 𝑸𝑷
along 𝑸𝑷.
• There are two ways in which this can happen: the first is when 𝑸𝑷 ሶ = 𝟎, i.e., when the linear velocity of point B and point A
are exactly the same. This type of motion is called pure translation.
• In the first way, the vector 𝑸𝑷 does not change. The second way is obtained by noticing that the vector𝑸𝑷 can change but
in a manner that keeps the length of 𝑸𝑷 unchanged. A moment’s thought would reveal that the only type of relative motion
possible between P and Q is the rotation of point P about point Q or vice versa: it keeps the distance between the two
unchanged but changes the orientation of the vector. This is the case of pure rotation.
• Any linear combination of the two types of motion, viz., pure translation and pure rotation about point is also a valid way
for the rigid body to move. Thus, given any point on the rigid body, the motion of a rigid body can always be expressed as a
sum of pure translation of that point and pure rotation about that point.
• Since this is true for both points P and Q, it may lead to the conundrum of how the body can simultaneously be seen to be
rotating about both points P and Q. The answer is that it is possible to view it this way since the two points are allowed to
possess different linear velocities.
 Y
Rotational motion of a rigid body
• Let us now consider that the point P is rotating with respect to point Q in the X-Y plane, i.e., about the Z-axis at
an angular velocity ω𝑧 . Let us, for the sake of simplicity consider a coordinate frame where Q is also the origin ω|QP|
so that 𝑸𝑷 = 𝑷. For a general position vector 𝑷, the linear velocity vector is tangential to 𝑷 and is given by QP P
ሶ ω𝑧 |𝑷|.
|𝑷|=
θ 90°
• In other words in the figure on the right, the velocity of the point P for any general orientation θ of the vector P Q X
ሶ
is given by 𝑷=-ω 𝑧 |𝑷|𝑠𝑖𝑛𝜃ෝ
𝒙 + ω𝑧 |𝑷|𝑐𝑜𝑠𝜃ෝ 𝒚
• Notice that |𝑷|𝑠𝑖𝑛𝜃= 𝑃𝑦 , the Y-component of the vector. Likewise,|𝑷|𝑐𝑜𝑠𝜃=𝑃𝑥 . Thus, the velocity can be written B
ሶ
as , 𝑷=−ω 𝑧 𝑃𝑦 ෝ ෝ
𝒙 + ω𝑧 𝑃𝑥 𝒚
• The above equation can be readily generalized when the object is rotating about Y and X axes by just changing
the labels for the axes in the figure above, while being mindful of the right-handedness of the coordinate
system. If we do that carefully, and ω𝑦 is the angular velocity about the Y-axis then the velocity of the point P
would be 𝑷=ω ሶ 𝑦 𝑃𝑧 ෝ
𝒙 − ω𝑦 𝑃𝑥 𝒛ො . Likewise, if ω𝑥 is the angular velocity about the Z-axis, the linear velocity
would be 𝑷= ሶ −ω𝑥 𝑃𝑧 𝒚 ෝ + ω𝑥 𝑃𝑦 𝒛ො .
• Thus, , for a rigid body rotation with angular velocities about all three axes, by superposition, the linear velocity
ሶ
is given by 𝑷=(−ω 𝑧 𝑃𝑦 + ω𝑦 𝑃𝑧 )ෝ
𝒙 + ω𝑧 𝑃𝑥 −ω𝑥 𝑃𝑧 𝒚ෝ + (ω𝑥 𝑃𝑦 − ω𝑦 𝑃𝑥 )ො𝒛
ሶ
• More generally, if Q is not the origin of the coordinate system, the linear velocity is given by 𝑸𝑷=(−ω 𝑧 𝑄𝑃𝑦 +
ω𝑦 𝑄𝑃𝑧 )ෝ
𝒙 + ω𝑧 𝑄𝑃𝑥 −ω𝑥 𝑄𝑃𝑧 𝒚 ෝ + (ω𝑥 𝑄𝑃𝑦 − ω𝑦 𝑄𝑃𝑥 )ො𝒛
 The cross-product
• We shall introduce a notational simplification. We notice that to obtain the linear velocity of a rotating object, we need to undertake
a weird type of multiplication between the entries of the angular velocity vector 𝝎=ω𝑥 𝒙 ෝ + ω𝑦 𝒚 ෝ + ω𝑧 𝒛ෝ and the position vector
𝑸𝑷=𝑄𝑃𝑥 𝒙 ෝ + 𝑄𝑃𝑦 𝒚
ෝ + 𝑄𝑃𝑧 𝒛ෝ to obtain (−ω𝑧 𝑄𝑃𝑦 + ω𝑦 𝑄𝑃𝑧 )ෝ𝒙 + ω𝑧 𝑄𝑃𝑥 −ω𝑥 𝑄𝑃𝑧 𝒚 ෝ + (ω𝑥 𝑄𝑃𝑦 − ω𝑦 𝑄𝑃𝑥 )ො𝒛

• The rules of multiplication are so weird, and the final expression so painfully lengthy, that we shall invent a new symbol to represent
it, namely the cross product “×”. So, by definition, 𝝎 × 𝑸𝑷=(−ω𝑧 𝑄𝑃𝑦 + ω𝑦 𝑄𝑃𝑧 )ෝ 𝒙 + ω𝑧 𝑄𝑃𝑥 −ω𝑥 𝑄𝑃𝑧 𝒚 ෝ + (ω𝑥 𝑄𝑃𝑦 − ω𝑦 𝑄𝑃𝑥 )ො𝒛.

• Note that unlike a regular product between two scalars, (or even the dot product between two vectors), the cross-product does not
commute: 𝝎 × 𝑸𝑷 ≠ 𝑸𝑷 × 𝝎. Instead, following the sequence of multiplications reveals that 𝝎 × 𝑸𝑷 = −𝑸𝑷 × 𝝎. So, we should
be very cautious about the sequence in which we represent the cross-product between two vectors.
• Likewise also note that for any three vectors a, b, and c, we have 𝒂 × 𝒃 × 𝒄 ≠ 𝒂 × 𝒃 × 𝒄, i.e., the operation is not associative
either.
• At the same time, this new type of multiplication also shares several common features with regular multiplication. For any three
vectors a, b, and c, we have 𝒂 × (𝒃 + 𝒄) = 𝒂 × 𝒃 + 𝒂 × 𝒄 , i.e., the operation is distributive.
𝑑(𝒂×𝒃) 𝑑𝒂 𝑑𝒃
• Likewise, for any two vectors a and b, we note that = × 𝒃+ 𝒂 × .
𝑑𝑡 𝑑𝑡 𝑑𝑡
Velocity of a rigid body

ሶ = 𝝎 × 𝑸𝑷 .
• In summary, the relative velocity of any point B with respect to a point A on the rigid body can always be written as 𝑸𝑷
ሶ =0
• It is easy to verify that this expression satisfies the relation 𝑸𝑷. 𝑸𝑷
• Defining 𝑷ሶ = 𝒗𝑷 , 𝑸ሶ = 𝒗𝑸 and 𝑸𝑷
ሶ = 𝒗𝑷|𝑸 , we note that for any two points A and B fixed to a rigid body, we can always write 𝒗𝑷 =
𝒗𝑸 +𝒗𝑷|𝑸 = 𝒗𝑸 + 𝝎 × 𝑷𝑸
• This equation, along with equations of constraint, is all that we need to obtain the velocities of an interconnection of rigid bodies
possessing one degree of freedom.
 Example
• Consider a 4 bar mechanism shown on the right. If the input bar (the crank) has an angular velocity P L2 Q
of 𝜔1 , obtain the angular velocity of the output bar and the connecting rod (bar 2). ω
• Approach: Notice that the first constraint is that, owing to the pin joint between bars 1 and 2 at the bar 2
point B, the linear velocity of bar 2 at this point is equal to the linear velocity of bar 1. The same L3 L1
argument also holds for the velocity of 2 and 3 at the point C. bar 3 bar 1
• The linear speed of point Q is |𝒗𝑸 | = 𝝎𝟏 × 𝑹𝑸 = 𝜔1 𝐿1 . θ2 θ1 R
O
• Next, we have 𝒗𝑷 = 𝒗𝑸 + 𝝎𝟐 × 𝑸𝑷. From the pin constraint, we also have 𝒗𝑷 = 𝝎𝟑 × 𝑶𝑷
• Though we do not know the numerical values of either 𝝎𝟐 or 𝝎𝟑 , we do know that 𝝎𝟐 × 𝑸𝑷 should
be perpendicular to 𝑸𝑷 while 𝝎𝟑 × 𝑶𝑷 should be perpendicular to 𝑶𝑷.
• So, from geometry, the tip of the 𝒗𝑷 should lie on the intersection of these two perpendiculars as
shown in the schematic on the right.
θ2
• Thus, we get |𝒗𝑷 | = |𝒗𝑸 |𝑠𝑖𝑛𝜃1 /𝑠𝑖𝑛𝜃2
𝒗𝑷|𝑸 𝒗𝑷
• Likewise, we get |𝒗𝑷|𝑸 | = 𝒗𝑸 𝑐𝑜𝑠𝜃1 + 𝒗𝑷 𝑐𝑜𝑠𝜃2
• Thus, we get 𝜔2 = ( 𝒗𝑸 𝑐𝑜𝑠𝜃1 + 𝒗𝑷 𝑐𝑜𝑠𝜃2 )/𝐿2 and 𝜔3 = 𝒗𝑷 /𝐿3

θ1
𝒗𝑸
Reference

J. L. Meriam, and L. G. Kraige, Engineering Mechanics: Dynamics, Wiley, 2013/7th Edition