1.

VIRTUAL DISPLACEMENT

A virtual displacement δ P is an assumed infinitesimal change of system coordinates occurring while

time is held constant. It is called virtual rather than real since no actual displacement can take place
without the passage of time. That is, a virtual (infinitesimal) displacement of a system refers to a change
in the configuration of the system as the result of any arbitrary infinitesimal change of the coordinates
δ Pi , consistent with the forces and constraints imposed on the system at the given time t.
The displacement is called virtual to distinguish it from an actual displacement of the system occurring in
a time interval d t during which the forces and constraints may be changing.
A definition of virtual displacement is presented that allows one to express them explicitly for holonomic
(velocity independent), non-holonomic (velocity dependent), scleronomous (time independent) and
rheonomous (time dependent) constraints. It is observed that for holonomic, scleronomous constraints,
the virtual displacements are the displacements allowed by the constraints. However, this is not so for a
general class of constraints.
In modern terminology virtual displacement is a tangent vector to the manifold representing the
constraints at a fixed time. Unlike regular displacement which arises from differentiating with respect to
time parameter t along the path of the motion (thus pointing in the direction of the motion), virtual
displacement arises from differentiating with respect to the parameter ε enumerating paths of the motion
varied in a manner consistent with the constraints (thus pointing at a fixed time in the direction tangent to
the constraining manifold). The symbol δ is traditionally used to denote the corresponding derivative:

∂
(1.1)
∂ε ε =0

The total differential of any set of system position vectors, Pi , that are functions of other variables, qj

Pi = Pi { q1 , q 2 , … , q m } (1.2)

and time t, may be expressed as follows:

∂ Pi m ∂ Pi
d Pi =
∂t
dt + ∑ ∂q
j =1
d qj (1.3)
j

If, instead, we want to consider the virtual displacement (virtual differential displacement), then:
m ∂ Pi
δ Pi = ∑ δ qj (1.4)
j =1 ∂ qj
This equation is used in Lagrangian mechanics to relate generalized coordinates, q j , to virtual work,
δ L , and generalized forces, Q j .
In analytical mechanics the concept of a virtual displacement, related to the concept of virtual work, is
meaningful only when discussing a physical system subject to constraints on its motion. A special case of
an infinitesimal displacement (usually denoted by d P ), a virtual displacement (denoted by δ P ) refers
to an infinitesimal change in the position coordinates of a system such that the constraints remain
satisfied.
It is also worthwhile to note that virtual displacements are spatial displacements exclusively. Time is
fixed while they occur. When computing virtual differentials of quantities that are functions of space and
time coordinates, no dependence on time is considered (formally equivalent to saying δ t = 0 ).

In case of a translational motion, the virtual displacement is commonly expressed by δ P (or δ P ), while
in case of a rotational motion of a rigid body the virtual displacement is commonly expressed by ε (or
ε ). Anyhow, virtual displacements are infinitesimal vector quantities.
The virtual displacement of a point P of a rigid body subjected to a rotational motion about an axis
passing through point O is given by:

δP = ε ∧ ( P−O ) (1.5)

2. WORK

The work of a force is given by:

L=F× s (2.1)
where vector s is the (finite) displacement of the point which the force is applied to.
The work of a torque (or moment) is given by:
L=M ×ϕ (2.2)
where ϕ is the (finite) angular displacement of the rigid body which the torque is applied to.
The Engineering Unit of work is [ N m ], or [ J ] (Joule).

3. POWER

The power, W, of a force is given by:

ds
W = F× = F× v (3.1)
dt
where vector v is the absolute velocity of the point which the force is applied to.
The power of a torque (or moment) is given by:
dϕ
W=M× = M× ω (3.2)
dt
where ω is the angular velocity of the rigid body which the torque is applied to.
The Engineering Unit of power is [ N m / s ], or [ W ] (Watt).

4. VIRTUAL WORK

The virtual work of a force is given by:

δL = F× δP (4.1)
while the virtual work of a torque is given by:
δL = M × ε (4.2)
The virtual work of a force, F, applied to point P of a rigid body that is subjected to a rotational virtual
displacement, ε , is given by:

δ L = F × ⎡⎣ ε ∧ ( P − O ) ⎤⎦ = ε × ⎡⎣ ( P − O ) ∧ F ⎤⎦ =
(4.3)
⎣ (
⎡ P−O ) ∧ F ⎤ × ε = MO × ε
⎦
In the case of a roto-transaltional virtual displacement of a rigid body, the virtual work of a force, F,
applied to point P, is given by the sum of the virtual works expressed by eqs.(4.1 and 4.3), while the
virtual work of a torque is only expressed by eq.(4.2).
4.1 VIRTUAL WORK PRINCIPLE

If the principle of virtual work for applied forces is used on individual particles of a rigid body, the
principle can be generalized for a rigid body: When a rigid body that is in equilibrium is subjected to
virtual compatible displacements, the total virtual work of all external forces is zero; and conversely, if
the total virtual work of all external forces acting on a rigid body is zero then the body is in equilibrium.
If a system is not in static equilibrium, D’Alembert showed that by introducing the acceleration terms of
Newton’s laws as inertia forces, this approach is generalized to define dynamic equilibrium. The result is
D’Alembert’s form of the principle of virtual work, which is used to derive the equations of motion for a
mechanical system of rigid bodies.
A “not rigorous” formulation of this principle and its application criteria are described below:
- the mechanical system that needs to be analysed (e.g. the global system or a sub-system) must be
isolated. That is, any possible external constraint must be removed and replaced with all the
corresponding reaction forces and moments;
- In case of a sub-system analysis, any possible internal joint that connects the sub-system to the other
parts of the global system must be removed and replaced with all the corresponding internal forces
and moments;
- For each rigid body, the actual distribution of the inertial forces must be replaced with the
corresponding inertial force vector, Fi G , and inertial torque, M i G .
- The virtual work δ L j associated with each force and torque ( j = 1, 2, …, N ) acting on the isolated
system (or sub-system) must be evaluated;
- The virtual work caused by friction forces that cause any energy loss must be taken into account;
- The sum of all the above mentioned virtual works must satisfy the following equation:
N

∑ δL
j =1
j =0 (4.4)

As said above, the Virtual Work Principle (VWP) can be applied to the global system and to any
sub-system as well. However, the equations obtained applying the VWP must be linearly independent of
each other.

5. POWER BALANCE

In accordance with a rigorous definition, a virtual displacement δ P is an assumed infinitesimal change of

system coordinates occurring while time is held constant. However, we can assume that it may be
associated with a virtual (infinitesimal) time interval δ t .
Then, the velocity associated with a virtual displacement δ P is given by:
δP
v = (5.1)
δt
Similarly, the angular velocity of a rigid body that is subjected to a virtual rotation is give by:
ε
ω= (5.2)
δt
The power corresponding to the virtual work δ L j is:

δ Lj
Wj = (5.3)
δt
Therefore, eq.(4.4) can be rewritten as:
N
δ Lj N

∑
j =1 δt
= ∑W
j =1
j =0 (5.4)

That is, in order to satisfy the system dynamic equilibrium, the sum of the powers of all the forces and
torques applied to an isolated system must be null, at any time. Eq.(5.4) can be written for the global
system and any sub-system as well. However, if these equations, written for more than one su-system are
processed at the same time, they must be linearly independent of each other. The unknowns of these
equations can be forces and torques, as well as kinematic parameters.

Sometimes, it can be useful to distinguish between:

- input power terms, Winput, (e.g. caused by driving forces and torques);
- output power terms, Woutput, (e.g. caused by resistance forces and torques);
- power lost caused by friction forces, Wlost;
- power terms associated with inertial forces and torques, Winertial.
Therefore, eq.(5.4) can be rewritten as:
Winput + Woutput + Wlost + Winertial = 0 (5.5)
The power associated with inertial forces and torques can be also expressed as:
dT
Winertial = − (5.6)
dt
where T is the kinetic energy of the system, or syb-system, that has been considered in the analysis.
In fact, if we consider the kinetic energy of a rigid body, subjected to a planar motion, whose mass and
barycentrical moment of inertia are m and JG, the kinetic energy can be expressed as:
1 1
T= m ( v G × v G ) + J G ( ω × ω) (5.7)
2 2
The derivative of the kinetic energy, with respect to time, is:
dT 1 1
= ⎡⎣ m ( aG × v G ) + m ( v G × aG ) ⎤⎦ + ⎡⎣ J G ( ω × ω ) + J G ( ω × ω ) ⎤⎦ (5.8)
dt 2 2
That is:
dT
= m ( aG × v G ) + J G ( ω × ω ) = Fi G × v G + Mi G × ω = − Winertial (5.9)
dt
Therefore, eq.(5.5) can be rewritten as:
dT
Winput + Woutput + Wlost = (5.10)
dt
Eqs. from (5.5) to (5.10) have been written considering positive the inlet power (or energy) and negative
the outlet and lost power (or energy). For a given mechanical system the inlet powers are those associated
with forces and torques that contribute to counteract resistance forces and torques.
Sometimes, the opposite convention is used for the outlet and lost power terms. In this case eqs.(5.5) and
(5.10) must be written as:
Winput − Woutput − Wlost + Winertial = 0 (5.11)
and
 dT
Winput − Woutput − Wlost = (5.12)
dt
In accordance with this particular convention, the power terms due to forces and torques must be written
as:
W = −F× v W=− M× ω (5.13)
If friction forces are negligible the Power Balance Equation of a mechanical system, alone, allows one to
evaluate the single unknown of the problem, e.g. a driving or resistance force (or torque), or a kinematic
parameter. Conversely, in case of significant friction forces, Power Balance Equations – written for the
global system or sub-systems – can be combined with a suitable set of cardinal equations of dynamics.
However, beware of avoiding linear combinations of the set of equations that you are going to solve.
These equations will contain several unknowns like internal actions and reaction forces and torques.