Constraints in classical mechanics are conditions that limit how an object or system can move. They determine which positions, directions, and motions are physically possible, often because of connections between objects, contact with surfaces, or restrictions imposed by the system’s geometry. A ball rolling inside a bowl, a pendulum swinging from a fixed support, and a bead sliding along a wire all move under constraints.
Constraints are important because they simplify the description of mechanical systems. Instead of tracking every possible motion an object could make, physicists can focus on the smaller set of motions allowed by the constraints. These conditions also help determine the forces acting on objects and provide the foundation for powerful methods such as Lagrangian mechanics.
What does a constraint mean in classical mechanics?
In classical mechanics, a constraint is a mathematical or physical condition that restricts the motion of one or more objects. It may limit where an object can go, how quickly it can move in a particular direction, or how its motion relates to that of another object.
Consider a simple pendulum: a small mass attached to a fixed point by a rigid, massless string. If the string remains taut and has length L, the mass must stay a distance L from the pivot. The mass can swing through different angles, but it cannot move freely toward or away from the pivot.
This restriction can be expressed mathematically as
x² + y² = L²
Here, x and y are the horizontal and vertical coordinates of the mass relative to the pivot, and L is the string’s fixed length. The equation defines a circle along which the mass must move.
Without the constraint, the mass could occupy any position in a two-dimensional plane. With the constraint, its position is limited to a one-dimensional path.
Constraints do not necessarily eliminate forces or determine the entire motion by themselves. Instead, they specify what motions are permitted. Newton’s laws or an equivalent formulation of mechanics must still be used to determine how the system moves under applied forces, gravity, and other influences.
Why constraints matter in mechanical systems
Constraints make it possible to describe complicated systems efficiently. A rigid body, for example, consists of many particles whose relative positions remain fixed. Tracking every particle independently would be unnecessarily complicated when the body’s shape and dimensions do not change.
By expressing the restrictions mathematically, physicists can reduce the number of independent variables needed to describe a system. These independent variables are called degrees of freedom.
A free particle in three-dimensional space has three translational degrees of freedom because its position requires three coordinates: x, y, and z. If the particle must remain on a fixed surface, its motion is restricted, and only two independent coordinates may be needed. If it must follow a fixed curve, one independent coordinate is generally sufficient.
Constraints also explain forces that keep objects on their permitted paths. A table supports a book against gravity, a taut string pulls on a swinging pendulum, and a track redirects a moving cart. These forces are associated with maintaining the relevant restrictions.
In many mechanical systems, constraints can be classified according to whether they depend on position, velocity, or time; whether they can be expressed as equations; and whether they permit or prevent certain types of motion.
The main types of constraints
Several classifications are used in classical mechanics. The most common distinctions are between holonomic and nonholonomic constraints, scleronomic and rheonomic constraints, and ideal and nonideal constraints. These categories describe different properties, so a single constraint may belong to more than one category.
Holonomic constraints
A holonomic constraint can be expressed as an equation involving the system’s generalized coordinates and, possibly, time.
Generalized coordinates are variables chosen to describe the configuration of a system. They might be Cartesian coordinates such as x and y, angles such as θ, or distances measured along a path.
A general holonomic constraint has the form
f(q₁, q₂, …, qₙ, t) = 0
Here, q₁ through qₙ represent generalized coordinates, t represents time, and f is a function describing the restriction.
The pendulum provides a straightforward example:
x² + y² − L² = 0
Because the string length is fixed, the coordinates must always satisfy this equation.
Another example is a particle constrained to move on a sphere of radius R:
x² + y² + z² − R² = 0
The particle may move over the sphere’s surface, but it cannot move away from that surface while the constraint remains in effect.
Holonomic constraints are particularly useful because they often allow some coordinates to be eliminated. For a pendulum, the angle θ can describe the mass’s position without requiring x and y as independent variables. This reduces the mathematical complexity of the problem.
Nonholonomic constraints
A nonholonomic constraint generally restricts velocities or possible motions in a way that cannot be reduced to an equivalent equation involving only coordinates and time.
A common mathematical form is
a₁(q, t)q̇₁ + a₂(q, t)q̇₂ + … + aₙ(q, t)q̇ₙ + a₀(q, t) = 0
The dotted coordinates represent time derivatives, or generalized velocities. The coefficients may depend on the system’s configuration and time.
An important example is a wheel rolling without slipping on a flat surface. The wheel’s translational speed and angular speed are related by
v = Rω
Here, v is the speed of the wheel’s center, R is its radius, and ω is its angular speed.
This equation expresses the no-slip condition along the direction of travel. In a more general rolling problem, additional relations connect the wheel’s orientation and motion. For a wheel moving and turning on a plane, these restrictions can be nonholonomic because the allowed instantaneous motions do not necessarily translate into a constraint on position alone.
Nonholonomic constraints require careful treatment. Not every equation involving velocity is nonholonomic: some velocity relations can be integrated to produce a coordinate equation. The distinction depends on whether the constraint can be reduced to a condition on configuration and time.
Scleronomic constraints
A scleronomic constraint does not depend explicitly on time. Its mathematical expression can be written without a separate time dependence:
f(q₁, q₂, …, qₙ) = 0
A pendulum with a fixed pivot and constant string length is a familiar example. Its geometric restriction remains unchanged as time passes.
A particle constrained to move on a stationary circular track also has a scleronomic constraint, assuming the track itself does not move or change shape.
The key feature is that the restriction is fixed in time. The object may move, but the permitted geometric relationship does not change.
Rheonomic constraints
A rheonomic constraint depends explicitly on time. This occurs when a restriction moves or changes as the system evolves.
For example, imagine a bead sliding along a straight wire that moves horizontally according to a prescribed position X(t). If the bead must remain on the wire, its horizontal position may satisfy
x − X(t) = 0
The allowed position changes with time because the wire moves.
Rheonomic constraints are useful for describing machinery with moving supports, objects guided by moving tracks, and mechanical systems driven by prescribed motions. They can transfer energy into or out of a system because the moving constraint may do work on the object.
The distinction between scleronomic and rheonomic constraints concerns time dependence, not whether a constraint is holonomic. A constraint can be both holonomic and rheonomic, or holonomic and scleronomic.
Ideal and nonideal constraints
Constraints can also be classified by how their associated forces behave.
An ideal constraint is one for which the constraint forces do no virtual work on allowed infinitesimal displacements. A virtual displacement is a hypothetical, infinitesimal change in configuration consistent with the constraints at a fixed instant.
For a smooth, stationary surface, the normal force acts perpendicular to the surface. A particle’s allowed instantaneous displacement lies along the surface, so the normal force does no virtual work.
This property often simplifies the equations of motion because the constraint forces need not be calculated explicitly when applying Lagrange’s equations in their standard ideal-constraint form.
A nonideal constraint does not have this property. Frictional contact, for example, can exert a force that opposes sliding and does work on the system. Such forces must be accounted for through an appropriate force or energy formulation.
The distinction depends on the specific model. A frictionless surface is commonly treated as an ideal constraint, whereas a rough surface with sliding friction requires additional treatment. Ideal constraints are useful approximations, but real mechanical contacts may involve friction, deformation, and energy dissipation.
Constraint equations and how they work
Constraint equations express the allowed configurations or motions of a system. Their mathematical form depends on the physical situation.
For a system described by generalized coordinates q₁, q₂, …, qₙ, a holonomic constraint can be written as
f(q₁, q₂, …, qₙ, t) = 0
If a system has n coordinates and m independent holonomic constraints, its configuration generally requires n − m independent coordinates. This count assumes that the constraints are independent and that the system is not at a special configuration where the restrictions change rank.
For example, a particle moving freely in three dimensions requires three coordinates. If it is restricted to a sphere, one independent constraint reduces the number of degrees of freedom from three to two. Two suitable coordinates, such as latitude and longitude within a suitable coordinate patch, can then describe its position on the surface.
Differentiating a holonomic constraint with respect to time gives a corresponding velocity relation:
∂f/∂q₁ · q̇₁ + ∂f/∂q₂ · q̇₂ + … + ∂f/∂qₙ · q̇ₙ + ∂f/∂t = 0
This equation follows from the chain rule. It shows how the velocities must relate to one another if the position constraint is to remain satisfied.
For the fixed-length pendulum, differentiating x² + y² = L² gives
xẋ + yẏ = 0
The equation means that the velocity is perpendicular to the position vector from the pivot. In other words, the mass moves tangentially to the circle rather than directly toward or away from the pivot.
Differentiating the velocity relation again provides an acceleration condition. For the same pendulum,
ẋ² + ẏ² + xẍ + yÿ = 0
This relation connects the particle’s acceleration to its current position and velocity. Such equations are useful when applying Newton’s second law while enforcing the constraint.
For a system with several interacting parts, the same process can produce a set of simultaneous equations. These equations, combined with force laws and initial conditions, determine the system’s motion.
How constraints produce forces
Constraints often require forces to prevent forbidden motion. These forces are called constraint forces.
For a pendulum, the string exerts tension along its length. That tension keeps the mass at a fixed distance from the pivot. For a book resting on a table, the table exerts a normal force that prevents the book from accelerating downward through the surface.
Newton’s second law still applies:
F = ma
Here, F is the net force, m is mass, and a is acceleration. In a constrained system, the net force includes both applied forces, such as gravity, and the constraint forces needed to maintain the permitted motion.
The constraint force is often not known in advance. It must be determined along with the motion.
Consider a mass moving on a frictionless horizontal table while attached to a fixed point by a taut string. The string’s tension points toward the fixed point and provides the centripetal force needed for circular motion. If the mass moves at constant speed v in a circle of radius R, the required inward force is
F = mv²/R
The string’s tension supplies this force in the idealized model. If the string breaks, the constraint disappears, and the mass continues along the tangent to its circular path, assuming no other horizontal forces act.
Not every constraint force does work. A normal force on a stationary, frictionless surface typically does no work because it is perpendicular to the object’s instantaneous displacement. Tension in an ideal pendulum string also does no work on the bob because the tension is radial while the bob’s instantaneous motion is tangential.
However, a moving constraint can do work. A moving support may pull an attached object along, changing its mechanical energy. Friction can also transfer mechanical energy into thermal energy. Whether a constraint force does work therefore depends on the type of constraint and the motion involved.
Degrees of freedom and the role of constraints
The degrees of freedom of a mechanical system are the number of independent variables needed to specify its configuration.
A free particle in three-dimensional space has three translational degrees of freedom. If its position is restricted to a fixed plane, it generally has two. If it must move along a fixed curve, it generally has one.
A rigid body in three-dimensional space has six degrees of freedom: three describing translation of a reference point and three describing orientation. If the body is fixed at a pivot, its translational motion is eliminated, leaving three rotational degrees of freedom, provided the pivot fixes a point but allows unrestricted rotation.
Constraints can also relate the motions of multiple objects. Consider two masses connected by an ideal string passing over a fixed, frictionless pulley. If the string remains taut and its total length is constant, an increase in one mass’s distance from the pulley must be matched by a decrease in the other mass’s corresponding distance.
If y₁ and y₂ measure the downward positions of the masses from their respective reference points, the constraint can be written, with suitable coordinate conventions, as
y₁ + y₂ = L
where L is the constant length available to the two straight string segments, assuming the portions around the pulley remain constant.
Differentiating gives
ẏ₁ + ẏ₂ = 0
and differentiating again gives
ÿ₁ + ÿ₂ = 0
Thus, the masses have equal acceleration magnitudes in opposite directions under these coordinate conventions. This relationship follows from the fixed string length rather than from the masses’ individual force equations.
The example illustrates a central benefit of constraints: they connect the motions of different parts of a system and reduce the number of independent variables needed to describe the whole system.
Constraints in Lagrangian mechanics
Lagrangian mechanics is an alternative formulation of classical mechanics that is especially useful for systems with complicated geometry or many interacting components.
Instead of beginning with separate force equations for every coordinate, the method uses the system’s kinetic energy T and potential energy V to define the Lagrangian:
L = T − V
For a system described by generalized coordinates qᵢ, the standard Lagrange equations are
d/dt(∂L/∂q̇ᵢ) − ∂L/∂qᵢ = Qᵢ
Here, Qᵢ represents a generalized force not already included through the energy terms, such as an appropriate nonconservative force. In a simple conservative system with no additional nonconservative forces, Qᵢ is zero.
For ideal holonomic constraints, one useful approach is to choose generalized coordinates that automatically satisfy the constraints. A pendulum, for example, can be described using its angle θ rather than independent Cartesian coordinates x and y. The fixed-length condition is then built into the coordinate choice.
When the constraints cannot conveniently be eliminated, they can be included using Lagrange multipliers. These are mathematical variables introduced to enforce the constraint equations while solving for the motion.
For holonomic constraints fⱼ(q, t) = 0, one common form of the equations is
d/dt(∂L/∂q̇ᵢ) − ∂L/∂qᵢ = Qᵢ + Σⱼ λⱼ ∂fⱼ/∂qᵢ
The sum runs over the constraints. The multipliers λⱼ represent the strength of the constraint terms in the equations, with their precise interpretation depending on the constraint equations and sign convention. In many ideal mechanical systems, they can be used to recover constraint forces.
This method is powerful because it allows physicists to solve for both the motion and the forces required to enforce restrictions. It is widely useful in problems involving pendulums, rigid bodies, linkages, and multibody systems.
Everyday examples of constraints in classical mechanics
Constraints are present in many familiar mechanical situations, even when they are not immediately obvious.
A ball rolling down a fixed ramp is restricted by the ramp’s surface. If the ball rolls without slipping, its translational and rotational motions are linked. For a rigid wheel or ball with radius R, the no-slip condition along a straight path is v = Rω. If the ball begins to slide, this relationship no longer holds, and friction and energy transfer must be considered accordingly.
A car traveling along a road is constrained by the road surface and by the contact between its tires and the pavement. The road prevents the car from freely moving vertically through the ground, while tire forces help control its acceleration and direction. During a turn, frictional forces between the tires and road provide the lateral force needed to change the car’s direction. The actual motion depends on the tires, road conditions, vehicle dynamics, and other forces.
A door rotating on hinges is constrained to rotate about an axis defined by the hinges, assuming the hinges are rigid and properly aligned. Its position cannot change arbitrarily in three-dimensional space. The hinges supply forces and possibly moments that prevent the forbidden translations and rotations.
A bead sliding along a fixed wire must remain on the wire’s path. The wire’s shape limits the bead’s position, while contact forces redirect its motion. If the wire moves, the constraint may become time-dependent, and the moving wire can transfer energy to the bead.
A two-mass pulley system demonstrates how a single constraint can connect distant parts of a mechanism. The string’s fixed length links the positions and accelerations of both masses. Their individual motions cannot be chosen independently, even though each mass experiences its own gravitational and tension forces.
These examples show that constraints arise from geometry, connections, and contact. They are not merely mathematical conveniences; they represent physical restrictions that shape the behavior of mechanical systems.
What happens when a constraint changes or fails?
A constraint remains valid only while the physical conditions supporting it continue to hold. If a string breaks, a contact surface separates, or a component deforms enough to invalidate the model, the corresponding constraint may no longer apply.
When a pendulum string breaks, for example, the mass is no longer required to follow a circular path. Its subsequent motion is determined by its velocity at the moment of release and the forces that remain, such as gravity. If the mass was moving when the string broke, its initial velocity is tangent to the former circular path.
Contact constraints require particular care. A smooth surface can prevent an object from moving into the surface, but it cannot pull an object toward itself when contact is lost. In a simple unilateral contact model, the gap between the object and the surface must remain nonnegative, the normal force must remain nonnegative, and the force must vanish when the bodies are separated. These conditions are often expressed through contact inequalities and complementarity relations rather than a single equality.
A rigid string also illustrates the difference between an idealized constraint and a physical limitation. A taut, inextensible string fixes the distance between its endpoints. A real string can stretch, become slack, or break. Once it becomes slack, the fixed-distance equality is no longer appropriate; the string can impose a maximum-distance restriction instead.
Understanding when a constraint applies is therefore as important as writing its equation. An incorrect constraint can lead to incorrect predictions even if the subsequent mathematics is performed flawlessly.
The limits of the constraint approach
Constraints are central to classical mechanics, but their usefulness depends on choosing a model that reflects the physical system closely enough for the question being asked.
Rigid-body models treat components as having fixed shapes. They work well when deformation is negligible, but they cannot fully describe bending, stretching, or vibration. Frictionless-contact models simplify calculations but may fail when friction substantially affects the motion. Ideal strings and pulleys ignore effects such as string mass, elasticity, and pulley friction.
Some constraints are also more difficult to handle mathematically than others. Holonomic constraints can often be incorporated through a reduced set of coordinates, whereas nonholonomic constraints may require equations that preserve velocity restrictions directly. Systems with intermittent contact, rolling, or collisions can introduce additional complications because the applicable constraints may change during motion.
These limitations do not make the method less useful. They clarify what a mechanical model assumes and when those assumptions should be revised.
Constraints provide a systematic way to connect geometry, motion, and force. By identifying which movements are possible and which are forbidden, classical mechanics can describe everything from a simple pendulum to complex machinery with a manageable set of equations. Their central role is to turn physical restrictions into mathematical relationships that can be used to predict how a system behaves.