Joint Constraints

A joint constraint's primary function is to remove degrees of freedom from the six possible motions (three translational, three rotational) between two bodies. The type of joint is defined by which DoFs are locked and which are free.

• Revolute/Hinge Joint: Locks 5 DoFs, allowing only 1 rotational DoF (e.g., a door hinge or elbow).
• Prismatic/Slider Joint: Locks 5 DoFs, allowing only 1 translational DoF (e.g., a piston or drawer slide).
• Spherical/Ball-and-Socket: Locks 3 translational DoFs, allowing 3 rotational DoFs (e.g., a hip or shoulder joint).
• Fixed/Weld Joint: Locks all 6 DoFs, creating a single rigid body from two.

Joints are implemented as mathematical equality or inequality constraints within the simulation's solver. These constraints are typically expressed as Jacobian matrices that relate the velocities of the connected bodies to the constraint forces (Lagrange multipliers) needed to enforce the joint limits.

• Holonomic Constraints: Define relationships between position coordinates (e.g., distance(p1, p2) = constant for a rigid rod).
• Non-Holonomic Constraints: Involve velocities (less common in basic joints).
• Inequality Constraints: Used to model joint limits (e.g., a hinge that cannot rotate past 90 degrees), turning the problem into a Linear Complementarity Problem (LCP).

Enforcing joint constraints accurately and stably is a central challenge. Solvers calculate constraint forces to correct positional or velocity drift that accumulates due to numerical integration errors.

• Penalty Methods: Apply spring-like forces to push bodies back into compliance; simple but can cause oscillation.
• Impulse-Based Methods: Apply instantaneous velocity changes at the time of constraint violation, common in real-time game physics.
• Constraint Force Algorithms: Compute exact forces using Lagrangian multipliers within a velocity-level solver (like Sequential Impulse) or position-based dynamics (PBD). Stability often requires baumgarte stabilization, a feedback term that adds a force proportional to the positional error.

Beyond passive limitation, joints are often actuated to drive simulation. This involves applying controlled forces or torques within the free DoFs of the joint to achieve target motion.

• Velocity Motor: Applies torque until a specific angular or linear velocity is reached.
• Position Motor/Servo: Uses proportional-derivative (PD) control to apply torque to move the joint to a target angle or position.
• Maximum Force/Torque: Actuators have limits, simulating real motor saturation. Implementing this requires solving a constrained optimization problem where the solver respects both the joint's kinematic constraints and the actuator's force limits.

Standard joint types model fundamental mechanical connections. Complex mechanisms are built by combining these primitives.

• Revolute Joint: Door hinge, robotic elbow/knee, wheels on an axle.
• Prismatic Joint: Hydraulic piston, sliding door, CNC machine gantry.
• Cylindrical Joint: 1 rotational + 1 translational DoF (e.g., a door knob or a sliding bolt).
• Planar Joint: Allows 2 translational DoFs and 1 rotational DoF within a plane (e.g., a puck on an air hockey table).
• Universal/Hooke's Joint: 2 rotational DoFs (e.g., a Cardan joint in a vehicle driveshaft).
• Fixed Joint: Welded seam, parts of a single rigid assembly.

In robotics and biomechanics, joint constraints are the building blocks of kinematic chains and articulated bodies. Their accurate simulation is critical for:

• Forward/Inverse Kinematics (FK/IK): Calculating end-effector pose from joint angles, or joint angles from a desired pose.
• Dynamics Simulation: Computing the motion of the entire chain under applied forces and torques using the articulated body algorithm.
• Control & Trajectory Optimization: Designing controllers that issue actuator commands to joints to achieve complex tasks like walking or grasping.
• Sim-to-Real Transfer: Faithful joint modeling (including friction, damping, and backlash) is essential for bridging the sim-to-real gap in robotic training.

A revolute joint permits rotation around a single, fixed axis, like a door hinge or an elbow. It removes all translational degrees of freedom, allowing only one rotational degree of freedom (1-DOF).

• Key Parameter: The joint angle, often limited by a lower and upper limit to prevent unrealistic rotation.
• Real-World Examples: Door hinges, robotic arm segments, human knees and elbows.
• Simulation Use: Critical for articulated mechanisms where parts swing or pivot.

A prismatic joint allows linear translation along a single, fixed axis, like a piston or a drawer slide. It removes all rotational degrees of freedom and translations in other directions, permitting only one translational degree of freedom (1-DOF).

• Key Parameter: The joint position or displacement along its axis, which can be bounded by limits.
• Real-World Examples: Hydraulic pistons, linear actuators, sliding cabinet drawers.
• Simulation Use: Models linear motion where components extend, retract, or slide without rotating.

A spherical joint permits free rotation around all three spatial axes (pitch, yaw, roll) but prohibits any translational movement. It provides three rotational degrees of freedom (3-DOF).

• Key Parameter: The orientation of the child body relative to the parent, often represented by a quaternion or Euler angles.
• Real-World Examples: The human hip or shoulder joint, a car's gear shift knob, a surveillance camera mount.
• Simulation Use: Essential for modeling limbs and other linkages that require wide-ranging, multi-axis rotation.

A fixed joint completely locks the relative position and orientation between two bodies, effectively welding them together into a single composite rigid body. It removes all six degrees of freedom (0-DOF).

• Key Parameter: The initial relative transform (position and rotation) is maintained permanently.
• Real-World Examples: Two pieces of metal welded together, a bolt firmly securing two plates, parts of a monolithic structure.
• Simulation Use: Used to simplify complex geometry by combining multiple shapes or to create permanent, non-breaking connections.

A cylindrical joint combines the motions of a revolute and a prismatic joint along the same axis. It allows for both rotation around and translation along a single axis, providing two degrees of freedom (2-DOF).

• Key Parameters: The joint angle and the joint position along the shared axis, each potentially having independent limits.
• Real-World Examples: A rotating door knob that also pulls out, a telescoping car antenna that can also spin, a screw inserted into a threaded hole.
• Simulation Use: Models complex mechanical couplings where sliding and twisting occur simultaneously on the same axis.

A planar joint allows translation within a defined plane and rotation around an axis perpendicular to that plane. It provides two translational and one rotational degree of freedom (3-DOF).

• Key Parameter: Motion is constrained to a 2D plane; the body cannot move along the plane's normal vector.
• Real-World Examples: A puck sliding on an air hockey table, a glass on a tabletop, a mouse on a mousepad.
• Simulation Use: Simulates objects moving on a flat surface, useful for 2.5D simulations or simplifying 3D contact for certain objects.