Rigid-Link Gaussian Kinematics

• Rigid-Link Gaussian Kinematics (RLGK) is a unified framework that employs Gauss’s principle of least constraint to derive equations of motion for interconnected rigid bodies.
• It bridges classical formulations—Lagrange, Newton–Euler, and Kirchhoff—with modern group-theoretic and dual quaternion approaches for mechanism analysis.
• The method enhances mechanism synthesis and computational dynamics by integrating algebraic invariants and geometric structures for robust spatial mechanism design.

Rigid-Link Gaussian Kinematics (RLGK) provides a unified variational formulation for the dynamics and kinematics of systems comprising interconnected rigid bodies. Instead of isolating Lagrange, Newton–Euler, or Kirchhoff frameworks, RLGK bases all on Gauss’s principle of least constraint. The resulting equations of motion naturally encompass all classical rigid-body dynamics formulations through judicious selection of velocity coordinates or “quasicoordinates.” RLGK also seamlessly connects with modern representation theory on $\mathrm{SE}(3)$ and projectivized dual quaternion models for mechanism constraint varieties, linking algebraic and geometric approaches to robot mechanism synthesis and analysis (Massa et al., 2016, Rad et al., 2016, Rad et al., 2015).

1. Gauss’s Principle and the RLGK Variational Functional

Gauss’s principle of least constraint identifies the physically realized accelerations among all kinematically admissible candidates as those minimizing the quadratic deviation from the free-body accelerations. For a chain of $n$ rigid bodies with unconstrained accelerations $\{a_i^*\}$ (subject only to external forces) and constraint-compatible accelerations $\{a_i\}$, the Gauss functional is:

$$Z = \tfrac{1}{2} \sum_{i=1}^n m_i \left\| a_i - a_i^* \right\|^2$$

The requirement $\delta Z = 0$ under all virtual variations respecting constraints yields the actual system acceleration (Massa et al., 2016).

When applied to an $n$-link rigid chain, each $a_i$ is determined from generalized coordinates $q$, their velocities $\dot q$, and accelerations $n$0 via: $n$1 where $n$2 is the twist-to-Cartesian Jacobian for body $n$3 and $n$4 groups the Christoffel-type velocity-dependent (Coriolis and centripetal) terms. The unconstrained acceleration is

$n$5

with $n$6 the spatial inertia. Substitution into $n$7 yields

$n$8

The stationarity condition $n$9 imposes a well-defined variational principle for the governed motion (Massa et al., 2016).

2. Unified Equations of Motion and Quasicoordinate Choice

The RLGK principle leads to the common form of rigid-body dynamics: $\{a_i^*\}$0 where the terms are constructed as:

• Generalized mass matrix:

$\{a_i^*\}$1

• Velocity-dependent (Coriolis/centrifugal) terms:

$\{a_i^*\}$2

• Generalized force/torque vector:

$\{a_i^*\}$3

The essence is that $\{a_i^*\}$4 can be chosen as arbitrary linear velocity coordinates—joint rates (Lagrange form), twist vectors (Newton–Euler form), or screw-velocities (Kirchhoff form), among others. Each classical rigid-body formalism is recovered by the appropriate selection of quasicoordinates and the corresponding Jacobians $\{a_i^*\}$5, maintaining mathematical equivalence at the level of $\{a_i^*\}$6 (Massa et al., 2016).

3. Group-Theoretical Structure and Trivializations on $\{a_i^*\}$7

The configuration space for a single rigid body is the Lie group $\{a_i^*\}$8, coordinatized as

$\{a_i^*\}$9

with $\{a_i\}$0 and $\{a_i\}$1. Tangent velocities reside in $\{a_i\}$2 as twists. There are two natural trivializations:

• Left-trivialization (body frame): Spatial inertia $\{a_i\}$3 remains constant for each link, and the Coriolis term $\{a_i\}$4 becomes sparse.
• Right-trivialization (world frame): Yields simplified articulated Jacobians for serial chains.

By choosing trivialization, one can often achieve desirable properties: either a constant mass matrix (useful for symbolic computation) or a skew-symmetric Coriolis matrix (beneficial for energy analysis). The compatibility of RLGK with group-theoretic approaches is central for efficient symbolic and numeric implementation in mechanism dynamics (Massa et al., 2016).

4. Constraint Varieties in the Projectivized Dual Quaternion Model

RLGK bridges with the dual quaternion formalism, particularly for kinematic chains (dyads, linkages). In this setting, rigid displacements correspond to points in $\{a_i\}$5 (projective dual quaternion space), constrained to the Study quadric $\{a_i\}$6. Dyads are represented as the intersection of 3-spaces $\{a_i\}$7 with $\{a_i\}$8, yielding regular ruled quadrics $\{a_i\}$9. The type of dyad (RR, PR, RP, or cylindrical) is distinguished by the incidence structure of “null” lines (generators) in $$Z = \tfrac{1}{2} \sum_{i=1}^n m_i \left\| a_i - a_i^* \right\|^2$$0 (with $$Z = \tfrac{1}{2} \sum_{i=1}^n m_i \left\| a_i - a_i^* \right\|^2$$1 the null cone) and the associated fiber-projectivity invariant $$Z = \tfrac{1}{2} \sum_{i=1}^n m_i \left\| a_i - a_i^* \right\|^2$$2 (Rad et al., 2016).

A summary of constraint incidence types is:

Dyad Type	Ruled Quadric $$Z = \tfrac{1}{2} \sum_{i=1}^n m_i \left\| a_i - a_i^* \right\|^2$$3	Null Incidence
RR	$$Z = \tfrac{1}{2} \sum_{i=1}^n m_i \left\| a_i - a_i^* \right\|^2$$4 contains	Null quadrilateral—4 lines in $$Z = \tfrac{1}{2} \sum_{i=1}^n m_i \left\| a_i - a_i^* \right\|^2$$5
PR/RP	$$Z = \tfrac{1}{2} \sum_{i=1}^n m_i \left\| a_i - a_i^* \right\|^2$$6 contains	2 complex conjugate null lines $$Z = \tfrac{1}{2} \sum_{i=1}^n m_i \left\| a_i - a_i^* \right\|^2$$7 1 real null line $$Z = \tfrac{1}{2} \sum_{i=1}^n m_i \left\| a_i - a_i^* \right\|^2$$8
Cylinder	$$Z = \tfrac{1}{2} \sum_{i=1}^n m_i \left\| a_i - a_i^* \right\|^2$$9 with fibers	All null lines collapse: $\delta Z = 0$0

Special coordinate choices and projective invariants discriminate among dyad classes and encode commutation and geometric singularities (Rad et al., 2016).

5. RLGK in Two-Link and Dyadic Mechanism Synthesis

For concrete illustration, consider a planar 2-link arm (two revolute joints), with joint angles $\delta Z = 0$1, link parameters $\delta Z = 0$2:

• Link center positions:

$\delta Z = 0$3

• Mass matrix (block form):

$\delta Z = 0$4

with explicit expressions for $\delta Z = 0$5, $\delta Z = 0$6, $\delta Z = 0$7 (see source for details).

• Coriolis terms:

$\delta Z = 0$8

• RLGK equation instantiated:

$\delta Z = 0$9

Mechanism constraint varieties for 2R dyads (two revolute axes) comprise regular ruled quadrics in $n$0 with null quadrilaterals (four null lines at $n$1, $n$2) (Rad et al., 2015). Parameterizations and ruling-family structure correlate with feasible joint motions. For exact linkage synthesis (e.g., interpolating four poses), the twisted-cubic theory on the quadric gives two admissible classes of 5R linkages, each characterized by a commutation relation among factors in a cubic motion polynomial (Rad et al., 2015).

6. Synthesis, Invariants, and Gaussian Complexes

RLGK’s correspondence with Gaussian complexes in projective kinematics associates 1- and 2-joint dyads with quadrics stratified by null incidence. The fiber-projectivity invariant $n$3 is crucial: for PR/RP dyads, $n$4 identifies the cylinder case (commuting joints), while for generic PR/RP, $n$5, distinguishing these from the RR case (null quadrilateral) (Rad et al., 2016). This stratification refines the traditional Gaussian classification, with each quadrics’ null content encoding the dyad’s kinematic mobility and singularity structure.

Admissible transformations (frame changes) correspond to projectivities fixing the Study quadric $n$6 and null cone $n$7. Such transformations carry RR, PR, RP, and cylinder dyad spaces into one another, preserving the nuanced distinction of ruling families and null incidence, which is essential for mechanism classification and synthesis (Rad et al., 2016).

7. Connections and Implications for Mechanism Science

The RLGK formalism provides an overview of geometric, analytic, and algebraic perspectives on rigid-body dynamics and kinematics. Its compatibility with coordinate-invariant, group-theoretic, and projective methods enables:

• Unified derivation of classical and modern equations of motion.
• Seamless translation between Newton–Euler, Lagrange, and Kirchhoff kinetic representations.
• Direct integration with dual quaternion and Plücker coordinate representations for linkage synthesis, classification, and pose interpolation tasks.
• Rigorous algebraic classification of mechanism constraint varieties via Gaussian complexes and null incidence.
• Provision of explicit geometric invariants for mechanism stratification (e.g., fiber-projectivity, null quadrilaterals).

These features make RLGK a central framework for both theoretical understanding and computational treatment of spatial mechanisms and their synthesis, as evidenced in the cited literature (Massa et al., 2016, Rad et al., 2016, Rad et al., 2015).