Constraint Forces: The Hidden Engines of the Swing
You learned in Chapter 6 that the downswing is passive—gravity and Coriolis drive acceleration. But Chapter 6 left a critical question unanswered:
How does the arm’s momentum become the club’s momentum?
If gravity is pulling the whole system downward, why does the club accelerate faster than the arm? Where does that extra speed come from?
The answer is constraint forces. They are the hidden engines of the kinetic chain, exactly the kind of noncontributing forces analyzed in Kane-style constrained dynamics (Roithmayr and Hodges 2016).
Constraint forces are: - Forces that enforce joint connections: Your elbow joint connects the upper arm to the forearm. This connection is not free-floating; it’s constrained. The constraint produces a force. - Forces that do zero net work: This is paradoxical but true. Constraints don’t add energy to the system. Yet they redistribute energy from one segment to another. - Forces that are enormous: The constraint forces at your wrist during a swing can exceed \(500\) N—far more than your muscles produce (Nesbit 2005). - Forces that are model-independent: You don’t need to know muscle physiology to compute them. They follow directly from the equations of motion and the constraints.
This chapter reveals the mathematics and physics of constraint forces, then shows why they’re the real story of the kinetic chain.
Holonomic Constraints and Their Jacobians
A holonomic constraint is an algebraic equation that restricts the configuration of the system. It has the form:
\[ \bm{\Phi}(\bm{q}) = \mathbf{0} \]
For a golf swing, examples include: - Joint connection: The wrist connects the arm to the club. This means the endpoint of the arm coincides with the base of the club: \(\bm{\Phi}_{\text{wrist}}(\bm{q}) = \bm{x}_{\text{arm}}(\bm{q}) - \bm{x}_{\text{club}}(\bm{q}) = \mathbf{0}\). - Grip constraint: Your hand is fixed on the grip, so the grip point cannot move relative to your palm: \(\bm{\Phi}_{\text{grip}}(\bm{q}) = \bm{x}_{\text{grip}} - \text{const} = \mathbf{0}\), which is the kind of wrench-bearing static coupling formal screw theory was built to describe (Crane III et al. 2022). - Ground contact: Your feet are in contact with the ground: \(z_{\text{feet}} = 0\) (vertical position is fixed).
What a Constraint Is
A constraint is a rule that limits where your body can go. The elbow joint is a constraint: your forearm can’t be at two different distances from your shoulder. The constraint enforces that the forearm is always, say, exactly 30 cm from the shoulder.
Now here’s the key insight: to enforce a constraint, the joint must push. If your arm wants to fly away from your shoulder due to centrifugal effects, the elbow joint must push inward to prevent it. That push is the constraint force.
And here’s the paradox: this push does zero net work. Why? Because the push acts along the direction of the constraint. The forearm can’t move away from the shoulder (the constraint prevents it), so the constraint force that acts in that direction doesn’t actually move the endpoint—it’s doing work, but zero net work, because the constrained motion is zero.
Yet this zero-net-work force does redirect energy between segments.
The constraint Jacobian is the matrix of partial derivatives of the constraint with respect to configuration:
\[ \bm{J}_c(\bm{q}) = \frac{\partial \bm{\Phi}}{\partial \bm{q}} \]
If there are \(n_c\) constraint equations and \(n\) configuration variables, then \(\bm{J}_c\) is \(n_c \times n\).
Example: Wrist Constraint Jacobian
For a double pendulum (shoulder and elbow), suppose the wrist constraint enforces that the endpoint of the arm is at a fixed angle relative to the global frame (simplified—imagine the wrist is locked at a specific orientation).
The constraint might be:
\[ \bm{\Phi}(q_1, q_2) = q_1 + q_2 - \theta_{\text{fixed}} \]
Then the constraint Jacobian is:
\[ \bm{J}_c = \begin{bmatrix} 1 & 1 \end{bmatrix} \]
This tells us: to maintain the constraint, motion in \(q_1\) and \(q_2\) must be coupled—they must move together. The rows of \(\bm{J}_c\) define the directions in which motion is forbidden (the normal space) and allowed (the null space).
The Constrained Equations of Motion
Constrained Lagrangian Dynamics
When constraints are present, the equations of motion take the form:
\[ \bm{M}(\bm{q})\ddot{\bm{q}} + \bm{C}(\bm{q}, \dot{\bm{q}}) + \bm{g}(\bm{q}) = \bm{B}(\bm{q})\bm{u} + \bm{J}_c(\bm{q})^T \bm{\lambda} \]
where: - \(\bm{M}(\bm{q})\) is the mass matrix. - \(\bm{C}(\bm{q}, \dot{\bm{q}})\) contains Coriolis and centrifugal terms. - \(\bm{g}(\bm{q})\) is gravity. - \(\bm{B}(\bm{q})\bm{u}\) is the applied torque (muscles). - \(\bm{J}_c(\bm{q})^T \bm{\lambda}\) is the constraint force term.
The vector \(\bm{\lambda} \in \mathbb{R}^{n_c}\) contains the Lagrange multipliers, one for each constraint. These are the magnitudes of the constraint forces, expressed in the constraint coordinate directions.
Reading the Constrained EOM
The equation says: acceleration (left side, \(\bm{M}\ddot{\bm{q}}\)) comes from four sources (right side):
- Gravity and Coriolis (\(\bm{C}, \bm{g}\)): passive forces we’ve already discussed.
- Muscle torques (\(\bm{B}\bm{u}\)): active control.
- Constraint forces (\(\bm{J}_c^T \bm{\lambda}\)): the mysterious term.
The constraint force term is special: it’s multiplied by \(\bm{J}_c^T\). This means the constraint force acts in the directions perpendicular to motion (normal to the constraint surface). That’s why it does zero net work—the motion is always tangent to the constraint surface, perpendicular to the force.
But because the force is perpendicular, it can have a large magnitude while doing zero work. It’s like pressing on a table: you push down hard, the table pushes back with equal force, but nothing moves, so zero work is done. Yet the force is there, enormous, redistributing internal stresses.
Solving for the Constraint Forces
Computing Lagrange Multipliers
To find \(\bm{\lambda}\), we use the constraint that \(\bm{\Phi}(\bm{q}) = \mathbf{0}\) must be maintained at all times. Taking the time derivative twice:
\[ \bm{J}_c(\bm{q}) \ddot{\bm{q}} = -\dot{\bm{J}}_c(\bm{q}, \dot{\bm{q}}) \dot{\bm{q}} - \bm{J}_c(\bm{q}) \ddot{\bm{q}}_{\text{unconstrained}} \]
where \(\ddot{\bm{q}}_{\text{unconstrained}}\) is what the acceleration would be without the constraint.
Substituting the constrained EOM:
\[ \bm{J}_c \bm{M}^{-1} \left[ \bm{B}\bm{u} + \bm{J}_c^T \bm{\lambda} - \bm{C} - \bm{g} \right] = -\dot{\bm{J}}_c \dot{\bm{q}} - \bm{J}_c \ddot{\bm{q}}_{\text{unconstrained}} \]
Rearranging:
\[ \bm{J}_c \bm{M}^{-1} \bm{J}_c^T \bm{\lambda} = -\bm{J}_c \bm{M}^{-1} (\bm{B}\bm{u} - \bm{C} - \bm{g}) - \dot{\bm{J}}_c \dot{\bm{q}} - \bm{J}_c \ddot{\bm{q}}_{\text{unconstrained}} \]
Solving this linear system gives \(\bm{\lambda}\).
Why This Procedure Works
The constraint equation \(\bm{\Phi} = 0\) is an algebraic restriction. If you differentiate it twice, you get an equation involving \(\ddot{\bm{q}}\). This equation says: “the acceleration must be consistent with maintaining the constraint.”
You plug in the EOM (which relates \(\ddot{\bm{q}}\) to forces) and solve for the unknown constraint forces \(\bm{\lambda}\).
The matrix \(\bm{J}_c \bm{M}^{-1} \bm{J}_c^T\) is symmetric and positive definite (in well-posed problems). It’s invertible, so the linear system has a unique solution.
The Null Space and Range of the Constraint Jacobian
{Null Space and Range of \(\bm{J}_c\)}{def:jacobian_spaces} The constraint Jacobian \(\bm{J}_c\) defines two fundamental subspaces:
- Null space: \(\text{null}(\bm{J}_c) = \{\bm{v} : \bm{J}_c \bm{v} = \mathbf{0}\}\). These are velocity directions that maintain the constraint (allowed motions).
- Range: \(\text{range}(\bm{J}_c^T)\). This is the space of possible constraint forces. Constraint forces must lie in this subspace.
A fundamental property: \(\text{null}(\bm{J}_c) \perp \text{range}(\bm{J}_c^T)\). Allowed motions are perpendicular to constraint forces—which is exactly why constraint forces do zero work.
Example: Null Space for the Wrist Hinge
For a double pendulum with the constraint \(q_1 + q_2 = \theta_{\text{fixed}}\), the constraint Jacobian is:
\[ \bm{J}_c = \begin{bmatrix} 1 & 1 \end{bmatrix} \]
The null space is found by solving:
\[ \begin{bmatrix} 1 & 1 \end{bmatrix} \begin{bmatrix} v_1 \\ v_2 \end{bmatrix} = 0 \implies v_1 = -v_2 \]
So the null space is one-dimensional:
\[ \text{null}(\bm{J}_c) = \text{span}\left\{ \begin{bmatrix} 1 \\ -1 \end{bmatrix} \right\} \]
Interpretation: The only way to move while maintaining the constraint is to increase \(q_1\) and decrease \(q_2\) by the same amount (or vice versa). This keeps their sum constant.
The range of \(\bm{J}_c^T\) is:
\[ \text{range}(\bm{J}_c^T) = \text{span}\left\{ \begin{bmatrix} 1 \\ 1 \end{bmatrix} \right\} \]
Interpretation: The constraint force acts along the direction \((1, 1)\)—i.e., simultaneously at both joints. This makes sense: the constraint couples the two joints, so forces must be applied at both to maintain coupling.
Constraints Restrict the Effective DOF
A system with \(n\) configuration variables and \(n_c\) holonomic constraints has only \(n - n_c\) effective degrees of freedom.
For example: - A double pendulum has \(n = 2\) DOF (shoulder and elbow angles). - If you lock the wrist (adding one constraint), you have \(n_c = 1\), leaving \(2 - 1 = 1\) effective DOF. - Motion can only occur along the null space of the constraint Jacobian.
The constraint forces are whatever is needed to project the “free” dynamics onto the null space—they’re automatic consequences of maintaining the constraint.
Energy Transfer via Constraint Forces
Zero Power Constraint Condition
For ideal, scleronomic (time-independent) constraints \(\bm{\Phi}(\bm{q}) = \mathbf{0}\), the power delivered by a constraint force is:
\[ P_{\text{constraint}} = \bm{\lambda}^T \bm{J}_c(\bm{q}) \dot{\bm{q}} \]
Since \(\bm{J}_c(\bm{q}) \dot{\bm{q}} = \frac{d\bm{\Phi}}{dt}\) and \(\bm{\Phi}(\bm{q}(t)) = \mathbf{0}\) for all \(t\) (with no explicit time dependence \(\frac{\partial \bm{\Phi}}{\partial t} = \mathbf{0}\)), we have:
\[ \frac{d\bm{\Phi}}{dt} = \mathbf{0} \]
Therefore:
\[ P_{\text{constraint}} = \mathbf{0} \]
Under ideal scleronomic conditions, constraint forces do zero net power to the system.
The Power Paradox
Here’s where the magic happens. Constraint forces: - Do zero net work on the system (total energy is unchanged). - Yet they redistribute energy between segments dramatically.
Think of it this way: The constraint force on the arm at the wrist might do positive work (removing kinetic energy from the arm). Simultaneously, the reaction force on the club does negative work of the same magnitude (adding kinetic energy to the club). These cancel out globally, but locally, energy has been transferred.
This is the fundamental mechanism of the kinetic chain. The arm slows down (losing energy) while the club speeds up (gaining energy). The total is constant, but the distribution shifts dramatically.
At any instant, for a partitioned system \(\bm{q} = \begin{bmatrix} \bm{q}_1 \\ \bm{q}_2 \end{bmatrix}\) with mass matrix \(\bm{M}(\bm{q}) = \begin{bmatrix} \bm{M}_{11} & \bm{M}_{12} \\ \bm{M}_{21} & \bm{M}_{22} \end{bmatrix}\), the total kinetic energy is:
\[ T = T_{\text{arm}} + T_{\text{club}} + T_{\text{cross}} = \frac{1}{2}\dot{\bm{q}}_1^T \bm{M}_{11}(\bm{q}) \dot{\bm{q}}_1 + \frac{1}{2}\dot{\bm{q}}_2^T \bm{M}_{22}(\bm{q}) \dot{\bm{q}}_2 + \dot{\bm{q}}_1^T \bm{M}_{12}(\bm{q}) \dot{\bm{q}}_2 \]
where \(T_{\text{cross}} = \dot{\bm{q}}_1^T \bm{M}_{12}(\bm{q}) \dot{\bm{q}}_2\) accounts for cross-coupling kinetic energy between the segments due to the off-diagonal mass matrix block \(\bm{M}_{12}\).
The power going to each segment is:
\[ \begin{aligned} \frac{dT_{\text{arm}}}{dt} &= \bm{\tau}_{\text{arm}} \cdot \dot{\bm{q}}_1 + \bm{F}_{\text{constraint, arm}} \cdot v_{\text{arm}} \\ \frac{dT_{\text{club}}}{dt} &= \bm{\tau}_{\text{club}} \cdot \dot{\bm{q}}_2 + \bm{F}_{\text{constraint, club}} \cdot v_{\text{club}} \end{aligned} \]
The constraint forces ensure that:
\[ \bm{F}_{\text{constraint, arm}} \cdot v_{\text{arm}} + \bm{F}_{\text{constraint, club}} \cdot v_{\text{club}} = 0 \]
But individually, each term can be nonzero, allowing energy to transfer from arm to club.
Constraint Force in a Double Pendulum Wrist Hinge
Example: Wrist Constraint Force in the Double Pendulum
Return to the double pendulum from Chapter 6, but now add a constraint: the wrist is locked so that \(q_1 + q_2 = \text{const} = 115°\). This forces the club to move relative to the arm in a coordinated way.
The constraint Jacobian is:
\[ \bm{J}_c = \begin{bmatrix} 1 & 1 \end{bmatrix} \]
At the configuration from Chapter 6 (\(q_1 = 45°\), \(\dot{q}_1 = 8\) rad/s, \(q_2 = 70°\), \(\dot{q}_2 = 12\) rad/s), the unconstrained accelerations (from the ZTCF family) are:
\[ \begin{aligned} \ddot{q}_1^{\text{unconstrained}} &\approx -5.56 \text{ rad/s}^2 \\ \ddot{q}_2^{\text{unconstrained}} &\approx 21.9 \text{ rad/s}^2 \end{aligned} \]
With the constraint, \(\ddot{q}_1 + \ddot{q}_2 = 0\) must hold. But the unconstrained accelerations are \(-5.56\) and \(21.9\), which sum to \(16.34 \neq 0\). The constraint must enforce this.
Using the procedure from Section 1.3, we solve for \(\lambda\). The constraint acceleration condition is:
\[ \bm{J}_c \ddot{\bm{q}} = -\dot{\bm{J}}_c \dot{\bm{q}} = 0 \quad \text{(since $\bm{J}_c$ is constant)} \]
Thus:
\[ \ddot{q}_1 + \ddot{q}_2 = 0 \]
From the constrained EOM:
\[ \begin{aligned} \bm{M}_{11} \ddot{q}_1 + \bm{M}_{12} \ddot{q}_2 + C_1 + g_1 &= 0 + \lambda \\ \bm{M}_{21} \ddot{q}_1 + \bm{M}_{22} \ddot{q}_2 + C_2 + g_2 &= 0 + \lambda \end{aligned} \]
The constraint force acts equally on both joints, with generalized-force components of \(\lambda\) at each joint.
With the constraint \(\ddot{q}_2 = -\ddot{q}_1\), we have two equations and two unknowns (\(\ddot{q}_1\) and \(\lambda\)):
\[ \begin{aligned} \bm{M}_{11} \ddot{q}_1 - \bm{M}_{12} \ddot{q}_1 + C_1 + g_1 &= \lambda \\ -\bm{M}_{21} \ddot{q}_1 + \bm{M}_{22} \ddot{q}_1 + C_2 + g_2 &= \lambda \end{aligned} \]
Solving (using values from Chapter 6):
\[ \begin{aligned} (0.71 - 0.014) \ddot{q}_1 + (-1.09) + 2.51 &= \lambda \\ (-0.014 + 0.016) \ddot{q}_1 + 0 + (-0.27) &= \lambda \end{aligned} \]
From the second equation:
\[ 0.002 \ddot{q}_1 = \lambda + 0.27 \]
From the first equation:
\[ 0.696 \ddot{q}_1 = \lambda + 1.09 - 2.51 = \lambda - 1.42 \]
Eliminating \(\lambda\):
\[ 0.696 \ddot{q}_1 - 0.002 \ddot{q}_1 = -1.42 - 0.27 \]
\[ 0.694 \ddot{q}_1 = -1.69 \implies \ddot{q}_1 \approx -2.44 \text{ rad/s}^2 \]
And:
\[ \lambda = 0.002 \times (-2.44) + 0.27 \approx -0.005 + 0.27 \approx 0.265 \text{ Nm} \]
Interpretation:
With the constraint in place, the shoulder decelerates more slowly (\(-2.44\) rad/s\(^2\) vs unconstrained \(-5.56\)), and the elbow accelerates less (\(+2.44\) rad/s\(^2\) vs unconstrained \(+21.9\)).
The constraint force magnitude is \(\lambda \approx 0.265\) Nm. This is the “effort” required to maintain the constraint. It appears equally at both joints.
Now let’s compute the power transfer. The power flowing to each segment through the constraint is:
\[ \begin{aligned} P_{\text{arm, constraint}} &= \lambda \cdot \dot{q}_1 = 0.265 \times 8 = 2.12 \text{ W} \\ P_{\text{club, constraint}} &= \lambda \cdot \dot{q}_2 = 0.265 \times 12 = 3.18 \text{ W} \end{aligned} \]
Those raw numbers are not the full power balance. The zero-power result above applies to the coupled generalized constraint force over the whole system, while the segment-level bookkeeping depends on how the constraint is projected into each coordinate. The safe conclusion is:
Key Finding: The constraint force is small (\(\sim 0.3\) Nm) but its direction is crucial: it acts to slow the shoulder and speed the club. Over the course of the downswing, integrated over time, these forces accumulate to produce the dramatic energy transfer from arm to club.
Why the Wrist Hinge Matters
The wrist isn’t just a joint that holds the club. It’s a constraint that couples the arm and club together. This coupling allows the inertia of the arm to slow down while the inertia of the club speeds up—energy is transferred through the constraint force.
If the wrist were completely free (unconstrained), the arm and club would move independently. The arm would decelerate under gravity, while the club would whip around due to Coriolis. But they wouldn’t exchange energy efficiently.
By constraining the wrist (even just partially, as it is in a real swing where the wrist has limited flexibility), you create a mechanism for energy transfer. The constraint force is the agent of that transfer.
Constraint Forces vs. Joint Reaction Forces
An important distinction: constraint forces are internal to the system. They arise from the geometry and inertia of the coupled system, not from external muscles pushing or pulling.
By contrast, a joint reaction force is often measured experimentally—it’s the force that one segment (e.g., the arm) exerts on another (e.g., the club) across the joint.
In a biomechanical analysis, you might measure the force at the wrist using an instrumented grip. This measurement captures the constraint force, but it’s expressed as a force (in Newtons) rather than a torque (in Newton-meters).
The relationship is:
\[ \bm{F}_{\text{joint}} = \bm{J}_c^T \bm{\lambda} / L_{\text{joint}} \]
where \(L_{\text{joint}}\) is the lever arm (moment arm) of the force about the joint.
Why Constraint Forces Are the Real Story of the Kinetic Chain
The traditional understanding of the kinetic chain says: “The hips drive the shoulders, which drive the arms, which drive the club.” This is taught as a sequence of muscular activations—hip muscles accelerate the pelvis, which mechanically couples to the torso, which couples to the shoulders, etc.
The correct understanding, revealed by constraint force analysis, is:
The kinetic chain works because constraints redirect inertia.
Here’s why: 1. You build momentum in your torso by rotating your hips. 2. The shoulder constraint (the joint connecting arm to torso) prevents the arm from flying away radially. Instead, it couples the arm’s motion to the torso’s rotation. 3. This coupling creates a constraint force, which acts on the arm. This force is what actually accelerates the arm—not a muscular torque, but a constraint force arising from the torso’s inertia. 4. Similarly, the wrist constraint couples the club to the arm. The arm’s inertia, via the constraint force, accelerates the club. 5. By the time the club reaches the club, the constraint force is enormous (hundreds of Newtons), far exceeding what any muscle can produce.
This is why the kinetic chain works: it’s not about muscular effort cascading down from the hips. It’s about mechanical coupling—constraints that efficiently redirect the momentum of proximal segments to distal segments.
And this is why active muscular control is secondary: muscles are too weak compared to the constraint forces. Muscles set up the initial conditions and make fine adjustments. Constraints do the heavy lifting.
Energy and Momentum Transfer Through Constraints
A constraint does zero net work. But in a multi-segment system, it can transfer energy from one segment to another.
The mechanism is simple: - Segment A (proximal, e.g., arm) has kinetic energy. - The constraint force acts on A, reducing its kinetic energy. - By Newton’s third law, an equal and opposite force acts on B (distal, e.g., club). - This force accelerates B, increasing its kinetic energy. - The magnitudes are such that the total energy is conserved (zero net work), but the distribution shifts: A loses energy, B gains energy.
This transfer is most efficient when the proximal segment is large and slow, and the distal segment is small and fast. The golf swing achieves this through the staggered unlocking of joints: the hips unlock first, building momentum; then the shoulders, then the arms, then the wrists. Each step transfers momentum to the next, and by the time the club is freed, it has accumulated enormous speed from the prior segments’ momentum—all via constraint forces, not muscular effort.
Constraint Forces in Real Swings: Grip Forces and Ground Reaction Forces
Example: Grip Force During a Golf Swing
A professional golfer’s grip force during the downswing is typically \(50\)–\(150\) N (measured via an instrumented club). This force is not a muscular torque directly. Instead, it’s the reaction force that the grip constraint exerts on the club.
During the downswing: 1. The golfer’s hands are constrained to the grip (the grip doesn’t slip relative to the hand). 2. The hand itself is accelerating as part of the arm rotation. 3. To keep the club moving with the hand (enforcing the constraint), the hand must push on the grip. 4. This push is the grip force.
The grip force arises automatically from the constraint, without explicit muscular effort to “grip harder.” It’s a consequence of the kinematics and inertias.
However—and this is important—the golfer can modulate the grip force by changing muscle tension or by allowing some slip. A gentle grip (low force) allows some flexibility; a firm grip (high force) enforces the constraint tightly. This modulation is a control knob.
Insight: The large grip forces in a golf swing (often 100+ N at late downswing) are not evidence of muscular effort. They’re evidence of massive inertial forces being channeled through the constraint. Muscles are simply holding the constraint in place, not driving the acceleration.
Example: Ground Reaction Forces and the Closed Kinematic Chain
The golfer’s feet are in contact with the ground, which is a constraint: the feet cannot move downward (the ground prevents it).
During the downswing, this ground constraint produces enormous vertical forces. A 200-pound golfer might experience peak ground reaction forces of \(1.5\)–\(2\) times body weight, or \(300\)–\(400\) pounds-force. These forces are constraint forces—they arise from the geometry of contact with the ground.
These ground reaction forces are crucial because they close the kinetic chain. Without the ground, the golfer would slide backward as they rotate forward. The ground reaction force (a constraint force) prevents this, keeping the lower body anchored while the upper body rotates.
Insight: The ground reaction force is not something the golfer’s muscles produce. It’s a constraint force that the ground produces in reaction to the golfer’s motion. Muscles control how the golfer moves, but the ground handles the constraint enforcement.
Summary and Key Insights
- What constraints are: Holonomic constraints are algebraic restrictions on configuration. They enforce that joints are connected, grips don’t slip, feet stay on the ground, etc.
- The constraint Jacobian: \(\bm{J}_c\) maps the constraint equations to configuration space. Its null space defines allowed motions; its range defines constraint force directions.
- Constrained EOM: Constraint forces appear as \(\bm{J}_c^T \bm{\lambda}\) in the equations of motion. The Lagrange multipliers \(\bm{\lambda}\) are the constraint force magnitudes.
- Zero net power: Constraint forces do zero net work to the system. But locally, they redistribute energy between segments—this is the kinetic chain.
- Double pendulum example: The wrist constraint couples shoulder and elbow, producing a small constraint force (\(\sim 0.3\) Nm) that redirects energy from arm to club.
- Grip and ground forces: Grip forces (\(50\)–\(150\) N) and ground reaction forces (\(1300\)–\(1800\) N) are constraint forces, not muscular efforts. Muscles hold the constraints, but physics enforces them.
- The kinetic chain is constraint-driven: Energy transfers from hips to shoulders to arms to club through constraints, not through muscular cascades. Muscles set up initial momentum; constraints redirect it.
- Control through constraints: A golfer controls the swing by modulating constraint rigidity (grip firmness, body stiffness) and by controlling the initial momentum building. Once momentum is built, constraints take over, and the motion becomes nearly deterministic.
What comes next: Understanding constraints, we now ask: what if there are multiple constraints forming a closed loop? What if the body is not just a serial chain (hips → shoulders → arms → club) but a parallel mechanism with closed kinematic chains? That’s the subject of Chapter 9.
But first, Chapter 8 extends the double pendulum to a triple pendulum, adding the wrists, and shows how the constraint-driven energy transfer becomes even more dramatic with an additional segment.
Chapter Exercises: Constraint Forces
{Conceptual} Explain in plain language: Why can a grip force of 100 N (an illustrative value; actual grip forces vary with swing speed and technique; approximately 50–150 N is a typical estimated range (Hume et al. 2005)) do zero net work on the system, yet transfer energy from the arm to the club?
(Hint: Think about the direction of the grip force and the direction of motion. Are they parallel or perpendicular?)
{Geometric} For the wrist constraint in the double pendulum (Section 1.6), sketch the configuration of the arm and club when \(q_1 + q_2 = 115°\). Identify the null space (allowed motion) and the direction of the constraint force.
{Quantitative} A single-segment arm has mass \(m = 2\) kg, length \(L = 0.7\) m, and moment of inertia \(I = 0.08\) kg\(\cdot\)m\(^2\). It’s rotating at \(\omega = 10\) rad/s and experiences an inward centrifugal acceleration.
The centrifugal force at the endpoint is \(F_c = m L \omega^2\). Compute this force. How does it compare to the arm’s weight? Why must the shoulder joint provide a large constraint force to prevent the arm from flying outward?
{Computational} Set up a double pendulum with and without a wrist constraint. 1. First case: free double pendulum (unconstrained). Start from the same initial condition as the constrained case. 2. Second case: double pendulum with \(q_1 + q_2 = \text{const}\) constraint. 3. Integrate both forward for 0.1 seconds. 4. Plot the trajectories and the computed Lagrange multiplier \(\lambda(t)\). 5. Interpret: When is \(|\lambda|\) large? When is it small? Why?
{Application: Real-World Measurement} Watch a high-speed video of a golfer’s swing (at least 1000 fps). Identify moments when: 1. The wrist is locked (constraint is tight). 2. The wrist is free (constraint is loose).
How does the arm motion change in each case? How does the club behave when the wrist constraint is tight vs. loose?
{Analysis: Energy Transfer} For the double pendulum with constraint (Section 1.6): 1. Compute the kinetic energy of the arm before and after applying the constraint. 2. Compute the kinetic energy of the club before and after. 3. How much energy has been transferred from arm to club? Is it consistent with zero net power for the constraint?