TECHNICAL ANALYSISv2HYUNDAI × BOSTON DYNAMICSDOCUMENT UM-KICK-02
Atlas Unseen Moves: Kick Mechanics
A physics-based reading of an anatomically anomalous free kick.
Abstract
At the moment of strike, Atlas's body is rotated 66° from the goal direction —
body forward azimuth 26°, line to goal 92°. The ball
departs at 81° azimuth, which is 55° off body forward —
an off-axis strike that loads angular momentum into the ball. Over the following 1.05 s of
flight the ball arcs to a peak lateral position of x = +2.4 m (well wide of the goal
frame) before the Magnus force curves it back through a 1.8-metre lateral arc
and into the post. The body-vs-kick mismatch and the resulting Magnus curve are the central demonstration: a
configuration human anatomy does not support, resolved by aerodynamic spin into a goal the body is not pointing
toward.
CINE · 03 Atlas free-kick — full take · 120 fps · f001 → f815
Source · Atlas(Walk_Twist_Kick)_CINE_03.mov
§1 · CONFIGURATION
The setup
Atlas's body is rotated 66° from the goal direction. The ball must curve through space to reach a goal the body is not pointing toward.
§2 · POWER + SPIN
Impulse transfer
Force, energy, and spin are delivered in a single 8.3-ms frame at the foot–ball boundary.
§3 · TRAJECTORY
The curve
Sidespin produces a 2.4-metre Magnus arc — the ball travels wide of goal, then bends back into the post.
§1The Setup — Body Faces Away From GoalPILLAR I · CONFIGURATION
At the moment of strike, Atlas's body is rotated 66° from the line to goal.
The pelvis forward axis points at 26° azimuth; the goal lies at
92°. The ball departs at 81° —
55° off body forward — then arcs through 2.4 m
of lateral overshoot before the Magnus deflection curves it back to the post. The setup is the wow: the body never
points where the ball ends up.
FIG · TOP-DOWN GEOMETRY
Atlas's body axis is 90° to the kick line. Atlas and ball both lie on the world X axis; the kick fires perpendicular along the Y axis toward goal.
Fig. 1 · The setup
Top-down geometry at f592. Atlas faces 26° while the goal lies at
92°; the ball launches at 81° and curves through a
2.4-metre lateral overshoot on its way to the post.
Measured from Atlas (Walk_Twist_Kick)_v03.blend at scene FPS 120.
§1.1Striking surface
Atlas's flat rectangular foot enables the off-axis strike. Where a human curved instep would constrain contact along
the body forward direction (Lees & Nolan 1998), the flat plate can deliver impulse at any angle to body
orientation — including the 55° off-axis angle measured here. The geometry of the
striking surface is what makes the setup possible.
Foot length
28cm
Heel to toe, Atlas D1 nominal. Rigid sole.
Foot width
10cm
Outside-of-foot contact area for redirection.
Pelvis yaw
26°
Body forward azimuth at f592.
Foot velocity azimuth
46.7°
Pre-contact (f591) plane of the strike.
Ball launch azimuth
81°
Post-impact direction of momentum.
Contact angle
34°
Foot motion vs ball departure. The geometry that generates spin.
Fig. 2 · Kick phases
Standard soccer kick phases (Approach → Plant → Loading → Strike → Follow-through). The Atlas sequence follows the
same five biomechanical phases as documented in human kickers, but the contact event (P4) substitutes a flat plate
for the curved instep.
Adapted from Lees, A. & Nolan, L. (1998). The biomechanics of soccer: A review.
J Sports Sci 16(3):211–234.
FIG. 1 · THE SETUP — BODY FACES AWAY FROM GOAL
Wide hero at strike (f592). Body forward at 26°, line to goal at 92° — an off-axis configuration of 66°.
§2Impulse TransferPILLAR II · POWER + SPIN
Across a single animation frame — 8.33 ms — the foot delivers the energy and
angular momentum that will define the next 1.05 seconds of flight. The strike is hard, off-centre,
and rotational. The off-axis impulse — body pointing at 26°, ball
launched at 81° — is the mechanism that imparts spin. A flat plate
striking a sphere off-centre generates angular momentum about a near-vertical axis.
Body-to-kick mismatch
55°
Body forward (26°) vs ball launch (81°). The mechanism that loads spin.
Foot impact velocity
14.81m/s
Pre-contact (f591). Peak: 31.77 m/s at f588.
Average contact force
1364N
J/Δt over the 8.33-ms contact frame. ≈ 1.36 kN.
Ball launch velocity
26.43m/s
Peak in flight: 32.39 m/s (117 km/h).
Energy to ball
150J
95% of foot KE lost in the 8.33-ms contact window.
Implied spin rate
6.4rev/s
≈ 384 rpm. High-spin free-kick regime.
FIG. 2 · THE STRIKE — CONTACT MECHANICS
Close-up at contact. Foot 14.8 m/s into the ball, departs at 26.4 m/s and 30.3° elevation, ~6.4 rev/s of induced sidespin.
Fig. 3 · Magnus effect
Pressure differential around a spinning ball produces a lateral force FM perpendicular to the flight direction.
Linear impulse
$$\vec{J} = \vec{F}\,\Delta t = m_\text{ball}\,\vec{v}_\text{ball}$$
§3The Curve — Ball Flight Through Space and BackPILLAR III · TRAJECTORY
The ball does not travel toward the goal. It travels past it, then curves back.
Peak lateral displacement reaches x = +2.40 m at t = 0.40 s
— well outside the goal frame — before the Magnus force pulls it back to x = 0
at the post. A spin-free projectile from the same launch state would cross the goal plane at
x = 2.57 m, missing by the full width of the goal.
Lateral curl magnitude
1.82m
From peak X back to goal centre. The signature of spin.
Peak lateral position
2.40m
At t = 0.40 s (f640). Outside the 7.3 m goal frame.
No-spin miss
2.6m
Where the ball would cross y = 11 m under pure ballistics.
Peak height
2.80m
Apex Z at f640. Clears the 1.8 m wall by 1.0 m.
Flight time
1.05s
Strike (f592) → settled in goal (f815).
Launch elevation
30.3°
Above horizontal, measured at f598.
FIG. 3 · THE CURVE — BALL FLIGHT TO POST
3/4 side-rear view. Ball departs the contact, peaks at +2.4 m of lateral offset, then curls 1.82 m back across the goal mouth.
Fig. 4 · Magnus deflection
Magnus deflection in a right-foot free kick. Outside-of-foot spin imparts Magnus force that curves the ball toward goal.
Side view · YZ plane
Gravity arc to apex, fall into goal. Dashed = no-spin reference.
measuredno-spinwall / goal
Top view · XY plane · the curve
Out to peak X = +2.40 m, then Magnus deflection bends the ball back to the post.
measured (curving)no-spin (straight)
Perspective · oblique 3D
3/4 view of the full curving trajectory. Measured (red) vs no-spin straight line.
The Atlas demonstration is a small-scale instance of the spinning-ball spiral
described by Dupeux et al. (2010). As the ball decelerates from 32.39 m/s through
the curl-back phase, its Magnus radius tightens — so the trajectory bends harder near the end
of flight, finding the post from outside the goal frame.
Dupeux, G., Le Goff, A., Quéré, D., Clanet, C. (2010).
The spinning ball spiral. New J Phys 12, 093004.
§4MethodologyDATA + ASSUMPTIONS
Source data
Blender 5.1 animation, 120 fps. World-space coordinates sampled from the rig pose at each frame; the ball trajectory is keyed with 9 control points and BÉzier (auto-clamped) interpolation.
Mass & size
mfoot = 1.5 kg (Atlas D1 distal-leg nominal). mball = 0.43 kg, r = 0.11 m (FIFA Law 2, size 5 midpoint).
Contact window
Δt = 1/120 s = 8.33 ms. Single-frame analog at scene FPS.
FM = ½ρACLv². Lift coefficient backed out of the observed lateral acceleration during the curl-back phase. Spin parameter Sp = ωr/v with CL ≈ Sp in the linear regime; implied ω = 6.4 rev/s.
No-spin reference
Drag-free projectile, a = (0, 0, −9.81) m/s², seeded with the post-impact state at f598.
Bones queried
head.nky · base · r_foot_IK_ctrl · l_foot_IK_ctrl
Preservation
Source .blend opened in --background mode and never written. MD5 hash unchanged before and after extraction.