Laurence J. Ryan

Foot speed, foot-strike and footwear: linking gait mechanics and running ground reaction forces

Running performance, energy requirements and musculoskeletal stresses are directly related to the action–reaction forces between the limb and the ground. For human runners, the force–time patterns from individual footfalls can vary considerably across speed, foot-strike and footwear conditions. Here, we used four human footfalls with distinctly different vertical force–time waveform patterns to evaluate whether a basic mechanical model might explain all of them. Our model partitions the bodys total mass (1.0Mb) into two invariant mass fractions (lower limb=0.08, remaining body mass=0.92) and allows the instantaneous collisional velocities of the former to vary. The best fits achieved (R2 range=0.95–0.98, mean=0.97±0.01) indicate that the model is capable of accounting for nearly all of the variability observed in the four waveform types tested: barefoot jog, rear-foot strike run, fore-foot strike run and fore-foot strike sprint. We conclude that different running ground reaction force–time patterns may have the same mechanical basis.

A general relationship links gait mechanics and running ground reaction forces

ABSTRACT The relationship between gait mechanics and running ground reaction forces is widely regarded as complex. This viewpoint has evolved primarily via efforts to explain the rising edge of vertical force–time waveforms observed during slow human running. Existing theoretical models do provide good rising-edge fits, but require more than a dozen input variables to sum the force contributions of four or more vague components of the bodys total mass (mb). Here, we hypothesized that the force contributions of two discrete body mass components are sufficient to account for vertical ground reaction force–time waveform patterns in full (stance foot and shank, m1=0.08mb; remaining mass, m2=0.92mb). We tested this hypothesis directly by acquiring simultaneous limb motion and ground reaction force data across a broad range of running speeds (3.0–11.1 m s−1) from 42 subjects who differed in body mass (range: 43–105 kg) and foot-strike mechanics. Predicted waveforms were generated from our two-mass model using body mass and three stride-specific measures: contact time, aerial time and lower limb vertical acceleration during impact. Measured waveforms (N=500) differed in shape and varied by more than twofold in amplitude and duration. Nonetheless, the overall agreement between the 500 measured waveforms and those generated independently by the model approached unity (R2=0.95±0.04, mean±s.d.), with minimal variation across the slow, medium and fast running speeds tested (ΔR2≤0.04), and between rear-foot (R2=0.94±0.04, N=177) versus fore-foot (R2=0.95±0.04, N=323) strike mechanics. We conclude that the motion of two anatomically discrete components of the bodys mass is sufficient to explain the vertical ground reaction force–time waveform patterns observed during human running. Summary: A basic relationship that links the motion of running to the ground forces applied enables practical, motion-based predictions of force–time patterns at essentially all speeds and regardless of foot-strike mechanics.

Design and testing of a high-speed treadmill to measure ground reaction forces at the limit of human gait

Investigations focused on the gait and physiological limits of human speed have been on-going for more than a century. However, due to measurement limitation a kinetic understanding of the foot-ground collision and how these dynamics differ between individuals to confer speed and limit gait has only recently begun to come forth. Therefore, we designed and tested an instrumented high-speed force treadmill to measure the forces occurring at the limits of human performance. The treadmill was designed to maximize flexural stiffness and natural frequency by using a honeycomb sandwich panel as the bed surface and a flexible drive shaft between the drive roller and servo motor to reduce the mass of the supported elements which contribute to the systems response frequency. The functional performance of the force treadmill met or exceeded the measurement criteria established for ideal force plates: high natural frequency (z-axis = 113 Hz), low crosstalk between components of the force (Fx/Fz = 0.0020[SD = 0.0010]; Fy/Fz = 0.0016[SD = 0.0003]), a linear response (R(2) > 0.999) for loading with known weights (range: 44-3857 N), and an accuracy of 2.5[SD = 1.7] mm and 2.8[SD = 1.5] mm in the x and y-axes, respectively, for the point of force application. In dynamic testing at running speeds up to 10 m s(-1), the measured durations and magnitudes of force application were similar between the treadmill and over-ground running using a force platform. This design provides a precise instrumented treadmill capable of recording multi-axis ground reaction forces applied during the foot ground contacts of the fastest men and animals known to science.

Impact forces during running: Loaded questions, sensible outcomes

Load carriage was used as an experimental tool to evaluate the ability of an anatomically-based, two-mass model of the human body to predict vertical impact and peak forces during running from only four inputs: body weight (W_b), contact time (t_c), aerial time, (t_a), and lower-limb acceleration (a₁). Simultaneous motion and force data were acquired from seven subjects during steady-speed trials (3.0-6.0 m·s^-1) on a custom, force-instrumented treadmill under three loading conditions: unloaded (1.0 W_b), 15% added weight (1.15 W_b) and 30% added weight (1.30 W_b). Model-predicted impact and peak forces corresponded with measured values, on average, to within 14.9±1.3% and 13.8±0.6%, respectively (R² best-fits=0.82 and 0.88, n=71). Ankle jerk and velocity data derived from optical position-time data suggest wearable sensor acquisition of the model-needed inputs is fully feasible. We conclude that the two-mass model offers a promising approach to quantifying running ground reaction forces using wearable technologies.