Simulation of tossing - biomechanics and motor control
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Tossing tasks, a kind of manual material handling (MMH), extends from lifting tasks. Unlike lifting task studies, there were few reports studying tossing tasks in term of Biomechanics. In this study, the tossing tasks were studied in terms of Biomechanics and Fitts’ law, and the tossing tasks were also simulated based on the lifting task simulation.
To validate the tossing task simulation, a 2×2×2×3 factorial experimental design (tossing distance, tossing height, weight of load, and tossing clearance) was established. Ten volunteer subjects were recruited from the students of Texas Tech University. A motion capture recorded their motion throughout their movement while doing the tossing tasks.
The first analysis was the comparison between the tossing and lifting tasks in order to validate that the tossing task simulation can be guided by the lifting tasks simulation. The comparison analyses were considered the joint angle trajectories, the load trajectories and velocities, and the compressive and shear forces of the L5/S1 joint. The comparison showed that similarities between both kinds of tasks were found in the first half of the activity duration. On the other hand, dissimilarities between both kinds of tasks were found in the second half of the activity duration. The differences of the joint angle trajectories were found because of the different objective at the end of the tasks. Subjects slowed the load velocity before it impacted on the table while they increased the load speed before the load was released from their hands. Although similarities and dissimilarities between tossing and lifting tasks were found, the tossing tasks could be separated into three phases with four key postures (initial, load-close to body, aiming, and releasing postures) which was similar to the lifting tasks. Therefore, the next step of this study was the evaluation of the four key postures by using the mathematical models. The second analysis was the validation of the four key posture evaluations. The objective function used for evaluation was body balancing. The center of mass (COM) of subjects while doing the tossing tasks was investigated to validated the objective function. COM investigation showed that if none of the subjects had COM beyond their feet, then body balancing can be used as the objective function. All key postures were evaluated by using body balancing with the constraints. The constraints of the 1st, 2nd, and 3rd key postures were the interaction between the human and the load position while the constraints of the 4th key posture were the load movements before and after the 4th key posture. The load movement before the 4th key posture was considered as the linear movement. The load movement after the 4th key posture was considered as the projectile movement which was affected by the tossing distance, tossing height, weight of load, and target clearance. The tossing distance and height significantly affected the projectile time and the initial angle of projectile. In addition, the tossing clearance significantly affected only the initial angle of projectile. With regard to the compressive force, only the tossing distance significantly affected the maximum compressive force on L5/S1 joint in the releasing phase. Moreover, all factors significantly effected the movement time. These investigations were considered for the 4th key posture evaluation. Therefore, the four key postures could be connected together to simulate the tossing task movement. The final analysis in this study was the validation of the tossing task simulation, which was comprised of the optimization for the positions in time frame of the 2nd and 3rd key postures and then the connections of the joint angle trajectories. The optimization methods applied three different objective functions (maximize the hand movement smoothness, maximize the smoothness of the center of the gravity (COG) movement, and minimize the muscular utilization rate (MUR)) in order to find three different positions in time frame of the 2nd and 3rd key postures.
The comparisons among three objective functions showed that the maximize the smoothness of the center of the gravity (COG) movement predicted the positions in time frame of the 2nd and 3rd key postures closest to the actual positions in time frame of the 2nd and 3rd key postures. To validate these optimization methods, all joint angle trajectories from four key postures were connected by using an 8th order polynomial equation.
The validation results showed that hand movement smoothness was the best predictor for arm movement; however, body balance was the best predictor for trunk and leg movements. As regards to the whole body posture, hand movement smoothness should be considered to simulate tossing tasks, because the important part of the tossing task movement in this study was the upper extremities. Since the subjects stood at a fixed point (no stepping) while doing the tossing tasks, the movement of the lower extremities had low variation. In this case, the lower extremities might be not concerned for tossing task simulations. Unlike the lower extremities, the upper extremities moved in a wider range and faster speed, including more varieties of hand movement compared to leg movement. In addition, the varieties of tossing factors and weight of load affected the varieties of hand movement more than the leg movement. Therefore, the movement of upper extremities should be a major consideration which could be simulated by the hand movement smoothness.