The mechanical rower: Construction, validity, and reliability of a test rig for wind braked rowing ergometers

• Published in: Journal of biomechanics 2020-06, Vol.106, p.109833-109833, Article 109833
• Main Authors: Mentz, Lennart, Engleder, Thomas, Schulz, Georg, Winkert, Kay, Steinacker, Jürgen M., Treff, Gunnar
• Format: Article
• Language: English
• Subjects: Accuracy; Coefficient of variation; Concept 2; Control algorithms; Displacement; Electric motors; Ergometers; Ergometry; Evaluation; Humans; Laboratories; Load cells; Mathematical analysis; Mechanical power output; Reliability analysis; Reproducibility of Results; Rowing; Sensors; Testing; Validation; Validity; Variables; Velocity; Water Sports; Wind
• ISSN: 0021-9290; 1873-2380
• DOI: 10.1016/j.jbiomech.2020.109833

Wind braked rowing ergometers are used worldwide for training and testing of rowers, but data on validity and reliability of the calculated mechanical power output are scarce. Studies published so far are based on data generated by human rowers, inevitably adding biological variability without any option to clamp particular variables like stroke structure or force. To this end, we developed a test rig for rowing ergometers aiming to generate valid and reliable stroke structures (i.e. force–displacement curves). Briefly, the rig consists out of a frame connected to the ergometer. The handlebar of the rowing ergometer is attached to a sledge that can be displaced on a linear drive by a motor that is controlled by torque curves which are derived from elite rowers. A load cell between handlebar and chain and an incremental linear transducer allow criterion measures of force and distance of displacement to calculate mechanical power output. To evaluate the validity of the machine generated force–displacement curves, three different stroke structures were compared to the respective human reference curves. To evaluate reliability, series of 50 consecutive strokes were performed for 10 times. Validity of the curves was indicated by small differences in stroke-force, -distance, and -work (≤ |−7.8|%) between machine generated and human generated curves. Mean power output of the test series was 445 ± 1 W with a coefficient of variation of 0.53% between series. Hence, the test rig allows to generate valid and reliable rowing strokes on wind braked rowing ergometers.