Abstract
Introduction: Added mass is the mass of water a swimmer has to accelerate in addition to his body during changes in velocity. It is an important concept in determining the total energy expenditure of human swimmers during unsteady motion. It is also important to know the added mass when using indirect methods to find a swimmer¿s passive drag. As added mass determination is dependent on specific measurement methods often unavailable to the common sports scientist, it would be worth exploring the possibility to estimate AM from body size measurements. The aim of our study was thus to find the added mass for human swimmers and its relationship with body size measures, and to construct predictive linear regression statistics. Methods: Eleven male adults (aged 25.2±4.9 years), 10 females (22.7±3.6 years) and 10 boys (13.7±0.8years) were included in this study. The subjects were connected to a 2.8m long bar with handles, attached with springs (stiffness k=318N/m) and a force cell. By oscillating this system vertically and registering the time period of oscillations it is possible to find the added mass of the swimmer, given the known masses of the bar and swimmer. Body weight and body length with arms stretched over the head was found using conventional methods, the frontal surface area, and the widest and deepest part of the swimmers body was measured by digitizing a still picture. Linear stepwise regression statistics were applied to find a suitable way of predicting added mass from body size measures. Results: A mean added mass of 17.0±3.8kg, corresponding to 26.3±2.9% of body mass was found for the whole group, and 21.4±2.5, 15.0±1.2 and 14.5±3.1kg for the men, women and boys respectively. For the whole group of subjects, an r2 value of 0.79 was reached for linear regression with body mass (BM). Including frontal surface area, reaching height (representing the underwater characteristic length) and fineness ratio did not enhance the r2. Including body width (BW) and body depth (BD) to the regression increased the r2 to 0.85. Added mass (AM) can thus be predicted by: AM=0,22*BM+0.25*BW-0.17*BD-2.4. Looking at males and females separately, r2 was 0.90 for males (including BM, BW and BD) and r2=0.56 for females (including only BM). Conclusions: By including body mass in the regression we could explain 79% of the variations in added mass for the whole group. Adding other body size measures such as body width and depth improved the accuracy of this prediction to a r2=0.85. For the rest of the variation in added mass other variables must be investigated, or is due to measurement errors. The human factor must not be forgotten, the position of the swimmer at the oscillating bar may vary, affecting the added mass to a small degree. It is concluded that a reasonable prediction of added mass needs only body mass as input, and is further enhanced by including body with and depth, but not for the females.