Secondary Logo

Journal Logo

Technical Report

A Novel Approach to Identify the End of the Concentric Phase During Ballistic Upper-Body Movements

Frost, David M1,2; Cronin, John B1,3; Newton, Robert U1

Author Information
Journal of Strength and Conditioning Research: January 2010 - Volume 24 - Issue 1 - p 282-286
doi: 10.1519/JSC.0b013e31819f1e7e
  • Free



Ballistic upper-body movements such as bench press throws, in which the load is thrown at the end of the concentric phase, have become a popular means of increasing the specificity of resistance training for sport, as researchers have suggested that they are able to extend the duration of the concentric phase spent accelerating (15) and increase velocity, force, power, and muscle activity (6,15) as compared with their non-ballistic equivalents. Consequently, the majority of researchers continuing to examine power production and the optimal training load for power development have diverted their attention toward ballistic efforts (3,2,5,13,17).

Analyzing the kinematics and kinetics of a movement, in which the displacement of the load continues to increase after contact with the upper limb has ceased, necessitates the identification of the last point of contact. Although total work has been calculated by using the change in barbell displacement (mass × height × gravity) (18), failure to identify the precise time or position that the load was released from the hands may lead to an underestimation of the mean velocity, force, and power produced. If an object is released with a positive vertical velocity, the mean concentric force applied to produce such velocity will be greater than the weight of the load or alternatively that produced with the non-ballistic equivalent (Fnet = m (v2v1)/t, where v2 > v1).

Compared with non-ballistic motion, for which peak displacement has been consistently used to define the end of the concentric phase (1,10,12), there remains to be agreement among researchers as to the most appropriate definition when the load is thrown (14,15,17). Several studies have disregarded the notion entirely and evaluated all means in reference to the point of peak barbell displacement (2-6), thereby skewing conclusions and reducing the validity of their results. For example, comparing a ballistic and non-ballistic bench press performed with a 30% 1 repetition maximum (1RM) load, Cronin et al. (6) reported mean concentric forces of 261.6 and 262.1 N, respectively. Because the load was thrown during the ballistic condition, it is not possible for the mean concentric forces to be identical (i.e., the velocity at the end of the ballistic effort was greater than zero).

The 2 methods that have been reported previously used the position (14) or force data (15) to indirectly identify the point of release. Newton et al. (14) defined the end of the concentric phase as the point at which the barbell (load) reached the position it was held in the hands before initiating the eccentric phase, whereas Newton et al. (15) searched the force data array for the time at which the vertical force dropped below 0 N. Theoretically, using the force plate will exhibit greater validity as 0 N corresponds to the instance at which only the individual's body weight is being sensed; however, this contention assumes uniform acceleration of all upper-body segments and, as stated previously, is an indirect method of measurement. Therefore, the objective of this investigation was to examine the effectiveness of 3 methods (1 direct and 2 indirect) to accurately define the end of the concentric phase and calculate the height that a barbell was thrown.


Experimental Approach to the Problem

Data from a larger study (11) were used to evaluate the potential variability in using dissimilar methodological approaches to identify the end of the concentric phase during ballistic upper-body movements. To investigate this research problem, the height that the barbell was thrown was calculated using 3 different pieces of equipment: a force plate, a position transducer, and a pressure switch.


Thirty men with a minimum of 12 months of resistance training experience and a 1RM bench press greater than their body weight were recruited to participate in the larger investigation (11). The subjects' mean (±SD) age, height, body mass, and resistance training experience were 24.9 (4.9) years, 1.79 (0.06) m, 80.6 (9.8) kg, and 5.6 (3.8) years, respectively. Before the commencement of testing, all subjects read and signed an informed consent and filled out a health questionnaire approved by the human ethics committee of the university.


Testing pertinent to this technical note was performed inside a standard power rack equipped with a magnetic particle brake (Fitness Technology, Adelaide, Australia), used to prevent negative displacement of the barbell subsequent to the point of release. A bench was secured to the center of a portable 0.92- by 0.92-m force plate (Quattro Jump Model 9290AD; Kistler, Winterthur, Switzerland) using a customized steel bracket. Footpegs extending horizontally from the end of the bracket were used to accommodate various foot positions so that subjects were not obliged to place their feet on the bench or the floor, thus maintaining a degree of comfort and ensuring an accurate reading from the force plate. Before each testing session, the force plate was calibrated and zeroed with the weight of the participant and bench.

A linear position transducer (PT5A-150; Celesco, Chatsworth, CA) with a signal sensitivity of 0.244 mV·V−1·mm−1 was secured to a wood plank and positioned approximately 1.5 m directly above the center of the barbell. The transducer was zeroed at the commencement of each repetition as the acquisition software recorded the initial displacement as 0.000 m.

Two 30-cm ribbon switches (151-BBW; Tapeswitch, Farmingdale, NY), with a pressure sensitivity of 2 N, were fastened to the underside of the barbell (Figure 1). Using a common power supply of 5 V, the 2 switches were connected in parallel and wired such that the output voltage would remain low unless both were released. All raw analogue signals were A/D converted using a 16-bit data acquisition board (PCI-6220; National Instruments, Sydney, NSW, Australia) and sampled simultaneously at 2,000 Hz. Custom-written Labview software (Version 8.1; National Instruments, Austin, TX) acquired, displayed, and stored all data for further analysis.

Figure 1
Figure 1:
The pressure switches secured to the underside of the barbell with Velcro tape.

Testing Procedures

A detailed description of the entire experimental protocol has been reported previously (11); therefore, only those procedures relevant to this research note are discussed herein. Four, single, ballistic repetitions, separated by 1 minute, were performed as explosively as possible at loads of 15, 30, 45, 60, 75, and 90% 1RM, with 3-minute rest given between each percentage. The 1RM was determined using a similar protocol to that outlined by Doan et al. (8). Loads were assigned in ascending order so as to make a systematic comparison across all subjects. Participants were instructed to lower the barbell as fast as they felt manageable, without touching the chest when transitioning from the eccentric to concentric phase. Any repetition that contacted the chest or failed to come within 0.05 m of the chest was disregarded and repeated after an additional 1 minute of rest. Subjects were required to keep the same grip and foot width for the entire testing protocol and ensure that contact was maintained between their hips and back with the bench and feet with the force plate for the duration of each repetition. All 4 repetitions were analyzed; however, only the 2 fastest (mean velocity) were used for comparative purposes.

Data Analysis

The raw displacement, force, and switch data were filtered using fourth-order, zero phase shift, low-pass Butterworth filters with cutoff frequencies of 100 Hz. Initiation of the eccentric phase was defined as the first instance of negative displacement, determined by searching the position array, backward from the point of the minimum displacement (end of eccentric phase). The conclusion of the concentric phase was defined by 3 separate calculation methods corresponding to the position, force, and switch data, respectively, (a) the position at which the barbell reached the position it was held at, with outstretched arms, before the onset of the eccentric phase (time at the position of the start of the eccentric contraction minus 150 milliseconds) (14); (b) the position at which the force fell below 0 N (15); and (c) the position at which the switch voltage rose above 1 V (both hands lost contact with the barbell). Height thrown was calculated as the peak displacement of the barbell minus the position at the point defined previously as the end of the concentric phase. If the position of peak displacement was greater than or equal to the position of the end of the concentric phase, the load was not thrown and the height was calculated as 0 m. Custom-written Labview Software (Version 8.1; National Instruments) was used to analyze the data.

Statistical Analyses

The height thrown was expressed as a mean and SD for each condition and evaluated for reliability with a coefficient of variation. A 2-way (condition × % 1RM) repeated-measures analysis of variance with Holm-Sidak post hoc comparisons was implemented to identify any significant differences. Pearson product moment correlations were used to assess the relationship between each calculation method. The statistical analysis was conducted using SigmaStat 3.1 (Systat Software Inc., Richmond, CA), and an alpha level of p ≤ 0.001 was accepted as significant.


The mean heights thrown for each calculation method are outlined in Table 1. When the position of the barbell was used to define the end of the concentric phase, the height thrown was 0.078 to 0.131 and 0.040 to 0.142 m greater than that calculated with the switch and force data, respectively (p < 0.001). No significant differences were found between the heights thrown as estimated by the force and switch data, with the exception of the 15% 1RM condition; using the force plate reduced the height thrown by 0.046 m (p < 0.001). The coefficient of variation, describing within-subject intertrial reliability, was below 5%.

Table 1
Table 1:
Pearson correlations between the height thrown calculated with a switch, force plate, and position transducer (Pos).*

Strong correlations (r = 0.83-0.90) were found between all 3 calculation methods at loads of 15 and 30% 1RM (Table 1). However, as the magnitude of the resistance being thrown increased, correlations between the position estimation and that of the force plate and switch were reduced to 0.19 and 0.31, respectively.


Using a potentially variable displacement such as the position of the barbell at arms' length was found to be an inappropriate method of identifying the point of release, as the height thrown was overestimated at every load tested when compared with that calculated by the force and the switch (Figure 2). Throwing or attempting to throw a load necessitates the proper sequencing of the upper- and lower-body musculature (16) to ensure that maximum force is produced and kinetic energy transferred to the barbell at the point of release (7). Consequently, individuals may seek to increase the magnitude of their scapular protraction (9), thus resulting in a larger concentric displacement and a time of release greater than that estimated by using the position at arms' length.

Figure 2
Figure 2:
The calculated height thrown (m) using the force plate (□), position transducer (JOURNAL/jscr/04.02/00124278-201001000-00042/ENTITY_OV0071/v/2017-07-20T235412Z/r/image-png), and switch (▪). Data are expressed as a mean (SD) for each load. *Denotes significantly different than the force plate and switch (p < 0.05, and #denotes significantly different than the switch (p < 0.05).

The only significant difference found between the heights thrown as calculated by the force plate and the switch was at the 15% 1RM load in which case the switch identified the end of the concentric phase before the force being reduced to 0 N (Figure 2). There was however an observable trend across all loads-the percentage difference between the 2 calculation methods was increased at each subsequent load tested (Figure 2). This finding implies that the barbell had not been released when the vertical force, as measured by the force plate, was reduced to 0 N.

Preceding the point of release, any contact between the load and the hands should be reflected as a vertical force (by the force plate) and barbell acceleration greater than 0 N and −9.81 m·s−2, respectively. Whereas the moment contact is lost, gravity remains as the only force able to influence the vertical kinematics of the load. This concept highlights the theoretical reasoning for using the force plate and a measurement of 0 N to define the end of the concentric phase during ballistic upper-body movements. However, such a contention assumes that the vertical forces being displayed by the force plate are a direct consequence of the acceleration of the load without giving equal consideration to the contributions from the upper- and lower-body segments also involved. All mass is subject to Newton's first and second laws of motion pertaining to inertia and acceleration. Thus, when attempting to project a load or implement into free space, the sequential accelerations of each body segment contribute to the impulse being generated, although the pattern is not as obvious in certain movements (i.e., simple vs. complex) or populations (i.e., athlete vs. nonathlete) as it is in others (16). Furthermore, the acceleration of all contributing body segments may not be uniform through the entire concentric range of motion, as the magnitude of the segment mass decreases from proximal to distal. Consequently, the most distal segment, which is also the last point of contact with the load, requires less opposing force to decelerate, and therefore, the acceleration of the hands may be reduced to a greater degree than that of the larger masses of the upper arm or torso, resulting in a point of release before the vertical force being 0 N. Conversely, if an individual is attempting to throw a load that requires the production of a large impulse and more momentum to be transferred from the body segments involved, the point of release may occur subsequent to a vertical force of 0 N. However, short of taking anthropometric measurements and evaluating the kinematics of each segment involved, it would be difficult to identify the magnitude of each contribution and thus using a switch may be more appropriate.

Strong correlations were found between the force and the switch at each load tested and between all 3 methods at loads of 15 to 60% 1RM (Table 1); however, given the possible variability in the sequencing or acceleration of each body segment involved, it is suggested that the switch be regarded as a more valid method of measurement than either the position transducer or force plate alone.

Although the variable of interest in this investigation was the height thrown, all variables dependent upon the time of barbell release (end of concentric phase) will be affected by the validity of the analysis method used. Figure 3 highlights the importance of using an accurate means of identifying the end of the concentric phase; a substantial portion of the force curve is disregarded and not included in all subsequent calculations when the position was used, or, conversely, included when the concentric displacement was terminated at a force of 0 N. Using an inappropriate position to define the end of movement may result in significant overestimations of the means and an inaccurate interpretation of those variables presented as a function of the concentric duration or displacement. Defined by the position at arms' length, the time to peak velocity (duration of the acceleration phase) was reported by Newton et al. (14) to occur at 96% of the concentric displacement when using a 45% 1RM load. Although the methodology and subjects used were different, further findings from our laboratory, in which a switch was used, suggest that for the same relative load, only 82% of the concentric phase is spent accelerating (11). Such findings highlight the degree of error that can occur if the end of the concentric phase is not appropriately identified.

Figure 3
Figure 3:
Force, position, and switch data for a representative subject attempting to throw a 15% 1RM load. HF, HP, and HS correspond to the heights thrown as calculated with the force, position, and switch data, respectively. The JOURNAL/jscr/04.02/00124278-201001000-00042/ENTITY_OV0071/v/2017-07-20T235412Z/r/image-png and □ shaded areas represent the section of the force curve either not included or included if the position and force, respectively, were used to calculate the time of release. 1RM = 1 repetition maximum.

Practical Applications

Using a potentially variable position such as arms' length failed to locate the end of the concentric phase with the same accuracy as either the switch or the force plate. If the objective is to provide an accurate assessment of the means and/or temporal variables with respect to the duration or displacement of the concentric phase, then a force plate or pressure switch should be employed to provide greater precision, although a direct means of measurement is recommended. These findings also highlight the importance of using sound methodology when devising the scientific protocol for future experiments. The sport scientist must ensure that the conclusions being made are representative of the experiment and not the product of poor science.


1. Asci, A and Acikada, C. Power production among different sports with similar maximum strength. J Strength Cond Res 21: 10-16, 2007.
2. Baker, D. A series of studies on the training of high intensity muscle power in rugby league football players. J Strength Cond Res 15: 198-209, 2001.
3. Baker, D. Comparison of upper-body strength and power between professional and college-aged rugby league players. J Strength Cond Res 15: 30-35, 2001.
4. Baker, D. Differences in strength and power among junior-high, senior-high, college aged, and elite professional rugby league players. J Strength Cond Res 16: 581-585, 2002.
5. Baker, D, Nance, S, and Moore, M. The load that maximizes the average mechanical power output during explosive bench press throws in highly trained athletes. J Strength Cond Res 15: 20-24, 2001.
6. Cronin, JB, McNair, PJ, and Marshall, RN. Force-velocity analysis of strength-training techniques and load: Implications for training strategy and research. J Strength Cond Res 17: 148-155, 2003.
7. Cross, R. Physics of overarm throwing. Am J Phys 72: 305-312, 2004.
8. Doan, BK, Newton, RU, and Marsit, JL. Effects of increased eccentric loading on bench press 1RM. J Strength Cond Res 16: 9-13, 2002.
9. Duffey, MJ and Challis, JH. Fatigue effects on bar kinematics during the bench press. J Strength Cond Res 21: 556-560, 2007.
10. Elliott, BC, Wilson, GJ, and Kerr, GK. A biomechanical analysis of the sticking region in the bench press. Med Sci Sports Exerc 21: 450-462, 1989.
11. Frost, DM, Cronin, JB, and Newton, RU. A comparison of the kinematics, kinetics and muscle activity between pneumatic and free weight resistance. Eur J Appl Physiol 104:937-956, 2008.
12. Jidovtseff, B, Croisier, JL, Lhermerout, C, Serre, L, Sac, D, and Crielaard, JM. The concept of iso-inertial assessment: Reproducibility analysis and descriptive data. Isokinet Exerc Sci 14: 53-62, 2006.
13. Lawton, T, Cronin, J, Drinkwater, E, Lindsell, R, and Pyne, D. The effect of continuous repetition and intra-set rest training on bench press strength and power. J Sports Med Phys Fitness 44: 361-367, 2004.
14. Newton, RU, Kraemer, WJ, Hakkinen, K, Humphries, BJ, and Murphy, AJ. Kinematics, kinetics, and muscle activation during explosive upper body movements. J Appl Biomech 12: 31-43, 1996.
15. Newton, RU, Murphy, AJ, Humphries, BJ, Wilson, GJ, Kraemer, WJ, and Hakkinen, K. Influence of load and stretch shortening cycle on the kinematics, kinetics and muscle activation that occurs during explosive upper-body movements. Eur J Appl Physiol 75: 333-342, 1997.
16. Putnam, CA. Sequential motions of body segments in striking and throwing skills: Descriptions and explanations. J Biomech 26: 125-135, 1993.
17. Thomas, GA, Kraemer, WJ, Spiering, BA, Volek, JS, Anderson, JM, and Maresh, CM. Maximal power at different percentages of one repetition maximum: Influence of resistance and gender. J Strength Cond Res 21: 336-342, 2007.
18. Wilson, GJ, Newton, RU, Murphy, AJ, and Humphries, BJ. The optimal training load for the development of dynamic athletic performance. Med Sci Sports Exerc 25: 1279-1286, 1993.

displacement; force; power; switch

© 2010 National Strength and Conditioning Association