Journal Logo


Effect of Skiing Speed on Ski and Pole Forces in Cross-Country Skiing


Author Information
Medicine & Science in Sports & Exercise: June 2008 - Volume 40 - Issue 6 - p 1111-1116
doi: 10.1249/MSS.0b013e3181666a88
  • Free


Progression in cross-country skiing is dependent on both arm-poles and leg-ski forces. To measure these forces under natural skiing conditions is a challenge. Traditional research technique has usually concentrated on temporal and kinematic changes as recorded with film or video techniques (e.g., (2,6,8)). These reports have revealed considerable information on both the traditional diagonal cross-country skiing as well as the free skating techniques.

Although the kinematic approach is usually easy to apply, especially for a short, selected distance covering a few steps, the recording of actual causal forces for propulsion and gliding is more difficult. Consequently, the literature is more limited about the development of force measuring devices, which can simultaneously record both pole and ski forces during natural situations. Ekström (2) was among the first to measure diagonal skiing forces. He used a portable telemetry-force plate system mounted on skis. This was followed by several other portable force systems for skis mostly based on strain gauge or beam load cell technology (e.g., (7)). Portable force systems were followed by Komi's (4) long (6 m) force plate array made of four parallel rows of 6-m-long plates to record both vertical (Fz) and anterior-posterior (Fy) forces. This system gave simultaneous force records of both skis (legs) and poles (arms) and together with kinematics and EMG parameters was considered important in revealing the complex application of forces during classical diagonal technique. The major finding of these studies (3,4) was the sequence of preloading and thrust phases, which started with hip extensors, going then to the knee extensors and finally to the ankle extensors. The reports gave also information regarding actual force values of skis and poles as a function of skiing speed and grip waxing conditions. However, this information must be considered insufficient because the measurements were performed on one skier only.

Although this earlier approach (3) was informative for understanding the classical technique more thoroughly, the system had also other limitations. It was felt that 6 m is not long enough to follow reliably fast speed skiing when only two consecutive steps could be recorded. Consequently, Babiel (1) measured more consequent steps for the free technique by using binding mounted strain gauge plates for ski Fz and Fy but without simultaneous records of pole forces. The measuring technique described in the present report was meant to allow simultaneous recording of pole and ski forces for more consecutive steps on a 20-m-long force plate system installed in an indoor ski tunnel, where snow and temperature and humidity are kept constant. The report gives details of results measured on expert cross-country skiers at different constant velocities of diagonal skiing.


Eight top level young Finnish male skiers (age 17.2 ± 1.0 yr, height 178.5 ± 3.0 cm, body mass 71.6 ± 6.8 kg; means ± SD) participated after being informed of the procedures and all the risks associated with this study, which was part of their training and testing programs, planned by their individual coaches, and approved by the Vuokatti Sports Institute. The subjects gave their written informed consent to participate in this study.

After measuring the maximal ski speed (100%) in each subject (MAX, 5.6 ± 0.4 m s−1), the subjects were asked to ski at the three different submaximal target speeds (65%, 75%, and 90% of their MAX skiing speed; SLOW, 3.7 ± 0.4; MID, 4.2 ± 0.5; and FAST, 5.0 ± 0.5 m s−1, respectively) on a stable 100-m-long low uphill (2.5°) portion of a special indoor Ski Tunnel in Vuokatti Sports Institute (artificial snow, air temperature −4 ± 0.5°C, air humidity 84% ± 4%, and no wind; Fig. 1).

A, The force plates were placed in the special indoor ski tunnel at 2.5° slope. B, A schematic view of the force plate configuration.

A custom-made 20-m-long special force plate system from the Neuromuscular Research Center, University of Jyväskylä, consisted of four rows of parallel series of force plates. It was mounted at the end of this uphill portion of the ski tunnel. When the skier continued his constant speed skiing over the force plates, the mean speed was calculated by dividing the 20-m distance between the photo cells with the time measured by the photo cell system (Fig. 1B). The ski and pole forces (vertical (Fz) and horizontal (Fy) directions) on the right and left sides of the body were recorded separately and simultaneously along the force platform system. The four row platform system consisted (Fig. 1) of 1-m-long individual plates, connected electrically in series, row by row. The two middle rows (for skis) were covered with snow similarly to the actual ski track. The metal surfaces of the outer two rows for the poles were covered with glued tartan type material, the surface of which was on the same level as the ski tracks (the two middle rows). Thus, the measurement conditions also corresponded for the pole plant and action to those of the hard (artificial) snow conditions. The natural resonance frequencies of the individual plates were 150 Hz in both Fz and Fy directions. After the force plate system was installed in place, the recorded error of measurements was less than 0.5%.

In addition to the kinetic data, surface EMG activities from the erector spinae (ES), rectus abdominis (AB), vastus lateralis (VL), rectus femoris (RF), and medial gastrocnemius (MG) muscles of the right side were recorded and amplified by using the EMG telemetric recording system (TeleMyo 2400T, TeleMyo 2400R, Noraxon, USA). These data were stored simultaneously to the personal computer via an AD converter (Sampling rate 2 kHz; Power 1401, Cambridge Electronics Design Ltd., England). The recording surface of the EMG electrodes was 21 mm in diameter, and they were placed on the skin at a 2-cm interelectrode distance. The skin was shaved, abraded, and cleaned with alcohol to secure an interelectrode resistance value below 5 kΩ.

During skiing, the following five phases were defined for a diagonal skiing stride of the right leg (Fig. 2): 1) free gliding (from the end of the left ski Fz to the point of the initial left pole contact), 2) free glide and pole contact (glide and pole: from the point of the initial left pole contact to the point of the lowest right ski Fz before rapid Fz rising), 3) preloading for a kick (preloading: from the point of the lowest right ski Fz before rapid Fz rising to the point of the zero level of the right ski Fy), 4) kick (from the point of the zero level of the right ski Fy to the point of the take-off of the right ski), and 5) swing (from the take-off of the right ski to the next right ski contact) phases (Fig. 2).

mple of records for all force plate rows of the right and left ski and pole forces in both vertical (Fz) and horizontal (Fy) directions. The analyses of the individual curves were centered on the division of the different functional phases as shown above. Please note that the beginning of kick phase is defined as the zero force for the R Fy_ski. This follows immediately the "braking spike (an arrow point)," which was used to trigger the averaging of the several consecutive steps.

EMG signals were first full-wave rectified and then filtered (Butterworth fourth-order low-pass filter, 75 Hz). Thereafter, for obtaining the averaged curves of the different steps over the whole cycle, the force signals in the two steps and the filtered EMG signals during the skiing were synchronized using the point of the braking spike of the right ski Fy. This braking spike point is very characteristic in shape and is visible for all records (an arrow in Fig. 2). The averaged force and EMG curves were integrated and averaged separately for the defined phases of a diagonal skiing stride and for each muscle together with force signals.


Values are presented as means and SDs unless otherwise stated. To analyze the differences between different ski speed conditions for measured and calculated parameters, a repeated one-way ANOVA was used with a post hoc least significant difference multiple comparisons. The Spearman rank correlation coefficient for polynomial regression analysis of variables was used to calculate the statistical significance of the relationship between the skiing speed and the parameters. The probability level accepted for statistical significance was P < 0.05.


Cycle characteristics with skiing speed.

Table 1 shows the cycle time characteristics with the testing protocol. The step cycle time was defined as the period from the beginning of the right ski contact to the next right ski contact. As expected, the whole cycle time, the pole plant time, the ski contact, and the swing times became shorter with faster skiing speed. During ski contact phase, the glide and pole and kick phase times were shorter with faster skiing speed, but the free glide and preload phase times did not show significant differences with faster skiing speed.

Cycle characteristics and skiing speed (mean ± SD).

Ski forces.

The sum of averaged vertical ski forces and pole forces over a step cycle remained the same between different skiing speed conditions (661.2 ± 62.9, 661.5 ± 65.1, 645.1 ± 87.8, and 649.7 ± 66.9 N for SLOW, MID, FAST, and MAX, respectively). When the ski and the pole forces were averaged for the functional phases shown in Figure 2, the results shown in Figure 3 were obtained. In the free gliding phase, the ski Fz was highest in SLOW and was subsequently reduced in MAX. The similar reduction occurred in the glide and pole phase (Fig. 3, right panel). The reduction of the ski Fz from the free gliding to the glide and pole phase was 13% ± 6%, 13% ± 4%, 22% ± 3%, and 32% ± 2% (SLOW, MID, FAST, and MAX, respectively).

The averaged Fz ski forces at different skiing speeds (SLOW, MID, FAST, and MAX) for the free gliding phase (left) and the glide and pole phase (right). Significantly different between conditions at *P < 0.05 and **P < 0.01.

During the preloading phase, ski Fy was naturally negative because the right ski had to be stopped for the optimal kick (Fig. 2). The negative ski Fy values were similar in all skiing speeds (Fig. 4A). The same was true for the ski Fz that remained, on the average, around 640 N across all speeds. During the subsequent kick phase, both ski Fy and Fz increased significantly as a function of the skiing speed as shown in Figure 4B. In the MAX condition, the averaged ski Fy and Fz during the kick phase were 150 ± 41 and 1018 ± 245 N, respectively.

The averaged horizontal and vertical ski forces (Fy and Fz) for the preloading (A) and the kick (B) phases at different skiing speeds. Significantly different between conditions at *P < 0.05 and **P < 0.01. Please note that for this preloading phase, both horizontal and vertical ski forces remained the same across the skiing speeds. The kick phase showed increase in both horizontal and vertical ski forces as a function of the skiing speeds.

Pole forces.

Figure 5 shows the averaged pole forces in both horizontal and vertical directions for the entire period of the left pole plant (see the two lowest curves in Fig. 2). These forces were relatively low (7-8% body weight) and changed slightly between skiing speeds, although the speed comparison reached the significance level of P < 0.05 between MAX and SLOW only.

The averaged horizontal (left) and vertical (right) pole forces for the whole left pole contact. Significantly different between conditions at *P < 0.05.

Propulsive force contribution of ski to pole.

Figure 6 presents the ski and the pole force ratios as calculated in two different ways. Figure 6A is based on the average ski and pole forces during their respective contact times. Figure 6B, on the other hand, shows the ratios of the ski and the pole impulses obtained for the entire cycle time. Irrespective of the ski speed, the Fz ratios were nearly the same in the two comparisons. The Fy ratio, however, increased significantly as a function of increasing ski speed.

Ratios of ski/pole forces in horizontal (Fy) and vertical (Fz) directions at different skiing speeds. In panel A, the forces were obtained by averaging the ski and the pole forces for their respective contact times. The results in panel B are the comparisons of force impulse ratios (ski force/pole force) by calculating them for the entire cycle time. The cycle time was defined as the period between ski contact to the next ipsilateral ski contact. Significantly different between conditions at **P < 0.01.

EMG activities.

Figure 7 (left panel) presents a representative example of one skier showing how his measured five different muscles were activated at different functional phases. The right panel of Figure 7 shows the group average values for these phases and different skiing speeds. For the free gliding phase, the AB muscle showed a significant pattern of increasing activation with increase in speed of skiing. All the other measured muscles did not respond in this manner. This pattern of muscle activation was also similar in the subsequent glide and pole phase. However, both the preloading and the kick phases were then obviously important for the VL, RF, ES, and MG muscles, but AB showed also very high activity in the kick phase, and it continued to maintain the activity levels at increasing manner also in the ski leg swing phase.

(Left panel) Examples of the analyzed EMG records for all examined muscles. (Right panel) The averaged EMG values during activities of these muscles as measured for various functional phases and different ski speeds. Significantly different between conditions at *P < 0.05 and **P < 0.01.


The present study was meant to examine systematically the changes of kinetics together with the EMG activations at different skiing speeds. The most important findings were that with higher skiing speeds, 1) the average ski Fz of the gliding and glide and pole phases decreased together with concomitant increase in AB muscle activation, 2) the ski braking Fy and Fz during preloading phase did not show any difference despite higher ES and leg muscle activation, 3) the ski Fy and Fz forces of the kick phase increased, 4) the pole forces were low and did not change across conditions, and 5) the contribution of the pole forces to propulsion decreased.

Gliding relative to skiing speed.

Although previous studies (3,5) have focused on the propulsive pole and ski force phase, the glide phase can also be important for skiing. As reported by Norman et al. (6), the elite skiers consistently obtained longer glide per stride and greater stride length. The present results, that the average Fz of the gliding and glide and pole phases decreased with higher skiing speed, can imply lower ski friction during the glide and glide and pole phases, supporting the results of Norman et al. (6). Further, the greater reduction of the ski Fz from the glide to glide and pole phases and together with AB muscle activation with higher skiing speed conditions may suggest a rapid downward movement of the center of mass during the glide and pole phases. This rapid movement can be important to produce the braking force effectively during the subsequent preloading phase.

Braking and kicking relative to skiing speed.

It is expected that one of the important features in diagonal skiing would be to stop the ski effectively for preparation of an important kick phase. Our results demonstrated that the braking ski Fy and Fz during the preloading phase did not show any skiing speed dependent difference (Fig. 4), although the EMG activation in ES and leg muscles increased with higher skiing speed (Fig. 7). In general, these higher muscle activations continued during the kick phase, and the patterns suggest that, especially for ES and VL muscles, high activations during the preloading phase prepare the skier to produce greater force during the kick phase as well as to stop the ski. In the MG muscle, however, averaged EMG values did not increase during the kick phase with higher skiing speed. It is likely that the MG muscle functions importantly to stop the ski. It is unfortunate that the present study lacks EMG normalization. Consequently, intermuscular EMG activity comparison cannot be done. In addition, future studies should also include comprehensive kinematic analysis together with kinetics and EMG recordings.

Pole forces relative to skiing speed.

In the present study, the average horizontal and vertical pole forces were 6% and 7-8% body weight, respectively. These values are relatively smaller than the values previously reported (5,7,8). In addition, the pole forces did not change much with higher skiing speed, and the Fy ratio of the ski force to poling force increased with higher skiing speed (Fig. 6). These results imply that the proportion of the propulsive ski forces increases with faster skiing speed, supporting the suggestion of Millet et al. (5) that the pole forces may serve other purpose beside the generation of the propulsive forces.


As a first part of a series of studies on classical technique cross-country skiing, the present report described the use of a 20-m-long force platform system to explore the nature of bilateral vertical and horizontal forces produced by pole and ski actions at different skiing speeds. Systematic force production patterns were observed, and they will serve as basis for future studies that combine more comprehensive EMG recordings with simultaneous measurements of kinematic and kinetic parameters. Especially important for understanding the real mechanisms of cross-country skiing will be to perform the detailed kinematic analysis together with the present force recording system. Such future studies will also concentrate on varying the grip waxing and physical characteristics of skis and poles as well as skiing slopes.

The authors thank the following persons for their assistance during the various phases of the study: Ms. Puttonen P, Roivas S, Mr. Ruuskanen M, Piirainen J, Tukiainen J, Hytönen M, and Kaipio-Sarja T. This study was supported by the European Union and municipality of Sotkamo via grants EAKR 62015 and ESR 85549/from State Provincial office of Oulu.


1. Babiel S. Studies on intra-individual variability of selected cross-country skiing techniques. Eur J Sport Sci. 2003;3:1-10.
2. Ekstrom H. Force interplay in cross-country skiing. Scand J Sports Sci. 1981;3:69-76.
3. Komi PV, Norman RW. Preloading of the thrust phase in cross-country skiing. Int J Sports Med. 1987;8(Suppl. 1):48-54.
4. Komi PV. Force measurements during cross-country skiing. Int J Sport Biomech. 1987;3:370-81.
5. Millet GY, Hoffman MD, Candau RB, Clifford PS. Poling forces during roller skiing: effects of technique and speed. Med Sci Sports Exerc. 1998;30(11):1645-53.
6. Norman RW, Caldwell G, Komi PV. Differences in body segment energy utilization between world-class and recreational cross-country skiers. Int J Sport Biomech. 1985;1:253-62.
7. Piece JC, Pope MH, Renstrom P, Johnson RJ, Dufek J, Dillman C. Force measurement in cross-country skiing. Int J Sports Med. 1987;3:382-91.
8. Smith GA. Biomechanics of cross-country skiing. Sports Med. 1990;9:273-85.


©2008The American College of Sports Medicine