Athletic shoes must provide at least enough traction to maximize performance and minimize slipping, but too much traction can potentially increase the risk of injury. Research suggests that traction on modern artificial turf can vary depending on cutting angle.
by Michelle Sabick, PhD; Benjamin Cooper, MS; Seth Kuhlman, MS; and Ronald Pfeiffer, EdD, LAT, ATC
The interaction between an athletic shoe and the surface on which it functions is complex, and the demands on the shoe-turf interface are often conflicting. In terms of performance, a high degree of traction is desired to allow the athlete to maximally accelerate, decelerate, and change direction.1, 2 In terms of injury prevention, however, traction should be low enough that loads transmitted to the lower extremity do not exceed the safe limits of the musculoskeletal system.2-4 Both epidemiological and biomechanical research suggest that higher traction is positively related to increased injury frequency.4, 5 Therefore, the most appropriate traction conditions depend on the requirements of the specific application.
Athletic surfaces appear to have an influence on injury rates. For example, Powell and Schootman used the National Football League (NFL) injury surveillance system to analyze relative risk of injury on AstroTurf and natural grass during 10 NFL seasons.6 They concluded that there was a tendency for AstroTurf to be associated with increased risk for certain types of knee injuries under very specific conditions. Although certain types of lower extremity injuries were sometimes associated with AstroTurf in the 1980s and 1990s, there are fewer studies evaluating the newer generation of rubber infill artificial turf surfaces. Older artificial turf surfaces consisted of relatively short vertically-oriented fibers of nylon (or similar materials) with a woven backing and were placed over a concrete surface. Newer rubber infill artificial turf surfaces are made of much longer nylon, polypropylene or polyethylene fibers that are partially in-filled with sand and/or rubber particles and more closely mimic the characteristics of natural turf surfaces. Meyers et al recently reported that the rates of certain types of injuries were significantly higher for high school football games played on rubber infill FieldTurf than on natural turf surfaces, although ligament injuries and injuries that resulted in three or more weeks of lost playing time were more common on natural turf.7 A biomechanical study by Shorten et al concluded that infill synthetic turf surfaces seemed to be close to natural grass in terms of limiting rotational traction but providing adequate translational traction.4
The term traction is often used to describe the interactions between shoes and surfaces. The traction coefficient is the ratio between traction force (Ff in Figure 1) and normal force (vertical ground reaction force, N) at a surface and can vary with time, normal force, contact area, surface grain direction, and velocity of the shoe relative to the surface.4, 8 In addition to variables describing the shoe-turf interface, traction coefficients can also vary with environmental variables such as temperature, humidity, and wear.9 Therefore, when measuring traction between a shoe and surface, the conditions tested should be similar to those observed in actual conditions to obtain relevant data. These complications make it difficult to measure traction in ways that are meaningful, repeatable, and appropriate.10
Traction coefficient can also be determined in a number of different scenarios. Static and dynamic traction coefficients represent subtle differences in shoe-surface interaction, depending on the relative movement velocity between the shoe and surface. Static traction coefficients represent the resistance to impending motion between the shoe and the surface, while dynamic traction coefficients represent the resistance to sliding or pivoting motions once they have already begun. Therefore, static traction coefficients can be interpreted as quantifying conditions in which the foot is fixed, and dynamic traction coefficients represent situations in which the foot is sliding along the turf.
The traction characteristics between a shoe and a surface can also be quantified for either rotational or linear motions. Many studies have investigated rotational traction, based on the assumption that most knee and ankle injuries occur from an excessive rotation of the upper body with the foot planted.1, 9, 11-13 Rotational tests generally consist of the front portion of a shoe being pressed into the turf surface and rotated about a vertical shaft.
Traction and Injury
Of course, many injuries can also be caused by the inability of the foot to slide or translate, as in the case of an inversion ankle sprain. In addition, since translational traction is a major component of athlete performance, it must be part of the overall equation defining the tradeoff between performance and injury prevention. Linear tests characterize the traction between a shoe and a surface when the shoe is slid across the turf while being pushed into the turf with a known force. These tests best represent the traction during rapid starts, stops, and changes of direction. There are relatively few of these studies in the literature,4, 14 especially on newer brands of turf surfaces, likely due to the technical difficulties in performing the tests in a repeatable way. More linear traction tests need to be conducted with realistic loads involving ground reaction forces equivalent to at least one body weight,14 especially on newer artificial turf surfaces.
The linear traction tests that have been performed to date generally simulate stopping motions. Although rapid stops have been implicated as a mechanism of ACL injury,15, 16 cutting motions have also been implicated in both inversion ankle sprains and ACL injuries,15-18 and are extremely important to athlete performance in sports such as football, soccer, and basketball. To date, no studies have investigated the effect of cutting direction on the traction developed at the shoe-turf interface on artificial turf surfaces. We conducted a study to investigate whether traction characteristics vary based on cutting angle in a representative sample of cleated athletic shoes on FieldTurf.
Shoes: Four pairs of U.S. men’s size 12 cleated football shoes were tested (Figure 2). The shoes selected spanned the spectrum of cleated shoe styles used by “skill” players in youth, collegiate, and professional American football. “Skill” players are responsible for advancing the ball from the line of scrimmage. Two of the shoes tested had molded stationary cleats (one with 9 cleats and one with 13 cleats), while one shoe had 7 detachable cleats and one shoe had a combination of 8 small molded and 7 detachable cleats. Such shoes are commonly worn by football players on both natural and newer artificial turf surfaces.
Artificial turf: A single installation of FieldTurf (FieldTurf Tarkett, Peachtree City, GA) brand synthetic turf was the testing surface. The testing zone on the turf surface was located between the 50-yard line and end zone, and roughly equidistant from either sideline on the playing surface of an American football field. All data were collected on the same day in a collegiate indoor football practice facility.
Instruments and apparatus: This study utilized a computerized testing device, the Boise State TurfBuster, which simulates the motion of a foot decelerating across the artificial turf surface (Figure 3) and can be fitted with a shoe. The normal force is applied via a pneumatic cylinder. The entire shoe and ankle shaft assembly is mounted to a cradle that moves horizontally through low friction bearings. The motion is controlled using a pneumatic actuator connected to the ankle shaft just above the ankle joint. Linear speed and motion are measured by a linear transducer attached to the actuator. All force and position data were collected at 250 Hz. Temperature and humidity of the turf’s surface were also collected using a hand-held thermometer and hygrometer.
Test conditions: To simulate differences in cutting angle, the shoes were tested in four different orientations relative to the direction of motion (0°, 30°, 60°, and 90°) (Figure 4). The cutting angle was set using a pin system located in the inner frame of the device. The shoes were oriented horizontally so that all cleats could engage with the surface and a vertical load of 900 N was applied through the ankle. The vertical load is approximately equivalent to one body weight for an average collegiate skill player. The shoe underwent 20 cm of translation at a rate of 10 cm/s. After each trial, the device was lifted and moved approximately two feet so that each test was performed on a fresh patch of turf. The device was also secured to an immovable object to prevent any relative movement between the turf surface and the device.
Data analysis: The following three pieces of data were collected for each test:
Peak Traction Coefficient– The peak value of the ratio of horizontal to vertical forces during the test. This value represents the greatest traction coefficient experienced during the shoe’s motion (Figure 5).
Dynamic Traction Coefficient– The ratio of horizontal to vertical force during the final 2 cm of motion when the velocity is constant. This value represents the traction coefficient between the turf and shoe when the shoe is moving at a constant rate relative to the turf (Figure 5).
Peak Resistive Torque – The peak value of torque resisted by the ankle while the shoe moves forward. This value represents the shoe’s tendency to rotate internally or externally while sliding forward.
Each of the variables was averaged over all the trials for that test condition. The independent variables were the shoe and cutting angle. The dependent variables were the peak traction coefficient, the dynamic traction coefficient, and the peak resistive torque. Repeated measures univariate analysis of variance (ANOVAs) were used to compare the means of the dependent variables among cutting angles and among shoes. Holm’s Sequential Selective Bonferroni Method post-hoc tests were performed to compare means in the case of a significant ANOVA.
Results and discussion
Peak traction values were very high, ranging from 1.7 ± 0.2 for the Super Speed shoe at 0° to 5.3 ± 0.5 for the Scorch shoe at 60°. Significant differences in peak traction coefficient were found among the cutting angles (p=0.000), but not among the shoes (p>0.05). Peak traction coefficient values increased significantly from 0º to 30º and again from 30º to 60º (Figure 6). Although the peak traction coefficient values tended to decrease between the 60º and 90º angles, the differences were not significant. Very similar trends were seen for the dynamic traction coefficient data. Peak resistive torque also varied significantly with angle, increasing in magnitude from 0° to 30° and from 30° to 60° for all four shoes.
Traction coefficients ranged from 0.54 to 1.45 in the same study. These values are much lower than those reported in the current study. However, the traction coefficients collected by Shorten et al were obtained with a load of only 529 N and at a higher relative velocity than in the current study. The authors also averaged the traction coefficient over the “initial period of shoe motion”, which would result in values much lower than the peak values reported here (see Figure 6). Because of the complicated nature of the shoe-surface interaction, traction values recorded in different testing conditions often cannot be directly compared. Our values of dynamic traction coefficient were approximately 1.2 for all four shoes for the 0° condition, very similar to those reported
by Shorten et al4 and McNitt et al.14
Implications for Athletes
The high traction coefficients we report suggest that, at least for the conditions under which we tested, athletes have plenty of traction available to perform stopping, starting, and cutting motions effectively. However, the traction coefficients approximately doubled from the 0° position to the 60° and 90° positions, resulting in traction coefficient values much higher than required for performance.4 These results suggest that athletes are at increased risk of injuries while performing 60° and 90° cuts. It might be possible to redesign shoes so that traction coefficients do not increase so dramatically with cutting angle, to eliminate this risk.
Since the number of cleats and the cleat surface area do not change with shoe orientation, there is no obvious reason for traction coefficients to change with the simulated cutting angle. One explanation for the pattern we saw could be the alignment of the cleats relative to direction of the applied motion. When the shoes are oriented in the 0° position, the cleats generally align such that the studs create a small number of independent channels in the infill material. As trailing studs pass through the previous studs’ path in the infill, they experience reduced resistance compared to the leading studs (Figure 7). When the shoe orientation is changed from 0°, the number of independent paths generally increases. Therefore, as the shoe rotates from 0º to 60º, more studs are exposed and create unique paths in the turf material, increasing the resistance to linear translation. We have termed this hypothesis the “trench effect” and have demonstrated in our laboratory that it can occur. Based on these results, cleat locations could be adjusted by shoe manufacturers to optimize cutting performance or decrease injury likelihood on a given turf surface.
One variable that has been underreported in the literature is the resistive torque generated during linear translation. Essentially, this is the tendency of the shoe to “fishtail” in either an internal rotation or external rotation direction while sliding. This phenomenon was described by Andréasson, who felt that a “balanced shoe,” which did not tend to rotate while sliding, might help prevent knee and ankle injuries.18 The take-home message of these data is that even when a shoe is not stationary on the turf, it could contribute to rotating the foot, which has implications for lower extremity injuries. Depending on the cleat pattern and the shoe orientation relative to the direction of sliding, a shoe could tend to rotate the lower extremity either internally or externally.
The shoe-surface interaction on artificial turf surfaces is extremely complex and difficult to fully characterize. Cutting direction had a much stronger effect on traction coefficient than did shoe model in this study. The data presented here demonstrate the effect of rotating the shoe, which simulates different cutting directions. Peak traction coefficients for all four shoes tested occurred when shoes were oriented at 60° relative to the direction of motion. Since the number and area of cleats engaged is not affected by orientation, other factors must govern the variation in traction with shoe orientation. We propose that a “trench effect” occurs, whereby traction is influenced by the number of independent rows of cleats aligned with the movement direction.
Our data suggest that, all else being equal, performance in cutting should be optimal at angles around 60°, since the peak traction coefficient values in all four shoes occurred at this angle. Of course, athletic performance is based on more than just shoe-surface interaction. Evidence suggests that movement patterns of athletes are influenced by traction characteristics,2, 5, 19 so mechanical testing of traction is not adequate for predicting athlete performance. The ability to perform optimal cutting motions is likely dependent on other factors influencing athlete kinematics and muscle activation patterns. Therefore, more work is needed to relate shoe-surface properties to athlete performance directly.
Michelle Sabick, PhD, is an associate professor in the department of mechanical and biomedical engineering at Boise State University and co-director of the BSU Center for Orthopaedic & Biomechanics Research. Benjamin Cooper, MS, is an exercise specialist at the Shepherd Center in Atlanta, GA, and formerly a graduate student in kinesiology and a research assistant in the Center for Orthopaedic & Biomechanics Research at Boise State University. Seth Kuhlman, MS, is a research associate in the department of mechanical and biomedical engineering at Boise State University and lab manager for the BSU Center for Orthopaedic & Biomechanics Research. Ronald Pfeiffer, EdD, LAT, ATC, is a professor and chair of the department of kinesiology at Boise State University and co-director of the BSU Center for Orthopaedic & Biomechanics Research.
The authors would like to acknowledge Intermountain Orthopaedics and the National Football League for partial financial support for this work.
1. Cawley PW, Heidt RS Jr, Scranton PE Jr, et al. Physiologic axial load, frictional resistance, and the football shoe-surface interface. Foot Ankle Int 2003;24(7):551-556.
2. Dixon SJ, Batt ME, Collop AC. Artificial playing surfaces research: a review of medical, engineering and biomechanical aspects. Int J Sports Med 1999;20(4):209-218.
3. Milburn PD, Barry EB. Shoe-surface interaction and the reduction of injury in rugby union. Sports Med 1998;25(5):319-327.
4. Shorten M, Hudson B, Himmelsbach J. Shoe-surface traction of conventional and in-filled synthetic turf football surfaces. in XIX International Congress on Biomechanics. 2003. University of Otago, Dunedin, New Zealand.
5. Nigg BM, Segesser B. The influence of playing surfaces on the load on the locomotor system and on football and tennis injuries. Sports Med 1988;5(6):375-385.
6. Powell JW, Schootman M. A multivariate risk analysis of selected playing surfaces in the National Football League: 1980 to 1989. An epidemiologic study of knee injuries. Am J Sports Med 1992;20(6):686-694.
7. Meyers MC, Barnhill BS. Incidence, causes, and severity of high school football injuries on FieldTurf versus natural grass: a 5-year prospective study. Am J Sports Med 2004;32(7):1626-1638.
8. Bowers KD Jr, Martin RB. Cleat-surface friction on new and old AstroTurf. Med Sci Sports 1975;7(2):132-135.
9. Torg JS, Stilwell G, Rogers K. The effect of ambient temperature on the shoe-surface interface release coefficient. Am J Sports Med 1996;24(1):79-82.
10. Frederick EC. Biomechanical consequences of sport shoe design. Exerc Sport Sci Rev 1986;14:375-400.
11. Villwock MR, Meyer EG, Powell JW, et al. Football playing surface and shoe design affect rotational traction. Am J Sports Med 2009;37(3):518-525.
12. Villwock MR, Meyer EG, Powell JW, et al. The effects of various infills, fibre structures, and shoe designs on generating rotational traction on artificial turf surfaces. Proc Inst Mech Eng P 2009;223(1):11-19.
13. Livesay GA, Reda DR, Nauman EA. Peak torque and rotational stiffness developed at the shoe-surface interface: the effect of shoe type and playing surface. Am J Sports Med 2006;34(3):415-422.
14. McNitt AS, Middour RO, Waddington DV. Development and evaluation of a method to measure traction on turfgrass surfaces. J Test Eval 1997;25(1):99-107.
15. Boden BP, Dean GS, Feagin JA Jr, Garrett WE Jr. Mechanisms of anterior cruciate ligament injury. Orthopedics 2000;23(6):573-578.
16. Cerulli G, Benoit DL, Lamontagne M, et al. In vivo anterior cruciate ligament strain behaviour during a rapid deceleration movement: case report. Knee Surg Sports Traumatol Arthrosc 2003;11(5):307-311.
17. Garrick JG. The frequency of injury, mechanism of injury, and epidemiology of ankle sprains. Am J Sports Med 1977;5(6):241-242.
18. Andreasson G, Lindenberger U, Renstrom P, Peterson L. Torque developed at simulated sliding between sport shoes and an artificial turf. Am J Sports Med 1986;14(3):225-230.
19. van Gheluwe B, Deporte E, Friction measurement in tennis on the field and in the laboratory. Int J Sport Biomech 1992;8(1):48-61.