| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Veterinary Anatomy, Faculty of Veterinary Medicine, Utrecht University, PO Box 80158, 3508 TD Utrecht, The Netherlands
2 Department of Agrotechnology and Food Sciences, Wageningen University, PO Box 17, 6700 AA Wageningen, The Netherlands
3 Department of Farm Animal Health, Faculty of Veterinary Medicine, Utrecht University, PO Box 80151, and
4 Department of Equine Sciences, Faculty of Veterinary Medicine, Utrecht University, PO Box 80152, 3508 TD Utrecht, The Netherlands
5 Faculty of Civil Engineering and Geosciences, University of Technology, PO Box 5048, 2600 GA Delft, The Netherlands
Corresponding author: P. P. J. van der Tol; e-mail: Rik.vanderTol{at}wur.nl.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Key Words: cattle lameness • claw disorder • animal welfare • biomechanics
Abbreviation key: COF = coefficient of friction, GRF = ground reaction force, RCOF = required coefficient of friction
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Because insufficient friction leads to most slip incidents, several studies, mainly in humans, have determined the slip properties of floors. The main factor in relation to slipperiness is the coefficient of friction (COF). The COF determines the horizontal (frictional) force that can be generated between the contact surfaces of 2 objects per unit of vertical force between these objects (Hall, 1995; Chang et al., 2001). The amplitude of this force depends on the character of the mechanical and molecular interactions between the 2 surfaces in contact. The COF is obtained by measuring the dimensionless ratio of the horizontal and vertical force just before the objects start to slide relative to one another (static COF) or during sliding at a given speed (dynamic COF). When sliding (dynamic), the magnitude of the dynamic COF remains constant and theoretically below the value of the static COF (Hall, 1995).
In cattle locomotion, the COF depends on the properties of the claw horn, the concrete floor, and their contact interface. Intervening fluids (e.g., water, slurry, etc.) affect its value. The COF was found to range from 0.25 to 0.54 using different measuring methods; the median value on concrete was about 0.3 (Webb and Nilsson, 1983; Phillips et al., 1998, 2000; Phillips and Morris, 2000, 2001). A COF to allow locomotory behavior was hypothesized to be at least 0.4 (Webb and Nilsson, 1983). Those researchers, however, did not provide experimental evidence for their postulate. To assess the risk of slipping in the cattle housing, one must know the minimum required coefficient of friction (RCOF) for the displayed locomotory behavior.
The cow exerts a force to the floor during foot contact that can be resolved in a vertical and horizontal component. In our study, the ground reaction force (GRF), a force equal but opposite in direction to the force the cow applies to the floor, was measured by means of a force platform. Whether the claw will slip is determined by the ratio of the horizontal and vertical force (Figure 1
). On slippery floors, cows will try to adapt their gait to ensure that they do not slip, by keeping the RCOF for the intended behavior below COF values of the floor.
|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
The following behaviors, which tend to occur frequently and involve relatively high frictional forces, were selected to analyze the GRF: 1) walking a curve, 2) starting, 3) stopping, and 4) straight walking, which was used as a reference. In a pilot study, stepping into a free stall (height = 13 cm) was analyzed, but it was found that GRF values were in the order of the values obtained at straight walking. Therefore, this behavior was not investigated further.
A Kistler force plate (type Z4852, dimensions 600 x 900 mm; Kistler Instrumente AG, Winterthur, Switzerland) was embedded in a concrete pathway to measure the vertical and the 2 horizontal (longitudinal and transverse) components of the GRF with a sampling frequency of 2500 Hz. The pathway and force plate were covered with a 5- to 6-mm thick rubber mat to ensure enough traction to allow natural locomotion and to prevent slippage. Covering of the force plate also prevented potential awareness of the system. The data captured by the force plate was converted from analog to digital and stored in a computer for later analyses.
Experimental Procedures
Walking trials.
The cows were repeatedly walked down the path, led by 2 experienced handlers, one in front of and the other behind the cow to control the direction of locomotion. The experimenter initiated the GRF acquisition by having the software start to sample for up to several minutes to measure a few trials in a row. The forelimb and hind limb on the same side of the cow were measured in a single trial. The trials continued until at least 3 measurements for each limb of the 9 cows were recorded.
Walking a curve.
Portable fences were used to set up a path with a 90°corner, in such a way that the force plate lay exactly halfway through the curve in position at the place where the feet would touch the ground and the forces for making the left turn were exerted to the ground. The forelimb and hind limb on the same side of the cow were measured in a single trial. The trials continued for all 9 cows until at least 3 measurements for all 4 limbs on the left curves were recorded.
Stopping trials.
The same procedure was used as in the walking trials. Just in front of the force plate the leading handler unexpectedly blocked the pathway to force the cow to stop with either the fore or hind feet on the force plate. For each pair of fore or hind limbs the trials continued until 2 measurements were made.
Starting trials
The cow was positioned on the force plate. While it stood still, an electrical prod was used to initiate a fleeing reaction. This behavior was performed twice, once while the 2 forefeet were standing on the force plate and once while the 2 hind feet were standing on the force plate.
Data Processing
For all behaviors, the GRF was determined during foot contact with the force plate. The force plate recordings were low-pass filtered (cut-off frequency = 25 Hz) by using a digitally fourth-order recursive Butter-worth filter to discard noise-related data and to prevent a phase shift over time.
The step cycle of a limb during walking can be described as a weight-bearing phase (stance phase) and a nonweight-bearing phase (swing phase). The stance phase starts with the foot landing on the ground and ends at push off (Leach, 1993; Schamhardt, 1998). During the stance phase, the forces caused by locomotory behavior are exerted by the foot to the ground.
During straight walking and walking a curve, the GRF recordings were resampled to 100% stance time and normalized for BW (expressed in N/kg of BW) to allow comparison of the data within and between cows. The RCOF per sample point was determined as the ratio of the resultant GRFhorizontal and the GRFvertical. Data were analyzed further at 3 specific times during the stance phase (Table 1, 2; see arrows in Figure 2): 1) at the time the horizontal force reached maximal amplitude (RCOF-hor-max); 2) at the time the vertical force reached maximal amplitude (RCOF-vert-max); and 3) at the time the RCOF reached maximal amplitude (RCOF-max).
|
|
|
The data from the stopping and starting trials were not resampled in time. For these behaviors, the maximum RCOF was determined during foot contact with the force plate.
No statistical procedures were applied because, in the case of sliding events, the "average" cow will probably not slip or fall; consequently, the highest RCOF values are of interest.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
. In Figure 2a, an example of the time course of the vertical and resultant horizontal GRF of one cow is shown. The time characteristics of normal walking are very similar for the fore and hind limb.
Figure 2b shows a typical example of the magnitude of the RCOF during 2 successive steps (forelimb and hind limb) of the same cow. For the hind limb, a relatively high peak (± 0.5) is present at the beginning (establishment of ground contact); subsequently, the RCOF decreases from about 0.3 to almost 0 at midstance (at 50 to 60% stance time, indicating a vertical force only). This is followed by an increase, reaching a RCOF value of about 0.3 at push off.
After combining all RCOF data during walking, the maxima determined were 0.83 for the peak at heel strike, about 0.7 at push off, and 0.4 in between the previously described times during the rest of the stance phase.
Walking a Curve
An example of the RCOF during walking a curve is shown in Figure 3. Average RCOF at the 3 specific times are shown in Table 2. The vertical GRF during walking a curve is similar to the one for normal walking. The main difference is that there is an additional horizontal force for making the curve that is directed toward the center of curvature of the path (the centripetal force = the "center-seeking force"). Because of this, the RCOF never decreases below 0.15; the values at the 3 times show some variation between cows, possibly indicating interindividual differences in speed that could depend on the amount of caution displayed by each cow. For example (Figure 3b), the initial peak RCOF is about 0.5, and for both limbs, it reaches a steady state of 0.3 until an increase of 0.35 at push off. In all RCOF data for the walking a curve trials, the greatest values obtained were 0.4 for the "steady state," a maximum of 0.85 at the beginning, and 0.6 at the push-off phase.
|
and 5). However, most cows responded quickly to the blocking by the leading handler. Their walking behavior was already adapted within 2 trials in such a way that the walking speed decreased a few steps before they reached the force plate. In the successful "stopping" trials, the greatest RCOF was 0.8, recorded just after heel strike.
|
|
), the greatest RCOF reached was about 0.55, with a corresponding resultant GRFhorizontal of 1.5 N/kg of BW. This accelerative force was about twice as great as the maximum GRFhorizontal in walking (Figure 2a).
|
illustrates that at certain times during all behaviors the RCOF reached values close to and beyond the estimated COF that a concrete surface can generate.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
To understand the cow’s frictional demands from floor surfaces, regularly performed locomotory behavior of cows having potentially high risk of slipping must be obtained. Because of practical implications, some types of risky behavior, such as cows in estrus mounting one another, could not be simulated in a standardized laboratory environment. The behaviors selected on the basis of a pilot experiment were walking, walking a curve, stopping, and starting. This study compared the friction provided by a floor surface and the friction needed for unrestricted locomotory behavior.
The resulting horizontal force was first determined so that the RCOF could be calculated. Because an ideal flooring surface should provide enough traction in any direction, the direction of the necessary frictional force was not calculated. The maximum RCOF ranged from 0.3 to 0.85 for the various behaviors.
Cham and Redfurn (2002) found in human walking trials that ramp angles of 5 and 10 degrees increased the RCOF, respectively, by 60 to 70% and 125 to 135%, compared with a level walking surface. Powers et al. (2002) determined that walking at a faster speed also increased the RCOF. Hence, if the cow has to perform her locomotory behavior on ramp angles or at increased speed, the probability of slipping and falling will increase as the RCOF increases up to or beyond the maximum COF values (Hanson et al., 1999; Cham and Redfurn, 2002; Powers et al., 2002).
At heel strike and at push off, very high RCOF maxima are present. At these times, during quadrupedal walking at a comfortable speed, the majority of the BW is already supported by other limbs. The center of body mass is already projected within the base of support formed by these limbs. Because of this postural stability, the limb of interest (being measured on the force plate) will be able to exert large horizontal forces to the floor, and (micro) slips need not lead to falls. In theory, however, such a slip might cause a sudden rotation of the limb, forcing the projection of the center of body mass outside the base of support, and the probability of falling will increase.
Tables 1 and 2 illustrate that during the main part of the stance phase, a peak in the RCOF occurred when the resultant horizontal force reached its maximum amplitude. This peak can occur at 2 times during a 100% stance phase, either at about 15% when a maximum decelerative force is present or at about 85% when a maximum accelerative force is present. Although not as large as the maximum RCOF during heel strike or push off, the amplitude of the RCOF may reach values approaching the maximum static COF possible on concrete floors (0.3). Postural stability depends then on stable weight bearing at this stage of the stance phase of the supporting limbs, and a slip at that time may easily cause a fall.
When walking a curve, the RCOF remains large for nearly all of the entire stance phase. Although the RCOF for walking and walking a curve is generally within the range of the maximum COF possible, the maximum COF found during heel strike or push off are often greater than that value (Figure 7). Faster walking might lead to incidents of slipping and falling on slippery floors (Powers et al., 2002), as is observed often in the cow housing.
The RCOF measured when the cow is startled reached values of at least 0.4, but some cows reached an RCOF up to 0.7. The maximum RCOF determined in the sudden stopping and starting trials were greater than for normal walking and mostly exceeded the values of the COF concrete floor surfaces generate. Therefore, sudden stop and start behaviors requiring large accelerative forces may be considered to be dangerous locomotory tasks in current free-stall housing systems having (slatted) concrete floors.
COF Determination
The COF between claws and several concrete floors was tested (Phillips et al., 1998; 2000; Phillips and Morris, 2000, 2001; Telezhenko et al., 2004) by means of drag-sled testing of an artificial cow on concrete having various surface roughnesses (solid, slatted, and grooved concrete). The values ranged between 0.25 for a smooth concrete floor to 0.54 for a new rough concrete floor. A COF of about 0.3 was found most frequently. The relationship between claw conformation and the COF was tested with various results. Phillips et al. (1998) determined that the COF was correlated with the volume of the claws of cattle; the larger the volume, the greater the friction possible on a concrete surface. The largest claws also had rougher claw-contact surfaces, and this might have biased the outcome of these COF measurements. However, the same researcher failed to demonstrate a correlation with claw conformation in a later study (Phillips et al., 2000). Significant differences were detected between the COF of concrete and the claws of several breeds of cattle (Phillips et al., 2000). The (interindividual) differences between breeds were due to the chemical composition of the claw-horn keratin and were probably mostly affected by the difference in moisture content of the claw horn. Phillips et al. (1998, 2000) used slaughterhouse material in which the differences in conformation of breeds could be biased because of varying rehydration of the claws used in the COF measurements. However, assessment of frictional properties and abrasion resistance of dehydrated and moisturized claw horn samples showed that the water content did not change the COF on strips of abrasive sand paper. In contrast, water content did significantly decrease the resistance to abrasion (Bonser et al., 2003).
The static COF prevents the initiation of a slip, when the claw is already planted on the floor (or at push off at the end of the stance phase). The dynamic COF needs to be sufficiently large to control some sliding of the claw, once placed on the ground (e.g., first claw-floor contact at the beginning of the stance phase). Both types of COF can be measured with the aid of many devices, but they often produce poor correlations (Chang et al., 2001). Those researchers stated in their review that drag-sled testing is only valid for dry and clean surfaces, and dynamic measurements are necessary to estimate the potential risk of slipping on contaminated surfaces. The main argument against these types of friction-measuring devices is that the dynamics of locomotion are often not incorporated in the testing procedure (Chang et al., 2001). It might be concluded that the COF is difficult to determine, and COF values must be used prudently in relation to the dynamics of locomotion (e.g., RCOF). The results of our study have been evaluated using a COF value of 0.3, found most frequently (Phillips et al., 1998, 2000; Phillips and Morris, 2000, 2001; Telezhenko et al., 2004).
Gait Adaptations
Slips occur from an inability to adjust the locomotory behavior to the frictional properties of a floor—in other words, to keep the RCOF below the COF possible on a floor. Not all slips will lead to a fall and subsequent injury. However, one must keep in mind that the slips occurring also may lead to injury without falling, from striking an object, or from muscular strain (Hanson et al., 1999).
Sometimes the cow needs considerable amounts of friction. Most falls originate from the lack of a stable claw-floor contact at heel strike, which therefore is probably the most critical moment during locomotion. In humans, it was shown that at the beginning of foot-floor contact, (imperceptible) microslips occurred that could lead to a fall (Hanson et al., 1999). In cows, it has been shown that after heel contact, the claw slips 21 to 36 mm on bare soil or dry concrete and twice as far (46 to 71 mm) on slurry-covered concrete (Albutt and Dumelow, 1987).
The main adaptation to slippery conditions seems to be an issue of stability. The projection of the body’s center of mass moving forward is kept more closely within the base of foot support. This postural change in locomotion results in less movement amplitude of the proximal joints, reduced forward velocity, lower acceleration at the propulsive phase, and less deceleration at the breaking phase. The postural change results in lower frictional requirements (i.e., RCOF) and is reflected in the GRF recordings (Grönqvist et al., 2001) and the kinematical experiments using cows (Phillips and Morris, 2000, 2001). It has been shown that cows are capable of determining the slipperiness of floors (Phillips and Morris, 2002). Altered locomotory behavior by the cow demonstrates less confidence in its slippery environment. It is tempting to suggest that this adjustment in behavior might be detrimental to the welfare of the cow when it endures a prolonged (housing) period of time (Metz and Bracke, 2003).
The COF should be sufficiently large. This can be achieved by increasing the surface roughness (Phillips et al., 1998, 2000; Bonser et al., 2003). As a side effect, wear and tear of the claws could become too much. In addition, it is often stated that a rough concrete floor is worn smooth during a few seasons and becomes slippery again (Faull et al., 1996). Another way of providing additional friction is by draping and deformation of one of the interacting surfaces. These possibilities depend on the elasticity of the surface material. Concrete hardly deforms under loading; therefore, its deformation will not provide additional friction. Although it could be suggested that wet and, thus, more elastic horn has better draping and deformation possibilities, wet horn samples (softer claw) did not display better slipping resistance and were subject to increased wear rates (Bonser et al., 2003). Hence, a better option would enable the claw to sink into the floor to generate more traction, as it would on natural softer surfaces. Moreover, such floors could prevent mechanical overloading because of smaller compressive stresses exerted to the claws and improve claw health (Benz, 2002; Livesey et al., 2003; Somers et al., 2003; van der Tol et al., 2003).
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
Received for publication February 24, 2004. Accepted for publication October 2, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  O. Hernandez-Mendo, M. A. G. von Keyserlingk, D. M. Veira, and D. M. Weary Effects of Pasture on Lameness in Dairy Cows J Dairy Sci, March 1, 2007; 90(3): 1209 - 1214. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Rushen and A. M. de Passille Effects of roughness and compressibility of flooring on cow locomotion. J Dairy Sci, August 1, 2006; 89(8): 2965 - 2972. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. B. Tucker, D. M. Weary, A. M. de Passille, B. Campbell, and J. Rushen Flooring in front of the feed bunk affects feeding behavior and use of freestalls by dairy cows. J Dairy Sci, June 1, 2006; 89(6): 2065 - 2071. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
= GRFhorizontal hind limb, 
= GRFvertical fore limb, and 
= GRFhorizontal fore limb) of a single stance phase during straight walking and b) the corresponding calculated required coefficient of friction (
= RCOF hind limb; 
= RCOF fore limb). The GRF are expressed in N/kg of BW (= 630 kg). The RCOF is dimensionless. The time when the claw contacts the flooring surface is expressed as a percentage of the total stance time. In addition, in this example, the 3 arrows indicate the 3 time points of interest of the hind limb during walking. In similar way, for all other limbs and trials, the results for straight walking are shown in Table 1
