Traction performance modeling of worn footwear with perpendicular treads

ABSTRACT The traction performance of the footwear deteriorates due to outsole wear which further increases the risk of slip and fall related accidents. To date, several studies have tested footwear tractions across several slippery conditions but only a few studies have attempted to assess their performance considering worn shoes. In this work, nine outsoles, with systematically modified tread geometries, were investigated, to study the effects of tread patterns in new and worn conditions on traction, across common slippery conditions. The outsoles were progressively worn in three wear cycles. Outsoles with increased worn regions generated lower friction and higher fluid pressures, indicating increased slipping risks. Also, diversion of fluid flow due to large worn regions produced high fluid accumulations at other locations over the outsoles. The methods and results are anticipated to help footwear manufacturers with the strategic design of tread patterns that can provide improved friction even when completely worn. GRAPHICAL ABSTRACT


Introduction
Nonfatal accidents such as, unintentional slips, trips, and falls are among the leading causes of hazards in the workplace, recreational, industrial, and during regular manoeuvres [1].Several lower limb problems such as, dislocations, ligament tears, and muscle inflammation are common consequences of slip and fall incidents [2].In the US, these issues have led the employees to avail emergency medical leaves, leading to a significant delay in work and an overall economic burden of over $170 billion [3,4].Reportedly, the primary reason for initiation of unintentional slips is related to an abrupt reduction in the traction at the shoe-floor interface [5,6].Therefore, it becomes essential to study the traction performance of footwears and other relevant characteristics to maintain adequate shoe-floor friction.
Several footwear related features such as, tread shape, type of flooring, outsole materials, shore hardness, and the presence of slippery contaminants are known to affect the available coefficient of friction (ACOF) at the shoe-floor contact [7][8][9][10][11][12].In particular, the presence of viscous pollutants on a floor substantially decreases the ACOF, which increases the risk of slipping [13].Also, the geometrical features of an outsole (i.e.treads) are crucial in comprehending its performance in dry and wet slipping circumstances [14,15].In a recent study by Yamaguchi et al. [16], the quality of an outsole to allow the fluid flow through the tread patterns during fluid-contaminated skidding was directly found to correlate with an increase in the ACOF.Additionally, prior researches [17][18][19] have linked elevated slipping risks with excessive fluid pressure and the development of a hydrodynamic liquid layer at the interface of the shoe and floor.In a study by Blanchette and Powers [20], tread groove dimensions and orientations were parametrically varied and found to significantly affect the slip-resistance of a footwear.In another study by Hale et al. [21], a tread element of a tennis shoe's outsole was considered to investigate the effect of tennis steps on friction and wear.It was reported that longer treads (oriented towards the sliding motion) showed reduced dynamic friction.It was also found that friction was influenced by the outsole tread's sliding history and the wearing behaviour.In another study by Goff et al. [22], Custom rubber samples with varying numbers of holes were used in trials to mimic the dimples found in tennis shoe treads.The static friction coefficient got significantly reduced as the ratio of holes to solid rubber increased until a critical ratio was reached.
Slipping hazards can be quantified with equipment such as tribometers or slip testers, whose designs and operating conditions vary widely [23].As the quantification of traction performance of footwear can be difficult, tedious, and time-consuming task, a few studies [8,24] have attempted to computationally model the contact mechanics of the interface.However, computationally, the performance of tread patterns during slipping in the fluid contaminant conditions have not been extensively explored.Hence, rapid testing and understanding the effect of varying treads on the friction on fluid contaminated flooring is essential.
The effectiveness of shoes to retain friction due to tread characteristics deteriorates with wear, and has a major impact on the ACOF at the shoe-floor interface [25,26].Worn shoes were found to increase the slipping risks upto 10 times than regular new footwear [26].Hemler et al. [27] analysed the impact of natural shoe wear on the frictional of slip-resistant shoes and studied the relationship between wear metrics and frictional performance.Worn footwear were reported to experience a reduction in ACOF of more than 25% when tested on different slipping conditions [11].Hemler et al. [17] investigated the effectiveness of worn slip-resistant shoes in fluid-contaminated environments and reported that increasing fluid pressures over the untreaded or worn region resulted in decreased traction.Only a few footwear were able to surpass the ACOF criterion of 0.3, which has been observed to significantly reduce slips [9].Hence, understanding the impact of wear to evaluate the traction performance of shoes will be beneficial in determining the replacement thresholds for these shoes and understanding the increasing risk of slipping.
In this work, traction performance of outsoles having horizontally oriented treads (i.e.perpendicular to the slipping direction) was extensively studied.In a previous study by Blanchette and Powers [20], three types of tread patterns i.e. parallel, oblique, and perpendicular were investigated for their frictional performance.To further analyse the effect of tread dimensions on the frictional outcome, the perpendicular (or horizontal) tread pattern was varied across its width and depth.Three different types of tread width (i.e. 3, 6, and 9 mm) and gaps (i.e. 2, 4, 6 mm) were considered.It was further reported that horizontal treads with 6 mm width showed high frictional outcome as compared to other pattern variations.In another previous study by Yamaguchi et al. [16], horizontal treads (with varying depths) were investigated for its traction performance in glycerol contaminated conditions.The considered rubber block had a tread width and gap of 3 and 2 mm respectively.It was found that treads with reduced depth retained their contact area, and thus resulted in low deformation hence, high frictional outcomes.In line with these studies, the current work considered an extensive investigation based on the horizontal tread patterns varied with several widths and gaps.
The developed outsoles were artificially worn in three progressive cycles and slip tested across dry and water contaminated flooring by employing a biofidelic slip tester.Furthermore, each new and worn footwear outsoles were computationally analysed by predicting the fluid pressures across the outsoles during wet slipping using a computational fluid dynamics (CFD) framework.The cost-effective procedures and the outcomes of this study are anticipated to help understand the traction behaviour of varying treads in new and worn condition.It will also help to determine footwear replacement thresholds due to its wearing.

Design and manufacturing of footwear outsoles
The geometrical design of the outsoles included in this work consisted of horizontally oriented treads (i.e.perpendicular to the slipping direction) based on the impressions of an original footwear.The selected footwear was a common Oxford style shoe (Model: FST KI-106 BLACK-42, Fausto, Uttar Pradesh, India).As formal shoes are majorly worn across the workplaces, this type of footwear was specifically considered in this work.Specifically, the Oxford style shoe was selected because it is one of the highly common formal shoes and have a consistent perpendicular tread design.The high availability of perpendicular patterns could be attributed to its ease in the development of the sole moulds.Footwear characteristics such as, tread dimensions and shore hardness were measured using a shore A durometer (Kern & Sohn GmbH, Germany) and digital depth gauge (Precision Instruments, India) respectively.These features were measured within the 50 mm linear range from the posterior point over the heel.Reportedly, the heel portion measuring 50 mm from the posterior point is the primary contact region during unintentional slips [5][6][7]9,28,29].The outsole of the original footwear had a tread width of 2 mm with 2 mm intervals (i.e.gaps).The outsole material was identified as polyurethane and had a shore A hardness of 60.The outsole geometry was modelled in a 3D CAD software (Fusion 360, Autodesk, US).To quantify the effect tread designs on the ACOF and fluid pressures, the tread dimensions were parametrically altered across its width (w) by 2 mm and gap (g) by 1 mm.Whereas, a constant depth of 2 mm was fixed for all the outsoles.A total of nine outsoles were generated as depicted in Table 1.
The CAD models of all the outsoles were converted into a positive mould to fabricate each outsole.The moulds were 3D printed with acrylonitrile butadiene styrene (ABS) using an enclosed 3D printer (CR-10, Shenzhen Creality 3D Technology, China).The developed positive moulds were then poured with liquid silicone material and left to dry for 2.5 h.Further, the fabricated negative silicone moulds were then filled with polyurethane (procured from Aditya Silicone, New Delhi, India) having hardness of shore A 60 to resemble the material properties of the considered original footwear.
To ensure the uniformity amongst the outsoles, the filled moulds were kept in an enclosed chamber to restrict the entry of any foreign pollutants.The moulds were left to cure for 8 h and after the curation was completed, the outsoles were removed from their respective moulds.Three samples of each pattern, resulting in 27 samples, were developed to check for repeatable manufacturing.The manufactured outsoles were measured for their dimensional and material replication accuracy.The criteria for dimensional tolerance (i.e. for tread width and gap) used was ±0.1 mm. Figure S1 shows the manufactured outsoles.

Wearing of outsoles
Progressive wearing of outsoles was performed to understand the effect of worn regions on the slip variables (i.e.ACOF and fluid pressure).Artificial wearing protocol was applied, which shortened the entire observation period for determining the treads' lifetime performance as suggested by Chang et al. [30].Outsoles were worn in three wear cycles namely, first, second, and third wear cycle which mimics the outsole topography of treads worn after 3 months, 6 months, and fully worn upto the outsole base material.The worn cycles were controlled based on a previous study by Hemler et al. [27] which simulated the wearing time based on the distance of abrasive belt grinder.Wearing of the outsoles were performed by attaching the developed outsoles beneath the footwear, placing the footwear at an angle of 17 ± 1°over the belt grinder (50 Grit, 3M Industries), and measuring the distances.The 17 ± 1°angle was based on previous studies [17,26] which considered this metric to wear the footwear.The outsoles were worn, worn burrs of outsole material were cleaned, and experimentally slip tested.Similar procedure was followed for the subsequent wear cycles.Figure S2 represents the worn outsoles after first, second, and third wear cycles whereas Figure 1 shows the consolidated images of an outsole (i.e.H6) after subsequent wear cycles.

Mechanical slip testing experiments
A whole-shoe biofidelic and portable mechanical slip testing device was employed to estimate the traction performance of the outsoles.The outsoles were attached to the bottom of the shoe and further placed to the slip testing device.The device running and testing was based on the ASTM F2913-19 guidelines [31] which represents the testing framework to assess the shoe-floor friction mimicking actual human slips.The detailed design and development of the implemented device is presented in a recent work by Gupta et al. [32,33].The device was found to be biofidelic and repeatable [34].For repeatability, the variations in the ACOF of one shoe was tested 10 times with different flooring-contaminant combinations.Additionally, the device was validated for its ability to differentiate shoes tested on similar slippery conditions, different floorings, and different contaminants.
During slipping, a sliding speed of 0.5 m/s, normal force of 250 ± 25 N along with the slipping angle of 17 ± 2.5°has been observed [35,36].Hence, an upper bound normal load of 275 N, slipping speed of 0.5 m/s, and sliding angle of 17 ± 2.5°was selected as the operating parameters of the slip testing device.The outsoles in new and worn conditions were tested across dry and water contaminated conditions on a common ceramic flooring.To mimic the slippery contaminated condition, 50 ml of water was spilled over the testing area.A total of 5 trials across each slipping condition was performed and the resulting ACOF was averaged.The surface roughness (R a ) of the flooring was measured as 38.45 µm using a digital surface profilometer (Precision Instruments, India).The R a was measured across 5 locations inside the area covered by the slip testing device and was averaged.Figure 2 shows the slip testing device used in this study.

CFD modelling for the prediction of fluid pressures
The prediction of fluid pressures during slipping simulation in the presence of water as a contaminant was done computationally by employing a software capable of solving fluid flows (Ansys Inc., US).The CAD models of the outsoles in new condition were directly imported to the software whereas, the estimation of the fluid pressures across worn outsoles were done by importing 3D scanned models using a 3D scanner (Intel RealSense, Intel, US).Scanned models of the worn outsole were post-processed for their boundaries and worn region refinement using a mesh editing software (Meshmixer, Autodesk, US).
Here, W is the strain energy potential function, µ is the initial shear modulus (Table 2), I is the strain invariant where, I is calculated by the software through deformation gradients.Up to the elastic limit, Young's modulus and Poisson's ratio were modelled, to simulate the linear behaviour of polymers (Table 2).Beyond this, hyperelastic modelling was invoked to simulate polymeric nonlinear stiffening effects.All the outsoles were meshed with 10-node SOLID 187 tetrahedral elements to facilitate ease in mesh convergence and enhance the models' accuracy [41][42][43].The outsoles were computationally warped at an angle of 17°and a force of 275 N was evenly applied over the upper surface of the outsole model.The deformed models were then imported in the CFD solver to estimate the fluid pressure across its treads.A comprehensive mesh convergence research with three distinct meshes at regular intervals was considered.The mesh with the lowest result variation (i.e.within 5%) was considered as the ideal mesh.Furthermore, to ensure accurate results around the worn regions of the outsoles, around 80% of the total meshed elements were ensured with an orthogonal quality of 0.9.Due to these settings, H1 was generated with 180,916 elements, H2 with 177,770 elements, H3 with 123,785 elements, H4 with 132,057 elements, H5 with 92,310 elements, H6 with 99,008, H7 with 107,274, H8 with 100,006 elements, and H9 with 100,451.
The wet slipping flow domain characterisation was done by applying an incompressible, steady-state turbulent flow system.For the conservation of mass and momentum, pre-coded equations including Reynoldsaveraged continuity and Reynolds-averaged Navier-Stokes were implemented respectively.To match the realistic slipping conditions, flow domain beneath the outsole was applied with a surface roughness of 38.45 µm.Moreover, slipping speed was mimicked   by applying a fluid velocity of 0.5 m/s towards the bent outsole.For the scaled-residual, a value of 1e-3 was selected as the convergence threshold.The SIMPLE algorithm was used as the foundation for the pressure-velocity modelling, with second order and second order upwind representations of pressure and momentum, respectively.Table 2 represents the numerical parameters applied for the slipping simulation.

Data analysis
The traction performance of the nine outsoles in new and worn condition was quantified by estimating the ACOF through mechanical slip testing in dry and wet slipping conditions.The fluid pressures across the new and worn outsoles were evaluated through the computational model.To accurately measure the worn region areas, the calculation of areas was done by importing the 3D scanned models in SolidWorks.
The quality of correlations between the worn region areas with the ACOF in dry and wet conditions, and fluid pressures was describes using correlation coefficient (R 2 ).For this study, R 2 < 0.5 was declared insignificant, 0.5 < R 2 < 0.7 was deemed moderate, and R 2 > 0.7 was deemed significant [9].The graphs presented in the upcoming sections shows the mean values whereas the error bars indicate the standard deviation.Also, the friction results are presented as mean ± standard deviation.

ACOF of footwear outsoles in new and worn condition
The friction values of the outsoles in new condition ranged from 0.15 ± 0.01 to 0.38 ± 0.01 across dry and wet slipping tests (Figure S3).Correlation between the ratio of tread area to gap area and the resulting dry ACOF showed a positive and significant R 2 of 0.72 (Figure 3).It suggests that outsoles with a lower ratio could result in high dry friction.During dry slip testing, traction values of the outsoles ranged from 0.31 ± 0.02 to 0.38 ± 0.01.H7 exhibited the highest ACOF value (i.e.0.38) and H9 showed the lowest (i.e.0.31).After H7, H4 and H1 showed similar traction behaviour which ranged from 0.35 ± 0.01 to 0.36 ± 0.01 whereas, the remaining outsoles (H2, H3, H5, H6, H8, H9) ranged between 0.31 ± 0.01 to 0.34 ± 0.01.Correlation between the ratio of tread area to gap area and the resulting wet ACOF showed an insignificant R 2 of 0.20 (Figure 3).Hence, no significant conclusion could be drawn when compared the ratio with wet friction outcomes.In wet slip testing, ACOF values ranged from 0.15 ± 0.01 to 0.20 ± 0.02 (Figure S3).Out of all the outsoles, H7 and H8 exhibited highest mean ACOF (i.e.0.20) whereas H6 and H3 showed lowest mean ACOF (i.e.0.15) during wet slip testing.After H7 and H8, H9 exhibited a higher mean ACOF value of 0.19.
The ACOF values of the outsoles after 1st wear cycle ranged from 0.11 ± 0.01 to 0.38 ± 0.02 (Figure 4).As compared to the outsoles in new condition, similar ACOF outcomes were observed for the outsoles after the initial wear in dry slipping conditions.Majority of the outsoles which had the ratio of 1.51 or above showed similar dry ACOF outcomes as compared to other outsoles (Figure 4a).Outsoles with lower area ratios showed varied outcomes across the dry slip testing.Only the outsoles H2, H4, H6 showed variations of 3%, 5%, and 6% respectively where H2 was observed with 0.32, H4 with 0.35, and H6 with 0.31.On the contrary, all the outsoles showed high variations when slip tested on water contaminated flooring (Figure 4b).The mean ACOF of worn outsoles in wet slipping conditions ranged from 0.11 to 0.16.
The ACOF values of the outsoles after the 2nd wear cycle varied from 0.10 ± 0.02 to 0.34 ± 0.01 (Figure 4).As compared to the outsoles after the 1st wear cycle, similar dry ACOF outcomes were observed for majority of the outsoles after the 2nd wear (Figure 4a).As compared to the new outsoles, except H3 and H9, all the remaining outsoles showed significant differences in the ACOF values.H3 and H9 both had wide treads, which could indicate the dependence of wide treads with retaining the outsole friction.Similarly, slip testing results in water contaminated flooring showed high variations when compared with new outsoles.The ACOF of worn outsoles in wet slipping conditions ranged from 0.10 to 0.13 (Figure 4b).
After the 3rd wear cycle, the ACOF values of the outsoles ranged from 0.07 ± 0.005 to 0.27 ± 0.01 (Figure 4).As compared to the new outsoles, all the outsoles showed significant differences in the ACOF values.Similar to new outsoles, slip testing results on water-contaminated flooring exhibited significant variations in the ACOF.In wet slipping conditions, the mean ACOF of worn outsoles varied from 0.07 to 0.15.Highest fluid pressure was experienced by H1 (i.e.625 Pa) and the lowest induced fluid pressure by H7 (i.e.453 Pa).Each outsole showed similar pressure contours with highly localised zones as formed near the worn location.

The predicted fluid pressures of footwear outsoles in new and worn condition
After the second wear cycle, the fluid pressure across the outsoles varied from 473 to 629 Pa (Figure S6).H1 experienced the highest fluid pressure (i.e.629 Pa) and H7 experienced the lowest fluid pressure (i.e.473 Pa).In a similar trend as compared to the outsoles after 1st wear cycle, each outsole showed similar pressure contours with worn locations exhibiting highly localised zones.The fluid pressure across the outsoles after the third wear cycle, ranged from 560 to 650 Pa (Figure S7).H3 exhibited the lowest fluid pressure (i.e.560 Pa) and H6 experienced the highest fluid pressure (i.e.650 Pa).

Effect of worn region area on ACOF and fluid pressure
The effect of worn region area after first, second, and third wear cycle on the ACOF in dry and wet conditions along with the fluid pressures was analysed by quantifying the correlation between them (Figure 6).The outsole worn region area after the first wear cycle ranged from 4.75 to 39.03 mm 2 .After the second wear cycle, worn region area ranged from 128.51 to 356.26 mm 2 .After the third wear cycle, worn region area varied from 508.11 to 980.58 mm 2 .
In dry slip testing condition, after first wear cycle, the worn region area showed insignificant correlation (R 2 = 0.23) whereas, it showed moderate correlation (R 2 = 0.53) with the ACOF values after second wear cycle (Figure 6a).For these conditions, no significant trend was reported.On the contrary, after third wear cycle, worn region areas correlated negatively and significantly (R 2 = 0.87) with the ACOF values.Negative trend meant the decrease in the ACOF in dry slipping conditions due to increase in worn region area.In wet slip testing condition, insignificant correlations were observed for the first and second wear cycle as it corresponded to R 2 values of 0.18 and 0.04 respectively (Figure 6b).On the other hand, after third wear cycle, significant (R 2 = 0.77) but negative trend was observed between the ACOF and worn region area.
In wet slipping simulations, the influence of increasing worn region areas on fluid pressure was quantified by estimating the quality of correlations between them.For the first and second wear cycles, worn areas and fluid pressure across the outsoles were found to be insignificantly and moderately correlated by exhibiting R 2 values of 0.03 and 0.51 respectively (Figure 6c).After the third wear cycle, a positive and significant correlation (R 2 = 0.86) was reported.Increase in the worn region area corresponded to increase in induced fluid pressures during the slipping simulations.

Discussions
This work studied the effects of new and worn footwear outsoles on their traction performance on dry and water contaminated flooring.Nine different outsole designs were considered which had horizontally oriented tread patterns.The outsoles were experimentally slip tested using a whole-shoe portable slip testing device whereas, the ability to dissipate the fluid through the tread channels were characterised by a combined computational framework consisting of static loading and CFD modelling.Significant impact was observed for the varying tread designs on the traction across dry and wet slip testing.Moreover, the effect of increasing worn region areas of the outsoles showed strong influence on the friction and induced fluid pressures.
In dry slip testing of new and initially worn outsoles, all the outsoles observed friction values more than 0.3, above which, the slip risk reportedly reduces [9].Outsoles with a lower (i.e. less than 1.5) tread area to gap area ratio showed high frictional outcomes in dry conditions.Outsoles with similar and lowest tread width exhibited increased ACOFs whereas the outsoles with larger tread widths reported reduced traction performances.Also, outsoles with high tread area to gap area showed increased worn region areas as compared to the others.These outsoles experienced increased worn regions due to increased apparent contact area which could have led to varied force distributions.Furthermore, these outsoles also experienced drastic reduction in the frictional outcomes both in dry and wet conditions.After subsequent wear cycles, majority of the outsoles were observed to perform below the threshold, which represented increased slipping risks.During wet slip testing, in each outsole segment (i.e.new and worn), generalised ACOFs were observed.This could be due to the formation of hydrodynamic fluid films at the interface between the contaminant and flooring.However, in each segment, treads with increased gap intervals showed increased ACOF values as compared to the ones with lesser gap dimension.This finding is consistent with previous studies by Hemler et al. [17] and Beschorner et al. [29].
Analysing the fluid pressure across new outsoles, the outsoles with large gaps exhibited lower pressure as compared to the ones with lesser gaps.The gaps could have allowed for the water to disperse easily with less fluid accumulations.A sudden increase in the fluid pressure at the entry and exit locations were observed in each contour, which could be due to the barrier created by perpendicular patterns.In the case of worn outsoles, localised and high pressure accumulation zones were observed on the areas surrounding the worn region.This could be due to the sudden deformity present over the outsoles due to wear.These observations show the increased slipping risks involved during testing in dry and water contaminated floorings.In the case fully worn outsoles, by analysing the fluid flow over the tread patterns, the fluid tried to escape the flow domain before entering the worn region area.This finding throws light on the strategic design of the tread patterns, which can provide enhanced traction performances even after getting fully worn.
The effect of worn region area on the ACOF in dry condition showed both insignificant and significant correlations depending on the wear cycles.First and second wear cycle observed the retention of treads to some level, which helped these outsoles generate considerably higher ACOFs as compared to third wear cycles.In these conditions, although the outsoles with lesser tread width showed higher ACOF, these outsoles also showed increased wear rates during subsequent wear cycles.Outsoles after the third wear cycle showed high increase in the worn region areas and exhibited reduced ACOF values.In this condition, absence of treads led to the outsole to behave as the base material.During wet slip testing, increased worn region areas led to increase in the area covered by the fluid film, which further led to increased fluid pressure and reduced ACOF.
Few limitations should be acknowledged.Based on previous work by Jones et al. [5] and Iraqi et al. [44] shore A hardness was reported as an appropriate parameter to compare the frictional performance of different footwears.Hence, based on these studies, this work only considered measuring the shore A hardness.Other material properties such as compressibility, durability, and stretching ability could further help understand the resulting friction and wear behaviour of the shoe outsoles.Furthermore, the statistical analysis of the variability between the outsoles was not performed due to large data set.Statistical analysis could have led to 360 comparative outcomes.Such high number of outcomes could have further led to the complexity and difficulty in understanding the results from the manufacturer's point.Future studies considering only a detailed statistical analysis could be used to further understand the variability of friction outcomes of several commonly available footwear treads.Other tribological factors such as surface profile characterisation, surface energy, and other scanning techniques (such as, SEM, AFM) might clarify the behaviour of the ACOF values across the considered conditions.Future studies quantifying these parameters will help understand the dry and fluid-structure behaviour of the footwear across the floorings under contaminant conditions.Also, as the primary focus of this work was to investigate the effect of tread width and gaps during dry and wet slips in new and worn conditions both experimentally and computationally, the roughness of the outsoles was not considered.Future studies measuring the roughness of the outsoles could further help correlate the traction performance of outsoles on different slipping conditions.
The current study was based only on wet slip simulations.Due to computational costs and complex flow patterns, it might be difficult to numerically model highly viscous contaminants such as floor cleaner and oil.In that case, physical experiments related to measuring the fluid pressures could be helpful in understanding the slipping scenarios in the presence of viscous contaminants.In future studies, the induced fluid pressures due to highly viscous contaminants (i.e.floor cleaner or oil) will be characterised through human slipping experiments.
In conclusion, outsoles with a lower (i.e. less than 1.5) tread area to gap area ratio showed high frictional outcomes in dry condition.Furthermore, increase in the worn area led to an increase in the maximum fluid pressure.This led to a significant reduction in the traction performance across both dry and wet slipping conditions.Outsoles, with increased worn regions, which generated lower ACOF and higher fluid pressures, indicated increased slipping risks.The current work reported considerably low dry friction outcomes after the second wear cycle as compared to first wear cycle.In this case, 40% of the outsoles showed dry friction values below the threshold of 0.3.This consideration was in-line with several other works, based on footwear slipping, in literature [7,9,[32][33][34].As per the wear protocol considered in our work, the second wear cycle corresponded to simulated 6 months of wear.Hence, a continuous usage period of 6 months could be suggested to the footwear manufacturers and public as a replacement threshold particularly for this type of tread design.Future works studying the effect of different tread orientations on wear or performing traction testing pilot studies involving different types of footwear could further help in determining the standardised replacement thresholds.Also, based on the above threshold value (i.e.0.3), traction performance of the outsoles across each wear cycles and slippery conditions were analysed.In future, human slipping experiments on range of contaminants, with such footwear, may be needed to effectively correlate with the chance or risk of slips.The results from this study are anticipated to provide footwear manufacturers the strategies and guidelines to design and plan the traction performances of the outsoles not only in the initial phases but after their wearing too.

Figure 3 .
Figure 3. Correlation of dry and wet ACOF of outsoles in new condition with respect to the area ratio.

Figure 5
Figure 5 represents the variation in the behaviour of the predicted fluid pressure on a single outsole (i.e.H6).In this case (H6), the maximum pressure increased from 523 to 650 Pa after 1st and 3rd wear cycle rspectively.The induced fluid pressure across the outsoles in new condition varied from 435 Pa to

Figure 4 .
Figure 4. ACOF of outsoles in worn condition: (a) Dry slip testing, and (b) Wet slip testing.

Figure 5 .
Figure 5. Induced fluid pressure of an outsole (H6) throughout its consecutive wear cycles.

Figure 6 .
Figure 6.Effect of worn region area on: (a) ACOF in dry slipping condition, (b) ACOF in wet slipping condition, and (c) fluid pressure.

Table 1 .
Details of the generated nine outsoles.

Table 2 .
Numerical parameters applied for the slipping simulation.