Development of an outsole compound for outdoor footwear.

In late 1993 the Outdoor Footwear group requested development of a new outsole compound for water sandals. The material used in the outsoles of our water sandals had been in use for many years as a fairly generic outsole material. This material was very durable in outsole use, and provided good traction on typical dry outdoor hard surfaces such as concrete and asphalt. This material provided adequate but not particularly good traction on wet rocky surfaces frequently encountered by those wearing water sandals. The Outdoor group wanted to develop higher perfonnance water sandals which required better traction on wet rock surfaces than our current product offered. This new outsole material also needed to have good durability, be non-marking, be reasonably priced and not create processing difficulties in production.

This development project had four phases. The first phase was initial material selection and screening. A large number of material formulations was made up in the process lab and submitted to various laboratory tests for evaluation. Mechanical traction testing and laboratory abrasion test results were considered the most important test results in this round of evaluation. Several compounds provided better wet traction than the standard material while maintaining good abrasion resistance. Phase 2 tests involved laboratory tests on prototype footwear produced in a factory using the candidate compounds. Based on these test results, the list of candidate materials was further reduced. Phase 3 testing involved field tests of the three best candidate materials using prototype sandals produced using nortnal production processes. Phase 4 testing involved laboratory abrasion, traction and physical tests of various minor variations of the best compound to n-dniniize cost and address sourcing issues in various offshore factories.

Phase 1 - initial laboratory testing - material specilmens

The first phase of this project involved development of some experimental compounds that would be expected to have improved wet traction (as compared to the control), as well as an acceptable level of abrasion resistance. As in development of a tire tread compound for which traction and abrasion resistance are important, blends of standard thennoset elastomers, such as polybutadiene, SBR, NR, nitrile, EPDM, halobutyl and polychloroprene were evaluated. Since a non-marking outsole was required, only silica, clays and other white fillers were included. Compounds were niixed on a laboratory mill using standard mixing procedures. Tensile sheets were cured at 160[degrees] C. The compounds were then subjected to the following battery of tests:

* Abrasion tests (Nike internal; Akron abrader);

* Durometer tests (ASTM - D2210; Shore A);

* 300% modulus test (ASTM - D412);

* Tensile test (ASTM - D412);

* Elongation test (ASTM - D412);

* Tear strength test (ASTM - D624);

* Mechanical traction test (using Nike's traction test device and standard protocol).

Mechanical traction tests were conducted by dragging rectangular sections of each material across a mica rock surface in both dry and wet surface conditions. The specimens were flat material specimens having no tread pattem. A vertical load of approximately 1,000 N was placed on the specimens during the traction tests. A Kistler force platform was used to measure the vertical force and the force in the direction of motion. The ratio of these forces is the coefficient of friction. Since the specimens were in motion during testing the coefficient of friction measured and recorded was the dynamic coefficient of friction.

Typical mechanical traction tests are conducted on clean and dry surfaces. Since this project was focused on developing an outsole material with improved wet traction characteristics, wet surfaces were also used in mechanical testing. This was accomplished by using a hand held spray bottle to apply water to the test surface. Water was sprayed after each trial to ensure that the surface was wet for each trial. Previous work has shown that the actual volume of water sprayed on the test surface does not need to be carefully metered or controlled. Excess water is pushed off the test surface by the specimens tested without affecting the measured coefficient of friction values. The physical test results for a represcntative sampling of the materials tested are presented in table 1.

Table 1 - physical properties of experimental outsoles

Specimen   Akron   Durometer   300%   Tensile,  Elongation   Tear
          abrsion,  Shore A  modulus,   MPa         %      strength,
          cc loss              MPa                          kg/cm

Control    0.38       73      4.22      8.33       450       49
Spec. A    0.39       69      4.71     11.97       781       52
Spec. B    0.47       69      2.75     10.03       937       43
Spec. C    0.36       70      7.26      9.31       430       43
Spec. D    0.22       68      4.41     20.01       950       61
Spec. E    0.20       66      9.81     12.45       370       53
Spec. F    0.28       68      5.98      9.02       650       45
Spec. G    0.08       55      3.13     13.53       720       49
Spec. H    0.19       67      3.73     13.73       600       62
Spec. I    0.28       62      7.65     12.25       460       41
Spec. J    0.19       52      4.90     11.76       610       36
Spec. K    0.38       57      5.59      8.24       370       34
Spec. L    0.33       48      5.20      8.14       480       28
Spec. M    0.51       55      7.06      8.92       425       38
Spec. N    0.27       55      4.02      9.81       580       40
Spec. O    0.48       48      3.43      8.43       650       33
Spec. P    0.12       66       -       12.85       260       42

The mecanical traction test results for a representative sampling of the materials evaluated on dry and wet mica rock surfaces are presented in figures 1 and 2.

Many of the materials tested had higher dynamic coefficient of friction values than the standard outsole material. Several of the prototype materials also had equal or better abrasion test results than the standard outsole material. The specimens tested were flat sections of material having no tread pattern. These mechanical traction test results are quite useful for comparing the relative performance of various materials for the test conditions used in the laboratory. However, the dynamic coefficient of friction values obtained for these specimens are only a rough indicator of how footwear made with these materials will perform. The outsoles of athletic footwear have tread pattems specifically designed for each product. Tread pattern design can have a significant effect on the in-service performance of athletic footwear outsoles just as it does for automobile tires.

An obvious question is whether viscoelastic properties, such as S'or tan [Delta], could be used to predict traction, especially wet traction properties. We suspect that the answer is yes; however, extensive testing to develop correlations between test conditions and product performance has not been done in the footwear industry.

Phase 2 - initial performance testing - product prototypes A number of candidate materials evaluated in phase I looked promising. Potential processing and material cost factors were then considered. This resulted in a further reduction in the number of materials of interest. This smaller number of candidate materials was screened further in phase 2 testing. Several additional candidate forinulations were developed by various production laboratories and included in the phase 2 evaluation. The phase 2 evaluation was based on initial performance testing of the first round of prototype products made with the candidate materials. One model of water sandal was used for the initial round of test specimens in phase 2. The samples tested were made in production facilities using standard production processes and the prototype outsole materials. Mechanical traction testing was used as the primary indicator of performance for the test specimens. Laboratory traction tests were conducted on a mica rock surface in both dry and wet surface conditions. Mechanical traction test results for the prototype sandals are presented in figures 3 and 4.

The differences in traction performance of the various outsole materials evaluated were more pronounced on the wet mica surface than on the dry mica surface. Note that all specimens had lower dynaniic coefficient of friction values on the wet mica than on dry mica. As discussed by Grosch (ref. 1), high levels of wet traction can be obtained. However, the actual area of contact between the rubber and surface is reduced, as compared to dry conditions, so the absolute friction in value is lower. Conducting mechanical traction tests on these sandals proved to be difficult due to the high traction performance and the upper design of these models. The upper design of these particular sandals did fit well on our standard footwear fixturing. Due to the compressed schedule for this project, it was determined that all future mechanical traction tests for this project would be conducted using the prototype outsole materials on shoe samples rather than on sandals.

During this testing phase, there was discussion as to whether niica was die most representative surface to test sandals or climbing shoes against, as the sandals would be used on a variety of rock surfaces. Samples of river rock, granite and basalt were collected, and traction testing was run under wet and dry conditions. Mechanical traction tests were conducted on the four rock surfaces using the control material, the most promising protomw material (material E) and two additional prototype materials. These tests were conducted using the prototype outsole materials on an outdoor shoe. The tracfion test results on dry rock and wet rock surfaces are presented in figures 5 and 6. Physical properties are shown in table 2. Two inter-esting results were obtained from this phase of testing. First, specimen E had the highest dry and wet coefficient of ftiction values on all four rock surfaces used and the control specimen had the lowest coefficient of friction values on all four rock surfaces tested. Secondly, the mica surface showed a better ability to distinguish between compounds. Therefore, the decision was made to use only the mica surface in future testing of outdoor outsole materials.

These tests were conducted using prototype footwear products, therefore the dynamic coefficient of friction values can be used to tell us something about the in-service performance of these prototypes. There is considerable debate about tht minimum coefficient of friction required between footwear and surfaces for humans to be able to walk without shpping. Jbere is wide variation in the kinematics and kinetics of wafidng in humans. The kinematics and kinetics directly affect the minimum coefficient of friction required for walking without slipping. In addition, humans adjust their gait biomechanics when confronted with different surface conditions. Most people have leamed to take shorter steps and push off with less force when wawing on icy sidewalks than when wafldng on dry sidewalks. Tbere is also disagreement on use of static coefficient versus dynamic coefficient of ftiction. In spite of the many disagreements in this area, a consensus is building for using a static coefficient of 0.5 as a reasonable minimum value for walking on dry, clean and level surfaces. Dynamic coefficient of friction values are generally lower than static coefficient of friction values, therefore all of the specimens tested in phase 2 appear to provide adequate traction for wal%g on dry mica rock surfaces. The mechanical test results for the control specimen indicate that some people could slip while attempting to walk or run on wet niica surfaces. Test results for specimen E indicate that most people should be able to walk and run without slipping on all fout rock surfaces in both wet and dry conditions.

Phase 3 - field testing - product prototypes T'he third phase of the evaluation process involved field testing the three most proniising new outsole material formulations. Twenty pairs of sandals were made in the factory using standard production processes. Five pairs of sandals were made with the standard outsole material to serve as control specimens and five pairs of sandals were made with each of the three prototype materials (prototype materials E, Q and R). All sandals were made in the same factory using the same upper designs and same outsole tread pattems. Eight subjects were recruited for the field test which took place in a local state park. Each subject used a pair of sandals with each type of outsole material for 15-20 minutes of waffing activity in two different areas of the park. One area was a river where the subjects walked along the bank of the river on both rocky and sandy areas. The subjects also walked into the river and across submerged rocks of various sizes and through submerged silty areas. The subjects also hiked up and back down a moderately steep dirt trail. The subjects were asked to rank the relative traction performance of the sandals and to rate the traction performance of these sandals relative to all sports sandals using a seven point scale. The order in which different outsole materials were evaluated by each subject was randomized among subjects.

There was strong agreement among the subjects as to the relative ranking of the traction perfonnance of the sandals evaluated. The sandals with outsole material E were perceived as providing the best traction in afl conditions evaluated. Most subjects felt that the best of the sandals performed better than other sport sandals they had previously used. Sandals with the control outsole material were rated as providing the least traction of the four specimens tested. These results were consistent with phase 2 mechanical traction test results.

Phase 4 - commerciauization The water sandals were scheduled for overseas production. The best prototype material (specimen E) was the compound selected to be commercialized. One of the base polymers used in compound E was a high styrene, aromatic oil extended SBR that was readily available in Asia. Also, it was determined that there might be some colored areas in the outsole, so it would be necessary to replace the aromatic oil to eliniinate the possibility of staining. Each of the three factories chosen for production developed a slightly different version, using locally available materials. In place of the staining polymer, a non-oe high styrene SBR was used, with a high styrene resin added to bring the styrene level up to that of compound E. Also, a non-aromatic oil was added (to the satne phr level) to develop a production ready material. Test specimens were cured and used in our phase 4 evaluation; the specimens were flat material specimens with no tread pattem. Therefore, these test results can not be directly compared to phase 2 test results which were based on testing footwear rather than material specimens. The mechanical traction test results for this last round of testing are presented in figures 7 and 8. Phyyical test results are included in table 3.

As can'be seen, there is some reduction in traction and physical properties, as compared to the development compound. However, after initial introduction, there was additional work to bring the formulations and properties into closer agreement with each other. Feedback from consumers confirmed improvements in wet and dry traction over the previous outsole compound, with its use expanded into hiking and waffing products. Additional work has continued in this area, with continued improvements in wet and dry outsole traction.

References

1. Grosch, K.A., "The rolling resistance, wear and traction properties of tread comp6unds, "Rubber Chemistry and Technology - Rubber Reviews, Vol 69, No. 3, p. 495-562, July-august 1996).