The Abrasion Subcommittee has been working to try to reduce the variation in abrasion testing both machine to machine and within machine. Testing has been done with the Martindale, Stoll Quartermaster and the ILE SCR 8000 abrasion testing machines. The Stoll and ILE tests were run under the current test methods. The Martindale method called for using a wool abrasive. The standard Martindale method was run in the first trials but this proved out to be not acceptable in testing socks as they contain nylon and the wool wore out before the nylon. We then ran the second round of tests using 320 grit sandpaper that was used in the ILE and Stoll methods.

The key lies in what is actually happening to the knitted fabric at the point of contact with the abrasive material. Abrasion testing of a woven fabric is much more repeatable than a knit because woven fabrics tend to be more stable and do not stretch as much as knits. After assessment of two different rounds of abrasion trials the following was noted:

• Standard deviation was not within expected parameters
• A correlation could not be established between machines due to excessive standard deviation.
• The methods are all showing results in a timely manner.
• Of major importance was that at the current rate of deviation, testing only a couple of samples is not a true reflectance of the overall product.

Trials are being run using a new abrasive material from the 3M Company called Trizact. Trizact is non-directional and is much more consistent than regular sandpaper. Initial results are promising, as the variation in number of cycles to sock failure has been reduced from 30% to 7% using Trizact and straight-line abrasion on the sock instead of the random pattern.

Design of Experiments (DOE) - At the last Abrasion Committee meeting held on June 21, 2000 Art Caldwell covered the timeline of the committee since its inception in March of 1998. Much work has been done and it clearly shows that standard abrasion test methods used for general textiles when applied to socks are not valid. Modifications of these methods have produced much better results. These modifications were made on a “gut feel” as to how to reduce variation in abrasion results. A formalized Design of Experiments (DOE) has been created to determine the factors that affect abrasion on socks and the level each factor contributes to the final results. This DOE is being performed at NC State on the Martindale and at the Hosiery Technology Center on the Stoll Quartermaster. The DOE factors are as follows:

Constant Factors

1. Using a “straight line” abrasion pattern.
2. Using Trizact as the abradant.
3. Mounting of abrasive material on the abrasion machine.
4. Abrading the bottom of the foot on the sock.
5. Choice of endpoint of the sock is when a hole appears and all yarns are cut.
6. Changing the abrasive material for each variable change.

Variable Factors

1. Mounting Tensions using a 7” plexiglass circle or the NAHM foot form.
2. Vary the weight applied to the sample. (9 and 12 kp for Martindale and 2 and 3 lb for Stoll)
3. Cleaning the lint off the sample (50 , 100 cycles for Martindale and 100, 200 cycles for Stoll).
4. This gives a total of 8 sets of conditions on two machines with 15 samples run on each condition and machine. This is a total of 240 socks tested.