Emerging of the Problem

The first sign of a problem was noticed in September of 2002. A board manufactured in August 2002 and exposed in the weathering box for only 860 h, showed cracks on the surface. Similar boards made in September 2002 showed cracks at the surface after 830 h in the box. These were isolated effects, and they were attributed to some technological deviations at the plant in Green Bay. However, based on those observations, in November of 2002 the amount of antioxidants was almost doubled compared to the original manufacturer s recommendations, and these boards with increased amount of the antioxidant were produced from November 2002 to January 2003. These boards were tested in the weathering box for 2000 h and were found quite satisfactory. 

Starting January 2003 GeoDeck boards were made without adding antioxidant, as the HDPE supplier had assured that the plastic had an unusually high amount of antioxidants. These boards were made until March-April 2003, and after that the manufacturing continued with another HDPE source but the additions of the antioxidant was not resumed. This was a dramatic mistake. Manufacturing without the antioxidant was continued until the beginning of October 2003. The plant QC system could not detect the problem because there was no procedure in place that would have been able to detect a low amount (or a complete absence) of antioxidants in composite materials. The procedure was developed and implemented in the plant QC system later (see below). 

In July of 2003 a new step was introduced at the plant, namely annealing of GeoDeck boards, in order to eliminate boards shrinkage. Initially, third-party kilns were employed to do the job, and in many cases boards were obviously overheated. They often became brittle and discolored. Being placed in the weathering box, they showed crumbling—in the worst case—even after 4 days. More detailed studies have shown that the lower the density, the less time it takes for board crumbling. Boards of density 1.06-1.07 g/caf crumbled after 4 days. Boards of density 1.08-1.10 g/cm crumbled after 7-12 days. Boards of density 1.15-1.16 g/cm did not crumble after a month. 

Density (Specific Gravity) of GeoDeck Boards in Pre-October 2003 

Density of GeoDeck boards manufactured in January-September of 2003 was low (Table 15.35) compared to those manufactured in 2002 and 2004 (see below). 

---