A crumbly running track

The insured company was in the business of installing running track surfaces for high school and collegiate athletic stadiums. The process was to grade the earth, lay down a layer of small, ground up, recycled rubber from tires, spray the surface with a mixture of latex and polyurethane, wait two days for the mixture to cure, and then use the track. After two weeks, a newly installed track at a high school was still sticky and the rubber bits were crumbing under the weight of the runners. GEI was contacted to determine the cause of failure of the curing process.

The components for building this track included crumbed rubber, a latex from Dow Chemical, and a polyurethane binder from an independent supplier (henceforth referred to as X). After several consultants examined the installation and prepared several reports, a final decision was made to remove what was clearly not an acceptable track and replace it with one that was acceptable. A consideration of applying some combination of treatments to yield an acceptable track surface was considered, but rejected.

We received a cross sectional piece of the failed track and two bottles of the polyurethane mixture that originated with X.

Since the polyurethane that was used was thought to be “bad” polyurethane, we sought to obtain  another sample of this polyurethane from the supplier that would be a “good” sample. The “good” urethane would be from a batch of material that had worked as expected during a track installation. This would have provided a basis for analysis and performance comparison for the “good” and “bad” urethane, and a possible determination of the failure cause. Unfortunately, after a discussion about the product and its use with X, the company did not respond to several subsequent requests for a sample of “good” polyurethane. Hence, we were left with the ability to examine only the “bad” polyurethane.
A copy of the X Technical Data Sheet for the spray binder product described the product and how it was to be applied. High temperature and high humidity were reported to increase the curing rate, and the reverse would slow the rate of cure. The application guide did not specifically warn against an application at relative humidities below 50%.
A qualitative test was done by spreading the X Spray Binder material on a surface and letting it cure. The conditions were indoors at about 85 degrees F and 85% Relative Humidity, which were well within the recommended range for the use of the product. After several hours standing, the material still had not cured sufficiently to prevent some flowing. After 24 hours curing, there was no discernible flow, but the material continued to be soft and sticky. After two weeks, the urethane appeared fully cured, but did not cure to a hard surface. The final cure was consistent with the product description that it is designed to have elastomeric properties. The product had a low boiling ketone-type smell, again consistent with the technical data provided by X, that it was a solvent based top coat, and the solvent should evaporate quickly in use.
During a number of discussions, we learned several important facts. The insured stated that he also later had performed a similar qualitative test on the batch of the X product. In his draw down test, the product did not cure as well as it should, in his opinion. Also, according to the insured, the latex component was a Dow product that had been used by them in the past with no difficulties. Thus, there was no reason to suspect the latex to be flawed, but there was reason to suspect the polyurethane.
An infra-red (IR) analysis of the polyurethane and the track samples showed little urethane present at the bottom of the failed track sample. It was essentially only rubber crumb and latex. The “bad” polyurethane was used as the control or standard sample for the IR.The IR of the track showed only some cured urethane on the surface of the rubber crumb (and no latex) at the top of the sample. The top was essentially just the loose rubber crumb. That is, the binders were essentially segregated with latex at the bottom and polyurethane on the rubber at the top of the failed track sample. Latex alone cannot bind the rubber crumb into an acceptable track, nor could urethane alone; hence the urethane/latex mixture was intended to provide the final binding of all the components.
The segregation of the polyurethane in the failed track sample meant only one of two possibilities. Either it was not applied in the correct manner to bind the track materials or the curing of the polyurethane did not bind the latex and rubber crumb as it was intended.
The application guide for the spray binder only states that the curing rate is slower in low humidity. This was an insufficient warning for the product’s use in the 20% Relative Humidity actually experienced at the time of application. This polyurethane was the wrong product, in the wrong place, at the wrong time. The track had the appearance of polyurethane sprayed on to a rubber crumb, from which all the latex had flowed down into the bottom of the track. The polyurethane could never have moisture-cured quickly enough to set all of the components.
All polyurethanes are not identical. There are numerous formulations and products with various properties. There were, however, several specific characteristics in the polyurethane provided by X that contributed to the final result. First, this was a moisture-cured product, meaning that it required water to cure,it was used in a very dry climate. Next, the X spray binder product was a proprietary formulation, apparently prepared on a batch basis, as needed. We did not receive any direct evidence that this particular batch of X spray binder polyurethane had ever worked successfully in the constructing of a running track.

To conclude, the X spray binder aromatic product did not perform adequately to bind the principal components, which lead to track failure. There was no evidence that this batch of product would have performed adequately even under ideal curing conditions.