Results and Discussion

Dynamically Strained Samples

During the summer, the dynamic strain on the samples was recorded and is shown in Fig. 2. In this figure, the average strain was 7 % with peaks of 16 % strain based on a zero-point strain of a warm day, May 10.

Sealant A shows deformation upon removal from the testing fixtures as shown in Fig. 3. The dynamically strained Sealant A samples showed a signifi­cant compressive or tensile set when first removed from the testing apparatus. After the samples were forced into their initial pre-exposure sample dimensions for a period of ten days (using gage blocks or clips), both replicates that expe­rienced either tension or compression were stable at the sample dimensions observed prior to exposure.

Stress-relaxation tests on these samples, shown in Fig. 4, reveal an interest­ing pattern. The baseline data for samples that did not experience summer exposure are shown as circles. The samples that experienced no strain during exposure (squares) increased in either modulus or stiffness with no change in curve shape. This finding most likely represents a change in the molecular



FIG. 4—The stress-relaxation modulus for Sealant A shown for prior to any exposure (Baseline), after exposure with no strain (0 % strain), after exposure to dynamic com­pression (shown in Fig. 2) (Summer/compression), and after exposure to dynamic ten­sion (shown in Fig. 2) (Summer/tension). The points represent the mean values and the error bars are the relative standard uncertainty.

weight between cross-links. The samples that experienced compression during the summer (inverted triangles) have a modulus lower than the zero-strain samples, but greater than the baseline, again with the same curve shape. The samples that experienced tension during the summer decreased in the modulus and had a much flatter curve shape compared with the baseline samples. This finding suggests that the sealant responded differently to exposure with or with­out strain, and the type of strain (compressive or tensile) also had an effect.

Sealant C had more dimensional stability as shown in Fig. 5 than Sealant A shown in Fig. 3. Both before and after exposure, the dimensions of the Sealant C did not differ within the experimental uncertainty. Again, after exposure, the sealant was forced back into the pre-exposure sample dimensions and main­tained for ten days.

The stress-relaxation data for Sealant C (Fig. 6) revealed that with just exposure and no strain, the sample modulus decreased, but was not signifi­cantly different within the experimental uncertainty. The samples that experi­enced dynamic compression during exposure showed a decrease in the modu­lus that was statistically different from both Sealant C samples with exposure and no strain or Sealant C without exposure. The samples that experienced tension during the summer showed larger decreases in the modulus. This result suggests that for Sealant C, there is a single mechanism that affects modulus change, but the introduction of both strain and the type of strain has a signifi­cant effect on the resulting modulus.