Characterization of Fatigue and Aging Effects

To complete the investigation of the assessment potential of the DMA material testing approach, initial studies to qualify and quantify fatigue and ageing effects were done. A permanent loading of the sealing joint (e. g., one that might result from the permanent (dead) load of the glass pane) under adverse condi­tions of + 80°C was simulated in the creep-test mode. According to Fig. 15, after

Copyright by ASTM Int’l (all rights reserved); Tue May 6 12:07:08 EDT 2014 Downloaded/printed by

Rochester Institute Of Technology pursuant to License Agreement. No further reproductions authorized.

FIG. 14—Curing kinetics for Sealants A and B at +30°C and 40 % RH (time-sweep mode).

20 h of loading, the typical creep deformation, and after an additional release period of 4 h, the material resilience, respectively, permanently viscous defor­mation could be determined for the three sealants that were exemplarily tested. A time-dependent stability ranking was determined for the sealants as follows: Sealant C > Sealant B > Sealant A. Converted to the geometrical conditions of a 12 x 6 mm2 joint (width x depth), the maximum measured viscous deforma­tion range from 1.6 mm (Sealant A) to 0.6 mm (Sealant B) down to 0.1 mm (Sealant C). Under a permanent load of 0.1 MPa from the time-deformation curve, it was apparent that Sealants A and B had distinctive visco-elastic mate­rial characteristics, whereas Sealant C exhibited predominating elastic behavior.

Even for a five times higher permanent load of 0.5 MPa, creep deformation of less than 0.3 mm without any remaining viscous effects after load removal were observed for Sealant C, underlining its special performance (not shown in figure here).

The potential to detect effects of artificial weathering was evaluated by com­parative measurements in temperature-sweep mode for Sealants A and B before and after 1300 h of weathering (see Fig. 16). Weathering was carried out accord­ing to ASTM C 1442-11 [13]. As can be seen from Fig. 16 for Sealant A, the com­paratively lower and steady course of the complex shear modulus G* after weathering indicates subsequent curing effects (induced by the weathering re­gime). A secondary transition temperature (T3) is no longer detected from the

Copyright by ASTM Int’l (all rights reserved); Tue May 6 12:07:08 EDT 2014 3

Downloaded/printed by

Rochester Institute Of Technology pursuant to License Agreement. No further reproductions authorized.

trace of tand indicating changes of the inner network. In addition, Sealant A exhibits, after weathering, a shift of the low transition temperature T1 to a 4° C higher value. The mechanical behavior of Sealant B, after 1300 h of weathering according to ASTM C 1442, is not affected (not shown in figure here).

The author recommends that in future studies the mechanical response of SSG sealants to superimposed artificial fatigue and ageing exposure within the dynamical spectrometer device itself should be evaluated. A first validation attempt in this direction was made in this study using superimposed independ­ently controlled shear and tension loading in combination with temperature and UV loading as a simplified durability test mode on the three sealants. The sealants were subjected to:

• ± 10% torsional deformation at a frequency f) of 0.1 Hz to simulate effects of daily temperature changes acting on the sealant joint,

• a constant normal force of 3 N to simulate the effects of permanent (dead) loading,

• continuous UV-light exposure (kmax. at 365 nm) with 452 Wh/m2, and

• constant temperature of + 80°C.

As can be seen from Fig. 17, Sealant A exhibited a considerable decrease in stiffness (50%) after 7200 load cycles that were meant to replicate daily temper­ature changes on the outside of a building. The behavior of parameter d (repre­senting the specimen length) under this aging regime indicates permanent flow. Although the stiffness loss rate decreased already, after around 23500 load

FIG. 17—Response of mechanical properties to superimposed loading (Sealant A).

Copyright by ASTM Int’l (all rights reserved); Tue May 6 12:07:08 EDT 2014 Downloaded/printed by

cycles, the remaining stiffness was only 35% of the initial value. The sealant was able to sustain in total 24000 load cycles before failing cohesively.

In contrast to Sealant A, Sealant B displayed much greater “durability" (see Fig. 18). After an initial stiffness decrease to adjust to the system tuning (to reach a balanced system state under loading), the stiffness steadily increased. Although not all structural processes are clarified, a subsequent curing is to be expected for the first test period (up to 3500 min) taking into account the results of the investigation of curing behavior. An explanation for the simultaneously occuring slight normal force decrease can be possibly a “wear-out" effect caused by fatigue, but this is compensated for, because of the normal force readjust­ment to 3 N. After more than 3500 min of loading time, a slight stiffness increase is correlated with increasing normal forces. This effect was interpreted as the beginning of an embrittlement of the cured sealant. This is also in agree­ment with the visual appearance indicating ageing effects by a shiny glazed specimen surface opposite to the UV source. After 39000 load cycles (daily tem­perature changes), no mechanical failure was detected. Although not all effects are, at this stage, fully evaluated, the dynamic-mechanical material test approach presented here provides an assessment of the sealants’ behavior under fatigue and ageing loading.