PSI - Issue 24

Michele Perrella et al. / Procedia Structural Integrity 24 (2019) 601–611 Author name / Structural Integrity Procedia 00 (2019) 000–000

609

9

Considering that the tested resin is suggested to be used for civil and industrial application in the range of temperature up to 35°C, another creep compliance master curve was obtained at 30°C (Fig. 9) by using the same procedure with the Arrhenius curve method.

0.00000 0.00020 0.00040 0.00060 0.00080 0.00100 0.00120 0.00140 0.00160 0.00180 Creep compliance J [ e /MPa]

1E+03

1E+04

1E+05

1E+06

1E+07

1E+08

1E+09

log(time) [s]

Fig. 9. Creep compliance master curve at 30°C.

The same level of deformation (about 10000 e ) was predicted by about 5 years at 30°C instead of 40 years at 25°C, thus highlighting how the long-term behavior of the tested resin is strongly dependent on temperature. 4. Conclusions Short-time measurements of creep compliance at different temperatures under constant stress are useful data for predicting models on the residual strength of mechanical structures. But, in civil field, where adhesive bonding is used for strengthening purposes the long-term behavior (up to 50 years) of the repairs becomes critical. Obtaining accurate experimental data for long-term strains under different loading conditions (stress and temperature) are extremely time consuming and cost intensive. Moreover, industrial applications require quick responses in order to provide with suitable solutions. Within this context, the well-known correspondence between time and temperature effects on deformation of polymers subject to stress conditions below the glass transition temperature offers a convenient method, i.e. the TTSP, to arrange the short-term test results into long-term data. In this paper an in house made test equipment was proposed for performing creep tests undergoing controlled temperature conditions. A dead weight load was imposed for obtaining a stress controlled test, while the strains were carried out by means of DIC technique and strain gauge data acquisition system. A good agreement between DIC and strain gauges data was obtained. Creep phenomenon was investigated on flat dogbone samples made of Sikadur30 epoxy resin subject to different temperatures (40°C, 35°C, 30°C, and 25°C) and to a fixed stress corresponding to the 20% of static tensile strength. A couple of tests were carried out for each imposed temperature showing a good repeatability and a data dispersion less than 7%, thus providing a satisfactory accuracy of the testing system. Experimental data highlighted the strong dependence of long-term behavior of the tested resin on temperature. Tests at 40°C showed high deferred strains (about 10000 e ) after 5 days as well as a relevant primary creep stage. This could be related to the proximity of the glass transition temperature of the resin. After the same time, the outcomes at 35°C, 30°C and 25°C tests provided strains of about 4000 e 1400 e and 800 e , respectively, and a strain rate significantly lower than the one resulting from creep experiment at 40°C. Based on these short-term results creep compliance master curve were obtained at 25° C and 30° C using the TSTP and the Arrhenius model, being the tests performed below the glass transition region of a commercial resin. In such a way, predicting creep strain response over about 5 years and 40 years could be achieved at 30 °C and 25 °C, respectively. The presented experimental data can be of interest for evaluating the deferred residual strength of industrial and civil applications of structural bonding.