Use of master curves based on time-temperature superposition to predict creep failure of aluminium-glass adhesive joints

Abstract

Advancements in materials technology and the use of innovative designs have led to extensive application of adhesive bonding techniques in the electric appliance industry. While the static strength of such joints is sufficient for the intended applications, long term durability remains a major concern, mainly due to creep effects. Conventional creep testing can be performed at the service temperature but it is a long test that can take decades, although it can be accelerated using high temperatures. In this work, glass-aluminium joints were studied under static and creep loads. Glass-aluminium specimens were subjected to creep testing at various temperatures. Using the time temperature superposition principle, the results of these individual creep tests were combined in a master curve that approximates the creep behaviour of the adhesive joint in a long time period. These master curves were used to guarantee a minimum service life of the joint.

Introduction

The durability of adhesively bonded joints is a research topic of significant importance, which has been gaining relevance as adhesive bonding is more widely adopted by several different industrial users, among them the home appliance manufacturers. In this industry, a recent drive to use high performance materials such as tempered glass and aluminium in home appliances combined with aesthetic concerns (i.e., no visible fasteners or brackets) has led to the implementation of alternative joining techniques such as adhesive bonding. However, the complex nature of the time-dependent (viscoelastic) behaviour of polymeric adhesives [1] leads to significant difficulties in making long term behaviour predictions. As these products will operate continuously for very large periods of time, creep becomes a significant design factor. Manufacturers are therefore demanding a simple and reliable method to determine the creep behaviour of an adhesive joint and ensure that their product will not fail during its work life [2]. However, the determination of creep damage is still a largely experimental process that requires a significant amount of time and resources to provide useful results. New equipment is still being devised with the aim of simplifying the process of creep testing adhesive joints [3]. Due to the time-scales involved, testing creep in real-time is impossible, as long term creep tests of the finished products are totally impractical, as the joints might have failure times measured in decades. Nonetheless, if the tests are performed at temperatures significantly above the service temperature, this will accelerate the creep process and might allow the determination of results in a practical amount of time [4]. The time-temperature superposition principle (TTSP) is one of the most commonly used methods, able to combine the data from shorter tests performed at various temperatures into a master curve, able to describe the creep performance of an adhesive joint [5]. This principle was first identified by the work of Leaderman [6]. This principle makes use of the fact that for many materials the creep compliance versus logarithm of time curves have the same shape for different temperatures, but the increases in temperature have the effect of shortening the time scale. Eventually it was determined by Tobolsky and Andrews [7] that this relationship can be used to combine individual creep compliance curves and assemble a curve that fully represents the creep deformation of an adhesive or material at any desired temperature.

To create a master curve a specimen must be subjected to a constant load at a certain temperature. The creep of this specimen must be registered in a graph with time given in logarithm scale. Several experiments are performed at different temperatures, registering a creep curve for each temperature. To construct the master curve, a reference temperature has to be first defined. The individual creep curves obtained at each temperature are then shifted along the time scale and, starting with the reference temperature, superposed to obtain a master curve. The adjustment of the curves must be such that there is a smooth transition between the individual curves. It is sometimes difficult to ensure that there is a smooth overlap of master curves and some materials might require not only the horizontal time shift, but also a vertical time shift. These materials are classified as thermorheologically complex materials [1], [7], [8].

It must be remembered though, that the TTSP creep study does not produce an exact model but an approximation and that the creep behaviour of an adhesive joint is the result of various coincident factors that act on a given adhesive joint. The creep results must be interpreted accordingly and sufficient margins must be introduced to account for these effects. The accuracy of master curves depends on the following factors [9]:

• Variation of the shift factors with temperature.

• Existence of the same creep mechanism at the different temperatures tested.

• The initial strain rate applied to the specimens.

• Variation in humidity.

• State of the polymeric material (glassy rubbery or on the transition zone).

• Rate of application of heat to achieve the desired temperature level.

The creep of highly plastic silicone adhesives was previously studied by Geiss and Voigt [10], and significant creep strain was identified. However, no master curve was derived from this work, therefore no long term durability prediction was made.

In this work, two silicone based adhesives were mechanically tested in joints to assess the joint strength they can provide. Surface treatments as well as curing times were also object of study. One of the two adhesives was then subjected to creep tests at various temperatures and, using the time-temperature superposition principle, this data was then used to build master curves and draw conclusions regarding the durability of the joint.