Fatigue and Durability of Silane-Bonded Epoxy/Glass Interfaces

Fatigue (slow) crack growth in epoxy/glass interfaces bonded with the silane coupling agent 3-aminopropyltriethoxysilane was studied under static and cyclic loading at 23°C, 95% RH using the double cleavage drilled compression test. Crack growth rates under cyclic loading were significantly greater than under static loading, in contrast to crack growth rate results in monolithic glass. After aging up to 34 h at 94°C in distilled water, the silane-bonded epoxy/glass specimens exhibited somewhat greater resistance to fatigue crack growth than the unaged samples; however, after aging at 98°C in distilled water and at 70°C in an aqueous KOH solution at pH 10, crack growth became cohesive and exhibited fractal behavior. Mechanisms for fatigue crack growth at silane-bonded epoxy/glass interfaces are proposed.     

Crack Propagation in Polymer Adhesive/Glass Sandwich Specimens

Moisture-assisted crack growth in polymer adhesive/glass interfaces was measured as a function of the applied energy release rate, G, using a four-point flexure test coupled with an inverted microscope. The specimens consisted of two glass plates bonded together with an epoxy or an epoxy-acrylate adhesive. It was found that cracks formed and grew on both interfaces if the glass surfaces were both smooth; however, roughening the surface of one of the glass plates increased the fracture resistance of the interface sufficiently so that crack growth occurred only on the remaining “smooth” interface (top or bottom). Finite element analysis was used to determine the G and ψ (phase angle) appropriate for the different crack geometries. It was found experimentally that crack growth rates for all crack geometries depended on the applied G via a power law relationship and that for a given applied G, crack growth rates were sensitive to the crack geometry. The results indicate that the primary driving force for moisture-assisted crack growth at a polymer/glass interface is the applied G at the crack tip and that the effect of the phase angle for the different crack geometries (13° to 54°) is negligible.     

Cyclic fatigue—crack growth along polymer/glass interfaces

Delamination of polymer/glass interfaces was studied under cyclic and monotonic loading using an interfacial, four-point flexure sandwich specimen. Specifically, crack growth rates along epoxy acrylate/glass interfaces were characterized over a range of velocities from 10⁻⁹ to 10⁻⁶ m/s as a function of low (10–20% RH) and high (75–80% RH) humidities. For low humidities, interfacial crack growth rates under cyclic loading are almost two orders of magnitude greater than those under monotonic loading with energy release rates G = G_max of the cyclic loading. At high humidities, interfacial crack growth rates under monotonic loading are approximately equal to cyclic fatigue crack growth rates at low humidity.     

Moisture-assisted crack propagation at polymer/glass interfaces

Moisture-assisted crack growth at various polymer/glass interfaces was measured as a function of applied strain energy release rate and relative humidity using a four-point flexure apparatus coupled with an inverted microscope. The specimens consisted of two glass plates bonded together with a thin layer of commercially available epoxy-acrylate, urethane-acrylate or epoxy adhesive. The crack front at the epoxy-acrylate and epoxy interfaces was relatively smooth and, above a threshold strain energy release rate, the crack growth rate was dependent on the applied strain energy release rate via a power law relationship. Crack growth along the urethane-acrylate interface was characterized by the development of finger-like perturbations along the advancing crack front. These finger-like perturbations grew until they reached a steady-state length. Once the fingers reached steady-state, the crack growth rate of the overallcrack front was dependent on the applied strain energy release rate via a power law function. With all the polymer adhesives crack growth rates increased with higher relative humidities.     

Crack Propagation in Polymer Adhesive/Glass Sandwich Specimens

Moisture-assisted crack growth in polymer adhesive/glass interfaces was measured as a function of the applied energy release rate, G, using a four-point flexure test coupled with an inverted microscope. The specimens consisted of two glass plates bonded together with an epoxy or an epoxy-acrylate adhesive. It was found that cracks formed and grew on both interfaces if the glass surfaces were both smooth; however, roughening the surface of one of the glass plates increased the fracture resistance of the interface sufficiently so that crack growth occurred only on the remaining “smooth” interface (top or bottom). Finite element analysis was used to determine the G and ψ (phase angle) appropriate for the different crack geometries. It was found experimentally that crack growth rates for all crack geometries depended on the applied G via a power law relationship and that for a given applied G, crack growth rates were sensitive to the crack geometry. The results indicate that the primary driving force for moisture-assisted crack growth at a polymer/glass interface is the applied G at the crack tip and that the effect of the phase angle for the different crack geometries (13° to 54°) is negligible.     

Fatigue and Durability of Silane-Bonded Epoxy/Glass Interfaces

Fatigue (slow) crack growth in epoxy/glass interfaces bonded with the silane coupling agent 3-aminopropyltriethoxysilane was studied under static and cyclic loading at 23°C, 95% RH using the double cleavage drilled compression test. Crack growth rates under cyclic loading were significantly greater than under static loading, in contrast to crack growth rate results in monolithic glass. After aging up to 34 h at 94°C in distilled water, the silane-bonded epoxy/glass specimens exhibited somewhat greater resistance to fatigue crack growth than the unaged samples; however, after aging at 98°C in distilled water and at 70°C in an aqueous KOH solution at pH 10, crack growth became cohesive and exhibited fractal behavior. Mechanisms for fatigue crack growth at silane-bonded epoxy/glass interfaces are proposed.     

Mixed-mode fatigue behavior of degraded toughened epoxy adhesive joints

Open-faced asymmetric double cantilever beam (ADCB) specimens of toughened epoxy-aluminum adhesive joints were aged either in a constant humidity environment or a cyclically changing environment to study the mixed-mode fatigue behavior. Under constant humidity environments, the fatigue threshold strain energy release rate initially decreased with aging time until it reached a constant minimum value for long times. In contrast, the crack growth rates continued to increase with aging time. It is hypothesized that at crack growth rates close to threshold the fatigue behavior is governed by the epoxy matrix, whereas at relatively high crack growth rates the fatigue behavior is governed by the loss of the rubber toughening mechanism. Increasing the aging temperature accelerated the degradation of the joints leading to a reduction in the time to reach the constant minimum value and increased the crack growth rates.Under a cyclic aging environment with intermittent salt spray, neither the threshold strain energy release rate nor the crack growth rates degraded until four weeks of aging. The superior fatigue performance of these joints compared to joints aged in constant humidity environments was due to the lower water concentrations in the adhesive while aging. This conclusion was supported by moisture uptake measurements of the adhesive in deionised and salt water environments that showed simple Fickian behavior at room temperature and dual-Fickian behavior at higher temperature. The salt water environment produced osmotic pressure that decreased the moisture concentration in the second stage of diffusion.     

Fatigue delamination growth under cyclic compression in glass/epoxy composite beam/plates

Results are reported on the fatigue growth of internal delaminations in glass/epoxy composite beam/plates subjected to constant amplitude cyclic compression. Because of compressive loading, these structures undergo repeated buckling/unloading of the delaminated layer with a resulting reduction of the interlayer resistance. A noteworthy feature of the problem is that the state of stress near the delamination tip is of mixed mode (I and II). The present combined experimental/analytical investigation for the glass/epoxy composites complements our earlier studies on delamination growth under cyclic compression in unidirectional graphite/epoxy specimens. Several configurations are studied with the delamination located at different depths (through the thickness) and with different applied maximum compressive displacements. The experimental data are correlated with the predictions from a combined delamination buckling/postbuckling and fracture mechanics model. A mode-dependent fatigue delamination growth law is used together with an initial postbuckling solution for the deformation pattern of the delaminated layer and the substrate, which does not impose any restrictive assumptions on the delamination thickness and plate length. The experimental data seem to be adequately correlated with the theory and the fatigue delamination growth is found again to be strongly affected by the relative location of the delamination through the plate thickness. Finally, a comparison of the cyclic growth rate in glass/epoxy specimens with the corresponding one in graphite/epoxy specimens of the same geometry and applied loading shows that the delamination would grow much faster in the graphite/epoxy specimens.     

The prediction of crack growth in bonded joints under cyclic-fatigue loading I. Experimental studies

The performance of adhesively-bonded joints under monotonic and cyclic-fatigue loading has been investigated using a fracture-mechanics approach. The joints consisted of an epoxy film adhesive which was employed to bond aluminium-alloy substrates. The effects of undertaking cyclic-fatigue tests in (a) a ‘dry’ environment of 55% relative humidity at 23°C, and (b) a ‘wet’ environment of immersion in distilled water at 28°C were investigated. In particular, the influence of employing different surface pretreatments for the aluminium-alloy substrates was examined. In addition, single-lap joints were tested under cyclic fatigue loading in the two test environments, and a back-face strain technique has been used which revealed that crack propagation, rather than crack initiation, occupied the dominant proportion of the fatigue lifetime of the single-lap joints. In Part II, the data obtained in the present Part I paper will be employed to predict theoretically the lifetime of the adhesively-bonded single-lap joint specimens.     

Cyclic Fatigue Crack Growth in Three-Point Bending PZT Ceramics under Electromechanical Loading

This paper investigates experimentally and analytically the cyclic fatigue crack growth in piezoelectric ceramics under electromechanical loading. Cyclic crack growth tests were conducted on lead zirconate titanate (PZT) ceramics subjected to dc electric fields, and a finite element analysis was used to calculate the maximum energy release rate for the permeable crack model. Based on bending experiments using single-edge precracked-beam specimens, cyclic fatigue crack growth rates are found to be sensitive to the maximum energy release rate and applied dc electric fields. Possible mechanisms for crack growth were discussed by scanning electron microscope examination of the fracture surface of the PZT ceramics.     

Subcritical Delamination in Epoxy Bonds to Silicon and Glass Adherends: Effect of Temperature and Preconditioning

The effects of temperature and preconditioning in deionized (DI) water and a cyan ink vehicle used in inkjet printer cartridges on the durability of glass/epoxy and silicon/epoxy systems have been investigated. A test matrix consisting of test temperatures, preconditioning temperatures, preconditioning times, and nature of adherends and adhesives was developed and a series of experiments was conducted using wedge test specimens (glass or silicon coupons bonded with epoxy) to investigate the subcritical adhesion performance of the glass/epoxy and silicon/epoxy interfaces. The glass/epoxy and silicon/epoxy interfaces were found to be relatively insensitive to temperature over a range of 22–60°, but significant temperature effects, more complex than suggested by time-temperature superposition (TTSP), were observed above 60°C, depending on the environmental chemistry and nature of the adhesive used.

Specimens made of silicon coupons bonded with epoxy were subjected to preconditioning in DI water and the cyan ink vehicle prior to wedge insertion to study the effect of prior environmental exposure. The wedge test data from preconditioned specimens were compared with standard wedge test results and the Si/epoxy interface was found to be insensitive to preconditioning in DI water but was affected significantly by preconditioning in the cyan ink vehicle. Plots of crack velocity versus applied strain energy release rate for particular sets of environmental conditions are presented and a comparison is made for different environmental conditions to quantify the subcritical debonding behavior of systems studied.     

Fatigue behavior of neat and long glass fiber (LGF) reinforced blends of nylon 66 and isotactic PP

This paper focuses on the study of the fatigue behavior of neat and long glass fiber (LGF) reinforced nylon 66/PP-blends. The fatigue was characterized using Parislaw plots in the stable crack growth acceleration range. The fatigue crack propagation (FCP) is presented as a function of the crack growth per cycle (da/dN), the amplitude of the stress intensity factor ΔK, and of the strain energy release rate ΔG. It was also of interest to compare the order of performance found in fatigue to that in the static fracture test. The fracture surfaces were characterized with SEM to determine the failure mechanisms. Further, thermographic camera recordings were used to study the size of a “heated” area (ΔT = 2°C) that developed around the crack tip during the cyclic loading of LGF-PP with different amounts of maleic anhydride grafted PP (PP-g-MAH). For the neat materials, a different order of performance was detected under static and cyclic loading. This was explained by the different failure mechanisms observed after static and cyclic fracture that were related to different stress states of the specimens during the fracture process. On the other hand, the LGF-blends showed a similar order of performance during the static and the fatigue test. This was explained by the observation that similar fiber related failure mechanisms occurred in the composite, both after failure caused by the static and cyclic loading, respectively. For the LGF-PPs with varying PP-g-MAH content, the order of performance in fatigue did not correspond to the size of the “heated area” around the crack tip. This was caused by a change in the composite failure mechanisms, which contributed differently to the size of the “heated area” and to the fatigue performance.     

Simulation of Mixed-Mode I/II Fatigue Crack Propagation in Adhesive Joints with a Modified Cohesive Zone Model

A model to predict fatigue crack growth in bonded joints under mixed mode I/II conditions is developed in this work. The model is implemented in the finite element software ABAQUS using the related USDFLD subroutine. The present model is based on the cohesive zone (CZ) concept, where damage develops according to the value of the opening/sliding at the bondline under static loading, and according to a cyclic damage accumulation law under fatigue loading. The damage accumulation law is obtained by distributing the cyclic crack area increment over the process zone ahead of the crack tip, where the cyclic crack area increment is calculated according to a Paris-like law that relates the crack growth rate to the applied loading. In this way, the experimental crack growth rate is related directly to damage evolution in the cohesive zone, i.e., no additional parameters have to be tuned besides the quasi-static cohesive zone parameters.     

High-Temperature Crack Growth in Monolithic and SiCw-Reinforced Silicon Nitride under Static and Cyclic Loads

Experimental results are presented on subcritical crack growth under sustained and cyclic loads in a HIPed Si₃N₄ at 1450°C and a hot–pressed Si₃N₄–10 vol% SiC_w composite in the temperature range 1300°–1400°C. Static and cyclic crack growth rates are obtained from the threshold for the onset of stable fracture with different cyclic frequencies and load ratios. Fatigue crack growth rates for both the monolithic and SiC_w-reinforced Si₃N₄ are generally higher than the crack growth velocities predicted using static crack growth data. However, the threshold stress intensity factor ranges for the onset of crack growth are always higher under cyclic loads than for sustained load fracture. Electron microscopy of crack wake contact and crack–tip damage illustrate the mechanisms of subcritical crack growth under static and cyclic loading. Critical experiments have been conducted systematically to measure the fracture initiation toughness at room temperature, after advancing the crack subcritically by a controlled amount under static or cyclic loads at elevated temperatures. Results of these experiments quantify the extent of degradation in crack–wake bridging due to cyclically varying loads. The effects of preexisting glass phase on elevated temperature fatigue and fracture are examined, and the creep crack growth behavior of Si₃N₄–based ceramics is compared with that of oxide-based ceramics.     

Fatigue crack growth law for ferroelectrics under cyclic electrical and combined electromechanical loading

New data sets of crack propagation in lead-zirconate-titanate DCB specimens under cyclic electric loading combined with a constant mechanical load have been obtained. Both an increasing mechanical load as well as an increasing field amplitude resulted in an enhanced crack propagation rate. The experiment was modelled with a Finite Element Analysis that used special crack tip elements and assumed a finite permeability of the crack. The calculations revealed a dielectric crack closure effect, explaining the experimentally observed threshold of fatigue crack growth for the electric load. Fracture quantities suitable for cyclic loading by electric fields above the coercive field were discussed and a Mode-IV intensity factor considered as appropriate. The resulting correlations were applied to the experimental results and a power law relationship for the crack growth rate versus the range of the Mode-IV intensity factor was found.     

Fatigue peeling at rubber interfaces

Abstract

A fatigue peeling test has been developed to evaluate the failure of rubber to rubber interfaces under cyclic loading. Results obtained through this method have been compared to those of a typical fatigue crack growth experiment. The results show that the trends between these two failure modes are similar with the peeling necessary to drive the crack being slightly higher than the strain energy release rate at the same crack growth rate. Cyclic and time dependent contributions to the fatigue crack growth behaviour have been calculated using this test for an styrene–butadiene rubber compound and the results appear to be consistent with previous work although the origin of the cyclic contribution remains uncertain. The influence of pressure at the interface during vulcanisation has also been investigated and it has been observed that the fatigue peel behaviour is proportional to the surface area of contact developed during the curing cycle.     

Rate Effects in Critical Loads for Radial Cracking in Ceramic Coatings

Rate effects in the Hertzian contact loading of model glass/polycarbonate and silicon/polycarbonate bilayers bonded by epoxy adhesive are examined. Glass is used because of its high susceptibility to slow crack growth, making this conventional contribution to the rate dependencies easy to distinguish. Silicon is used as a control material with effectively no slow crack growth. Abrasion damage is introduced into the undersurfaces of the brittle coating layers to provide controlled flaws for the initiation of radial cracks from flexural stresses introduced by the contact loading. Critical loads are measured as a function of loading rate. Comparative flexural strength tests on free-standing abraded specimens show a pronounced rate dependence in the glass but none in the silicon, entirely consistent with slow crack growth effects. The glass/polycarbonate bilayer critical load data show a similar trend, but with stronger loading-rate dependence, suggesting an extraneous contribution to the kinetics from the adhesive/substrate. The silicon/polycarbonate bilayer data also show a loading-rate dependence, albeit much smaller, confirming this last conclusion. Data from cyclic contact tests on the glass/polycarbonate bilayers coincide with the loading-rate data on lifetime plots, eliminating the likelihood of a mechanical component in the fatigue response. It is concluded that the adhesive/substrate contribution is viscoelastic in nature, from energy-dissipating (but noncumulative) anelastic deformation during the cyclic loading. Critical load tests on bilayers with different exposures to external water show no influence of external environment, suggesting that internal moisture is responsible for the slow crack growth in the glass-coating bilayers.     

Static and Cyclic Fatigue of Alumina at High Temperatures: II, Failure Analysis

Failure mechanisms of an alumina, tested at 1200°C under static and various cyclic loading conditions, were examined. Slow crack growth of a single crack is the dominant mechanism for the failure in specimens under cyclic loading with a short duration of maximum stress at all applied stress levels, as well as at high applied loads for static loading and cyclic loading with a longer hold time at maximum stress. At low stress levels, failure of static loading and cyclic loading with a longer hold time at maximum stress might occur by formation and/or growth of multiple macrocracks. More importantly, for all the given loading conditions. The viscous glassy phase behind the crack tip could have a bridging effect on the crack surfaces. A simplified model for calculating effective stress intensity factor at the crack tip under static and various cyclic loading demonstrated a trend consistent with the stress–life data.     

Crack-Bridging Analysis for Alumina Ceramics under Monotonic and Cyclic Loading

Frictional degradation of grain-localized bridges behind a crack tip has been recognized as the major cyclic fatigue mechanism in alumina ceramics. Such a fatigue mechanism implies that the crack growth resistance ( R ) curve behavior during cyclic fatigue is different from that of monotonic loading due to the reduction in crack-tip shielding. A recent crack-bridging theory based on crack compliances is used to study the bridging stresses under monotonic loading and during cyclic fatigue. The bridging-stress distributions of two coarse-grained aluminas under monotonic loading are determined using compliance measurements. Because the interlocking grain bridges at the crack wake are subject to frictional damage from cyclic loading, the bridging-stress distribution evaluated during cyclic fatigue is distinct from that for monotonic loading. These results indicate that it is incorrect to incorporate the R -curve behavior from monotonic loading to the analysis of cyclic fatigue of alumina ceramics.     

Fatigue behavior of filament‐wound glass fiber reinforced epoxy composite tubes under tension/torsion biaxial loading

A study of filament‐wound glass fiber/epoxy composite tubes under biaxial fatigue loading is presented. The focus is placed on fatigue lives of tubular specimens under tension/torsion biaxial loading at low cycle up to 100,000 cycles. Filament‐wound glass‐fiber/epoxy tubular specimens with three different lay‐up configurations, namely [±35°]_n, [±55°]_n, and [±70°]_n lay‐ups, are subjected to in‐phase proportional biaxial cyclic loading conditions. The effects of winding angle and biaxiality ratio on the multiaxial fatigue performance of composites are discussed. Specimens are also tested under two cyclic stress ratio: R = 0 and R = −1. The experimental results reveal that both tensile and compressive loading have an influence on the multiaxial fatigue strength, especially for [±35°]_n specimens. A damage model proposed in the literature is applied to predict multiaxial fatigue life of filament‐wound composites and the predictions are compared with the experimental results. It is shown that the model is unsuitable for describing the multiaxial fatigue life under different cyclic stress ratios. POLYM. COMPOS. 28:116–123, 2007. © 2007 Society of Plastics Engineers