Currently there are no standard tests using strain gauges to monitor strain in an adhesive layer. However, structural monitoring capabilities, where strain-sensing devices (e.g. Fibre Bragg Grating) are embedded in materials, are the subject of research in many organisations. Strain gauges can be attached to adherends and will measure the strain in these adherends. The usefulness of such measurements may be limited except in the cases where changes in joint performance are manifested in measurable changes in the adherend strain. One such application is back-face strain gauging of thin lap-shear joints where crack growth in the adhesive layer can be monitored through strains measured by gauges bonded to the external surface of the adherends at the overlap. Strain gauges have been used for determining the onset and growth of localised damage in bonded structures.
Strain gauges are generally limited to the measurement of strains less than 10%. Biaxial rosettes are available for measuring longitudinal and lateral strains. Large strain gauges are preferable as alignment and handling is easier, and they average out local strain variations. Local strain variations can cause premature failure of the strain gauges. Correct alignment of strain gauges is important, as significant errors can be caused by careless application of the strain gauges to the specimen. Errors of 15% can occur from a 2° misalignment [4].
The adhesive used to bond the strain gauge should be capable of withstanding the test environment for the complete duration of the test. Most adhesives are sensitive to moisture (and other chemicals), which can often preclude bonding prior to specimen conditioning. Moisture attack of an adhesive and strain gauges will occur from the top, edges and in the case of polymeric materials through the substrate beneath the gauge. The situation is exacerbated at elevated temperatures. It is therefore important to ensure that the adhesive selected for bonding the strain gauge and associated electrical wiring is suitably encapsulated.
The strain gauges are usually bonded to the specimen following moisture conditioning (i.e. immersion in water or exposure to humid environments). However, bonding the strain gauge to the specimen may require heat and pressure, which will induce drying out of the conditioned specimen. To avoid drying out, room temperature curing anaerobic adhesives have been used and have proved satisfactory for bonding strain gauges to moisture conditioned specimens. For hot/wet conditions, a high temperature anaerobic adhesive can be used provided the application temperature does not result in thermal damage to the adhesive joint (i.e. adhesive and adherend). Although anaerobic adhesives have good moisture, solvent and temperature degradation resistance, these adhesives are known to attack certain plastics. Hence, precautions need to be taken when selecting these materials for use with plastics or fibre-reinforced polymer composites. Cyanoacrylates (or super glues), which are sensitive to surface moisture and low pH levels, are unsuitable for environmental testing. Strain gauge manufacturers can provide information on adhesive selection and procedures for protecting strain gauges.
For cyclic loading, it is essential that the fatigue life of the strain gauges, over the operating strain levels, should be well in excess of the life expectancy of the test component. Autogenous (self-generated) heating can degrade the mechanical properties of the adhesive bond between strain gauges and the specimen. This can result in small errors in strain measurement, thus requiring correction of the data to account for the temperature rise. Measurements should also be carried out to determine the magnitude of creep within the strain gauge adhesive.
An approximate measurement of strain and hence stiffness can be obtained from measuring the crosshead displacement of the test frame. The strain is the ratio of crosshead displacement and the initial grip separation. Hence, any slippage within the loading train will produce errors in the strain measurement. The strain values obtained from crosshead measurements will differ from the actual strain in the central region of the specimen.
Stiffness measurements directly obtained from the crosshead movement need to be corrected to take into account the stiffness of the loading train. This can be a difficult task as the specimen size and geometry, and the deformation behaviour of the specimen need to be taken into account. Given the small adhesive layer deflections that occur even at large strains owing to thin bondlines, the accuracy of strains determined using crosshead displacements must be considered suspect and used only for qualitative purposes.
LVDTs are recommended in preference to monitoring crosshead movement. These devices provide a direct reading of the moving part and can be attached at any point on the structure as required. LVDTs tend to be used to monitor global rather than localised deformation. Accurate alignment is essential otherwise measurement errors will occur and the movement of the device can be restricted.
ESPI is a non-contact technique capable of measuring and monitoring non-uniform strain fields at high resolution. The system (see Figure 5) can measure the deformation and thus the strain under mechanical and/or thermal loads along the three material axes (i.e. 3-D strain measurement). ESPI systems are capable of measuring local deformation with a resolution of 0.1 mm, equivalent to 200 microstrain for a 0.5 mm thick bond.
The technique needs minimal specimen preparation and is capable of inspecting areas ranging from 25 mm2 to 600 mm2, but capital outlay for equipment is generally prohibitive for most test facilities. The technique can be used to measure strain distributions in complex geometries, and for checking finite element analysis. Figure 6 shows a typical speckle pattern resulting from out-of-plane deformation. Strain distribution in a double-lap joint at different loads is shown in Figure 7. Further details on the technique with illustrated case studies are given in reference [5]. Interferometry techniques are not routine and are thus unlikely to be suitable for mass screening programmes. Similarly the technique may not be suitable for cyclic testing.
Figure 5: ESPI System Layout [6]
Figure 6: A typical speckle pattern resulting from out-of-plane deformation [6]
Undeformed (top); deformed (middle); and difference (bottom)
Figure 7: Strain distribution in a double-lap joint at different loads [5]
(a) 2.5 kN, (b) 10 kN, (c) 15 kN and (d) 20 kN
