PSI - Issue 13

Christopher Schmandt et al. / Procedia Structural Integrity 13 (2018) 799–805 C. Schmandt, S. Marzi / Structural Integrity Procedia 230 (2018) ECF22

800

2

Though, Dillard (2010) delineates advances in polyurethane structural adhesive bonding and Loureiro et al. (2010) describe various advantages of polyurethane adhesives compared to epoxies such as better damping properties, fatigue resistance or energy consumption under impact, for instance. Due to these benefits of hyperelastic adhesives, their application field has spread in recent years beyond the standard application case of windshield bonding to other structural applications concerning the main car body. The polyurethane adhesive to be investigated here is currently used by car manufacturing industry to bond carbon-fiber-reinforced plastics to the aluminum chassis. Due to its viscoelastic characteristics, loading rate is significantly influencing the material behavior of hyperelastic adhesives. Therefore, topic of the present work is investigating effects of loading rate on the peel fracture behavior. 2. Methods Investigations on mode I fracture behavior were carried out by performing DCB tests on a hyperelastic one component polyurethane adhesive system, Sikaflex® UHM, which is very tough and possesses high modulus. All substrates were made of AlZn5,5MgCu (3.4365), which is a high-strength aluminum alloy. Fig. 1 shows dimensions, loading conditions and measuring locations regarding acting force F , substrate rotation θ and crack opening displacement u of the used DCB specimens. Within manufacturing, bonding surfaces of substrates were sandblasted with fine-grained corundum. After degreasing with butanone, Sika® Primer-207 was applied at room temperature to the bonding surfaces by usage of melamine foams. Adhesive and prepared substrates were slowly warmed up to a temperature of 50 °C before applying the adhesive layer. Spacers made of PTFE were used to achieve a defined layer thickness of 3.0 mm and were removed after completed curing process. Specimens were fixed by screw clamps during the entire curing process for at least two weeks with a constant room temperature of 23 °C and relative humidity of approximately 50 %. Afterwards, surplus adhesive at the edges of the specimen was cut off by use of a utility knife. A pre-crack was caused by inserting a razor blade parallel to the bonding surfaces at the beginning of adhesive layer. The layer thickness at the position of pre-crack was determined by caliper gauge measurements. The specimen was loaded by crosshead movement of an electromechanical tensile testing machine. Special clamping devices, equipped with high resolution incremental rotary encoders, were utilized to measure substrate rotation θ at the position of load introduction (Fig. 2 (a)). The tensile force F , acting on the substrates, was measured by a load cell of 5 kN capacity. The crack opening displacement u was captured by a line sensor system. Thereby, white measuring flags were attached to upper and lower substrates at the position of initial crack tip. The line sensor accurately detects variations in contrast between white edges of measuring flags and black background. The line sensor system was oriented in direction of specimen’s longitudinal axis to avoid errors resulting from rigid body movement of the specimen during increasing loading. Additionally, a camera system was set up at right angles to specimen’s longitudinal axis, focusing on the specimen’s edge to visualize crack propagation (Fig. 2 (b)).

Fig. 1. DCB specimen with introduced pre-crack, dimensions and locations of loading and measurement.