International Journal of Fracture 69: R57-R62, 1994/1995. © 1995 Kluwer Academic Publishers. Printed in the Netherlands. R5 7 INFLUENCE OF ANVIL SUPPORTS AND OVERHANGS IN DYNAMIC ANALYSIS OF THREE POINT BEND TESTING P.R. Marur and P.S. Nair ISRA Satellite Center, Bangalore 560 O17, India Tel: 91-80-526 62 51; Fax: 91-80-5585 407 K.R.Y. Simha Indian Institute of Science, Bangalore 560 012, India Optimum design of dynamic fracture test rigs demands a thorough apprecia- tion of beam vibration under impact. Analyses invariably presume rigid anvils, and neglect overhang effects. The beam response predicted analytically and numerically in this paper highlights the significant role of anvil rigidity and beam overhangs on the impact dynamics of three point bend (3PB) specimens. Recently, the authors presented a dynamic analysis of 3PB specimens using tup displacement as the input to a lumped mass-spring model [1]. It was concluded that using tup force directly in the model lends poor support for the experimental results. The deficiency seen with the 'tup force approach' was attributed to the uncertain contact mechanics of the interaction between the tup and the beam. However, a subsequent analysis [2] indicated that the tup load well reflects these effects in itself, and if the sensors are close to the loading end of the striker and the anvils are rigid enough, the tup force can be directly used in lumped mass-spring models to predict the Kd;time history. Hence, investigations were continued to resolve the issue, which in turn, directed the attention to anvils and overhangs. Impact experiments were carded out in an instrumented pendulum-strike system. Specimens with different dimensions and material properties were tested, but only the results pertaining to an aluminium 6061 T6 specimen (AI-10) tested with initial impact velocity of 1 m/s, are discussed here. The specimen was 172 mm long, 25.4 mm wide and 6.35 mm thick. Support span used was 152 mm, and len.gth of the initial crack was 5 mm. The impact process was modeled by a spnng-mass as described in [1], with the measured tup force applied as forcing function. Influence of anvil rigidity was demonstrated by testing the specimens with different anvils. The support used in the previous study is shown in Fig. 1 along with the improved design. It can be noted that the old design using roller suppports is more compliant than the modified version. When the specimen was tested with the old anvils, the beam response predicted using the tup force in the inertial model was in disagreement with the measured data as reported in [1]. Repeating the test in the new system with modified anvils yielded the tup and anvil force plots shown in Fig. 2. The computed and measured Kd~(t) plots are Int Journ of Fracture 69 (1994)