CP620, Shock Compression of Condensed Matter - 2001 edited by M. D. Furnish, N. N. Thadhani, and Y. Horie © 2002 American Institute of Physics 0-7354-0068-7/02/$ 19.00 DYNAMIC TENSILE RESPONSE OF ALUMINA-AL COMPOSITES R. Atisivan,1 A. Bandyopadhyay1 and Y. M. Gupta2 Institute for Shock Physics and School of Mechanical and Materials Engineering Institute for Shock Physics and Department of Physics Washington State University, Pullman, WA 99164 Abstract. Plate impact experiments were carried out to examine the high strain-rate tensile response of alumina-aluminum (Al) composites with tailored microstructures. A novel processing technique was used to fabricate interpenetrating phase alumina-aluminum composites with controlled microstructures. Fused deposition modeling (FDM), a commercially available rapid prototyping technique, was used to produce the controlled porosity mullite ceramic preforms. Alumina-Al composites were then processed via reactive metal infiltration of porous mullite ceramics. With this approach, both the micro as well as the macro structures can be designed via computer aided design (CAD) to tailor the properties of the composites. Two sets of dynamic tensile experiments were performed. In the first, the metal content was varied between 23 and 39 wt. percent. In the second, the microstructure was varied while holding the metal content nearly constant. Samples with higher metal content, as expected, displayed better spall resistance. For a given metal content, samples with finer metal diameter showed better spall resistance. Relationship of the microstructural parameters on the dynamic tensile response of the structured composites is discussed here. used to develop the method to create tension in the specimens. The method to produce tension in the sample subjected to uniaxial strain loading was similar to that described earlier by Barbee et al. (5) In this method, the test sample was surrounded by a tightly fitted material of nearly the same shock impedance. The function of the surrounding material was to minimize the radial release waves from propagating into the sample. This is termed as 'momentum trapping'. Unlike quasi-static tests, surface effects play a negligible role in these dynamic experiments, since the tension is produced inside the sample. Because microdamage in brittle materials take place at very rapid rates, these experiments provide a means of initiating microcracks and controlling the load duration such that these cracks do not coalesce into large-scale cracks and rupture the material. In this research, the role of metal content and microstructure on the dynamic tensile response of alumina-Al composites were studied. INTRODUCTION Three dimensionally honeycomb porous mullite ceramic pre-forms with interconnected pore structure were fabricated via the indirect fused deposition process. With this approach, it is possible to control the size, distribution and connectivity of the pores in the resulting porous preform. The porous ceramic structures were then infiltrated with molten Al metal by pressureless reactive metal infiltration to form interpenetrating phase alumina-Al metal-ceramic composites. The details of processing of these structured composites has been described elsewhere. (1, 2, 3) This approach can be used to design a part using CAD in which microstructural features as fine as 50 microns can be incorporated and fabricated directly. This paper discusses the dynamic tensile response of these composite materials. In this work, the behavior of the composite material under high strain rate loading was examined using plate impact experiments. The work of Gupta et al. (4) was 697 EXPERIMENTAL PROCEDURE 120- Cylindrical alumina-Al composite specimens with 30 mm diameter and 25 mm long were fabricated to examine the dynamic tensile response. All the samples were processed under identical conditions. Three sets of composite samples were fabricated for the dynamic experiments with metal contents of: 26, 32 and 42 (vol%). These compositions were used to examine the effect of metal content on the spall behavior of the composites. Two more sets of samples with nearly same metal concentration (~39 vol%), but having different metal and ceramic widths were fabricated. These samples were used to examine the effect of microstructure on the spall behavior of the composites. The samples with 32% metal content were chosen for the velocity interferometry measurements under two different projectile velocities, which provided the sample response under uniaxial loading. The response for the 32% metal content was used for estimating stresses in all other samples. .—v 100- £ "o> > £ 40- Time (jj.s) Figure 1. The particle velocity profiles for the two the VISAR shots. 1614- 12- To characterize the sample response under uniaxial strain, particle velocity measurements were carried out with VISAR measurement using a laser interferometry system. (6) These measurements were used to estimate the initial compressive stress state in the composite samples, when subjected to dynamic loading. Measurements were made at two different projectile velocities, 139 and 197 m/s. Particle velocity histories for the two experiments are shown in Fig 1. The horizontal lines in Fig 1 represent the average particle velocities. The average particle velocity for experiment 1 was calculated to be 66±2.6 m/s. The average particle velocity for experiment 2 was 92±5.1 m/s. The average and standard deviation values were calculated for the time interval between O.OSjis to 0.5ns. 10- £ 55 6H 4- 20.00 a = 149.9 u 0.02 0.04 0.06 0.08 0.10 0.12 Particle Velodty (mm/ps) Figure 2. The P-u diagram for the composite sample. is assumed to be a homogeneous material, the shock speed in the composite can be calculated from the second Rankine-Hugoniot (R-H) jump condition and is found to be 4.98 mm/jis. Using a linear elastic deformation assumption, the shock impedance of the composite was calculated as 149.9 kbar/(mm/|is). As this value is fairly close to the impedance of Al 6061T6, it was believed that the aluminum holder minimized the propagation of radial release waves into the samples. The particle velocity measurements obtained using samples with 32% metal were also used for all other samples. Fig 2 shows the P-u diagram for the composite that was obtained from the two VISAR experiments. The particle velocity used to construct this plot was calculated by determining the change in the particle velocity from the projectile velocity. This P-u plot was used to estimate the initial compressive stressstate in the recovery experiments. The slope of the P-u plot for the composite gives the shock impedance of the composite. Since the composite 698 Recovery Experiments: A schematic of the experimental configuration is shown in Fig 3. Recovery experiments were performed on polished composite samples cut parallel to the build layers direction. Each sample had the shape of a frustum with 25 mm as the smaller diameter and a taper of 8°. The machined samples were 6 mm thick. Shock waves were generated by impacting 3 mm thick Al-6061 impactors. The Al impactor was mounted onto a projectile and accelerated to a velocity of ~0.140 mm/|is using a 100 mm compressed gas gun. The projectile velocities were measured just prior to impact using a method requiring piezo-electric pins. Samples failed due to tensile stress that was generated in the sample. Due to the taper, the sample leaves the Al holder just after impact and is collected in a recovery chamber at the end of the gas gun filled with soft rags. Test sample Influence of metal content: The metal content in the samples was increased between 23 to 39 vol%. The thickness of the metal strands was kept constant at 300 Jim. The ceramic thickness was varied from 330 Jim to 825 |im to change the metal content in the composite. The projectile velocity was nearly constant (~ 140 m/s) for all the experiments. It was found that except 39% metal content sample, all other samples split into two pieces. Samples with lesser metal contents split into two pieces along the diameter. These results indicate that the tensile stresses were generated inside the sample as expected. Moreover, samples with higher metal content have better spall resistance. This is likely due to the fact that the toughness of the composite increases with increase in metal content. The ceramic is inherently less resistant to spall compared to the metal. As the relative amount of metal in the composite decreased, the spall resistance decreased as well. The experiment with 39% metal content was repeated five times to verify that the results are reproducible. In two experiments, the samples broke into two pieces but the rest of them were intact in one piece. Influence of metal width: In this set of experiments, the width of the metal strands were varied from 300 to 400 |im, keeping the metal content nearly constant (39 vol%) in the composite. As the metal thickness was increased, the thickness of the ceramic strand was increased to keep the amount of metal content constant, while the number of metal strands in a given sample size decreased. It was found that except the composite with 300 Jim metal width, all other samples split into two parts. This result suggests that finer microstructure provide higher toughness. Aluminum 6061 Figure 3. The target assembly with alumina-Al composite sample RESULTS Two sets of experiments were performed. The first set of experiments examined the effect of metal content on the tensile response of these composites. In the other set of experiments, the metal width of the composites was varied while keeping the metal content nearly constant. The consequent effect of the microstructure on the failure behavior was examined. 699 Failure Analysis: The recovered samples with 39% metal were mounted in an epoxy, and cut along the diameter. The sample was polished, and examined under an optical microscope. Figure 4 shows the damage accumulation due to tension created in the sample. The tensile crack can be observed to have propagated through the interior of the sample. It can also be seen that the crack had propagated predominantly through the ceramic rich region in the composite. Similar results are observed in other cases as well. Figure 5 shows a scanning electron micrograph of a fracture surface with 32% metal content that splitted in to two pieces. In general, the failure analysis suggests that the failure was initiated primarily in the ceramic phase and then propagated through the ceramic and the metal phases. CONCLUSIONS An experimental method to subject the structured composite samples under dynamic tension was established. Using particle velocity measurements, the shock impedance of the composite sample with 32 vol% metal was calculated to be 149.9 kbar/ (rnm/jis). The shock velocity was inferred to be 4.9 rnm/jis. The behavior of composite samples under dynamic tension Figure 5. An SEM micrograph of a fracture surface for 32% metal content samples that split into two pieces. ACKNOWLEDGEMENTS Authors would like to acknowledge financial supports from the US Department of Energy contract number DEFG0397SF21388 and the Office of Naval Research grant number N00014-98-01-0550. We would also like to acknowledge experimental support from Kurt Zimmerman and Jens Darsell of Institute for Shock Physics. Direction of impact REFERENCES 1. Bandyopadhyay, A. "Functionally Designed 3-3 Mullite-Aluminum Composites," Advanced Engineering Materials, 1 [3-4], pp. 199-201 2. 3. Figure 4. (Top) Fractograph of Sample tested at 145 m/s. (Bottom) The enlarged optical micrograph of the encircled portion in Figure 4a. 4. was examined. Composite samples with higher metal content showed a better spall resistance. For a given metal content, samples with finer metal width showed better spall resistance. Fractographic analysis showed extensive cracking in the ceramic rich region and relatively less damage was observed in the metallic phase and at metal-ceramic interface. 5. 6. 700 (1999). Atisivan R., S. Bose and A. Bandyopadhyay, "Porous Mullite Preforms via Fused Deposition," Journal of the American Ceramic Society, 84 [1], pp. 221-23 (2001). Soundararajan R., R. Atisivan, G. Kuhn, S. Bose and A. Bandyopadhyay, "Processing of MulliteAl Composites," Journal of the American Ceramic Society, 84 [3], pp. 509-13 (2001). Gupta Y. M., P. D. Horn and J. A. Burt, J.Appl.Phys. 76[3], 1784-88, 1994. Barbeejr. T. W., L. Seaman, R. Crewdson, and D. R. Curran, Journal of Materials, 7[3], 393401, 1972. Barker L. M. and R. E. Hollenbach, J. Appl. Phys. 43, 4669-4675, 1975.
© Copyright 2025 Paperzz