The primary questions in dynamic fracture of nominally elastic materials are the appropriateness of the stress-intensity factor characterization of the crack tip fields and the possible material-dependent relation between the dynamic fracture toughness and the crack tip velocity. Despite the availability of substantial amounts of experimental and analytical results, these issues are yet to be unambiguously resolved, primarily because of the contradictory nature of some of the results. With the suspicion that at least part, if not most, of this confusion is due to the inappropriate application of a strictly asymptotic analytical result to real life experiments, a new optical technique has been developed to help resolve some of these issues. This technique is called the "Coherent Gradient Sensor" (C.G.S.), and is capable of measuring gradients of out of plane displacements when used in a reflection arrangement, and gradients of stresses when used in transmission. By using this technique in conjunction with high speed photography, we plan to investigate the above basic issues. Concurrently, a parallel effort is underway to numerically simulate the dynamic experiments in order to correlate with the results obtained by C.G.S.
This work concentrates on the experimental study of damage evolution in thick composite plates subjected to quasi-static compression and impact loaded by high-speed projectiles. Issues investigated include the formation of localized damage zones in individual plies (king bands) as well as areas of high-speed delamination between plies. In addition, dynamic crack initiation and growth in unidirectional thick composites is also studied. High-speed photography and optical interferometric methods are used to observe the phenomena described above in real time and to facilitate the formulation of failure criteria in heterogeneous multilayered systems.
A 3-D continuum model recently developed by Gurson and Rice is used to investigate the fracture initiation process in highly ductile metals. In such ductile solids cracks typically initiate and propagate in a "tunneling" mode through the specimen thickness. Initiation is then followed by the formulation of extensive shear lips on the specimen surface.
The model views crack growth as a consequence of the nucleation growth and coalescence of microvoids in front of the main crack and is able to capture the basic phenomenon of tunneling in three dimensions. The associated experiments use a newly developed technique called the "Coherent Gradient Sensor" (C.G.S.), as well as "Shadow Moire", to study the tunneling process and to investigate the applicability of a variety of fracture criteria.
Future plans include the extension of the damage model concept to dynamic deformations. Here we plan to extend the classical constitutive models of viscoplasticity to include damage. We plan to develop damage evolution laws applicable over a wide range of strain rates and temperatures. The newly developed constitutive description will then be applied to numerical modelling of dynamic crack growth in metals.
The field of dynamic crack initiation and growth in bimaterial interfaces is virtually unexplored from both the experimental and analytical points of view. We plan to conduct a series of dynamic experiments in PMMA-steel, PMMA-glass, and PMMA-Al3O bimaterial fracture specimens. High speed optical and thermographic techniques will be used in this study. Experiments concentrate on the dependence of the bimaterial toughness on near tip mode-mix and crack tip velocity. The numerical modelling concentrates on the study of the effect of near tip three-dimensionality in the interpretation of experimental (optical) measurements performed by means of high speed photography. Particular emphasis is given to crack tip velocity regimes which are transonic with respect to one of the bimaterial constituents. The experiments are followed by analytical investigations of crack tip fields in the transonic regime and by detaled numerical studies of the same phenomenon.
A drop weight tower is used to provide high loading rates for the study of the crack initiation process in highly ductile metals. The effect of rate sensitivity on the value of the dynamic J integral (a parameter characterizing the near tip plastic strains at initiation) is investigated at a number of loading rates by means of high speed photography and the optical methods of reflected caustics, Shadow Moire, and the Coherent Gradient Sensor interferometer. The experimental results are compared to three-dimensional, elastic-plastic, numerical simulations of the impact event. Materials tested: HY100, 4340 Steels, Titanium and Aluminum alloys. High speed infrared temperature sensors are used to measure temperature generated during crack initiation. The near crack tip temperature rise is related to the value of the J integral at initiation.
Here we use a high velocity air-gun for the asymmetric "shear" loading of prenotched metal specimens. It has often been observed that low impact velocities (up to 30 m/s) result in the generation of dynamic cracks at the tip of the prenotch. If, however, the impact velocities are increased, the mode of failure switches from that of fracture to that of dynamic (adiabatic) shear banding.
This phenomenon of loading rate dependent failure mode transition is not well understood. Here we use both optical and high speed infrared diagnostics to measure dynamically both deformation and temperature fields at incipient failure. The experiments are also modelled by extensive thermoviscoplastic finite element computations which provide a benchmark for comparison with the experimental measurements. Materials to be tested include steels and titaneun alloys, as well as a variety of metallic glasses synthesized at Caltech. The goal of this study is to provide a measure of the resistance of different materials to the initiation and growth of adiabatic shear bands.
The terminology "Metallic Glasses" or "Amorphous Metals" refers to a class of solids that exhibit a metastable amorphous atomic structure. Metallic glasses can be formed by a process of very rapid quenching of a melt that "freezes" the microstructure and does not allow for the establishment of the classically observed crystalline structure. In the past the requirement of rapid quenching has limited the site of metallic glass specimens and has hampered the potential of these solids for structural applications. However, recent advances in the casting of such solids have made it possible for the first time to produce large enough samples suitable for mechanical testing. The project concentrates on the investigation of the extraordinary quasistatic and dynamic mechanical properties of metallic glasses. The initial investigation concentrates on glass systems involving Zr, Al and Ti.
Measurement of Transient Crack Tip Deformation Fields Using the Method of Coherent Gradient Sensing, Krishnaswamy, S., Tippur, H.V. and Rosakis, A.J., Journal of the Mechanics and Physics of Solids, Vol. 40(2)pp.339, 1992.
Two Optical Techniques Sensitive to Gradients of Optical Path Difference: The Method of Caustics and the Coherent Gradient Sensor (CGS), Rosakis, A.J., to appear in Experimental Techniques in Fracture, Vol. III, (J. Epstein, ed.) 1993.
Temperature Rise at the Tip of Dynamically Propagating Cracks: Measurements using High Speed Infrared Detectors, Zehnder, A.T. and Rosakis, A.J., to appear in Experimental Techniques in Fracture, Vol. III (J. Epstein, ed.) 1993.
Full Measurements at the Dynamic Deformation Field Around a Growing Adiabatic Shear Band at the Tip of a Dynamically Loaded Crack or Notch, Mason, J.J., Rosakis, A.J., and Ravichandran, G., Journal of the Mechanics and Physics of Solids, Vol.42(11), pp. 1697, 1994.
Quasistatic Constitutive Behavior of Zr41.25Ti13.75Ni10Cu12.5Be22.5 Bulk Amorphous Allous, Bruck, H.A., Christman, T., Rosakis, A.J., and Johnson, W.L., Scripta Metallurgica et Materialia, Vol. 30, p. 429, 1994.
The Dynamic Compressive Behavior of Beryllium Bearing Bulk Metallic Glasses, Bruck, H., Rosakis, A.J., and Johnson, W.L., Journal of Materials Research, Vol. 11(2) pp. 503-511, 1996.