Dynamic Failure Modeling

Home > Research > Dynamic Failure of Materials > Dynamic Failure Modeling

The group has developed custom software accounting for multi-body finite kinematics contact, finite deformation plasticity, temperature effects, fragmentation and comminution. An example is the development of ceramic models based on grain level representative volume elements (RVE) of ceramic microstructures. The model was employed to interpret both normal impact soft-recovery and pressure-shear soft recovery experimental results. The numerical simulations were based on a 2-D stochastic finite element analysis. Normal plate impact velocity histories obtained in earlier studies were used to assess conditions under which the cohesive fracture model could capture failure mechanisms experimentally observed. The analyses showed that in order to properly model damage kinetics a stochastic distribution of grain boundary strength and detailed modeling of grain morphology are required (Zavattieri and Espinosa, Acta Mat., 2001). Moreover, it was determined that compressive wave attenuation at stress levels below the Hugoniot elastic limit, a counterintuitive finding that preoccupied the ceramic community in the early 90’s, was the result of grain boundary relaxation in shear due to the presence of a glassy phase. In these simulations, compression-shear properties independently identified for glass were employed in the cohesive law describing the grain boundary constitutive law. Overall compression wave decay and nucleation of microcracks at triple grain junctions naturally emerged from the simulations. These studies were significant because the simulations were compared to experimental data containing information on crack initiation and kinetics as observed in plate impact velocity histories and electron microscopy studies performed on recovered samples.

Figure 1: Schematics of microcracking at grain boundaries using an irreversible interface cohesive law (left). Application to fragmentation and pulverization (right).

 

Some of the group’s early research focused on continuum/discrete models for the high-strain rate response of advanced materials (including brittle and composite materials). This included the development and implementation of models and numerical algorithms in finite deformation FEM codes using parallel programming. The models included: i) Adaptive remeshing techniques based on the optimization of element size and shape (including refinement and coarsening) with mapping of state variables within a finite deformation framework, ii) Continuum/Discrete models, based on fracture and damage models together with a multibody contact-interface methodology to capture crack initiation, growth, coalescence and interaction between fragments, and iii) the combination of both adaptive remeshing and the continuum/discrete model to capture delamination and fracture in fiber reinforced laminate composites.

Figure 2: Computational techniques for mesoscopic modeling of failure: (a) Examples of adaptive remeshing technique based on optimization of element size and shape according to local material behavior. Examples include impact problems of rod penetration, high-speed machining and ballistic penetration. (b) Continuum/discrete models for fragmentation in brittle materials, (c) Combination of (a) and (b) for delamination of glass reinforced composites under ballistic impact.

 

Personnel 

  • Horacio D. Espinosa (PI)

 

Collaborators 

  • P. Zavattieri, Purdue University, IN, USA

 

Publications 

 

 

Robert R. McCormick School of Engineering and Applied Science
McCormick Home | Northwestern Home | Northwestern Calendar | Accessibility | Contact | Emergency Plan | Maps
© Robert R. McCormick School of Engineering and Applied Science, Northwestern University
2145 Sheridan Rd., Evanston, IL 60208-3100 | Phone: (847) 491-5220 | Fax: (847) 491-8539
Email: webmaster@northwestern.edu | Last modified: | Legal and Policy Statements