Multiscale Modeling of Concrete Failure under Blast and Fragment Loading using a Coupled LDPM-RKPM Method

While single scale models (either discrete or continuum) provide useful insight into the physics of concrete fragmentation process they become intractable for practical structural application in particular for Very High Strength Concrete (VHSC) having small aggregate size and Fiber Reinforced Concrete (FRC) where resolving the details of fiber-concrete interactions is critical to assess the structural performance under extreme conditions. To successfully address the complexities of predicting the performance of VHSC and FRC, Prof. Fish and Dr. Yuan from MDS, LLC with expertise and record of accomplishments in developing integrated multiscale design software for heterogeneous material systems have teamed up with Prof. JS Chen from UCLA who has extensive experience in fracture and fragmentation of concrete structures subjected to blast and fragment loading. In Phase I, we will: 1. Assess existing computational models including LDPM, RKPM and reduced order homogenization models for simulating blast and fragmentation loading on VHSC, FRC, and other types of special concrete materials and cementitious composites; 2. Identify shortcoming of existing methodologies and formulate predictive multiscale strategy for future development based on combination of LDPM and RKPM in which LDPM provides the lower order approximation of cement-aggregate behavior, while RKPM is introduced as an enrichment of LDPM to account for the effects of small aggregate and fiber reinforcement. 3. Perform preliminary verification and validation studies that compare the feasibility of the proposed multiscale methodology including: (a) calibration against test data in collaboration with DTRA and ERDC and (b) initial demonstrations on out-of-plane compression tests and low velocity impact tests. In Phase II, the LDPM-RKPM approach will be combined with the computational continua methodology to account for dynamic effects in RVE problem and the eigendeformation approach to maximize computational efficiency by constructing residual free fields.