We introduce a novel algorithm – ConvNEAT – that evolves a convolutional neural network (CNN) from a minimal architecture. Convolutional and dense nodes are evolved without restriction to the number of nodes or connections between nodes. The proposed work advances the field with ConvNEAT’s ability to evolve arbitrary minimal architectures with multi-dimensional inputs using GPU processing.
more »
« less
Work Optimization Predicts the Evolution of Extensional Step Overs Within Anisotropic Host Rock: Implications for the San Pablo Bay, CA: Evolution of extensional step overs
- Award ID(s):
- 1219919
- PAR ID:
- 10058010
- Date Published:
- Journal Name:
- Tectonics
- Volume:
- 36
- Issue:
- 11
- ISSN:
- 0278-7407
- Page Range / eLocation ID:
- 2630 to 2646
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Polymer solution processability in extensional-flow dominated operations is strongly influenced by polymer conformation and solution phase behavior. Cosolvent addition can be used to tailor polymer conformation and solution phase behavior to yield formulations that are amenable to processes such as spraying and atomization, coating, and fiber spinning. The addition of N,N-dimethylformamide (DMF) to aqueous poly(N-isopropylacrylamide) (PNIPAM) solutions induces unique phase behavior and microstructure formation, yet the effects on solution processability have remained unexplored. In this work, the effect of DMF cosolvent content on the rheology (both shear and extensional) and microstructure of PNIPAM solutions is investigated. While all examined PNIPAM solutions exhibit nearly Newtonian steady shear behavior regardless of DMF content, the same solutions exhibit varying degrees of extensibility. Surprisingly, the extensional relaxation time increases by more than twenty-fold with increasing DMF content in the water-rich regime. In the DMF-rich regime, however, solution extensibility dramatically decreases. Interestingly, this unique variation in extensional flow behavior does not scale as expected based on changes in the measured intrinsic viscosity and radius of gyration. Instead, a mechanism is proposed that relates the extensional flow behavior to the solution microstructure, which is found to vary with DMF content in light scattering measurements. In the water-rich regime, DMF molecules are proposed to bridge PNIPAM chains via hydrogen bonding and hydrophobic interactions, resulting in physically crosslinked aggregates. In extensional flows, these aggregates behave like a polymer with higher apparent molecular weight, increasing the extensional relaxation time. In the DMF-rich regime, non-bridging DMF molecules increasingly solvate individual PNIPAM chains; consequently, more individual chains are stretched in extensional flows, leading to a reduction in the extensional relaxation time. These findings demonstrate that interactions between components in these ternary systems have unexpected but significant implications in solution extensional flow behavior. Additionally, in the case of PNIPAM/DMF/water, the processability of polymer-containing formulations can be modulated for spraying or for fiber spinning applications just by varying cosolvent (DMF) content.more » « less
An official website of the United States government

