Search

We propose Neural Cellular Automata (NCA) to simulate the microstructure development during the solidification process in metals. Based on convolutional neural networks, NCA can learn essential solidification features, such as preferred growth direction and competitive grain growth, and are up to six orders of magnitude faster than the conventional Cellular Automata (CA). Notably, NCA delivers reliable predictions also outside their training range, which indicates that they learn the physics of the solidification process. While in this study we employ data produced by CA for training, NCA can be trained based on any microstructural simulation data, e.g. from phase-field models.

The CrMnFeCoNi high-entropy alloy is highly printable and holds great potential for structural applications. However, no significant discussions on cyclic plasticity and fatigue damage in previous studies. This study provides significant insights into the link between print processes, solidification microstructure, cyclic plasticity and fatigue damage evolution in the alloy fabricated by laser powder bed fusion. Thermodynamics-based predictions (validated by scanning transmission electron microscopy (STEM) energy dispersive X-ray spectroscopy (EDX)) showed that Cr, Co and Fe partition to the core of the solidification cells, whilst Mn and Ni to the cell boundaries in all considered print parameters. Both dislocation slip and deformation twinning were found to be responsible for plastic deformation under monotonic loading. However, the former was found to be the single dominant mechanism for cyclic plasticity. The surface finish helped to substantially delay the crack initiation and cause lack-of-fusion porosity to be the main source of crack initiation. Most significantly, the scan strategies significantly affect grain arrangements and grain dimensions, leading to noticeable effects on fatigue crack propagation; in particular, the highest resistance crack propagation was seen in the meander scan strategy with 0{\deg} rotation thanks to the most columnar grains and the smallest spacing of grain boundaries along the crack propagation path., Comment: 24 pages, 16 figures

Polynorbornene-based decomposable polymer which can be patterned with ultraviolet or electron-beam radiation is used to create micrometer-scale fluidic channels. Silicon nitride substrates are used to fabricate nanophotonic wavegide and resonators operating in the visible range of the spectrum. Fluidic channels generated by thermally decomposing the polymer through the oxide cladding is used to deliver ultra-small amounts of florescent samples to the optical sensors.

Silicon photonics has emerged as the leading candidate for implementing ultralow power wavelength division multiplexed communication networks in high-performance computers, yet current components (lasers, modulators, filters, and detectors) consume too much power for the femtojouleclass links that will ultimately be required. Here, we propose, demonstrate, and characterize the first modulator to achieve simultaneous high-speed (25-Gb/s), low voltage (0.5VPP) and efficient 1-fJ/bit error-free operation while maintaining athermal operation. Both the low energy and athermal operation were enabled by a record free-carrier accumulation/depletion response obtained in a vertical p-n junction device that at 250-pm/V (30-GHz/V) is up to ten times larger than prior demonstrations. Over a 7.5{\deg}C temperature range, the massive electro-optic response was used to compensate for thermal drift without increasing energy consumption and over a 10{\deg}C temperature range, increasing energy consumption by only 2-fJ/bit. The results represent a new paradigm in modulator development, one where thermal compensation is achieved electro-optically., Comment: 23 pages, 5 figures

In this paper, we propose a complete synchronization algorithm for continuous phase modulation (CPM) signals in burst-mode transmission over additive white Gaussian noise (AWGN) channels. The timing and carrier recovery are performed through a data-aided (DA) maximum likelihood algorithm, which jointly estimates symbol timing, carrier phase, and frequency offsets based on an optimized synchronization preamble. Our algorithm estimates the frequency offset via a one dimensional grid search, after which symbol timing and carrier phase are computed via simple closed-form expressions. The mean-square error (MSE) of the algorithm's estimates reveals that it performs very close to the theoretical Cram\'er-Rao bound (CRB) for various CPMs at signal-to-noise ratios (SNRs) as low as 0 dB. Furthermore, we present a frame synchronization algorithm that detects the arrival of bursts and estimates the start-of-signal. We simulate the performance of the frame synchronization algorithm along with the timing and carrier recovery algorithm. The bit error rate results demonstrate near ideal synchronization performance for low SNRs and short preambles., Comment: Copyright 2013 IEEE

Solving free riding and selecting a reliable service provider in P2P networks has been separately investigated in last few years. Using trust has shown to be one of the best ways of solving these problems. But using this approach to simultaneously deal with both problems makes it impossible for newcomers to join the network and the expansion of network is prevented. In this paper we used the game theory to model the behavior of peers and developed a mechanism in which free riding and providing bad service are dominated strategies for peers. At the same time newcomers can participate and are encouraged to be active in the network. The proposed model has been simulated and the results showed that the trust value of free riders and bad service providers converge to a finite value and trust of peers who provide good service is monotonically increased despite the time they join the network., Comment: 7 pages IEEE format, International Journal of Computer Science and Information Security, IJCSIS November 2009, ISSN 1947 5500, http://sites.google.com/site/ijcsis/