Deep Neural Network a Posteriori Probability Detector for Two-dimensional Magnetic Recording

Primary author: Jinlu Shen
Faculty sponsor: Benjamin Belzer, Krishnamoorthy Sivakumar

Primary college/unit: Voiland College of Engineering and Architecture
Campus: Pullman

Abstract:

The magnetic recording channel in hard disk drives is a binary inter-symbol interference (ISI) channel that typically adopts a linear minimum mean square error (MMSE) equalizer with partial response (PR) signaling followed by a trellis-based detector such as Bahl-Cocke-Jelinek-Raviv (BCJR) or Viterbi. In two-dimensional magnetic recording (TDMR), an array of heads read data from multiple adjacent tracks in order to equalize inter-track interference (ITI), which is severe in high density recording. The multi-track effects combined with pattern-dependent noise inherent to HDD recording channels lead to a trellis state explosion problem, when an auto-regressive model is used for pattern dependent noise prediction (PDNP). The detector complexity grows exponentially with ISI channel length I and noise predictor order L, and becomes impractical for more than two tracks.
As a solution, we propose a novel deep neural network (DNN). The DNN detector replaces the typical Viterbi-PDNP or BCJR-PDNP, directly outputs log likelihood ratios of the coded bits and iteratively exchanges them with a channel decoder to minimize decoded BER. Three DNN architectures are investigated – fully connected DNN, convolutional neural networks (CNN), and long short-term memory (LSTM). The DNN’s complexity is limited by employing MMSE equalizer pre-processing. The best performing DNN architecture, CNN, is selected for iterative decoding with a channel decoder. Simulation results on a realistic media model shows as much as 30.47% detector BER reduction, and as much as 21.72% areal density gain compared to a conventional system.

Additive Manufacturing Using Liquid Metal

Primary author: Steven Peyron
Faculty sponsor: Arda Gozen

Primary college/unit: Voiland College of Engineering and Architecture
Campus: Pullman

Abstract:

Metal 3d printing has played a role in rethinking our manufacturing methods. Using the study of eGaIn and the numerical model of filamentary metal alloys developed by Dr. Gannarapu et al[1] we are going to evaluate the further nonnoble metals and alloys. We will be examining the oxide skin’s effect on the filamentary shape and strength in the subsequent metals and metal alloys. With that information further research on layer interactions of the oxide skin and the thermofluidic flow of the metal alloys and metals at the mesoscale. We have confirmed the layer interactions of eGaIn act like that of a liquid and the oxide skin does not maintain individual layers while liquid. The next step is to print with a metal that is sold at room temperature. We will start with fields alloy and move on to high-temperature metals.

Design and modeling of a microfluidic platform for portable electrochemical analysis

Primary author: Daniel Molina
Faculty sponsor: Cornelius Ivory

Primary college/unit: Voiland College of Engineering and Architecture
Campus: Pullman

Abstract:

A microfluidic platform for electrochemical analysis of flowing solutions was developed, consisting of an acrylic chip and three removable microelectrodes, each housed in a high-resistance plastic tube. The electrodes can be removed independently for cleaning, polishing or replacement. The sensing microelectrode is a 100-µm diameter platinum disk, located flush with the upper face of a 150 µm x 20 µm x 3 cm microchannel, smaller than previously reported for this type of electrodes, and with a total volume of 90 nanoliters, which minimizes the size of the samples required. The platform was evaluated by oxidizing a potassium ferrocyanide solution, a well-known electrochemical probe, at the sensing electrode. The electrical current signal increases with increasing applied potential until it reaches a limiting current. The value of this limiting current increases with the flow rate of the solution, so a better signal/noise ratio can be achieved at higher flow rates.
Numerical models can help us make predictions and serve as design aids without having to iterate physical prototypes. While microdisk channel electrodes have been simulated numerically before using a finite difference method in an ideal 3D geometry, here we predict the limiting current using finite elements in COMSOL Multiphysics®, which allowed us to easily explore variations in the microchannel geometry that have not previously been considered in the literature. Experimental and simulated currents showed the same trend but differed by 41% in simulations of the ideal geometry, which improved when channel and electrode imperfections were included.

Impact identification on concrete panels using a surface-bonded smart piezoelectric module (SPM) system

Primary author: Ayumi Manawadu
Faculty sponsor: Pizhong Qiao

Primary college/unit: Voiland College of Engineering and Architecture
Campus: Pullman

Abstract:

Structural damage assessment after a truck/barge collision is crucial to preserve the integrity of aging concrete bridges, even if there is no apparent damage on the surface. However, given the size of bridges, it would be expensive to analyze the whole structure at once. Therefore, the location and magnitude of the impact should be determined promptly to identify critical areas that require further damage assessment. Such systems help to determine timely corrective action to avoid catastrophic failure. Nevertheless, there is no in-situ cost-effective monitoring technique to carry out this task. Thus, wave-based piezoelectric sensor systems are a promising alternative for real-time impact detection of concrete structures.

Surface-bonded smart piezoelectric modules (SPM) are used to investigate the impact response on concrete panels regarding impact location, impact force, projectile mass, and projectile velocity. Theoretical models based on a spring-mass system and Reed’s model are developed and then validated using numerical and experimental investigations. The main parameters used in this approach are the time of flight and the amplitude of the propagating waves.

The method successfully determined the impact location and magnitude of impact, with an error of 6.40% and 2.73%, respectively. Further, the mass and velocity of the projectile were also successfully computed. Such an evaluation helps to prioritize impact events and to recognize more effective repair techniques. The results demonstrate that the surface-bonded SPMs provide a basis for the development of a cost-effective in-situ real-time non-destructive technique to analyze the impact-response of concrete members.

Three-dimensional(3D) printing conductive material on fabric

Primary author: Dan Liang
Faculty sponsor: Hang Liu

Primary college/unit: Voiland College of Engineering and Architecture
Campus: Pullman

Abstract:

3D printing is an additive manufacture technology, which prints material in a layer by layer mode. Because 3D printing technology has lots of advantages, such as personalized customization, lower parts cost, accurate production, increasing material utilization rate, and accelerating prototyping and manufacturing by eliminating costly model, it becomes a fashionable technology in various industries. In the textile industry, there is great potential to impart advanced functions to traditional textile materials by 3D printing, such as electrical conductivity, sensitivity in heat and chemicals, and shape memory. FDM (fused deposition modeling) is the most used type of 3D printing. However, FDM printed part will increase stiffness and decrease adhesion to textiles. Direct inkjet writing (DIW) has better adhesion and flexibility to fabric than FDM. In this research, the DIW of conductive polymers on three textile fabrics (100% cotton, 100% polyester, and 50%cotton/50% polyester blend) was explored. The resistivity change with tensile stretching was evaluated and compared to FDM. The electricity resistivity change of printed fabrics with abrasion test were measured. The research adopts the PLA, PEO, and PCL as part of the printing material and carbon nanotube as the conductive material. Resistivity measurement, tensile test, abrasion test, FTIR, and TGA are implemented in the research. Overall, FDM printed fabrics had lower resistivity (high conductivity) compared to DIW printed ones. However, the FDM samples are more brittle than DIW. The resistivity of DIW printed fabric is more stable compared to FDM during the tensile test. DIW printed samples performed better during abrasion test.