Anthony D. Santamaria

Machine learning: A crucial tool for sensor design

Sensors have been widely used for disease diagnosis, environmental quality monitoring, food quality control, industrial process analysis and control, and other related fields. As a key tool for sensor data analysis, machine learning is becoming a core part of novel sensor design. Dividing a complete machine learning process into three steps: data pre-treatment, feature extraction and dimension reduction, and system modeling, this paper provides a review of the methods that are widely used for each step. For each method, the principles and the key issues that affect modeling results are discussed. After reviewing the potential problems in machine learning processes, this paper gives a summary of current algorithms in this field and provides some feasible directions for future studies.

Liquid Water Interactions With PEFC Gas-Diffusion Layers: The Effect of Vibration

Efficient removal of liquid-water from gas-diffusion layer (GDL) media in polymer-electrolyte fuel cells (PEFC) is critical to achieving reliable, high power operation. Studies have shown that PEFC systems exposed to vibration suffer from performance loss; however, fundamental understanding as to why is lacking. This work investigates vibrations ranging from 1–100 Hz and their effect on wetting behavior as well as liquid adhesion to the GDL surfaces. Vibrations are found to redistribute water in a way that raises its barrier to removal. An increased frequency reduced droplet contact area and height while expanding the wetting diameter. Moreover, while vibrations did not change the adhesion force, the increased wetting diameter resulted in a larger net force required to achieve droplet detachment.Copyright

Characterizing nitinol wire bond strength in silicone for artificial skin muscles

This study examines the bonding force between Nitinol wire and silicone polymer. The purpose is to understand the feasibility and limitations of using nitinol wire as an actuator in artificial skin. This study improves upon previous work, which studied silicone embedded nitinol and expands its scope by looking deeper into the wire/polymer configuration, and manufacturing process which can lead to air bubbles at the wire/silicone interface. Prior results were less consistent due to the presence of tiny bubbles in samples, which lead to variable maximum pulling forces between tests. This study addresses this issue, and improves the manufacturing process, so that the capacity of the bonding between wires and polymers can be increased.This is accomplished by using an injection molding method preceded by a vacuum stage.When samples are manufactured and tested with the improved method there is a significant improvement in the strength and consistency of results. A maximum pull force improvement of 32% was seen in the vacuum prepared samples. This lays the foundation for developing computer simulations of the artificial skin using experimentally verified data. Future work will continue to address the manufacturing process, material variants, as well as checking the effect of different wire diameters and materials. All this data will go into developing a predictive numerical model using commercial finite element analysis software, which will assist in the creation of more complex shapes of controllable artificial skin. These complex wire/polymer configurations will be used to address current biomedical issues, such as facial paralysis, and assisting burn victims, or in humanoid robotics.

Role of GDL Surface Wettability and Operating Conditions in Liquid-Water Removal From NSTF Catalyst Layers

Proton-exchange-membrane fuel cells (PEMFCs) are becoming the center of attention as an alternative power source for automotive and stationary applications due to their ability to produce high power densities under rapid load changes. Even though a significant amount of research on PEMFCs has been conducted over the last few decades, the cost associated with platinum (Pt) catalyst remains a barrier for their commercialization. For instance, the US Department of Energy has a target to reduce the use of Pt in PEMFCs to 0.125 mg/cm2 or less, while maintaining rated stack power densities. A path to achieve this target is the use of thin catalyst-layers (CLs), such as nanostructured thin-film (NSTF) CLs [1]. Figure 1 shows a SEM image of NSTF CL showing the catalyst coated whiskers on a microstructured substrate after sputter coating with a PtCoMn alloy. The lengths of the whiskers are in the order of a few microns; however, the thickness of an NSTF CL inside a PEMFC sandwich is often in the range of 500 nm. Due to its submicron thickness, NSTF CL has low liquid-water capacity which can make low temperature operation a challenge. At low temperatures, NSTF CLs experience severe water flooding [2]. Unfortunately, sometimes low-temperature operation is unavoidable and for rapid startup relatively high current densities may be necessary. Therefore, a successful water balance is required to avoid water-flooding in the NSTF CLs.© 2014 ASME