Rodrigo Sarlo

Spectral Analysis and Characterization of a Membrane-Based Artificial Hair Cell Sensor

A fully hydrogel-supported, artificial hair cell (AHC) sensor based on bilayer membrane mechanotransduction is designed with sensitivity and versatility in mind. Thanks to fabrication improvements from previous generations, the sensor demonstrates peak current outputs in the nanoamp range and can clearly measure inputs as high as 2k Hz. Characterization of the AHC response to base excitation and air pulses show that membrane current oscillates with the first three bending modes of the hair. Output magnitudes reflect of vibrations near the base of the hair. A 2 DOF Rayleigh-Ritz approximation of the system dynamics yields estimates of 19 N/m and 0.0011 Nm/rad for the equivalent linear and torsional stiffness of the hair’s hydrogel base, although double modes suggest non-symmetry in the gel’s linear stiffness. The sensor output scales linearly with applied voltage (1.79 pA/V), avoiding a higher-order dependence on electrowetting effects. The free vibration amplitude of the sensor also increases in a linear fashion with applied airflow pressure (18.4 pA/psi). Based on these sensitivity characteristics, an array sensing strategy for these sensors is proposed.Copyright

Optimal Parameter Identification for Model Correlation Using Model Reduction Methods

Classically, to achieve correlation between a dynamic test and a Finite Element Model (FEM), an experienced engineer chooses a small subset of input parameters and uses a model updating technique or engineering judgment to update the parameters until the error between the FEM and the test article is acceptable. To reduce the intricacy and difficulty of model correlation, model reduction methods such as the Discrete Empirical Interpolation Method (DEIM), and dime are implemented to reduce the scale of the problem by reducing the number of FEM parameters to its most critical ones. These model reduction methods serve to identify the critical parameters required to develop an accurate model with reduced engineering effort and computational resources. The insight gained using these methods is critical to develop an optimal, reduced parameter set that provides high correlation with minimal iterative costs. This can be seen as a particular approach to sensitivity analysis in the model updating community. The parameter set rankings derived from each method are evaluated by correlating each parameter set on five simulated test geometries. The methodology presented highlights the most valuable parameters for correlation, enabling a straightforward and computationally efficient model correlation approach.

Directional Sensitivity Analysis of a Hydrogel-Supported Artificial Hair Cell

An artificial hair cell sensor imitates the function of cilia in natural hair cells in order to detect surrounding fluid displacement. Here, a novel structure for creating artificial hair cell sensors uses established methods of creating lipid bilayers at the interfaces of millimeter scale hydrogel shapes. This paper describes the fabrication of the sensor components and the manner in which they are assembled and tested. The hair’s vibration can be detected by monitoring changes in the current produced by mechanical fluctuations in the bilayer. The cross-sectional geometry of the hair can be changed to enable directional sensitivity. Spectral analysis of the sensor current response indicates that frequencies and magnitudes change when a flattened hair is excited in different directions. Finally, the sensor is shown to become more sensitive with applied potential across the bilayer. Results agree with similar studies on this phenomenon.© 2013 ASME

Modal Parameter Uncertainty Estimates as a Tool for Automated Operational Modal Analysis: Applications to a Smart Building

The knowledge of modal parameter uncertainties derived from operational modal analysis (OMA) can greatly improve automated decisions by providing information about the quality of the modal identification. Yet so far, this information has been largely ignored in continuous monitoring studies on civil infrastructure, especially with respect to buildings. In this paper, we implement an automated version of Covariance Based Stochastic Subspace Identification on a highly instrumented smart building. An expansion of the technique estimates uncertainty bounds for all modal parameters. Through a series of full scale experiments, we demonstrate how uncertainties are valuable tools in various contexts of automation. These include the identification and removal of badly-fitted modes, the identification of periods of high signal-to-noise ratio, and the validation of reference sensors selection.

Improved model correlation through optimal parameter ranking using model reduction algorithms: Augmenting engineering judgment

As the complexity and scales of dynamic models increase, novel and efficient model correlation methodologies are vital to the development of accurate models. Classically, to correlate a Finite Element Model (FEM) such that it matches a dynamic test, an experienced engineer chooses a small subset of input parameters that are surmised to be crucial, sensitive and/or possibly erroneous. The operator will then use engineering judgment, or a model updating technique to update the selected subset of parameters until the error between the FEM and the test article is reduced to within a set bound. To reduce the intricacy and difficulty of model correlation, a methodology is proposed to provide a quantitative parameter importance ranking using a model reduction algorithm applied to a parameter sensitivity analysis. Four model reduction algorithms are studied in this effort, the Discrete Empirical Interpolation Method (SVD-DEIM), Q-DEIM, Projection Coefficient and finally Weighted Projection Coefficient. These model reduction methods identify and rank critical parameters, enabling the selection of a minimum set of critical correlation parameters. This reduced set of parameters results in reduced computational resources and engineering effort required to generate a correlated model. The insight gained using these methods is essential in developing an optimal, reduced parameter set that provides high correlation capability with minimal iterative costs. To evaluate the proposed parameter selection methodology, a representative set of academic and industry experts provided their engineering judgment for comparison with the methodology presented. A comprehensive investigation of the robustness of this methodology is performed on a simple cantilever beam for demonstration. The scale of the model has expressly been chosen to allow for all potential ranking variations to be evaluated so that these ranking methods can be understood relative to the true optimal ranking. The ranking robustness to incorrect engineering judgment, resulting in uncertainty in the assumed size of the design space and, therefore, the error bounds, is investigated. The methodology presented identifies the most useful parameters for correlation, enabling a straightforward and computationally efficient model correlation approach as compared with other methods. To quantify the ranking quality, a metric, the Correlation Norm Error, is developed. For the problem discussed, blind random assessments result in a Correlation Norm Error of 413.3. Engineering judgment has been shown to improve upon blind random assessments, reducing the Correlation Norm Error to 334.3. The best performing model reduction method, Q-DEIM using 10 FEM runs as the input, was able to identify the optimal ranking correctly, reducing the Correlation Norm Error to zero.

Measuring Violin Bow Force During Performance

Violin bowing is a complex skill and controls a majority of the sound produced by the instrument. Yet despite significant interest in the modal analysis of violins, comparatively little work has been done to study the complexities of this “input.” In this work, we have used fiber optic strain sensors, a modern strain sensing technology, to test a novel method for measuring bowing force during violin performance. Gaining greater insight into how violinists vary bowing force to create sound could not only lead to better violin excitation methods for modal analysis but might also key in discovering indicators of violin quality and musician preferences. Live performance testing was performed by a professional violinist on two violins of differing quality and at different volume levels. The results showed a log-linear relation between bowing force and volume. In addition, a bowing gesture named the “average down-up stroke” was found by averaging several similar gestures. Its duration was observably longer for a high quality violin compared to a fair quality sibling. Such a measure could be adapted for various gestures and styles and subsequently be explored as a potential indicator of violin quality or player preference.

A Neural Network Approach to 3D Printed Surrogate Systems

The geometry of a Stradivarius violin was recently replicated through additive manufacturing, but nevertheless failed to produce a tone of professional quality. Due to the material limits of additive manufacturing, it is clear that purely geometric replicas are unlikely to create violins of comparable sound. We propose a “surrogate” system approach, which tailors some combination of a structure’s material and geometric properties to mimic the performance of a target system. Finite element (FE) methods can approximate the vibrational performance of a violin or similar structure with high precision, given its specific physical properties and geometry. Surrogate systems, however, require the solution of the inverse problem. This can be achieved through artificial neural networks (ANN), a powerful tool for non-linear function estimation. As a stepping-stone to the violin problem, we first developed a surrogate method for simple beam structures. A neural network was trained on 7500 randomized beams to predict a thickness profile for a set of desired mode shapes and frequencies. Numerical simulation shows surrogates with good performance (<8 % modal error, <18 % frequency error) for target structures with a similar degree of thickness variation to that used in training the neural network. Performance improves dramatically (<2 % modal error, <7 % frequency error) for slightly less complex target structures.

Airflow Sensing With Arrays of Hydrogel Supported Artificial Hair Cells

The hair cell is a biological sensor that uses microscopic hair-like structures to detect delicate motions of surrounding fluid. Inspired by this principle, we have created an artificial hair cell (AHC) sensory method based on biomolecular transduction for sensing spatial variations in air flow. The key feature of this method is the use of one-dimensional arrays built from modular AHC units which measure local velocity at different points in a flow profile. Each of the AHC units uses thinly extruded glass fibers as mechanical receptors of air velocity. Hair vibrations are converted to current via hydrogel-supported lipid bilayer membranes through their mechanocapacitive properties. Preliminary tests with linear arrays of three AHC units attempt to measure the air source profile with varying position and intensity. Each unit was fabricated with a hair of different length, giving it a unique vibrational response. This technique was inspired by how organisms use hair cells with tuned responses to mechanically process flow stimuli. A significant challenge in processing the sensors’ output was the limitation of one input channel on the current measurement unit, thus each sensor output had to be sent over the same channel. When several AHC units are excited simultaneously by an airflow, the resulting signal is a superposition of each sensor’s individual response. To separate the signals back into their individual measurements, the Hair Frequency Response Decomposition method is developed, which maps the spectral content of a combined output to the location of excitation in the array. This method takes advantage of the AHC’s high signal-to-noise ratio (compared to other membrane-based AHCs) and linear output response to flow velocity. Results show that the bilayers’ consistent spectral responses allow for an accurate localization of sensor excitation within the array. However, temporal variations in bilayer size affect sensitivity properties and make accurate flow velocity estimation difficult. Nevertheless, under stable bilayer conditions the measured velocity profiles matched closely with theoretical predictions. The implementation of the array sensing method demonstrates the sensory capability of bilayer-based AHC arrays, but highlights the difficulties of achieving consistent performance with bio-molecular materials.Copyright