Non-Contact Ultrasound Setup
As an undergraduate researcher in the URSA program (Undergraduate Research for Scientific Advancement), I worked with the Roesler Group to engineer and iteratively refine a prototype housing for an ultrasound-based sensing device used to detect Leaky-Rayleigh waves in fresh concrete.
For context, prior research in the lab established that the measured amplitude of Leaky-Rayleigh waves is linked to fresh concrete set time; analyzing these waves would thus enable us to make inferences about a concrete sample’s properties, such as set time, without physically contacting the sample. The issue was that the existing test setup was essentially a large, stationary collection of exposed wiring, loosely integrated sensors, and redundant support components that was difficult to move and impractical for field use.
My role was to consolidate that laboratory setup into a compact, self-contained, and transportable prototype that could preserve the required sensor geometry while being much easier to assemble, move, and eventually deploy outside the lab.
Over multiple design cycles, I designed and fabricated iterations of modular component housings, including those for an ultrasonic speaker, 6x parallel MEMS microphones, an NI-USB 63636 data-acquisition card, and relevant internal wiring. I used Autodesk Fusion 360 assembly software to design individual components and used FDM 3D Printing with CURA to rapid prorotype and test designs. Some notable considerations across these iterations are listed below:
-
TOLERANCING: I designed test-tolerance components to account for the tolerancing of various joinery methods (i.e., for dowel pins and heat-threaded inserts) to ensure components were rigid with respect to each other.
-
CAD: the primary design considerations regarded the generation of CAD component housings such that they would be easy to assemble, easy to manufacture via 3D printing (i.e., consider supports and positioning on the print bed), and easy to deploy in the field over a sample of concrete that was freshly laid. Components also had geometric constraints with respect to each other; i.e., research had established that the speaker must have 12-degree incidence and the MEMS be suspended 5mm over a concrete surface. Each subsequent iteration optimized these considerations.
-
VIBRATIONS & ACOUSTICS: I experimented with the effect of various acoustic buffers (neoprene, EVA foam) to dampen unwanted vibrations in the stucture; this was important due to the sensitive nature of the MEMS microphones. I selected PVA in rapid prototype fabrication as it propogated the least vibrations out of available print material.
My CAD design and prototype housing earned me 1st Place (Best Overall Poster and Presentation) in the URSA Undergraduate Research Synposium (Spring 2025). Following this, I was offered a paid position with the Roesler Group to continue this work with an emphasis on improving the device's data acquisition pipeline.
The system processed 100 cycles of 16-pulse measurements over a 400-microsecond time-averaging window, and repeated this to supply real-time data to a computer. I improved the algorithm by first analyzing from an electrical standpoint, re-soldering all electrical connections, removing leftover redundant wiring from research. and shielding all connections with EMI tape. I improved the algorithm by first implementing a Savitsky Golay filter to reduce vibrational interference, and then using the NIDAQmx, pandas, and SymPy Python libraries to automate data-collection to organized excel sheets for subsequent analysis and storage. Across this process, I troubleshot both any electro-mechanical issues and software issues, with the help of NI-aided software like NI MAX.
