Dupler, Ellen2019-09-172019-09-172019-07https://hdl.handle.net/11299/206705University of Minnesota M.S.E.E. thesis. July 2019. Major: Electrical/Computer Engineering. Advisors: Lucy Dunne, Sarah Swisher. 1 computer file (PDF); vii, 155 pages.Textile-based strain sensors are first defined with examples of various sensing mechanisms and applications, focusing on on-body smart garments for biomonitoring. A current lack of research in the textile substrate influence on sensor performance is noted, with a thesis investigation outlined to highlight key variables that may be important for successful sensor design. Two conductive thread stitch-based strain sensors are chosen for the textile-based strain sensors and two fabric substrates (2-way and 4-way stretch) are used to investigate their influence on sensor performance. Part 1 investigates if fabric strain properties change due to the attachment of sensors and how the sensor performance changes due to fabric choice and attachment angle. Part 2 uses the recommendations for textile choice, stitch geometry of the sensor, and sensor placement based on Part 1 results to create a 3-sensor, 60° strain rosette. Between the two versions of rosettes fabricated, the 4-way fabric and chainstitch geometry, the strain rosette is proven to improve the overall sensor performance in predicting force, displacement, and force direction. This rosette is characterized and using machine learning model algorithms, model-fitted for future garment based strain sensing applications.ene-textilesmart clothingstitched sensorstrain sensorwearable sensorwearable technologyThesis or Dissertation