Automated Post-Processing for Material Characterization Using Matlab
Material characterization is a key component to understanding material behavior, is essential for planning future experiments, and is necessary for obtaining accurate input data for numerical simulations. Creating a comprehensive material profile generates massive quantities of data from several sources (material testing system, Digital Image Correlation, data acquisition device, etc.), which often require significant efforts to post-process. Data navigation and creating essential figures, i.e., depicting the results, can be time consuming. A single Excel file is sufficient for a limited number of specimens. Thorough research, however, requires diverse experimentation, repetitions of each test case, and vast data records. Thus, data storage, demanding calculations, and graphical results can make Excel inefficient, causing tedious document or figure alterations.
In this project, MATLAB coded scripts were developed to interpret data sets, from various load displacement tests, and streamline the post-processing procedure. The scripts were initially written for uniaxial testing data using versatile function commands rather than hard coding equations into the script. This flexibility allows for future developments that will include analyzing microtube tension-inflation and biaxial cruciform tests.
Two main scripts, implementing unique project functions, were created to encapsulate the program goals. The presenter also integrated individual functions from MATLAB File Exchange, utilizing experienced programmer knowledge to slightly alter existing MATLAB functions for the program’s specific goals.
The first script focuses on a single specimen analysis compiling all required calculations to fully understand the experimental results. The beginning of this script contains a rudimentary menu consisting of a list of variables, which allow the user to dictate the desired outputs. The menu’s options dictate relevant sections of the code and allows the program to bypass large portions of the script, shortening the processing time. The script then opens raw data by asking the user to select all necessary files in a program generated window. A combination of for loops, Boolean statements, and string-matching functions allow MATLAB to distinguish the selected file types, find the required data within each file, and extract it using dynamic field generation within structures, depending on which files were initially opened. It then calculates several vital material analysis results including: Young’s Modulus, engineering and true stress-strain curves, as well as Lankford Coefficients (anisotropic R-values). The single specimen analysis also utilizes smoothing functions to generate a plot of the noisy, instantaneous R-value data overlaid with a clean, curved trendline. Finally, this script saves all analytical results into an Excel file. This specimen summary document has set floating numbers in the sheet’s cells, reducing Excel’s processing requirements by excluding integrated formulas. This allows users to create separate figures in Excel, if desired, instead of using the program’s built-in graphing capabilities. The script circulates through the graph menu to plot, save, and export the user requested figures. These graphs are saved as JPEGs and MATLAB figure files in organized program output folders. Initial script development was validated against SS316L uniaxial tensile test, results that were previously post-processed in Excel.
The second script gathers inputs by prompting the user to select summary Excel workbooks, generated by the first script. The summary workbooks are used to combine experiment results, i.e., repetitions of experiments or a representative sample of experiments, into single figure outputs efficiently. The combining script also allows users to calculate stress ratios by comparing data between specimens.
The aid of programing knowledge reduced the post-processing time for a single specimen from an hour to a matter of minutes. The scripts will be reused in future research by New Hampshire BioMade, a $20M/5-year NSF EPSCoR award, when CP-Ti is subjected to similar tests during the next phase of the project. The code’s flexible design accommodates multiple geometries, machine outputs, and materials while prioritizing minimal user inputs per test analysis and requiring fewer computing resources than previous post-processing methods. Overall, the scripts assist researchers in continuing experimentation more efficiently and can be expanded for use in other scientific and engineering applications.
Automated Post-Processing for Material Characterization Using Matlab
Category
Poster Presentation
Description
Session: 16-01-01 National Science Foundation Posters - On Demand
ASME Paper Number: IMECE2020-24948
Session Start Time: ,
Presenting Author: Marguerite Kennish
Presenting Author Bio: Marguerite Kennish is a non-traditional student. She returned to school to further her personal development and push towards new career goals. As an older undergraduate adjustment to the university classroom was at first challenging but has become a place for her to flourish.
Marguerite is a rising Junior at the University of New Hampshire, pursing her bachelor’s degree in Mechanical Engineering. She currently attends as a New Hampshire BioMade transfer scholar. Over the past two summers Marguerite has also worked as an undergraduate researcher for NH BioMade at the John Olson Manufacturing Center, assisting material testing, and MATLAB coding to aid in post processing collected data.
Marguerite is also a Great Bay Community College Alumni with an associate degree in Engineering Science. She graduated with highest honors, and recipient of the school’s award for outstanding achievement in Engineering Science.
Marguerite’s is a bright student, and she is excited to participate in future opportunities.
Authors: Marguerite Kennish John Olson Manufacturing Center
Elizabeth Mamros University of New Hampshire