Session: 01-06: Shape Memory Alloy

Paper Number: 110692

110692 - Validation of Smanalytics: Comparison of Automatic and Human Analyzed Shape Memory Alloy Test Data 

Over 60 years of shape memory alloy research and development, properties of these alloys have been measured through a variety of testing methods, often customized by the organization performing the test.  Two of these test methods have recently been standardized (ASTM E3097 – Standard Test Method for Mechanical Uniaxial Constant Force Thermal Cycling of SMAs (UCFTC), and ASTM E3098 - Standard Test Method for Mechanical Uniaxial Pre-Strain and Thermal Free Recovery of SMAs (UPFR)), and are used to measure strains and transformation temperatures under constant loads or under free recovery.  In developing the two ASTM standards, ruggedness testing consisting of two replicates of a set of eight tests was performed for both test methods (UPFR and UCFTC).  However, there are no uniform programs or software packages for post processing test data to extract property data. Each organization that performed the testing analyzed their own data using their own custom manual methods to extract the required values for strains, applied and maximum stresses, and transformation temperatures.  At that point, there exists the possibility of variability in the data from: sample geometry and size differences, test setup (including loading, heating, and cooling methods, and capability to attain required rates or holds), precision of post processing technique, and person-to-person variability in processing the data.  To address the last two factors, namely precision of technique and person-to-person variability, and provide a more streamlined analysis approach we have developed a software suite named SMAnalytics to allow automated and uniform analysis of shape memory properties from raw test data.  The suite covers both new ASTM test methods, as well as many modifications to the method, including loading in martensite versus austenite, or performing multiple thermal cycles at stress.

For this study, UPFR and UCFTC ruggedness test data from four different organizations including NASA GRC was collected.  Sample geometry varied from flat dogbone to round dogbone style samples depending on the organization’s test setup and capability.  The data provided included the raw test data as well as the required strain, stress, and transformation temperature values measured by the institution performing the testing.    The results generated by each organization were compared with each other, to look at variability arising from the differences in testing and analysis methods used by each institution.  Then, each organization’s raw data was re-processed using NASA’s automated analysis software without human input to the fitting process for the required measured values. These automatically generated results were compared to those from each institution in order to see how well the results from SMAnalytics compared to those generated by a human, and to help determine where variability arises in the measured values.  The working of the software, and the results of the analysis and comparison to those from each institution will be presented.

Presenting Author: Glen Bigelow NASA Glenn Research Center

Presenting Author Biography: Glen Bigelow is a Materials Research Engineer at NASA Glenn Research Center in Cleveland, Ohio working on developing new shape memory alloys since 2005.