Numerical and experimental evaluation of dynamic shear response of a high entropy alloy
Vallejo Rodríguez, Luis (2021)
Vallejo Rodríguez, Luis
2021
Materiaalitekniikan DI-ohjelma - Master's Programme in Materials Engineering
Tekniikan ja luonnontieteiden tiedekunta - Faculty of Engineering and Natural Sciences
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2021-12-14
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-202112089012
https://urn.fi/URN:NBN:fi:tuni-202112089012
Tiivistelmä
In this thesis two multi-principal element alloys, also known as high-entropy alloys (HEAs), were studied to evaluate their shear responses under quasi-static and dynamic loading conditions by means of geometries designed with the purpose of promoting shear localization at the gage section. Cantor and Al0.3CrFeCoNi HEAs can replace the alloys used nowadays by mining, construction, railroads, automobile and defense industries. A common phenomenon on materials used in these applications is shear localization and adiabatic shear band (ASB) generation, which has been scarcely studied; however, only shear specimen geometries that did not allow the direct observation of shear localization have been used.
With the aim of speeding up the study of shear geometries and designing of new ones, a first simulation approach with the Finite Element Methods tool Abaqus is done based on the manufacturing conditions of the HEAs (2 mm thick Cantor plate and 8 mm Al0.3CrFeCoNi plate) and the available equipment for the characterization the material response (Split Hopkinson Pressure Bars, Split Hopkinson Tensile Bars and 2D-DIC). The designs did not include the input and output bars from the SHB set-ups and afterwards, they were adjusted to quasi-static loading conditions and modified to optimize shear localization at the gage section. The resulting specimen geometries are variations of the S-shaped specimen with a thin gage section (Cantor HEA, tested in SHTB), and the flat-hat specimen (Al0.3CrFeCoNi HEA, tested in SHPB); five specimens of each design were manufactured and tested in quasi-static and dynamic loading conditions, with shear strain rates ranging from 0.0025 s-1 to 11000 s-1 for the Al0.3CrFeCoNi specimens and from 0.001 s-1 to 8500 s-1 for the Cantor specimens. The data from the experiments was processed with Matlab. The tests were coupled with 2D-Digital Image Correlation tool for analyzing strain field evolution in the gage section while the experiments were running to compare the strain fields obtained from there with those from the simulations. The data from the DIC measurements was processed with LaVision software.
In both DIC and Abaqus, the design used for testing the Cantor HEA concentrated the strains at the gage section, uniformly throughout its thickness and forming a shear band; however, as the yield stress does not increase with the strain rate in dynamic loading conditions, more tests are required. The design used for testing the Al0.3CrFeCoNi HEA concentrates most of the strains at the gage section also, but the strains also spread out of the gage section and are not uniformly distributed throughout the specimen thickness because the gage section was not milled down as in the case of the Cantor specimen design. Furthermore, the material response of Al0.3CrFeCoNi HEA was not in concordance with that from previous investigations, except for the yield stress obtained at low strain rate, which was considerably higher, decreasing considerably the strainrate sensitivity of the alloy. Regarding the shear stress – shear strain curves from the model, using fitting models such as Johnson-Cook model, studying the mesh size sensitivity and including fracture modes in the simulations would be recommended for better agreement with the experimental curves.
With the aim of speeding up the study of shear geometries and designing of new ones, a first simulation approach with the Finite Element Methods tool Abaqus is done based on the manufacturing conditions of the HEAs (2 mm thick Cantor plate and 8 mm Al0.3CrFeCoNi plate) and the available equipment for the characterization the material response (Split Hopkinson Pressure Bars, Split Hopkinson Tensile Bars and 2D-DIC). The designs did not include the input and output bars from the SHB set-ups and afterwards, they were adjusted to quasi-static loading conditions and modified to optimize shear localization at the gage section. The resulting specimen geometries are variations of the S-shaped specimen with a thin gage section (Cantor HEA, tested in SHTB), and the flat-hat specimen (Al0.3CrFeCoNi HEA, tested in SHPB); five specimens of each design were manufactured and tested in quasi-static and dynamic loading conditions, with shear strain rates ranging from 0.0025 s-1 to 11000 s-1 for the Al0.3CrFeCoNi specimens and from 0.001 s-1 to 8500 s-1 for the Cantor specimens. The data from the experiments was processed with Matlab. The tests were coupled with 2D-Digital Image Correlation tool for analyzing strain field evolution in the gage section while the experiments were running to compare the strain fields obtained from there with those from the simulations. The data from the DIC measurements was processed with LaVision software.
In both DIC and Abaqus, the design used for testing the Cantor HEA concentrated the strains at the gage section, uniformly throughout its thickness and forming a shear band; however, as the yield stress does not increase with the strain rate in dynamic loading conditions, more tests are required. The design used for testing the Al0.3CrFeCoNi HEA concentrates most of the strains at the gage section also, but the strains also spread out of the gage section and are not uniformly distributed throughout the specimen thickness because the gage section was not milled down as in the case of the Cantor specimen design. Furthermore, the material response of Al0.3CrFeCoNi HEA was not in concordance with that from previous investigations, except for the yield stress obtained at low strain rate, which was considerably higher, decreasing considerably the strainrate sensitivity of the alloy. Regarding the shear stress – shear strain curves from the model, using fitting models such as Johnson-Cook model, studying the mesh size sensitivity and including fracture modes in the simulations would be recommended for better agreement with the experimental curves.