A Study On Ultra-small Specimen Mechanical Test Methods

Small specimen test methods are widely used for the cases that the material volume is limited and thus cannot satisfy the requirement for preparing standard specimen.In recent year,with the development of measurement techniques and promotion of test requirements,specimen dimension has reduced to a more miniature scale that is known as ultra-small specimen.Compared with the well-studied miniature test methods with“larger” small specimens,ultra-small specimen test techniques are confronted with more serious challenge.Therefore,it is needed to conduct thorough and systematic research for ultra-small specimen scale.Aimed at the three ultra-small specimen test methods,namely small tensile,small punch and nanoindentation tests,this thesis proposed pertinent optimization and development,and verified the accuracy and feasibility based on both the pristine and ionirradiated materials.First,the geometry design of small tensile specimen was optimized to reduce the strain measurement errors due to thickness non-uniformity,which can drastically eliminate the scatters in the test results.Then,data correlation and correction methods were developed respectively,in order to improve the accuracy and reliability of the extracted true stress-strain curves from small punch and nanoindentation tests.Finally,the three miniature test methods with optimization or development were used to obtain the bulk properties from ion irradiated materials,in order to further evaluate their equivalence and reliability in ultra-small specimen scale.The main conclusion of this thesis is listed as follows:1)Geometry optimization for small tensile specimen,namely reducing the lengthto-width ratio of the parallel section properly,can drastically decrease the strain measurement errors due to thickness non-uniformity,and eliminate the scatters in the test results.Compared with small punch and nanoindentation tests,small tensile test suffers much more scatters in the test results when using ultra-small specimens.The FEM simulation shows that,when the specimen thickness is smaller,slight thickness nonuniformity can lead to significant inhomogeneity in strain distribution.That means the existence of a low strain area in the parallel section,which can lead to the scatters in the measured elongation.Geometry optimization by shortening the parallel section length properly can efficiently suppress the ratio of the low strain area in the parallel section,and thus significantly reduce the scatters in the test results.On that basis,digital image correlation(DIC)method was used to trace the local strain in the weakest section directly.This method can essentially avoid the influence of the low strain area on strain measurement,so is capable to further reduce the scatters in the test results.2)Development of data correlation and correction methods can improve the accuracy and reliability of the extracted true stress-strain curves from indirect test methods.Small punch and nanoindentation tests need inverse finite element method(i FEM)for data correlation.The proposed hybrid searching algorithm based on predictor-corrector strategy can give consideration to both the convergence accuracy and efficiency.The prediction stage searches for the parameters approximately by using genetic algorithm with global convergence capability,in order to approach the global optimum or a better local optimum.Then,the correction stage further searches the parameters by using Nelder-Mead algorithm with higher convergence rate,in order to achieve accurate convergence.Other the other hand,nanoindentation test usually accompanies with additional hardening effects.Data correction methods were proposed to remove such effects due to indentation size effect and strain rate mismatch.The indentation size effect is corrected by inducing the extrapolated results of Nix-Gao model to the expression of flow stress under a spherical indenter,which is achieved by means of correlating the geometrically necessary dislocation(GND)density underneath the spherical and Berkovich indenters.The strain rate mismatch is corrected by using i FEM analysis based on strain rate sensitive constitutive model,in which the strain rate sensitivity exponent is obtained from indentation strain rate jump test.The simulation and experiment results show that,the proposed data correlation and correction methods have good accuracy and reliability for true stress-strain curve extraction for the indirect test methods.3)The successful application in ion irradiation field shows that,the three test methods equipped with the optimization and development have good equivalence and reliability in ultra-small specimen scale.Reduced activation ferrite/martensite(RAFM)steel specimens with the thickness of 0.1 mm were ion-irradiated at the irradiation platform of heavy ion accelerator in Lanzhou,and successfully obtained ?0.2 dpa irradiation damage that is quasi-homogeneously distributed through thickness.Although the mechanical properties have drastically changed after irradiation,the extracted true stress-strain curves exhibit good consistency between the three methods.This indicates that all the three methods are workable in ultra-small specimen scale,and there exists huge potential to study irradiation effect in bulk materials via high-energy ion irradiation.In summary,this thesis reveals the main issues in ultra-small specimen mechanical tests,and promotes their developments in both theoretical and practical aspects.Therefore,it is believed that this thesis has both scientific and engineering significance.