【作者】 王春亮；

【导师】 王滨； 杨力；

【作者基本信息】 机械科学研究总院 ， 测试计量技术及仪器， 2007， 硕士

【摘要】 纳米压痕试验方法是在传统的布氏和维氏硬度试验方法的基础上发展起来的新兴的力学性能试验方法，它通过连续控制和记录加卸载时的载荷和位移数据，可以得到材料压痕硬度、杨氏模量、压痕蠕变、压痕松弛和断裂韧性等力学性能指标。美国MTS公司生产的纳米压痕仪是目前世界上最先进的纳米力学性能测试仪器之一，是微纳观力学性能测试的重要手段。本研究着眼于纳米压痕试验方法的影响因素，压痕硬度、马氏硬度和维氏硬度之间的关系，纳米压痕试验结果的不确定度，以及纳米压痕试验方法在薄膜试样中的应用等，其中的部分工作，根据目前的资料国内尚未开展。通过对纳米压痕试验影响因素的研究，结果表明a)对于纳米范围内的压痕试验，接触零点相对移动1％引起压痕硬度的相对误差约为4％，压痕模量的相对误差约为2％；b)压头面积函数必须定期校合，否则会使得测试结果出现误差，而且面积函数改变引起的压痕硬度的相对误差约为压痕模量的2倍；c)计算接触刚度时，卸载曲线拟合参数可选择在20～80％之间，而以往有关资料所介绍的是25～50％；d)连续刚度(CSM)法可以在连续加载过程中获得压痕硬度和压痕模量等压痕深度的连续函数，从而大大方便了薄膜材料的力学性能表征，但其测试结果与卸载法得到的结果不一定完全吻合，尤其在压痕硬度的测量上；e)采用玻氏压头和维氏压头进行纳米压痕试验时，相邻压入间距应保持在最大压入深度的25倍以上。在对压痕硬度、马氏硬度和维氏硬度之间关系的研究后发现，对于不同压头、以及不同硬度等级的试样，它们之间的关系不尽相同。本研究发现，对维氏压头而言，维氏硬度值约为压痕硬度值的0.0877倍，马氏硬度值约为压痕硬度值的0.8333倍；对玻氏(Berckvich)压头而言，维氏硬度值约为压痕硬度值的0.0901倍，马氏硬度值约为压痕硬度值的0.9009倍。本研究选择了高纯度熔融石英和高硬度GCr15钢样品，通过压入深度为2000nm的压痕试验，对样品的表面均匀性和深度均匀性进行了统计研究。高纯度熔融石英由于表面光滑(表面粗糙度R_a约为0.006μm)，又是非晶材料，各向同性，因此表面均匀性和深度均匀性都很好。采用变异系数表征时，压痕模量的表面均匀性为0.56％，深度均匀性为0.83％，压痕硬度表面均匀性为0.73％，深度均匀性为0.99％；采用F检验的统计方法更证明了在大多数情况下，该样品的不均匀性对压痕模量和压痕硬度试验结果的影响可以忽略不计。高硬度GCr15钢样品的表面粗糙度也可以达到约0.008μm。但由于其组织为多相合金，其中的隐针马氏体尺寸约为3μm，对于压入深度较小时，压痕硬度试验结果受其影响比较大，对于压入深度较大时，压痕试验结果受其影响相对比较小，采用F检验的统计方法证明了该样品的不均匀性有可能导致不同部位压痕模量和压痕硬度试验结果的偏差。影响纳米压痕测试结果的不确定度的因素很多，特别是由于纳米压痕试验设备和试验过程的复杂性，使得定量给出B类标准不确定度的评估十分困难，本研究通过实验室之间的比对试验研究，得出了采用纳米压痕试验方法测量熔融石英样品压痕模量的重复性限为2.96％，复现性限为9.33％，压痕硬度的重复性限为4.48％，复现性限为11.35％；测量Gr15钢样品压痕模量的重复性限为9.31％，复现性限为22.36％，压痕硬度的重复性限为15.67％，复现性限为26.60％。同时应用重复性限和复现性限的方法定量评估了纳米压痕试验方法的测量不确定度。该方法大大简化了纳米压痕试验结果的不确定度评估。从其不确定度评估结果得知，虽然接触零点确定和压头面积函数校准等对纳米压痕试验中压痕硬度测试值的影响大于压痕模量测试值(约为2倍)，但试验中压痕硬度和压痕模量的总不确定度确相差不大(压痕硬度的相对合成不确定度略大于压痕模量的相对合成不确定度)。即使对均匀性很好的样品，不同实验室间的试验结果仍可能存在一定的差异，这些差异可能来源于a)ISO 14577-1：2002对试验条件(如设备的校准，试验过程的控制等)规定得还不够严密；b)目前参加比对的实验室对该标准的理解或执行存在差异；c)各制造厂家生产的设备彼此间差异很大，准确度不一致等。随着试验次数的增多，纳米压痕试验结果的扩展不确定度减小，但试验次数超过5次后，试验次数对扩展不确定度的影响不大。本文还研究了几种现有的测试方法对于薄膜样品测试结果的影响，得出对于小于200nm的超薄膜，采用DCM组件中DCM CSM hardness & Modulus for Thin Films方法所得的结果相对比较可靠。在薄膜样品的测试中还发现不同硬度等级的薄膜，薄膜厚度对压痕模量和压痕硬度的影响程度也不一样，压入深度和薄膜厚度的比值是影响压痕模量和压痕硬度试验结果的重要因素。通过对厚度为10.4μm的氮化样品表面白亮层的试验结果分析，氮化样品白亮层的模量约为306GPa，硬度约为14.3GPa。解决了传统的测试手段无法对氮化样品白亮层性能进行测试的难题。更多还原

【Abstract】 Nanoindentation testing is a newly-developed experimental method for mechanicalproperties on the basis of those traditional experimental methods including Brinell’s andVickers’ methods for hardness. Through continually controlled and recorded the data ofload and displacement during the periods of loading and unloading, Meso-mechanicalmethod can provide many mechanical property indexes including instrumentalizedindentation hardness, Young’s modulus, indentation creep, indentation relaxation andfracture toughness etc. The nanoindentation produced by American MTS Corp. is one ofthe most advanced instruments in the world for nano-mechanical property measurement.Nanoindetation is a method for micro- and meso- mechanical property measurement. Thepurpose of our research is to make some contribution to the development ofnanoindentation through the study on the influencing factors, relations amongnanoindentaion hardness, Martens hardness and Vickers hardness, the uncertaintyanalysis for the experimental results of nanoindentation, and applications of thin-filmtesting samples in nanoindentation. And some of the work which has been done in thispaper was still not carried out by others according to domestic present reporting.Based on the study on the influencing factors in nanoindentation experiments, the resultsshowed that: a) In terms of nano-scale indentation, 1% relative displacement of contactzero will lead to an about 4% relative error to nanoindentation hardness, and an about 2%relative error to modulus, b) The intender area function should be checked periodically, orthe measuring result will be inaccurate. The relative errors to hardness resulted from thearea function are twice as those resulted from modulus, c) During the computation process,the fitting parameter of unloading rigidity curve can be selected in the range from 20% to80%. d) Indentation hardness and indentation modulus were obtained from the function ofdepth though Continuous Stiffness Measurement (CSM). It’s a good method for testing themechanics of thin films. But the testing result is not as same as that from unloading method,especially in indentation hardness, e)The adjacent separation for indentation should beover 25 times of the largest depth for indentation in terms of Berckvich’s and Vickers’indenter.By studying on relations among nanoindentation hardness, Vickers hardness and Martenshardness, it was found that the relations varies with different indenters and samples ofdifferent hardness levels. When using Vickers indenters, the values of Vickers hardnessare about 0.0877 times of those of nanoindentation hardness and Martens hardness areabout 0.8333 times of nanoindentation hardness; while to Berchvich indenters, values ofVickers hardness are about 0.0901 times of nanoindentaion hardness and Martenshardness are about 0.9009 times of nanoindentation hardness. High-purity fused silica and high hardness steel GCr15 were choosed in this study. Statisticresearch on the indentation with the depth of 2000nm were carried out to study its surfaceand depth uniformity, it was found that the degree of roughness of high-purity fused silicacan reach 0.006μm for its smooth surface. Thus, the surface uniformity and depthuniformity of high-purity fused silica were good due to its amorphous and isotropy. Toevaluate by coefficient of variability, the surface uniformity of Young’s modulus was 0.56%,the depth uniformity was 0.83%, and the surface uniformity of nanoindentation was 0.73%,the depth uniformity was 0.99%. According to F formula, the distribution of Young’smodulus and hardness of the samples can be considered to be uniform. In terms of highhardness steel GCr15, despite its only 3μm hidden needle-type martensite, its surfaceroughness can reach about 0.008μm, mainly for its multi-phased alloy structure. Thus, thenanoindentation will be affected largely, and the influence to indentation depth is relativelysmall. According to F formula, the surface Young’s modulus of steel is asymmetrical.As many factors should be considered in the uncertainty of the result of nanoindentation,through comparison experiment and it found that the repeatability of the tested fusedsilica’s indentation modulus is 2.96% and the reproducibility is 9.33%. The repeatability ofthe tested fused silica’s indentation hardness was 4.48% and the reproducibility was11.35%. The repeatability of the tested GCr15 steel’s indentation modulus was 9.31%, thereproducibility was 22.36%. The repeatability of the tested GCr15 steel’s indentationhardness was 15.67% and the reproducibility was 26.60%. It is obvious that the uncertaintyof GCr15 steel is a bit higher than that of fused silica. Meanwhile, uncertainty ofnanoindentation experiment were attained though repeatbility and reproducility. It’s asimple method on evaluating the uncertainty of nanoindentation experiment. Though theinfluence which come from the change of original point and vary of area function toindentation modulus is 2 times of indentation hardness, the values of total uncertainty onindentation hardness is almost no different on indentation modulus. It found that: becausea) ISO 14577-1:2002 does not have strict rules for experimental condition (such as theequipment adjust and the control of experimental process, b) different labs may havedifferent understanding and execution on the ISO standard, c) equipment of differentfactories differs a lot from each other and thus has different accuracy, the test result willhave some difference. The influence from the times of test can be ignored when the testingtimes are over five.Lastly, the influence of the current several methods to test results of thin film samples wasstudied. The results showed that for 200nm super-thin films, the method of DCM CSMhardness & Modulus for Thin Films in DCM tum out a relatively reliable result. In theprocess of testing thin-films, it was found that the influence of indentation hardness andindentation modulus was different from the thickness of thin-films. The ratio of indentationdepth to thin-films’ thickness is the most important factor in testing of the indentationhardness and the indentation modulus. Through the experimental analysis of the 10.4μmsurface white-bright layers of nitridized samples, it can be found that nanoindentation is arelatively good method to test the white-bright layers. Meanwhile, the results also show that the modulus of white-bright layers of nitridized samples was about 306GPa, the hardnessdegree was about 14.3GPa, which provides an effective analysis method for themechanical properties testing of nitridized layers.更多还原