| In recent years,with the rapid development of Chinese railways,the applied basic theory of wheel-rail relationship,has become more and more important.The wheel-rail adhesion-creep characteristics directly affect the traction/ braking performance and running quality of high-speed trains.The adhesion-creep characteristics are characterized by the adhesion coefficient and creepage.These two parameters need to be measured by the adhesion test.Internationally,the measurement uncertainty is usually used uniformly to express the quality of the measurement results.China National Accreditation Service for Conformity Assessment(CNAS)puts forward clear requirements for measurement uncertainty evaluation.The measurement uncertainty evaluation methods stipulated in domestic and foreign standards mainly include GUM(Guide to the Expression of Uncertainty in Measurement,GUM)and MCM(Monte Carlo method,MCM).According to the literature review,investigation and analysis of examples,it is found that GUM has certain limitations in the use of conditions.The traditional MCM is only for the case where the input quantities are independent of each other.To this end,some scholars have carried out MCM research with correlation,but they are all for the case where the correlation coefficient matrix is positive definite.There is a non-positive definite correlation among the longitudinal creepage inputs in the adhesion test of the high-speed wheel-rail interface test rig,which makes it impossible to evaluate the measurement uncertainty based on the traditional MCM.To solve this problem,an iterative correction algorithm is proposed in this paper.At the same time,combined with relevant mathematical transformations,the implementation steps of MCM considering non-positive definite correlation are sorted out.According to the adhesion test method of the high-speed wheel-rail interface test rig,the main sources of measurement uncertainty of the adhesion force coefficient and creepage are analyzed.Then,the measurement uncertainty of the adhesion force coefficient,the adhesion coefficient of different speed levels and the longitudinal creepage were evaluated and studied.The main results and conclusions obtained are as follows:(1)For the measurement uncertainty of the adhesion force coefficient at a speed of 300km/h,GUM and MCM were used for analysis and evaluation,and the evaluation results increased with the increase of the adhesion force coefficient.The standard uncertainty and expanded uncertainty obtained by MCM are 2% ~ 8% and 6% ~ 16%smaller than those obtained by GUM,respectively.The evaluation results of GUM failed to pass the verification of MCM.Therefore,the uncertainty of the adhesion force coefficient should be evaluated by MCM.(2)Then combining with the evaluation results of the measurement uncertainty of the adhesion force coefficient,The measurement uncertainty of the adhesion coefficient at 120km/h,200 km/h,250 km/h,300 km/h and 350 km/h was analyzed and evaluated,respectively.There is no obvious change in the evaluation results with the increase of speed.(3)For the measurement uncertainty of the longitudinal creepage at a speed of300km/h,GUM,MCM and MCM considering non-positive definite correlation were used for analysis and evaluation respectively.The evaluation results of MCM are seriously deviated from the reality.The evaluation results of GUM and MCM considering non-positive definite correlation were compared and analyzed.The uncertainty obtained by the two methods do not change with the increase of the creepage.The standard uncertainty and expanded uncertainty obtained by MCM considering the non-positive definite correlation are 8%and 21% smaller than those obtained by GUM,respectively.The evaluation results of GUM failed to pass the verification.Therefore,the uncertainty of creepage should be evaluated by MCM considering non-positive definite correlation.The evaluation of the measurement uncertainty of the adhesion coefficient and creepage can provide a supplement for the data analysis of the adhesion test and a method reference for the evaluation of the measurement uncertainty of other parameters. |