دانلود رایگان مقاله اثر barreling در حین انجام آزمایش فشاری محوری نمونه مکعبی

Abstract

For scarce materials, such as archaeological wood, cubic samples are often used instead of standardized prisms for mechanical tests, since the elasticity can be determined in all three directions within a single sample, but with such samples barrelling makes it difficult to identify the elastic properties. The purpose of the present study is firstly to numerically investigate the effects of barrelling in cubic samples during compressive testing; secondly to numerically investigate and compare barrelling on isotropic and transversely isotropic material parameters; thirdly to compare four strain measurement techniques using digital image correlation, strain gauges and direct readings from the testing machine and finally to estimate the error due to barrelling by implementing the experimentally obtained orthotropic material parameters to the numerical model. The presented relative errors provide information when the perturbation caused by barrelling is negligible or significant for various materials and strain measurements. As an example, the results of compressive tests on waterlogged archaeological oak impregnated with polyethylene glycol are discussed.

4. Conclusions

This study was initiated due to the problem of obtaining accurate data for the waterlogged PEG-impregnated oak timber of the Vasa ship, and the fact that the material was available in the form of cubic samples. This study has validated the use of cubic samples, and the magnitude of the errors in the elasticity constants arising from the friction between platens and the sample when cubic samples are subjected to compressive tests has been quantified. Taking into account the magnitude of these errors, it will be possible to estimate the right engineering constants of material and to design suitable support for the ship. Barrelling in cubic samples subjected to compressive testing is a phenomenon that can influence the estimation of the engineering elastic constants. The magnitude of the error depends on the type of material anisotropy, viz. isotropy, transverse isotropy and orthotropy, which are representative of most load-bearing materials: Isotropic material is affected most by the barrelling. For example, the measured Young's moduli was overestimated by up to 20% for a material with a Poisson's ratio of n ¼ 0.45. The estimation of Poisson's ratio gives less errors than with other materials. Transversely isotropic material shows an overestimation of Young's moduli in both the L and T directions but less than the isotropic material. The error is larger for the LT and TT plane in the estimation of the Poisson's ratio. Orthotropic material was simulated using the experimental input data for the experimentally characterised archaeological wood. The result shows an overestimation of Young's moduli in all principal planes of orthotropic wood, but the error is similar to that in the TT plane in the transversely-isotropic material, which is less than 1.5% error. T