Effects of sample and boundary imperfections on calibration of monotonic and cyclic constitutive models in sand

像许多其他科学和工程领域一样，岩土工程工程师使用实验室测试来确定土壤的工程特性，因此地基、水坝、隧道和其他结构可以用土壤来设计和建造。像所有其他实验室试验一样，这些试验也有缺点，特别是在试样边界与土壤相互作用的方式以及这种相互作用如何影响结果方面。**然而，所有岩土工程实践都假定这些试验在其边界条件下是完美的。标本的缺陷，如不均匀性，已经被发现，但解决方案一直是创造更均匀的标本。没有人知道怎样才算足够均匀，也没有人量化非均匀性的影响，也没有人量化边界条件对结果的影响。本研究将使用分析工具来估计这些不完美性对从这些测试中获得的参数的影响，并将产生计算工具来解释这些不完美性对工程师为设计创建的模型的影响。这项工作得到了加拿大软件公司RocScience的支持，这使他们比其他岩土工程软件公司有了优势，这些公司仍然认为他们的输入数据来自完美的实验室测试。

Like many other fields of science and engineering, geotechnical engineers use laboratory testing to determine**the engineering properties of soils, so foundations, dams, tunnels and other structures can be designed and**constructed with, in, or on soils. Like all other laboratory tests, these tests have shortcomings, specifically in**the way boundaries of specimens interact with the soil and how this interaction affects the results.**However, all geotechnical engineering practice has assumed these tests are perfect in their boundary conditions.**Imperfections of specimens, such as non-uniformities, have been recognized but the solution has always been a**race towards creating ever more uniform specimens. Nobody knows how uniform is uniform enough, or has**quantified the effects of non-uniformity, or boundary conditions on the results.**This research will use analytical tools to estimate the effects of these non-perfections on the parameters**obtained from these tests, and will produce computational tools that account for the effects of these**imperfections on models that engineers create for design. The work is supported by RocScience, a Canadian**software company, giving them an advantage over other geotechnical engineering software companies that still**assume their input data come from perfect laboratory tests.