<abstract xmlns="http://www.w3.org/1999/xhtml">In this work, the complex microstructure of the soil solid, at the microscale, is modeled by prescribing the spatial variability of thermal conductivity coefficient to distinct soil separates. We postulate that the variation of thermal conductivity coefficient of each soil separate can be characterized by some probability density functions: <italic>fCl</italic>(<italic>λ</italic>), <italic>fSi</italic>(<italic>λ</italic>), <italic>fSa</italic>(<italic>λ</italic>), for clay, silt and sand separates, respectively. The main goal of the work is to recover/identify these functions with the use of back analysis based on both computational micromechanics and simulated annealing approaches. In other words, the following inverse problem is solved: given the measured overall thermal conductivities of composite soil find the probability density function <italic>f</italic>(λ) for each soil separate. For that purpose, measured thermal conductivities of 32 soils (of various fabric compositions) at saturation are used. Recovered functions <italic>f</italic>(λ) are then applied to the computational micromechanics approach; predicted conductivities are in a good agreement with laboratory results.</abstract>

In this work, the complex microstructure of the soil solid, at the microscale, is modeled by prescribing the spatial variability of thermal conductivity coefficient to distinct soil separates. We postulate that the variation of thermal conductivity coefficient of each soil separate can be characterized by some probability density functions: fCl(λ), fSi(λ), fSa(λ), for clay, silt and sand separates, respectively. The main goal of the work is to recover/identify these functions with the use of back analysis based on both computational micromechanics and simulated annealing approaches. In other words, the following inverse problem is solved: given the measured overall thermal conductivities of composite soil find the probability density function f(λ) for each soil separate. For that purpose, measured thermal conductivities of 32 soils (of various fabric compositions) at saturation are used. Recovered functions f(λ) are then applied to the computational micromechanics approach; predicted conductivities are in a good agreement with laboratory results.

Recovery of microstructure properties: random variability of soil solid thermal conductivity

Wrocław University of Science and Technology, Faculty of Civil Engineering, Wrocław, Poland

Studia Geotechnica et Mechanica

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

In this work, the complex microstructure of the soil solid, at the microscale, is modeled by prescribing the spatial variability of thermal conductivity...

Recovery of microstructure properties: random variability of soil solid thermal conductivity

Data publikacji: 18 kwi 2016

Zakres stron: 99 - 107

DOI: https://doi.org/10.1515/sgem-2016-0011

Słowa kluczowe
simulated annealing, heat transfer, homogenization, saturated soil

© 2016 Damian Stefaniuk et al., published by De Gruyter Open

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Recovery of microstructure properties: random variability of soil solid thermal conductivity

Data publikacji: 18 kwi 2016

Zakres stron: 99 - 107

DOI: https://doi.org/10.1515/sgem-2016-0011

Słowa kluczowesimulated annealing, heat transfer, homogenization, saturated soil

© 2016 Damian Stefaniuk et al., published by De Gruyter Open

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Słowa kluczowe
simulated annealing, heat transfer, homogenization, saturated soil