Abstract
Precast concrete lightened sandwich panels are widely used building elements. They are made by two concrete wythes separated by a layer of lightweight material: the central layer is inhomogeneous due to the presence of concrete ribs which tie the external wythe and act as thermal bridges.
Computation of thermal transmittance of sandwich panels is clearly described in European Standards, but in many cases it requires numerical simulations to determine the linear transmittance ψ associated with lightweight material-concrete interfaces in the inhomogeneous layer. Although simple, these simulations represent a critical issue for many panel manufacturers and they would much rather prefer correlations to compute ψ.
In this work we present a correlation based on an artificial neural network (ANN) to estimate linear trasmittauce values for current Italian sandwich panel production. Five input parameters are considered: rib width, lightweight material conductivity, and thickness of the three panel layers. To obtain the data which are necessary to train and test the ANN, a fast and accurate Spectral Element Method is used to solve Laplace equation in the neighborhood of a rib. 5460 ψ values are collected which ensure an accurate network response.
Export citation and abstract BibTeX RIS
Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.