Giakoumakis, A;
(2006)
Estimating annual buildings ground floors heat losses using a one-dimensional (1-D) numerical model.
Doctoral thesis , UCL (University College London).
Preview |
Text
Giakoumakis_thesis.pdf Download (21MB) | Preview |
Abstract
In this work, an estimation of the annual buildings ground floors heat losses by means of numerical simulations of two different geometrical models (constructional details of buildings ground floors), using a 1-D numerical model, is attempted. Given the three-dimensional (3-D) nature of the heat transfer through the ground, the annual ground floor heat losses are first estimated using a 3-D model, constructed and simulated with the thermal analysis computer programs: "TRISCO" & "VOLTRA". Then, the 3-D model is converted to the 'respective' one-dimensional (1-D) one and the 'equalization' of the two models - for the both cases (geometrical models) - as far as the annual ground floor heat losses per unit surface area are concerned, is done by changing the values of the various simulation parameters of the used computer programs. Furthermore, since the various simulation tools, such as "TAS" thermal analysis software, generally simulate all heat transfer processes in one dimension - those through the ground floors included - and model the soil depth, in particular, to be: 1m, an estimation of the possibly introduced, in this 'methodology', errors is made, by comparing the respective results derived from the 3-D & 1-D numerical models. As far as the 'equalization' of the 1-D & 3-D numerical models is concerned, the results in question 'revealed' that, the (1-D numerical model's) soil depth, primarily and the soil thermal conductivity (A), secondly, are the most significant simulation parameters for the achievement of this aim. Regarding the errors possibly introduced in the process of estimating the annual buildings ground floor heat losses using a 1-D numerical model (with a soil depth value of: 1m), it is shown that, the size of these errors - for the specific models examined in this work - is approximately: -38% for the first and: +59% for the second one and, furthermore, that, the definition of the 'proper' soil depth value depends on the specific numerical model (its geometry, configuration and simulation parameters), as well as, the soil type and its thermal properties, the thermal conductivity being the most significant one. However, given the limited capabilities of "VOLTRA" thermal analysis computer program and of the PCs used for the numerical simulations, as well as, the great differences between the various numerical models (regarding their: geometry, configuration, simulation parameters and soil type/ thermal properties), a generalization of the conclusions presented hereby cannot be defended. Keywords: Annual buildings heat losses via the ground Numerical simulations Thermal analysis computer programs "TRISCO" & "VOLTRA" 'Equalization' of the 1-D & 3-D numerical models Estimating the annual buildings ground floor heat losses using a 1-D numerical model "TAS" thermal analysis software Soil depth (of the 1-D numerical model) Soil thermal conductivity.
Type: | Thesis (Doctoral) |
---|---|
Title: | Estimating annual buildings ground floors heat losses using a one-dimensional (1-D) numerical model |
Open access status: | An open access version is available from UCL Discovery |
Language: | English |
Additional information: | Thesis digitised by ProQuest. |
UCL classification: | |
URI: | https://discovery.ucl.ac.uk/id/eprint/1568369 |
Archive Staff Only
View Item |