Aldo Orioli

On the reliability of ASHRAE conduction transfer function coefficients of walls

The transfer function method recommended by the ASHRAE can be considered the most modern tool currently available for the thermal analysis of a building. It is particularly well suited for use with a computer as it makes it possible to describe with great accuracy the internal heat gain of walls using a small set of coefficients. The present paper shows how to calculate sets of coefficients diverse from that proposed by Mitalas who first developed the method and on the basis of an unequivocal criterion, to prove the advantages in using them. The authors also investigated some of the paramount mathematical and physical aspects which affect the approximation degree of the ASHRAE method and, with reference to the optimal number of poles and coefficients, they tried to offer some certainties for a more conscious use of it.

The ASHRAE clear sky model. An evaluation in the Mediterranean Zone

Abstract The ASHRAE clear sky model in the version proposed by G. L. Powell is here applied to a yearly set of data for Palermo with small scale time intervals. The model is shown to yield computed values generally higher than the measured ones. New values for the atmospheric extinction coefficient are derived and proposed for the application to the Mediterranean atmosphere. A successful verification is performed using hourly data of Genoa, a location belonging to the same meteoclimatic area.

An analysis of direct solar transmittance and atmospheric turbidity in the Mediterranean climatic belt

Abstract The empirical model developed by King and Buckius for the direct transmission in clear sky is applied to hourly data of two Yugoslav locations exhibiting typical meteoclimatic characteristics common to the whole Mediterranean belt. The occurrence of “very clear”, “clear” and “very hazy” hours, after the usually accepted definitions, is determined and discussed. The degree of accuracy of the previsional model is investigated performing the comparison with the corresponding values obtained directly by energetic procedure. A general overestimate of the transmittance is experienced with notable deviations in meteorological conditions affected by the presence of dense clouds which behave as screens to the direct solar beam and which indeed the model neglects. Better results have been obtained testing the methodology by reduced samples, i.e. excluding hourly data with high experimental values of the ratio diffuse/global radiation, or adopting average daily values.

Analysis of the accuracy in modelling of transient heat conduction in plane slabs

Abstract The accuracy achieved by modelling unidimensional, thermal conduction using lumped representation is assessed. The transient responses of one surface of a plane, infinite, homogeneous slab when the other surface is subjected to unit step or linear ramp excitations of temperature or heat flux are analysed. The results are expressed in dimensionless form and are compared as a function of the Fourier variable. The lumping errors are assessed by the calculation of the percentage error. Graphs collecting the possible pairs m.β by which a prefixed percentage error is found are shown and discussed.

An Improved Algorithm For Thermal Dynamic Simulation Of WallsUsing Z-transform Coefficients

The Transfer Function Method (TFM), recommended by American Society of Heating, Refkigerating and Air-conditioning Engineers (ASHRAE), is one of the most modern tools available to solve heat transfer problems in building envelopes and environments. TFM utilises Z-transform to solve the equations system that describes the heat transfer in a multi-layered wall. Due to an analogy with an electric circuit, it is possible to write the equations system in a matrix suitable to be solved by computer. Authors carried out an analysis on an historical building placed in the south of Italy to test the reliability and the quality of the thermal dynamic simulation using TFM. The analysis is performed using some control systems tools like Bode plots, step response, root and poles location. Results show clearly that, for very massive building, a simply application of TFM fails but if the poles and the residuals are re-ordered in a different position, it will be possible to solve the numerical problem. Furthermore, the analysis shows that the choice of the sampling period is a significant factor to obtain a reliable simulation.

A computer oriented method for the analysis of non steady state thermal behaviour of buildings

Abstract A method for the analysis of non steady thermal behaviour of buildings, based on the Laplaces transform, is proposed. The study is performed in a general way attaining the expressions of the solutions in a synthetic form by using matrix analysis. It is verified that the generic unknown quantity, expressed as a ratio of determinants, satisfies the conditions of inverse Laplace transform so that the theorem of residue can be applied. A procedure for finding the thermal responses time variation, well suited to personal computer implementation, is identified, which does not require the analytical development of the system of equations describing the thermal balances. This method, here applied to a thermal structure assuming constant values for the involved parameters, shows distinctive features and appears to be a viable alternative to other more popular techniques.

A more reliable relation between Angström and Linke atmospheric turbidity parameters

Abstract This work analyses the correlation between Angstrom and Linke atmospheric turbidity parameters (β and T ) which are commonly used in studies about the atmospheres behaviour since they efficaciously monitor a point turbidity state and model the aerosol attenuation of solar radiation. Starting from the usual linear relation β = a + bT , a more reliable one is here derived by hourly data of three coastal locations with typical Mediterranean meteoclimatic characteristics, for values of the experimental ratios of diffuse and global solar radiation ⩽0.3. The expression found notably differs from the starting one as shown by the time variable.