Measuring system with a dual needle probe for testing the parameters of heat-insulating materials

Abstract

The paper presents a prototype of a measurement system with a hot probe based on a transient line heat source method, designed for testing the thermal parameters of heat insulation materials. The proposition is to use an auxiliary thermometer (dual needle probe) and a trained artificial neural network to determine the parameters of thermal insulation materials. The data extracted from the simulation of a nonstationary two-dimensional heat conduction model inside a sample of material with a dual needle probe trained the artificial neural network (ANN). The significant heat capacity of the needle probe is taken into account in the model. To solve the system of partial differential equations describing the model, the finite element method (FEM) was applied. The ANN is used to estimate the coefficients of the inverse heat conduction problem for a solid. The network determines the values of the effective thermal conductivity and effective thermal diffusivity on the basis of the temperature increases of the hot probe and the auxiliary thermometer. All calculations, such as FEM, and training and testing processes of the ANN, were carried out in the Matlab environment. The results of the experiment are also presented. The proposed measurement system for testing the parameters is suitable for temporary measurements in a building site or factory.

Measurement of the effective thermal conductivity of particulate materials by the steady-state heat flow method in a cuvette

“…Obviously, the effect of reducing the crosssectional area of the specimen through which the vertical heat flux travels is taken into account when calibrating the cuvette by the experimental points (figure 3, equation ( 3)). The contact thermal resistances when measuring in the cuvette are also obviously taken into account by the calibration 9 . We used the measurement technique with a cuvette, described above, to study the effective TC of particle beds [31,32] − of particles of various natures (polymers, ceramics, metals, etc), − of particles of various shapes (from debris to spherical), − of particles of various sizes (from micropowders to coarse grains), − in the range of effective TC from 0.06 to 0.7 W (m K) −1 .…”

“…In the hot wire method, for example, the wire is heated at a constant rate and the temperature change ΔT over time τ is recorded at a point close to it. The temperature response recorded after a given time delay is linearised for an infinitely long and thin heat source in the coordinates ΔT-ln(τ), and the TC λ of the medium is defined from the slope of this line [9]. Because the basic equations of TC for these models are derived on the assumption of an infinite medium of heat transfer, fairly large amounts of material are required in this case.…”

Measurement of thermal diffusivity of insulating material using an artificial neural network

“…The use of neural networks to determine thermal properties can be found in many publications [14][15][16][17][18][19]. Those publications illustrate the use of different concepts of measurement methods for different examined materials and substances.…”

Bulletin of the Polish Academy of Sciences: Technical Sciences

“…ANNs, which constitute an important element of the current solution method, have been previously considered an interesting alternative method to solve general IHTPs alongside genetic algorithms (GA), particle swarm optimization (PSO), and proper orthogonal decomposition (POD) [18]. Their successful application has been demonstrated in the case of inverse and optimization problems involving conduction [19][20][21][22][23][24][25][26][27], radiation [28][29][30][31], and convection [32][33][34], but also in power engineering [35][36][37]. In particular, Krejsa et al assessed their potential and discussed various strategies regarding their application for the inverse problem of heat conduction [19].…”