Estimation of two-sided boundary conditions for two-dimensional inverse heat conduction problems

A hybrid numerical algorithm of the Laplace transform technique and finite-difference method with a sequential-in-time concept and the least-squares scheme is proposed to predict the unknown surface temperature of two-sided boundary conditions for two-dimensional inverse heat conduction problems. In the present study, the functional form of the estimated surface temperatures is unknown a priori. The whole time domain is divided into several analysis sub-time intervals and then the unknown surface temperatures in each analysis interval are estimated. To enhance the accuracy and efficiency of the present method, a good comparison between the present estimations and previous results is demonstrated. The results show that good estimations on the surface temperature can be obtained from the transient temperature recordings only at a few selected locations even for the case with measurement errors. It is worth mentioning that the unknown surface temperature can be accurately estimated even though the thermocouples are located far from the estimated surface. Owing to the application of the Laplace transform technique, the unknown surface temperature distribution can be estimated from a specific time.     

Solution of inverse heat conduction problems using maximum entropy method

A solution scheme based on the maximum entropy method (MEM) for the solution of one-dimensional inverse heat conduction problem is proposed. The present work introduces MEM in order to build a robust formulation of the inverse problem. MEM finds the solution which maximizes the entropy functional under the given temperature measurements. In order to seek the most likely inverse solution, the present method converts the inverse problem to a non-linear constrained optimization problem. The constraint of the problem is the statistical consistency between the measured temperature and the estimated temperature. Successive quadratic programming (SQP) facilitates the maximum entropy estimation. The characteristic feature of the method is discussed with the sample numerical results. The presented results show considerable enhancement in the resolution of the inverse problem and bias reduction in comparison with the conventional methods.     

Two dimensional steady-state inverse heat conduction problems

In this work we estimate the surface temperature in two dimensional steady-state in a rectangular region by two different methods, the singular value decomposition (SVD) with boundary element method (BEM) and the least-squares approach with integral transform method (ITM). The BEM method is efficient for solving inverse heat conduction problems (IHCP) because only the boundary of the region needs to be discretized. Furthermore, both temperature and heat flux at the unknown boundary are estimated at the same time. The least-squares technique involves solving the equations constructed from the measured temperature and the exact solution. The measured data are simulated by adding random errors to the exact solution of the direct problem. The effects of random errors on the accuracy of the predictions are examined. The sensitivity coefficients are also presented to illustrate the effect of sensor location on the estimated surface conditions. Numerical experiments are given to demonstrate the accuracy of the present approaches.     

Virtual boundary element method in conjunction with conjugate gradient algorithm for three-dimensional inverse heat conduction problems

In this article, the virtual boundary element method (VBEM) in conjunction with conjugate gradient algorithm (CGA) is employed to treat three-dimensional inverse problems of steady-state heat conduction. On the one hand, the VBEM may face numerical instability if a virtual boundary is improperly selected. The numerical accuracy is very sensitive to the choice of the virtual boundary. The condition number of the system matrix is high for the larger distance between the physical boundary and the fictitious boundary. On the contrary, it is difficult to remove the source singularity. On the other hand, the VBEM will encounter ill-conditioned problem when this method is used to analyze inverse problems. This study combines the VBEM and the CGA to model three-dimensional heat conduction inverse problem. The introduction of the CGA effectively overcomes the above shortcomings, and makes the location of the virtual boundary more free. Furthermore, the CGA, as a regularization method, successfully solves the ill-conditioned equation of three-dimensional heat conduction inverse problem. Numerical examples demonstrate the validity and accuracy of the proposed method.     

Numerical method for hyperbolic inverse heat conduction problems

The Laplace transform technique and control volume method in conjunction with the hyperbolic shape function and least-squares scheme are applied to estimate the unknown surface conditions of one-dimensional hyperbolic inverse heat conduction problems. In the present study, the expression of the unknown surface conditions is not given a priori. To obtain the more accurate estimates, the whole time domain is divided into several analysis sub-time intervals. Afterward, the unknown surface conditions in each analysis interval are estimated. To evidence the accuracy of the present method, a comparison between the present estimations and exact results is made. Results show that good estimations on the unknown surface conditions can be obtained from the transient temperature recordings only at one selected location even for the cases with measurement errors.     

Radial integration boundary element method for heat conduction problems with convective heat transfer boundary

A new boundary domain integral equation with convective heat transfer boundary is presented to solve variable coefficient heat conduction problems. Green’s function for the Laplace equation is used to derive the basic integral equation with varying heat conductivities, and as a result, domain integrals are included in the derived integral equations. The existing domain integral is converted into an equivalent boundary integral using the radial integration method by expressing the normalized temperature as a series of radial basis functions. This treatment results in a pure boundary element analysis algorithm and requires no internal cells to evaluate the domain integral. Numerical examples are presented to demonstrate the accuracy and efficiency of the present method.     

An analytical solution for two-dimensional inverse heat conduction problems using Laplace transform

An analytical method has been developed for two-dimensional inverse heat conduction problems by using the Laplace transform technique. The inverse solutions are obtained under two simple boundary conditions in a finite rectangular body, with one and two unknowns, respectively. The method first approximates the temperature changes measured in the body with a half polynomial power series of time and Fourier series of eigenfunction. The expressions for the surface temperature and heat flux are explicitly obtained in a form of power series of time and Fourier series. The verifications for two representative testing cases have shown that the predicted surface temperature distribution is in good agreement with the prescribed surface condition, as well as the surface heat flux.     

A semi-analytical boundary collocation solver for the inverse Cauchy problems in heat conduction under 3D FGMs with heat source

Abstract

In this article, the inverse Cauchy problems in heat conduction under 3D functionally graded materials (FGMs) with heat source are solved by using a semi-analytical boundary collocation solver. In the present semi-analytical solver, the combined boundary particle method and regularization technique is employed to deal with ill-pose inverse Cauchy problems. The domain mapping method and variable transformation are introduced to derive the high-order general solutions satisfying the heat conduction equation of 3D FGMs. Thanks to these derived high-order general solutions, the proposed scheme can only require the boundary discretization to recover the solutions of the heat conduction equations with a heat source. The regularization technique is used to eliminate the effect of the noisy measurement data on the accessible boundary surface of 3D FGMs. The efficiency of the proposed solver for inverse Cauchy problems is verified under several typical benchmark examples related to 3D FGM with specific spatial variations (quadratic, exponential and trigonometric functions).     

Estimation of unknown heat source function in inverse heat conduction problems using quantum-behaved particle swarm optimization

The estimation of temporal dependent heat source in transient heat conduction problem is investigated. A stochastic method known as quantum-behaved particle swarm optimization (QPSO) is used to estimate the heat source without a priori information on its functional form, which is classified as the function estimation by inverse calculation. Because of the ill-posedness of this kind of inverse problems, Tikhonov regularization method is applied in this paper to stabilize the solution. Numerical experiments indicate the validity and stability of the QPSO method. Comparison with the conjugate gradient method (CGM) is also presented in this paper.     

An inversion approach for the inverse heat conduction problems

The inverse heat conduction problems (IHCP) analysis method provides an efficient approach for estimating the thermophysical properties of materials, the boundary conditions, or the initial conditions. Successful applications of the IHCP method depend mainly on the efficiency of the inversion algorithms. In this paper, a generalized objective functional, which has been developed using a generalized stabilizing functional and a combinational estimation that integrates the advantages of the least trimmed squares (LTS) estimation and the M-estimation, is proposed. The objective functional unifies the regularized M-estimation, the regularized least squares (LS) estimation, the regularized LTS estimation, the regularized combinational estimation of the LTS estimation and the M-estimation, and the regularized combinational estimation of the LS estimation and the M-estimation into a concise formula. The filled function method, which is coupled with the Broyden–Fletcher–Goldfarb–Shanno (BFGS) algorithm, is developed for searching a possible global optimal solution. Numerical simulations are implemented to evaluate the feasibility and effectiveness of the proposed algorithm. Favorable numerical performances and satisfactory results are observed, which indicates that the proposed algorithm is successful in solving the IHCP.     

A finite element analysis for inverse heat conduction problems

This paper presents an efficient inverse analysis technique based on a sensitivity coefficient algorithm to estimate the unknown boundary conditions of multidimensional steady and transient heat conduction problems. Sensitivity coefficients were used to represent the temperature response of a system under unit loading conditions. The proposed method, coupled with the sensitivity analysis in the finite element formulation, is capable of estimating both the unknown temperature and heat flux on the surface provided that temperature data are given at discrete points in the interior of a solid body. Inverse heat conduction problems are referred to as ill-posed because minor inaccuracy or error in temperature measurements cause a drastic effect on the predicted surface temperature and heat flux. To verify the accuracy and validity of the new method, two-dimensional steady and transient problems are considered. Their surface temperature and heat flux are evaluated. From a comparison with the exact solution, the effects of measurement accuracy, number and location of measuring points, a time step, and regularization terms are discussed. © 1998 Scripta Technica. Heat Trans Jpn Res, 26(6): 345–359, 1997     

A fitting algorithm for solving inverse problems of heat conduction

The paper presents an algorithm for solving inverse problems of heat transfer. The method is based on iterative solving of direct and adjoint model equations with the aim to minimize a fitting functional. An optimal choice of the step length along the descent direction is proposed. The algorithm has been used for the treatment of a steady-state problem of heat transfer in a region with holes. The temperature and the heat flux density were known on the outer boundary of the region, whereas these values on the boundaries of the holes are to be determined. The idea of the algorithm consist in solving of Neumann problems where the heat flux on the outer boundary is prescribed, whereas the heat flux on the inner boundary is guessed. The guess is being improved iteratively to minimize the mean quadratic deviation of the solution on the outer boundary from the given distribution.The results obtained show that the algorithm provides smooth, non-oscillating, and stable solutions to inverse problems of heat transfer, that is, it avoids disadvantages inherent in other computational methods for such problems. The ill-conditioning of inverse problems in the Hadamard sense is exhibited in that a very quick convergence of the fitting functional to its minimum does not imply a comparable rate of convergence of the recovered temperature on the inner boundary to the true distribution.The considered method can easily be extended to nonlinear problems.Numerical calculation has been carried out with the FE program Felics developed at the Chair of Mathematical Modelling of the Technical University of Munich.     

A technique for uncertainty analysis for inverse heat conduction problems

A technique is presented for the uncertainty analysis of the linear Inverse Heat Conduction Problem (IHCP) of estimating heat flux from interior temperature measurements. The selected IHCP algorithm is described. The uncertainty in thermal properties and temperature measurements is considered. A propagation of variance equation is used for the uncertainty analysis. An example calculation is presented. Parameter importance factors are defined and computed for the example problem; the volumetric heat capacity is the dominant parameter and an explanation is offered. Thoughts are presented on extending the analysis to include the non-linear problem of temperature dependent properties.     

Solution of inverse heat conduction problems using Kalman filter-enhanced Bayesian back propagation neural network data fusion

This paper presents an efficient technique for analyzing inverse heat conduction problems using a Kalman Filter-enhanced Bayesian Back Propagation Neural Network (KF-B²PNN). The training data required for the KF-B²PNN are prepared using the Continuous-time analogue Hopfield Neural Network and the performance of the KF-B²PNN scheme is then examined in a series of numerical simulations. The results show that the proposed method can predict the unknown parameters in the current inverse problems with an acceptable error. The performance of the KF-B²PNN scheme is shown to be better than that of a stand-alone Back Propagation Neural Network trained using the Levenberg–Marquardt algorithm.     

Estimation of time-dependent heat flux and measurement bias in two-dimensional inverse heat conduction problems

This paper presents the results from the adaptive estimator developed to estimate time-dependent boundary heat flux in two-dimensional heat conduction domain with heated and insulated walls. For the estimation, the algorithm requires only the temperatures measured at the insulated walls. In addition, the estimator also predicts the bias in the measurements. In modeling the system, it is assumed that the input flux and bias sequence dynamics can be modeled by a semi-Markov process. By incorporating the semi-Markovian concept into a Bayesian estimation technique, the estimator consists of a bank of parallel, adaptively weighted, Kalman filters. Computer simulation results reveal that the proposed adaptive estimator has improved estimation performance even for step changing heat flux and measurement bias.     

Some comments on the sensitivity to sensor location of inverse heat conduction problems using Beck's Method

It has been shown that two inverse heat conduction problems are not always identical despite the dimensionless time steps based on the distance from the heated surface to the sensor are the same. First, the sensitivity to measurement errors is twice for an interior sensor than for a sensor at the insulated surface even if both the dimensionless time steps based on sensor depth and the dimensionless measurement errors are equals. Now, considering an experiment where the dimensional random measurement errors are more likely to be the same for every sensor, the sensitivity to measurement errors is inversely proportional to the sensor depth for a constant dimensionless time step. But it is important to point out that the dimensional time steps are proportional to the square of the sensor depth. Thus the closer the sensor, the smaller the dimensional time step and then the largest the sensitivity to measurement errors. However, the best sensor location for a given dimensional time step and a given dimensional variance in the temperature measurements is near the heated surface. The results presented herein apply to all IHCP algorithms and are of interest for the comparison of IHCP methods. They are only valid when small dimensionless time steps are used which usually is required.     

Analysis of conduction heat transfer problems in steady, 2D regions with circular holes using a special boundary integral method

A special boundary integral method developed for two-dimensional regions containing circular holes is used to calculate temperature and heat transfer on the boundaries of several selected regions. The geometrical configuration of the region is arbitrary and convective boundary conditions are assumed. An important feature of the method is analytic representation of temperature and its normal derivative on the interior circular holes in the form of a harmonic series. This makes the application of the boundary integral method convenient and free from conditioning problems associated with small interior boundaries. Heat transfer from circular isothermal interior holes are calculated for several illustrative examples using three terms of the harmonic series representation for heat transfer at each of the circular boundaries. The results are presented and discussed.     

Fourier analysis of conjugate gradient method applied to inverse heat conduction problems

The convergence and regularization mechanism of the conjugate gradient algorithm applied to inverse heat conduction problems are studied within the context of a Fourier analysis, for a square enclosure subjected to an unknown time-varying heat flux on one side, and to known boundary conditions on the remaining sides. Analytic solutions are derived for the Fourier components of the unknown flux over a given time interval. The convergence rate of the algorithm is thereby shown to depend essentially on the time frequency of the data. Numerical solutions are also presented to describe in details the convergence process and solution regularization power of the conjugate gradient method, when the unknown heat flux contains many frequency components and the measurement data are noisy. It is found that an unknown time-dependent heat flux may be satisfactorily recovered using a single sensor even when the temperature field becomes two-dimensional, and that the sensor should be placed in a symmetric manner for better results.     

Implementation of an analytical two-dimensional inverse heat conduction technique to practical problems

Two improvements to practical implementation of a solution to the two-dimensional inverse heat conduction problem are presented. The first concept is useful for experimental data with strong or irregular fluctuations in time. The second procedure improves the spatial resolution for problems where the source of the surface heat flux distribution is moving along the surface. The method is tested against analytical solutions and data from quench cooling experiments. Both procedures are found to enhance the quality of the inverse solution results.     

Particle Swarm Optimization-based algorithms for solving inverse heat conduction problems of estimating surface heat flux

An inverse analysis of estimating a time-dependent surface heat flux for a three-dimensional heat conduction problem is presented. A global optimization method known as Particle Swarm Optimization (PSO) is employed to estimate the unknown heat flux at the inner surface of a crystal tube from the knowledge of temperature measurements obtained at the external surface. Three modifications of the PSO-based algorithm, PSO with constriction factor, PSO with time-varying acceleration of the cognitive and social coefficients, and PSO with mutation are carried out to implement the optimization process of the inverse analysis. The results show that the PSO with mutation algorithm is significantly better than other PSO-based algorithms because it can overcome the drawback of trapping in the local optimum points and obtain better inverse solutions. The effects of measurement errors, number of dimensionalities, and number of generations on the inverse solutions are also investigated.