Numerical Inversion Techniques

any transform can, in theory, be inverted provided the behavior in the complex plane is such that I F(s) 1 - 0 as I s 1 - . All physical processes must satisfy this condition, so in principle an inversion always exists, although the mathematics to find it can often be intractable. For this reason, numerical techniques have been developed, which have wide applicability for practical problems. We discuss these approximate, numerical techniques in the next section. 

Previous sections dealt with the analytical development of Laplace transform and the inversion process. The method of residues is popular in the inversion of Laplace transforms for many applications in chemical engineering. However, there are cases where the Laplace transform functions are very complicated and for these cases the inversion of Laplace transforms can be more effectively done via a numerical procedure. This section will deal with two numerical methods of inverting Laplace transforms. One was developed by Zakian (1969), and the other method uses a Fourier series approximation (Crump 1976). Interested readers may also wish to perform transforms using a symbolic algebra language such as Maple (Heck 1993). 

---

Evaluate in the time domain the time-amount course by applying traditional inverse Laplace transforms [359] or numerical inversion techniques [353,360],... 

Numerical Inversion Techniques 387 Table 9.4 Comparison Between the Numerical Inverse Obtained by the Zakian Method and the Analytical Solution for Example 9.17 ... 

C. LONG DIVISION. The most interesting and most useful z-transform inversion technique is simple long division of the numerator by the denominator of The ease with which z transforms can be inverted by this technique is one of the reasons why z transforms are often used. 

Eq. 4 is amenable to solution techniques based on the numerical inversion of Laplace-transformed equations these calculations can be performed rapidly and are therefore suitable for calibration. In Figure 1, typical soil/bentonite column predictions are shown to highlight the effect of the influent mixing zone on the spatial contaminant distributions for low-flow systems. The simulation results, which were generated for column conditions described by Khandelwal et al. (1998), indicate that the mixing zone has a significant influence on the shape of the spatial contaminant distribution and, therefore should be considered explicitly in estimating sorption parameters from spatial column data. 

The last term in Eq. (A-3) is the Laplace transform of E(a) Aweq(a), where a plays the role of t in Eq. (A-4) and t plays the role of s in Eq. (A-4). Thus the problem is reduced to this experimental observations give us the Laplace transform of the desired distribution function [modified by multiplication by Aweq(A)], We want the distribution function. The required calculation is a numerical inverse Laplace transform. This is clearly feasible, and numerical techniques are discussed in the literature 56). It is not a simple matter to carry out, however, and the accuracy requirements on the data are likely to be stringent. No direct method of computing the distribution from frequency response is known, although the step response can be computed from the frequency response by standard techniques. In view of the foregoing discussion, it appears that in principle, at least, the distribution can be computed from the frequency response. 

The techniques that avoid numerical inversions, correct the MMDs in a single step, and are based on either 1) rotating the linear calibration (when only a mass chromatogram and a linear calibration are available) or 2) modifying the interdetector volume shift (when molar mass sensitive detectors are employed). Their main advantage is that they produce smooth and unique solutions. Their limitation, however, is that they produce only approximate solutions. 

The membranes of practical usefulness arc, on the other hand, prepared by the phase-inversion technique. We have learned that the top selective layer of the asymmetric membrane prepared by the phase-inversion technique has a microscopic structure that depends on the conditions under which the membranes are prepared. Obviously, Che numerical parameters associated with the membrane transport equations should depend on the microscopic structure of the membrane and also on the conditions under which membranes are prepared. 

It has probably not escaped the attention of the reader that methods based on the ZORA Hamiltonian will be difficult to implement because the potential appears in the denominator. There are two possible approaches that could be taken to the evaluation of the inverse terms. One is to use numerical integration techniques where the potentials are tabulated on a grid. Formation of a function of the potential such as an inverse power is then a trivial problem. This kind of approach is suited to density functional... 

To practitioners in reservoir engineering and well test analysis, the state-of-the-art has bifurcated into two divergent paths. The first searches for simple closed-form solutions. These are naturally restricted to simplified geometries and boundary conditions, but analytical solutions, many employing method of images techniques, nonetheless involve cumbersome infinite series. More recent solutions for transient pressure analysis, given in terms of Laplace and Fourier transforms, tend to be more computational than analytical they require complicated numerical inversion, and hence, shed little insight on the physics. 

The BRM has been used to study the physics and the limitations of the various pulsed field techniques. In the last sections of this article, we discuss intermittent field, crossed fields and field inversion techniques in the framework of the BRM, comparing analytical as well as numerical work with some experimental results. 

Commonly used functions are listed in Table 1.3. In the integral form, the functional fitting involves extensive numerical integration techniques. With the development of the Laplace inversion technique, because of the arbitrariness in choosing a proper function, functional forms are used primarily for theoretical modeling and simulation. 

Adjugate Matrix of a Matrix Let Ay denote the cofactor of the element Oy in the determinant of the matrix A. The matrix B where B = (Ay) is called the adjugate matrix of A written adj A = B. The elements by are calculated by taking the matrix A, deleting the ith row and Jth. column, and calculating the determinant of the remaining matrix times (—1) Then A" = adj A/lAl. This definition may be used to calculate A"h However, it is very laborious and the inversion is usually accomplished by numerical techniques shown under Numerical Analysis and Approximate Methods. ... 

The spectral method is used for direct numerical simulation (DNS) of turbulence. The Fourier transform is taken of the differential equation, and the resulting equation is solved. Then the inverse transformation gives the solution. When there are nonlinear terms, they are calculated at each node in physical space, and the Fourier transform is taken of the result. This technique is especially suited to time-dependent problems, and the major computational effort is in the fast Fourier transform. 

With respect to the spreading calibration, several methods have been suggested e.g. (6-1 ) Numerous techniques have been proposed for solving the inverse filtering problem represented by Equation 1, with different degrees of success e.g. (it,15-19) Only references (M, (l8) and (I9) make no assumptions on the shape of g(t,x). 

In this work, an inverse filtering technique based on Wiener s optimal theory (1-3) is presented. This approach is valid for time-varying systems, and is solved in the time domain in mtrix form. Also, it is in many respects equivalent to the numerically "effl- lent" Kalman filtering approach described in ( ). For this reason, a... 

All the system response curves in frequency and time domains were calculated numerically from equations that are much too involved to reproduce in detail here. Transfer functions in Laplace transform notation are easily defined for the potentiostat and cell of Figure 7.1. Appropriate combinations of these functions then yield system transfer functions that may be cast into time- or frequency-dependent equations by inverse Laplace transformation or by using complex number manipulation techniques. These methods have become rather common in electrochemical literature and are not described here. The interested reader will find several citations in the bibliography to be helpful in clarifying details. 

To explore the full potential of what is sometimes called dynamic NMR 54>, i.e. NMR studies involving site and ligand exchange, is beyond the scope of this chapter and the reader is referred to numerous reviews 40 41 54 75 82). Only a few examples of the application of this technique can be given here, e.g. in the study of ring inversion, rotation about single bonds and inversion at nitrogen. 

---