Differential Equations in Applied

The little book, Differential Equations in Applied Chemistry (Fig. 6), has been unsung for years, though all through an era when few engineers truly understood calculus it showed off powerful mathematical tools, among them (in the 1936 second edition) numerical solution methods. It connected with chemist J. W. Mellor s 1902 Higher Mathematics for Students of Chemistry... 

Figure 6. Differential Equations in Applied Chemistry, 1923, by Frank Lauren...

Equation (16) is a differential equation and applies equally to activity coefficients normalized by the symmetric or unsymme-tric convention. It is only in the integrated form of the Gibbs-Duhem equation that the type of normalization enters as a boundary condition. 

It must repeatedly have been remarked, however, that these equations are not in themselves sufficient to lead to a complete solution of the problems to which they have been applied. This arises from the fact that they are differential equations, in the solution of which there always appear arbitrary constants of integration (H. M., 73,101, 121). Thus, the relation between the pressure of a saturated vapour and the temperature is expressed by the differential equation of Clausius ( 80) ... 

The modeling of steady-state problems in combustion and heat and mass transfer can often be reduced to the solution of a system of ordinary or partial differential equations. In many of these systems the governing equations are highly nonlinear and one must employ numerical methods to obtain approximate solutions. The solutions of these problems can also depend upon one or more physical/chemical parameters. For example, the parameters may include the strain rate or the equivalence ratio in a counterflow premixed laminar flame (1-2). In some cases the combustion scientist is interested in knowing how the system mil behave if one or more of these parameters is varied. This information can be obtained by applying a first-order sensitivity analysis to the physical system (3). In other cases, the researcher may want to know how the system actually behaves as the parameters are adjusted. As an example, in the counterflow premixed laminar flame problem, a solution could be obtained for a specified value of the strain... 

For properties of Bessel functions, see for example, Walas, Modelling with Differential Equations in Chemical Engineering, 1991). Applying the condition at the center, Eq (4),... 

Equation (10) is a diffusion equation which applies equally well for matter or heat. The solution of this equation has been studied by many workers. As with differential equations in general, one arbitrary constant is required for each derivative. Since the diffusion equation has partial derivatives, the arbitrary constants have to be functions of the variable which is not involved in the derivative. The diffusion equation requires two functions of time at fixed values of space coordinates (the boundary conditions) and one function of distance at a given time (the initial condition). These have already been established [eqns. (3)—(5)]. It is possible to proceed to solve the diffusion equation for p(r,t) now and then calculate the particle current of B towards each A reactant and so determine the rate of reaction. 

Overall our objective is to cast the conservation equations in the form of partial differential equations in an Eulerian framework with the spatial coordinates and time as the independent variables. The approach combines the notions of conservation laws on systems with the behavior of control volumes fixed in space, through which fluid flows. For a system, meaning an identified mass of fluid, one can apply well-known conservation laws. Examples are conservation of mass, momentum (F = ma), and energy (first law of thermodynamics). As a practical matter, however, it is impossible to keep track of all the systems that represent the flow and interaction of countless packets of fluid. Fortunately, as discussed in Section 2.3, it is possible to use a construct called the substantial derivative that quantitatively relates conservation laws on systems to fixed control volumes. 

Different problems are modeled by two-point boundary value differential equations in which the values of the state variables are predetermined at both endpoints of the independent variable. These endpoints may involve a starting and ending time for a time-dependent process or for a space-dependent process, the boundary conditions may apply at the entrance and at the exit of a tubular reactor, or at the beginning and end of a counter-current process, or they may involve parameters of a distributed process with recycle, etc. Boundary value problems (BVPs) are treated in Chapter 5. 

The Dimensionless Parameter is a mathematical method to solve linear differential equations. It has been used in Electrochemistry in the resolution of Fick s second law differential equation. This method is based on the use of functional series in dimensionless variables—which are related both to the form of the differential equation and to its boundary conditions—to transform a partial differential equation into a series of total differential equations in terms of only one independent dimensionless variable. This method was extensively used by Koutecky and later by other authors [1-9], and has proven to be the most powerful to obtain explicit analytical solutions. In this appendix, this method will be applied to the study of a charge transfer reaction at spherical electrodes when the diffusion coefficients of both species are not equal. In this situation, the use of this procedure will lead us to a series of homogeneous total differential equations depending on the variable, v given in Eq. (A.l). In other more complex cases, this method leads to nonhomogeneous total differential equations (for example, the case of a reversible process in Normal Pulse Polarography at the DME or the solutions of several electrochemical processes in double pulse techniques). In these last situations, explicit analytical solutions have also been obtained, although they will not be treated here for the sake of simplicity. 

One advantage in the sequential approach is that only the parameters that are used to discretize the control variable profile are considered as the decision variables. The optimization formulated by this approach is a small scale NLP that makes it attractive to apply for solving the optimal control with large dimensional systems that are modeled by a large number of differential equations. In addition, this approach can take the advantage of available IVP solvers. However, the limitation of the sequential method is a difficulty to handle a constraint on state variables (path constraint). This is because the state variables are not directly included in NLP. 

B. L. Hou, and Y. P. Sun, Applying mechanization for soving nonlinear boundary value problem of ordinary differential equation in reaction engineering, J. Computers and Applied Chemistry 23(3) (2006) 255-259. 

Where the Schrodinger or Dirac equations apply, quantization will appear from sundry mathematical conditions on the existence of physically meaningful solutions to these differential equations. However, in nonrela-tivistic terms, spin does not come from a differential equation It comes from the assumptions of spin matrices, or from "necessity" (the Dirac equation does yield spin = 1/2 solutions, but not for higher spin). So we must posit quantum numbers (see Section 2.12) even when there are no differential equations in the back to "comfort us." This is especially true for the weak and strong forces, where no distance-dependent potential energy functions have been developed. 

To express this in more formal mathematical terms, we cast a partial differential equation in variational form (or integral form or weak formulation) by multiplying by a suitable function, integrating over the domain where the equation is posed and applying Greens theorem. Thus, if we consider the reaction-diffusion equation... 

There are several reasons for going first to this level of generality for the n-compartment model. First/ it points out clearly that the theories of noncompartmental and compartmental models are very different. While the theory underlying noncompartmental models relies more on statistical theory/ especially in developing residence time concepts [see/ e.g./ Weiss (11)]/ the theory underlying compartmental models is really the theory of ordinary/ first-order differential equations in which/ because of the nature of the compartmental model applied to biological applications/ there are special features in the theory. These are reviewed in detail in Jacquez and Simon (5)/ who also refer to the many texts and research articles on the subject. 

Therefore, while it is relatively easy to calculate the shear stress at the surfece of the rotating cylinder from Equation 3A.2, one can only derive an expression for the difference in shear rates at the surfaces of the inner and outer cylinders from the basic equations of flow. Additional work is required to calculate the corresponding shear rate yi and there have been several approaches to determination it. One approach has been to apply infinite series solution to the differential equation in 3A.12. 

The solution procedure for solving heat conduction problems can be. summarized as (1) formulate the problem by obtaining the applicable differential equation in its simplest form and specifying the boundary conditions, (2) obtain the general solution of the differential eqnation, and (3) apply the boundary conditions and determine the arbitrary constants in the general solution (Fig. 2—40). This is demonstrated below with examples. 

The same line of reasoning also applies to difierential equations. Typi- cally, differential equations have multiple solutions that contain at least one arbitrary constant. Any function that satisfies the differential equation on an interval is called a solution of that differential equation in tiiat interval. 

The method of zonation was applied to the energy and material conservation equations. Based on centered finite difTerence approximations, this method can transform three partial differential equations in radial distance and time to ordinary differential equations in time only. Following this, the ordinary differential equations were solved by using Crank-Nicholson algorithm. On the basis of this, the volumetric fluxes of those tar-phase and total volatile phase components were integrated with time by using in roved Euler method to evaluate overall pyrolysis product yields, and afterwards the gas yield can be deduced. 

Finally, in 65 there is for the first time a detailed presentation of the exponential fitting. This is an excellent book in which one can find reference work for the exponential fitting applied to differentiation, to integration and to the solution of differential equations. In chapter 2, some mathematical properties are studied and the mathematical theory of exponential fitting is presented. 

To compute unsteady flows, the time derivative terms in the governing equations need to be discretized. The major difference in the space and time co-ordinates lies in the direction of influence. In unsteady flows, there is no backward influence. The governing equations for unsteady flows are, therefore, parabolic in time. Therefore, essentially all the numerical methods advance in time, in a step-by-step or marching approach. These methods are very similar to those applied for initial value problems (IVPs) of ordinary differential equations. In this section, some of the methods widely used in the context of the finite volume method are discussed. 

The method of lines involves converting the governing equation (equation (5.1)) to a system of coupled ordinary differential equations in time by applying finite difference approximations for the spatial derivatives. The governing equation (equation (5.1)) can be converted to its finite difference form as follows ... 

Steady state mass or heat transfer in solids and current distribution in electrochemical systems involve solving elliptic partial differential equations. The method of lines has not been used for elliptic partial differential equations to our knowledge. Schiesser and Silebi (1997)[1] added a time derivative to the steady state elliptic partial differential equation and applied finite differences in both x and y directions and then arrived at the steady state solution by waiting for the process to reach steady state. [2] When finite differences are applied only in the x direction, we arrive at a system of second order ordinary differential equations in y. Unfortunately, this is a coupled system of boundary value problems in y (boundary conditions defined at y = 0 and y = 1) and, hence, initial value problem solvers cannot be used to solve these boundary value problems directly. In this chapter, we introduce two methods to solve this system of boundary value problems. Both linear and nonlinear elliptic partial differential equations will be discussed in this chapter. We will present semianalytical solutions for linear elliptic partial differential equations and numerical solutions for nonlinear elliptic partial differential equations based on method of lines. 

Spatial coordinates and integrating numerically in time. In this chapter, we apply finite differences in one of the directions (x), convert the governing equation and boundary conditions in x to finite difference form. The resulting system of coupled nonlinear boundary values problems (second order ordinary differential equations in y) are then solved using Maple s dsolve numeric command for boundary value problems (see chapter 3.2.8). 

Linear first order hyperbolic partial differential equations are solved using Laplace transform techniques in this section. Hyperbolic partial differential equations are first order in the time variable and first order in the spatial variable. The method involves applying Laplace transform in the time variable to convert the partial differential equation to an ordinary differential equation in the Laplace domain. This becomes an initial value problem (IVP) in the spatial direction with s, the Laplace variable, as a parameter. The boundary conditions in x are converted to the Laplace domain and the differential equation in the Laplace domain is solved by using the techniques illustrated in chapter 2.1 for solving linear initial value problems. Once an analytical solution is obtained in the Laplace domain, the solution is inverted to the time domain to obtain the final analytical solution (in time and spatial coordinates). This is best illustrated with the following example. 

---