Nonlinear similarity matrices

Eqn (23) is a second order nonlinear difference equation the Jacobian of which is easily established as a regular tridiagonal matrix with a dominating diagonal, similar to system matrices found in the simulation of distillation columns. The analytical derivation of the Jacobian and the Newton-Raphson iteration is trivial. In figure 3 is shown an example where the intermediate pressures are plotted as functions of the total pressure drop across the line segment. The example is artificially chosen such that all e-parameters are the same, i.e. ... 

Linear prediction and state-space methods are grouped together here because, although the philosophy behind the two methods is different, the mathematics involved is very similar. In both cases, the data-fitting problem is made linear by constructing a matrix from the observed data points, and the model equation is then solved by linear means. The nonlinear model parameters are... 

Some interesting points can be made from Figure 6.3, for which the data of Docherty et al. (17) have been used and which shows the plot of the contributions of various excited states in the sum over states indicated in equation 7 to P ec for 4-amino-4 -nitro-frans-stilbene. First, most of the nonlinearity occurs because of the charge-transfer resonance associated with the lowest optical transition in the molecule. Second, the seventh excited state drastically reduces the nonlinearity of the molecule. Careful examination of the matrix elements is required to determine whether this reduction is caused by reverse charge transfer or by an unfavorable transition moment between two of the low-lying excited states. Third, the calculation converges with inclusion of only a fraction of the total number of excited states included in the calculation. Judicious use of the last observation could save tremendous amounts of computer time in evaluating classes of similar molecules. 

It follows from the preceding results that the electro-optical properties of molecules in degenerate electronic states should have unusual temperature dependence, which is absent in the case of nondegenerate states. Even for nondipolar degenerate electronic states (e.g., for states in which the reduced matrix elements of the dipole moment are zero) for certain relationships between the vibronic constant and the temperature, there may be a quadratic dependence of the Kerr effect on p, similar to that observed in the case of molecules that are simultaneously anisotropic polarizable and possess a proper dipole moment. The nonlinear dependence on p under consideration is due exclusively to the vibronic interaction that redetermines the vibronic spectrum and leads to different polarizability in different vibronic states. This dependence on p has to be distinguished from that which arises due to the nonzero value of the dipole moment in the initial ground electronic state (e.g., as in the case of the E term in molecules with D3h symmetry). The two sources of the... 

As discussed earlier, while the scale of the fillers is substantially different, nanocomposite materials concepts and technology are very similar to those of conventional composite materials. This is clearly demonstrated in the case of new thermosets for nonlinear optical (NLO) applications, " " where nanocomposite of liquid crystalline thermosets, IPNs, and simple filled thermosets are evaluated. Tripathy et al. discussed four different ways to prepare nonlinear optical polymers. (1) The polymer matrix is doped with NLO moieties in a guest/host system (2) In side-chain polymer systems, NLO polymers with active moieties are covalently bonded as pendant groups (3) In the main chain polymer, the chromo-phores are incorporated as parts of the main polymer backbone to enhance the temporal stability of the NLO properties and (4) Stability of the optical noninearity in sol-gel-based thermosets is related to... 

In the previous sections, we derived general correlation function expressions for the nonlinear response function that allow us to calculate any 4WM process. The final results were recast as a product of Liouville space operators [Eqs. (49) and (53)], or in terms of the four-time correlation function of the dipole operator [Eq. (57)]. We then developed the factorization approximation [Eqs. (60) and (63)], which simplifies these expressions considerably. In this section, we shall consider the problem of spontaneous Raman and fluorescence spectroscopy. General formal expressions analogous to those obtained for 4WM will be derived. This will enable us to treat both experiments in a similar fashion and compare their information content. We shall start with the ordinary absorption lineshape. Consider our system interacting with a stationary monochromatic electromagnetic field with frequency w. The total initial density matrix is given by... 

It follows from the definition (4.16) that, for any real value of the parameter a, the operator t is self-adjoint and the matrix (4,32) is hermitean. It is evident that the matrix < ] 11 f> > in (4.32) for z = E in some way must be similar and sometimes identical to the Bloch matrix IH I <( p>, which is easily verified by considering the power series expansions in V. It is then also clear that, as an alternative to the nonlinear Schrodinger equation H0 = 0H0, one may use multi-dimensional partitioning technique, which treats the eigenvalue problem firom a rather different point of view. It is also applicable to the case when -instead of an exact degeneracy of order p - one has p close-lying unperturbed reference states It should also be remembered that,... 

Controllability of Linear Systems It is possible to determine if a system of linear differential equations is controllable or not. Although reactive systems found in AR theory are generally nonlinear, the underlying concepts are similar and shall be useful for later discussions. In 1959, Rudolf Kalman showed that specifically for a linear, time-invariant system, it is possible to determine whether a system is controllable by computing the rank of a special controllability block matrix, E (Kalman, 1959)... 

---