Curvilinear boundaries

One category of finite-difference method uses a rectangular grid. In this approach one covers the specified layout with a grid, or mesh, as shown in Figure i 15.2a. When curvilinear boundaries are involved, it is possible to sample the... 

Assume the transverse cut of the curvilinear boundary of Figure 4.9, appearing in the secondary mesh, with n = hu, hv, T a normal unit vector in the (u, v, w) general coordinate system. The covariant components of E 1 = and fields... 

The first of these is the Reynolds equation in its simplest form the second equation covers the variation of viscosity as a function of pressure the third expression gives the film thickness, where R is the radius of the equivalent cylinder and i j is the combined displacement of the two solid boundaries. The equivalent cylinder treatment is a way of generalizing and simplifying the geometry of curvilinear boundaries if x is small enough relative to R,... 

For azeotropic mixtures, the main difficulty of the solution of the task of synthesis consists not in the multiplicity of feasible sequences of columns and complexes but in the necessity for the determination of feasible splits in each potential column or in the complex. The questions of synthesis of separation flowsheets for azeotropic mixtures were investigated in a great number of works. But these works mainly concern three-component mixtures and splits at infinite reflux. In a small number of works, mixtures with a larger number of components are considered however, in these works, the discussion is limited to the identification of splits at infinite reflux and linear boundaries between distillation regions Reg° . Yet, it is important to identify all feasible splits, not only the spUts feasible in simple columns at infinite reflux and at linear boundaries between distillation regions. It is important, in particular, to identify the spUts feasible in simple columns at finite reflux and curvilinear boundaries between distillation regions and also the splits feasible only in three-section columns of extractive distillation. 

A three-dimensional body limited by two curvilinear surfaces is called a shell if a distance called a thickness of the shell between the afore mentioned surfaces is small enough. We assume that the thickness is the constant 2h > 0. The surface equidistant from the surfaces is called a mid-surface. Thus, a shell can be uniquely defined introducing a mid-surface, a thickness and a boundary contour. 

In this section we analyse the contact problem for a curvilinear Timoshenko rod. The plastic yield condition will depend just on the moments m. We shall prove that the solution of the problem satisfies all original boundary conditions, i.e., in contrast to the preceding section, we prove existence of the solution to the original boundary value problem. 

Subdivision or discretization of the flow domain into cells or elements. There are methods, called boundary element methods, in which the surface of the flow domain, rather than the volume, is discretized, but the vast majority of CFD work uses volume discretization. Discretization produces a set of grid lines or cuives which define a mesh and a set of nodes at which the flow variables are to be calculated. The equations of motion are solved approximately on a domain defined by the grid. Curvilinear or body-fitted coordinate system grids may be used to ensure that the discretized domain accurately represents the true problem domain. 

In a suspended carbon nanotube, in addition to the purely Coulomb energy, we also have the nanomechanical corrections. Generally, these corrections make the relations between V and Vg, which describe the boundaries of Coulomb blockade regions, non-linear. Consequently, the Coulomb diamonds in suspended nanotubes are not diamonds any more, but instead have a curvilinear shape (with the exception of the case Cr = Cr = 0). Their size is also not the same and decreases with Vg - Thus, the mechanical degrees of freedom affect the Coulomb blockade diamonds. However, since these effects originate from the nanomechanical term which is typically a small correction, its influence on Coulomb diamonds is small as well. For the E-nanotube, these effects do not exceed several percents for typical gate voltages. 

The well-known Princeton model with a vertical -coordinate, a curvilinear horizontal grid adapted to the coastline, a turbulent closure of the order of 2.5 was used for the studies of the BSGC in [58]. Eighteen levels were specified over the vertical and the horizontal spacing was about 10 km. Similarly to [48], various combinations of the surface boundary conditions were specified. The model started with the wintertime climatic temperature analysis salinity fields [11] and three years later reached a quasi-stationary regime in the upper 200-m layer. 

Another possibility for modeling packed-bed reactors involves the use of a so-called unit cell approach where a suitable periodic structure in the packing is identified and subsequently used to define the boundaries of the computational domain. Due to the geometrical complexity the fluid flow (and other relevant equations have to be formulated and solved in curvilinear coordinates. In fact this approach has been followed for example, by Guj and De Matteis (1986) who used a MAC-like scheme (Welch et ai, 1965) to solve the Navier-Stokes equations. For random packings the unit cell approach becomes much more difficult due to the fact that a suitable periodic structure is difficult to define. 

Figure 8 Boundary conditions and curvilinear structured meshing.

Consider current-potential distribution in a curvilinear Hull cell (Subramanian and White, 1999 Chapman, 1997 ). The governing equations and boundary conditions are ... 

Figure 3.1 Schematic diagram showing concentration boundary layer surrounding u cylinder (or sphere) placed in a flow carrying diffusing. small particles. Curvilinear coordinate. v, taken parallel to the surface, is measured from the forward stagnation point A. Particle concentration rises froin zero at / = (I +f(p almost to the mainstream concentration (for example, to 997r of the niainstrcani value) at the edge of the boundary layer.

The solution of these equations by means of standard eigenfunction expansions can be carried out for any curvilinear, orthogonal coordinate system for which the Laplacian operator V2 is separable. Of course, the most appropriate coordinate system for a particular application will depend on the boundary geometry. In this section we briefly consider the most common cases for 2D flows of Cartesian and circular cylindrical coordinates. 

An analytic solution for such a problem can thus be sought as a superposition of separable solutions of this equation in any orthogonal curvilinear coordinate system. The most convenient coordinate system for a particular problem is dictated by the geometry of the boundaries. As a general rule, at least one of the flow boundaries should coincide with a coordinate surface. Thus, if we consider an axisymmetric coordinate system (f, rj, ), then either f = const or r] = const should correspond to one of the boundaries of the flow domain. 

Because this inner boundary-layer region is infinitesimal in thickness relative to a, all curvature terms that appear when the equations of motion are expressed in curvilinear coordinates will drop out to first order in Re thus leaving boundary-layer equations that are effectively expressed in terms of a local Cartesian coordinate system. 

An approach somewhat related to the broken-path method, but more accurate, has been employed by Johnson (1972). It also makes use of curvilinear coordinates (s, p) chosen with constant curvature, and divides the intermediate region into several sectors. In each of these the variables s, p are separable. Calculations were done with the amplitude density technique, matching functions and derivatives at each sector s boundary. The potential was that of Porter and Karplus, and the local vibrational motion was assumed to be harmonic. Good agreement was found with Diestler s at low energies. 

---