Handout (153.5 kB)
solution of Weng-Taylor turbulence model. In the Weng-Taylor
model the eddy viscosity coefficient is defined from additional
phenomenological consideration, which constitutes a turbulence
closure, and nonlinearly depend on velocities. In such type models
sharp vertical boundary layers causes difficulties for traditional
numerical methods. In this work a new numerical method is
proposed, based on analytical representation of Weng-Taylor model
solutions. It is shown that these analytical solutions of
constituted initial boundary value problem can be resolved by
additional solutions of system of ordinary differential equations.
This system of equations is solved numerically, by using
exponential/polinomial type substitution for turbulent fluxes. The
obtained numerical solution, precision and effectiveness are
compared to solution by using our numerical propagator method.
The propagator method exploits an approach of a non-standard
representation of the time derivative, by applying the derivative
to the solution given as the product of two functions, where one
of them is a propagator function. Propagator function is chosen in
nonlocal way, and with respect to the solution, a new finite
volume difference scheme is obtained. Stability of the scheme is
investigated, by using analytical solutions/estimations of
Weng-Taylor model. It is shown, that stability restrictions for
the propagator scheme becomes more weaker comparison to
traditional semi-implicit difference schemes. There are some
regions of the model coefficients, for which elaborated propagator
difference scheme becomes absolutely stable. It is proven that the
scheme is unconditionally monotonic. The scheme has the first
order in time and the second order truncation errors in space.