I am solving the advection equation which is hyperbolic. Some basic techniques (shown on the page  http://www.cse.illinois.edu/iem/pde/discadvc/) are either of large oscillation or numerical diffusion.

I am striving to find a method without or with little of these problems. Or is there any post correction method to use?

Thank you all!

What is the best way to numerically solve the transport equation like(the least numerical diffusion phenomenon):

df/dt+v df/dx=0

 Thank you

Two-point tensor appears in some of continnuum mechanics reference books.

1. In wikipedia, two-point tensor is interperated as:" Two-point tensors, or double vectors, are tensor-like quantities which ... " 

2.Physically speaking, a tensor is a quantity that should not be changed by coordinate transformation while two-point tensor seems to have essential relationship with coordinates.

This is a basic question regard to elsticity fem.

How to compute traction boundary integration numerically in elasticity fem under general situation? That is:

\int_{\Gamma} [N][t] d\Gamma,   here [N] is shape function and [t] is surface traction

 My E-mail: shengxu.xia@yahoo.com

Thanks a lot

Is there any available fem source code for plasticity theory?

 Thanks alot!

what are the books/papers that give the basic idea about dislocation-crystal plasticity theory?

the problem described by these equations is: at time=0, N particles are dropped onto an infinite plane to diffuse. C(r,t) represents the concentration density of particles

The solution is very simple, but I want to see the procedure.

Thank you very much!

My own DEM(Discrete Element Method) program's solution for static problem seems strange.

I use my program to calculate the displacement of a cantilever under static pressure:

1. it seems my program shows convergence for static problem as the mesh gets denser. As the mesh grows denser(that means the particle becomes smaller), the displacement(of y direction)  of the free tip becomes smaller.

2. if the particle is big, the displacement of the free tip is greater than the theoretical solution, but if the particle is small enough,  the displacement gradully becomes smaller than the theoretical solution.

Is there any computing module which can compute a heap of disk elements in DEM(discrete element method)?it'd be better if it is written in matlab or C/C++. I'm doing research on the combined DEM-FEM. I need such module so as to save my time.

Thank you