Assembling a Trilinear Form

Question

Dear all,

I have a mixed function u, splitted in c, mu, n defined on a ME = mixedspace([V,V,V]) where V is a P1 space and v,w,t are the test functions.

Is there a way to assemble the matrix from
a = c* n* v* dx ?

If I do A = assemble (a) , it gives me errors.. of course.

Thanks,
Cristina

Comments

This works just fine for me using dolfin1.5. The form a you're trying to assemble will assemble into a vector, not a matrix. Could you be more specific (code example, error messages)?

---

mesh= RectangleMesh(-3, -3, 3, 3, 2, 2)
V    = FunctionSpace(mesh,"Lagrange",1)
ME   = MixedFunctionSpace([V,V,V])
u = TrialFunction(ME)
c,mu,n = split(u)
v,w, t = TestFunctions(ME)
a = c*n*v*dx
A = assemble(a)

Traceback (most recent call last):
File "versione3.py", line 234, in
A = assemble(a_n)
File "/Applications/FEniCS.app/Contents/Resources/lib/python2.7/site-packages/dolfin/fem/assembling.py", line 193, in assemble
..........
File "/Applications/FEniCS.app/Contents/Resources/lib/python2.7/site-packages/ufl/algorithms/check_arities.py", line 54, in product
raise ArityMismatch("Multiplying expressions with overlapping form argument number {0}, argument is {1}.".format(x.number(), x))
ufl.algorithms.check_arities.ArityMismatch: Multiplying expressions with overlapping form argument number 1, argument is v_1.

If I do u = Function(ME) it works and in gives me a vector as you suggested. The problem is that I should lump all the L2 product (in order to have the convergence of the loop) and I don't know how to do it. I have this term (cn, v) in my equation and I don't know how to treat it.
Do you know what is the meaning of A = assemble(cn0vdx), with n0 a given vectors? That obtained A is a matrix..

---

Is this a bug? I am having the same problem with dolfin 1.6