This can not be used in adaptive simulations though,
because the ASMu2D::constrainNode method is not implemented yet.
Removed the virtual method setQuadratureRule in SIM[123]D.
Instead a common implementation is added in SIMbase.
git-svn-id: http://svn.sintef.no/trondheim/IFEM/trunk@1560 e10b68d5-8a6e-419e-a041-bce267b0401d
<boundaryconditions> tag. The new tag takes all the old formats for
specifying surface pressure variations in addition to expressions.
This makes the <constantpressure> and <linearpressure> tags in the
elasticity solver obsolete and have therefore been removed.
Function expressions are now also available for Dirichlet conditions.
git-svn-id: http://svn.sintef.no/trondheim/IFEM/trunk@1558 e10b68d5-8a6e-419e-a041-bce267b0401d
Added the possibility to also include the time ("t") in the expressions.
Added some more function parsing methods in Functions.C
git-svn-id: http://svn.sintef.no/trondheim/IFEM/trunk@1557 e10b68d5-8a6e-419e-a041-bce267b0401d
This also includes most of the nonlinear formulation options
previously only available as command-line arguments.
The old command-line options are still supported, however, if
also specified on the input file, the command-line options will
be overrridden (this in contrast to the more general options).
git-svn-id: http://svn.sintef.no/trondheim/IFEM/trunk@1556 e10b68d5-8a6e-419e-a041-bce267b0401d
This is to allow for specification of the axiSymmetry and plane strain
options on the XML input file instead of as command-line options.
git-svn-id: http://svn.sintef.no/trondheim/IFEM/trunk@1555 e10b68d5-8a6e-419e-a041-bce267b0401d
Added a virtual getPrioritizedTags method in SIMbase scope, since that is
where the keywords involved are known. Removed pointless invokation of the
handlePriorityTags method from within the application-specific blocks.
git-svn-id: http://svn.sintef.no/trondheim/IFEM/trunk@1553 e10b68d5-8a6e-419e-a041-bce267b0401d
simplifies the cmake rule for >= 3.2, while keeping the old for version
< 3.2. hopefully the old block can be removed in not too far a future
git-svn-id: http://svn.sintef.no/trondheim/IFEM/trunk@1547 e10b68d5-8a6e-419e-a041-bce267b0401d
Use indices instead of duplicating the for loops.
Use .resize() instead of incrementally resizing vectors through push_back.
Exploit the tensor product nature of the algorithm to reduce code duplication
git-svn-id: http://svn.sintef.no/trondheim/IFEM/trunk@1541 e10b68d5-8a6e-419e-a041-bce267b0401d
introduce a perknot variable instead. this is the true meaning
of the entity and we avoid the confusing increase/decrease cycle
git-svn-id: http://svn.sintef.no/trondheim/IFEM/trunk@1540 e10b68d5-8a6e-419e-a041-bce267b0401d
1) encountering a partitioning with a different number of processors is
not an error. we should just silently continue. this allows having
several different partitionings specified in the file
2) a refactoring missed the fact that nGlPatches is read from the
partitioning, thus all nodes have to process all lines, not only its part
git-svn-id: http://svn.sintef.no/trondheim/IFEM/trunk@1529 e10b68d5-8a6e-419e-a041-bce267b0401d
Fixed: Index calculation for refinement error and printout of global norms.
This also reverts the previous fix r1517 which I don't think was correct.
git-svn-id: http://svn.sintef.no/trondheim/IFEM/trunk@1518 e10b68d5-8a6e-419e-a041-bce267b0401d
this is not a proper solution, but since we have these evil globals and
no virtual interface in ASMbase, this will have to do for now
this broke adaptive simulations
git-svn-id: http://svn.sintef.no/trondheim/IFEM/trunk@1509 e10b68d5-8a6e-419e-a041-bce267b0401d
