<br>Hi Paul,<div>It is possible to perform the online phase of rom approach in nek5000? After the computation of pod basis, for example,</div><div> The reduced stiffness matrix will be of form Z^T A Z, Where A is the stiffness matrix of full discretization and parameter ind., Z is the matrix obtain columnwise by pod basis. That implies to do not extract the stiffness matrix, but to perform a new directed computation of reduced system inside nek5000 and select the parameter by datafile to obtain the online snapshot.</div>
<div>In this case the reduced stiffness matrix will be , of course dense, but cheaper. However , it is needed to work around in order to have ah matrix component and perform pointwise the tensor product z^t a z ( maybe with routine mxm). Maybe with new rb-hybrid approch it is possible to extend on the A constructed by block refering each one to a physical domain. </div>
<div>Best regard</div><div>Davide</div><div><br></div><div><br>Il martedì 18 febbraio 2014, <<a href="mailto:nek5000-users@lists.mcs.anl.gov">nek5000-users@lists.mcs.anl.gov</a>> ha scritto:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
Hi Giuseppe,<br>
<br>
It occurs to me that the right way to address your<br>
problem is to dump out the unassembled matrices, which<br>
are all block diagonal, with full blocks. Then, in addition,<br>
it's easy to write out the matrix that assembles the submatrices<br>
into the full sparse matrix. That part should be relatively<br>
easy to handle in a framework that is designed to work with<br>
the global index set. (Nek doesn't deal with the global<br>
indices and for the size matrix you want, in parallel, it<br>
wouldn't be easy to generate them.)<br>
<br>
So, basically, Nek would produce<br>
<br>
A_L = block_diag {A^e} _{e=1}^E<br>
<br>
and Boolean assembly matrix Q such that<br>
<br>
A = Q^T A_L Q<br>
<br>
The elemental matrices, A^e, are completely full. Thus, for<br>
N=9, corresponding to 10 x 10 x 10 = 1000 points in a given<br>
element, you would have 1 million nonzeros in each matrix.<br>
(For undeformed geometries, some of the matrices are sparser.)<br>
The matrix Q^T is rectangular and consists of columns of the<br>
identity matrix. (See, e.g., Deville, F., & Mund, 2002).<br>
It probably wouldn't take much effort to code up the output<br>
routines for this plus some matlab code to demo how to<br>
assemble the stiffness matrix.<br>
<br>
Paul<br>
<br>
<br>
On Thu, 13 Feb 2014, <a>nek5000-users@lists.mcs.anl.gov</a> wrote:<br>
<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Dear users and developers,<br>
I am using NEK5000 for a 3d unsteady simulation with mixed Dirichlet and<br>
periodic bc. For the post-processing (a POD-based dynamics) I need the global<br>
mass and stiffness matrices as built by NEK5000 on my mesh. I would also need<br>
the matrix having as entries (phi_i, grad(phi_j)) where phi are the basis<br>
polynomials. Is it possible to have NEK build these matrices and then save<br>
them on file? The ultimate goal is to import them on a PETSc program to<br>
perform some algebraic manipulations. I already found a way to import the<br>
simulations' results. Surfing the code, I have found some 1-d routines, but<br>
I don't know how to extend them to my needs. Thank you in advance for any<br>
help or hint.<br>
Best regards,<br>
Giuseppe<br>
<br>
______________________________<u></u>_________________<br>
Nek5000-users mailing list<br>
<a>Nek5000-users@lists.mcs.anl.gov</a><br>
<a href="https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users" target="_blank">https://lists.mcs.anl.gov/<u></u>mailman/listinfo/nek5000-users</a><br>
<br>
</blockquote>
______________________________<u></u>_________________<br>
Nek5000-users mailing list<br>
<a>Nek5000-users@lists.mcs.anl.gov</a><br>
<a href="https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users" target="_blank">https://lists.mcs.anl.gov/<u></u>mailman/listinfo/nek5000-users</a><br>
</blockquote></div><br><br>-- <br><div dir="ltr"><div>Davide Baroli, PhD student</div><div>MOX - Modeling and Scientific Computing</div><div>Mathematics Dept.</div><div>Politecnico di Milano</div><div>Via Bonardi 9, 20133 Milano, Italy</div>
</div><br>