add nb with inno benchmark and time independent heat equation

[nickgoldring]

We have concluded our thermal simulation benchmarking with Innocenzi using a faster, time-independent formulation. The results are provided in the notebook rslaser_fenics_innocenzi_ti.ipynb found in the examples/thermal directory.

[bruhwiler]

@dtabell -- please review this PR today.

[dtabell]

@nickgoldring I see several items that you need to address:

• lower the mesh density from 120 to a value that will execute quickly; include a comment about why you want 120 for a “production run”
• the various section titles have inconsistent depths; for example, “Execute simulation” and “Innocenzi benchmarking” are both subsections of “Define simulation parameters”
• under “Innocenzi benchmarking”, after the first two graphics, the following three cells emit errors on np.savetxt, np.loadtxt, and then on the data not loaded
• in the section “Divide data into n longitudinal slices”, all graphics appear empty
• almost all sections/subsections/etc. would benefit from the addition of markdown comments that describe what you are doing in the given section/subsection/etc.
• in the section “ABCD matrix calculation”, I don't see the image <img src="figures/gaussian_duct_abcd_mat01.png">
• in the last section, all but the first couple of cells emit errors