Smoothing scale improvements

[brbass]

Summary

• Adds a damping factor to the ideal-H iteration in both SPH and ASPH smoothing scales, which helps convergence for difficult initial distributions.
• Adds a check in SpheralController that limits the initial neighbor count to a reasonable value before iterateIdealH begins.
• Supports extra packages during initial H iteration and a flag to force Voronoi tessellation.

---

ToDo :

• Annotate RELEASE_NOTES.md with notable changes.
• Create LLNLSpheral PR pointing at this branch. (PR#)
• LLNLSpheral PR has passed all tests.

[jmikeowen]

What is the purpose of this method? It appears to be forcing H to be smaller if there are more points than would be implied if the kernelExtent was 1 and points are perfectly arranged. Not all kernels have kernelExtent=1 though.

[brbass]

If you start off with a large density gradient, for instance, a single point can encompass tens of thousands of others. This starts of iterateIdealH with a reasonable number of neighbors. If you don't think this function will be generally useful, we can remove it. If we want to keep it, then we should include the kernelExtent.