Review all density effect correction files distributed with EGSnrc

[ftessier]

We ought to verify all the *.density density effect correction files distributed with EGSnrc, under HEN_HOUSE/pegs4/density_corrections/, in light of the update to the estar.f code for ICRU Report 90. Also, in recent years we found that the NIST web interface to generate ESTAR data suffered some issues, specifically in regards to the user input for the mean excitation energy (I-value). It is unclear how the EGSnrc density correction files were generated originally, so at any rate a full review is warranted. Updated data should be consistent with ICRU Report 37, ICRU Report 90, and otherwise generated with the estar.f Fortran program, icru90 version.

See related #573 and #705 as well.

[ftessier]

@bourgouina reminded me that one broad change in the icru90 version of estar.f done by @mainegra was the removal of the 13% boost in the I-value for mixtures (it only applies to molecular compounds). This was confirmed through discussion with the ESTAR code authors. So indeed most of the mixtures in the database will be affected. I also think we should really add the estar code to the repository to allow anyone to generate their own density files (I just have to sort out any licence or copyright issue!).

[ftessier]

@mainegra if we can add estar to the repo, then I would suggest integrating the commit history for estar, and then just develop it further inside the EGSnrc repo. Another one for the chicken coop/! 🤓 . In time of course, ideally the density effect corrections would be calculated on the fly, like for pegsless runs!

[mainegra]

As a first step is great. But I think it should be integrated to EGSnrc's data generation mechanism.