Prohibit burning in shock zones by default

[brady-ryan]

Prohibiting burning in the numerically broadened shock structure is a common prescription in stellar detonation simulations (some double detonation examples: Townsley 2016, Shen 2018, and Rivas 2022). This practice stems back to Fryxell 1989, with a robust analysis using a Type Ia domain in Papatheodore 2014. I have also probed this treatment in 1D using Castro's subch_planar problem setup and have attached an image of the results at t = 0.180s, where we can see that the flames converge with increasing resolution towards a prescription that prohibits burning in zones containing the shock front. The mechanism in which Castro uses to detect shocks can be seen in the Appendix of Zingale 2024. At the moment, Castro does not prohibit shock burning by default, which perhaps should be changed to ensure accurate nuclear modeling.