OpenFOAM 8 janafThermo 'out of temperature range' warnings can fill a disk (GBs per run) and cap T silently

Symptom: on a 30 GB cloud VM running several 2-D compressible cases (detonationFoam V2.0 on OpenFOAM 8, about 17k cells each, 3 species with GRI-Mech 3.0 JANAF thermo), the root disk hit 100% and apt-get failed with 'not enough free space'. du showed each case directory at 2 to 3 GB. Almost all of it was log.run.

Cause: GRI-Mech 3.0 thermo polynomials end at Thigh = 3500 K for many species. Any cell hotter than that (post-shock burnt gas at 3,800 to 4,300 K here) prints a multi-line FOAM Warning from janafThermo::limit ('attempt to use janafThermo<...> out of temperature range 200 -> 3500; T = ...') on every evaluation of every step. It also slows the run, and the reported temperature is clamped: the solver printed min/max(T) = 200, 3500, so hot-gas properties were evaluated at 3500 K.

Fixes: - Use thermo data valid to 5000 or 6000 K for the hot species (for example Burcat/NASA 9-coefficient fits refit to JANAF ranges), or raise Thigh only where the polynomials stay physical. - Redirect solver output through a filter, for example solver 2>&1 | grep -v -A3 'janafThermo' > log.run, or write only every Nth step, so the log cannot grow unbounded. - Watch df -h during multi-case sweeps on small cloud disks. If results are already post-processed, it is safe to truncate the logs: keep the tail with tail -c 20000 log.run > log.run.tail; rm log.run. - When comparing geometries, note that the cap biases every case alike. Absolute hot-gas temperatures above Thigh are not trustworthy.