Better choice of bracketing constant in the EAGLE cooling bisection code.

examples/CoolingRates/README

@@ -11,8 +11,9 @@ The test may be run by:
 python plot_cooling_rates.py
 
 where X is the redshift at which the cooling rates are evaluated and Y
-is the base 10 logarithm of the hydrogen number density. Different
-metallicities may be specified in testCooling.yml
+is the base 10 logarithm of the hydrogen number density expressed in
+cgs (i.e. cm^-3). Different metallicities may be specified in
+testCooling.yml
 
 Running with -z 3 -d -4 should reproduce the Fig.4 of Wiersma+09 with
 the added Compton cooling contribution.

src/cooling/EAGLE/cooling.c

@@ -56,7 +56,7 @@ static const float explicit_tolerance = 0.05;
 static const float newton_tolerance = 1.0e-4;
 static const float bisection_tolerance = 1.0e-6;
 static const float rounding_tolerance = 1.0e-4;
-static const double bracket_factor = 1.0488088481701515; /* sqrt(1.1) */
+static const double bracket_factor = 1.5; /* sqrt(1.1) */
 static const double newton_log_u_guess_cgs = 12.30103;   /* log10(2e12) */
 
 /**