One 24 hour cycle won't do anythign too bad as long as the light isn't super strong... you'd see damage if it was right away.
DLI (daily light integral) is the key here.
looking at any of these factors alone without referencing all of them together is the wrong way to understand light. Area of coverage (relative to 1 m^2) hours of operational, rate of photons produced need to be considered together. These add up to DLI (simplified explanation. so read the wiki, get the gist. the math can be referenced with a DLI Table, so you don't have to memorize it).
35-40 DLI based on output of the light (spec sheet umol/s) is what matters. If you can give that over 12 hours or 18 hours, it doesn't matter, the same yield will result. So, if you do calculate this based on your light, you might find you only need a 16/8 cylce, or maybe due to heat you are better off running it at 80% power and 18/6 cycle etc etc.. More than one way to skin this cat, but it always should add up to 35-40 DLI -- whether autoflower or photoperiod. Some trial and error is needed, but this is a tight window to start.
if heat is an issue i'd work it out for 18hours per day and run the lights at night as much as you can. lights off during hottest 6 hours of day.
Running light longer while still giving same DLI will reduce heat released per second. Though maybe the heat is so extreme that avoiding midday completel with a 14/10 cycle might result in lower temps while lights are one... you do what is best for the context, but still add up to 35-40 dli.