Some people have success by treating the academic job like a 9-5 office job. You simply come to the office at 9 and leave at 5, teaching your classes when they are scheduled, going to meetings, and spending the rest of the time at your desk grading papers, preparing class, and writing your articles. This only works if you are actually working during those hours and actively schedule research time. Also, you must not limit your 9-5 schedule to the 15-week semester. You'll want to do this at least 45 or 50 weeks of the year, not 30.
The most common way of scheduling time in academia is the spill-over model, where you might not be working all of the working day, but you are also working until late at night, or most of the time over a weekend, etc... The disadvantage of this model is that you are never not supposed to be working, and yet you might not be organizing the hours very efficiently or keeping track of how much time you are spending.
My current schedule is to put in 10 or 11 hour days on T/R, work about 7-8 on Wednesdays, and sporadically the rest of the days in two or three hour segments. That seems to work out.