How much grad students 'should' work (other than the 20 hours of teaching/research on which their salary is based), is one of those unanswerable questions that varies with the student and the research and whether the equipment gods are happy and so on. I've known and advised grad students who worked 9-5 (or 8-6), had a life/family, and did quite well. This is very rare; these people are super-efficient and have well-defined projects that don't rely on balky equipment and less efficient co-workers. Much more common are research projects that require some (to a lot) of night/weekend time. There are also many examples of students who tried to work 9-5, weren't super-efficient, and who flamed out because of their lack of progress with research.
With the exception of one of my current students, I work more hours than any of the others, and (with one other exception) this is OK with me. I have one student who works an insane amount because he wants to, but (with the one other exception) my other students have more of a balance between time at the department and time off campus. They are always telling me about concerts or crazy parties they went to, or a weekend trip they took, or a hike/bike-ride they went on, or a non-science book that they read, and I like hearing about their other interests. They are doing interesting and productive research and they know I am satisfied/happy with their work.
As for the one who isn't currently working much or well, we have an appointment to discuss the situation next week: is it a time-management problem, family/health etc. problem, lack of interest, lack of something else? I can deal with some of those in terms of adjusting the student's research program and/or my expectations. If it's a lack of interest or motivation, I want to know sooner rather than later.
4 hours ago