Simply showing up for class is not a sufficient condition for learning, especially in a difficult subject like engineering.  For this reason I place a strong emphasis on homework in my teaching and grading.  I believe that the combination of lectures and working out problems on your own is simply the best learning method that has been invented.  Its not perfect, of course, but it works better than the approach preferred by most students, which might be called "iPod-assisted osmosis-in-absentia".  While I certainly encourage discussion of homework problems with me, the TAs, or fellow students, there is little chance for true learning unless you read the book(s) and struggle with the problems individually.  My job as a teacher, along with the TAs, is to help and guide your learning process.  But without your active participation you will, quite frankly, just waste your time and a rather large amount of someone's money on this course. 

Note my emphasis on the phrase "intensive, solitary struggle" in the quote above.  This process is not easy.  If you approach a homework assignment with the expectation of polishing it off it 10-15 minutes, you will not have success.  If the subject were that easy it wouldn't be taught in a University.  Nobody is born with this knowledge; it is acquired by dedicated and determined effort. 

For maximum benefit you must read the textbook and struggle with the homework in concert with lectures and discussions. Lectures, discussions, office hours, and homework due dates have been sequenced accordingly.  Therefore I have a zero-tolerance policy on late homework.  If it is late, it gets a zero, no excuses.  I've instructed the TAs not to accept homework anytime after the specified time on the due date.  Please note that if an assignment is due by 5:00 PM, that by definition means we expect it to be turned in before 5:00 PM, and thus it will be declared late if turned in at 5:00 PM. 

On the subject of excuses, any intelligent adult will already know that excuses rarely work.  In fact, merely attempting to offer an excuse in my course may result in a further loss of points unless it is truly meritorious.  To calibrate your senses, note that slow or broken watches, broken bike chains, oversleeping, travel delays, apartment evictions, jail time, deaths/illnesses of third parties, pet-related incidents, an "awesome swell" at Campus Point, the sudden breakup of an intimate relationship, non-fatal bike accidents, flying solder blobs affecting vision, outpatient medical issues, and parking problems are NOT valid excuses for missing or late assignments. (I didn't make these up, they have all been tried on me over the past 17 years.)  My advice is to channel your creative energies to figuring out how to finish the assignment and get it to me on time, rather than developing a creative excuse.   See "The dog ate my disk" for a colorful essay on the subject (I have yet to experience the "throwing up blood" defense). 

As far as grading is concerned, I try to insure that each homework and exam problem is graded by the same person for consistency.  But invariably I get many complaints about homework and exam grades.  Students often seem surprised to learn that sloppy work, or failure to clearly indicate the answer or show supporting evidence of original work, may result in especially poor grades.  Sometimes I even get complaints from people who readily acknowledge they did the problem wrong, but believe they deserve a better grade because of some perceived injustice in grading relative to other students in the class (so-and-so got 2 extra points, so I should too).  That is absurd.  There is no conspiracy here, the grade you get is simply a reflection of the work on the page, and is not a function of someone else's grade.  Furthermore, grading in engineering courses nearly absolute:you either solve the problem correctly, or you don't.  Engineers do not remain employed for very long if their solutions don't work.   In coursework we will occasionally allow for partial credit because some mistakes might be considered inconsequential if the person has clearly demonstrated a mastery of the underlying physics.  But it is dangerous to depend on partial credit, because 1) some problems are easier than others, so the expectation of solving it correct is proportionally higher; and most importantly 2) such credit is critically dependent on how well you documented your thought process on the page.  Graders are not clairvoyant; a messy scribble in the margins does not constitute evidence of work deserving partial credit.  So, unless there is a clear case of an oversight in grading (such as not adding up the points correctly, or the grader overlooking a problem altogether), beware that I have very little patience for grading complaints.

A last note regarding "grading on a curve" or "course scaling": I will if I have to, but it will never LOWER your score.   In other words, if everyone in the class does well I am not going to artificially force the class grades to fit a normal (Gaussian) distribution.   Quite frankly it seems that Mother Nature and/or our Office of Admissions seem to encourage that distribution, but personally I would like nothing better than to have a class with all A's.  It hasn't happened yet, but I'm still optimistic....