Feedback on MATH35001 exam Jan 2014
-----------------------------------

Overall the paper was done quite well -- many students scored
extremely high marks, though there was also quite a "long tail" of
not-so-spectacular marks, and, sadly, a few students seemed to collapse
completely and obtained single-digit scores...

Q1: Common mistakes: forgetting that action = re-action, i.e. 
    traction on solid = - traction on fluid; not evaluating
    traction on boundary; not actually drawing any conclusion 
    from computing the curl of the fluid (and, again, not evaluating
    it on the boundary).

Q2: People tended to lose points because their argument as to why
    u = U f(eta) wasn't sufficiently rigorous. A fair number
    then tried to solve the variable coefficient ODE with an
    exp(lambda y) ansatz!

Q3: Derivation of parallel flow equations (with proper arguments 
    for each step) was generally done quite badly.

Q4: Dimensional arguments often rather incomplete (e.g.: yes, we 
    need 1/r "outside" the function, but why can it not appear
    "inside"; etc. As often with "show that" questions, people
    were sometimes a bit too quick to (allegedly) reach the
    result -- I do read the entire argument, you know...

Q5: A large number of students felt the need to derive a solution
    to the biharmonic equation from first principles even though 
    question (c) said explicitly to construct it from the "catalogue" 
    of solutions in cylindrical polars provided. They typically
    didn't get anywhere. Practially everybody who didn't fall for that
    hurdle tended to get lost in the ensuing algebra, but I tended
    to mark attempts at trawling through this generously.