Numerical integration vs. integrals that are zero

[Thomas Okken]

"Suspiciously" quick was a poor choice of words, but what I meant is that it terminates without attaining the requested estimated relative error, and without hitting some arbitrary maximum number of iterations. For integrals like these, Free42 keeps going until it reaches 20 iterations, and I'm looking for a strategy to make it terminate more quickly, but without changing the UI. (Mathematica lets you choose between specifying a relative or absolute error, which would take care of this problem, but the 42S has only the single parameter ACC.)