In a gravity-less world without friction, starting with a stationary needle, would one be able to make it spin by first flicking it, to cause it to twirl (so that its axis of rotation is perpendicular to its long axis), and then tap its ends repeatedly in some way so as to cause the twirling motion to transform into spinning (i.e., so that the axis of rotation is lined up with the needle's axis)? (I posted a long and somewhat involved variant of this question on the "homework/introductory physics" section over a week ago -- though it's not a homework problem -- without getting any response. It was titled "A simple rotational problem." I thought that posing it in this simpler way, I might at least get a comment or two.) One possibility would seem to be to have the twirling needle be scooped up by the wide end of an accelerating funnel, so that, in effect, the needle would be "going down the drain," making a series of reorienting, frictionless contacts with the sides of the funnel. Let's say the funnel approaches the needle so that its axis of symmetry is almost (but not quite) lined up with the needle's angular momentum axis (off only a bit, so that the needle doesn't get lodged in the funnel). The question is, when the needle comes out of the funnel (whose bottom is just barely wider than the needle), so that its own axis has been forced to rotate almost 90 degrees, so as to basically line up with its original angular momentum axis, will it be spinning? I say yes, and that it'd be spinning much faster than it twirled, owing to the much smaller moment of inertia in this orientation. But, I have an expert who disagrees with me (Professor Eugene Butikov, of the Univ. of St. Petersburg, Russia), who says the needle would have zero spin about its long axis, and would immediately go back to twirling, but at an angle determined by the last contact it makes as it leaves the funnel.