Monday, June 24, 2013

Timescales, averaging, and baseball

Please pardon the summer blogging slowdown - it's been a surprisingly busy couple of weeks, between an instructor search, working on papers, proposal stuff, and trying to write more on my big long-term project. 

Thanks to an old friend for pointing me to this link, which does a great job looking at why a knuckleball is so erratic in its flight from pitcher to batter.  For non-Americans:  In baseball, a pitcher throws a ball to a catcher, while a batter attempts to hit the ball.  There are several types of pitches, depending on the pitcher's grip on the ball (which has seams due to the stitching that holds the leather cover on), the throwing motion, and the release.  A fastball can reach speeds in excess of 100 mph (161 kph) and typically spins more than 1000 rpm.  In contrast, a knuckleball can drift by the batter at a leisurely 70 mph yet be nearly unhittable because of its erratic motion.   A knuckleball barely spins, so that it may complete only 1-2 revolutions from leaving the pitcher's hand to reaching the batter.  This means that the positioning of the seams is absolutely critical to determing the aerodynamics of the motion, and no two knuckleballs move the same way.  In physics lingo, a knuckleball has almost none of the orientational averaging that happens in basically every other pitch.  I propose the definition of a new dimensionless parameter, the Wakefield number, \(W\), that is the ratio of the ball's period of revolution to its time-of-flight from pitcher to batter.   A knuckleball is a pitch with \(W \sim 1\).


paris said...

Interesting professor! by the way, can you recommend an updated replacement for Ashcoft and Merrmin?
I need a good solid state physics book written for this century.

Anonymous said...

Sounds interesting. Emailed this idea to Alan Nathan yet? Also, I bet Professor Tezduyar in Rice's MechE department would have some ideas about it.

Anonymous said...

Since Doug hasn't answered yet, let me propose a provocative but (in my experience) widely accepted claim: Ashcroft & Mermin is abysmal dreck, but in a weird Goedelian twist there can exist *no* solid state book which is both good AND semi-complete.

Certainly not all of the modern research in mature subfields (particles, CM, AMO, for example) can be captured in textbook format, but the intellectual foundation and framework of say, particle physics, lives in Peskin & Schroeder -- which is also a very good book. A unifying language for ALL of condensed matter (that is useful both to future experimenters and theorizers) just doesn't exist. Maybe it's silly to pretend it does.

My semiconductor friends swear by Yu & Cardona. My budding young CM field theorists learn to recite chapter and verse of Chaiken & Lubensky and then Fazekas with bits of Auerbach. I'm sure they'll be fine.

DanM said...

My comment runs along different lines: the Wakefield!?! That's nearly heresy. If you're gonna pick a knuckler, pick a good one; and also one who did not pitch for Evil Incarnate. May I suggest Charlie Hough? Then baseball could have its own H index.

Douglas Natelson said...

DanM, how about Niekro? Statistically the best knuckleballer in MLB history. Then we could have the N index.

Douglas Natelson said...

Anon@3:55, well, you were provocative. I quite like A&M (except for the fact that it ends in 1975). Calling it abysmal and dreck is waaaaay over the top. There are far worse books. More modern alternatives include Ibach and Luth, and Marder; interestingly, both of those had very poor editing and high densities of typos in their first printings, which have now been corrected. Chaikin and Lubensky does a great job summarizing statistical mechanics in Ch. 3, and convinced me that liquid crystal phases are real, honest-to-goodness thermodynamic phases (something about which I'd been skeptical when a student).