Why does the shower curtain blow in despite the water pushing it out (revisited)?
I am a professor at the University of Massachusetts-Amherst. I thought I should bring this to your attention. I think I have settled the shower curtain question decisively.
I knew we'd get to the bottom of it someday. By the way, I can appreciate your contacting me, Dave, even though your research has already been written up in the New York Times and the Boston Globe. It's all very well to have the adulation of the masses. But one wants the respect of one's peers.
The original question, first raised in this column more than 20 years ago, was why the shower curtain blows up and in when the pressure of the water is pushing it down and out. Ignorant parties may think: Like I care. But we can be certain these people would have tossed out that icky bread mold, too.
Initially I was reluctant to tackle the problem with so many other world crises clamoring for my attention. But eventually I allowed that the phenomenon could be attributed to the Bernoulli effect, the well-known principle that explains how airplanes fly: as the velocity of a fluid increases, its lateral pressure decreases. The water flowing out of the shower nozzle entrains the surrounding air, and the resulting decrease of pressure perpendicular to the direction of flow ( following this, are we?) pulls the shower curtain in.
I don't claim this answer represented any bold leap; it was one of two prevailing theories at the time. Predictably, I soon heard from adherents of the other, known as the chimney effect. They argued that air heated by the hot shower water rose and pulled cool air in from below, taking the curtain with it — convection, in other words. As proof, they advised running a cold shower, apparently believing that the curtain would remain motionless. I did as suggested. The shower curtain still blew up and in, though not quite as vigorously. OK, geniuses — you were saying?
At this point I consulted Jearl Walker, who, sad to say, muddied the waters. Jearl, a physicist who then authored the Amateur Scientist column in Scientific American, attributed shower curtain motion to the Coanda effect, the tendency of a fluid in motion to adhere to a surface or, in this case, the tendency of a surface (the curtain) to adhere to a fluid in motion (the flowing water). A bit of a reach, we may agree in retrospect, but when you've got a guy like Jearl, who plunged his hand into a pot of molten lead on the Johnny Carson show to demonstrate the Leidenfrost effect, you tend to believe what he says. Anyway, there the matter rested for lo these many years.
That brings us to Professor Schmidt, who decided, The devil with mere opinion, it's time for the facts. This being the 21st century, he pursued them using computational fluid-dynamics software from Fluent Inc., which you'll remember also played a pivotal role in unraveling the Guinness Stout conundrum, i.e., why the bubbles float down, not up.
I thought it curious that Fluent should figure in two of the great controversies of our time, so I did what any good journalist would have done. I asked Dave, ever so politely, whether he was on the take. His reply: (1) I am not currently on the payroll. (2) But you never know. (3) However, the main thing is, I'm buds with these guys, have drunk Guinness with them, helped develop the part of the software dealing with spray, which directly addresses the issue here, and besides, I know how to use this software and I don't know how to use the competition's, so off my case. Dunno if that's good enough for Sam Donaldson, but it's good enough for me.
Back to our story. Dave created a computer model that divided a shower into 50,000 tiny tetrahedral cells, with the water blasting at eight gallons a minute for 30 seconds. He then started his home PC chugging away on the necessary 1.5 trillion calculations (no lie), which took two weeks. He would have been done sooner except that he could only work on the job at night — he was running Linux on the computer and every morning his wife made him reboot.
But finally he got an answer, which, as it turns out, is none of the above. What happens is, the water spray creates a sideways vortex in the shower stall, like an undershot waterwheel. (One hopes Slug's drawing makes this reasonably clear.) The center of the vortex, like the eye of a hurricane, is a low-pressure area, which sucks in the shower curtain, somewhat in the manner of a centrifugal pump. (Dave and I argued about this analogy, but I'm convinced it's reasonably close.) So forget Bernoulli, chimneys, and the reverse Coanda effect — what it's really all about is a vortex. You'll sleep better tonight now that you know.