Key Biscayne

Key West is interesting for a little while and while I was there something happened that was right out of a Blake Edwards movie, so dreams do come true. It was nice. Since the last time I was there the place has gained and lost a reputation for boozing and drugs.

It used to be that no one who went there was just passing through, because there is nowhere farther to go. Jimmy Buffett and cruise ships changed all that. Now when the cruise ship hits the dock a thousand people come ashore and drink and shop like it was the Embarcardero in San Francisco. The whole north end of Duvall Street is shoulder to shoulder. Some of the other keys are kind of nice if they aren't too flat and don't have an RV Park. And John Mellencamp was recently sighted down there according to the paper. He did it right, is my guess.

Hunter Thompson killed himself a few days ago, he used to spend a lot of time in Key West. His writing is like the 60's, three layers of foliage and even if you could see through that, there is still no way to know if there is a tiger lurking underneath. They say he got the suicide from Hemingway who lived on Key West, but I think he probably had a terminal liver disease and wanted to end the pain.

Contrast Thompson with Ken Kesey's books. Kesey's early work is like a tour through the tropics at the garden club, and at the end they have a tiger that roars on command. I wanted to mention Kesey because of his "Merry Pranksters" book - which is about some people touring the country for no good reason, in the 1960s, when nobody did anything like that, and it was a suspicious thing to be doing.

Well anyway....Why is it that I think I ought to be writing about fluid dynamics. I went to engineering school and took a bunch of their classes but I didn't graduate. In fact I missed a few really basic classes in engineering. What I did instead was learn more math than any applied person ought to know, so if I can't read a fluids book to find out whats going on, then who can? Fluid dynamics is interesting, so my mind is made up.

After a lengthy college career I worked for a long time in computer software development, first as a technical writer, then for 4 years as a programmer, then for 5 years as a network engineer/tools/release expert, and finally working a few years in quality assurance. I met four really smart engineers, Jorge Wes Dave and Kersti, three really good managers, Eric Bryan and Tom, and a few other people with particularly outstanding skills like "getting things done in an organization - Stephanie", "teaching people how to get along - Woody", and "passing along knowledge to younger workers - Charlie" and "dreaming up something way ahead of its time - Steve". But one of the most interesting things I was exposed to was a story told to me one day by my uncle. My uncle Jim had started as a computer programmer way back in 1964, back in the day when for memory they used magnetic drums similar to ones in older analog Xerox machines.

Young Jim was the new kid on the block, all the other programmers had used the computer during the war, or had been involved in building the thing. So after a few days Jim got to go near the machine with an application that read the MICR magnetic encoding off of a stack of checks. He put the punch cards in the hopper and compiled the program, and he put the checks in the feeder. He told me he noticed people peeking at him while he did this. Then he pushed "Go" and the first check zipped along the runners and was routed correctly and then everything stopped. It turned out he forgot to allow for a mechanical delay in the mechanism that fed the checks. After the first check was read the computer forgot to wait for the next check to come by. It decided there were in fact, no more checks.

Putting in a delay for some mechanical operation is not an obvious thing to think of. You don't normally have to account for the fact that the computer is so much faster than the mechanical runners. Timing these kinds of things and getting used to the quirky nature of computer driven machinery led to the science of "systems" which is what I was supposed to learn in college.

I learned a lot about it, and years went by and I was talking to my Uncle again, and he exclaimed that I was talking about something he was familiar with. I was just talking about ways to eliminate wobble in car tires at high speeds, or how radios sometimes make squelching noises. I was talking about feedback and automatic control, and for me those were easy topics, but to technical people in my uncle's generation it was not something he really understood. He was familiar with some people who built a high pressure valve that was regulated with a rotating weight system. These kind of rotating governors were invented for use on steam engines, and they also made cotton gins practical. But my Uncle Jim's reaction told me that these things were a mystery to him, while for me I was familiar with all the quirky ways the governor could interact with the engine it was controlling. That is what systems is all about - the interactions between complicated things.

A lot of people will tell you systems is about viewing things like a black box, and measuring inputs-outputs, or drawing charts of flows, or measuring rates and levels. Those things are good tools but it might be that "interactions between complicated things" is the best definition of systems. In Jim's check processing example the interaction broke down because the computer was working in milliseconds and the check feeding mechanism was working in tenths of seconds. There needed to be synchronization, and Jim accomplished this by making the computer wait a fixed amount of time. This isn't "synchronization" but it solved the problem by making sure the check was waiting there before the computer looked for it. It could be foreseen that maybe there is a case where two checks are waiting, but thats another lecture!

Now step back and see if there is another way to look at the system Uncle Jim was in. There was also interaction between people and the computer. Not much can happen between the one person reading in the deck of punch cards and stacking the checks. But the part of the system that Jim was really excited about was the interaction between the veterans and the college guy, with respect to using the computer. Likely the veterans felt threatened by the college kid, just as the college kid felt judged by the veterans. But what happened was good for everybody. When the computer-check reader failed the veterans felt vindicated - "there is no substitute for experience" And the college kid's failure made him a member of the group - "you can only learn by making mistakes" This is what systems is all about - the interactions between complicated things!

Now what does this have to do with fluid dynamics? Well not much really. A few years ago I listened to the audio tape of John Glenn's book The most interesting thing to me were two or three stories he told about being a test pilot. In one of the stories they were working on a new jet, say the F-86, and there was a problem with the machine guns overheating. Someone designed a heat sink to put on the barrels and when Glenn tested the guns they no longer got hot, but the leading edge aerodynamics changed dramatically. A resonance was created when he fired the guns, causing a rippling wave to go through the wing. The wingtip was whipped about when the wave reached it, causing what is called PFOA - parts falling off the airplane. This was a life-threatening interaction between a well designed wing and some add-on hardware that was never tested in the wind-tunnel. The best solution turned out to be synchronizing the guns so they did not fire in order. No regular firing order, no wave, no PFOA.

Another Glenn example had to do with the air intakes for jet engines at high sub-sonic speeds. This is a tough math problem, and in the early 1950s there were no computers good enough to solve it. Some of these planes would fly along and as soon as you turned or dove, the airflow into the engine would choke off and the engine would stop. Another example is in the 1970's they added wing tip fuel tanks to the B-52. One day they found out that at a certain rate of descent the new fuel tanks caused flutter in the wings that could resonate and shake the wing apart. Until they developed a new spoiler that could dampen the flutter they had to live with the dangerous wing. Pilots could easily stop the flutter by pulling up out of the descent, but they only had about ten seconds to do this once the flutter started. Having a man-machine system that requires human input at least every ten seconds is a dangerous thing. People get fatigued, they get distracted, they get bored. This system required constant vigilance which is stressful for human beings.

So with John Glenn's machine gun problem the solution was to test fire the guns and make sure they didn't fire in order. With the engine intake problem the best solution was to redesign the intake. For the B-52 wing tanks they temporarily required the pilot to be careful during descent, and the permanent fix was a new spoiler that dampened the flutter.

All these things illustrate that fluid dynamics as applied to airplanes is not well understood. Things seem to fail when you least expect it, and over-confidence is not a good thing to have. One of the stories that came out of Edwards test pilots is that the pilots did not trust the new hardware to fly safely. If someone had made an ad-hoc fix to the plane using duct tape, then it was probably safe to fly. The way I heard the story is they always used green duct tape at Edwards, to tape down an access panel that was a little loose, or to close off a seam in some faulty slat design, or to hold a hydraulic line in place on the cockpit floor. If it didn't have any green tape on it, it wasn't ready to fly.

Anyway, the whole conversation with my uncle led me to conclude that the stuff I knew was not obvious to everyone, and most people took it for granted. For example most people would not realize that there were two systems involved in the check processing application - 1) the new programmer interacting with the computer, and 2) the group of computer programmers interacting with the new guy about passing along computer knowledge. Since its not obvious, I thought it might be a good idea to write some of these ideas about systems down, and that's why Im writing this blog. The fluid mechanics is a good place to start, since that has been on my mind for awhile, and is fresh in my mind.

And if you get a chance, stop by Wal-Mart. And for Edwards sake, get some green duct tape. It will make me feel a whole lot better if you do.