The Idea Factory

by Pepper White

  "What's that?" I asked.

  "It's the lab data acquisition system. The sensors mounted on the engine send their data to the little computer here. Then it downloads the data files to the mainframe computer upstairs for analysis."

  Simple enough. "So how's it going so far?" I asked.

  "Pretty good," she said quietly. "The only problem is that my partner on the project is a real dork from Georgia Tech. I don't know how he got accepted."

  Maybe he talked bicycles with West.

  "Well, I hope everything works out all right," I said. "I've got to go back to work. Every minute of progress now means I'll finish that much sooner and be able to get on with my life."

  I looked at the chrome-plated shaft on the front of the machine, the shaft Nick and I'd unveiled when we removed the cylinder the day before. It had a crack about one-third of the way around it, right near the shoulder where the nonreturn pawls caught it. The force of the compressions had taken their toll. I unscrewed the chrome section from the rest of the shaft and carried the baseball-bat-size shaft to where Nick was drilling a hole in a piece of metal.

  "Problem, Nick. The shaft's got a crack in it. Is there any way we can fix it? I don't want the thing to break off a year from now when I'm almost ready to finish my experiments," I said.

  "We could leave it alone-it'll probably stay straight until it breaks-or I can weld a bead around the whole shaft. That'll make it stronger, but it'll probably bend the shaft a little. It's up to you."

  "If the shaft bends, can we bend it back?"

  "We can try," he said.

  "OK. Let's go for the weld."

  We went to the welding area and Nick put on his little welding cap and the leathers that made him look like a torturer in the Spanish Inquisition. He handed me a set of leather gloves, a cap, and a fiberglass welding helmet with a treated glass slit through which to look at the weld. The glass treatment allows you to see the weld but not much else.

  "OK, Cap'n. You hold it and turn it slowly, and I'll weld it. If you hold the end you'll be fah enough away an none of that metal will hit, so you won't need leathers."

  I put on my helmet, as did Nick. The bright flash of the weld contacted the metal at the shoulder and I slowly turned the shaft. It took about two minutes for Nick to weld all the way around it, layering the temporarily molten metal like cake frosting.

  "There. That should do it," he said. "Now before you do anything else with it, put it in that bucket of water over there."

  I put the welded end in the bucket, and it flashed the water like a hot frying pan in a sink.

  "Keep it in there for a while. That guy's hot all right," Nick said.

  After the steaming died down we put the shaft in the lathe. Nick set up a dial indicator, a little round gauge with a pin coming out of one side and a needle on the gauge. If you push the pin in, the needle moves around and tells you how far the pin is pushed in. Nick put the pin against the welded end of the shaft just as he'd put the cutting tool against the piece of steel.

  "Time to assess the damage, Cap'n," he said, as he turned the shaft around. The needle went way up, to fifteen thousandths on the dial, then back to zero, and then up to fifteen thousandths and back to zero as he turned the shaft around. "Not too bad," he said. "Now for the hammah."

  "What are you going to do?" I asked.

  "Give it a tap to make it straight again."

  He turned the shaft so the high spot was at the dial indicator pin, moved the dial indicator away, and gave the end of the shaft a quick, sharp tap with the plastic-headed mallet. He set up the indicator again and the high spot was down to four and a half thousandths.

  "I think we bettah quit while we're ahead, Cap'n," he said. "This guy's straight enough and he'll last another twenty years."

  "OK, Nick. Thanks."

  That night the temperature outside was below zero and the key wouldn't go into my Kryptonite bicycle lock. The metal in the lock had shrunk-not much, maybe a thousandth of an inch or two, but enough to prevent the big, warm key from fitting.

  At the coffee shop I asked for a cup of hot water, a plastic bag, and a rubber band.

  I wrapped the plastic bag around the lock, secured it with the rubber band, and slowly poured the hot water over the plastic. Then I quickly removed the plastic, put the key into the lock, unlocked the bike, and rode home to bed.

  The MIT motto is "Mens et Manus"-"Mind and Hand."

  C H A P T E R

  8

  The Taskmasters

  And Pharoah commanded the taskmasters, saying ye shall no more give the people straw to make brick. ... Let there more work be laid upon the men, that they may labor therein; and let them not regard vain words....

  EXODUS 5:6-9

  Schedule:

  Spring '82: 2.651 The Internal Combustion Engine (Heywood)

  2.999 Independent Study (Greene)

  2.996 Thesis

  February 15, 1982

  The Energy Laboratory conference room is inside the fourth floor on Amherst Street across from the Sloan School. The rheostatcontrolled lights were at full brightness while the Sponsors of the Research ate their choice of the best pastries MIT's catering service could muster. Even the coffee was good.

  It was sponsor-stroking time. Philip Hughes, an Australian, was director of engine research for Caterpillar. Sharma, originally from India, had published a path-finding paper on the theory of diesel combustion in the early 1960s and headed up engine research at John Deere. Albert Lee was working his way up the ladder at Cummins Engine Company-he was the last person to work on the RCM before me. And one of my office mates for the year was Jean Questois, a rising star from the Paris office of Renault. They were engineers in the most literal sense of the word.

  No pinstripes here; brown and light blue were the clothing colors of choice.

  Team Sloan had more members. The captain was Professor John B. Heywood, director of the Sloan Lab, graduate of University of Cambridge, recipient of honorary degrees. West was one lieutenant; the other was Chet Yeung. Chet was a new assistant professor, whose Ph.D. was from the Aeronautical Engineering department at MIT and whose bachelor's degree was from the one institute in the world that is better than MIT-Cal Tech.

  I was a blighter, as Snoopy called the World War I trench footsoldiers as he flew over them in his Sopwith Camel. So was Scott Rogers, my new lab partner. Scott would do the bulk of the computer work, I the bulk of the experimental work. The third blighter was Ben Radovsky, who worked on the square piston engine in the cell next door. Our mission-and we had no choice but to accept it-was to convince the men in light blue and brown that their money was well spent.

  Professor Heywood brought the meeting to order promptly at 9:00. "I'd like to thank you all for coming here today. Our students have been working very hard during the past month and a half, and I think you'll be pleased with the progress that they've made to date." His British accent made him sound even smarter than he was. I wanted to be like him, to command the situation.

  "We'll start the meeting with a presentation by Pepper White, on the progress on the mechanical aspects of the rapid compression machine experimental programme," he said.

  My knees were shaking. After I've been through a few meetings like this, I thought, any job interview should be a piece of cake.

  Professor Heywood continued smoothly, "Then Scott Rogers and Chet Yeung will discuss some thoughts on the computer model of diesel combustion we'll be developing. After lunch, we'll hear from Ben Radovsky and then take a tour of the lab. Before Pepper starts, though, I'd like to remind you all that we hope to expand the consortium programme to include more sponsors, and if you know of colleagues at other engine companies or oil companies who might be interested in participating in the programme, we'd be most appreciative of your putting us into contact with them."

  I put the stack of transparencies with the cardboard frames like the ones Gyftopoulos used next to the overhead projector and turned the projector on while he finished.
I stood next to the screen until my cue came.

  Ben dimmed the lights. All the faces around the conference table were lit up by the light of the screen; only the faces showed against the darkened background of the room, as in a theater. The eyes were focused, intelligent, good, and decent, since there is no room for lies in engineering. In science you can lie and fudge the data because you don't have to make anything work. In engineering the product is the proof of your honesty.

  First slide. "I'd like to talk about our general objectives, our specific short-term objectives, some of the things we've accomplished in the past several weeks, and the schedule for the coming months."

  My knees started to shake less as the audience smiled and listened attentively. I was using the right lingo.

  "First the general objective: We'd like to get the machine up and running and determine the range over which we can conduct our experiments. That means we need to determine the maximum pressure and temperature we can obtain from the machine, and also how high a swirl rate we can obtain, with and without combustion."

  Next slide-diagram of machine. As you can see, the machine has three chambers. The front is, of course, the combustion chamber; behind it is the driving air chamber, connected to a compressed air storage tank; farther back on the shaft is the snubbing chamber, filled with transmission fluid and rings of varying inner diameters. The piston in the snubbing chamber slows the shaft motion down during the compression stroke. By varying the order of the rings, we can make the motion of the shaft and piston simulate that of an actual diesel. The nonreturn pawls will lock the shaft in place when it has moved fully forward. We thus have the ability to look at combustion occurring at constant volume." The bright faces were still smiling. I'd learned a lot of words in the past six weeks.

  I went on to tell them about taking the machine apart and putting it back together, about fixing the shaft with Nick, about measuring and drawing a new piston with new bronze-Teflon onepiece piston rings, about having it machined and feeling it fit into the shaft. About the first unsuccessful test of the whole machine.

  At the end there was a warm round of applause, and on the way to the coffee and Danish table, West said, "First class." Maybe I'm not such a dummy after all, I thought. Maybe I just didn't know the system during the first term, the system that encourages dropping any course in which you're below average.

  Before the lunch break, Sharma asked for monthly progress reports from us. Oh no, more work, I thought. Evidently so did Professor Heywood.

  "They're bright young people, but there are many demands on their time and it might impede their progress on the research," he said.

  "Do you mean to tell us that the people you're producing for us won't be able to write a brief paragraph once a month saying what they've accomplished and where they're headed?" Sharma retorted.

  Professor Heywood hedged without flinching. He was still in control, the quintessential grantsman. "Perhaps the reports could be prepared every two months," he said. He knew that he would have to review them, that his secretary would have to type them, that he might have to meet with the whole group just to discuss that one subject. It could easily kill a morning that would be better spent writing his book.

  "Every two months would be fine with me," Sharma said, and the others agreed.

  February 17

  Professor Allen Greene's office was in Building 7, on the second floor overlooking Mass. Ave. and the student center across the street. Following up on our conversation after his Energy Engineering lecture in early October, he'd agreed to take me on as an independent study student in January. I remembered the incident in The Paper Chase when Hart was honored to be asked by Kingfield to help him prepare a scholarly article; I thought this might be close to equivalent.

  Greene ran the Institute for Applied Systems Engineering (IASE), a midcareer retraining ground for the engineering profes sion founded to do what Harvard Business School and the Kennedy School of Government do for their respective trades.

  IASE is one of the mechanisms by which MIT maintains its close links to industry. Mid- to upper-level engineers or research and development managers come to MIT, take a few classes, maybe do a small project, meet a lot of professors, and later hire the professors they meet as consultants to figure out how to solve their problems.

  Like me, Greene was an outsider. The IASE was his fiefdom, apart from the mainstream departments, apart from the traditional tenure track. His Ph.D. was from the University of Illinois, and his positions at Union Carbide, the World Bank (director of technology development), and Ohio State (dean of engineering) did not quite measure up to the MIT professor goal-getting tenure and starting a multimillion-dollar company. He was good enough for me, though.

  When I entered the antechamber of his office, his secretary was typing little blue index cards with his appointments and things-to-do on it. To hyperachieve you have to be hyperorganized.

  His phone was busy when she first tried to buzz him, but after a few minutes of her telling me how hypnosis was helping her quit smoking, he buzzed her to send me in.

  At four in the afternoon his office was bright with the western sun, bright enough that he'd turned off the fluorescents above his desk. His office was corporate-looking, like Gyftopoulos's, only bigger, with two couches in an L around a coffee table in addition to the desk, chairs, bookcases, and blackboard.

  He shook my hand and said, "Sorry to keep you waiting; that was the president of Ford on the phone and I don't like to cut him off. Here, have a seat." He sat on one couch, I on the other. "How are things going for you here?"

  I didn't tell him about being on probation. "Oh, pretty good, I guess. I'm doing experimental work in diesel combustion at the Sloan auto lab. Professor Heywood heads that up, and I'm also taking his engine class this spring, plus the independent study with you."

  "John Heywood is very good," he said. "And your experimental work will help you understand what's possible and realistic, even if it's a little on the researchy side. We can try to balance that with some real-world design problems. I'd like you to work up a computer example for my book as your major project, but first we'll take a few weeks to go over some problem-solving techniques I've developed over the years."

  He talked to me with respect, as a near-peer or at least a future peer. It was refreshing. I wondered whether he was a nice guy or if I was improving. He opened up the manila folder on the coffee table and said, "Let's go up to the blackboard and ..."

  Oh, no. Not this again.

  .. I'll work an example to show you how to solve steam cycle problems." He drew a grid on the blackboard and a diagram of a steam power plant-a boiler where the water is heated by the flame or by the nuclear reactor; a turbine that is forced to spin by the high-pressure steam generated by the boiler; a condenser to turn the low-pressure steam back to water; and a pump to push the water into the boiler. This is how the world economy is powered-anything that uses electricity is effectively a little fire. He put little hash marks at various points in the diagram and numbered them.

  "What you want to do is start at one point in the cycle and work your way around. For each point in the cycle, you put a line in the table. Each column in the table is a variable, like pressure, temperature, energy, entropy. In any design problem you'll be given certain constraints and your job will be to devise a workable solution that satisfies the constraints," he said.

  The two of us filled in the table of known quantities for the cycle, leaving blanks for the unknowns. The approach was orderly, methodical, simple.

  The session was a tutorial in the Oxbridge sense: at Oxford and Cambridge there are no classes; instead students meet regularly with tutors. I wished MIT encouraged more of this, but it's very time-consuming for the professor, and if he spends his time teaching and explaining things, there won't be time for the real work of bringing in research funds, consulting, publishing and delivering papers, making a name for oneself. Besides, if you can't figure out everything by yourself, you probab
ly don't belong at MIT. And they wonder why students kill themselves.

  Greene said, "It'd be a good exercise for you to finish this between now and our next session. First do it by hand and then we'll see about putting it on the computer. I've also got another problem for you, the balloon-filling problem. I want you to figure out how the volume for a balloon will vary as a function of time if you fill it with a large constant-pressure tank of air. And as a third item, try to come up with some examples or analogies for energy that isn't useful because of high entropy. You can make an appointment with my secretary for ten days to two weeks from now for our next session."

  I picked two weeks. I might need every day.

  February 19

  The steam cycle was easy. The balloon was a different story. It was a combination of thermodynamics and fluid mechanics, with that most dreaded of all things, a "deformable control volume." Usually volumes are fixed, and at steady state-i.e., things don't vary with time. A 3-inch-diameter pipe will typically stay 3 inches in diameter, and you can draw an imaginary box around it and account for mass, energy (whatever that is), and entropy (really whatever that is) by invoking the appropriate conservation equations. Or as they say at Harvard, all the gazintas gots to gazatta. Everything that goes in has to come out.

  Not so in the case of a balloon filling with air from a large tank at constant pressure. There the volume is always changing, until the pressure in the balloon is equal to the pressure in the tank, or until the balloon pops. This is the advanced course. I began my attempt at a solution.

  Step 1. Define problem: Draw picture. Think about what the problem involves physically. He just stated it verbally but I need to say it with symbols.

  Assume a pressure, say 3 pounds per square inch above atmospheric pressure. Look for key words in his problem statement. Large referring to the tank means that it's not affected by what happens to the balloon-i.e., the imaginary tank keeps pumping air out at a constant rate and never empties. Now think about what will happen. The balloon will fill up until the pressure inside it equals the pressure in the large tank. Those are initial and final states. The balloon is going to stretch, and energy will be stored in the balloon both in the air pressure and in the tension of the rubber.