Sunday, September 09, 2007
Failure
When we are in the US most of our time is spent seeing doctors, or waiting to see doctors, or having operations in hospitals or clinics or nice stuff like that. To amuse ourselves without spending too much money we spend Saturday going to Yard Sales. We buy a lot of books and send them down to the Caribbean in barrels along with stuff that's hard to get down there, because a used paperback book costs about what a laborer gets for a day's work. If we don't want to keep them we donate them to the public library.
Once in a while we run across a real gem. Yesterday I found a copy of To Engineer Is Human by Henry Petroski. It is subtitled "the role of failure in successful design" and his view of the history of engineering considers not only that failure is inevitable but that the thoughtful analysis of failure is a stimulus to the evolution of engineering design.
I sat down and, pretty much, read it through. I have to say that I enjoyed it more than any book that I have read in recent years.
Part of that is nostalgia, of course, I not only had, and used at MIT, a K&E LogLog Decitrig slide rule like the one he uses as an example, but I had a book of log tables from the Bureau of Standards for pushing some calculations a couple of decimal places. When I worked at Yale I bought for our project an early Marchant electronic calculator that was as big as a Monroe mechanical calculator, cost $1500 and had the same power as the ones you can now get in a dollar store. Petrosky uses the sliderule as an example of the historical limitations of technology.
But new technology brings in new limitations. While in graduate school I did part time technical writing on the Wright YJ67 jet engine. Evidently their engineers wrote so badly that the company shipped the raw information to the small firm I worked for and we wrote their reports. In one case I ran across a diagram which showed the temperature distribution in the combustion chamber With the temperatures stated to 4 decimal places. It may be that their engineers believed it, but nobody who actually tried to measure temperature in a combustion chamber would take more than two decimal places seriously. It was probably the result of a computer simulation and computers don't get to take measurements.
I also did a failure analysis of a fuel control for that engine. We got the job because our overhead was lower than anybody else. I taught myself what it was about, hired a college student for the summer, and got a commendation from the Air Force. We got it to perform the mission by saying that the pilot had to push a test button every few minutes.
Nobody was happier than I was when the Air Force scrapped that engine.
I also taught a course on nuclear safety for prospective engineers for the nuclear freighter Savannah where one student was transferred from graduate school in maryland and a couple of the others had GED certificates and 25 years at sea. The chairman of the department asked me not to go too heavy on safety because one of my class was a shop steward and they didn't want him to know too much. I did go heavier than I might have if he hadn't said that, but I was happy when the NS Savannah was decommissioned.
I also learned something about brittle fracture. When I was cleaning up toxic waste sites I had a client who had leaded gas on the groundwater under his gas station. He hadn't had leaded gas in those tanks for twenty years but it was downgradient of a landfill where the National Guard had reportedly dumped drums of leaded gas.
The State forced him to exhume those gasoline tanks and the state inspector found a leak by pounding on a rust spot with a sledge on an extremely cold winter day. This was two weeks after the tank had passed a standard leak test. But the daily cover on the landfill was excavation material from the "Big Dig" and I was the only one who wanted to look into that carefully. After all, if they found the excavated material was contaminated it would have added to the "Big Dig" deficit.
The state later took away my license as a site cleanup engineer. The expert witnesses against me at the hearing couldn't cite anything I had done contrary to regulations. but, as the Board's lawyer said, I hadn't been charging my clients enough. Since the Board was mostly my business competitors that was a terrible crime.
As an example, there was an oil spill at a church I had belonged to. Another engineer had estimated $70,000 to explore for contamination. I provided my services pro bono and worked with volunteers from the congregation. We did the job for $600.
The State Board took away my license on the spot, so I retired. My lawyer charged me more than my income for that year and never really understood what was going on.
That sort of thing should explain why I enjoyed the book. I had been trained as a nuclear physicist but I had done a fair amount of engineering work, at least partly because my father was a tool & diemaker. When I worked for the engineering firm anything that didn't fit precisely within one or another specialty was given to me to try. Since I didn't have any preconceived ideas I could sometimes figure out a cheap way to accomplish the task.
Even my Ph.D. thesis was less interesting for the physics it contained than the design of the instrument I did it with. People all over the world were interested in that.
But I'm not in that business any more as you can see from http://index.karleklund.net
But if you run across one of Professor Petroski's books, grab it. Even a non-engineer will find it interesting.
Once in a while we run across a real gem. Yesterday I found a copy of To Engineer Is Human by Henry Petroski. It is subtitled "the role of failure in successful design" and his view of the history of engineering considers not only that failure is inevitable but that the thoughtful analysis of failure is a stimulus to the evolution of engineering design.
I sat down and, pretty much, read it through. I have to say that I enjoyed it more than any book that I have read in recent years.
Part of that is nostalgia, of course, I not only had, and used at MIT, a K&E LogLog Decitrig slide rule like the one he uses as an example, but I had a book of log tables from the Bureau of Standards for pushing some calculations a couple of decimal places. When I worked at Yale I bought for our project an early Marchant electronic calculator that was as big as a Monroe mechanical calculator, cost $1500 and had the same power as the ones you can now get in a dollar store. Petrosky uses the sliderule as an example of the historical limitations of technology.
But new technology brings in new limitations. While in graduate school I did part time technical writing on the Wright YJ67 jet engine. Evidently their engineers wrote so badly that the company shipped the raw information to the small firm I worked for and we wrote their reports. In one case I ran across a diagram which showed the temperature distribution in the combustion chamber With the temperatures stated to 4 decimal places. It may be that their engineers believed it, but nobody who actually tried to measure temperature in a combustion chamber would take more than two decimal places seriously. It was probably the result of a computer simulation and computers don't get to take measurements.
I also did a failure analysis of a fuel control for that engine. We got the job because our overhead was lower than anybody else. I taught myself what it was about, hired a college student for the summer, and got a commendation from the Air Force. We got it to perform the mission by saying that the pilot had to push a test button every few minutes.
Nobody was happier than I was when the Air Force scrapped that engine.
I also taught a course on nuclear safety for prospective engineers for the nuclear freighter Savannah where one student was transferred from graduate school in maryland and a couple of the others had GED certificates and 25 years at sea. The chairman of the department asked me not to go too heavy on safety because one of my class was a shop steward and they didn't want him to know too much. I did go heavier than I might have if he hadn't said that, but I was happy when the NS Savannah was decommissioned.
I also learned something about brittle fracture. When I was cleaning up toxic waste sites I had a client who had leaded gas on the groundwater under his gas station. He hadn't had leaded gas in those tanks for twenty years but it was downgradient of a landfill where the National Guard had reportedly dumped drums of leaded gas.
The State forced him to exhume those gasoline tanks and the state inspector found a leak by pounding on a rust spot with a sledge on an extremely cold winter day. This was two weeks after the tank had passed a standard leak test. But the daily cover on the landfill was excavation material from the "Big Dig" and I was the only one who wanted to look into that carefully. After all, if they found the excavated material was contaminated it would have added to the "Big Dig" deficit.
The state later took away my license as a site cleanup engineer. The expert witnesses against me at the hearing couldn't cite anything I had done contrary to regulations. but, as the Board's lawyer said, I hadn't been charging my clients enough. Since the Board was mostly my business competitors that was a terrible crime.
As an example, there was an oil spill at a church I had belonged to. Another engineer had estimated $70,000 to explore for contamination. I provided my services pro bono and worked with volunteers from the congregation. We did the job for $600.
The State Board took away my license on the spot, so I retired. My lawyer charged me more than my income for that year and never really understood what was going on.
That sort of thing should explain why I enjoyed the book. I had been trained as a nuclear physicist but I had done a fair amount of engineering work, at least partly because my father was a tool & diemaker. When I worked for the engineering firm anything that didn't fit precisely within one or another specialty was given to me to try. Since I didn't have any preconceived ideas I could sometimes figure out a cheap way to accomplish the task.
Even my Ph.D. thesis was less interesting for the physics it contained than the design of the instrument I did it with. People all over the world were interested in that.
But I'm not in that business any more as you can see from http://index.karleklund.net
But if you run across one of Professor Petroski's books, grab it. Even a non-engineer will find it interesting.
# posted by Karl Eklund @ 2:37 PM 
