Development Nightmares

Have an interesting story about a development project?

Lee_Teschler — Wed, 25 Jul 2007 20:06:30 GMT

Development Nightmares chronicles humorous anecdotes from readers who have developed products and lived to tell about it. Submit a 200 word or less story to this forum and if we use it in one of our electronic newsletters, you will earn a quick $50 for your efforts.

Perplexing Paint Phailure

JimT_MD — Wed, 25 Jul 2007 20:06:33 GMT

We were fabricating contractors in Fort Worth, TX who had an order for some machinery parts to go to Australia. The order had very specific and rigid painting specifications, complete with the exact paint to use and how to apply it. We did everything exactly as specified with many quality checks along the way. We were happy with the good looking results using the Epoxy Zinc special paint. We wrapped and packed the parts very lovingly in custom made crates and sent them off to the land down under. Much to our horror we were advised that when the crates were opened the paint had literally fallen off the parts and was laying in the bottom of the crates in sheets. After weeks of investigation we found that the shot blasting operation left an invisible coating of powdered metal on the surface of the parts. We had actually painted the dust and did not get any primer on the parts at all. With no adherence to the metal the dust layer let the paint fall free from the surface of the parts.

How to make a sheet metal match

M_Husted — Wed, 25 Jul 2007 20:06:33 GMT

I have a problem that for the past two years can not come up with any answer. First let me explain the application: We make fireplaces, about 150,000 per year, each one uses four refractory panels and during the mixing of these panels we install a wire mesh as a binder for each panel. These wire mesh components cost about .70 each and the math comes to about $420,000.00 per year in wire mesh. I did some testing and found if I install a few hundred strips of metal about the size of a paper match into the mix for each panel, the binding is equal to the mesh. I have plenty of scrap metal from 22 to 30 gauge and my only problem is to find a paper shredder on steroids, since I need about 2 million metal matches per day. I have talked to many companies that build machines and nobody has the answer. If anyone could solve this problem there is a round of golf at pebble beach in your future.
M Husted
VP engineering

"Does it really make that much difference?"

Nermal — Wed, 25 Jul 2007 20:06:32 GMT

I received a very expensive assembly from engineering for analysis. The
small SMA connectors had rotated within the slotted panel when
"engineering" department had put everything back together. They were trying
to torque the nuts to 10 to 12 inch pounds when the nuts and connectors
rotated within the panel.

When I removed the lid the semi-rigid coax cables had rotated along with
the connector and fractured the solder connections on the PC card. The SMA
connectors were not damaged but the milled slots in the aluminum panel were
damaged.

I reviewed the history: "connectors turns when subjected to 8 or 9 inch
pounds". I looked at the manufacturers spec: "torque 10 to 12 inch ounces!"

I called project and confirmed the torque units that were used was in fact
in inch pounds. The person from project further indicated that they went
through considerable effort to find an appropriate torque gauge. The reason
the unit was re assembled in the engineering laboratory (by engineering
personnel) was to prevent further damage to the assembly. It had already
seen its 3rd rework cycle and would have to be scrapped (about $35k + my
department overhead + value added labor).

I explained to them that they had used the wrong torque units. Their
response: "is there really that much difference!" (between inch ounces vs.
inch pounds)

Recalibrate

D_Lee_MD — Wed, 25 Jul 2007 20:06:32 GMT

Some number of years ago I was called in to investigate a situation where a batch of springs (4.5-in diameter) had been rejected by our QA department because they had the wrong spring constant. The vendor of the springs disputed our findings and had their own test results to support their claim.

Normally, we have a pretty strict calibration program and our gage calibrations are traceable to NIST (National Bureau of Standards, back then). Indeed, I found that our spring gage had been sent out to have its calibration checked and certified. Our test results had come from a different organization that had used an Instron machine.

So, I talked with the test person who ran the Instron machine. He showed me how he ran the test and got the same results that he previously got.
"How do you calibrate the Instron?" I asked.
"I push this button," he said, pushing a button on the Instron machine.
"What about an external standard," I asked.
"You don't need one," he said, "it's internally calibrating."

Sure enough, an external standard showed that his Instron was out of calibration. I think Instron would probably have said that the test person needed to have his understanding recalibrated.

Drawing blanks

Lee_Teschler — Wed, 25 Jul 2007 20:06:32 GMT

An acquaintance of mine who ran a small job shop found he could fill some of the shop's slack time by stamping out washers thanks to a temporary industry shortage. With no international business experience, he advertised for more such work in a publication that reached all of North and South America. Soon he had landed an order stamping out simple spacers and welch plug blanks for an obscure engineering firm in South America.

He couldn't understand why anyone would want specialized blanks when some were available off-the-shelf that might work at much lower cost. But it was a sizeable order, so he didn't say anything. The customer gave him detailed requirements for several specific types of blanks in terms of materials, thickness, weight, edge trim and so on. When he asked about the end-use application, the customer was evasive, as if protecting a proprietary technology.

It didn't take too long for the real story to come to light. He eventually figured out that several Latin American and European countries had coins and subway tokens that were suspiciously similar to the spacers he was stamping out for a fraction of a cent each. Inquiries turned up no records of the "engineering" company he was dealing with. Though he wasn't sure he was doing anything wrong, he politely declined any further orders.

Thom M., Roseville, Mich.,

Half-Baked Idea

Lee_Teschler — Wed, 25 Jul 2007 20:06:32 GMT

Our controlled reaction rate mixing and curing vessel was an established product, which had gone through many revisions. To keep costs and weight down on the 3-foot-high unit it was made mostly of sturdy industrial plastic, but with some complex digital controls powered by 110 Vac. Our engineers knew certification tests well, so approval from the testing labs for the latest round of minor customer-requested improvements was assured. And for the first time we didn't have to do the paperwork, as it had now been handed over to the newly-established Global Quality department.

Testing schedules were tight, but ample as it always passed the first time. So you can imagine our astonishment when a noticeably provoked call came from the VP of Engineering telling us for the first time in 12 years it had failed the tests.

We rushed over to the shipping dock to inspect the returned unit and couldn't believe our eyes. There was nothing left but a big blob. It looked like a complete, Pizza-the-Hut style meltdown, little lights and switches poking out from the gray plastic mozzarella like pepperoni. Even the hi-temp ball valves were contorted and heat-welded into the plastic case. All we could think was some major transformer failure.

The explanation was actually simpler. A marketing whiz-kid had consorted with a new-hire liberal arts Global Quality wonk. They looked inside the unit and, seeing nothing flammable, had decided that 'fire-resistant' would be a dandy additional marketing feature. So they checked that box on the required tests. The agreeable, ever-cooperative UL Labs had cheerfully baked it at 1,200 degrees for 4 hours in their testing oven, after which it didn't work all that well.

Wilbur K., Tuscaloosa,

Regressive Testing

Lee_Teschler — Wed, 25 Jul 2007 20:06:32 GMT

A few years ago I joined the program maintenance team of an application software company that shall remain nameless. The first week I was reviewing trouble reports and noticed the software did not work reliably when used over a network with standard file-sharing. Customers were obviously frustrated and incensed, but the company explained that the product was out of major revision, and only guaranteed for stand-alone use.

It so happens I had worked on this problem at my previous job. I knew that a few pathological packages that didn't do disk access properly were prone to erratic network file-sharing failure. So I looked up that section of the code and fixed the problem in the next release.

Three weeks later I get a panicked and slightly irritated call from the Quality Control department. The new release was going through routine quality assurance 'regression' testing. I asked if there was somethine wrong wtih the changes I made. The reply was that the code worked fine, but that was the problem. The company had a client-server version of the product that was ten times the price, with a yearly renewable license instead of the one-time purchase price. If this older product started working reliably in network mode, nobody would buy the client-server version.

Management wanted to force people off the old product, while making it look like they were honoring the maintenance agreement.
QC had written a separate test to check the file-share network performance. The purpose of their so-called 'regression' testing was to make sure the software still didn't function correctly.

I restored the defective code and left the company shortly thereafter.

Howard T., Cincinnati

Dispensing with accuracy

Lee_Teschler — Wed, 25 Jul 2007 20:06:31 GMT

I once worked at an equipment company that made a dispenser which filled and discharged a chamber from the city water line for chemical mixing. It had to be very accurate. We designed a new and improved model that used optical water level sensors for 'full' and 'empty,' driven by an inexpensive chip that also operated the valves.

The unit worked perfectly at the factory, but soon after shipping we began to get angry complaints from customers that the unit was erratic. It would fail to drain, or hang up, or overfill. We dragged out the test stand and went over and over the thing, testing relentlessly. It was so simple we just couldn't see what could go wrong, and could not get anything to fail.

Finally we sent engineers into the field. After many false leads one of them noticed an extra unit in the water supply feed piping. When he asked about it, he was told it was an aspirator for premixing one chemical with the water, and that many of our customers used them. Of course, it was shut off during our tests. It didn't register at first, but after several more weeks of headscratching and devising some specific tests, the extra chemicals turned out to be the problem. The chemical shifted the index of refraction and surface tension of the water just enough that our optical sensors were getting fouled up at the mounting angles we used.

So after all that agony and gnashing of teeth, a simple change to a different type of level sensor cured the problem.

Dave B., Eastlake, Ohio,

TLAs

Gil_MD — Wed, 25 Jul 2007 20:06:31 GMT

We had our annual all-hands engineering meeting at a fancy winery in the hills of Saratoga, CA. One of the VPs proceeded with an introductory PowerPoint presentation of the days agenda. We were going to talk about current and future projects for the coming year. On one slide were several acronyms that I wasnt familiar with and I raised my hand to ask what they stood for. After explaining what the acronyms meant, he said there would be lots of TLAs and if anyone had any questions, just ask. I asked the guys to either side of me what a TLA was and they didnt know either. The next day, one of the other engineers came by my cubicle and said he found out what a TLA is. He said TLA stands for Three Letter Acronym. I almost fell out of my chair laughing. I couldnt believe there was an acronym for acronym. Only in corporate America!

The $5000 Hammer

rdf — Wed, 25 Jul 2007 20:06:31 GMT

Before the M1 Abrams main battle tank, there was the M60. It served with honor for well over 20 years before being replaced by the revolutionary turbine powered M1. The M60 was powered by a magnificent V-12 Diesel engine. As an employee of the engine manufacturer, I served as Liaison Engineer with the Army Tank Automotive Command. One of my responsibilities was the qualification under military specs of any new or changed components that went into the engine.

Now, much has been said about the proverbial $5000 military hammer that you and I can purchase from the helpful hardware man for $8. The jist of the talk usually suggests that suppliers are ripping off the military and therefore us tax payers, too. But there's much more to the story, as I found out during my first month on the job.

Late one Friday, I got a call from the manufacturer of the starter used on the M60 engine. As you can immagine, the starter was a large unit designed specifically to meet military standards and specifications. It had been in production ever since the first V12 was built, and was used no where else. So I almost panicked when the nice lady on the other end of the phone line announced that the starter we'd been using for more than 20 years would no longer be available in only two weeks time. "Don't worry", she said, "We'll send you one that's just as good". I tried to reason with her, pointing out that there was a rigorous qualification procedure that had to be completed before we could accept the new starter. She was unfased and responded that we could do what we liked, but the next shipment of starers in two weeks would be completely different than those we'd been using. She did, however, agree to overnight a couple prototypes for us to test.

As I dug into the qualification requirements, I found that the starter would have to be cycle tested under various stressful conditions including tests at -20 degrees F, underwater testing, dirt ingestion testing, high temperature testing, drop tests, repeat starting and several others that escape me now. It became painfully clear why the hammer cost $5000. It didn't have anything to do with the materials of construction. It had more to do with the cost of the airplane to carry out the drop test from 3000 feet, and the cold room required to test at -20 degrees! Thankfully, the new starter wouldn't require an airplane to complete its drop test. But the other tests looked like they'd take about 3 months to complete. A quick discussion with the VP of Engineering got me all the overtime and help I needed and I started around-the-clock testing. I slept in a dynamometer cell on two occasions, and didn't sleep at all many days. It took a painfull and expensive effort by 8 or 10 of us, but after two weeks, the new starter was qualified. The new starters arrived on time and proceded directly to the production line where they slid into the assembled engines like nothing at all had happened.

Learn to communicate

rdf — Wed, 25 Jul 2007 20:06:31 GMT

In college we were told over and over again that good communication skills were imperative to success as engineers. This was realy driven home when I started work for a major heavy equipment manufacturer. Unfortunately, there was an impending strike when I started in the dynamometer lab. The union members were all on edge and looking for ways to "get back" at managment, including those of us in the trenches with them. One particular engineer that I worked with was known for his inability to communicate well. After a long day of turbocharger sizing on his dyn engine, he left a note for the second shift technician. It read: "Remove the turbo from the engine and install the turbo on the bench". Given the sour attitude that prevailed amongst the techs at the time, you can imagine what the engineer found when he entered the dyn room next morning. Sure enough, the tech had removed the turbo from the engine and bolted it securly to the bench, right next to the brand new turbo to be tested that day. When he complained, the engineer was told he'd have to wait in line for service since the tech had done exactly what was written on the request.

Busted!

Lee_Teschler — Wed, 25 Jul 2007 20:06:31 GMT

About 15 years ago I worked for a company that made mail sorting equipment. While our equipment was reliable, some of the temps and contract employees who were involved had to travel to an out-of-state facility with the project engineers to test the machinery for post office officials.

This was during football season, so one of the traveling temps faxed the weekly "confidence points pick-em" for all the traveling employees back to the home office. Unfortunately, he did it using the postal fax machine. He was busted along with the other participants for transmitting gambling material across state lines.

Needless to say, when the company's general manager found out (also a participant), that was the end of the football pool for everyone.

Mike Olijnyk, Chicago, IL,

Invalid Tensile Test

Bob_S — Wed, 25 Jul 2007 20:06:31 GMT

About 40 years ago I inherited project engineering responsibility for a fiberglass jet engine seal housing and was immeadiately informed by the quality and materials liasons that the specified material had failed the required tensile test. Reading the report confirmed that the yield stress of all the "dogbone" samples was way below the minimum requirement. Upon opening the packet of tensile specimens, I observed that all of them broke off at one end, not in the rectangular reduced-area section. The lab technician had dutifully measured the width and thickness of this section, calculated the area, and divided this value into the force value the tensile-test machine provided, giving a totally meaningless PSI value. The cause of the "yielding" of the specimens was obvious as the holders were two vice-like jaws with sharp teeth that litterally crushed and fractured the large ends of the "dogbones." No wonder they broke there! The whole test had to be repeated by another lab that had the proper holding fixtures, and the material easily passed.

Knock, knock

rdf — Wed, 25 Jul 2007 20:06:31 GMT

I worked for many years developing engine components. During that time I met and worked with a number of very talented people. One was a dyno tech named Pete. Pete was very good at his job, but he was a practical joker. Because of his talent, Pete often ended up showing the ropes to new techs. On one occasion, Pete was instructing a very nervous new employee about doing dynamometer spark maps. The new guy was shaking in his boots because the engine they were mapping was a one-off prototype that cost close to a quarter million dollars. Pete explained that in order to hear the first signs of knock, it helped to wear head phones. They blocked most of the background noise but still let the high frequency knock through. As the new tech sat at the console with his head phones on, Pete stood beside and just behind him so he could reach the dyn controls if necessary. The test began. As the tech dialed in more timing, Pete produced a box of Tic-Tacs and shook them behind the new-hire's head. He immediately dialed timing out and recorded the knock point. They proceded to the next point and continued the same procedure. Again, Pete shook the Tic-Tacs well before the engine realy knocked. This continued until a complete, but wholy unreal spark map was generated. Then, Pete challenged the new tech to recheck all the points by himself and left the room. Proud of his success and surprised at the ease with which the map was generated, he gleefully began repeating the test. Poor guy tried point after point with no repeat data. Pete (and others) stood outside the dyn room and laughed hartily. When the new tech finally gave up, he emerged from the dyn room with a puzzled look on his face. He was a little ticked off when he found us all laughing at his dilema. But Pete saved the day when he explained that the "new" spark map was indeed the correct one.

Saturate Magnetics

Kscheste — Wed, 25 Jul 2007 20:06:31 GMT

In the late 70's we were developing a Variable Reluctance Indicator for aircraft. I was a julior engineer and the designer of the instrument. My bosses who were electrical engineers inststed that we build the laminations used in the indicator out of a super magnetic material. I kept asking why our comeptitor didn't use the super magnetic material. I was told I didn't know what I was takling about. As it worked out when I was talking to the lamination manufacturer, I asked if laminations could be made out of common magnetic material. He said he would use that material for set-up and send me the pieces. As the deadline for the project neared, the prototype was built using the super magnetic material with the EE's touting how superior the instrument would be. The Thursday before the prototype was due at the customers the prototype was tested. It was in ful magnetic saturation so that no matter what was done, it would not work. The next day the company CEO had us in a meeting to determine what was needed for recovery. I pulled my shoe box of laminations out and suggested we try them. The CEO said we had until 7:00 PM Sunday to produce three instruments. A technician and I started working immediatel after the meeting and at 5:00 AM on Sunday succeeded in getting the right output from the instrument. The conclusion was that we got the contract and recieved a patent for our efforts. The simplest approach is sometimes the right answer.

Keith C., Chicago.

Changing Times

Lee_Teschler — Wed, 25 Jul 2007 20:06:30 GMT

Back in the late 70's our latest model of nuclear safety monitoring system used multiple minicomputers for redundancy. For coordination, they used a type of memory that could be shared among several CPUs. We wrote all the code according to the manuals, but when we tested it, it didn't work right. The systems guys went over it and over it; we spent thousands of man-hours chasing the problem.

We swapped in several different memory units to make sure it wasn't a hardware problem.

Turns out, the hardware was misdesigned, and just plain did not work. The vendor failed to inform us that this hardware had never been sold to anyone, and never tested.

We were dumbfounded. In that era, it was unheard of for an industrial computer vendor to deliver hardware that had not been thoroughly tested. It never crossed our mind that the hardware had never before been used, and the software manuals were basically a tech writer's conjecture.

That's a mistake no engineer would make today. Times certainly have changed. Industrial computer hardware then was above suspicion; now we assume hardware and software have something wrong somewhere. The snafu was a harbinger of mounting pressure on the minicomputer industry; within a decade most of these highly touted companies were gone.

John P., Dexter, Mich.

Random Error

Lee_Teschler — Wed, 25 Jul 2007 20:06:30 GMT

Some years ago we made monitoring systems and were asked to do one for a large boiler. The plant engineers emphasized that 5-digit accuracy was crucial, as even the smallest error could mean millions in operating costs.

George and Charlie were in charge of data collection and A/D conversion. Both guys were very reliable with many years of experience, so we had no worries in that regard.

The customer insisted on an unusual full pre-installation demo at our factory, squeezing our normal schedule. Two weeks before the demo Charlie took a new job in another country, leaving George to finish up. But the week before the test George died in a boating accident. We were sure he was finished with the system, and the numbers looked right, so we went ahead with the demo. The system was then installed and went through certification, which the plant engineers signed off on.

Years later I bumped into Charlie, and we got to talking. It turns out the system was not finished after all. The least-significant BCD digit from A/D conversion wasn't working right, so to meet the deadline Charlie dummied it in by software with cleverly manipulated random numbers for a realistic demo, until George could fix the hardware problem on-site before final testing. Obviously, George never made it to the site.

As far as anyone knows, that's how the system ran until it was decommissioned. The customer never complained, so apparently accuracy wasn't all that important after all.

Simon S., Melrose, Mass

Early Out Sourcing 1977

Carl_MD — Wed, 25 Jul 2007 20:06:30 GMT

A small company in Kansas was one of the first to build digital watches. They wished to be fully independent of out side suppliers and as they had experience building quartz crystals for the military they started a production line for quartz watch crystals. They also had a policy of putting managers with out previous experience in charge so they would not be limited by preconceived notions of what doesnt work. After a year of not quite making the process work they contacted a consultant from Texas. He looked at our process and said the best we could do was $3 each and we should throw everything away and buy the crystals from the Japanese for a $1 each. We said we knew if he came and worked with us we could match the Japanese. For a high price the consultant agreed to come to our plant and rebuild our production lines and promised to be making quality crystals in six months, at $3 each. Sure to his promise we were making the crystals for $3 at the end of six months. The consultant took his money and went back to Texas. We threw all the equipment in the dumpster and started buying from the Japanese, for $1 each.
Carl.

Tickets to Ride

Lee_Teschler — Wed, 25 Jul 2007 20:06:30 GMT

When times got tough in the automated inventory control business (pickers, stackers, label readers, etc.) the company at which I once worked decided to try its luck at revenue-collection systems for turnpikes and bridges.
We were told to get started on a turnpike toll collection system without a contract, as the fix was in; the new VP of sales's brother-in-law headed the turnpike commission.

So we built a prototype system complete with a ticket spitter that printed the Turnpike Authority name and logo to prevent fraud.

But the opposing party got wind of the graft and organized a surprise visit to inspect our operation. They could find nothing on the books to connect our prototype equipment to turnpike contracts. But near the end, one of them reached over and pushed the 'test' button, which caused the spitter to start spewing out tickets with incriminating imprints.

The marketers and execs ran around wildly, stomping on the tickets and collecting them, and substituting preprinted anonymous tickets, while the engineers snickered at the sight from outside the soundproof room.
Fortunately, no incriminating tickets got loose, and we got the contract and installed the equipment at ten times the usual margin. The VP then quit and moved to another company. Two years later our 'inadequate' system was again replaced at nearly four times the price by his new company. In fact, it was replaced every few years until the brother-in-law retired.

Name withheld to protect the guilty.

High-speed goop generator

Lee_Teschler — Wed, 25 Jul 2007 20:06:30 GMT

I once worked for a manufacturer of high-speed packaging machinery. We installed a cup-filling machine in the middle of a production line populated with equipment from several other manufacturers, and we were all busily debugging our respective systems.
The customer's plant manager was the sort who manages by intimidation, and part of the routine was for him to call supplier personnel into his conference room a couple of times a week for a ream job. There he'd complain about the rate of progress, call us all idiots, and generally engage in other unproductive and morale-damaging activity. After a few weeks this was wearing pretty thin.
Toward the end of the job, our equipment was pretty well dialed in, but some of the downstream equipment was not ready for full-speed operation. This was particularly true for the case packer which consolidated all of these cups and put them into cartons . So on this particular day, Mr. Manager, having heard all of our explanations, concluded his rant with, "I want your machine running at full speed starting right now and I don't want it stopped until I say it is ok."
Fine by me. I warned the guys ahead of us to have plenty of product ready. I let everyone know we were instructed in no uncertain terms to do a full-speed test run of undefined duration. We then set the machine running at exactly the rated speed.
Cups went into the case packer where cartons of this stuff were promptly smashed at the rate of four a minute, sending geysers of icky, slimy dairy product to the ceiling. People were running around frantically begging us to stop. "I'm sorry, but I have specific instructions from Mr. Manager that this machine is not to be stopped or slowed down until he gives the ok." By the time he showed up, purple as a beet, we were awash to our ankles in goop, and most of the other workers had abandoned the area.
I, at least, was excused from future ream sessions.

anonymous, to protect the guilty

Disposable Keyboards

Machinery_Guy — Wed, 25 Jul 2007 20:06:30 GMT

I worked for a custom machine builder in the late 1980's when the market for personal computers seemed like it would never end and it seemed that all the big multinationals had a division targeted to making home PC's.

We had obtained a multi-million dollar contract to fully automate the assembly of the keyboards. It was a huge multi-station palletized line with robots, bowlfeeders, dye sublimation machines, testers, and a host of other third party systems.

The customer came to visit us several times as the system evolved. It seemed that each time they showed up at our factory, the size of the group got larger and larger as more people within the customer organization jumped onto the bandwagon. At the preliminary run-off the group had become so large that they had charted a Grayhound bus to bring the crowds of operators, technicians, engineers, and QC personnel that were to run the line. The machine was a wonder to watch. There were high fives all around and we set up a date for final acceptance run-off.

Then abruptly everything became quiet, calls were not returned. The run-off date passed and no one showed up.

Turns out that there had been a massive wave of layoffs that had decimated the customers rank and file employees.

The reason for the layoffs?
Keyboards had never been life tested.
The keyboard worked wonderfully.
It did everthing it was expected to do.
Except for the fact that the keyboard only lasted 2 months before it became totally unusable.
Fortunately the customer after months of legal wrangling paid us for the work done. We had to scrap the entire machine.
Millions of dollars wasted and dozens of jobs lost, because of one simple mistake.

Heart Monitor Attack

Lee_Teschler — Wed, 25 Jul 2007 20:06:30 GMT

As head of the medical equipment department for a small rural hospital, I often had to repair equipment myself to keep costs down, and uptime up. When a cardiac monitor broke, I quickly traced the problem and ordered some replacement main signal amp boards. When they arrived the technician swapped one in, but it didn't work properly. We switched it with a board from a known-good unit, but that board didn't work. Worse yet, the known-good unit suddenly didn't work either, even when its original board was put back. We put another brand-new spare in it, but that didn't help. We got another good unit, but it went bad too. The more boards we changed, the worse the situation got.
Turns out these boards were so sensitive that fingerprint oils across two traces would throw the amplifier off. The overweight technician was continually munching on potato chips while working, and always gripped the board the same way to install it. After consulting with the manufacturer, we found the boards had to be washed with a particular solvent, and installed while wearing special gloves.

DTB, Dayton

It doesn't take a degree

rdf — Wed, 25 Jul 2007 20:06:30 GMT

Several years ago, I headed the development of a system utilizing a small air motor. The air motor was connected to its internal gearbox using a thin, large diameter steel nut, secured to one half with a small wire locking ring and threading onto the other half. A single key engaged locating grooves in both of the aluminum halves to prevent rotation. The nut was torqued to 45 lb-ft at the factory. Testing went without a hitch and the product was released to production. Shortly thereafter, we began to get air motors back with the thin nut loose on its threads and the locating grooves in the two halves badly pounded out by the key. We immediately notified the manufacturer there might be a problem during their torquing of the nut. Quite arrogantly, they responded with "proof" in the form of torque records that they had never shipped an air motor with a loose nut. After many contentious phone calls and emails, they finally agreed to send engineers to meet with us. The meeting was attended by one of our technicians, not a degreed engineer, but one of those hands-on type guys that just seem to have a knack. The engineers quarreled for some time before the tech. finally interupted. In his hand, he held two of the wire rings that secured the nut on the motor half. "Look closely at these. One is from a motor that didn't come apart and one is from a motor I just junked and replaced because of the damage done when the two halve worked loose." Upon close inspection, a small burr was noted on both ends of one of the rings, but nowhere on the other one. The room went silent. I think every engineer there realized what had been going on. Later tests confirmed our suspicions. The burr on the ring dug into the aluminum body of the air motor when the nut was torqued. The opposite burr dug into the nut. The spring steel ring deformed under torque because it couldn't rotate with the nut. The resultant clamping load was far less than it should have been after 45 lb-ft of torque, even though the torque instruments indicated correctly. A cost reduction change to the forming of the wire ring had produced burrs that weren't present on earlier rings. The manufacturer reluctantly made changes that quickly solved the problem.

I hate when this happens

Lee_Teschler — Wed, 25 Jul 2007 20:06:30 GMT

One day the main rotor blades of a military helicopter we made were partially unpinned and folded back along the fuselage after a mission off an aircraft carrier, as was the routine. In the process, one of the blades broke neatly at the cuff attachment point and fell to the carrier deck.
Everyone was dumbfounded. Naturally, a thorough investigation ensued. During hover at the approach to the carrier deck, there is substantial compressive stress on the top surface of the blade. A fatigue crack had developed at the top root, but it was kept closed by the compressive stress. When the craft was parked, the blade was no longer stressed in a way to keep the crack closed. So upon blade folding, enough tension developed at the crack to make the blade snap.
The fatigue crack was a nice clean brake right along the attachment cuff, to which the blade was adhesive bonded. Further investigation showed that a scribe mark had been placed along the edge of the cuff. This mark eventually grew into a fatigue crack.
Now the question was: Why was there a scribe mark on the blade? It turned out that an enterprising worker in the blade shop didn't like the way the adhesive oozed slightly beyond the cuff and out onto the blade surface. The excess adhesive looked a bit messy and created the impression that the utmost workmanship had not been used building the blades. So he took it upon himself to use an Exacto knife to trim the excess adhesive away. In so doing, he was unaware that he was scribing an infinitesimal scratch in the blade. The scratch is what grew into a fatigue failure.
Well, talk about a product recall. All blades built around that time had to be inspected, and the company put out an engineering bulletin for blending out any scratch found. And as a further preventative measure, a rule was established prohibiting any unauthorized sharp objects in the blade shop.
Tanner G., Lansing, Mich.