Engineering Problems

There have been several news articles recently about the problems with the Boeing 737 MAX, including two planes that crashed – one in Indonesia, one in Ethiopia. It appears that the root cause of the problem is the fact that because the plane has been stretched (the reason for the MAX part of the plane’s name) the nose of the plane has a tendency to tip down and it requires sophisticated software to keep the plane’s attitude nose up. When one of the external sensors is not working properly, the software is not handling the situation properly and the plane keeps going nose down until it is unrecoverable and the plane crashes.

I have a background in engineering and many years of experience in writing computer software, but the intricacies of this are difficult for even me to totally understand. So, I thought I’d write about some less complicated transportation problems that might illustrate the kinds of issues that engineers deal with and how simple things can lead to complicated problems. These two examples are real ones that people I knew had to deal with about 40-50 years ago.

The Oil Filter

A car manufacturer (I seem to recall it was Ford, but that’s not important to the story) came out with a new model of a popular car that had a more powerful engine than the prior year. A friend of mine bought one of these new cars and loved it. After a month or so it was time for the first oil change. The dealer he bought it from included with the car a free first oil change (this is a good marketing strategy as if you can get in the habit of going to them for this kind of service you are more likely to keep coming back – and making money for the dealership). So, he scheduled his new car for an oil change. Figuring that it would not take too long, he stayed in the waiting room where he had a view of the service bays.

He could see his car on the lift and watched the mechanic unscrew the oil drain plug in the bottom of the engine and drain out the used oil. But then the mechanic seemed to be under the car for an extended period of time and he began calling others over until nearly everyone in the garage was gathered underneath his car. Wondering what was going on, my friend went to the service desk and inquired if there was a problem with his new car. The response that he got was that they were unable to remove the oil filter!

Because this was a new model with a newer engine, the engineers who design it have to take into account all the various parts of the engine and its attachments to ensure that everything will both fit properly in the engine compartment and also be able to be taken out. They had done so properly. But the space in the engine compartment is pretty tight (as anyone who has looked under the hood of most modern cars can see). Thus, the amount of space between some components is quite small. The oil filter is generally near the bottom of the engine and in this case was right up against the engine frame with a fraction of an inch clearance.

But the engineers had forgotten a small point – that when you need to change the oil filter it must be unscrewed and that means that there must be enough clearance between the filter and whatever it is next to (the frame in this case) to allow that extra fraction of an inch for an unscrewed filter. But they had not done so, so there was not enough room to remove the filter.

If the item that interfered had been some other component, they could have loosened the other component – that would have been bad enough. But in this case the other component was the car’s frame. Thus, the only way to change the oil filter was to loosen the motor mounts, use a jack to push up the entire engine that fraction of an inch, unscrew and change the filter, then lower the engine and retighten the motor mounts.

The consequence of that minor mis-calculation meant that most people would not be able to do their own oil change, that they could not take it to a Jiffylube or some other such establishment, and that the labor and time involved were not just a simple 15 minutes, but a couple of hours of work! Fortunately for my friend, the first oil change was free, but the long-term prospects were not great.

The dealer notified the manufacturer who very quickly got their engineers involved. The solution – make a new oil filter that was a fraction of an inch shorter so it could be unscrewed in that tight space. They also had to change the specifications in the owner’s manual, get a large quantity of the filters manufactured so they could be available on the assembly line for all cars not yet built, then contact all the car owners to schedule a recall – each of these cars needing to be put on a lift, the motor mounts loosened so the filter could be replaced, etc. It was a massive effort for the next several months – and all because the engineers forgot that unscrewing a $4 oil filter means you need a little bit of clearance!

The Computer Reset

I began teaching computer courses to adults in an evening program in 1980. My students were working adults who took classes three nights a week for ten weeks. In the first week of one of these classes a few of the students were talking together before class started and one of them shared that he had had a car problem on the way to class that night.

His route to school meant that he had to take one of the limited access roads in the Lehigh Valley (22, 309, or I-78). Some of the access ramps on the former two are quite short, in fact a few of them have stop signs at the end of the ramp so you have to stop, wait for a clear spot in the traffic, then accelerate rapidly to match the flow of traffic as there are no merge lanes. Also, these roads are all concrete and not always in the best shape (PA roads are notorious for potholes). It was at one of these short ramps that he had the problem.

As he was accelerating rapidly so that the traffic behind him did not run him over, his engine suddenly slowed to an idle speed and the RPMs dropped. He initially thought that the engine had died and began quickly looking for a place to pull over – but there are usually no spots for doing this either, so he was beginning to panic. Then, just as quickly as it began, the crisis ended and the engine once again began working properly and he was able to get back to the required speed. It was this moment of panic that he was relating to the others in the room.

Things went well for several classes, then one night it happened again – the engine RPMs quickly dropped and the car’s speed decreased accordingly, then after a few seconds everything began working again. Now he knew it was not a fluke and that there was a real problem. He took it to a car dealer but their tests showed nothing wrong. He took it to a local mechanic who could also find nothing wrong. The car was working properly all the rest of the time except under these times of acceleration on the short ramps – when the problem reoccurred a few more times over the coming weeks. What was going on?

Finally, he had another mechanic look at it – and this new mechanic found the problem. As on most cars, there are a number of wires in the engine compartment that are connected a number of “computers” that govern the operation of the engine. One of these sensor wires had come loose from the bundle that it was part of and was somewhat floppy. It would bounce around a little – especially on the bumpy sections of the PA roads when the car was at higher speeds and it had developed a small section where the insulation around the wire had worn through. Now when it made contact during one of those bounces it was shorting out. This was giving a phony signal to the computer to which it was connected.

At high RPMs, the phony signal generated from the bouncing on the roads gave conflicting information to the computer and the computer didn’t know how to handle the conflict. So, the solution, programmed by the engineers, was to “reset” the computer, then process the various sensor signals one at a time until it took the appropriate course of action. But this reset meant that the engine speed was also reset (to idle) until the computer could process the conflicting signals. And of course, by then, the loose wire was no longer in contact and shorting out, so the computer properly interpreted everything, including the fact that the driver was pushing on the accelerator pedal, and the car began working properly.

While the best solution would be for the engineers to do something other than a reset and drop the vehicle speed instead of just maintaining the current speed. But in this case, the mechanic just did the next best thing – wrapping some electrical tape around the bare spot in the wire and securing it so that it wouldn’t flop around anymore.