Murphy’s Law

This past weekend I was organizing some older pictures on my computer and I came across a project that I did a few years back. This project was making a replacement flywheel for a model steam engine. The steam engine is completely wrecked, but making a replacement flywheel posed a challenge that involved the use of the lathe and milling machine. Not only that, but it required using a turret index chuck on the milling machine.

I started out with the broken flywheel and a piece of aluminum cylinder stock. I ground a HSS tool blank into a nice radius and put the aluminum blank into the lathe.

The project would be done in 3 operations:

1. In the lathe, cut one side of the flywheel, and hub. The hub includes a section protruding far enough out from the rim to cut a pulley into it.
2. Move the piece to the index turret on the lathe, Use the index turret to rotate the piece and drill 6 radial holes from the rim to the hub.
3. Insert stainless steel spokes in each of the 6 holes, using super glue to hold them.
4. Move the piece back to the lathe, and part the rim and hub (now secured to each other by the spokes) from the remainder of the aluminum stock cylinder.
5. Using a 4 jaw chuck and index micrometer, center the flywheel in the chuck with the unfinished side out. Finish this side of the flywheel.

Although it is only 5 steps, coming up with this process took quite a bit of thinking and consideration. Each step must take all subsequent steps into account, there were several problems encountered along the way which had to be resolved, such as how to secure the spokes into the rim and hub. We had several options including heat/cold shrinking a steel band on the outside of the aluminum rim. In the end we chose to use super-glue. The “perfect” conditions for using super glue are rare, but this was one of them: the surfaces to be bonded were smooth, clean, and in close proximity. Most people don’t realize that super-glue requires closely fitted surfaces, and the less glue the better the bond.

Since the holes for the spokes were drilled with a drill the same diameter as the spokes themselves, the error in the bit leaves a space on the order of .0001″-.0005″ between the outside of the spoke and the inside of the hole. This space was perfect for super-glue. Each spoke was placed in the holes and super-glue was dripped on to the edge, immediately pulled into the interface by capillary action! I was at first skeptical about the super-glue since the final machining operation required holding the rim of the wheel while cutting the hub with the machine tool. During this operation, all cutting forces had to be transfered through the spokes, which were glued to the hub and rim. I suppose the proper term is bonded, and those spokes were surely bonded to the rim and hub as they had no problem holding up to the cutting forces on the final operation.

Click on the first picture to browse the photos. Unfortunately I only have pictures from the first operation.

Last night after my small group studies, we all carved pumpkins. My camera only had enough power to snap a few pictures of my work, but everyone else’s pumpkin turned out pretty good as well. I was trying to make a tree silhouette with the sun rising on the horizon, but it turned out like this… It still looks pretty cool.

You can click the picture to see a larger version (and again to see it full screen).

(Click here to see it with the flash on)

I haven’t been to the shop in quite a while, but I managed to get over there today around 3pm. My grandpa had been working on the diesel compressor that we use for high-volume tasks such as sand blasting or sometimes spray painting. Over the past year it has been giving us problems to the point where we were ready to junk it (it was built in the 70’s after all). It would start fine most of the time, but then after about a minute there would be some struggling and smoking and it would shut down. When this happened, there was a pressurization somewhere in the engine that would make it impossible to turn the crank at all. After a few minutes, the pressure would bleed off and we could crank it back up again.

The other problem we encountered off and on was a complete loss of compression. We would crank it and then all of a sudden, there was compression and it would fire up and run. This lasted several months, then a few weeks ago it lost compression and we never got it back.

This week my grandpa and one of his renters took the valve cover off, and found that one of the valve keepers was completely broken up in pieces in the bottom of the valve cover. To find a valve keeper, my grandpa went to the place that he bought the compressor from. As of a month ago they had scrapped all of the compressors of the same model…. But lucky for my grandpa he ran into an old friend of his that said there was one more compressor that wasn’t scrapped with the others and was in the dumpster to go out with the next load. It was a 2 cylinder (ours is one cylinder) but the valve components should still be compatible.

So the scrap yard guys tore it apart and retrieved the parts that my grandpa needed for the repair, plus some extra parts- valves, keepers, rocker arms and push rods. The keepers worked great. We put the engine back together and fired it up. There was plenty of compression and the tank came up to pressure wonderfully, but just as it reached full pressure, white smoke poured out of the exhaust, and the engine died. The culprit was the mysterious over-pressurization.

After some discussion, we decided that it must be the compressor intake cutoff, and took that part off. Everything seemed to look good, we sprayed some blaster in it, moved the plungers, put it back together, and fired it back up. Same result. Stall with over-pressurization.

Still determined that the intake cutoff was the problem, we decided to pull a bolt that appeared to set the pressure for the cutoff. When we started to clean it, a hole appeared where we thought the bolt was just dirty. It turns out that a mud dobber had covered this hole, which was the atmospheric relief for the cutoff regulator. We put it back together and it started, and cycled the cutoff just like it did when we first bought it over 10 years ago.

It amazes me that a simple stinging insect can plug a hole with enough mud to resist over 100 psi of pressure, and send us on a wild goose chase to fix a 40 year old compressor. As I told my grandpa- it is great that these older machines are so simple to fix, but the catch 22 is that when a problem like this comes about, that simplicity compels you to fix it, even though it turns into a long drawn-out project!

My boss finally got his high velocity duct test contraption set up yesterday and showed us some interesting results. It’s the first time I’ve seen an example of static vs dynamic pressure in a working fluid. (even UF’s Fluid dynamics lab didn’t show anything like this)

the contraption is set up like this:
1/6 hp forward curved centripetal fan > 50′ length of 7″ diameter steel sheet metal duct > static regain diffuser (cone with inlet at 7″ and smoothly increases to about 20″ in 4′ of distance)> flow meter

there are 6 test ports along the 50′ length of pipe to test static pressure. He turned on the fan, and starting at the fan connection to the pipe, measured the static pressure at each test port:
length from fan: 0′ 10′ 20′ 30′ 40′ 50′ exit after regain diffuser
static pressure: .33″ .22″ .11″ .008″ -.1″ -.2″ atm (0″)
CFM flow at the meter: 870 cfm

Then he picked up the fan cutsheet and handed it to us. It reads 805 cfm @ 0.0″ water

so there are a few questions floating around now:
1) how does a fan move more air through a pretty high static pressure than it is rated to move at 0 pressure drop?
2) how do they test the fan?
3) there is obviously a ~.5″ friction loss through the pipe at ~870 cfm, but does the fan see .5″ or .3″ or is the fan actually seeing .1″ (.33+(-.2″))?

When the static regain diffuser is pulled off of the end of the pipe, and everything else is kept the same, the pressure readings look like this:
.46″ .39″ .32″ .22″ .11″ -.05″ ( the last test port was 2″ from the open-air discharge)
we can’t measure the air flow, but i bet it’s less than 870…

So the static regain diffuser is truly sucking air through the duct because of it’s conversion of dynamic pressure to static pressure, but does the fan see that? Or does the fan only see that the output static pressure is .33″ or .46″?

I don’t know… it’s going to take more research than I’m willing to commit at the moment.

During disassembly, there were several parts that were broken because of poor handling on my part. Several more parts were already broken, and several parts missing completely. Except for the Flywheel, which I will elaborate on further in another article, the most disturbing part was the cam shaft. After several weeks of soaking, it still seemed hard fast.

So I put a little leverage on it, and of course it snapped right off. I dropped the nut with the broken end of the shaft into a small can with blaster and left it for several weeks.

One night after an unusually depressing day, I felt like taking some frustration out by freeing that nut. This was probably not a very good idea since I had the notion to get it free no matter what it took. I pulled it out of the blaster, put it in the vise, and pulled out the torch to prepare to use heat to get it free. Before I fired up the torch I thought I would try to get it out as it stood, so I stuck the easy-out into the hole (nicely provided already for the governor actuator shaft) And put a nice 12″ adjustable wrench on it. It started to turn with almost no effort! BUT it was turning the wrong way! If I had studied this just a little bit closer, I could have noticed that it was a left hand thread, not a right hand thread, and prevented the next bit of headaches.

So after freeing the nut from the shaft, the next task was to get the shaft put back together. My grandpa suggested using silver solder, so that was what I tried.

The first attempt to solder this together seemed like a good idea, but failed terribly. We cut a ring of silver solder sheet the same size as the shaft and put a long bolt through the center of the long end of the shaft, slipped the solder ring over the bolt then put the broken end of the shaft over that- sort of like a sandwich. A spring was used to apply pressure to the broken end as the solder melted. I used the torch to heat the whole assembly to the melting point of the silver solder, but after the solder melted and I removed the heat, the broken piece just fell right off again. Somehow my grandpa was able to repair it the next day while I was at work. I believe that he heated the pieces first separately, applied solder to them (tinning them) then put them together with heat. It worked beautifully. After cleaning the threads, it is as good as new.