Sorry, ronmar I did indeed misquote you. You said generator weight, not "engine weight." My only excuse was I continue to think in terms of balancing an engine, and so read what you said inaccurately. Apologies.

Continuing, I don't believe that the mass and acceleration of the piston and rod alone account for all of the vertical forces involved, so I would very much have liked to see you calculate what they were, for accuracy sake. Saying they would have to be 50 lbs to cause the case to move vertically is not something I necessarily disagree with, but it isn't relevant to me since I see the the added weight's contribution as part of the picture, not all of it.

We started out with a statement that we had 8.8 pounds of weight opposing 1000 lbs. (1% you said) and by now have moved on to say it was 135 pounds of force opposing 750 lbs. A very different picture of the forces involved, and an important conceptual modification. But the 135 figure is also wrong and needs to be increased, I think you agree.

If you don't mind, in the interest of everyone's else's general understanding, I'll try a lazy method of calculating it, please object loudly and correct me if it doesn't seem accurate or fair:

You took 65% of the conrod mass as rotational and gave a figure for that as 13.64 pounds (I believe). I'd like to grab the remaining 35% if that's okay and call it reciprocal. That seems to be 7.34 pounds (unless I misunderstood your earlier statement.)

Now if the 8.8 pounds you calculated earlier yields 135 pounds of force and we add another 7.34 pounds to that weight, we might guess that the force increases proportionally. So that looks like another 113 pounds or so of force. Total force tending to lift is now 248 pounds. Not enough to do it yet on its own, of course, as you say, but very substantial.

However, what goes up must come down. We also have a force of 248 pounds force headed down every tenth of a second or so and through the crankcase trying to compress whatever is beneath the engine. If the engine is bolted solid to massive concrete, bedrock, etc. we naturally can't bounce.

If it's not, we may bounce more or less depending on what we are on and how attached. That is usually because of an equivalent form of spring compression beneath the engine, and it is caused by the engine appearing to weigh 1000 pounds  one moment, and 500 pounds about a tenth of a second later to whatever it is on.

In a worst case resonant scenario, our spring will store 1000 pounds of energy, and release that against an apparent 500 pound weight every tenth of a second. In this situation, without damping, as in the case of the Tacoma Narrows Bridge which destroyed itself in a moderate wind, our engine may toss itself overboard.

That's unlikely to occur, unless someone intentionally tuned for resonance, or was incredibly unlucky in their choice of mounting. Luckily with damping, restraint, and/or a different spring resonance, we will see less vertical movement, but not necessarily no vertical movement, unless an engine is completely restrained from downward compression.

Restraining is not the same thing as balancing. You can definitely stop vertical and horizontal movement by preventing it, without balancing at all. Is that desirable? That's an open question. Noise and power pulsing transmitted through the ground is often reduced by a compressible mounting system.

I have not added in calculations for the effect I also asked about because of the offset center of mass of the genset .The reciprocating forces would then act on a fairly long moment arm offset to the side of the engine. But we've already got enough going without it to get things moving vertically. I think it isn't necessary to make the case further.

An engine can obviously move vertically, depending on many factors. The forces involved are not trivial -- an 8.8 pound piston and conrod can generate far more force than the numeric relationship of its weight to the overall weight of an engine.

My original question that started this was whether chalking on top of the flywheel would help in balancing an engine's reciprocating masses. I didn't understand why one would chalk in front of the engine. I do now understand that in some specific cases rotary balance is poor, (as in the case of a non-counterbalanced engine) and a particular engine is suffering from excessive fore and aft movement. If that engine is normally mounted to be restrained vertically, and ground pulsing is not a problem, then balancing fore and aft 100% for that case is fine.

My engine is well restrained fore and aft, and less restrained vertically. It has a buried railroad tie in graduated sand mount. The ties cannot move horizontally but are freer vertically. The genset frame is welded heavy channel.

I do have a somewhat objectionable amount of ground transmitted pulsing, as well as channel edge vertical flexing and I wondered if balancing could help reduce this. Balancing fore and aft didn't make much sense to me in this situation. Adding a rubber mat under the channel, as sometimes described to reduce noise might add to the vertical component, so I wondered if I could get better vertical balance with the chalk method on top of the flywheel. Hence the original question.

The proof of the pudding is in the eating. The real way to tell what is happening is to try it. At some point I'd like to loosen the mount and add rubber underneath, then chalk from above, to check periodic vertical movement, and if present, add weights to shift the balance percentage as described elsewhere. If not, report to ronmar, on my fool's errand.

As a side note, and reason why I haven't so far, one difficulty I have is that the ground actually moves vertically now in the vicinity of the engine, so a steady rest placed on the ground is going to be well, less than steady. I'll have to figure out a way to get a stable marker before I can get any results, one way or another..

"Without these weights the engine had tremendous lifting force at pretty low speed  Shocked, I couldn't believe how much force it lifted with!"

Hello Leland. Hmmm, lifting. Would you say that was more than .001"?

The piston/wrist pin weight is about 8.8#, that combined with the mostly vertical moving part of the rod means that only about 1% of the all up weight is trying to lift the the generator up and down.  It just won't move with that little energy input, especially considering it is a smooth sinosoidal acceleration/deceleration.  It would have to impart greater than 1000# of force to even start to move the assembly vertical.

Sorry, I don't buy it.

You think the fully absorbed impact force of an nine pound mass at velocity is still nine pounds? So since you are able to divide 8.8  by 1000  (for your full genset) with a calculator, and hit the percent key, the force of the piston rod assembly is proportional to its stationary weight? And so therefore can't move the genset? Uhhhhhh, no.

Sorry. Here's an experiment, drop an 9 pound iron ball from say 15 feet onto a scale, while taking a high speed movie of the scale dial. Then run it back frame by frame. and tell me if it reads 9 pounds for the whole sequence.

Look, just think for a second. What do you think is the point of the Lister's cast-in counterweights? Don't tell me two sizable chunks of cast iron at about 20" diameter are simply there to oppose a couple inches of crank rotary mass!

Sorry, yes, those counterweights are designed to oppose the vertical reciprocating forces. Of course they do an approximate job, since they're rotary --as I already pointed out. So you necessarily get horizontal oscillation if you put your engine on rollers.

Okay, we're determined to chalk this rollered engine in front, not on top. Hmmm, how do we re-balance to get the chalk even?

Well, we need to actually saw the Lister balance weights off, and add a small weight equal to the crank's swinging mass at only a few inches out on a spoke. Just like the crank.

Gosh, that's a big weight savings. We've exactly countered the crank, and rid the engine of those destabilizing rim weights pushing the engine horizontally back and forth across the rollers, making chalk spots on the rim. Of course our engine started jumping off the rollers and pounding them into the ground, which wasn't too good for chalking either, since the chalk got smashed when the engine tipped over.

I'm a little confused here. Why are we balancing fore and aft instead of up and down? The piston is moving up and down. The flywheels have balance webs that are (I thought) supposed to counteract the vertical reciprocating weight of the piston and rod. Because these counterweights rotate rather than reciprocate like the piston, they will naturally create fore and aft movement if the engine is unrestrained in this direction. Putting the engine on rollers will show this up. But fore and aft movement would be expected, wouldn't it?

I don't see how we can balance both ways. If we reduce the fore and aft movement to zero, we would have to remove the counterweights and make the flywheels themselves balanced, wouldn't we? Then the engine would be unbalanced vertically. That's where the real pounding takes place, isn't it?

So, shouldn't we be placing our chalk on top of the flywheel not in front of it. And shouldn't the engine be restrained horizontally on vertically springy mounts to check the balance --- the proper amount and location of flywheel counterweights to compensate for the piston and rod?

Just want to thank bschwartz for the water pump -- it arrived yesterday. Really wonderful folks on this forum!

I didn't have a chance to take it completely apart. I took off the pulley, and the cover plate on that end, and the outlet plate on the other end. The pump is identical to mine on the exterior.

What I could see so far was a well packed (greased) ball bearing, unlike mine which had little grease, or, more accurately,  black sludge and bits of metal in mine.

I didn't remove the snap ring retainer in the casting -- on mine the bearing was not a slip fit in the casting, it would require pressing out, but it also had a retainer ring. Not sure why that is.

There is no retainer on the shaft at  the pulley end, so something internal is preventing removal of the impeller and shaft assembly. Unless the impeller can be removed from the shaft, the only way to remove the shaft is out the impeller end. But I'm not sure what is retaining the shaft inside. A careful pull by hand didn't work. Hesitant to try to press it out, yet.

Looking at the impeller, I can't tell if it is pressed onto the shaft or screwed on. The end of my impeller looks a little different than the new one. The new one might have a round nut on the end, mine looks pressed. Will take and post pix when I get a chance.

To answer questions:

@M61hops, yes I thought of that, too, but there's no evidence of a sleeve bearing in the casting in mine. The hole through the casting is much larger diameter than the thick shoulder on the shaft. There was never a sleeve bearing in my water pump. Maybe bschwartz's version will have one, and mine was left out. Also the thick portion of the shaft isn't machined, so doesn't look like it would run in a bearing (unless it was supposed to be machined!).

There was a sort of off-center hole in a "bulkhead" or wall in the casting at the impeller end. But hard to believe that part of the casting was intended as a bearing, or that it would have been worn so wide. The edges didn't look like there was recent wear -- it looked like sand cast finish. It was maybe 1/8" thick. Look at the third photo above, I think you can see the hole. Hard to imagine as a bearing.

On the shaft, I do see a machined area, about 1/2" long right about where you'd expect a second ball bearing at the impeller end. There's also a recess in the casting that looks like it would take the outer ring of a ball bearing there. But no groove for a retaining snap ring (as there is at the pulley end). The machined areas on the shaft at both ends look the same, so I'm leaning toward the idea that there should be ball bearings at both ends.

True, a sleeved bearing would make more sense considering the grease cup location, though -- right in the middle of the shaft housing. Hard to figure what that's for. It would take about 20-50 cups full to fill the casting space and reach either bearing! My housing was empty of grease.

If I can figure out how to take apart the new pump without wrecking it, I'm sure we'll have an answer. It feels good. With the ends off there's no side-play I can feel nor end play. Something is acting as a bearing at the impeller end for sure. (Maybe a bearing!) It does turn a little stiffly -- probably the seals when turned by hand without water in the pump

@bruce: yup. I already have some electrical circulators on-hand, but am interested in figuring this mechanical METRO pump out anyway. It's a challenge. Even if I do go eventually with an electrical or thermo-siphon system, we'll have the information about how these pumps work -- or if they don't -- why they don't, and what can be done about it.

Just curiosity.

 

That first pic with the broken bearing is of the pully side of the housing?  What is that thing on the shaft right next to the back side of the impellar?  It looks like it might have supposed to have been a bearing of some sort.  It looks lke there is supposed to be some type of O-ring on the back of the impellar plate.  Is this correct?  I don't think any part of the impellar would be used as a bearing sutface.  Because of the nature of a centrifugal pump, the center of the impellar must be clear to allow fluid to enter the pump at the center of the impellar disc. 

I don't think I have heard anything positive about one of these pumps, or of anyone getting any long term service out of them...     

Thanks ronmar. To answer your questions:

1.) Yes, the first pic is the pulley side of the water pump
.
2.) The thing on the shaft next to the impeller is the seal. It is a rubber dome on one side and a sheet metal disk on the other side. The rubber side is shaped like a dome with a hole in it. The hole side bears against the back of the impeller normally. The back of the impeller has a ring of plastic bearing material inset into it (what you thought were O-rings) that bears against the rubber cup. At least that's what it appears like to me.

There is no bearing at all at this impeller end of the shaft. I believe there should be one and it was left out by the assembler. But that's just my guess.

I am curious, because I have the means to rebuild the pump (mill, lathe, etc) if I just understood how it is supposed to be. I hope someone here is familiar with water pumps and how they should be inside, or has actually opened one of these particular ones up.

I realize that the originals weren't very good or long lived, but there's no reason why I can't rebuild it so it works well and lasts long, given proper machining, good bearings, and seals. Mine must be unusually bad. It lasted 5 minutes with a (probably) missing bearing . I imagine others at least have all their bearings!

The big problem I'd like to solve is just finding out if there is a missing bearing, or not. Maybe someone has a picture of one, or manual page, or just the commercial experience of rebuilding water pumps and can help with information on how this one should be.

Thanks Ronmar. That's good news about the 1" dia thermosiphon.

I'd planned on breaking in under load for a lot longer (I do have a ST-5 belted to the engine), but the pump self-destruct put an end to it sooner than planned. I didn't want to run electrical loads until I knew more about out of the box RPM, etc.) I think the 2x4 against fulcrum and flywheel was doing a good job of loading the engine. I was able to drop the RPM, and I probably could have stalled it with more pressure. That's a brake on the flywheel with a 10' lever. I checked the revs with an optical tach I had already. Put a piece of reflective tape on the flywheel rim. As set up it ran 677 RPM. I dropped that with the governor adjustment to 650.

Vibration wasn't too bad. Subframe is on rairoad ties in wetted mixed size sand, per utterpower mention. It was amazing how quietly the engine ran after the water pump seized! Most of the mechanical noise was coming from that bad bearing.

I'm still curious about that single bearing at one end. Is that normal? I'll post pictures later.