Rick's Overbalanced Water Wheel Concept

[rickoff]

Okay now. Onward ho! As to the construction problems that I mentioned in my previous post, here's what I ran into. You may remember, back in post #40 I had stated a concern about whether the 3/8 x 6 inch shaft would be long enough to allow adequate space between the uprights so that the pipes, when assembled to the wheels, will not run into any interference from the uprights when the wheels are turning. I had calculated the overall width of the wheels assembly to be about 5.5 inches, and the space between the uprights to be about 6.375 inches, but that wasn't quite accurate. For one thing, while the Medium Density Fiberboard that both wheels were made from is listed as being 1/4 inch thickness, as seen from the label photo in post #52, the actual thickness is about 0.260, so that added 0.020 extra width. Also, while I had estimated that there would be an intervening space of 1 inch between the two wheels, the actual length of the rounded top pipe caps I had used was actually 1.082 inches, so that added 0.082. Lastly, the interference which came into play was at the mid-point of each upright, where the long stabilizing links were fastened. These are fastened to the inner sides of the uprights, and between the links being 1/8 inch thick, and the attachment screws adding another 1/8 inch, that decreased the 6.375 inch clearance between the uprights to 5.875 inches. Now you'd think that the actual 5.782 width of the wheel assembly would still allow about 0.046 of clearance to each of the two uprights at the mid-point link connection, but that didn't hold true, as the interference occurred due to the U-shaped plastic clamps which hold the pipes against their support blocks. While I could have solved that problem by re-positioning the long stabilizing links to the outer sides of the uprights, I didn't want to do that because the outer side, instead of being flat, is a channel, and thus attachment at the outer side would not have offered the stability and rigidity that was needed. So, I knew that my best option would be to look for some flat top pipe caps to use rather than the rounded top ones which I had installed. Thank goodness I had decided to install the rounded top caps with some silicone sealer instead of PVC pipe cement! I was able to remove the rounded top caps easily by twisting them with my channel lock pliers and pulling them off. I was able to order some flat top pipe caps online, and when I received and measured their height I found it to be 1.022 inches, or 0.060 less than the rounded top caps. I wanted to add even more clearance, though, so decided to cut these down to 0.750 inch height. The inside bore of the cut down flat top caps features a shoulder at a depth of 0.500 from the open end, so this required that I also cut the short pipes which extend from the elbows, and through the wheel thickness, down to where they would protrude less than 0.500 inch past the inner surface of each wheel. I decided to make that 0.400 inch to leave about a 0.100 space where I would lay a bead of silicone sealer. Here's a photo showing the completed cap and pipe cutting revisions:

Inner-pipe-end-cap-revision-details.jpg
Since the pipe caps for each wheel are staggered 30 degrees apart from each other in the rotation, and the wheels are locked to the shaft, no cap will come into contact with another. This revision would allow the flat top of one wheel's pipe cap to rest against the inner surface of the opposite wheel, except for the fact that each of the flanged shaft collars which I had made by epoxying a shaft collar to a heavy flat washer (as also shown in post #52) had an overall width of 0.500 inch, forcing a 1.000 inch intervening space. I decided to take the assembly apart and remove one of the shaft collars from its washer, so that I would decrease that space by the width of the shaft collar, or 0.375 inch. When I did that, I found that the collar came off the washer quite easily, and too easily in fact. The epoxy bond was not nearly as effective as I had hoped it would be. So, to avoid messing around any further, I looked for pre-made one piece flanged collars on the Internet. I did find a few that were made with a 3/8 inch shaft size bore, but they were very expensive - around $30 to $45 each. I noticed however that what came up in my searches more often, for flanged shaft collars, and at a far more reasonable price, were ones made to fit a 8mm shaft (0.315 inch), which of course would have a bore 0.060 inch smaller than what I wanted. Seeing as these came in a set of 4 for just $10.99, and already had mounting holes pre-drilled, I decided to get these and modify them as necessary. That only required passing a 3/8 inch drill bit through the shaft bores, and by passing a 11/64 inch drill through the mounting holes I was able to re-use the mounting screws which I already had. Here's a link to where I found these flanged shaft collars at Amazon.com. Notice that each of these collars uses 2 set screws, which makes for a very secure hold on the shaft. I only used 3 of these when assembling the wheels to the shaft, as I reused on of my heavy washers as the inner hub on one wheel. Even though this allowed bringing the wheels much closer together, it was not quite enough so that the flat pipe caps would rest against the inner wheel surfaces. There remains a 1/16 inch space. The reason for that is illustrated in the below shown photo. The rather large Nylock nuts securing the heavy washer (at right) come into contact with the shaft collar seen at left. I didn't expect this to occur, since the same size Nylock nuts secure the flanged shaft collar seen at left, and these do clear the shaft. The only explanation, of course, would be that the factory drilled mounting holes on the outer right flanged collar were drilled closer together at the factory where made, and were probably from a different production run than the flanged shaft collar seen at left. This situation could be cured easily enough by reversing the screws holding the washer and the outer flanged shaft collar to the wheel at right, but I haven't done that yet, and it's really not a pressing matter because I now have plenty of rotational clearance and the wheels are securely mounted to the shaft with no wobble.

Flanged-inner-shaft-collar-to-washer-spacing.jpg

[RAMSET]

Rick
Time well spent [in the studio... sounds great !!....I have a smile that just won't quit now....

Your song link again. https://onedrive.live.com/?authkey=%...%21130&o=OneUp [hope this longer link works]

Good to see you back at the build too.
respectfully
Chet
PS
I think you should link the music in your Tag area below each post..you guys sound great.

[rickoff]

Here's a photo showing an outer flanged shaft collar attached to a wheel, with the shaft extending outward, through a plain shaft collar, and into the bronze bushing held by the wood block attached to the upright.

Outer-wheel-and-shaft-connections-at-upright.jpg
This view, showing the socket head screws attaching the outer flanged collar to the wheel, makes it clear how reversing these screws would solve the contact problem shown in the last photo of my previous post. When mounting the shaft to the uprights, I first set the outer plain shaft collars at a position that would allow a 3/8 (0.375 inch) entry into the bronze bushings on each side. I also applied a small amount of light oil to the inside bores and flange faces of the bronze bushings. With the two screws removed from one of the short stabilizer link bars at its attachment point to the upright, it was easy to pull that upright away from the opposite upright enough to slip the shaft into the bushings and then re-insert the stabilizer link's attachment screws. Next, I loosened one of the plain shaft collars and slid it inward on the shaft just enough to leave about a .005 inch gap between its outermost side surface and the bushing flange, just so as to relieve any frictional pressure between the collars and flanges.

With that done, the next step was to align the two wheels properly. You may have noticed the 30 degree radian line, in the above photo, that I had added to the outer surface of the wheel. That would be the vertical line drawn between the two 60 degree radian lines. I extended that line over the top edge of the wheel, and also extended a 60 degree radian line over the top edge of the opposite wheel, so as to make proper alignment of the wheels a simple task. here's a photo showing those markings.

Alignment-marks-at-30-and-60-degrees.jpg
After loosening the set screws of the outer flanged shaft collar which mates to the heavy washer at the inside surface of the wheel, I rotated that wheel to the point where the edge marks aligned. With the wheels in correct 30 degree offset alignment, I then snugged up the set screws of the outer flanged shaft collar to lock it to the shaft. I like the idea of having the two set screws at each flanged collar, as that will make it highly unlikely that the alignment could go out of whack.

Next, here's a photo of a front view showing wheel and pipe details on both of the wheels, and in the space between them.


As you can see in the photo directly above, I reconfigured the wheels so as to place all the clear pipes on one wheel, and the white PVC pipes on the other. Previously, I had placed just two clear pipes on each wheel. I figured that placing all the clear pipes on one wheel would make for a better visual effect, making it easier to follow the flow of water in the tubes during rotation. This required making two new clear pipe assemblies, and one of these two is found at upper right of the photo. You can see that the 3/4 slip x 3/4 FPT connector on this new pipe is different than the older type shown at lower right. Both are made by Lasco, and have the identical part number, but they have simply changed the design to drop the 0ctagon ring and use 6 lengthwise protrusions instead. The new one is also 1/16 inch shorter than the old one. Like everything else, this shows that a simpler design using less material saves the manufacturer some costs, even though we pay the same or more for the new item. I don't think this is going to be a problem, though. In making the pipe extensions, I will first measure every pip from the elbow junction to the extreme outer end of the 3/4 slip x 3/4 FPT connector fitting. Although I have figured that the extension pipes will basically be cut to 8 inches length before adding the remaining fittings at each end, the most important factor will be ending up with each full length pipe being as close to identical as possible. The pipes already mounted on the wheels will have slight variances in their measured lengths, so if a pipe is 1/32 inch longer than others, I would cut its extension pipe to 7 and 31/32 inches length, for example. Conversely, if a pipe measures 1/16 inch shorter than the others then I would cut its extension pipe to 8 and 1/16 so as to equal its overall length to the others. I have already measured and cut the extension pipes for the white PVC pipes and have cemented the fittings to them. Tomorrow I may have time to install those extensions, but it may be another day or two before I get around to measuring and cutting the clear pipe extensions. In any case, I should have all the extensions completed and mounted sometime this week. The next step after that phase is completed will be checking the balance of the wheel, and making adjustments if necessary to compensate for any out of balance condition. I'll probably made a video showing how I do that. Basically, that phase will involve spinning the wheel into rotation, marking the location where it stops rotating, and doing this several times to see if it tends to stop in the same position. If it does, the heaviest portion of the wheel assembly would be at the bottom, where marked, and some amount of weight would need to be added directly opposite, at the top of the wheel. Then spin testing would resume to determine if that corrected the imbalance, or if more (or less) weight should have been added. Once decent rotational balance is achieved, I'll be ready to start adding in the water. I plan to start that phase by adding an amount which would precisely equal half of each pipe's capacity, and as I mentioned earlier, the way to do that is to completely fill one of the pipes and then dump all its water into a graduated cylinder so as to determine the full volume amount. Once that is known, we simply halve that amount for the filling of each pipe. I'll plan on adding some coloration, such as blue food coloring, to the water in the clear pipes, to make the water flow more easily discernable.

[rickoff]

Hi, Ramset. Glad to hear that you enjoyed the song that I gave a link to. We have ten songs on the CD, and this isn't necessarily the best one. It all depends upon what people like, of course. Maybe I'll throw in another song link soon.

Something I forgot to mention in my last above post is that the use of the FPT connectors at the ends of the existing pipes may prove problematical. Originally, I had expected to use these on wheels half the size of these ones, and wouldn't have needed extensions, but with the larger wheels you can see that extensions definitely should be added in order that the build will look similar to the concept diagram shown in post #2 of this thread. If I had planned to use two-piece pipes, rather than single but longer ones, which wouldn't make much sense, I would have used slip joint unions to join the pipes instead of terminating the first pipe with a 3/4 inch female pipe thread. Whereas a slip joint union would have a through-pass bore the same size as the pipe bore (which happens to be 0.815 inch), the inner bore of the male threaded connector for the extension pipe is only 0.750 inch, so there is a reduction of .065 inch in bore size at that point. This will have some amount of negative impact upon water flowing outward into the extension pipes, and again when flowing back inwards towards the wheel. I'm hoping that this negative impact will not be great enough to cause failure of the wheel to self-rotate. I thought about using a hair dryer to heat up the end connectors so that I could remove them and replace them with slip joint unions. That's a well known trick that plumbers use to separate a cemented fitting from PVC pipe, and though it would work well enough on the white PVC pipe, the clear pipe appears to have a much lower heat tolerance. I tried this method on one of them and the pipe started to distort when I attempted to twist off the fitting. Thus, I'll have to use the threaded connectors and see what happens. Hopefully the flow reduction won't be that great, but it also means that to pass over that smaller bore section the pipes will have to be at a lower angle to accomplish that when water is flowing outward, and a higher angle than would otherwise be necessary when flowing inward. Time will tell, and I hope this doesn't cripple the build, as I'd hate to have to scrap these and make up all new pipes. The clear pipes are very expensive, too, so that makes a re-do quite unappealing.

By the way, here's a photo showing a side view of the wheel assembly mounted on the support stand, with the clear pipes facing the viewer. Sorry for the background distractions, but that was the best place to situate the apparatus at that time due to daylight lighting conditions. More photos, and some video, will come along soon.

[rickoff]

Shown below is side B of the wheel assembly, the opposite side to that shown in my last above post. This view shows the extension pipes installed and tightened, which completes the construction aspects for this side. Rotation, in this view, would of course be in a clockwise direction, while a view of side A would show a counter-clockwise rotation. The plugs shown at the outer ends of the white PVC pipes were only installed finger tight, and simply to keep out any dust or other foreign matter. Hopefully I'll be able to work on measuring and cutting the clear pipe extensions tomorrow.


No water will be added to the pipes until I test the rotational balance, and adjust that as necessary. Currently, the balance appears to be quite good, and no doubt that had much to do with careful layout of the wheels, on-target cutting of the wheel bores, and matching the overall lengths of each pipe. When water is added to the pipes, this will of course throw the wheel assembly off balance, but in this build that will be a good thing because it will induce rotation. How well the wheel rotates, and whether or not rotation is continuous, remains to be seen. The most advantageous amount of water to add to each pipe is also an unknown at this point, but it seems reasonable to assume that it will be half or less of the capacity of a pipe, and less may prove to be better. Less water equates to less applied rotational force, but could be more advantageous when water lies at the inward end of a pipe because its center of gravity would be closer to, or inside of, the 8.5 inch radius of the circle that the wheel's bore holes are centered upon. Since the pipes are straight, and the circle is curved, there is a certain length of pipe that the circle arcs away from (outward) before finally crossing back under the pipe, and that section will have the greatest advantage in keeping water as close to a vertical centerline of the wheel as possible, thus lessening counter-rotational force. I may end up doing several experiments with varying water fill amounts before finding what works best. The primary objective will be to achieve self-induced rotation due to one half of the wheel (divided by a vertical centerline) being heavier than the opposite half. In other words, no matter what position the wheel is in, it should begin to rotate if it is not being held in place. Such self-induced rotational movement would of course prove that the side rotating downwards is heavier than the opposite side, and that the heavier side causes an "overbalance" condition to exist. The second objective will be to achieve continuous rotation, and for that to occur the primary objective must first have been met, and the overbalance condition must continually (or near continually) exist at every point in the wheel assembly's rotation. That, of course, will be the most difficult objective to attain, but I do remain hopeful that this can be done. As I have stated earlier on, increasing speed of rotation is not an objective. The wheel assembly, if continuing to rotate, will rotate slowly. This is desirable, of course, because it allows time for water to transverse the length of the pipes when moving outward or inward due to a pipe's angle of incidence. If speed increased then centrifugal force would prevent water from flowing back inwards when otherwise appropriate due to the pipes's angle of incidence. Thus the wheel assembly would self regulate the speed of it's rotation. Slow but smooth rotation would be the best possible result, though if the hammer effect in outward flow is substantial enough then movement may tend to be somewhat jerky. Inward flow would have less of a hammer effect because of the curve of the pipe elbows, which would direct much of the force against the side of the wheel.

[BroMikey]

You sure are a stickler for detail with good mechanical skills. That is so important and hard to find.
Nicely done Rick I like the frig channel choice.

[thaelin]

Hi Rick:
Glad to see you back with us. I keep getting the feeling you should have heavier support for the wheels on this. That thin and small of an angle bracket scares me. I have played with the movement of water in a container and it can get out of hand very quick. Water "hammers" can be very destructive on the containment. Just my thoughts.
But as usual, your build is right up to par with the others you have done.

Say hi to the other members of the group, loved it. Want the CD when ready.

thay

[thx4]

@Rickoff
A big thank you for continuing to make us dream, the model should soon tell us more.

[rickoff]

Glad to see some people are following the build as it approaches completion, and to see your comments.

Bro Mikey: the channel stock used for the uprights is actually built for use with shelving. Screws would be passed through the 3 pre-drilled holes into a wall, and shelf brackets would be inserted into the slots. These do look similar to refrigerator shelving channels, but are more heavy duty. The thickness of the metal is 0.070 inch.

Thaelin: From reading in this thread you are most likely aware that I originally set out to create a build using wheels half this size, and the original pipe assemblies without the extensions that I have added. The plans changed when I realized that I couldn't achieve the desired 68 degree angle with the smaller wheels, and I wanted to build this as close to the design shown in the concept drawing of post #2. I had planned to use the 3/8 inch X 6 inch shaft with the small wheels, and I still think that is sufficient for this larger build. In building the supporting structure, I basically tried to use items which I already had on hand, as much as possible, in order to cut down on expenses. I realize that the 1/8 inch x 1 inch aluminum stabilizer links between the uprights and the L-bracket anchor pads are the weaker elements of the build, but I do think they will work adequately for testing purposes. As I said in an earlier post, I would have preferred using a wider base, but this was the largest that could be made from the birch plywood which I had on hand, left over from another project. With these relatively small bore pipes, I don't believe that the water hammer effect will be all that substantial. I'm guessing that it may cause some amount of jerkiness, rather than having a smooth rotation, but I expect rotation will be very slow - perhaps 6 rpm or less, so I'm not very worried about the unit self-destructing. If things go well enough with the prototype, I'll plan on building a larger model which would have the potential of being more than just an interesting conversation piece. For that I might use 3 inch or 4 inch PVC pipes and would of course construct a heavy duty support fixture, probably using my dodecagon wheel design and the heavy duty flanged bearings shown in post #11.

thx4: Thanks for your kind and encouraging words. Yes, this build will soon have something to tell us one way or another, and I'm quite sure of that. Since my last above post, I have been working on completing the "clear" pipe extensions, and I also decided to go ahead and reverse the screws that I spoke of in post #61, and which caused the clearance problem seen in the last photo of that post. That solved the problem, but of course required dismounting the wheels from the support structure and removing one of the wheels from the shaft. Now the two wheels are gapped about 3/4 inch apart, and sandwich the flat pipe caps quite nicely. I was able to complete the clear pipe extensions yesterday, and mounted those today after getting the wheel assembly back into place on the support structure. Here's a photo showing the completed prototype:



I'm sure you will agree that this really is beginning to look a lot more like the concept drawing from post #2. As you look at the bluish tinted "clear" pipes, you will notice that the outer extension pipes appear to be a different color than the inner pipes, but darker coloration is simply caused by the brownish color of the wheel blending with the light blue pipe coloration. When rotated, the closest that the pipe plugs come to the wood base is about 1.750 inches, and that would be at the point where a plug is in line with both uprights. I checked the balance by spinning the wheel assembly several times, and marking the lowest portion, each time it stopped, with a piece of masking tape lined up with an upright. I expected that the wheel assembly would show some amount of imbalance, but on each spin it stopped in a different place and never rolled back in the opposite direction or rocked back and forth any amount whatsoever. It was hard to believe I could have been this lucky. I'm going to check again tomorrow to make sure that the shaft is not misaligned with the bearings, and that the outer shaft collars are not causing unnecessary friction against the flanged bronze bearings, but it did appear to rotate easily. If that looks okay then I'll fill one of the pipes full of water and dump it into a large graduated cylinder so I can determine the actual water volume capacity, which will the allow me to fill each pipe to a percentage of that amount. Fifty percent is what I have assumed will be a reasonable volume to start with, and I'll see how that lays and reacts in one of the pipes before making a decision on what may actually be the best starting fill. The first try at getting the right amount of water fill will really be just a wild guess, and likely may need to be adjusted to improve results. There are several unknowns, and only time and experimentation will tell the tale.

Best to all,

Rick

[citfta]

That is a beautiful build Rick! As always with your builds it is easy to see the craftsmanship you put in to what you do. I can hardly wait to see the results. Do you plan to video the final results so we can see it in action?

Take care,
Carroll

[rickoff]

Hi Carroll,

Good to see you are watching, and thanks for the compliment, though there's really not a lot of craftsmanship involved. I think that any handyman/experimenter could build this with very little difficulty. I try to let people know exactly what I am doing, as far as materials, methods, and measurements go (the 3 M's). Yes, I will be recording some video showing what happens after all the pipes are filled to 50% capacity. That may not be the "final results," though, as I may have to do some experimenting with water fill percentages to see what works best (if at all, of course). I may prepare a video tomorrow showing how water flows in a single clear pipe as the wheel assembly is turned. I think that might be interesting to watch. I was able to figure the 50% capacity water fill for one of the clear pipes, and marked black dots upon the pipe at 3 intervals which represent the following: The 50% fill point, 1/2 the measurement from the outer end of the pipe to the fill point, and another dot at 1 and 1/2 times the measurement from the outer end to the fill point. From the outer end going inwards, this provides the outer filled portion's center of gravity point, the mid or "fill point" (which is the same point whether the pipe's outer end is facing straight up or straight down), and the inner filled portion's center of gravity point. These points were then duplicated on the remaining 11 pipes.

To estimate the 50% capacity water amount, I filled one of the clear pipes to its maximum capacity (to top of the FPT threaded fitting at the outer pipe end) while the pipe was held vertically. Next, I installed the pipe plug and tightened that down until there was a 6 millimeter gap between the underside of the hexagon plug head and the top of the FPT fitting. Of course that forced some water to spill out because of displacement. Before removing the plug, I rotated the pipe to a horizontal position, and when I did so I saw an air bubble move across the pipe. The reason for the air bubble is that the plug is made with an inner recess. I then turned the pipe upwards again, removed the plug, slipped a 250 ml (milliliter) graduated cylinder over the top and rotated the pipe downwards until the water was emptied out. The amount of water in the graduated cylinder was at the 216 ml mark, so I halved that to 108 ml for an initial fill, reinstalled the pipe plug, and marked the fill line both with the pipe straight up and straight down. Here's a photo showing the result, which came out close but not reaching the center point between these fill lines until I added more water. After knowing where that center point lay, I removed the plug again and added enough water to reach the center point, which turned out to be an additional 2 ml, bringing the total to 110 ml.



As you can see in the above photo of the "I" pipe, the mid point was situated at the pipe's support block, so I removed the plastic U-clamp in order to get a better view. The lower line was drawn at the initial water fill level with 108 ml water added, and the top line was where the fill level was with the pipe inverted. The center dot is the mid point between those marks, and where the water settled after adding an additional 2 ml.

Next, I measured the distance from the top of the FPT fitting to the fill line dot shown in the above photo, and this measurement came out to 304.5 mm. With that known, I halved that amount to find the center of gravity for the outer portion, which comes out to 152.25 mm, and tripled that to find the inner portion's center of gravity point. The below photo shows the markings, and the related measurements.


Note how much closer the inner center of gravity is to the elbow fitting than it is to the FPT fitting at the outer end. This of course occurs because water flows through the curve of the elbow, and then an additional 0.850 inch through a short pipe as it passes through the 0.250 inch thickness of the wheel and terminates at the inner surface of the short pipe's flat end cap another 0.600 inch along. This feature brings the inward center of gravity point inside of the 8.5 inch radius circle which the wheel's pipe bores are centered on. You can see how the arc of that circle moves away from the right side of the pipe for some distance before passing back beneath the pipe and meeting the elbow at the next bore hole. This inward advantage of being closer to the center of the wheel was one reason why I chose a pipe lay angle of 68 degrees from its radian line. The other reason was to allow the pipe to become horizontal well before its radian line reaches the 12 o'clock position, thus allowing its water to begin traversing towards the outer end of the pipe. Conversely, this also allows a pipe at the bottom of the wheel to go horizontal well before its radian line is in alignment with the holding fixture upright. I'm not sure how many degrees before the radian is vertically aligned with the upright, but that shouldn't be too difficult to approximate. What counts, of course, is the point at which the water traverses, which may not be exactly when the pipe goes horizontal.

Now here are some things to consider:

• A pipe lay angle less than 68 degrees would allow water to traverse even sooner, but would also mean that the outward center of gravity would be brought closer to the center of the wheel, so it seems to me that the positive advantage is canceled by the negative one.
• What effect will the coupling of two sections of pipe have on performance? As water passes through that joinder point, it will be forced through a section of smaller inside diameter for a short distance, and then continue through the remaining section of normal inside diameter. This will happen as water flows outward, as well as when it flows inwards. We all know that water passing from a larger section of pipe to a smaller section results in an increase of pressure. In this instance, it will also result in a certain amount of restriction, because there is no force behind the flow other than gravity. Will the pressure increase help to get some water sent to the outer or inner end sooner than it would if a straight, single piece pipe was used? By the same token, would a straight one piece pipe allow full water traversal to occur sooner?
• At any particular point of rotation, there will be four pipes with water at their outer ends providing the rotational force, while there will be 8 pipes that will NOT be aiding rotation. Two of those 8 pipes, for instance, might have their centers of gravity aligned with the holding fixture's uprights, thus not aiding or impairing rotation. At least 6 pipes, though, will at all times be working to impair rotation, as they will have centers of gravity at their inward ends, and be located at the other side of a vertical centerline from the 4 pipes supplying rotational force. It would seem that since the 4 pipes providing rotational force have their centers of gravity much farther outward from the center of the wheel assembly than the 6 to 8 pipes opposing rotation, we should (theoretically at least) be able to overcome such opposition.
• As I stated earlier in this post, the amount of water used for this 50% of capacity water fill will be 110 ml in each pipe. A water volume of 110 ml weighs 110 grams, or 3.88 ounces, or 0.2425 pound. Thus, knowing the weight of the water fill, and knowing where the centers of gravity are located, the applied force (torque) can be determined for any pipe, at any point in its rotation by determining how distant its center of gravity is to the wheel's vertical centerline, and summing the values for pipes at one side of that centerline, as compared to the sum of values for the opposite side will (theoretically, at least) determine whether or not an overbalance condition exists at that point of rotation, as well as how much overbalance torque is available. Will that amount be enough to turn the wheel? Will the sums of values, changing during rotation, remain heavier at all times, or nearly at all times, in the 4 pipes providing rotational force than in the pipes impairing rotation? Only time, measurements, calculations, and testing will tell.

EDIT - April 14, 2020: The 110 ml water fill stated in this post was found to be incorrect later on when I removed water from tubes. I had written down several numbers in a notepad, and evidently used the incorrect one. The actual initial fill was 106 ml, after which I added 2 ml to bring the total to 108 ml.

[rickoff]

Hi folks,

I have a preliminary video ready for you to view now. It's about 11 and 1/2 minutes in length. It was just something I made off-the-cuff, so to speak, while experimenting, so it isn't anything very exciting to watch. I started by demonstrating rotational action with just the water in tube "I" for motive force, and then added an equal amount of water to tube "C," which is 180 degrees opposite, and parallel to tube "I." After that, I was preparing to add water to the remaining 4 clear tubes, but after filling just two of them I noticed that there was an unwanted drip onto the platform. I knew that I hadn't spilled any water, so looked directly up from where the droplet had fallen and could see that it had come from one of the flat end caps. Evidently I hadn't applied enough silicone sealant to that cap. I stopped filling, and drained both of the newly filled tubes, knowing that I'd have no choice but to dismount the wheel assembly from the holding fixture, separate the wheel with the clear tubes from the shaft, and repair as necessary. While the first 2 tubes that I filled show no signs of leakage, and I could probably just reseal all the remaining ones with silicone, I've decided that I might as well go ahead and use PVC cement to ensure that the possibility of any leakage is eliminated. I'll be working on that tomorrow, and will probably have the second video ready to view by April 1st. For that one, I'll have all of the clear pipes filled. I don't think that will be enough to achieve continued rotation, but i think it should prove interesting. After that, I'll fill all the white tubes and we'll see what happens. Here's the link to the video: https://youtu.be/w1aZzRD5NSU

[rickoff]

Back again. I went ahead with removing the wheel assembly from the stand, separating the 2 wheels, removing all the flat end caps, cleaning out the silicone sealer, replacing the caps using PVC cement, and reinstalling the wheel assembly on the stand. After doing that, I filled the 4 remaining clear tubes and set one of the tubes to a horizontal position at top of the wheel. As water flowed outward in that tube, I let go to see if there would be any rotational force applied. The wheel assembly did start to move for about 30 degrees and then stopped. I didn't think that would look very interesting, so didn't make any video. Instead, I filled all the white PVC tubes with 110 ml of water to match the fills in the clear tubes. I hoped that this would result in at least a somewhat longer sweep, but it didn't. That really didn't surprise me, as I never expected that a 50% water fill would somehow end up being the optimal amount. It was just an initial starting point, and one that would allow for easily calculated fill reductions at other percentages for future experimentation. Tomorrow I will start doing some critical thinking on the issue, and determine what I want to try next. I definitely will do some measurements first to allow calculating the applied torque for each side of the wheel from the vertical centerline. When I used the concept drawing in post #2 to figure applied torque, I was basing that upon 3 inch tubes with a water fill for each tube having a 5 pound weight. Those calculations showed there would be about 6 foot-pounds of overbalance torque at the left side of the centerline. Since the initial fill weight for the prototype was 0.2425 pound (less than a quarter pound), the prototype would have less than 5% (0.2425/5=0.0485 or 4.85%) of the water weight figured in the concept drawing, even though the comparative fill level is higher in the prototype. In looking at the concept drawing, it is quite apparent that the level of a 5 pound fill was considerably less than what a 50% fill would have been, which is what I used for an initial fill in the prototype, and that could prove to be an advantage, as it moves the center of gravity further outward in tubes providing motive force, and further inward in tubes opposing that force. Of course that would also result in less water weight at both the inward and outward end of the tubes, and that could prove to be a critical factor in a small build like this one, since a certain amount of weight is necessary just to begin turning the wheel due to friction. The bronze bushings obviously don't provide the nearly frictionless rotation that some expensive bearings would allow. I also need to figure how much of a hindrance to water flow is being caused by the reduction of cross sectional area at the pipe fittings joining the inner and outer sections of the tubes. In a pipe not having those connections, and the resultant bore reduction, I do believe that full water transfer inward and outward would occur sooner. As mentioned earlier, I did worry that this would have a negative effect that might impair rotation. One thing is certain, which is that there are many factors to consider. If I can fairly accurately compute applied torque at each side of the centerline for each of 6 intervals of say 5 degrees rotation for a total of 30 degrees, which is the distance between successive pipes, this would give a pretty good idea of what is happening at any particular moment of rotation. I would also then be able to determine what effect reducing the amount of water fill from the current 50% to say 45%, 40%, 35%, 30%, or less, would have on the balance condition. If reducing the fill amount does prove to be advantageous, which I suspect may be the case, then I'll try a series of reduction experiments.

Edit - April 14, 2020: As an edit to the last previous post states, the 110 ml amount of water fill was found to be incorrect. The actual final amount placed in the clear pipes was 108 ml. Thus, when I placed 110 ml in each of the white PVC tubes, this factor alone would have caused an off balance condition which may have proved detrimental.

[rickoff]

After taking some time to think about what I wanted to do next, I decided to drain the water from all the tubes. When I did that, I started by draining one of the clear tubes and found that rather than the 110 ml amount which I had stated earlier, there was actually 108 ml. As I have explained in edits at the bottom of my last 2 previous posts, the 110 ml stated was taken from figures I had jotted down in a notepad. The initial fill amount had actually been 106 ml, to which I added another 2 ml, bringing the actual half-point fill amount to 108 ml. I checked a couple more of the clear tubes, and they too had 108 ml fill levels. Thus, by filling all the white PVC tubes with 110 ml, I was actually throwing the fill levels off balance. This 2 ml difference, per tube, most likely would have been enough of a factor to cause rotation failure in a build this size. Rather than fill all tubes with a corrected amount, and try again, I decided that I wanted to compare computations for centers of gravity for the 50% fill level with a lesser fill level to see what the actual difference works out to. As I mentioned in my last previous post, a lesser fill amount would place the centers of gravity further outward in tubes moving downward from the top horizontal position, and further inward on tubes moving upward from the bottom horizontal position, and this of course would seem to prove beneficial. The only drawback to doing so, of course, would be the water fill weight reduction factor, which would provide less torque for rotation. In a build with high performance bearings this wouldn't be much of a problem, but with the bronze bushings that I am using I have to take into account the amount of drag that must be overcome just to initiate rotation. I could test that amount of drag with the tubes unfilled and the wheel assembly being in relatively perfect balance, but with water fills in place a drag test would not be reliable. I can only be certain that adding the water fills to all tubes would cause a greater drag effect at the bearings than would be present with unfilled tubes, and of course this becomes a performance limiting factor.

In deciding what water fill amount I would like to use as a comparison to the 50% fill amount, I decided to go with a 25% fill amount (quarter fill) for the comparative calculations. Thus, I marked off the 25% centers of gravity in the clear tubes by adding a black dot at each of the inward and outward ends, at the distances shown in the first photo below, which is simply a modification of an earlier tube fill photo. Note how the inward center of gravity, moved 76.125 mm further inward, now falls upon the 90 degree elbow, while the outward center of gravity is moved 76.125 mm further outward.



Next, I decided that I wanted to be able to capture images of the wheel centered on the holding fixture upright so that I could draw a line upwards through the center of the upright to divide the wheel as closely as possible to it's vertical centerline, for purposes of taking measurements to the centers of gravity at differing angles of rotation. I needed to draw some marks upon the wheel circumference, and attach some kind of fixed indicator to one of the uprights so that it would overhang the angle markings. To do this, I positioned tube I at the top horizontal position and made the first mark at the top of the wheel's circumference to establish that as the zero degree marking. Since I had already divided the face of the wheel into a 30 degree radian section, I extended those radian line to the circumference and took a millimeter measurement between those marks, on the circumference, which happened to be precisely 160 mm. I wanted to mark off lines at each 3 degree angle of rotation from my zero mark, until reaching 30 degrees of rotation. Since this would require drawing ten lines to the right of my zero marking, each line would be 16 mm further along the circumference. For an overhanging indicator, I simply wedged a steel coat hanger into the upright of the white PVC side of the apparatus, bent the wire down over the top of the wheel, and cut it to an appropriate overhang, as shown in the next below photo, which also shows the degree markings.



Then all I had to do was bend the overhanging wire enough to overhang the zero mark when tube "I" was in its exact horizontal position to obtain my first photo of the angle series. The reason why I only marked off 30 degrees of rotation is because at the 30 degree point tube "J" would then be in the top horizontal position, and thus I could use this 30 degree span to determine all desired measurements and calculations for each 3 degrees of rotation, and these figures would be the same no matter which clear tube was being referenced. The next below photo shows The wheel set at the zero reference point, and with the vertical centerline superimposed at the center of the upright and extended upwards to the top of the photo.



You will notice that the overhanging indicator wire is seen slightly to the right of the centerline at top of the wheel. Evidently, when I scribed that first mark, I may not have precisely located the top dead center of the wheel. That really doesn't matter, though, as long as the angle markings are correct and the wire indicator is positioned directly above the zero marking when tube "I" is at top horizontal. I took the series of photos from a distance of 8 feet, with the camera lens at the same height as the wheel shaft so as to minimize any distortion effects caused by the lens, and simply zoomed in just enough to frame the photos as I wanted them to appear. I shot these at maximum resolution, so the actual photos are larger than what is shown. This will make it easier to take precise measurements from the wheel's vertical centerline to the center of gravity markings on the tubes. Since the white PVC tubes cannot be fully seen, I'll need to take another series of photos starting at the point where tube "J" is in the top horizontal position, which is the last position in this 30 degree series. That way, clear tube "K" will be in the same starting position as tube "J" is at the zero point shown in the above photo.

I reduced all eleven of the first 30 degree series to smaller photos so as to create the photo collage shown below. My next step will be to take the measurements which will compare applied torque at either side of the wheel's upright centerline for each 3 degrees of rotation. Then I'll proceed to the second 30 degree series in order to have a complete overview of applied torque for any tube at any 3 degree point of rotation. This will take me some time to accomplish, so the results will probably not be available for both series until several days from now, but I will post whatever I do have as soon as it is ready.


While this method of determining the measurements and figuring the calculations will not be absolutely precise, I think it offers a fairly reasonable analysis, at least as a starting point in determining what is going on, and helps to determine the differences in overbalance (if any) between a 50% fill and a 25% fill. The centers of gravity do shift within the tubes as the wheel rotates, and I'll have to take a look at that factor to determine how much of a shift there is at any 3 degree position of rotation so that I can come up with a percentage figure that can be used to modify my table of results. At that point of analysis, we should have a very accurate estimate of balance conditions.

[rickoff]

Hi folks, I know that it has been a long time since I posted updates here, and I'm sorry if I have disappointed anyone who had been following the thread. So as not to repeat a lengthy explanation for my whereabouts, and what I have been doing, please see my current posts from the past couple of days at this link. I will definitely be pursuing a continuation of this thread, and will show in diagrams and video how my Overbalance Water Wheel can be made to continuously rotate using only the force of gravity. I often wondered if anyone had built my device and followed through on that with experiments. I do realize, however, that most people are not going to invest the time and resources into building a device until someone proves it will work, so my aim is to do exactly that. I will also be continuing my Pipe Dream Magnetic Motor Generator project, and for reasons explained in my above link, I will first devote available time to completing that project. I will, however, also be popping in here as time allows to explain the modifications that I see as being necessary to achieve rotation.

Best regards to all, - Rick