Making A Yo-Yo ???

[randyc]

(The following project may seem trivial but I thought that it might be of interest, especially to those of us with old machinery - no DROs, just the dials on the cranks and maybe a travel indicator on the lathe.  It was initially posted on another forum a few years ago.)

My nephew, Kurt, when he was about 12 years old and interested in "yo-yo’s", showed me a rough sketch, including a few dimensions and the desired weight of his ideal device. This isn't about old wooden "Duncans" or newer plastic devices illuminated with flashing LEDs. Kurt was envisioning a modern version made of metal, including features like a ball-bearing axis and low-friction side pads. Apparently these types of yo-yo's are necessary for advanced tricks and are readily available although costly.

Kurt did some internet research and purchased a few parts for his prospective design. One part was a ball-bearing with the outer race ground in a shallow concave arc, to guide the tether string toward the center of gravity. He also obtained a pair of low friction pads (shaped like washers, made of "Teflon" or a similar low-friction material) that fit into shallow grooves on either side of the hub interior. The purpose, he told me, is to reduce friction of the tether string against the side surfaces of the yo-yo.

We discussed means to increase the spin time of the yo-yo. The yo-yo is a simple flywheel so we want the greatest amount of mass distributed as far away from the spin axis as possible to maintain momentum. Kurt's original concept evolved from a pair of aluminum discs to a four-piece design consisting of thin aluminum hubs attached to rims with more mass. The choice of rim materials could include most of the ferrous or copper-based alloys, since these are about three times the density of the aluminum selected for the hub material.

This is a cross-section of the design that we conceived (chamfer/radii not shown). The yo-yo is 2.300 inches diameter x 1.625 from rim-to-rim.


  yoyo1.jpg (Size: 6.88 KB / Downloads: 76)

There isn't much in the design that is dimensionally critical other than the ball-bearing interface, the low-friction pad grooves and the fit between the hubs and the rims. But it's important that the yo-yo be well-balanced and symmetrical, according to my nephew. I arbitrarily decided that non-critical tolerances should be +/- .005 inches which is not normally a difficult requirement.

This design incorporates multiple odd-angle angular transitions and accounting for backlash (measured and written down) is necessary. It was helpful for us to consider the sequence of events. This can be especially important when dealing with structures that aren't orthogonal or rigid and the thin hub sections of this yo-yo project aren't very rigid or orthogonal.

Diameters of the body parts and rims are too large to fit in a normal collet; the workpiece needs to be held in a chuck, a pot-collet, a mandrel, etc and it's desirable to perform as many operations as possible without flipping or removing the part and losing concentricity.  "Wobble" is not good - the parts need to be concentric for maximum conservation of energy and "smoothness" of spin, right?

I planned a sequence of operations that seemed sensible, progressing through the various machining processes. The yo-yo hubs are flimsy and might deflect somewhat under turning or boring pressures. This requires sharp HSS tooling – carbide isn’t a good choice.

During this project, I found that high spindle speeds caused the O.D. of the aluminum hubs to centrifugally expand enough to affect the diameter dimensions. For example: without moving the cutting tool location and changing the spindle RPM from 400 to 950 produced a measured difference in the turned diameter of almost .001.

This wasn't due to a "spring" cut, it was definitely caused by centrifugal expansion of the perimeter of the workpiece due to the flexible cross section.

Even for flimsy structures, this wouldn't normally be a problem but it CAN be for precise work. A lesson for me was that we frequently change spindle speeds when transitioning from roughing to finish cuts - sometimes increasing RPM for the finish cut. This could affect accuracy in irregularly-shaped workpieces with thin cross-sections.

The finished diameter of the aluminum hub was fairly critical because it needed to accommodate a shrink fit with the rim. Keeping this in mind, had I spun up the spindle for the final cut (after roughing at a low RPM), the expansion of the O.D. may have resulted in the finish diameter being SMALLER than the design value.

The parts may have been sufficiently out of tolerance so that the rims might not have shrunk enough to fit the hubs properly if the rims had already been produced. In this case, the slight dimensional change was not enough to affect function.

Sorry for the diversion, I just thought that this was an interesting fact to keep in mind.

Both of the hubs were roughed turned to shape in a 3-jaw chuck. The bearing and friction pad features were finished to final size. A 1/4 square HSS trepanning tool was ground to produce the friction pad grooves in one pass, the dimensions of which were shown in the initial sketch above. The next step was obtaining a preliminary idea of their weight.

  yoyo6.jpg (Size: 15.95 KB / Downloads: 74)

The balance in the photo is claimed to have the capability of resolving tenths of a grain. This was an interesting exercise: comparing the calculated weight of an object with a complex shape to the actual weight. (Recall that Kurt wanted a specific weight for the yo-yo.)

So far, so good.  Although the hubs were overweight (they have only been rough turned) they were within hundredths of an ounce of being equal.


  yoyo11.jpg (Size: 15.84 KB / Downloads: 74)

Continuing, rather than drilling a pilot hole for the #8-32 tapped hole in the center of the yo-yo halves, the hole was bored for maximum concentricity with the other features. All this was finicky work but the project was a natural for an Emco "Compact Eight" lathe that I purchased some thirty years ago.

Most of my projects prior to retirement involved workpieces that fit easily within the envelope of the Emco, 8 x 18 inches. (I have a larger lathe but it is the same age as me; and functions about as well as me, ha-ha-ha.)

Looking through my small stock of materials, I didn't find anything suitable for yo-yo rims except steel but I thought the yo-yo would look cooler with rims made from a copper alloy. Other than junk yards, there are no metal suppliers in this part of northern California - suppliers are about four hours drive south or east.

After noting the outrageous prices that hobby-oriented internet metals suppliers charge for their treasure (not to mention shipping costs) I contacted a friend nearby, who operates a one-man job shop, and described my problem.

He has always been generous with parts and materials, the use of his machinery, his time and his suggestions based on years of knowledge and experience as a millwright (now retired). He rang my doorbell a few days later, with 25 pounds of various materials he considered appropriate for the yo-yo project.

For density, I selected a naval bronze blank. I drilled and then bored (to finish I.D. dimension) in a 3-jaw chuck, then parted off two rims. I was surprised at how tricky cutting this material turned out to be. I substituted a couple of cutting tools, varied spindle RPM, feeding the carriage by hand until an acceptable combination was determined by feel.

Note that copper alloys usually like cutting tools with zero rake or even slightly negative rake.  Unusual coolants are said to be effective - I've read of using everything between milk and bacon grease !

  yoyo3.jpg (Size: 20.35 KB / Downloads: 74)

This is a situation where large Morse taper drills are very useful, the one in the photo is 1 inch diameter. Unfortunately one requires horsepower to take full advantage of big drills - I had to feed lightly with the small Emco to prevent slipping the drive belts. A boring bar in a special holder is visible to the left of the drill in preparation for finishing the ID of the work.

The bored blanks were faced slightly over the desired width and the O.D. was left oversize. The bore dimension had been previously determined for a shrink fit with the hub O.D. based on the thermal expansion characteristics of the two different materials and the desired overall dimensions. After boring to the finish dimension, the rims were parted from the workpiece and set aside:


  yoyo9.jpg (Size: 24.59 KB / Downloads: 73)

Balance, concentricity and symmetry requirements suggested that the yo-yo be completed on a mandrel after the parts were semi-finished. The mandrel is in two pieces, a removable one that threads onto a fixed stub held in 3-jaw chuck (the threaded part of the mandrel accommodates the tapped holes in the yo-yo hubs).

The removable part of the mandrel would be supported by the tailstock center for greater precision when holding the assembled yo-yo. I wanted to perform the final finishing operations with the yo-yo completely assembled so that all features would be concentric. The hubs were to be secured between the mandrel parts, separated by the ball bearing (used only for a spacer) and steadied by the tailstock.

Making the mandrel was a routine task but an important one and a few items are worth noting. I know that it's hard to take this project seriously - a YO-YO, for goodness sake ! The interests of teen-age boys (except in some areas) are unpredictable but definitely compelling. The project was also intended to be educational for Kurt, of course.

The mandrel was made from a scrap of CRS, cut to length on the ubiquitous small Chinese horizontal bandsaw. During sawing, when I need to attend to other tasks, it's not practical to apply cutting lube to the bandsaw blade.

Cobbling up a drip-coolant for the saw has been low on the priority list so I devised a cheap trick to lubricate the blade while otherwise occupying myself. I save old candle stubs (electrical power is sometimes intermittent during the winter so there are usually a few stubs laying around) and place the stub next to the workpiece being sawed. The saw blade picks up enough candle wax to keep the blade happy.


  yoyo2.jpg (Size: 21.65 KB / Downloads: 74)

Here's a roughing pass on the fixed mandrel illustrating that this little Austrian-made machine, because of it’s hefty construction and excellent ratio of swing to length, can take heavy cuts that many larger lathes cannot. In this photo, for example .250 depth of cut at .007 IPR on the steel workpiece - over 1-1/2 cubic inches per minute of material removal.

A sharp HSS tool, a rigid tool holder and some sulphur-based pipe threading oil has always provided good results on steel for me (actually, as you can see, it cuts pretty good without coolant but it's harder on the cutting tool). As a hobbyist, I rarely use carbide cutting tools, usually only for very specific tasks.

Going from one inch diameter to .162 diameter required only two passes of a sharp tool with this little sturdy lathe. However, I should note that, being retired, cutting efficiency isn't normally of much importance to me.


  yoyo7.jpg (Size: 18.31 KB / Downloads: 74)

The second pass produced the finish diameter and was followed by a thread chasing head to produce 8-32UNC-2A threads. (Although I just implied that machining time isn't that important, I do NOT have the patience for single-pointing small diameter threads when I can chase them in fifteen seconds.)


  yoyo12.jpg (Size: 20.42 KB / Downloads: 73)

The faces of both mandrel parts starting about .050 from the rim were slightly rebated (can be seen below) so that the yo-yo hubs could be gripped at the perimeter of the faces, minimizing flexing during subsequent operations.  This is barely visible in the photo of the mandrel above.

One might wonder why this would be necessary, and frequently it isn’t, but the several different facing operations and the interfaces (between hubs, bearing races and mandrel surfaces) suggest that perpendicularity and concentricity MIGHT otherwise be a problem. Making peripheral contact - as opposed to an uncertain contact across a wide surface - is likely to produce the most satisfactory results. (Besides, it takes only a few seconds.)


  yoyo17.jpg (Size: 14.61 KB / Downloads: 73)

This is the completed mandrel – doesn’t look like much but is quite functional and quite necessary for this application:


  yoyo8.jpg (Size: 18.15 KB / Downloads: 72)

From this point on, the mandrel will not be removed from the chuck – this was the main reason for planning the sequence of operations in advance. Critical mandrel dimensions are now concentric with the lathe spindle. The yo-yo parts, when carefully mounted on the mandrel, are expected to also be concentric when they are turned to finish dimensions.

The aluminum hubs were mounted to the mandrel and finish-turned. At this point some may wonder why these thin parts turned from solid bar were not singing like a tuning fork and producing distinct chatter marks.

In fact, the parts WERE singing like a tuning fork whenever the boring tool approached areas of the workpiece that stimulated natural resonant frequencies

There are techniques for minimizing chatter marks resulting from this problem. One method is to pack internal cavities of a workpiece with putty or modeling clay - this works well for thin-section workpieces if the clay can be packed in such a manner as to prevent it being flung out.

In this case, dipping my flux brush in cutting oil and holding the brush firmly against the surface opposite the cutting tool was adequate to dampen the resonance and produce an acceptable finish.

The next photo shows the OD’s being turned for a shrink fit with the rims while mounted on the mandrel. Obviously this would have been MUCH better performed with chip-breaker geometry, eliminating the messy, stringy pile-up caused by the sharp tool.


  yoyo13.jpg (Size: 18.27 KB / Downloads: 66)

A shrink fit is a good way to achieve a near-permanent joint between several parts and for this yo-yo, it’s not a complicated operation. Aluminum has a linear temperature coefficient of about 23 ppm/degree C/inch while bronze is about 19 ppm/degree C/inch. (1 ppm = 1 part per million = .000001).

Wanting the simplest method of shrink-fitting the parts, I placed the aluminum hubs in the refrigerator freezer and the bronze rims in the kitchen oven. The aluminum hub will shrink by about .001 inches. The bronze rim will expand by about .009 inches so when the two parts are at their temperature extremes, the interface clearance will be about +.010 and should easily be joined.

After 30 minutes temperature soak, the parts were removed; they were easily fitted together and totally secure after reaching room temperature. The completed halves were returned to the mandrel, and were rough-turned and faced so that the difference in weight between the two could be evaluated again.

After weighing, the halves were reinstalled on the mandrel. The final lathe operation was completed by facing both sides of the rims flush to the hubs and turning the rim OD’s to finish dimension. Removing the halves from the mandrel and re-weighing, the two parts were within 5 grains of one another (0.01 ounce). The O.D. of the yo-yo was intentionally left about .125 oversize so that Kurt could test drive it before further modification.


  yoyo14.jpg (Size: 18.95 KB / Downloads: 74)

After trying out the yo-yo, Kurt liked the "feel", the diameter and width but felt that removing a tiny bit more weight would be helpful. I should mention at this point that the boy is GOOD at this stuff, he performs tricks that are W-A-Y more complex than those I recall from my boyhood and competes at state level (who knew that there were yo-yo competitions ?).

At this point, the yo-yo is very difficult to fixture since the walls are thin and flex easily. I made a collet-like split clamping device from a slice of 4 x 4 timber, bored in the vertical mill to closely fit the O.D. of the yo-yo halves and then split diagonally, part-way through, on the table saw.  (Returning the block to the original position on the mill re-centers the mill spindle with the collet and therefore centers the yo-yo when it is gripped.)


  yoyo5.jpg (Size: 14.75 KB / Downloads: 73)

Installing the halves in the split block, clamping the block in the mill vise and applying some side pressure with a "C" clamp held the thin parts so that 7/16 diameter holes could be plunged through the hubs with a 2-flute end mill. I wanted to produce eight holes in each hub (because it would look cool) but Kurt stopped me at four holes - he wanted to try the yo-yo again before plunging the remainder of the holes.


  yoyo18.jpg (Size: 20.86 KB / Downloads: 74)

We reassembled the device and Kurt pronounced it perfect, in terms of size, balance and spin time. Here's the yo-yo prior to de-burring, polishing and coating with shellac - all to be performed by Kurt. The teflon washers and the ball bearing have been installed and this thing rotates for a l-o-n-g time !


  yoyo15.jpg (Size: 19.34 KB / Downloads: 74)

The project was intended to accomplish the following:

• Acquaint my nephew with a few fundamentals of metal turning

• Produce a unique device - not necessarily better than a commercial product but a custom-designed, hand-crafted product that Kurt would value because of his personal contributions to the design and construction

• Introduce a performance advantage by composite construction: lightweight hub + heavier rims = more stored energy hopefully allowing the boy to perform more elaborate "tricks"

(It goes without saying that the project took twice as long as I'd estimated.)

Cheers,
randyc

[haroldmulder]

Very nice project and great write up. I think somebody will be spinning that yo-yo for years to come.