Final Specifications

Comparison of Initial and Final Critical Dimensions
Location	Average Final Value [in]	Initial Value [in]	Difference (Final-Initial)
Max Diameter of UFO	2.495	2.461	0.034
OD of Snap Fit	2.004	1.990	0.014
ID of Snap Fit	2.031	2.000	0.031
Alien Height	0.688	0.68*	.008
Alien ID Snap Fit	0.256	0.260	-0.004
Base UFO OD Snap Fit	0.248	0.250	-0.002
Height of Dome	0.694	0.688	0.006
Width of Dome	1.623	1.580	0.043
String Gap	0.09	0.100	-0.01
Total Width of Yoyo Half	1.161	1.138	0.023
Total Mass	65.1 [g]	64.68 [g]	.44 [g]
Maximum Rotation Speed	~2500 rpm (measured)	1271.4 rpm (estimated)	1228.6 rpm
Tolerances	Final Tolerances (all tolerances not listed here are +/-.0075) Note: Dome width tolerance:+/-.025		Initial Tolerances (all tolerances not listed here are +/-.005)
Location	(+) Value [in]	(-) Value [in]	(+) Value [in]	(-) Value [in]
OD of Snap Fit	0.005	0.005	0.005	0.0
ID of Snap Fit	0.005	0.005	0.0	0.005
Alien ID Snap Fit	0.005	0.005	0.0	0.005
Base UFO OD Snap Fit	0.005	0.005	0.005	0.0

Each dimension was adjusted slightly through the design and manufacturing process. The measurement that were not measured on all 110 parts were measured on a random selection of 10 parts. The numbers presented here are averages of the collected measurements. Below is an explanation for the change in each dimension.

Max Diameter of UFO

Upon inspecting the parts, it is clear that the top part tends to jut out slightly more on the bottom part. There are two likely causes for this. The first with a lesser contribution, is the minimal shrinkage of the outer diameter of the top part, when slight dishing occurs on the sloped side of the ring. In addition to the slightly increased mold size to accommodate for shrinkage, the top ring may be slightly larger than the bottom ring that is more prone to shrinkage with such a large mass. More importantly however, is the significant interference of the snap fit, an average  close to .03 inches. This could make the thin top ring stretch and flare outwards, accounting for the increased total diameter.

OD and ID of Snap Fit

In initial testing runs, the snap fit was far too loose. It could be pried apart by hand and did not stay assembled when dropped. Initial steps taken were to increase the interference. This helped but it still did not remain assembled when dropped. We decided to add an undercut to the snap fits. We then discovered that the mold for the top ring had what was most likely a draft angle on the snap fit. When this was removed in addition to the undercut and the increased interference, we had a significant interference that required an arbor press to assemble. This was the process that led to the parameter change in this dimensions.

Alien height

The alien height was close to the initial parameter. This minor variation is likely the cause of minimal shrinkage when the shrinkage accounted for by the mold was 2%.

Alien ID Snap Fit

This dimension stayed quite close to specifications. The minimal change here is likely due to the choice of Injection Molding parameters.

Base UFO OD Snap Fit

This dimension stayed quite close to specifications. The minimal change here is likely due to the choice of injection molding parameters. Despite the large mass of the UFO and the long cooling time, the OD of the alien snap fit is in a geometry that allows little variation, making shrinkage variation minimal.

Height of Dome

This dimension stayed quite close to specifications. The slight variation may be due to the location where the dome was cut from the thermoform, since a drop gauge was used to measure the height.

Width of Dome

This dimension seems to be extremely far from the specification. However, this is likely a function of measurement method rather than the change of the specification. On the curved surface of the thermoform, it was hard to determine a location that should be measured for the width of the dome. The measurement method we settled on provided relatively consistent results, but perhaps not at the exact location of the width of the thermoform. Other measurements methods were considered. If the dome were die punched, the measurement would be a function of both the variability of the die cut and the dome. In addition, the calipers would squish the pliable thermoform, also causing inconsistency in measurements.

String Gap

This dimension is relatively close to the original dimension. A significant change that took place was that during assembly, we discovered that with a ⅝ inch set screw, the aliens were pushed out of their snap fit. Therefore, a ½ inch set screw was used. This shorter set screw may have allowed for extra tightening, causing the .01 inch change in dimension.

Total Width of Yoyo Half

The slight difference in dimension here is due in part to the dishing on string gap side of the base part.  When using the drop gauge, it was apparent that the dishing on that surface was contributing to the total height.

Total Mass

The total mass of the yoyo about half a gram greater. This is likely due to the addition of the set screw, spacer, and string.

Maximum Rotation Speed

The initial calculation assumed that 30% of the energy was lost to friction, and that the initial speed was 1 m/s. After testing the yo-yo, it is clear the the estimated 1 m/s is too low. Considering the yo-yo can easily sleep for about a minute, a 30% energy loss to friction is probably also quite high. Both of these poor approximation could account for the large difference in rpm.

Tolerances

Most tolerances were close to the initial specifications. The tolerances of these parts were determined by calculating 3σ. After performing the calculations, it is clear that the snap fits did not have one-sided tolerances. However for the top and bottom snap fit, the tolerances were almost irrelevant, considering how large the interference was. For the base and the alien, a very large interference was hard to assemble because the geometry of both the base and the alien do not allow for stretching or squishing. In addition, if the parts were only nominally the same size, the fit was not snug enough. In this scenario, the plethora of aliens available to us made it easy to perform what we dubbed “the apple method” which was switch out the alien if the press fit was too challenging, or the alien felt too loose. In addition, for several of the parts with only 10 sample measurements, the tolerance was higher than .005. With a larger data set, the tolerance is expected to drop. In addition, for reasons mentioned in the width of dome section, the dome width had a higher tolerance.