Wednesday, May 13, 2015

Deliverable 5

Deliverable 5 due May 13, 2015


Part 1 - Team Video

Part 2 - Summary of Cost Analysis

In creating a model to compare prototyping and manufacturing, it becomes clear that there is a clear tradeoff between parts produced and cost. For prototyping, a team looks to minimize waste, increase speed, and identify possible improvements and redesigns. In the manufacturing stage, the team looks to reduce cost as much as possible per yo-yo. Manufacturing is more of a profit-driven process, whereas prototyping still remains in the design stage.

With respect to costing the two processes, several factors are important. In our model, we included the sunk cost of the machines, which we valued at 50,000, 30,000, 60,000 and 15,000 for the lathe, mill, injection molding machine and thermoform, respectively. We estimated our mass to be approximately 50 grams, and our volume to be 30 cc, and a cost of .1 cent average per gram of raw material. For the turning and milling of our molds, we predict about 1.5 kWh per mold. For injection molding and thermoforming, we predict about .2 kWH per part, or 100 kWs. For time/labor, we estimate that each yo-yo to completion (once the molds have been made), will take approximately 5 working minutes, where each employee is on average $10 per hour.

All of these factors have been put into a matlab script, and plotted as cost per yo-yo versus number of yoyos produced.




Part 3 - Reflection

One of the characteristic components of our yo-yo is the spokes part that holds the steel balls. To get the effect of an overbalanced wheel we wanted the compartments for the balls to be as large as possible. We found out early on that a big constraint on our design was the locations of ejector pin holes in the injection molding machine. In order to remove each spokes part from the machine the pins and the material need to line up. This not only meant making our compartments smaller, but we also had to increase the thickness of the walls so the pins can slide in the grooves of the mold. In the end, these changes didn't affect the functionality or aesthetics of our design and I think our yo-yo would do well in mass production.

Part 4 - Recommendations

Although manufacturing a yo-yo is a great experience that teaches several different manufacturing processes, students never get the chance to see many of the other methods covered in the course in practice- for example welding, metal casting, or deposition methods.  This material is covered in problem sets and lectures, but not hands on.  Though it is significantly more time intensive for each student to learn to weld (and not practical to add to the course), live examples are much more engaging and informative for students than problem sets.  The field trips to manufacturing companies are helpful, but if possible, it would be better to have these demonstrations (welding, laser sintering, etc.) on campus so that more students will have the chance to attend.  These demonstrations could be in the place of, or in addition to the problem sets and exams, dependent on whether there is any predicted additional time commitment from students.

In summary, 2.008 is a fun and informative course, and the most valuable skills learned from the class are how to navigate many different manufacturing processes.  However, it is one thing to know the manufacturing theory and another to see a process in action.  It is more valuable to see how people and machines work to create products than to do problem sets on beam bending that have already been covered in other mechanical engineering courses at MIT.

Thanks for a great semester!


Thursday, May 7, 2015

Deliverable 4 for Friday 5/8/2015

Blog 4

1.1 Final Yo-Yo

Below is a photo of our final yo-yo. The inner spokes and bearings were inspired by an perpetual motion wheel while the body was designed to be sleek and minimalist. Some key features include a clear cover that reveals the ball bearing moving inside, spokes held in place by locating pegs, and insert molded nuts in the body.  We were happy with how the functionality of the retaining ring press fit and the locating pegs on the spokes turned out even after cooling. In future iterations, we could choose to make the body a different color, because the we had trouble completely purging the previous color and our white has a bit of an orangish tinge. It also shows dirt from manufacturing. We also could have used larger ball bearings, that would allow the yo-yo to function more as an infinite motion wheel.

Final assembled yo-yo, top view

Final assembled yo-yo, side view
1.2 Final Components

1.2.1 Retention Ring

The retention ring was made to hold the thermoformed cover in the body. It had a small fillet on the top surface both because of tooling specifications and aesthetic reasons. It took us two different iterations of machining to fine-tune the snap fit, but overall we were pleased with the final part fit and visible finish. Some minor improvements could be made to the surface finish on the sides of ring, which are hidden in the final assembly.

Retention ring

1.2.1 Body
The body was designed to be low profile and have multiple mating parts. The inner diameter of the wall press fits with the retention ring. The locating pegs and outer diameter of the hub help hold the spokes in place. We were pleased with how the locating holes aligned with the final injection molded spokes consistently, and that there was no detectable warping due to part thickness. In future iterations we would potentially make the outer wall thinner for aesthetic reasons and choose a different color that hides dirt better. 
Body

1.2.1 Thermoformed Cover
The thermoformed cover was chosen to be clear so that the motion of the ball bearings inside would be visible. We were happy with the tolerances of how the outer diameter of the ring interacted with the retention ring to make an effective press fit. In future iterations, we would potentially make some design changes to make it easier to align on center every time. 

Cover
1.2.1 Spokes
The spokes were designed to have cavities for the ball bearings to roll about freely in, and to mate with the body with both locating pegs to prevent rotation as well as a press fit in the center to align it on center every time. We were pleased with the accuracy of the dimensions of this part. In future iterations, we could modify the design to prevent warpage on the thicker areas where the spokes meet in the center. 

Spokes

1.2.1 Ball bearings
We chose ball bearings based mainly on appearance, we wanted them to look cool and in proportion to the rest of the yo-yo. We were happy with how freely they rotate in the current design, and in future iterations could choose some with slightly more mass in order to make the infinite motion wheel more functional. 

Ball bearings 

1.2.1 Shaft Collar 
The key feature of the shaft collar is that is rotates on the bearing but is a close enough press fit that the bearing will not shift dramatically and dislocate the string. We were happy with the final fit we achieved, but we could improve this element by speeding up the assembly time; we had chosen shaft collars that were slightly smaller than necessary so we had to first buckle and then crimp every one to get it to press fit with the bearing. In the future, we could choose a shaft collar with a slightly larger diameter. 

Shaft collar + bearing assembly. 
2 Table of Specifications

2.1 Table



2.2 Discussion

As can be seen from the table above, there was minimal error between the specified feature dimension, and the final dimension that was measured on the part. With regards to all of the injection moulded parts, it appeared that the expected shrinkage was greater than the shrinkage that eventually occurred, resulting in slightly oversize parts. Particularly with respect to the body and spoke interface, which was about 1/2 inch in diameter, it appears that shrinkage was almost entirely negligible. Numbers were closer on the retention ring, which was of a larger initial diameter, and therefore more susceptible to shrink. Due to a lack samples to measure and relatively loose design specifications, there was no attempt to guess what dimension the cover would shrink to. Consequently, the cover ended up being a fair degree smaller than was expected.

3 Retention Ring: In Depth

3.1 Summary of Findings
Since blog 3, the retention ring has changed dimensions to be thinner such that our press fit was tighter.  The ejector pins now touched the part on an outer ring, which during optimization sometimes caused the retention ring to fail to eject properly.  The retention ring experienced a change in cooling time from 9 seconds to 15 seconds for parts 61 - 70, which resulted in no detectible shift in either the part data or the Shewart Xbar chart to indicate that any change had occurred.

Retention Ring Run Chart 

Retention Ring Histogram 

 Retention Ring Shewart x-bar chart

The UCL and LCL were found to be 2.228 inches and 2.180 inches respectively, and the mean was found to be 2.204 inches.  


Retention Ring Process Capability

3.2 Link to PDF


Friday, April 24, 2015

Deliverable 3 for Monday 4/27/2015. Process Optimization

The process optimization of the retention ring mold was highly involved; from the first part, it was clear a major redesign was in order. There was simply no way that the retention ring would fit within the outer diameter of the Yo-yo body and initially it was unclear as to whether this was the result of an error in design or in fabrication.

The design of the original retention ring mold had been dictated by the location of ejector pins. This aspect of the design worked very well, as there was no difficulty removing the finished part from the mold. Unfortunately, this also prevented any alteration of the diameter of the ring without designing an entirely new mold and ejection system. Despite the amount of work and machining this presented, it was decided that it would be better to make an alteration to the ring than it would to the body of the yo-yo, which would have to be expanded dramatically to accept the larger diameter ring.

We changed the mold to a thinner ring by decreasing the outer diameter, and installed a runner system to a larger diameter ejection ring that ran at a radius that transected the sprue (see figure below). The principle was that an outer ring of ejector pins would strike the removal ring, and in the process the retention ring at its center clear of the mold. Though this created visual defects in 3 places rather than one on the ring, it was of little concern as the outer walls of the ring would be hidden anyways.

From the first test, we observed a somewhat unexpected behavior. The thin outer ring intended to be struck by the ejector pins was cooling significantly more rapidly than the retention ring its self. This meant that for short cooling times, the ring had a tendency to stick to the cavity mold and would pull the retention ring most of the way clear, leaving the rest of the work to the sprue ejector pin. As cooling time increased, the retention ring began to hold more resolutely and removal became difficult. The thin runners to the outer ring would shear on impact, leaving the retention ring wedged. Accordingly we kept the cooling time at a relatively short 9 seconds, with a 4 second screw feed delay time.

Testing our finished retention ring, we found that we had underestimated the degree of shrinkage, and the ring dropped easily into the Body of the yo-yo. Luckily, this had been expected in our original design, and the interference provided by the body could be adjusted by a simple turning operation. The outer diameter of the body core mold was progressively turned down until a strong interference fit was achieved.

Pressing the ring was difficult, and required the added force of an arbor press. This was good, as it means that we will likely avoid the risk of spilling ball bearings even if the yo-yo were to be dropped or thrown.

With regards to the body, it was discovered later that an outdated drawing had been used to create the mold. This explained the original difficulty of fitting the ring. This also resulted in a string gap that was larger than expected, and a problem by which the string would slip next to the bearing and bind. Both of these problems were solved by inserting a small brass shim behind the magnet that was holding our nut guide into the mold. This decreased the string gap, and held the bearing in closer proximity to the sides of the yo-yo, preventing the string from slipping.

Ejection of the body with wider walls was initially disastrous due to the degree by which the walls clamped to the core mold. After trying a variety of cooling times ranging from 9-35 seconds, it was determined that slowing the ejector speed to 15%, and ejection force to 45% was sufficient to remove the part without either blowing out or severely deforming the back.

No changes were needed for the Spokes or the Cover.

Monday, April 6, 2015

Deliverable 2 for 3/6/2015: Retention Ring Details etc.

Hello There!

This post will be highlighting the work our team has completed on one part (the retention ring molds) and what we think will be our schedule and time frame moving forward with our YoYo.

Retention Ring Mold

Description:
The mold pieces for the retention ring have a fairly simple design. The core has a depression that extends to the outer radius measurement of the ring from the center of the stock. Also, this is where the gate is located for the combined mold. The cavity also has this radius as it’s outside most edge, but also has a circular section left blank by the lathe. This creates the inner diameter for molding the ring piece. In addition to that, the cavity mold has holes at the radius from the center to further assist in removing the injection-molded part from the mold.

Cavity:


Core:





Shrinkage Calculation:
We measured and found an average shrink of 10% of critical dimensions. Due to a geometric limitation of ejector pin location, rather than scaling the yoyo up to shink to our final design, we've decided to scale our entire design down by 10% to accommodate the shrink of the retention ring.


We have started our injection molding with the retention ring, so as to double check our expected shrink, and proceed from there. These updates will be addressed in the upcoming post.

Manufacturing process:

For the core, after the initial facing of the stock piece, the outer stock of the piece is removed, leaving a centered circular extrusion. This corner is then sharpened and two grooves are cut into the piece: one at the outer edge and one at the inner edge of the ring. Next, the stock is moved to the Mill and the center drill makes initial locator depressions for the ejector pins, which are finally drilled out of the piece.


Step
Operation
Machine
Tool
Justification
1
Facing cut
Lathe
1
Leveling and initial stock reduc-
tion
2
Facing cut
Lathe
1
Removing outer stock
3
Rough cut
Lathe
9
Sharpen inner corner for Reten-
tion Ring
4
Finish cut
Lathe
8
Cut groove for outer edge for
Ring
5
Finish cut
Lathe
7
Cut groove for inner edge for
Ring
6
Drill
Mill
13
Center Drill for location place-
ment
7
Drilling
Mill
17 Drill
Drill Ejector Pins

For the cavity, once the piece has been faced, the center depression is removed and smoothed over. Grooves are then created on the inner and outer edges of what will be the retention ring.
Step
Operation
Machine
Tool
Justification
1
Facing Cut
Lathe
1
Leveling and initial shock reduction
2
Rough Cut
Lathe
10
Cutting to depth of cavity
3
Finish Cut
Lathe
10
Smoothing of cavity body
4
Finish Cut
Lathe
7
Create a groove on the inner side for retention ring
5
Finish Cut
Lathe
8
Create a groove on the outer side for retention ring

 Manufacturing Time Estimate

Process
Time
Units
Machining
x
Hours
Turning
3
Hours
Molding
2.5
Hours
Forming
1.7
Hours
Assembly
1.7
Hours
Total
x
Hours

Time estimates were generated as a mix of Mastercam time estimates and
educated estimation from research. Mastercam produces time estimates based
on its generated toolpaths, and additional multiples of constant time were added
for each tool change necessary during the fabrication of a part. This was less
time for turning operations, as the tools are pre-loaded on the turret. Injec-
tion moulding time was determined through research and discussion with Dave.
It was assumed that injection time and cooling time scaled linearly with the
volume of the part, while ejection remained constant. An additional hour per
injected part was allotted for tuning tuning the process, with respect to injec-
tion pressure, packing time, cooling time, and temperature.

Thermoforming was assumed to be a constant time process of 1 minute, in-
cluding heating and application of vacuum. Assembly was similarly assumed
based o_ of process complexity, though as the operator becomes more skilled,
the times are expected to drop. There is the possibility of some additional ma-
chining to tune parts, but this should be absorbed in some of the extra time.
The estimated time should not change the schedule of the team, though there
is an incentive to start machining as quickly as possible.

Currently, we do not anticipate much reworking to be done for our molds as we already scaled them for the retention ring shrinkage. All time estimates currently stand where we have them before.





Other Molds and Optimization:

Spokes Core and Cavity
  • Key Details
    • Not split down the center but where the initial fillets ended on one side. 
    • Bottom and top face fillets removed to decrease milling time and make mold possible to create

Yoyo Body cavity and core
  • Key details
    • As mentioned before, Body and the parts were scaled down 10% to match up to spokes.
As of right now, the team is working on optimizing the retention ring and will soon move on to optimizing the other pieces. These advances will be discussed in our next post. Until next time!

~Walking The Dog Yoyo Team