Waffle Production Completed!

Breakfast is served!



Screen Shot 2016-05-09 at 2.15.59 PM

Key features of the plate include the snap-fit inner ring and the smoothed outer curvature of the plate edge. We consider both of these features a success because the snap-fit works perfectly with the waffle design, and the outer plate edge is smooth and comfortable to the hand.


Screen Shot 2016-05-09 at 2.16.09 PM

Key features of the waffle include the snap-fit outer ring and the syrup storage cells. The snap-fit worked really well, and the part resembled a waffle. Opportunities for improvement include the color of the waffle (there was considerable variation) and warpage due to shrinkage on the sides of the waffle.


IMG_3244 2

Key features of the butter include rounded edges to resemble real butter, and being small and a cute touch to the assembly. The thermoformed butter ended up with nice round edges and snapped well with the waffle and plate. They made the waffle look more adorable as well!

Assembled yo-yo




Key features of the assembled yo-yo include the two halves, string, axle spacer, and the shoulder bolts. We consider the final assembled yo-yo a success, particularly for beginners (like us). The waffle yo-yo is easy to use and spins well. Opportunities for improvement include adjusting the string gap and assembly design so that the yo-yo can perform sleeping and other tricks.

Design Specs Vs. Measured Specs

Item Design Value [in] Measured Value [in] Explanation of Discrepancies
Max Diameter of Yoyo (plate) 2.630 2.630 Within spec
OD of waffle 2.200 2.203 Within spec
OD of snap fit (ID of waffle) 1.950 1.951 The outer diameter of the snap fit averaged to be 1.951”, which is slightly larger than the design value, but still within specification for the part.   
ID of snap fit (plate extrusion) 1.960 1.961 The inner diameter of the snap fit was actually out of the specifications that we had originally designed. However, the plate mold was not re-machined to fix this discrepancy because the snap fit with the waffle fit extremely well. Thus it was determined that the discrepancy was negligible.  
Height of waffle 0.400 0.402 Within spec
Height of plate extrusion (nut) 0.200 0.167 The height of the plate extrusion was not within spec, mostly due to the shrinkage of the plastic after injection molding. This dimension was not critical to the function of the yo-yo, and the slightly out of spec final dimension is acceptable.
Height of rib from plate bottom 0.300 0.299 Within spec
Thickness of plate (inner) 0.090 0.092 Within spec
Thickness of plate (middle ring) 0.130 0.132 Within spec
Thickness of plate (edge) 0.080 0.084 Within spec
String gap 0.100 0.09 The average measured string gap was slightly smaller than the designed value, and this was due to a decision to tighten the assembled yo-yo’s shafts slightly more than expected, in order to create a more snug/secure assembly.
Total width of yoyo half 0.53 0.544 The discrepancy between the design value and the measured value for the total width of the yoyo half is likely due to slight deviations in the thicknesses of each component of the yoyo; although each of the components are within spec, together, the cumulative variations cause the total width to be out of spec. This was not seen as a problem.
Total mass 0.13 lbs 0.13 lbs Within spec
Moment of Inertia 0.08 lbs*in^2 0.08 lbs*in^2 Within spec
Estimated Maximum Rotation Speed 1798.2 RPM 1798.2 RPM Within spec
Tolerances (all tolerances not listed here are +/- 0.005”)
Location (+) Value [in] (-) Value [in]
OD of snap fit 0.005 0.00
ID of snap fit 0.0 0.005


untitled shoot-017Findings about Plate from Written Deliverable 4 

For the plate, we measured the inner diameter of the snap-fit between the plate and the waffle (which was the inner rib of the plate). It was determined that the average value for the inner diameter was 1.961”, which was actually slightly larger than the specified design value. Overall, the range of measurements for the inner diameter varied from 1.960” to 1.965”, and although these measurements were technically out of the limits set by the design specifications, no change in design or mold was made for the plates (since the plates fit extremely well with the waffle snap-fit). It was thus determined that the originally tight tolerances set on the snap-fit were not as necessary, and the range of inner snap-fit diameters achieved were acceptable.

Although a disturbance in the production system (an increase in the injection pressure) was introduced during the manufacturing of parts numbered 50-60, there was no detectable difference in the measurement of the inner snap-fit diameter. This led us to the conclusion that the designed part was decently robust and could withstand parameter changes in the manufacturing setup.

Summary of Cost Analysis for Prototyping Versus Manufacturing

For 100 Parts:

Item Cost Quantity Total
Mold Blanks $18.62 2 $37.24
Thermoform Blank $3.12 1 $3.12
Resin $3.90 7 $27.30
Thermoform Plastic $0.25 120 $30.00
10-24 Hex Nuts $0.01 100 $1.00
Axle Spacers $0.49 50 $24.50
Shoulder Bolts $0.77 50 $38.50
Design Labor $50.00 40 $2000.00
Machining and Programming Time $25.00 60 $1500.00
Injection Molding Run Time $15.00 5 $75.00
Thermoforming Run TIme $15.00 2.5 $37.50
Instruction and Overhead $35.00 40 $1400.00
Total $5,174.16

For 100,000 Parts:

Item Cost Quantity Total
Mold Blanks $18.62 2 $37.24
Thermoform Blank $3.12 1 $3.12
Resin $3.90 7000 $27300.00
Thermoform Plastic $0.25 100000 $25000.00
10-24 Hex Nuts $0.01 100000 $1000.00
Axle Spacers $0.49 50000 $24500.00
Shoulder Bolts $0.77 50000 $38500.00
Design Labor $50.00 40 $2000.00
Machining and Programming Time $25.00 60 $1500.00
Injection Molding Run Time $15.00 5000 $75000.00
Thermoforming Run TIme $15.00 2500 $37500.00
Instruction and Overhead $35.00 40000 $1400000.00
Total $1,632,340.36

Key Plot:

Screen Shot 2016-05-09 at 2.26.26 PMChanges for Mass Production

For mass production, the waffle core and cavity would need to be adjusted. Currently, there is a lot of warpage due to shrinkage on the sides of the waffle due to uneven thickness. The core would need to be redesigned so that the entire waffle has even thickness. We also had an issue where the waffle would sometimes get stuck to the cavity mold. Currently, the issue was resolved by creating a burr on the core mold. This would not be an ideal solution in mass production. Instead, the molds could be designed with larger draft angles or the ejector pins could eject the waffle from the cavity. The plate and butter production required a significant amount of human/manual labor. For the butter, a lot of labor was put into cutting the thermoforming material, placing it into the machine, and cutting out the final part. For mass production, the entire process could be automated to save time and human capital.

Team’s Constructive Recommendations

We all agreed that we found the most useful and interesting aspect of 2.008 to be the labs and applying what we were learning during lecture to design and manufacture 50 yo-yos. It was well-planned that the machining, injection molding, and thermoforming lectures were towards the beginning of the semester so we could directly apply what we were learning to designing the molds and adjusting injection molding and thermoforming parameters. In contrast, the general organization of 2.008’s pop-quizzes, problem sets, and quizzes has a lot of room for improvement. The pop-quizzes were often on small details in the reading. If the pop-quizzes were actually meant to just check attendance and reading completion, the questions could have been better tailored so that anyone who did the reading could answer the questions, not just those with extraordinary memory. The problem sets were sometimes frustrating in that some of the material was not covered during lecture. Particularly for micromachining and thin films, the lecture went through the process very quickly for a subject that not many students have had prior exposure to. This topic definitely could have been addressed more in depth. Additionally, in the future, the materials needed for studying for the quizzes should be posted well in advance, not a couple hours before the exam (solutions to problem sets covered in Quiz II). Finally, the quizzes given would seriously benefit from some proofreading by either the professor or the TAs — we were consistently given quizzes with multiple errors, typos, and confusingly worded questions that the TAs would sporadically alert us to during the actual taking of the quiz. We think that this manner of writing and administering quizzes is extremely unprofessional, and we hope that the administration of the class spends some time to improve this aspect of the class for the future.



Process Optimization and First Production Run!

Plate Process Optimization

Process sheet for injection molding:

Screen Shot 2016-04-25 at 1.57.58 PM

For the plate, it took a bit of experimentation with the injection molding machine to determine the appropriate process parameters. We started out with a shot size of 20mm, but soon determined that the mold wasn’t filling completely, and there was a short shot every time we injection molded. Bumping the shot size up to 22 mm seemed to produce the best results, with no visible defects. We also experimented with the injection pressure, eventually settling on 500 psi, which produced a plate that had no warping.

Waffle Process Optimization

Process planning sheet for waffle:

Screen Shot 2016-04-25 at 1.58.09 PM

As with the plate, a lot of experimentation was necessary before we arrived at proper parameters for making the waffle. Initially, our shot size was much too large (32mm) and we had difficulties with the packing phase. This, in combination with inadequate holding pressure, led to undesirable flash. To alleviate this problem, we reduced holding pressure, increased packing time, and reduced shot size. After minor adjustments of these parameters and increasing the injection speed from 50% to 70%, our waffles looked very edible.

Summary of Overall Process Optimization


Once we had the basic process parameter settings down for the plate, our biggest problem was trying to minimize the sharpness of the edge of the plate. We originally thought that the sharp edge was caused by a bit of flash, so we tweaked a bunch of the process parameters to try to eliminate this feature. Unfortunately, none of the changes in process parameters made a difference in reducing the sharpness of the edge of the plate, and we determined that the issue was primarily caused by the location of the parting line on the piece — it would be extremely difficult to completely eliminate the sharp edge, given the design at that time.

With this in mind, we re-machined the core mold, this time extending the curvature of of the plate so that there would be a smoother edge. We then ran the molds on the injection molding machine, using the our original process parameters. Since the change we made to the plate was not drastic, the original process parameters worked well, and those became our final process parameters to use with the re-machined molds.

Screen Shot 2016-04-25 at 1.58.21 PM


After injection molding a few plates and waffles, we realized that the snap fit was too loose. The waffle component did not shrink as much as we thought it would. We re-machined the core mold of the waffle to get a better snap fit with an overlap of 0.01 inch, whereas before, there was no overlap in between the ID of the waffle and OD of the plate. With the re-machined core mold, the snap fit was tight; we were not able to get the waffle off the plate.

Since our waffle component has uneven thickness, during the process optimization stage, we wanted the part to have the least warping without introduction of defects such as flash or short-shot. When we played around with parameters, we encountered problems balancing pressure and injection speed. The mold was filling very slowly and occasionally had cases of short-shot. When we increased the pressure, flash occurred. To address this, we opened up the runner to allow more area for the plastic to flow in. After doing this, our parameters were easier to optimize. We found an appropriate balance of injection speed and pressure (70%, 500 psi) that addressed both short-shot and flash issues. The shot size of 27mm was optimal for the packing time of 6 seconds without causing flash. Ultimately we were able to achieve consistent results with these parameters.

During our production run, we encountered another problem when trying to achieve the perfect golden-brown waffle color. We discovered that mixing two colors was much more consistent than mixing three colors. When we mixed cream, brown, and yellow, our color ranged from light brown and cream to olive green and dirty yellow. We decided to stick with just the cream and brown plastic pellets.

Screen Shot 2016-04-25 at 1.58.30 PM



When the only setting we could change was temperature, we found that 625oF worked well with our part. When the display was fixed, we tried to optimize time-wise so production could go just as smoothly but in a shorter amount of time. First, we decreased the heat time from 30 seconds to 22 seconds, and the form and open delay times from 15 seconds to 10 seconds. With those settings, even though production was faster, the base of the butter was not clearly defined enough to snap together with the waffle and plate. Next, we increased the heat time to 25 seconds, but the base was still not well defined enough. We increased the temperature until 640oF, at which point the base of the butter was well defined enough to fit with the rest of our parts, and the top of the butter that’s exposed in the yoyo was still round enough to give it a soft buttery look.

Final thermoforming settings

Not pictured: butters 11-120

Plate Mold

Mold Description

We chose to highlight our body mold, which is used to create the plate feature in our waffle-themed yo-yo. The plate mold consists of two parts: the core and the cavity. Since the outside of the plate has a clear curvature, we chose to make the parting line between the core and the cavity molds along the edge of the curvature — to minimize visibility of the parting line on the final produced plastic part. The cavity mold is straightforward — it is a circular cavity with two sections of different depths. This ultimately forms the underside of the finished plate. The core mold was slightly more complicated, with a raised middle that creates the circular rib on the finished plate, where we will snap our finished injection molded waffles.

Shrinkage Calculations

The dimensions that we determined for the mold are based on an estimate of 2% shrinkage, which we determined by measuring similarly sized molds to their finished parts in lab. To obtain a 2.500” diameter in a finished plastic plate, the mold outer diameter was designed to measure 2.550”. To account for shrinkage and achieve a toleranced snap fit, the outside diameter of the plate’s rib feature was designed to measure 2.000”, while the inside diameter of the waffle was designed to measure 1.990”. With shrinkage, the two parts should cool at 1.960” on the plate and 1.950” on the waffle, and create a snap fit. The thickness of the plate was not a major concern in scaling our part up, because due to its current thickness it would be sure not to bend or fail. However the molds measure an inner plate thickness of 0.120”, which would shrink to 0.118”. From the bottom of the plate, the rib according to the mold is designed to be 0.306” tall, though after shrinkage, the final rib should measure 0.300”.

Manufacturing Process

We first manufactured the plate core mold on the lathe, using tools 7, 8, and 9 to create the grooves in the part. Next, we used the 0.125” drill to drill the space for the hex nut. Then, we used the mill to drill the ejector pin holes in the core mold, finishing up with the drill press to ream out the holes. For the plate cavity mold, we first used tool 10 on the lathe to create the necessary features. Then, we used the center drill and the 0.234” drill to bore the sprue hole. Finally, the we used the mill to create the path for the runner.

Step-by-step Process Plan

Plate Core Mold Process Plan

Step Operation Machine Tool Justification
1 Insert stock Lathe N/A Need stock to produce mold
2 Load G-code Lathe N/A To initiate program
3 Zero tool Z axis Lathe 9 So the lathe knows where measurements are relative to
4 Rough groove Lathe 9 Tool 9 has a full radius tip of 0.0385”, which was the right size to bore out the begin the trepan
5 Finish trepan Lathe 7 Tool 7 is good for performing cuts in close quarters, and its point radius of 0.01” was good for finishing the trepan corner.
6 Finish trepan Lathe 8 Tool 8 is good for performing cuts in close quarters, and its complementary geometry to tool 7 made it the right choice to finish the other corner of the trepan.
7 Bore nut hole Lathe 0.125” drill Bore hole in the middle of the piece to create room for the hex nut.
8 Remove piece and clean up Lathe N/A To leave machine ready for the next use

Plate Cavity Mold Process Plan

Step Operation Machine Tool Justification
1 Insert stock Lathe N/A Need stock to produce mold
2 Load G-code Lathe N/A To initiate program
3 Zero tool Z axis Lathe 10 So the lathe knows where measurements are relative to
4 Roughing Lathe 10 Removes material in a facing operation according to the curve we designed. Tool 10 is a boring tool with a small enough radius to produce the curvature and sharp edges that the plate silhouette requires
5 Finishing Lathe 10 Tool 10 is also capable of finishing, when set to a slightly lower feed rate.
6 Drilling Lathe Center Drill Creates an initial hole in the back of the plate cavity
7 Drilling Lathe .234” diam drill Finishes the hole drilled in the previous step, drilling through the entire width of the mold; this creates a channel for the nut to be held in place and the plastic to flow around it
8 Contour Mill ⅛” Ball end mill Creates runner from sprue hole to the back of the part

Manufacturing Time Estimate

Manufacturing Step Estimated Time Justification
Fabrication of Plate Cavity 2m 19.31s Projected machine time from Mastercam
Fabrication of Plate Core 3m 3.43s Projected machine time from Mastercam
Fabrication of Waffle Cavity 12h 30m 25s Projected machine time from Mastercam
Fabrication of Waffle Core 2m 9.18s Projected machine time from Mastercam
Fabrication of Butter Core 1 h 9 min 43.25s Projected machine time from Mastercam
Process Optimization Plate 1 hour 1 hr should be enough time to make minor changes to the mold as well as tweak process parameters on the injection molding machine.
Process Optimization Waffle 1 hr Assuming only minor changes need to be made to the hold, 1 hr should be enough time to fix the mold as well as tweak process parameters on the injection molding machine. If the mold has to be remade, however, then we will require significantly more machine time.
Process Optimization Butter 1 hr Similar to the Waffle, 1 hour should be sufficient for small mold changes as well as process parameter tweaking, but more time will be necessary if the mold needs to be entirely remade.
Final Production for 1 Plate 45 seconds This was a generous estimate of how long we spent for each cycle on the injection molding machine.
Final Production for 1 Waffle 45 seconds We expect the waffle production to take a similar amount of time to the plate.
Final Production for 1 Butter 45 seconds We expect the thermoforming to be a quick process, so 45 sec for each cycle should be reasonable.

The changes made to this schedule are mostly in the final production for each part — we changed the time estimates (to produce one of each part) to 45 seconds. Based on our first testing runs of injection molding with the plate, 45 seconds is a generous estimate of how long each cycle takes on the machine. Since each part will vary (especially for the cooling time), we leave our final estimate at 45 seconds, which should hopefully be reasonable. I don’t think any of these changes will affect our team schedule. The biggest challenge for our schedule will be if we need to remachine the waffle mold.

Photographs of Manufactured Molds


Plate Cavity Mold

Screen Shot 2016-04-04 at 2.54.50 PM

Plate Core Mold

Screen Shot 2016-04-04 at 2.54.39 PM

Overall Design and Specifications

We will be spending the next six weeks preparing breakfast: a golden-baked waffle on a plate, complete with a pad of butter. Our yoyo will be breakfast you can spin at any time of day, making mealtime unlimited fun! Each half of the yoyo consists of one waffle, one plate, and one slab of butter placed in the middle of the waffle.

Each half of our waffle-inspired yo-yo contains three standalone parts: the injection molded base plate, the injection molded waffle, and the thermoformed butter. Specifically, the injection molds for the base plate will be machined entirely on the lathe, the injection molds for the waffle will be machined on both the lathe and the mill, and the thermoform mold for the butter will be machined on the mill. Both the plate and the waffle will be molded with polypropylene, and the thermoformed butter will be molded with clear polystyrene. Once molded, the plate, butter, and waffle come together in a sandwich-like fit — the butter is inserted into the assembly from the underside of the waffle, which allows it to have long thermoformed flanges and supports that will be hidden from view. The end of the butter’s flanges rests on the rim of the plate, and the waffle snaps into place around the same rim, effectively sandwiching the butter’s supports between the waffle and the plate. Once each half of the yo-yo is completed, the two halves are held together by a sleeve.

In order to make sure our breakfast assembly would fit together well, we considered tolerances and shrinkage in our dimensioning. Draft angles and part thicknesses were taken into account across each component to ensure consistent manufacturability and to minimize unwanted effects in our final parts.

The plate is designed to be manufactured by injection molding. The bottom of the plate

is a filleted extruded part, which creates the surface parallel to the windings of the yoyo string. The fillet prevents the plastic from slicing through the string or any wayward fingers. The curvature of the plate was established to emulate the silhouette of a dinner plate. On the surface of the plate we’ve built a circular rib, sized to press-fit with the waffle part. We also built an inner cylinder to surround the hex nut, where our shaft will screw in. We’ve removed material between the cylinder and the rib, where we would expect to see material deformation had we not removed as much material as possible. And lastly, we drafted the inner edge of the rib at three degrees which will ease its removal from our mold.

The waffle part was carefully scaled down from the dimensions of the real waffle while also taking into account the limitations of our mill tools. The waffle holes and filleted edges are all machinable with a 1/16” minimum tool diameter while maintaining the delicious look of a real waffle. Draft angles were incorporated in all of the holes and edges to assist in part ejection. To minimize warping, the thickness of the part was kept to a consistent 0.1”.

The butter part was designed while carefully considering all the parts around it and also the properties of the thermoforming process. The butter extends into a disk of radius 0.97”, so that it could be sandwiched between the waffle and the plate, which have radii of 0.98” at the point where they snap together. The disk also has 8 ribs to make it more sturdy and for the fit to be more snug. The height of the butter above the disk base is 0.25”, which is slightly taller than the waffle, which is 0.24” thick. This allows the butter to pop out a little above the waffle to give it a more realistic look.

Table of Specifications

Item Value [in] Measurement Method
Max Diameter of Yoyo (plate) 2.500 Calipers
OD of waffle 2.200 Calipers
OD of snap fit (ID of waffle) 1.950 Calipers
ID of snap fit (plate extrusion) 1.960 Calipers
Height of waffle 0.400 Calipers
Height of plate extrusion (nut) 0.200 Calipers
Height of rib from plate bottom 0.300 Calipers
Thickness of plate (inner) 0.090 Calipers
Thickness of plate (middle ring) 0.130 Calipers
Thickness of plate (edge) 0.080 Calipers
String gap 0.100 Calipers
Total width of yoyo half 0.53 Calipers
Total mass 0.13 lbs Scale
Moment of Inertia 0.08 lbs*in^2 Solidworks
Estimated Maximum Rotation Speed 1798.2 RPM Hand calculation
Tolerances (all tolerances not listed here are +/- 0.005”)
Location (+) Value [in] (-) Value [in]
OD of snap fit 0.005 0.0
ID of snap fit 0.0 0.005

Our project timeline and progress are recorded in this Gantt chart and Google Doc!

2.008 Yoyo


Anjali Krishnamachar and Kira Schott will assume the task of designing and machining molds for the injection molded base plate, as well as carrying out the injection molding process to manufacture one hundred base plates. Grace Li and Wen Zeng will assume the task of designing and machining molds for the injection molded waffle, and they will manufacture one hundred waffles. Finally, Tracy Cheng will assume the task of designing and machining the thermoformed butter mold, and she will manufacture one hundred thermoformed butter parts.



Hi! We’re a 2.008 Thursday morning lab group, and we’re waffly excited to get started on this yo-yo project.

From left clockwise, we are Tracy, Kira, Anjali, Wen, and Grace.

Anjali enjoys doing pull-ups, procrastinating from her work by baking banana bread, and going for long romantic walks by the Charles! Kira loves to run and play outside which is a contributing factor to why she’s a member of Women’s Lacrosse at MIT. She lives in Theta on campus, does Camp Kesem, and hails from Connecticut, USA. Grace enjoys shooting videos of waffles in her free time. Wen is a breakfast enthusiast and photographer from Los Angeles. Tracy likes dancing, making things, playing the piano, and laughing a lot.