During my Junior year of college, I had to take many challenging engineering classes. Two of which was MECH 340, for Mechanical Engineering Design and MECH 308, for Finite Element Analysis (FEA). In the previous semester in the fall, I took CIVL 311, for Strengths of Materials, where you discussed deflections, stress and many other structural or strength fundamentals, and it was hard. During that time, I had some classmates (in other classes, not in 311) and students tell me that MECH 340 was CIVL 311 on steroids and it would be a lot more hard, thankfully, it was not. In 340, there was a lot of mechanical aspects which were discussed such as gears, more stresses, bearings and other components, and using the knowledge from that, had to complete a final project. This final project was to build a coffee grinder.

A couple ideas were thought of on how to build this grinder. The first was made from steel and had a structure/tower like design. It had an industrial feel but felt that it would be too heavy and hard to assemble. Its also steel and could rust, which for a food related item, is not good. The second was made to get a more clean look, and it featured a small opening to insert the coffee beans, This idea/design helped transition to the final design with small modifications being made long the way. However, during assembly, the hole was ditched since it would be hard to load the coffee and some may get stuck. A door which opened exposing everything was more convenient, and the professor liked the clean look and how it all closed up. The base of the grinder more or less stayed the same throughout the project.



The final design was a modified version of the second idea with a few changes. These changes were either done before the part was made or after it was made, in the sense of misreading values or building the part then adding it when it wasn’t on the original drawings. The final design which we came up with is seen below.

The general idea of the activity to start the grinder was to design a pair of shafts, (input and output) which were connected to each other with a pair of bevel gears. Our main task was to find which size shaft would be best fit for the job; with the given materials and how long the shaft would last based on the stresses and endurance life. Using MS excel, the moments, reactions, stresses and factor of safety were found, using this value, we can see if the device would or would not have infinite life, and if it didn’t, how many cycles it would have before it broke. In the preliminary analysis based on rough estimates on the forces which may be seen while the machine is in use, the diameter at each region can be found. For simplicity, we can find the largest bearing which is needed and use that at each location. Once that value was found, we would move up to the next nominal sized bearing, which in this case was 12mm. However, since the pinion’s face was just over 9mm, a 12mm shaft diameter at that part would not work, to account for this, a reduced section was added which allowed a larger bearing to be used but a smaller section which the pinon would be pressed on.
Once the lengths and diameter of the shafts were found, design of the grinder could be started. Using SolidWorks, I designed the grinder, starting with the shafts and the bearing blocks, then I thought of the design for the structure of the grinder. I wanted to enclose the device to contain all the grindings and make it look more clean/nicer looking. The first design had a more industrial look but it would be made from steel and bolted together, a small drawback to this would be it would be heavy and at the end, since this a food related product, a material which does not corrode or cause any health concerns would be chosen.
Adjuster
The main objective of the grinder was to grind coffee, however, many grinders are adjustable so you can fine tune it to get the best tasting coffee. This was an optional requirement for the activity, but if we are going to do it once, might as well do it right and challenge ourselves. After a lot of thinking, I thought of a way on how we can adjust the grind, unlike common coffee grinders where you adjust the burr on the shaft to make it more fine or coarse, we opted to adjust the outer part of the grinder with a giant nut. The part would be press-fitted into the adjuster where it can be raised or lowered to adjust the ground size. This made it easier to run calculations since the region where the force was applied was at the same point. A small recess was cut in the top of the adjuster to make the top of the grinder flush, and if the press fit wasn’t tight enough, it will hit the side of the recess and stop its rotation/allow it to work still.
The original idea was to cast both parts due to their size and finish it up on the lathe, but there were large enough pieces of scrap which worked. and one had a lot of material removed already so it made the process a lot faster.





For the adjuster nut, it was turned and bored on the lathe and was finished on CNC mill. Since the threads are very deep, (2 1/2 – 4 UNC) compared to most threads, the shop did not have the proper tooling to make that deep of a cut, so the part had to be done on the CNC where a rotary V cutter was used. The instructor helped make the G-code for it.

Shaft and Bearings
During the manufacturing of this component, we received the stock for the shafts, however, there was a slight problem, it was too small. The shaft’s OD was 12mm, and the original plans which we had required at least 14mm, so we had to move down to the next nominal size, which was 10mm, this made the shaft not reach infinite life, but it was still enough to operate properly and not break, and get a lot of use before it did. This also made it a tiny bit easier for the burr since the bore for it was already at 10mm. Unfortiently, this simplicity caused a small problem, since the bore for the bearings and the burr would in theory be at the same size, we would need to be able to slide the bearing onto the shaft to its shoulder while still having enough space for a press fit for the burr, this distance would be around 4in and it may be hard with the tools which we have in the shop. After the shaft was turned, during a test fit with the bearings which we got, (6900-2Z), we found that the ID of the bearing was smaller than the ID for the burr, due to this, this eliminated the possibility of the burr press fit. With assistance from one of the shop technicians, we decided to turn the shaft to where the bearing would slide on and braze the burr on to the shaft. However before we could do that, a few other pieces had to be on the shaft since the OD of the burr would be larger and they would not be able to be installed afterwards. Unfortiently, the plans were misread and the shaft was made too long so we had to grind the braze and remove the burr, luckily, nothing broke
| Bearing |
| Bearing Block |
| Base Cover |
| Grinder Burr (Outer) |
| Adjuster |
| Burr |
Power Transfer
Another design decision which we made were how the gear and pinion were going to be attached. For the input shaft, we opted to the original idea to press fit it, this allowed us to get a good feel for the gear mesh and adjust it by pressing it further onto the shaft. For the output shaft, we decided to use a set screw, this would make it easier to assemble the project since that part can be added and locked in place once everything was set up. Since the bearing block B looked too close to the gear, a small slot was made to give it extra clearance.

Once this was done, the shafts were more or less done. But we still needed a handle to power the device. Using aluminum, I milled a handle and drilled a hole which slid onto the input shaft where a screw can be used to tighten it onto the shaft and hold it in place. In total, there were two lever arms which were machined. The first one was made by eye but the slot/counterbore where the screw set was too wide and too deep. Having it too wide didn’t affect it a whole lot besides it’s appearance but since it was too deep, it cut though the side of the hole which mounted onto the shaft and when the screw was tightened to where it was able to turn and grind, we noticed that the part had permanently deformed and failed. Since aluminum isn’t as elastic as steel and it tends to be buckle in some cases, this is why it failed. To improve and fix this, a new handle was made at the same size but the slot for the screw was large enough for the screw head and the counterbore was not as deep. At the same time, we decided to stay with the machined finish instead of buffing it and not add any decorative slots. I personally like the look from the cutter on the part and it matches the other parts which were made.
Originally this part was going to be casted out of aluminum then cleaned up on the mill but we decided to mill the part since it was a lot faster, 1hr vs 2-3hrs and if something happened, which it did, it would be easier to redo the part. There were also stress and deflection worries for it and that it would fail.



We also had the opportunity to use a bronze handle for the project, but since the part was extremely heavy, we didn’t want to add any extra stresses or strain to the device from its weight, so we decided to turn one instead. We wanted to use it for the bling factor but our eyes were bigger than our brains, luckily this was an easy part to replace or change out. To allow the handle to rotate during use, a long bolt was used to attach the handle to the lever, which is threaded for the bolt. However, if the handle bolt isn’t tight, it may unscrew itself during use.
Structure
For the project, the design of what the grinder would look like is up to the team, to design on SolidWorks and build. Since I was very enthusiastic to start the project and wanted to keep the product, I was the one who designed the project. I was also the one who had the most experience with the CAD software. I started designing this during activity 4 for the class. Knowing that the metal shop had a large supply of hardwood countertops, I decided to use that as the choice of material for the grinder. These materials were walnut and maple, I picked them since they were hardwoods, had a nice contrast with one another, and once finish was added, would make the project pop, which it did. In addition, the finger joint pattern on it was cool and it can also symbolize dark and light roast coffee.
The base consisted of walnut, and the maple was for the top part. The wood was originally 1 1/2″ and 2″ thick, so the pieces were resawed on the table saw. Using the bandsaw in this case would’ve been better, but the blade on it was dull and the table saw was the next option, but it made a clean cut so at the end, it worked. The parts were all cut on the miter saw and table saw, all holes were drilled on the drill press with a forester bit, luckily the bearings which were purchased were a hair smaller than one of the bits so it made a good fit. In addition to using the saws to cut the pieces, a drum sander was used to bring the parts into thickness, however the scale was off so the parts are 1/32″ thicker than they should be.
For this part, a CNC router was used to cut a few of the holes, these locations were: the front and rear panel for the opening for the drawer (for grain match) and where the adjuster sticks out, and the holes for the cover of the base and the hole for the adjuster nut. The CNC was used for those two holes since there wasn’t a drill bit large enough. Next to that, a laser engraver was used to do the decals, warnings and other labels, and luckily, the person who was running the engraver noticed a mistake in our spelling; course and coarse are not the same.



All the parts were drilled, countersunk and screwed together, with around 50 screws. However with the stack up from the output shaft, we had to use dowels to hold the grinder to the base, and is held on with a tiny amount of glue, however for travel/carry, its best to hold it by base, not the top.
Finishing Touches
After all the parts were done, there were some finishing touches which had to be done. One idea which I had was to make a scoop, something small which can be used to insert the coffee beans without the user overfilling. So by eye, I made one, with a ball and flat end endmill, I also buffed it to a shine. In addition to the scoop, a mount was made to attach it to the base, this was made similarly to its mating part on the mill. The scoop is very thick and is overkill.



To protect the grinder, wipe on poly was applied to it, it also helped bring out the color in the wood, about 1.5 coats were applied.


In addition to the bearings being purchased, a drawer knob was also purchased, the original idea was to turn one on the lathe, but we opted to buy one instead. A second knob was purchased with the order too incase we had enough space for the adjuster pin to lock the adjuster while in use, but there was not enough space so a small knob was turned on the lathe and slid onto the bolt. A small funnel was 3D printed to help direct the beans into the grinder. This had a small cut in it to be able to be added onto the shaft after everything was together, however that made to more complicated and it was not easy; it looked better on paper.






Decisions and Revisions
Since the start of the project, there were a handful of design changes which were made, either intentional or unintentional. Many of the revisions can be seen earlier in the post.
The top part of the grinder was changed from a tower like design to the one which was built. A few changes to how the front panel was mounted and what it had one it was also changed. The Decal was enlarged, the hole was deleted and it was hinged on the grinder, not screwed directly on. A safety/how to part was also added to the inside of the front panel. The locations of the screws were also changed, in the design, they were mainly added to see how close to the bearings they would be and not have it interfere.
The shaft diameter and bearings were also changed, the diameter was originally meant to be 12mm but was changed to 10mm, respectively , the bearings were changed for the new shaft diameter but a smaller size bearing was chosen by the professor compared what we originally wanted, this did not affect the operation of the machine. The addition of the set screw to the gear was also done to combine the two parts and the burr was brazed on instead of being press fit.
The adjuster was originally meant to be casted then turned but a suitable piece of scrap was found and to lock it, it was going to have straight knurls on it but was changed to slot where a pin could be inserted. The hole for the burr was also changed and made a bit wider while machining it, this made it easier and gave more space for the funnel. The nut was machined similar to its male counterpart, but it was made a tiny bit thinner and it had to be finished on the CNC, not on the manual lathe.
Similar to the adjuster, the lever was going to be casted but was not due to the amount of work needed to make it and other stress and deflection concerns. A more rectangular one was made on the mill but with the part being too thin in a few places, it caused it to buckle and be scrapped. Compared to the 2nd one the final one did not have decorative slots and the location where the screw was located was thicker and only accounted for 1 screw, not 2, which is why the 2nd one had a wide area and had most likely failed.
The handle at one point was planned to be a bronze part which was found in the shop, but due to its weight and the idea that it could bend the shaft, damage the bearings or strip some of the screws, we opted to turn it on a wood lathe instead. This also made it match the rest of the assembly. I wanted to cast the parts because a can and it may look cool but we chose to make the parts another way.
The drawer walls were made thicker and a drawer knob was purchased. Additionally, the front was finished on the CNC to get the grain match from the butcher block instead of individually cut on the saws. A small bin was added to catch the grounds and a small slot was cut on one of the sides after being glued together to allow the screw for the knob to be countersunk and not stick out allowing the bin to fit.
Finite Element Analysis
As mentioned at the start of the post, one of the classes I took was FEA, and for that class there was a final project which you had to use FEA on and run a simulation on SolidWorks to test and compare your data and for that project, I decided to use the coffee grinder.
To simply the simulation, I created a second assembly with only bare essential parts and had the majority of it hidden. The simulation was used to compare the deflection in the shaft during use and the stresses at the fillets. To see if the locations of max stress which was specified earlier can be seen and determined and if the machine would fail in a jam case scenario.
There were two results which were gathered when the simulation was run, one for von misses stresses and another for deflection, on the x and y axis. The deflection on the output shaft from the x direction was similar to the hand calculations if we simplified the shafts to cantilevered beams with an overhanging load on each end. By doing this, we can use Beam equations to determine the deflections. However the deflection in the input shaft in the y direction had some error and wasn’t as accurate as the output shaft, this was most likely from how the simulation was run since the simulation was run backwards and the fixed end to make it jam was on the input shaft, not the output. I wanted to replicate where the forces were coming from similar to the original lab assignment.


For the von misses stresses, they were more close to their hand calculation counterpart. When the simulation was done, the max stress location was on the gears, so in theory, the gears would most likely fail before the shafts do, but that will be determined if it happens. In the preliminary calculations for the assignment, the locations of the stresses were calculated from the location of the moment in that region, knowing this, probes can be placed in those regions to find the stresses. Due to this being a rotating shaft, there will be a median and absolute moment which correlates in a minimum and maximin stress which are 180 degrees from one another. But the stress viewed from the probes may be a bit off since finding the location of maximum stress took a lot of trial and error to see which spot had the highest region.





Comparing the data calculated to what was simulated, some of the results were close and others were not. How I set up the simulation, I made the input shaft end fixed and added the forces to the output shaft just like the coffee grinder activity. Excel was used to make all the matrices and find the deflect while the data from the stresses were reused with the appropriate shaft diameters. Percent error was not calculated from these indifferences but the results were close enough.
Final Remarks
All in all, this was a fun project, I got to use a lot of large machines to make a small one. I enjoyed working with my team, everyone did their part and we were able to complete it. I made around 80% of the parts and designed the grinder, the other members help make the shafts and 3D prints the parts which needed to be printed, and they also helped make the final lab submission. This was my first physical group project in probably years due to Covid-19 and it was nice being back. At the same time to the lab submission, I made a user manual and a brochure and the packet which we submitted was over 80 pages long, this also included the drawings for each part. My classmates and professor were all amazed by the design and everyone’s design was unique and different, at the same time how everyone made the projects were cool, some 3D printed the whole project, others made it on the CNC mill and some used a waterjet. It was funny how after everyone was stressed to take the final (and died from it), we were all in the back of the class showing off our grinders and making some coffee.



































































