What is this?
The ME14 Transmission project is the first design and fabrication project in the caltech Mechanical Engineering curiculum. The idea is to work in teams of three to design and fabricate the best fixed ratio transmission possible.
Once built, the transmissions are tested using a known motor and load. The test apparatus tracks a number of parameters, but the two critical values are "Seconds to reach 250RPM" and "Maximum speed reached in 250 seconds". The latter divided by the former provides a final score for the competition in the form:
Score = [Max Speed] / [Time to 250]
Team Goals:
Before starting our design process, we got together to talk about our priorities as a team. Based on that conversation, we came up with three guiding attributes (below) for our transmission design. We also decided that although we're very willing to work outside the box we don't have any interest in creating something groundbreaking for it's own sake. Design innovations should decrease fabrication time, improve reliability, or increase professionalism/polish.
Projected score as a function of gear ratio
- Completely functional.
- Highly Reliable.
- Built to a high standard of professionalism and polish.
- *NOT* groundbreaking for groundbreaking's sake.
Gear Ratio:
Estimated time to 250 RPM as a function of gear ratio
One of my big projects this first week was drawing up a comprehensive model of the transmission's motion systems. I first defined moments of inertia for all of the rotating components of the system, and then used standard values for spurr gears and angular contact bearings along with the drag model we were given for the test apparatus to determine the parasitic torque as a function of speed. That, along with the stall torque, operating voltage, and max speed of the motor let me predict the gearbox's performance across a range of different gear ratios.
Within our allowed ratio range (5-7) I found that performance improved slightly for smaller values. However, difference in scores was small enough to be dominated by secondary factors like shaft alignment and bearing efficiency. With that in mind, we went ahead and selected a pair of high precision stainless steel gears from SDP. Our large gear has 52 teeth, and the smaller gear has 20 teeth. That gives us a stage ratio of 2.6 and an overall ratio of 6.76.
We expect a score of 8.04
An early draft of the central module using placeholder gears. The model is fully parametric, and shifts to accommodate different gear ratios.
General Structure:
Traditionally ME14 transmissions are built as a series of vertical walls, screwed to a base-plate and connected with rods, or another plate, on top to provide rigidity. This structure has the benefit of being easy to CAD, and fairly straight forward to machine since all the components are rectangular. However, it also seems relatively weak, difficult to adequately align, and fairly inelegant in that it's really only applicable to this exact competition.
The transmission casing and transmission stand. The number of screws was later reduced to 6 per circle in deference to our budget.
I choose instead to design our transmission as a series of stacked circular plates, separated by spacers and aligned by 1/4" steel rod. The gear shafts are held between the rods, and the entire assembly is kept under tension by four 1/4-20 cap screws. This module then slots into a larger aluminum cylinder with end-caps to keep it isolated from dust and damage. Jack then designed stand to hold the cylindrical housing in place and interface with the test apparatus.
If successful we will be the first team to substantially deviate from the standard design in the history of the course. My hope is that this structure will prove more rigid, more reliable, easier to assemble, and more compact than the traditional framework. It should be noted that this design will be substantially more difficult to machine. However, we decided it was better to spend time making things correctly once, rather than spending time aligning the components by hand. Since everyone involved has a fair amount of machining experience, we have the luxury of making that choice.
Materials Selection:
ME14 transmissions have historically been fabricated out of acrylic. Both because it is easy to use, and because it is provided by the shop. We decided to use aluminum. Working in metal means being able to hold much tighter tolerances, take heavier cuts (without worrying about chipping), and achieve a much higher level of overall polish. The one drawback to using aluminum for the main casing is that it means you cannot see into the transmission. Since our goal is polished, and strong, this is not an issue.
For gears, we choose stainless steel. High carbon steel would have been preferred. However, there were none available from an allowed supplier in the size range we wanted. Using 316 stainless, we have a factor of safety of around 9.6, well more than sufficient.
The drive shafts, alignment rods, and transfer shaft are all composed of O-1 tool steel. This was to accommodate our high-precision bearings, and because we wanted experience with the material. (4140 would have been a few cents cheaper and would likely have worked just as well).
Shaft Retention and Damping:
Our gear train and shaft retention system. Note that while all relevant components are included and accurately positioned the gear collar diameters are not true-size.
One of our big focuses for this design is minimizing vibration. With that in mind, we have decided to use o-ring spacers and retention clips to constrain each of the shafts axially. This lets us dampen any axial vibration we might get from shafts. It also allows us to easily and repeatably pre-load our bearings. This is important for ensuring the bearings wear in reliably, and is very important for the kind of high-precision bearings we have chosen.
To provide additional vibration isolation, the alignment plates will be connected to the outer housing with press fit o-rings. The housing structure, with it's nylon riders should also help isolate the entire structure from the base of the test fixture (which tends to vibrate when the wheel runs).
Oil Proofing:
Early in our discussions, we debated making the transmission fully water-tight. That only requires slight modifications to the IO bearings, and the addition of some shaft seals. However, we decided it was inadvisable to pursue that kind of feature without more time for testing. Despite that, we have taken care to keep the design as close to water-tight as possible.
Depending on the final transmission performance, we may take advantage of this to fill the inner housing unit entirely with oil. This would be a performance hit, but could also prove invaluable if we end up running in the transmission. If the transmission were ever to be used for serious torque transfer, oil filling would be required to protect the gears and wick away heat. It should be noted, that since we are using shielded rather than sealed bearings, we expect the unit to leak slowly. Fortunately, the rate of loss should be low enough to be inconsequential for out purposes.