The file is on version 59 at this point and I have 4 more files that I completely scrapped. Ive done a lot of failing but finally success!
To make the experience fit your profile, pick a username and tell us what interests you.
while designing a tooth profile for a strain wave gear I have found that your main goals should be to; first, make it so its comfortable for the teeth to mesh so you don't lose a whole bunch of energy to friction and its easy for the wave generator to mesh the teeth, and second you want to make sure the teeth have a hard time slipping over each other.
One of the ways they tend to collide is through over or under expanded stator inner diameters. The number of gear teeth that will fit onto a given diameter is not perfect and there will be small little gaps between the teeth that will cause the inner teeth that are theoretically perfectly next to each other to be slightly out of phase with the outer gear ring. If you go to mesh your teeth by using a simple move tool and find that the teeth are completely overlapped extremely soon after the crest of the strain wave you will likely have meshing problems and your performance will be affected. You can fix this by expanding or contracting your stator by increasing or decreasing the flex distance. However if you decrease the flex distance you risk tooth tip collision which puts you back at square one. The solution? get rid of the tooth tips on your stator.
The second part of the game is getting tall tooth walls that aren't too skinny. This is more one of those things that is just the right answer and not that complicated to understand compared to the tooth meshing issues. you don't want your tooth walls to be excessively thin and tall because there is a point of diminishing return where it will negatively impact the number of teeth that you can have meshed while adding more teeth. even if they are tall and hard to slip you'll have little to no moving torque or your drive might completely fail because of how hard they would be to mesh and how much the flex spline would have to flex to prevent TT collisions. Another thing to watch out for is that triangular teeth will fail, they will slip right over each other with little to no effort, trust me I tried everything. It is beneficial to make your tooth walls curve upwards so its harder for them to slip over each other.
Every strain wave gear is different for every situation, my suggestion is to make a test strain wave gear with as many tooth independent parts as you can so you can rapidly iterate tooth design.
This section will be mostly stats and it's not gonna do a deep dive into the mistakes I made along the way. If you are intending to read this whole writeup and you fear redundancy, the information in this section will be covered again, so feel free to skip.
The flex spline of my design has 114 teeth on the flex spline and 117 teeth on the stator giving the design has a +3 tooth difference between the flex spline and stator, I had a bit of an epiphany and realized that the number 3 is bigger than 2, this trades gearing ratio for holding torque.
This is perfect for our application because we need most of all is for our HD to hold its position under high loads with minimal deflection. I am still testing holding torque but I have broken several test arms without making the actual drive fail, up to a peak of 52Nm. while this does have quite a bit of flex due to the plastic drive shaft and more specifically the plastic load arm. I will be casting a load arm soon from aluminum and testing to failure.
I do not know precisely what the moving torque is because I haven't finished my test bench with a load cell yet but I will update this when I do.
The project in question is strain wave gear, sometimes called a harmonic drive (this is a brand name and for the sake of explanation it makes a little more sense to call it a strain wave gear). It's a complex gearing mechanism that allows for really high gear reduction with no backlash and excellent holding torque. So how does it work? Well I'll get to that, chill out and sit back down man. If you like reading keep going, however if you're more of the youtube type; this is a good explanation.
The strain wave gear is a complex gearing mechanism that functions by having a stator ring of teeth (light tan) and an inner ring of teeth with a thin inner wall that allows the ring to be flexed into an ellipse; this part is referred to as the flex-spline (dark tan). The stator has a set number of teeth more than the flex spline for now let's make that number two, this causes the teeth to align in a way that they can mesh in two spots 180 degrees apart. In these areas a rotating cam called the wave generator is used to flex the flex-spline into an ellipse, meshing the two sets of teeth together. Then as the wave generator rotates, it meshes the next set of teeth, moving both of the peaks of the strain wave. By the time that wave generator gets back to the same spot, the internal teeth have meshed into 2 more slots than the external teeth, and thus relative motion is created between the flex spline and the stator, in less fancy terms; wave generator go brrrrrrr, flex spline rotates really slow in the other direction. So why would you need it? Jeez you ask a lot of questions kid. Well, all sorts of things, this is designed to be used for the rotary axes of a CNC mill, but it could be used for whatever you want. Because of its internal mechanics, it is impossible to back drive and is highly resistant to failure, it has no backlash and an extremely high gearing ratio for its size. This makes it precise and robust, perfect for moving exactly where you want it to and staying there until told otherwise.
I have not completely written up all of these sections and I have not completed testing completely so you will be getting updates as I finish stuff, whether that's writing a section or finishing some sort of test. I appreciate your interest :).
Here is the outline
Prerequisite: What and Why
Quick Stats: For The Engineer On The Go
Section 1: Tooth Design Fueled By An Excessive Amount Of Caffeine
Section 2: How Flexy Can One Flexure Get
Section 3: Wave Generators, and the Question of Why I'm Not a NASA Engineer
Section 4: Everything Else
Section 5: 3d Printing Is Going To Institutionalize Me
Section 6: Testing, And The Reason I'm Not A NASA Engineer.
Become a member to follow this project and never miss any updates