I clicked because I was going to say something about the saw cut on your shafts, but now I have to compliment your video. That was well done, and your communication was high quality.
besides bevels being the worst gear of all, hardest to manufacture, hardest to make accurate, with the correct form... without a generator. sure, FDM works for experimenting but yeah... run out here just defeat the purpose. was that backlash test performed on every tooth? why mitre gears? with a hunting ratio? would want non-hunting for even wear. only saving grace of miter gears is you dont have to alter the setup of the gleasson to make the pair... i made a 105T FDM gear for my lathe years ago. it worked, but if you actually tried running a nut up the freshly cut M1 thread? (yes, it was missing the most important and most expensive to replace gear) the thread gets skewed. any inaccuracy in the gear train, leadscrew has nonconstant velocity, and the resultant pitch it cuts is not consistent. they look good, theres only a tiny little subtle bit of warping, the typical "grung grung" sound of unmatched teeth... and that transfers down to a pitch that may vary by 0.02mm on every revolution... i aint got the tools for measuring pitch accuracy! i finally machined a real one up with a dividing head, and yeah... that fixed that. to the tolerance of the machine, anyway. not precision but no worse than any other lathe i guess... .
It would probably be more accurate if you use a friction model and identify the viscous and Coloumb friction coefficients as viscous friction depends on the angular velocity.
Bro how do you not have more subs…very solid content
please show some of the output data/graphs!! amazing work
This is amazing to see. Would you be willing to publish your test results to show different materials and print settings?
Those motors are beefy af with a badass setup
I will like/comment on every video because you deserve to be recognized by the algorithm
I clicked because I was going to say something about the saw cut on your shafts, but now I have to compliment your video.
That was well done, and your communication was high quality.
As a personel gear lover, I love this video
GLaDOS - "You must really love to test."
Excellent videos. Thank you for sharing.
You sir are a genius to this ol' guy. Will do my best to help your algorithm. Thank you for sharing your details.
Excessive overkill? I see excessive run out. Something that must be eliminated before any usable data can be generated.
besides bevels being the worst gear of all, hardest to manufacture, hardest to make accurate, with the correct form... without a generator.
sure, FDM works for experimenting but yeah... run out here just defeat the purpose. was that backlash test performed on every tooth?
why mitre gears? with a hunting ratio? would want non-hunting for even wear. only saving grace of miter gears is you dont have to alter the setup of the gleasson to make the pair...
i made a 105T FDM gear for my lathe years ago. it worked, but if you actually tried running a nut up the freshly cut M1 thread? (yes, it was missing the most important and most expensive to replace gear)
the thread gets skewed. any inaccuracy in the gear train, leadscrew has nonconstant velocity, and the resultant pitch it cuts is not consistent.
they look good, theres only a tiny little subtle bit of warping, the typical "grung grung" sound of unmatched teeth... and that transfers down to a pitch that may vary by 0.02mm on every revolution... i aint got the tools for measuring pitch accuracy!
i finally machined a real one up with a dividing head, and yeah... that fixed that. to the tolerance of the machine, anyway. not precision but no worse than any other lathe i guess... .
Thanks!
Are the test results publicly available?
cyclic loading might be better for the wear test
Very Professional. Collaborate with Stuff Made here / This Old Tony.
It seems printing with 100% infill for the foundation of the root of the tooth would shift your failure mechanic to the next weakest point.
What's this for?
It would probably be more accurate if you use a friction model and identify the viscous and Coloumb friction coefficients as viscous friction depends on the angular velocity.
It calibrates the friction at various speeds then interpolates between them for the compensation
Cool’