[Mike Everman]

Close it? Naw, this is a very active development thead that quite nicely stays on topic!

[georgebarr]

bemfarmer,

This is regarding your spreadsheet for the hypo and epi cycloidal gears. If using the same R, r, N, and E variables, are the resulting calculations for the hypo and epi cycloidal gears considered matching gears? Are they to be used as a set?

Also, how accurate are these numbers and calculations. I would like to know before I CNC mill some aluminum. Have others actually CNC milled parts and tested them using these calculations/spreadsheet?

How is the gear wear. Are the two gears rubbing each other which creates wear over time or are they rolling against each other.

Some suggestions. You should combine both the hypo and epi cycloidal calculations in one spreadsheet so that they both use the same R, r, N, and E variables. Also, I modified my version to have an increment of 0.001 for the t value. This will provide more accuracy when I import the points into SolidWorks.

Thanks for the spreadsheet. This will make my work much easier. I am trying to make a 4th axes and using cycloidal gears for the backlash free reduction at 100:1 using 2 stages.

Also, in case there are any spreadsheet updates, which web site will you post those updates? Do you have a personal web site?

Thanks,

[georgebarr]

I did just figure out how to draw the lobes correctly so that they always are in contact! You have to adjust the diameter of the valley on the rotor with the eccentricity.

Yep i know how to do the eccentric shaft in a lathe.. But i don't have access to a lathe for the moment

Zoidberg, in post #72 of this thread and your 3 photos of the cycloidal gears, what are is the diameter of Y (pin rollers of lobe), eccentrity that was used, and diameter of X (of the outter gear)?. I think you posted a blueprint that has some of these X, Y, and eccentricity formulas. I am trying to fix and make changes to a spreadsheet that I downloaded from this thread.

Thanks,

[bemfarmer]

bemfarmer,

This is regarding your spreadsheet for the hypo and epi cycloidal gears. If using the same R, r, N, and E variables, are the resulting calculations for the hypo and epi cycloidal gears considered matching gears? Are they to be used as a set?

Also, how accurate are these numbers and calculations. I would like to know before I CNC mill some aluminum. Have others actually CNC milled parts and tested them using these calculations/spreadsheet?

How is the gear wear. Are the two gears rubbing each other which creates wear over time or are they rolling against each other.

Some suggestions. You should combine both the hypo and epi cycloidal calculations in one spreadsheet so that they both use the same R, r, N, and E variables. Also, I modified my version to have an increment of 0.001 for the t value. This will provide more accuracy when I import the points into SolidWorks.

Thanks for the spreadsheet. This will make my work much easier. I am trying to make a 4th axes and using cycloidal gears for the backlash free reduction at 100:1 using 2 stages.

Also, in case there are any spreadsheet updates, which web site will you post those updates? Do you have a personal web site?

Thanks,

Georgebarr, From the Joong-Ho Shin paper, the "hypo" and "epi" profile are NOT a matched set. There are no corrections to the lobe peaks and valleys, as there are in the zincboy Python generator.
It's been a hobby project, untested, I have not cut anything. Did manage to draw up a gearset in Alibre, as a learning project.
For "hypo" cycloid profile, the matching interior "ring gear" with cylindrical rollers, can be built in CAD using the N, R and r. (For a completely different reducer, with inner "epi" profile, a "ring gear" with cylindrical rollers is on the outside.)
(There are two papers by R. Dolchinkov et. al. (free on web,) which do have a
matched set of epi and hypo, with no cylinders. But I have been unable to complete an excel spreadsheet with all of the gargantuan formulas.(so far). The 2 papers have typos with subscripts, plus vs minus, and sin cos swaps.) So after a while, it gets real confusing. Which are the typo's, and which are the correct formulas? (The math is pretty rough to follow.)
The "starik/astarik" Russian website posts (of the zincboy formulas) are very good, with lots of google translates, and following the links.
Not sure about the backlash for homemade cycloid reducers, unless a person uses the very nice harmonic drives?
There some interesting "ball cycloidal reducers" as well.
Sorry, no personal web page.

[wildwestpat]

Withdrawn as my post was out of sequence

[georgebarr]

Below is a picture of the Hypocycloidal gears I made via a spreadhseet and Solidworks. I found a document on this thread (I think) that states that the 2 matching gears should have the same eccentricity and same diameter for the lobes and the gear diameter are incremented/decremented by a factor of the eccentricity, and other info. This seems to be the key to make a matching cycloidal gear set. I modified bemfarmer's spreadsheet to take the above into account. I will try to upload my spreadsheet version in 1 week. I still have to add both gears data points to the spreadsheet.

Is there a CAD to simulate the movement of these gears to verify they will work? Either Solidworks cannot do that or have not figured out how to do this.

[JerryFlyGuy]

George, Solidworks should be able to check these for you. If you use a physical interferance test and leave the stop on coincidence out of the equation you should be able to rotate the gears and see them mesh [granted they are not both fixed or mated such that they cannot turn] If you have the premium version of solidworks you could add a motion study allowing them to rotate and then even input that into an FEA study and check for stresses on the meshing faces.

fwiw

JFG

[bemfarmer]

Enclosed is an Excel sheet with "Generalized Cycloid Drive Equations," for
generating "Epi" and "Hypo" cycloidal profiles in Alibre, by importing points and generating splines. They are NOT a matched set, (Unless someone can
make the "hills equal the valleys"(?)

[bemfarmer]

After a great deal of study, have come to the conclusion that the Dolchinkov/Nikolov/Nikolov/Latinovic equations for Hypocycloidal and Epicycloidal gear profiles, in "Synthesis of Hypocycloidal Gears" and
"Meshing Characteristices of Hypocycloidal Gears", (both free on web via google), are the SAME as the equations in "Gear geometry of cycloid drives" by Chen, Fang, LI, and Wang. This should not be suprising, as the same general mathematic methods (equations of meshing, etc.) are use by both.
The disquise is the relation between Rz and Zb.
Center Distance equals eccentricity. e = m/2(1-x)
λ= (1-x) = K1 = +/- eZb/Rx(Zb-Zg).
Since Dolchinkov has derived the formulas to transform the hypocycloidal ring gear into a MATCHING internal "epi-trochoidal" profile, all that is left to do to replace the internal pinion with cylindrical teeth, is to correct the typos, and add the internal profile to the spreadsheet. (My next project.)

[kenbarls]

Enclosed is an Excel sheet with "Generalized Cycloid Drive Equations," for
generating "Epi" and "Hypo" cycloidal profiles in Alibre, by importing points and generating splines. They are NOT a matched set, (Unless someone can
make the "hills equal the valleys"(?)

More post like this mate. :wave:

[Herbertkabi]

Hereby concept of Russian Hybrid car found out from morning news,
Turn attention to this planetary gear,
perhaps some ideas for backlash free rotary motion
[nomedia="http://www.youtube.com/watch?v=XY-ABZjVoeU"]YouTube - ё-мотор[/nomedia]
cheers,
herbert

[Mike Everman]

If nothing else, that's the most impressive animation I've seen.

[BillTodd]

If nothing else, that's the most impressive animation I've seen.

Very impressive animation...

But, how else are they going to persuade investors to plough millions of Rubles into an old, failed and rehashed technology?

Ignoring the sealing problems, the pursuing piston toroidal engine is just a piston engine; it offers something in the way of reduced friction but little in the way of improved breathing . Instead it concentrates the waste heat into a smaller space (I suspect that the pistons rely greatly on the incoming charge for cooling - thus decreasing efficiency further).


I wonder if the big chunky gears have more to do with avoiding aliasing in the animation than backlash (BTW the cycloidal thing in the front is the oil pump).

Bill
[edit] coincidently , I've just come across this pursuing piston toroidal with a very clever Cardan joint drive (these guys are really very creative)
http://www.pattakon.com/pattakonRotary.htm

[Mike Everman]

Like the Segway or the dirigible, it's a "yes, but why?". Just because it can be done differently...oops, ok, I'm really guilty of that kind of thinking!

This one has more of a sealing problem than the wankel, and that failed in production. Grinding perfect toroidal grooves in mating parts makes my head hurt, and the sealing of all that makes these pistons transmit their force through the chamber wall seems like some hand waving. While an interesting problem, it's not going in the right direction, of simplicity.

Here, I need to put my head on the floor, reach through my legs and scratch my backside..

[handlewanker]

Hi all, I saw this design in the Popular Mechanics back in the late 50's, only then it had the "pistons" shaped like square vanes in a drum like chamber, and had the arms driving the vanes on the outside by two geared together cranks from the central shaft, same principle of two pistons chasing each other without having to reverse direction to exhaust or induct etc.

Never heard anymore about it, so now the patent has expired it's probably being "invented" again, but this time with the complication of round pistons with a curved skirt length.

BTW, on the subject of backlash free motion, I was examining my Macson hand surface grinder down feed mechanism the other night to see what had gone wrong for the previous owner who sold it after a catastrophic event when the grinding head did a sudden down plunge and threw the job and half the grinding wheel across the shop.

I'm about to commission the grinder, after having many other irons in the fire, and now it's the grinder's turn to see the light of day.

The point is, the wheel head slide has a spring mechanism that counteracts the weight of the wheel head in an upward direction, and applies a force against the down feed screw so keeping the job against the screw flanks going up or down.

At some time or other the cable end, attaching the spring mechanism to the slide, had become detached from the wheel head, and only the slight stiffness of the wheelhead slide held it in place.

Having no bias to counteract the weight of the wheel head the slide dropped down in the middle of a grind, dug into the job and the lot went into orbit.....one dismayed guy, but my luck as I bought it cheap.

It will be a simple job to reconnect the spring mechanism and remake the cable end, but now I think I'll reinvent the mechanism and use a counterweight on the end of a longer cable through the column and through a hole in the table, about 10 KG weight, which will give a constant bias no matter if the slide is at the top or bottom of it's travel.

The fact that the slide is under constant reverse force to the screw feed is a typical backlash solution.

This means any slide going forward needs a driving force to push it back against the forward motion of the screw, and on reversing the motion needs to follow the screw as it retreats from the driving face.

A stepper motor running in the opposite direction with just enough power to maintain screw flank reverse face contact would do the job OK I would think, which is the same as my surface grinder wheelhead spring/counterweight compensation.

The problem with rotary tables is the rotation has no end, so a spring loaded compensator would not work, but the constant force from a stepper reverse drive would be constanly on.
Ian.

[Herbertkabi]

£ ÍÏÔÏÒ :: ÷Ù ÓÍÏÔÒÉÔÅ ËÁÎÁÌ: anon70707 :: :: ÷ÉÄÅÏ ÎÁ RuTube -
Сергей Стиллавин - "�-мобиль": первые ощущени� от реального прототипа
Hmm ... I have no idea how they will to write Ё-мобиль by Latin alphabet,
I mean cyrillic Ё ... like Yo
cheers,
herbert

[PEU]

Will machine a new cam next week

Back into the reducers thread

The new cam never materialized and that piece was abandoned, now I would like to try again but in steel and the parts precision wire cut. I have a couple of questions:

1) what would be the ideal material, amutit-S (AISI O1)? AISI D3? other?
2) should I made pieces size on size or must use some tolerance?

Regarding 2) if tolerance is needed how much backslash it introduces?

Anyone tried wirecutting these cycloids before? results?


Pablo

[aystarik]

here is my hub. milled with a single set-up for accuracy...

[georgebarr]

Hello,

I have my version drawn up using a MS Excel formula calculations and converted to a CAD model. I plan to cut these gears using the waterjet method. My question is, what is the tolerance for these gears (e.g. 0.005")? This information is what I need to give to the waterjet service company?

Thanks,

[PEU]

Waterjet can't keep the needed tolerances for these parts I think.