wild and crazy gear teeth 2

[Windward]

Is it possible? Meaning, would it work?

In the attached file, the white parts are standard spur gears. I used 20 tooth, 20DP, 25 degree PA gears, but the gear size does not matter.

Then I extended the lands, as shown in black. These extensions are curved, but with a smaller radius than the tooth faces where they meet. They are not separate parts, just cut that way into the gear blank. I am assuming that this smaller radius would prevent any interference with normal operation of the otherwise standard involute teeth.

These extensions use one-half of the normal clearance, just an arbitrary choice to get started. So there is still some clearance. I am not worried about it if nobody else is. I suppose the gears could be designed with normal clearance, resulting in a slightly increased center distance, which might weaken the teeth a little, but that would be acceptable if it is necessary.

If this design would work, I also suppose that the problem would be cutting the gears. I don't know enough about it. I hope to get some comments and advice from the experts. Wouldn't these extensions increase the tip strength a little?

I am thinking that this design would allow easier meshing of the gears, not by sliding them together axially as in a car transmission, but by a radial approach. That is, in the same plane, with decreasing center distance until they meet and then mesh.

If the gears are not turning as they approach, and have normal lands which are about flat, they will sometimes be prevented from meshing because the lands meet. That stops the approach and the gears can't mesh. It helps if the gears are turning, the slower the better, but still the lands can meet and prevent meshing. Then it is just "grinding the gears" until the lands eventually separate enough to allow meshing.

That can take some time, because friction between the lands tends to hold the gears in the same orientation, and they turn together as though they were in mesh. If there is not much load on the gears, they will just rotate together without meshing until something good happens. If the gears are not turning in this situation, you are stuck. They will never mesh.

Now it will be clear why I would like to have these round land extensions. They should allow the teeth to slide by each other almost every time the gears are to be brought into mesh, instead of jamming way too often with flat lands.

A typical synchro-mesh would not work here, because the meshing direction is radial, not axial. If there is synchro-mesh for the radial direction, I have never seen one. If there is one, it would have to be extremely simple for my application, cheap and of very low volume and weight and high efficiency.

OK, it is a bicycle gearbox. If anyone gets any ideas, I can say with confidence that there is still a long way to go.

 

[spigor]

Without going into detail (unless somebody wants it) - I have found that using a 20 degrees tool is not practical for this application, because of insufficient tooth height.
It turns out that standard 20DP 30PA gears should be just enough to meet your requirements.
The 18 teeth has almost no undercut without any addendum modification.
Based on your comments regarding quality and the type of application, I assumed no addendum modification is needed at this prototype stage, because if that was used the driven gear would not be the same as the driver gear with the same teeth number.
The worst case scenario of a full rounded tool tip (since we don't know the exact parameters of the tool) gives a rounding coefficient of .12, which at .112WD leaves a good margin to exclude the root intereference.

Here are the specs for the 18T gear:
20DP 30PA .112WD
18T
X=0 /no addendum modification/
Major dia.:1.000
Minor dia.: .776
Number of teeth spanned: 4
Span measurement: .5180
Circular top land width: .0143

These are nominal parameters, the tolerances are to be chosen as tight as possible but reasonable, taking the following remarks into account:
Decreasing the major dia. will increase the top land width and reduce the contact ratio, I propose -.005
Minor dia.: max value showed
Span measurements: for a general purpose gearing a tolerance would be in the range of -.0017/-.0029
Circular top land width: tolerances for the span decrease also that

Here are the specs for the 40T gear:
20DP 30PA .112WD
40T
X=0 /no addendum modification/
Major dia.:2.100
Minor dia.:1.876
Number of teeth spanned: 7
Span measurement: .9773
Circular top land width: .0177
Span measurements: for a general purpose gearing a tolerance would be in the range of -.0024/-.0039

A 18T/40T gear set like that would have center distance of A:1.4500
as all such gear sets created by a substraction and addition of the same number of teeth, e.g. 22T/36T

To calculate the nominal span measurement you could use

http://www.meshingwithgears.us/cgi-bin/

http://wwwboard/span.cgi

If you need to change the number of teeth spanned, add or substract m*Pi*cos(a)=.13603495231756633879455586932316
Possibly you could use this:

https://khkgears.net/new/gear_calculator.html

/I've never tried it, but looks interesting/

GROB rolling: the material has to flow up the flank, could be problematic with such narrow top land, but GROB can have a solution to that. GROB rolling produces very quality and strong gears, QC6 (old AGMA 10) and is able to roll material up to 1200 MPa, which is around 36 HRc, so just as needed. The rolls are form tools (so you need multiple sets for a teeth number range) and are not cheap, but very productive, might be an option for a high volume production if other technical reasons won't exclude it.

I've done my share in a bicycle gearbox design and manufacturing and you seem to underestimate the requirements of the gears that you need. I think anything worse than QC8 gears won't do the job, but hey, this is just my opinion and you decide.

Now you get the gears specified, you can ask for quotes and choose what's best.

If you choose for gear hobbing, here's where you could get the tool from (I think they also have a rental program):

http://ashgear.com/pdfs/hs30.pdf

Good luck!

 

[spigor]

I forgot: the top land width is not a part of the specification, I've calculated it for your reference.
You need to specify:

20DP 30PA .112WD
Number of teeth
X=0 /no addendum modification/
Major dia. w/tolerance
Minor dia. max
Number of teeth spanned
Span measurement w/tolerance
Quality class

 

[mfgenggear]

@spigor
30 degree full addendum gears are more difficult to manufacture because the tool or grinding wheel apexes, being there done that,
when customer specify these types of gears I have to change the tooling hobs or grinding wheels to a different base pitch in order to grind.
we can agree and disagree but that has been my experience, I don't under stand your logic 20 & 25 degree doesn't work for this type of application.
please state with facts why it doesn't work. there are many gear design out there that do.
20 or 25 degree will give more land on the tips. and that is the reason for using profile shift is to improve the lands on the tips.
in addition recess gears (profile shifting) is to change the line of action, and to make the pinions stronger for tooth bending,
and to change the line of action for wear.
In the USA the preferred method for measurement is Measurement over and between wires. I have actually done both as I am also a trained Fabricator and precision inspector.
I would rather use inspection over wires. when it is possible. when gears get to large it does become impractical. this not the case here.
remember gears that are ground require stock on the gear teeth profiles, and compounds the issues above. normally .005 stock per surface cause
the tools to apex even more causing more issues with manufacturing. this cause the root to apex.
the object of a gear box is to transfer smooth, non vibration ( Total Composite error) and able to handle the torque, and stress.
the object is to have error free, good surface finish gears. with no interference.

 

[Windward]

I will have to take some time to study the latest posts. It is a lot of information from spigor and an interesting reply from mfgenggear, all much appreciated. Spigor, I hope you will say more about your experience with bicycle gearboxes.

If the quality of the gears I am proposing is your only doubt, I would certainly use whatever quality is required. I am wondering whether you think my gears might be too small in any case. Compared to my design, Pinion and Effigear use larger gears at lower RPMs. I think there is no question that the same loads could be transmitted with smaller gears at higher RPMs. I do not believe that the higher RPMs I am using would produce unacceptable noise, vibration and harshness. It has not been a problem in my operating model, but I don't have a prototype yet and no road tests. My gears would weigh less and I suppose they would cost less. Why would Pinion and Effigear use such low RPMs requiring larger gears?

My arrangement has a lower Q factor than Pinion's hefty 6.85 inches for all units. Effigear does not list their Q factor, but their gear arrangement is similar to Pinion's so I suppose their Q factor is about the same. I would call it a dimension, not a factor, but maybe something was lost in translation.

Both Pinion and Effigear require a special frame. My unit would use existing frames and bottom brackets without modification and would bolt on without difficulty. I believe this is an important advantage. It would allow most existing bicycles to be easily converted. My unit could also be designed for a special frame, thus taking up a smaller volume than a conversion would.

Both Pinion and Effigear shift their gears with cams located inside one of the axles. These special axles must be a significant part of the cost of the unit. My method is simpler.

Probably no one here is much interested in hearing why I think I have a better way. But if you will indulge me a little, spigor, I would like to know what you think about these things. It would be especially interesting to learn more about your experience in this field.

 

[spigor]

@mfgenggear
I agree that using the 30PA tools presents problems that you have pointed out. Knowing about that problems I tried to find a way to avoid using the 30PA tooling by replacing it with the 20PA tooling. I remind that one of the requirements in this application is top land width .01 to .02. I was led to believe that standard tooling is preffered by the OP to keep the costs down, at least for the prototype stage. A 20DP 30PA standard tooth proportions profile meets the top land width requirement.
By my calculations I have found that a standard 20PA deep depth tool would cut a profile with correct top land width, but if two such gears are meshed, there would be a collision as not enough root clearance is created by the 20PA tool. I tried to avoid that collision by truncating the tip diameters and confirmed that it can't be done, because the top land width requirement is not met /and also the contact ratio drops/. So, I stated that using a 20PA standard tooling is not practical for this application.

The recess action gears - I have written in my post why I propose not to use that at least for the prototypes. I was led to believe that the OP wants to keep the costs down, and with the recess action an 18T driven gear would be different than an 18T driver gear.

My approach for the serial production would probably be 20PA tooling with special pitch and tooth proportions.

There are some good calculators for the MOW out there, so there is no problem to use that.

I hope you would investigate this and the earlier posts to find enough proof to conclude if I have defended my case, or provide evidence, preferably in form of calculation results as these can be followed, showing that I'm wrong.

@Windward
I want to see the result of the discussion above first, before I reply.

 

[mfgenggear]

@Spigor
not saying you are wrong,
check out these pics of the pinion gear train. epicycle, helical gear teeth & spur gears of different DP
these guys worked on this for 8 years

https://biciclub.com/cambios-internos-pinion-una-nueva-era/6

I double checked your data for the 18T 20DP 30PA
and it will work, but getting touchy.
chordal tip .0120-.0109
.0279-.0288 tooth space at @ .780 diameter
.096 wires
M.O.W. 1.0448-1.0433, CTT = .0765-.0755
span meas. over 4 T = .5163-.5154

18T 20DP 25PA for ref
chordal Tips~ .0226-.0215
chordal Tooth Space @ .817 diameter .0487-.0496
.096 wire
M.O.W 1.0449-1.0432 CTT= .0765-.0755

as to not give out to Much info on this board
if interested @windward & @spigor

contact me at

https://rockytopusa.com/

 

[spigor]

I know the Pinion gearbox and think it is very interesting, especially the way they used a planetary to drive the gear selector shaft so it rotates unisono with the gear shaft and is able to change its relative position to it to select another gear.
The minus tolerance of the major diameter will increase the chordal tip, .0120 at 1.0 becomes .0159 at .995

So Windward has now lots of information what are the pros and cons of the 30PA or 25PA designs.

Here's my share:

https://youtu.be/KEIpydvlmF8

I propose that we take this opportunity to set up a rule, how much information may be given out for free on this board, so it ain't ruining somebody's business. I ask you for your input on this one. Thanks.

 

[Windward]

spigor, your nuseti bicycle is a fine invention. Excellent job. Most interesting to me is the gearbox with sixteen speeds, easiest possible shifting, extremely high torque capacity yet amazingly compact. My hat is off to you for the gearbox alone. I am disappointed that we can't see how it works.

It is a mystery to me why your bicycle has yet to make its mark. The only explanation I can think of is the cost.

Thanks for the valuable information you and mfgenggear have generously provided here.

 

[spigor]

Windward,
you can see it here:

https://patents.google.com/patent/EP2567888A1

Please note, that I was responsible only for the technical design and manufacture of the gearbox, everything else, including but not limited to functional and optical design and marketing, was made by the Project Owner. He used to have a vision and energy to make it all happen, great guy. I liked that project very much, but the market obviously though otherwise. But I can tell you it was a hell of a pleasure to ride that bike. If you pushed the lever a little it would instantly shift one gear up or down, if you pushed all the way it would shift two gears, very easy, like it encouraged you to shift. In fact the shifting was pleasant on its own.
I wish you good luck with your design.

 

[mfgenggear]

Very well done Spigor on the gear box, and thank you Windward
that is a nice bike , I am to puzzled.