Hi Art,
This isn't a bug but I could use some help on the use of "Shaper Control" This setting defaults to 12 and goes from there down to 2. My understanding is that this setting affects the shape of the tooth that is calculated and put out in gCode.
If I am getting a rounded tooth (see photo) where do I go with changing the setting in the shaper control? Or am I even adjusting the proper variable. Currently I just pick a new number and see what happens and if that doesn't work then a different number - not very "professional".
The little fidget shown in the photo does what its supposed to (nothing practical) in terms of meshing but it does have some rounded teeth. I have the gear point distance set at 0.05 and that gives me a nice smooth CNC movement but maybe its this variable I need to fuss with.
Hopefully you can shed some light.
John
Shaper Control
Re: Shaper Control
Are the teeth on the rack supposed to change like that?
Re: Shaper Control
John:
No, the teeth shouldnt look like that on the rack, they shoudl be all the same. Did they look like that in the DXF? If not,
your cnc is messing on somewhere , looseness or backlash.. if the DXF does look like that or the rack on the project screen
looks like that, please post the project fiel so I can duplicate your numbers and find the bug. :0
Thx
Art
No, the teeth shouldnt look like that on the rack, they shoudl be all the same. Did they look like that in the DXF? If not,
your cnc is messing on somewhere , looseness or backlash.. if the DXF does look like that or the rack on the project screen
looks like that, please post the project fiel so I can duplicate your numbers and find the bug. :0
Thx
Art
Re: Shaper Control
Hi John:
Nice work!, and congrats on being the firt to show us a working rack. :)
The number on the shaper is the number of teeth on the shaper. As you change it, take note the
picture of the shaper on the bottom of the screen has less teeth.
How to use that? Imagine a gear thats perfectly round, and your going to create it by rolling that round disk in a gear shaping machine,
which basically uses that shapper as a punch, as the disk rotates the "shaper" also rotates just as if it was the pinion, but every small change
in rotation of the main disk, the shaper gets again punched through the original disk.
This works well, but the number of teeth on the shaper has an effect which accumulates depending on the way the two gears mesh.
Imagine instead of a round disk, the disk has a concave section, like the modified ellipticals. If you were puching that "shaper" through
the disk all the way around, youd find if the shaper has too many teeth when you go to puch at the concave section of the disk it will
punch area's not involved with the current tooth being punched. Reducing the number of teeth on the shaper will make it smaller so in
the concave section it doesnt punch area's it shouldnt.
A similar effect is fond if the shaper has too few teeth. In a convex part of a gear it will punch a wider area thus rounding tooth off.
This is not an error though, just a setting. In reality, the number of teeth has both a good and bad effect on any noncircular gear and
its all a function just how noncircular that gear is.
Experiment with some of the higher modified coefficients and youll find that when the gear has the highest discontinuities around its
perimeter, you need less teeth on the shaper. You normally dont need any higher than 12 as it will give a nice involute result on a round disk,
but you definitely need fewer as the gear gets more and more out of round.
Its normal in a good eliptical for some teeth to look rounded off, this is because the pressure angle of the tooth, which is static in a round
gear changes as you go around a noncircular gear. The shaper control is why the GTA ellipticals are far superiour to the normal elliptical tab.
The general tab is done with osculating circle theory, which attempts to make each tooth as if it were on a round disk, so a 22 tooth elliptical
has 12 different teeth on it , all based on 11 different diameter circles. The GTA ellipticals though, would have 22 teeth, with each tooth
not only being different, but each side of each tooth is different as well to take into account the varying pressure angle. Those tooth sides
will be normal involutes ONLY on circular portions of a gear, and since no section of the ellipticals are truly round, none of them will actually
be perfect involutes, but instead a modified changing involute, a very complex curve near impossible to define mathmatically ( though I do have formula's
that can do that). But the point is the curve result on each tooth of a shaped noncircular gear is something non-intuitive, but ensured to be a better fit
than any other method of making one. So if you have rounded teeth, thats fine, just a matter of the meshing being a bit different.
What IS intuitive about this process is when to use a lower tooth count on the shaper, if you find non-smooth irregulaties in the toothed gear, or gaps
when the other gear doesnt mesh, the tooth count is too high and a radial concavity is clipping the shaper. Reduce the tooth count till it doesnt.
Having too low a tooth count will usually work fine, but the teeth will be more rounded off as the roots of the shaper have a wider effect, that
is to say the roots of the shaper cut more than just the tips of the teeth on either side, they cut into the actual tooth face.. again ,not an issue,
but the resultant gear fits a bit more loosely into its mesh. Using 2 teeth for example can very often make the gear look like an epicycloidal
as the tip rounding goes circular due to the true circular nature of the process.
I know this is hard to explain, but your doing great, playing with the kitchen in that way , even if guessing , is what the GTA was designed for, not
only do you get nicely meshing gears, but it leads to an intuitive sense of whats going on. Youll find yourself thinking automatically that you dont like a
result, even though it meshes and lowering the shaper makes it "look" better. In reality a designer needs an in depth understanding of cause and effect
to wrap their mind around what their going to do on a shaper machine, the GTA instills that depth of knowledge on a much lower level where you get
to understand it without any math study. ( which is always a prime design consideration whenever I add a new toy. )
Yell if you need help or advice, I really love to see the program used as you are doing
Nice work!, and congrats on being the firt to show us a working rack. :)
The number on the shaper is the number of teeth on the shaper. As you change it, take note the
picture of the shaper on the bottom of the screen has less teeth.
How to use that? Imagine a gear thats perfectly round, and your going to create it by rolling that round disk in a gear shaping machine,
which basically uses that shapper as a punch, as the disk rotates the "shaper" also rotates just as if it was the pinion, but every small change
in rotation of the main disk, the shaper gets again punched through the original disk.
This works well, but the number of teeth on the shaper has an effect which accumulates depending on the way the two gears mesh.
Imagine instead of a round disk, the disk has a concave section, like the modified ellipticals. If you were puching that "shaper" through
the disk all the way around, youd find if the shaper has too many teeth when you go to puch at the concave section of the disk it will
punch area's not involved with the current tooth being punched. Reducing the number of teeth on the shaper will make it smaller so in
the concave section it doesnt punch area's it shouldnt.
A similar effect is fond if the shaper has too few teeth. In a convex part of a gear it will punch a wider area thus rounding tooth off.
This is not an error though, just a setting. In reality, the number of teeth has both a good and bad effect on any noncircular gear and
its all a function just how noncircular that gear is.
Experiment with some of the higher modified coefficients and youll find that when the gear has the highest discontinuities around its
perimeter, you need less teeth on the shaper. You normally dont need any higher than 12 as it will give a nice involute result on a round disk,
but you definitely need fewer as the gear gets more and more out of round.
Its normal in a good eliptical for some teeth to look rounded off, this is because the pressure angle of the tooth, which is static in a round
gear changes as you go around a noncircular gear. The shaper control is why the GTA ellipticals are far superiour to the normal elliptical tab.
The general tab is done with osculating circle theory, which attempts to make each tooth as if it were on a round disk, so a 22 tooth elliptical
has 12 different teeth on it , all based on 11 different diameter circles. The GTA ellipticals though, would have 22 teeth, with each tooth
not only being different, but each side of each tooth is different as well to take into account the varying pressure angle. Those tooth sides
will be normal involutes ONLY on circular portions of a gear, and since no section of the ellipticals are truly round, none of them will actually
be perfect involutes, but instead a modified changing involute, a very complex curve near impossible to define mathmatically ( though I do have formula's
that can do that). But the point is the curve result on each tooth of a shaped noncircular gear is something non-intuitive, but ensured to be a better fit
than any other method of making one. So if you have rounded teeth, thats fine, just a matter of the meshing being a bit different.
What IS intuitive about this process is when to use a lower tooth count on the shaper, if you find non-smooth irregulaties in the toothed gear, or gaps
when the other gear doesnt mesh, the tooth count is too high and a radial concavity is clipping the shaper. Reduce the tooth count till it doesnt.
Having too low a tooth count will usually work fine, but the teeth will be more rounded off as the roots of the shaper have a wider effect, that
is to say the roots of the shaper cut more than just the tips of the teeth on either side, they cut into the actual tooth face.. again ,not an issue,
but the resultant gear fits a bit more loosely into its mesh. Using 2 teeth for example can very often make the gear look like an epicycloidal
as the tip rounding goes circular due to the true circular nature of the process.
I know this is hard to explain, but your doing great, playing with the kitchen in that way , even if guessing , is what the GTA was designed for, not
only do you get nicely meshing gears, but it leads to an intuitive sense of whats going on. Youll find yourself thinking automatically that you dont like a
result, even though it meshes and lowering the shaper makes it "look" better. In reality a designer needs an in depth understanding of cause and effect
to wrap their mind around what their going to do on a shaper machine, the GTA instills that depth of knowledge on a much lower level where you get
to understand it without any math study. ( which is always a prime design consideration whenever I add a new toy. )
Yell if you need help or advice, I really love to see the program used as you are doing
Re: Shaper Control
Hi Art,
I didn't save the original file that produced that rack - Sorry. When I went to reproduce it - I didn't think I changed anything but I must have because the next two racks came out perfectly.
Thanks for info. on the shaper that really helps, although it will take some time to soak in.
One more little item - In the Output manager there is a check box for "rounded corners". Mine is off but I would like to know when to use it.
Thanks.
I didn't save the original file that produced that rack - Sorry. When I went to reproduce it - I didn't think I changed anything but I must have because the next two racks came out perfectly.
Thanks for info. on the shaper that really helps, although it will take some time to soak in.
One more little item - In the Output manager there is a check box for "rounded corners". Mine is off but I would like to know when to use it.
Thanks.
- Attachments
-
1% inspiration 99% try, try again
Re: Shaper Control
Oops forgot something else.
The tabs don't appear to have any clearance and that makes them really tight on assembly.
John
The tabs don't appear to have any clearance and that makes them really tight on assembly.
John
1% inspiration 99% try, try again
Re: Shaper Control
Hi John:
>> In the Output manager there is a check box for "rounded corners".
That check tells the output Gcode algorithm that its OK to concatenate small lien segments into curves where
it finds it appropriate. It also asks the routine to make the outside corners round in the toolpath ( still making sharp corners, but with a rounded path ). This doesnt always take effect, it depends highly on the bit used and the code. Its basically an optimisation, so feel free to use it, its unlikley to affect you badly and may help speed things up a bit in the cut.
Oh, the tabs ARE a bit tight, but thats by design, better to sand a bit than try to add material to make a
tabbed box fit tightly.. They are in fact zero clearance items...:)
Art
Art
>> In the Output manager there is a check box for "rounded corners".
That check tells the output Gcode algorithm that its OK to concatenate small lien segments into curves where
it finds it appropriate. It also asks the routine to make the outside corners round in the toolpath ( still making sharp corners, but with a rounded path ). This doesnt always take effect, it depends highly on the bit used and the code. Its basically an optimisation, so feel free to use it, its unlikley to affect you badly and may help speed things up a bit in the cut.
Oh, the tabs ARE a bit tight, but thats by design, better to sand a bit than try to add material to make a
tabbed box fit tightly.. They are in fact zero clearance items...:)
Art
Art
Who is online
Users browsing this forum: No registered users and 8 guests