GearHeads Corner
October 25, 2020, 05:47:11 PM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Search Login Register  
Pages: 1 2 [3]
  Print  
Author Topic: Bevel Gear Stock Prep  (Read 15368 times)
0 Members and 1 Guest are viewing this topic.
ArtF
Administrator
Hero Member
*****
Posts: 5663



View Profile
« Reply #30 on: April 04, 2011, 01:16:15 PM »

Mark:

 No problem. When I finished bevels I came to understan d exactly why home machinist s consider
them a true bitch to do. Even using technique s such as we're attemptin g just the setup is a bitch. So I also expect debugging and full testing to take a very long time. Feel no pressure at all,
just let me know when you find something works or doesnt. I appreciat e all such input.

Thx
Art
Logged
markwoodard41
Newbie
*
Posts: 25


View Profile
« Reply #31 on: April 05, 2011, 08:23:39 AM »

Art,  I just couldn't leave this alone.  I am quite intrigued .  I resized the blank to the thickness specified on the project screen and recut the gear.  I maintaine d the face angle etc.  I did have to make an adjustmen t to the X=0 position.  I initially zeroed as proscribe d but still didn't get a cut completel y through.  So I adjusted the X such that the cutter overhung the stock by the diameter of the cutter.  In other words, I touched off of the back bevel and set the X DRO to be -0.123 (diameter of my cutter)  I had observed that the gcode was moving off the rear of the piece -0.123 so I took advantage of that clearance and moved it to the X+ side.  Below is a pic of the results.

You can see that the cutter still doesn't make a clean exit in the X+ direction .  Also note the negative involute shaping.  There is a step there that is not evident in the positive involute shaped face.

I did create and cut the mating gear but didn't do a real accurate job creating the blank nor zeroing the Y axis.  I got in a bit of a hurry.  I got a gear but it also doesn't have teeth cut through the entier face and shows the same shaping character istics. 

I think I will hold off on more testing until you can advise.  My guess is that if you can provide a more accurate DFX file, I can stop trying to manually adjust the thickness . It just adds one more variable for you to consider.  I can only guess at the effort required on your part to code something like this so let me know how I can help further.

Mark


* DSC_0307.JPG (463.07 KB, 2304x1536 - viewed 518 times.)

* DSC_0308.JPG (400.94 KB, 2304x1536 - viewed 531 times.)
Logged
ArtF
Administrator
Hero Member
*****
Posts: 5663



View Profile
« Reply #32 on: April 05, 2011, 08:44:38 AM »

HI Mark:

 Yes, those look pretty bad.. ( amazing that minr run so nice.. )

  Ill schedule a bevel update and look at both the dxf's and the generatio n code to see what may be wrong.. Bevels are truly a bitch to debug and diagnose. .
Ive learned to be very wary of them. Ill let you know when Im ready for more testing..

Art
Logged
markwoodard41
Newbie
*
Posts: 25


View Profile
« Reply #33 on: April 05, 2011, 09:59:09 AM »

Thanks Art.  I have a friend who wants me to check out the GT timing pulley function.  I'll pull the DFX into a CAD appl, create a solid and craft some code for a 4th axis implement ation.  Wish me luck.

Mark
Logged
ArtF
Administrator
Hero Member
*****
Posts: 5663



View Profile
« Reply #34 on: April 05, 2011, 02:11:02 PM »

Mike:

  So far, Ive verified the drawing in the DXF is correct. The GearThick ness on the projects page is really just the face width when tilted down as seen from the X axis..

  I need to consider finding a good way to simulate this beast..th at'd make it easier to see where Im screwing up.
Art
 
Logged
ArtF
Administrator
Hero Member
*****
Posts: 5663



View Profile
« Reply #35 on: April 05, 2011, 02:47:42 PM »

Mark:

  I see the issue. Its going to take a full rewrite of the Bevel code to fix. Another bad assumptio n on my part.
The shape of a tooth changes as you rotate it upwards to the bevel plane. A guy making one on a gleason for example,
knows to aplpy various formulas to correct the tooth shape.. BUT, in actual fact as I create the bevel for display, I am
bending the tooth upwards and therby changing its shape. When I do tap code though, I didnt take this new shape into account
properly.
   So the displayed bevel is pretty good , the tap output is wrong..

  Good news is, I see the error.. bad news is it will take me a while to fix as it requires new code.. the old wont do for this new shape..
requires a few more formulars to compute the differenc es.. Ill let you know when Im ready.

Thx
Art
Logged
markwoodard41
Newbie
*
Posts: 25


View Profile
« Reply #36 on: April 05, 2011, 07:34:06 PM »

Glad to see that you found something afterall.  That is good news although it does cause you work. I'll stand down until I see an alert from you to download a fix. 

I did cut two timing pulley on my A axis using GM DFX file and a CAD/CAM appl.  I cut a GT5 and an XL timing pulley using maple wood blanks for starters.  They look pretty good in appearanc e.  Too soon to tell how well the contours fit a belt.  I did wrap an XL belt around the wooden model.  I felt just the barest perceptib le slop when I held the belt and tried to rotate the pulley.  I think I will compare that to a commercia l pulley to see if I am being truly objective in my observati ons.  I just don't have one handy without taking my machines apart.  Not something I am going to tackle tonight.

Good luck and thanks again.
Mark
Logged
markwoodard41
Newbie
*
Posts: 25


View Profile
« Reply #37 on: April 19, 2011, 08:05:41 AM »

Art,  I know you are focused on your new laser (pun intended) but I was curious if you have had any chance to look further into the bevel gear problem.   I am building a drill spindle for my CNC machine to complemen t my high speed router spindle and was thinking using beveled gears to achieve the gear reduction and orientate the drill motor convenien tly at right angles to the spindle axis.

Let me know if I should eliminate bevel gears from considera tion for the near term.

Mark
Logged
ArtF
Administrator
Hero Member
*****
Posts: 5663



View Profile
« Reply #38 on: April 19, 2011, 08:32:13 AM »

Hi Mark:

  Id eliminate it near term.. Im still puzzling over where I went wrong. I have found one problem in my theory that isnt too hard to correct,
and thats the root clearance .. Im convined I screwed up there. The root is clearing within an angular spacve that is equal to the virtual gear thats being beveled.. BUT when the teeth are tilted to the bevel, the diamter isnt the same as the vurtual gear.. so the angle of the root is wrong. I can correct this, BUT Im still a bit stumped as that would actually mean a slightly larger root clearance .. and your gears looked like they needed a slightly less root clearance . Im also a bit confused as to why mine worked well, even the root looked good. Of course I do mine different ly.
When I create a blank..Im lazy.. I create a foam cone of the addendum angle.. then I find a point on the cone where the diameter is equal to the outside dimension of the gear. I then cut the cone at that point and by the gear width ahead of that point.zer o my mill just behind that point and cut it... ( You were right, the tool wasnt meant to sit half over the tooth, it was meant to start directly behind it..)

 Due to the difficult y in testing of bevel code, Im still doing thought excersize s to see where I may have messed up. As we discovere d
early on, even discussin g the subject leads to mental confusion on how the angles interact in the creation of the blanks. I want to make sure I have a proper mental image when I tackle it again.. Im close..bu t still a bit confused. I hope to redo it soon though so I can at
least put it to bed. GM is pretty far along, the spoking and such is much better, so Ill release this week a new version, and if not too many things bump up in priority due to bugs, then Ill start the rewrite. But I cant promise anything near term.. so I wouldnt want you going down an experimen tal road where what you need is a solution.



Art
 
Logged
markwoodard41
Newbie
*
Posts: 25


View Profile
« Reply #39 on: April 19, 2011, 10:08:34 AM »

Thanks for the quick response, Art.  There are other options that I can consider.  I just thought it would be neat to try bevels again on a real project.  I'll press on.  I may just try to modify my setup as you described but maybe not.  We'll see.

Mark
Logged
Pages: 1 2 [3]
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!