45 degree worm M2.5

[ArtF]

Phil:


  Set your final depth as a positive number.. In 4th it translates to "final Thickness" and a negative number had an unknown effect.. (Ill fix that..).
Had me wondering why it worked here.. till I saw the final depth.. Now Im wondering if that affected other cuts, it basically caused a mirror of the path on my system...

Art

[philbur]

Then I'm confused. What has final depth got to do with the blank thickness, final depth should be fixed at the distance the root is below the top of the blank (the z = 0 point). Why does the CNC processing redefine the blank thickness This is confusion total confusion!!!!!!

Phil :)

[ArtF]

Phil:
 
  Not so confusing when you consider its history. We started with 2.5D only. Now consider that a person designs a gear with a facewidth of 5mm.
So they go to cut it in 2.5D. Its unlikley they will use 5mm as their depth of cut, the actual thickness considered is taken from the object, but is overridable by "Max Depth.". This allows a person with that 5mm thick gear to cut it from 12mm material or whwtever they wish. So far so good, now he can cut 5mm or 12 mm or whatever, by using the "Max Depth." field.

  Now consider the 4th axis. "Max Depth" has the exact same function, it allows for a thicker cut. Its unfortunate that the term to a gear designer
refers to the depth from tip to root, but then that's not a setting that would be usefull at all, who wants to cut a tooth to a variable depth, a tooth
is always cut to its design depth or it simply isnt usefull. But having the "Max Depth" work  in 2.5D and in 4th axis having the same effect made sense.

  As Ive said this was all done and explained at the time as a holder, a better system than GM's, that would be redone in  grander scale
when all else in GT caught up.  Its simply a handy way of making the gear thicker without having to go back and redesign it. I did fail
however to catch negative numbers.....

  Most of the settings will become more intuitive when its recoded, but "Max Depth" is simply a thickness modifier. You cannot, nor I suspect would you want to control the depth of a tooth in the toolpath generator.

Art

[philbur]

Of course, I had assumed that total depth related to the cut depth in 2.5D and had nothing to do with the 4th axis. It's difficult for the user to guess that you also use total depth in 4th axis to vary the gear thickness to avoid having to go back to the design stage. Seems to violate the whole principle of what you design is what you get. Total confusion for the user comes from throwing all parameters in one box  then later using a parameter for another total unrelated, none intuitively function. Clearance seems to be another example. This is all very frustrating.

I've been using negative DOC since the start but I don't recall ever see a blank that was wider than the cut or have gcode that had an x move less than the blank width ( I would remember that) until today. I see that the my same helical input today doesn't generate the same gcode as yesterday with respect to the x axis travel?

Phil  :(

[philbur]

Art, here's the M2.5, 4 tooth spur

Phil :)

[ArtF]

Hi Phil:

  The paths HAVE changed for sure over the course of this. When I have my fingers in there I tend to change things I see that arent right.
The negative DOC wouldnt have affected you till I fixed it yesterday. Till then only the spurs were affected by DOC. SO it likely didnt affect
anything you did up till now anyway.
  The X axis travel should have stayed the same... at least if the DOC is set to the same width you had set for the helical yesterday.
The DOC should fill in with the designed thickness when you select the gear, its only an override if one wishes. ( Kinda handy to have though..)

  Ill run a few more checks. It wont be long before the tabs get separated and Ill rename variables them to stop any confusion and allow settings for
some of the stepping values.  Once it has its own tab the 4th wont have to share variables with the 2.5D

(Your photo didnt take :)... The forum is pretty picky about used names and such.. if you post a photo, name it something weird .)

Art

[philbur]

Regarding the missing photo. When I post with an illegal format attachment it rejects the whole post. When I correct the attachment and try to repost it says its a duplicate post and rejects it. So I copied the text, closed the site then reopen, pasted the text and posted it to see if it would be rejected before going to all the trouble of adding the attachments. It accepted the post so then I modified the post by adding the correct attachments. I tell you all this in the hope that you can change the site so it doesn't reject a repost when illegal attachments have been corrected.

I on a couple of occasions I have rearrange the text a bit but it's not fooled so easily. :'(

Phil :)

[philbur]

Art, I thought we pretty much had a solution in post #49 and were only trying to avoid the tool marks on the tan shaving by lowering the cutter a little bit?

Phil :)

[ArtF]

Oh I agree, #49 looked pretty good. And we figured raising the bit a bit woudl stop the problem..wish I knew what changed..

(Ill find it..obviously I messed up something in other fixes I was doing..) :)

Art

[philbur]

And we figured raising the bit a bit woudl stop the problem..wish I knew what changed..

I though lowering the tool was the solution. It puts the flank at the tangent point, not the tip?

Phil :)

[ArtF]

Phil:
 
  I was speaking of the rooting. Raising the rooting would have had the effect of stopping it from
touching the tooth profile. But yes, I did also lower the tangent point. Or to bne more specific , I stopped it from calculating the tangent height each pass, and locked it to  the base circle height.

  Really, other than stopping the rooting from going outside the balnk I did nothing else, yet
you went from a fairly good cut to a very very bad one. Im still analysing, but Im still confused.
Take that last spur cut, its asymetrical..that should be impossible unless the Y wasnt zeroed dead straight up form the center. I havent suggested that because I doubt you did that. So Im still plotting
the toolpath to see where it hits..so far, it doesnt, but the plotting isnt automated so it takes me awhile, Ill find it yet..

Art

[ArtF]

Phil:

  Im away most of today, but I have uploaded a new version. It relabels the depth and thickness overrides, and when you press the GetFromObject button you go back to the actual design thickness
and such of the gear. There is a new checkbox that turns off the cleanup section of the rooting,
this may have caused trouble with any undercuts.

  I still cannot find what is causing the variations, but it is warming fast here so it wont be long. So far each tool simulation I do shows no collision, but I have decided after all this to make a simulation
on screen so I can watch the damn thing cut in real time.. :)

Art

[philbur]

The spur was very small, 15mm OD, and I used the largest possible end mill 3mm, as this was the size I used for the helicals (for comparison). Maybe the cutter size and the relatively small number of root passes is the cause of the ugly spur?

Phil :)

[ArtF]

Phil:

  Could be.. when you make anything small errors get magnified considerably.
Hurts me head either way. I really need to write a true simulator so we could watch it
work and see where it fails..

Art

[philbur]

Art. having a square ended tool dancing around in a confined space trying to produce what should be a convex profile was always going to be tough ask. Surely a bullnose cutter would stand a much better chance and at the same time be able to do the tan shaving and have the potential to minimise tool markers. Basically this is effectively 3D milling, which is bullnose territory.

Phil :)