Standardized test shapes

Here's the place for discussion related to CAM/CNC and the development of the Path module.
RatonLaveur
Posts: 162
Joined: Wed Mar 27, 2019 10:45 am

Re: Standardized test shapes

Postby RatonLaveur » Sun Jun 02, 2019 4:26 pm

Hello everyone,

Given that we're opening the can of worm labled machine accuracy, here is a little bit of basic reading material i found interesting, https://www.google.com/url?sa=t&source= ... gIyScZQTYK

Perhaps it fits the conversations.

At any rate it feels to me like any or a subset of the shapes used for CAM testing will necessarily fit the bill of machine accuracy testing.

J.
freman
Posts: 375
Joined: Tue Nov 27, 2018 10:30 pm

Re: Standardized test shapes

Postby freman » Sun Jun 02, 2019 5:23 pm

Herbert:
Use only the "Export" Icon to export to g-code, the way File -> export don't work
Thanks for the detailed reply Herbert. Dog bone seems to have done that nicely but it throws out a few wide paths off to apparent infinity. I can no longer drag the object using scroll wheel.

Where do I find this icon? I hovered all the icons in Path WB and I don't see anything like that.
[EDIT] I think you are referring to the Post Process icon, which I was doing from the menu. This does NOT produce an output file, though it seems it should because if I do not configure output it prompts with a dialogue.
Select the wall of the hole and use the Pocket OP
Well that's exactly what I'd expect to do but it did not work.

Code: Select all

Pocketing Select Mode
Traceback (most recent call last):
  File "/svn/freecad-build/Mod/Path/PathScripts/PathOpGui.py", line 170, in updateData
    self.panel.updateData(obj, prop)
  File "/svn/freecad-build/Mod/Path/PathScripts/PathOpGui.py", line 961, in updateData
    page.pageUpdateData(obj, prop)
  File "/svn/freecad-build/Mod/Path/PathScripts/PathOpGui.py", line 242, in pageUpdateData
    self.updateData(obj, prop)
  File "/svn/freecad-build/Mod/Path/PathScripts/PathPocketShapeGui.py", line 325, in updateData
    self.setExtensions(obj.Proxy.getExtensions(obj))
  File "/svn/freecad-build/Mod/Path/PathScripts/PathPocketShapeGui.py", line 302, in setExtensions
    baseItem.appendRow(self.createItemForBaseModel(base[0], sub, edges, extensions))
  File "/svn/freecad-build/Mod/Path/PathScripts/PathPocketShapeGui.py", line 244, in createItemForBaseModel
    ext0 = _Extension(self.obj, base, sub, label)
  File "/svn/freecad-build/Mod/Path/PathScripts/PathPocketShapeGui.py", line 70, in __init__
    self.switch = self.createExtensionSoSwitch(self.ext)
  File "/svn/freecad-build/Mod/Path/PathScripts/PathPocketShapeGui.py", line 82, in createExtensionSoSwitch
    wire =  ext.getWire()
  File "/svn/freecad-build/Mod/Path/PathScripts/PathPocketShape.py", line 199, in getWire
    return self.extendEdge(feature, edges[0], direction)
  File "/svn/freecad-build/Mod/Path/PathScripts/PathPocketShape.py", line 161, in extendEdge
    return extendWire(feature, Part.Wire([e0]), self.length.Value)
  File "/svn/freecad-build/Mod/Path/PathScripts/PathPocketShape.py", line 121, in extendWire
    l0 = (ePts[0] - endPts[0]).Length
<type 'exceptions.IndexError'>: list index out of range
Traceback (most recent call last):
  File "/svn/freecad-build/Mod/Path/PathScripts/PathOpGui.py", line 98, in setEdit
    self.setupTaskPanel(TaskPanel(vobj.Object, self.deleteObjectsOnReject(), page, selection))
  File "/svn/freecad-build/Mod/Path/PathScripts/PathOpGui.py", line 109, in setupTaskPanel
    panel.setupUi()
  File "/svn/freecad-build/Mod/Path/PathScripts/PathOpGui.py", line 953, in setupUi
    self.panelSetFields()
  File "/svn/freecad-build/Mod/Path/PathScripts/PathOpGui.py", line 925, in panelSetFields
    page.pageSetFields()
  File "/svn/freecad-build/Mod/Path/PathScripts/PathOpGui.py", line 224, in pageSetFields
    self.setFields(self.obj)
  File "/svn/freecad-build/Mod/Path/PathScripts/PathPocketShapeGui.py", line 215, in setFields
    self.setExtensions(self.extensions)
  File "/svn/freecad-build/Mod/Path/PathScripts/PathPocketShapeGui.py", line 302, in setExtensions
    baseItem.appendRow(self.createItemForBaseModel(base[0], sub, edges, extensions))
  File "/svn/freecad-build/Mod/Path/PathScripts/PathPocketShapeGui.py", line 244, in createItemForBaseModel
    ext0 = _Extension(self.obj, base, sub, label)
  File "/svn/freecad-build/Mod/Path/PathScripts/PathPocketShapeGui.py", line 70, in __init__
    self.switch = self.createExtensionSoSwitch(self.ext)
  File "/svn/freecad-build/Mod/Path/PathScripts/PathPocketShapeGui.py", line 82, in createExtensionSoSwitch
    wire =  ext.getWire()
  File "/svn/freecad-build/Mod/Path/PathScripts/PathPocketShape.py", line 199, in getWire
    return self.extendEdge(feature, edges[0], direction)
  File "/svn/freecad-build/Mod/Path/PathScripts/PathPocketShape.py", line 161, in extendEdge
    return extendWire(feature, Part.Wire([e0]), self.length.Value)
  File "/svn/freecad-build/Mod/Path/PathScripts/PathPocketShape.py", line 121, in extendWire
    l0 = (ePts[0] - endPts[0]).Length
<type 'exceptions.IndexError'>: list index out of range
ControlSingleton::showDialog: Can't show  Gui::TaskCSysDragger  since there is already an active task dialog 
If I change your T4 in any of the paths it breaks them and even reverting to T4 does not clear it. If I change even the speeds on your T4 it breaks !

I'm starting to realise this is too flaky to be of much use at this stage. :(
freman
Posts: 375
Joined: Tue Nov 27, 2018 10:30 pm

Re: Standardized test shapes

Postby freman » Mon Jun 03, 2019 7:37 pm

OK, I'd better dump this here in case it goes tits-up again.

This is a working model with paths. I reduced the step height to 2mm to test on ply wood before risking metal. If your parametric CAD is in order this should be simple enough to scale back to 10mm later. ( I won't be betting on the paths no falling apart completely though ). :D :D
Attachments
diamond-circ-sqr-2clean.FCStd
(77.88 KiB) Downloaded 7 times
chrisb
Posts: 16861
Joined: Tue Mar 17, 2015 9:14 am

Re: Standardized test shapes

Postby chrisb » Mon Jun 03, 2019 10:01 pm

freman wrote:
Mon Jun 03, 2019 7:37 pm
:D :D
It's not completely clear to me: did you finally succeed to create GCodes?
freman
Posts: 375
Joined: Tue Nov 27, 2018 10:30 pm

Re: Standardized test shapes

Postby freman » Tue Jun 04, 2019 7:09 am

Yes, having re-asserted the tool speeds, I have what looks like useable GRBL Gcode. I have not cut the part yet because it was throwing errrors but I suspect that is Universal Gcode Sender messing about. [ Confirmed there was a 50 char line length set in UGS, not sure why. Set to 127 fine. ]
freman
Posts: 375
Joined: Tue Nov 27, 2018 10:30 pm

Re: Standardized test shapes

Postby freman » Tue Jun 04, 2019 9:24 am

OK a clean run produces quite credible results. A few notable defects:

1. Big gouge from central hole before face-milling the top 'diamond' face. Why didn't it lift clear before moving ?
2. Remaining material at edge of top face and circular face. Am I testing my h/w or the s/w : too many variables.
3. Remaining fluff in centre of one of circular segments. Again h/w or s/w ?
4. Slight machining marks on top surface due to compromise in choosing path. A spiral cut produces a nice finish on wood but fails to work at 45 deg.

First impression is that it's pretty good but maybe still too experimental for daily use.

Unless the gouge is me not using FreeCAD properly, I can't afford that to happen when cutting metal !
Not much help cutting a test piece to test the machine when I don't know whether defects come from the CAD.

Can anyone comment on that gouge?

Thanks.

PS not that different path tools were used intentionally to cut the circular segments to compare results. Some better than others.
Attachments
diamond-circ-sqr-3.FCStd
(78.9 KiB) Downloaded 5 times
test_piece.png
test_piece.png (901.76 KiB) Viewed 505 times
herbk
Posts: 1401
Joined: Mon Nov 03, 2014 3:45 pm
Location: Windsbach, Bavarya (Germany)

Re: Standardized test shapes

Postby herbk » Tue Jun 04, 2019 11:34 am

Hi freman.
if i use "keep tool down" i have had some errors like that at the past, not actuall, - but i don't use keep tool down very often.
Look at the gcode you get, last move of each operation should be a move to "Z save hight", the gauge looks like it's missing at the end of the Op for the hole.

The left material looks a bit like caused by backlash

The fluff im in mind its caused by a wrong tool for wood and a to low spindle speed (wood wants rpm's) ;)
Gruß Herbert
User avatar
sliptonic
Posts: 1549
Joined: Tue Oct 25, 2011 10:46 pm

Re: Standardized test shapes

Postby sliptonic » Tue Jun 04, 2019 12:57 pm

You have material left at the top probably because your stock was configured larger than your part and you didn't add mill face operation to clear down to it.

The center gouge is troubling. Are your clearance and safe heights high enough?
freman
Posts: 375
Joined: Tue Nov 27, 2018 10:30 pm

Re: Standardized test shapes

Postby freman » Tue Jun 04, 2019 1:09 pm

thanks for the comments.

"left hand " =? , it if is the outside edge of the base contour it's not a problem that's just the cut out path, the outside is gash material. What concerned me more was the edge of the cylinder which has a fine slither of material not removed.

400Hz = 24000 rmp tool speed , HSS four flute end mill. More a tool for metal I admit.
The center gouge is troubling. Are your clearance and safe heights high enough?
I had a look at the paths and the face mill op for the top is the last operation. It had "use start point" checked. Not sure why? I maybe because that was initially the first operation. I gave some more thought to cutting order and more or less reverse the op. order.

Could be more of a gotcha than a bug.
freman
Posts: 375
Joined: Tue Nov 27, 2018 10:30 pm

Re: Standardized test shapes

Postby freman » Tue Jun 04, 2019 4:42 pm

I tried increasing the level of each slab from 2mm to 3mm. I went fairly well and the model of the part body remains coherent.

However, some of the paths don't work. Some do. I went through the offending paths to reset the depths to the new positions but there is still residual material on the four corners and the circular segments.

The sim shows about the same thing as I see on real cut.

I have tried to totally remove the offending paths, and recreate them from scratch. I still seem to get the same problem. It seems that there is still some hangover from the 2mm version of the model.
Attachments
diamond-circ-sqr-2-resid.FCStd
(1000.01 KiB) Downloaded 8 times