mirror of
https://github.com/ajisaacs/OpenNest.git
synced 2026-11-17 23:27:40 -05:00
docs: drop internal release-gate and job references from operator docs
Common-line cutting and the CI Fiber output docs called the machine check a release gate and named the shop's machine and coupon; they now say the single cut is not yet proven on a machine and how to check it. An engine example no longer names an internal job, support-tab settings name the Cutting Parameters dialog, the Avalonia colour doc mentions its Options scheme row, and the post doc says replan. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
1 parent
d4103a3b9f
commit
16ba05afec
6 files changed
+18
-18
No files matched your search
@@ -90,7 +90,7 @@ order, member by member, each member's holes first:
|
||||
- **Shared cut** (`CommonLine` motions, programs planned before the owner cut its own edge): the
|
||||
cut-off form with a real lead-in, no `G41`/`G42`.
|
||||
|
||||
**Unconfirmed on the machine** (release gate, see [common-line cutting](common-line-cutting.md)):
|
||||
**Not yet confirmed on a machine** (see [common-line cutting](common-line-cutting.md)):
|
||||
|
||||
- compensation start-up on a remainder whose lead-in starts in void (the kerf the owner's cut
|
||||
opened) and runs along it, and on the owner's lead-in collinear with its shared edge;
|
||||
@@ -98,7 +98,7 @@ order, member by member, each member's holes first:
|
||||
uncompensated, since the post writes no `G40` (the macros cancel compensation);
|
||||
- whether `L4` runs correctly before an older program's uncompensated shared cut, as for cut-offs.
|
||||
|
||||
Verify on a 3/16 stainless coupon pair on the CI FIBER 4020 and record the result here.
|
||||
Verify these on a test pair before production use.
|
||||
|
||||
## Configuration and boundaries
|
||||
|
||||
|
||||
@@ -288,7 +288,7 @@ The post cannot write a [common line cut once](common-line-cutting.md#cutting-ea
|
||||
it compensates every feature from its own winding and, in `KerfMode.PreApplied`, offsets no
|
||||
geometry, so a shared cut would be compensated into a part and a remainder could not be offset on
|
||||
its perimeter's side. It refuses such a nest before writing
|
||||
anything. Set the plate to cut its common lines twice and plan cutting again to post it here.
|
||||
anything. Set the plate to cut its common lines twice and replan to post it here.
|
||||
|
||||
### Kerf compensation
|
||||
|
||||
|
||||
@@ -5,11 +5,11 @@ one cut: programs are nominal geometry and the controller offsets each contour b
|
||||
(`G41`/`G42`), so both parts' facing edges are cut on the same centerline. Placing parts this way
|
||||
saves the part spacing along every shared edge, which matters on costly plate.
|
||||
|
||||
**Release gate.** Cutting a common line once is not yet confirmed on a machine. Before release it
|
||||
needs a 3/16 stainless coupon pair on the CI FIBER 4020: measure both parts' shared edges against
|
||||
nominal and check that compensation starts cleanly in the opened kerf and cancels where a
|
||||
remainder ends in it (see [CI Fiber output](cincinnati-ci-fiber-post-output.md#common-lines)).
|
||||
Until then, **Plate > Arrange > Cut Common Lines Twice** keeps the double cut.
|
||||
**Not yet proven on a machine.** Cutting a common line once has not been checked on a real cut.
|
||||
Before relying on it, cut a test pair: measure both parts' shared edges against nominal and check
|
||||
that compensation starts cleanly in the opened kerf and cancels where a remainder ends in it (see
|
||||
[CI Fiber output](cincinnati-ci-fiber-post-output.md#common-lines)). **Plate > Arrange > Cut
|
||||
Common Lines Twice** keeps the double cut.
|
||||
|
||||
## Groups
|
||||
|
||||
|
||||
@@ -59,16 +59,16 @@ valid at box contact.
|
||||
#### Squeezing the last sheet
|
||||
|
||||
The fit rules pack into the whole sheet, so when the last sheet has room to spare they spread the
|
||||
remaining parts out and rarely find the shortest strip that holds them. On N26-2A5 (14 door stops
|
||||
and 14 lock plates on a clamped 26 x 26 sheet) every rule stopped 8.0 high, while the hand layout
|
||||
remaining parts out and rarely find the shortest strip that holds them. On one real job (14 door
|
||||
stops and 14 lock plates on a clamped 26 x 26 sheet) every rule stopped 8.0 high, while the hand layout
|
||||
holds all 28 under 7.5 and leaves a half-inch-deeper drop. So when some rule places all the
|
||||
remaining demand on a sheet, Rectangles packs that demand again into a bin cut shorter, and
|
||||
separately narrower, running every rule and mode at each size. It keeps the tightest complete
|
||||
layout along each axis that beats the whole-sheet ones, and the stock loop then picks among all of
|
||||
them by cost and offcut as before (`SheetSqueeze`).
|
||||
|
||||
Whether a greedy packing fits is far from monotonic in the bin size (N26-2A5 packs at 7.5 but not
|
||||
at 7.62 or 7.74), so the sizes probed are stack heights, not halves of a range: sums of the parts'
|
||||
Whether a greedy packing fits is far from monotonic in the bin size (that job packs at 7.5 but
|
||||
not at 7.62 or 7.74), so the sizes probed are stack heights, not halves of a range: sums of the parts'
|
||||
spaced sides along the axis, starting at the bin edge or at the far side of a clamp the layout
|
||||
reaches. The probes walk down from the best extent so far, spread over the probes left, and a probe
|
||||
that fails never raises the floor. With more than 2048 stack heights the walk steps through the
|
||||
|
||||
@@ -75,8 +75,8 @@ read `PartColorPalette.Current`, so editing a scheme's array later changes nothi
|
||||
it is applied again. WinForms draws with per-thread GDI+ pens from `GdiSchemeTools`; the Avalonia
|
||||
plate canvas draws with Skia colours from `SkiaSchemePalette`. The Avalonia app
|
||||
applies the scheme saved as `ActiveColorScheme` in the shared user-settings file
|
||||
when it starts, and recolors a nest's drawings when it opens one; it has no
|
||||
Options screen to change the scheme yet. Its plate canvas builds part paths from
|
||||
when it starts, and recolors a nest's drawings when it opens one; its Options
|
||||
dialog changes the scheme ("Color scheme:"). Its plate canvas builds part paths from
|
||||
`OpenNest.Desktop.Rendering` (a port of the WinForms `GraphicsHelper` layer routing:
|
||||
etches stroked apart from cuts, tabs filled but not cut, break cuts never filled), so
|
||||
the etch, cut and fill rules above hold there too; WinForms still draws with
|
||||
|
||||
@@ -82,7 +82,7 @@ Cutting inspects a proposal, the overlay shows the proposal's parts on the same
|
||||
|
||||
## Support tabs in Plan Cutting
|
||||
|
||||
Cutting Settings > Tabs > "Add tabs to hold parts in the sheet" sets
|
||||
Cutting Parameters > Tabs > "Add tabs to hold parts in the sheet" sets
|
||||
`CuttingParameters.SupportTabs`, and its two choices say which parts need a tab. Like every
|
||||
cutting setting they are kept per machine thickness (Tools > Machine Configuration > Tabs &
|
||||
Pierce), so thick plate can go untabbed; see [machine cutting defaults](machine-cutting-defaults.md):
|
||||
@@ -122,7 +122,7 @@ never on a reflex or cusp vertex). Ranking and lead checks are unchanged.
|
||||
## Parts the tab alone holds
|
||||
|
||||
A tabbed part stays in the sheet on its tab, and a part short enough does not need the bed at
|
||||
all: the tab holds it wherever it goes. Cutting Settings > Tabs > "Tab alone holds parts up to"
|
||||
all: the tab holds it wherever it goes. Cutting Parameters > Tabs > "Tab alone holds parts up to"
|
||||
sets `CuttingParameters.TabHeldMaxLength`: on a plate with a support bed, a part the bed cannot
|
||||
hold level whose length (the long side of its outline's minimum-area bounding rectangle,
|
||||
`SupportAnalyzer.Length`) is at most that, bound included, is still tabbed, but
|
||||
@@ -140,7 +140,7 @@ does not read cutting settings and still shows it red, since it would tip withou
|
||||
|
||||
## Tabbing by size
|
||||
|
||||
Cutting Settings > Tabs > "By size, or with no support bed, tab parts whose narrow side is" sets
|
||||
Cutting Parameters > Tabs > "By size, or with no support bed, tab parts whose narrow side is" sets
|
||||
`CuttingParameters.TabSizeMinWidth` to `TabSizeMaxWidth` (stored under their earlier JSON names
|
||||
`noBedTabMinWidth` and `noBedTabMaxWidth`, so saved settings and nests read unchanged). Support
|
||||
tabs go by size under "Tab parts by size", and under the support grid choice on a plate with no
|
||||
@@ -204,7 +204,7 @@ the plate.
|
||||
## Slug slices
|
||||
|
||||
A cutout's slug drops onto the bed too, and one the bed cannot hold level can tip up into the
|
||||
head's path. Cutting Settings > Cutouts > "Slice slugs the support grid can't hold level" sets
|
||||
head's path. Cutting Parameters > Cutouts > "Slice slugs the support grid can't hold level" sets
|
||||
`CuttingParameters.SliceTippingSlugs` (off by default, kept per machine thickness like every
|
||||
cutting setting, and written to settings and nest files only when on). With it on and a support
|
||||
bed, Plan Cutting cuts each such slug into pieces that hold no contact before it cuts the cutout,
|
||||
|
||||
Reference in new issue
Block a user