Abbott Analytical Products

Reference:

• LinuxCNC knowledge base library.
• FreeCAD user manual and its Path Workbench.
• OpenSCAD user manual, CheatSheet, and highly useful insight.
• pyCam documentation.
• Bezier
  • surfaces,
  • shape,and useful.
  • example.
• HeeksCAD/CAM/CNC.

Note: ** Some browsers do not allow links to these sites from their anchors.

Brass Credit Card:

Brief Summary of Project:
From 2022: (FreeCAD and LinuxCNC) A friend requested the fabrifcation of a credit card formatted nameplate for one of our mutual associates. Having gained experience since the 2018 Brass Nameplate project, FreeCAD was employed for design and generation of gcode. FreeCAD is an opensource CAD tool that allows the design of objects and subsequent gcode (program for executing the actual LinuxCNC milling of the object.

Image	Comment
	This image shows the final delivery of the Credi Card Style nameplate. The design was rendered using FreeCAD 18.4 running under Ubuntu 16.04 (64Bit). The FreeCAD design artifacts and final gcode programs are available for downloading form the GitHub link given below. https://github.com/trooker/credit_card_style_3line_nameplatete
Lower Level Detail
FreeCAD View	Snapshot From FreeCAD Design Image Discussion These images are of the "Engrave" task definition showing the "G" of the third line. In this instance the designer has elected to sequnetially "highlight" one of the 1324 endges required to define the three lines of charcters. The image also shows the "Engrave" task five depths that FreeCAD rendering of the design created. This rendering yielded the first gcode program for milling the Credit Card Three Line design.
	This image displays the LinuxCNC program driving the FreeCAD produced gcode on a Sherline 8580 NexGen CNC in a simulation mode. The engraving bit sat idle while the user observed the simulation run. Based upon the simulation run a decision was made to "modify" the original FreeCAD gcode. The modification employed various used defined variable for x-, y-, and z-axis speeds, a "safe height to which the bit was raised in the z-axis for relocation. The modification also consolidated the actual engrave loop for milling the three lines of text.
	First pass using the engraving bit used for projects in 2018 and 2019. This first attempt was rule unacceptable quality. That led to a search for a "tighter" engraving bit using the same edges. Introducing the new Dremel Engraver tool to the project required a new entry in the tooltable and minor changes to FreeCAD artifacts as well as the gcode.
Helpful Approaches	These modifications greatly improved the original gcode from FreeCAD
Flatness Check	A gcode program was written which verifies that the four corners of the brass stock are flat with material secured using the project hold-down fixturing. This project's flatness test reuses gcode fragemnts from the Fishing Lure lip checks.
Major Tweak to gcode	The original FreeCAD engrave process for the three lines of textrequired 6196 separate movements to make five separate levels of depth cuts. In other words the engraving bit would would follow the same path to create the three lines of text five separate times. The only difference would be the increasing depth to which the engraving bit would plunge. Using the LinuxCNC subroutines to control the program flow (the movement of the engraving bit) an engraving loop was created using "one pass" of the original FreeCAD gcode and a logic while less than to check to see which loop to do next or to break-out of the loop. This essentially allowed a "single step through a depth, halt, check". In that manner when the results appeared to be "satisfactory" the milling phases was terminated.
Results	The final approach that was found to be helpful during the actual milling of the Credit Card Style Nameplate was adding the same type of "single step through a depth, halt, check" approach was implemented for the Contour-Fillet-Tab to generate the appearance of the credit card style.

Brass Nameplate:

Image 2019: Second Brass Design Production demonstrating movement down learning curve.
In 2019 the usage of FreeCAD replaced pyCAM and HeekCNC. Two different engraving styles and bits were employed to allow client to select preferred style. Finaly, c-shaped adjustable hold-down brackets greatly improved fabrication quality. FreeCAD provided a "Tab" contouring feature that gave extra control during the final milling.

Brief Summary of Project:
From 2018: (OpenSCAD/PyCAM/HeeksCNC/LinuxCNC) A brass nameplate for use on the display of a model of the USS Constellation was milled. It was a simple three line affair with two drill holes of brass tacks set on a thin oval brass blank. It was a very real learning experience in homing, tool heights, work offsets, tabs, and Heeks CAD/CAM tool. There were multiple runs made in wood and scrap copper. But there was only one piece of scrap brass created. That run demonstrated the need for tabs as the holddown fixturing failed. The final run was a step process involving three tool changes. In order to do that at the skill level at that time, three separate NGC programs were sequentially run.

Image 2018: First Brass Design Prototype displaying hold-down failure and need for tabs.

From 2018: Discussion of Process Flow
The gcode shown below is ordered in the sequence used to mill the project. Note that that engrave and blank milling employed the usage of user defined variables and o-code flow control looping.

1. Engrave text
2. Center drill two holes along centerline of raw blank.
3. Milling blank in four passes with tabs.

Tool Table Entries

• T4 P4 Z1.627 ;1/8 flat end mill 3/8 holder
• T5 P5 Z0.9975 ;center drill#2 in collet
• T7 P7 Z0.7865 ;router engrave bit collet

Design to gCode Process Flow
1> Use OpenSCAD to create ellipse shape.
2> Using the 2D shape export it as a SVG image from OpenSCAD.
3> Invoke pyCAM and set units to inches.

Plexiglass Heart:

Discussion of Process Flow
The gcode shown below is ordered in the sequence developed/used to mill the project. Note that that engrave and blank milling employed the usage of user defined variables and o-code flow control looping. The Heart Blank goes one step further using scaling in the XY plane.

Heart Blank Design Supporting Artifacts

1. Spreadsheet rendering of the classic Heart Curve.
2. SVG file exported from Veusz based on classic Heart Curve.
3. HeeksCNC project file yielding gcode.

1. pyCAM produced Engrave Text.
2. HeeksCNC generated Milling Blank.

Tool Table Entries

• T7 P7 Z0.7865 ;router engrave bit collet
• T16 P16 Z2.27855 D0.1275 ;1/8 x 3/8 double 2 flute

Design to gCode Process Flow

Heart Component: A suitable set of equations representing the classic Heart Curve was found. A spreadsheet was used to assmble a graph of the outline of the heart. The data points of the outlined heart were convert to a *.csv format. The Veusz graphic tool was used to create an XY plot of the *.csv data. Once that was done the figure was exported as an *.svg file. The *.svg file was imported to HeeksCNC. As an *.svg the heart shape is basically a 2D figure. But before the heart shape could be used it had to have the extraneous Veusz introduced axis information removed. Once that was done processing continued. There is no extrusion necessary for a 2D figure, since the same effect can be achived in a Profile Operation. After setting the tooling and tabs gcode was generated. The "raw" gcode was modified to incorporate o-code program flow control. Scaling of the XY was achived using global replacement.

Engrave Text Component: Prior engrave experince with pyCAM led to using that tool for the engrave text. A pyCAM session was started. The default model deleted The the units was switched to "inches". The Edit bar menu item was sleected which resulted in "Engrave text" option becoming available. Selecting the "Engrave text" option yielded the Engrave text wizzard. Add "Love" and select Apply to create a new model, Text Love. Adjust the sizing as necessary. The 3D view should now display. At this point the development to gcode should follow the typical tools, process, bounds, task, toolpath simulate/export process to yield the appropriate gCode for the "Love" text.

Milling Time: Open a LinuxCNC session. Load and execute the Heart Blank gcode. Once that set of code finishes load the Engrave Text. Inspect the Engrave Text to insure it is placed as needed. Relocate/place the Engrave Text as needed using the Touch Off button process for X/Y axis for P1 G54. When satisfied execute.

Simple Wooden Fishing Lure:

Brief Summary of Project:
(OpenSCAD/HeeksCAD/CNC) Simple Fishing Lure: shape showing the space between the object and the walls of the pocket. The object rest on the plane-Sketch. The plane-Sketch surface will serve as the "tab/tag" holding surface for milling of the second side. Details are provided at this link.

Image 3: Prototype executed with scrap material.

Waterline Fishing Lure:

Brief Summary of Project:
(OpenSCAD/Blender/FreeCAD) The STL/mesh file was grabbed from the earlier HeeksCNC effort. It was imported and converted into shape/solid and eventually into a set of waterline NGC code. Waterline refers to the milling pattern that follows the contours of the shape being milled. A run of saw dust was made using this design to create the left side of the lure. The stock was then flipped and a second run executed. The results demonstrated the "feasibility" of the design. But it clearly exhibited a need for additional layers of waterline cuts and a need for closer control of the "flipped" origin. This original design was discarded because too much "manual" effort was required to reconcile the right/left side mis-match. Plus the shape was not an acceptable "lure".

Refinement of the waterline by increasing the granularity of the contour lines (depth of cut from -0.025 to -0.010) did not provide a suitable quality to justify moving forward with this approach.

Discussion of Process Flow

The same STL that was imported to the HeeksCNC lure project was reused. The import process steps for FreeCAD is the same as that used for the box/lid logo. Once the

The code shown below is ordered in the sequence used to mill the project.

1. The "starter" CAD file builds the STL
2. FreeCAD Waterline Fishing Lure design project.
3. Waterline operation yields the surface shape of the lure
4. Finishing operation removes material from around lure forming pocket

Tool Table Entries

• T16 P16 Z2.27855 D0.1275 ;1/8 x 3/8 double 2 flute

Collection of Helpful Images

• OpenSCAD hull() shape rendered with a sheet rectangle inserted.
• Waterline processing under LinuxCNC looking down from Z axis.
• Base plane under LinuxCNC looking down from Z axis.

Design to gCode Process Flow
This project reuses an existing OpenSCAD object. This object originally integrated a simple hull function in union with flat planar surface. Once the object was designed and rendered it was exported as a *.STL file. The file was imported into FreeCAD as a STL mesh entity. Sequentially that entity is converted to a shape, solid, refined shape, and finally a usable path shape. Once the path shape was defined a "Job" was created with an ocl waterline operation and a finishing operation. The finishing operation first used a pocket shape to remove stock material from region between the lure and a bounding box. The finishing operation completed its run by cutting an outline contour outline of the lure. Tabs were not used since the planar surface was to serve as a base/wall between side 1 and side 2. As originally planned side 1 would be milled first. Side 1 must also have index registration points to facilitate flipping and milling of side 2. This will require extending the registration points through the work piece to what will become side 2 milling surface. (Insure that the registration points do not mar the surface of the region being milled.)

Side 1:
The waterline operation and the finish operation would run two separate gcode programs under LinuxCNC-Axis. But first the finish job would be run without the stepper motors being engaged. Once this "virtual" mill job was completed LinuxCNC would display the tool path executed except indicating a red tool path . Then load the waterline. Us the "Touch Off" to jockey X P1 G4 until a suitable waterline aligns with the "finish" representation. Repeat for the Y P1 G54. Continue to jockey the waterline until a "best" balance appears. Try zooming/panning for better resolution and jockey. Once satisfied run the waterline gcode program. After the waterline program executes, load and run the finish program.

Side 2:
With Side 1 complete, flip the work piece/stock. Use the previously set index registration points of Side 1 set the physical location of the work piece. Start by loading/running the finish gcode without the stepper motors engaged. Using the completed finish tool path align the waterline gcode and execute with the stepper motors engaged. Then complete side 2 by executing the finish gcode.

Conclusion As originally conceived the lure was to appear as a symmetric hull figure. If all went well, side 1 and side 2 should appear symmetrically spaced when viewed axially along the Z-Y and/or Z-X planes. The difficulty in alligning the two sides precluded further movement down this avenue.

Complex Surfacing of Wooden Fishing Lure:

Brief Summary of Project:
(OpenSCAD/FreeCAD) A more nartual looking fishing lure has been designed using OpenSCAD with Bezier curve surfaces. The imported STL mesh file was found to be too complex for PC resources. As a result no usable FreeCad shape was produce.

(OpenSCAD/Blender/FreeCAD) The STL model design effort of the lure shape was imported to Blender. This work is in process.


Image 5: Complex Fishing Lure in OpenSCAD (Flat Nose)


Image 6: Complex Fishing Lure Imported to FreeCAD (Round Nose).

Discussion of Process Flow
Subsequently discovered OpenSCAD "hull() function" and Bezier curved surfacing.

Design Supporting Artifacts

• Design code combines lure component shapes.
• Creates the surface patch for the under-belly of lure.
• Renders a generalized lure shape/surface created.

Design to gCode Process Flow (OpenSCAD/FreeCAD) The generalized shape of a "crank" fishing lure was created using OpenSCAD. That shape was cut to create a more realistic lure. The two components were then assembled into the lure shape. That shape was then converted to an *.stl file for export. It was then imported itno FreeCAD where the complexity of the mesh shape became obvious. It proved to be impractical to proceed along this approach.

A number of alternate approaches to obtaining a complex solid surface shape are being explored.

Wooden Box with Lid:

Brief Summary of Project:
The intent was to create suitable wooden boxes for general gifting. The project originally started by looking at creating puzzle boxes with dovetail edges. But reality and CAD skill set collided. Once reality and complexity of FreeCAD's cubes and Path work bench began to manifest itself, a decision made to slowdown. Then began the effort to make presentable wooden boxes with pockets and lids with matching pads. The lids were to also have an appropriate logo/design artwork of minor complexity. That level of crafting was captured in the following project.

Discussion of Process Flow

• (FreeCad) The first run resulted in a usable box/lid. This run involved learning about the complexities of Sketches and setting constraints. The default Zig-Zag pattern produced overly tedious clean-up work/sanding of the pocket wall.
• (FreeCad) A second run has been designed to mill more material from the stock for the pocket/pad. This run employed the original Sketch from the earlier effort with resizing of the pocket/pad. The Offset milling pattern was employed. Simulation under LinuxCNC demonstrated a cleaner approach. The simulation of the milling showed an issue with the placement of the pocket and lid. Re-positioning of the pocket and lid within the virtual stock resolved that issue. By re-positioning the milling was simplified. It also supported the pocket-pad matching without having to use the mirror or to set other design features.
• (OpenSCAD/FreeCAD) A "radiant sun" image was designed in OpenSCAD as a 3D object and exported to FreeCAD as an STL mesh model. The import went well resulting in body, shape, solid, path shape, and pocket_shape.

Box/Lid Design Supporting Artifacts

• The supporting FreeCAD and OpenSCAD design project files for the box/lid are download-able as one zip file.
• The mis-directed dovetail design approach project files are available as a zip file.

gCode for Milling
The resulting NGC gcode from FreeCAD and associated tool table are shown below. The current sizing is for a 6x4x1.5 box. The pocket and pad are roughly 5x3. The pocket is set to 1.3 inches depth. The pad is set to .375 inches.

1. Box with pocket.
2. Lid with pad. The thickness of the lid is left to the miller's choice.
3. Radiant Sun pocket.

Tool Table Entries

• T3 P3 Z1.273 D0.09375 ; 3/32 end mill Short Sherline 7002 for "radiant sun" logo.
• T18 P18 Z4.459 D0.233 ;1/4 router deep pocket for pocket, pad.
• Large surface removing bit via manual operation of the mill.

Design to gCode Process Flow
(OpenSCAD) The original box/lid involved dovetail edges and a circular pocket as shown the box and lid CAD drawings. These drawings led to exportable *.stf files.

(OpenSCAD/HeeksCNC) Importing the two *.stl files from above produced the box_tab and the box_slot projects. Although their where reflective of the intended design objective, they appeared to be beyond the skill level available for the project.

(FreeCAD) As a good launch into FreeCAD a simpler "box/lid" with pocket/pad matching was initiated. As described and documented above gcode was produced directly from FreeCAD and saw dust created. After receiving a request for two more wooden box/lid this project was measured as a success.

Path Forward

• The box lid needs ot be tweaked just a bit to reduce the need for sanding the leading/lagging edges.
• The next generation of box pocket/lid will employ a dovetail fit to slide the lid open/close.
• Additional "logos" will be added ot the collection.

Evolving a Practical Wooden Fishing Lure:

Brief Summary of Project:
(FreeCAD) A nartual looking fishing lure can designed using the various workbenches of FreeCAD. This avoids the "resource hogging" associated with improting STL files of lures generated externally.

• Welling Lure Designshop
• Converting simple sketches(https://www.youtube.com/watch?v=7oOWvC80dNw) into complex solids
• Useful sketch and BSpline (https://owncloud.gwdg.de/index.php/s/HS2FD9s49wQs4fI) information/guidance.

Discussion of Desgin Process Flow
An original conceptual drawing of a fishlure must be manifested as a draft drawing. This "lure draft" serves as the "2D container". The defined "lure draft" provides the outline and serves as the boundary for the additional "shape defining layers". For this version of design four shape defining layers were generated. At this point the shape defining layers were jockeyed into appropriate locations to provided the required shape in 2D (X/Y Plane) and constrained (see Image 9 below). Using the following Best Practices the complexity of multiple sketches in a drawing surface is reduced.

1. Practice with the sketch features before attemtping a "for record" run.
2. Constrain as each sketch feature (lines, circles, BSpline, eclipses) is drawn.
3. If a sketch is not immediately needed, hide it.
4. Draw the structure of sketches in reverse order of the offset to be used like a pyramid. The smallest sketch, the circle, was assigned to the top, The largest sketch, the outer trace approximation of the Bezier shape was destined for the the bottom.

Code Artifacts for Milling
The resulting NGC gcode from FreeCAD and associated tool table are shown below. The current sizing is for a 6x4x1.5 box. The pocket and pad are roughly 5x3. The pocket is set to 1.3 inches depth. The pad is set to .375 inches.

• FreeCADC project file suitable for execution or downloading.
• Waterline gcode adjusted for milling lure shape.
• Finishing gcode for milling outline and removing material from pocket.

Design to gCode Process Flow With a set of gcode (Waterline and Finish) it was time to move to the mill. The Waterline gcode (with the "tweaks") ran without complaint the first time.




The Finish gcode required placement via Touch Off of X/Y G54 inorder to algin with the previously milled lure.



After making the touch off adjustment, the gcode ran without incident.




Observations
There were two issues observed with the Waterline gcode.

1. Moderate: The FreeCAD Waterline executed as a "topographical contour" cut rather than as a waterline/level cut. The expectation was that it would remove all the stock material within the basin of the lure shape for the Z depth value. Since the prototype lure mill was executed in soft wood the End Mill/Spindle had no significant difficulty. Workaround The work around is to return to the Waterline cGode and use the approach that has previousl been employed to assmeble a "composite" of the lure's Waterline toolpaths into one o-code loop. Then on each pass of the loop the step-down cut value would be decremented to meet the water line/level. This will increase to execute each exection time. But it will save the stress on the mill.


Original Observation/Issue of single mill cut per depth.




Post Adjustment to gCode using o-loop flow control and user defined variables.
For a level multiple toolpath cuts per depth remove material along "waterlevel" of work basin.






2. Minor: The Waterline gcode failed to "remove" material ("excess material) from the shape of the lure in three places. Workaround The excess materail was removed with little effort using tools on-hand in the shop. But it is risky.

1. Minor: The four Tabs failed to materialize as the outline contour executed. Workaround This was acceptable at this level of prototype. But it needs to be examined before moving to far forward. This aspect will become increasingly important as the lure moves into a left side/right side mirroring design/job.

Creating a Realistic Wooden Fishing Lure:

Brief Summary of Project:

This project, 016_fishlure, employed previously gained FreeCAD knowledge, experimentation with FreeCAD Draft and Parts workbenches, and new milling skills to produce a very realisitic shaped crank style lure.

Image 11: Realistic Wooden Lure ready for painting, hardware, and water trials.

Field Testing The lure created from the following process has been "pond tested".

Discussion of Process Flow
The FreeCAD project files, resulting gcode, and salient configuration files for the Sherline Next Generation 8580 with Rotary Table/Right Angle are given below in compressed file format.

1. FreeCAD project files for Lure and Finish Operations.
2. Compressed file collection of gcode for milling the lure objects.
3. Configuration files in compressed format to support milling the lure and finish operations.

Tool Table Entries

• T16 P16 Z2.321 D0.1275 ;1/8 x 3/8 double 2 flute

"stepper.var" Entries

Discussion of Design Process Flow

See the earlier crank lure for a preliminary discussion and first baby steps towards a realistic crank lure design.

Side-1/Side2: The initial steps in the design followed the same basic process as the earlier crank lure desgin for Side-1. The "tricky" point was creating a symmetrical matching object, Side 2, that could be milled. This side would need to be milled from the same work stock but with a different orientation. This "effect" was generated during the design phase by mirroring the Side-1 solid model. That sounded simple. But multiple passes at attempting to produce Side-2 proved to be a problem. The solution involved capturing the toolpath for Side-1 as the "model" of the solids surface. The generalized design process steps are given below:


Side-1

Design to gCode Process Flow With a collection of nine gcode files the project was moved to the mill. The Probe file was replaced with a level and a "scribe" point engraver to verify centerline on top. Then the work stock was flipped and the bottom centerline was verified. Verification of both top and bottom centerlines involved moving in the X axis along the centerline and "touching at several points. The combination of the 4-jaw chuck, right angle, level, centerline, and engraver/probe produced excellent results. Sequence of Milling Steps:

• Level, Setup Centerline, and Verify
• 089.01_01_016_970_fishlure_waterlineEye.ngc
• 089.02_07_016_495_fishlure_waterlineFinish.ngc
• 089.03_01_016_980_fishlure_flipper.ngc
• Shutdown LinuxCNC and Restart As Is
  With Side-2 "now" facing the end mill
• 089.04_01_016_970_fishlure_mirrorEye.ngc
• 089.05_05_016_495_fishlure_mirrorFinish.ngc
• 089.03_01_016_980_fishlure_flipper.ngc
• Shutdown LinuxCNC and Restart As Is
  With Side-1 "now" facing the end mill
• 089.07_04_016_999_fishlure_TabContourCutOut.ngc
• 089.08_05_016_999_fishlure_boxCutOut.ngc

Through-Wire Wooden Fishing Lure:

Brief Summary of Project:

This project, 021_fishlure, employed previously gained FreeCAD knowledge, experimentation with FreeCAD Parts.Revolve workbench.tool, incorporation of four flute endmills, and cable wrapping loops. The results, lure style 021.4, is a stronger, realistically shaped crank style lure. Fish actually took an interest in this prototype.

Pond testing, lure confirmed the need for a lip. Various lip designs were attempted using materials such as plexiglass and thin sheet metal. Finally a potential, feasible 3D printer design was discovered. However, that design was determined to be too difficult to insert/mate with the 021.4 style lure body. A modification of that 3D printer design and of the 021.4 style lure was employed. A discussion of the modifications to the lip's 3D design can be found as a "make" of the original lip design. The FreeCAD project 021 was modified to include a pocket cut on the bottom of the lure to support mating with the lip with a modest negative three degree drop. The documents for the 3D printer lip and the pocket cut are provided further below in this section.

Field Testing The 021.4 style lure has completed "pond tested".

Discussion of Process Flow
The FreeCAD project files, resulting gcode, and salient LinuxCNC support files for the Sherline Next Generation 8580 with Rotary Table/Right Angle are given below in seven part compressed file rar format. A high level view of contents of the compressed files is given directly below. Links in the next section lead to the downloadable seven compressed files.

1. A file called dirlist which provides a working manifest of the zip file.
2. Folder containing a more detailed a working checklist for moving through the milling process.
3. FreeCAD project files for Lure and Finish Operations.
4. The full collection of gcode for milling the lure objects.
5. A folder containing supporting program modules with images which require loading to the LinuxCNC machine folders.
6. The Cura 3D printer design project and gcode are available from the lucca82 Make Project at Thingiverse. That site also contains a discussion directly related to the 3D Printer aspect of the lip for the lure.

Download Actifcats

1. Part1
2. Part2
3. Part3
4. Part4
5. Part5
6. Part6
7. Part7

Tool Table Entries

• T21 P21 Z1.273 D0.045 ;dremel scribe engraver added 20190502
• T22 P22 Z4.013 D0.0625 ;dremel 1/16 drill
• T23 P22 Z4.013 D0.0725 ;dremel 3/32 drill
• T24 P24 Z2.264 D0.1275 ;1/8 x 3/8 4flute
• T25 P25 Z0.0 D0.25; 1/4 x 38 2 Flute End Mill Hand Powered

"stepper.var" Entries

Discussion of Design Process Flow

See the earlier crank lure and realistic lure for a discussion of the maturing steps in crank lure design. Instead of using a Side-1 and Side-2 (Right Half and Left Half) the perspective was changed to Dorsal {Top} Half and Belly {Bottom} Half. This change allows the compartmentalization of the design. The Dorsal Half has both eye buds and the primary shape of the lure. The Belly (Bottom} Half features include the Through Wire, the Belly Hook Loop Wire, and Alignment Pins Holes.


Dorsal Half/Belly Half:

The initial steps in the design differed from the earlier crank lure of Project 016. The Dorsal or Top Side was based upon an ellipse 2D shape. Using this shape as the outline, a continuous multipoint DraftWire 2D line was drafted from the left most point touching the x-axis to its right most point touching the x-axis. Using the Draft bench-revolve, the 3D Dorsal Half was created. With the Dorsal Half defined, it was then copied/pasted as the Belly Half. With the Belly Half shape now available, the Belly Half was modified from within the Draft workbench to generate the "desired" shape.

From that point forward the Dorsal Half and Belly Half were available for creation of Jobs and the associated *.ngc code. The Through Wire Processing was originally attempted as a separate CAD/CAM CNC gcode process. But common sense that it was better to create the Through Wire grooves, drill Belly Holes, Alignment Holes, "custom fit" craving of gaps for wire wraps manually, and adding a pocket cut to provide mating surfaces with the plastic lip. NOTE: For this early prototyping of the 021.4 style lure, the pocket cut was executed post-bandsaw separation because the need for the lip had not been confirmed. The difficulty in the alignment curved body lures presented with this deferred pocket cutting of the prototype strongly suggested that this pocket needs to be cut prior to the waterline surfacing while the centerline and factory edges of the woodstock can be employed.

A bandsaw was employed to separate the Dorsal Half and Belly Half from the workstock. The Dorsal Half and the Belly Half were then placed in a vise (with sacrifice wooden bumpers) to allow the drilling of the two alignment holes. The original centerlines (now significantly shortened) of the Dorsal Half were used for the drill operation setup.

Once the Alignment Pin Holes had been completed the two halfs were separated and then re-assembled. During the re-assembly, the Belly Half was now placed above the Dorsal Half such that the Alignment Pins could be inserted to allow exposure of the Belly Half's split face matching surface. With the Belly Half's matching surface exposed its centerline could be roughly verified between the head and tail points along the x-axis. The lures Through Wire groove was then scribe to a sufficient depth to allow the wire to just reach the top of the surface. The Belly Hook hole and its associated wrap "bore" hole were drilled. The accompanying "howto zip package" will provide a more detail set of steps as well as the pocket cut gcode and additional files to support this new operation.

The two lure halves were removed from the vise. The Through Wire Loops and Belly Hook Loops were created, assembled, and installed. Clearance "bore" holes were identified and created as needed to allow a "good" fit between the two halvs. This required a couple of assembly attempts using the Alignment Pins and then gently separating. During separation the both halves were examined to locate any points of conflict that might require additional "boring/cutting".

Once a "good" fit was achieved with the two halves and alignment pins were epoxied bonded. The excess lengths of the alignment pins were allowed to protrude from the Belly Half while being tangent with the Dorsal Half's surface. Painting coats were added as each dried. Avoid placing paint on the lip mating surface of the lure. Once the lure has dried the lip is ready to be epoxied into place. After that mating joint has cured a slanted slightly off-centerline pre-drill hole should be made such that the lip is pulled toward the mating surfaces of the lure. A small wood screw should then be fully inserted and then sealed. The body of the lure, immediate mating of the lip, and the lip's screw should then be receive sealant coat. The lure hardware was added after a "full" sealant cure in a solar oven.

Dorsal Half