Automatic Generation of Milling Toolpaths with Tool Engagement … · 2010. 10. 1. · However,...

Automatic Generation of Milling Toolpaths with ToolEngagement Control for Complex Part Geometry

Alexandru Dumitrache Theodor Borangiu Anamaria Dogar

Centre for Research & Training in Industrial ControlRobotics and Materials EngineeringUniversity Politehnica of Bucharest

IMS 2010

Alexandru Dumitrache (CIMR) Milling Paths with Tool Engagement Control IMS 2010 1 / 36

Outline

1 Overview3D Laser Scanning SystemCNC milling issuesTool Engagement Angle

2 Related work

3 Proposed algorithm and experimental resultsMilling parts for algorithm evaluationTraditional toolpathsProposed algorithmResults

4 Conclusions


3D Laser Scanning System Overview

Robot Controller Rotary table controller

4-Axis CNC Milling Machine

1

23

4

5

7

6

Data acquisition module (PC)

Instantaneous encoder positions

Laser probe location with respect to

scanned workpiece

USB

Point Cloud COM

PC

6-DOF VerticalRobot Arm

Laser probe

Rotary table

Scanned workpiece


3D Laser Scanning System Overview

6-DOF ArticulatedRobot Arm

4-Axis CNCMilling Machine

Laser Probe

Scanned Workpiece

Rotary Table


Main issues for CNC milling

CNC millingAutomatic CNC toolpath generationMilling parts with complex geometryMinimizing CNC milling time without overheating the cutting tool

Proposed solutionDepth map modelling of design part and raw stockNatural representation for 2.5D milling operationCan be extended to 4-axis milling


Tool Engagement Angle

The amount of sweep subentended by each cutting edgeas it engages and leaves the stockProportional to the cutting forcesIt is known to increase at internal corners in the toolpath

r

D

Milling tool Raw stock Tool engagement

∗

r


Tool Engagement Angle←→ Stepover

On straight line toolpaths, TEA (θ ) is proportional to stepover (s):

s =1+ sin(θ −90◦)

2, 0≤ θ ≤ 180◦ (1)

θ = 90◦+ arcsin(2s−1) , 0≤ s≤ 1 (2)

0

3760

90

120143

180

0 10% 25% 50% 75% 90% 100%

Tool

Eng

agem

entA

ngle

(deg

.)

Stepover (%)


Related work

Coleman (2006) explains the problem well, with intuitive examplesStori and Wright (2000): modified offset toolpath for convexcontoursBieterman (2001) replaced contour-parallel toolpaths with asmooth spiral, nearly circular at pocket center, and slowlymorphing into the part shape as it gets closer to the partIbaraki et al. (2004) removed the convexity requirement fromStory and Wright’s approachWang et al. (2005) defined a set of quantifiable metrics which canbe obtained by pixel simulationUddin et al. (2006) applied the Ibaraki’s approach for improvingtolerance on the finishing part by offsetting it nonuniformly, so thatthe finishing step is done at constant tool engagementRauch et al. (2009): constraints-based trochoidal toolpaths


Milling parts for algorithm evaluation


Milling parts – depth map representation


Traditional toolpaths

0 20 40 60 80 1000

30

60

90

120

150

180

Too

l Eng

agem

ent A

ngle

0 20 40 60 80 1000

30

60

90

120

150

180

Too

l Eng

agem

ent A

ngle

Direction-parallel toolpaths

Contour-parallel toolpaths


Trochoidal Step (SprutCam v7)

0 20 40 60 80 1000

30

60

90

120

150

180

Too

l Eng

agem

ent A

ngle

0 20 40 60 80 1000

30

60

90

120

150

180

Too

l Eng

agem

ent A

ngle


Contour-parallel toolpaths with trochoidal step



0 20 40 60 80 1000

30

60

90

120

150

180

Too

l Eng

agem

ent A

ngle

0 20 40 60 80 1000

30

60

90

120

150

180

Too

l Eng

agem

ent A

ngle



Tool engagement peak



0 20 40 60 80 1000

30

60

90

120

150

180

Too

l Eng

agem

ent A

ngle

0 20 40 60 80 1000

30

60

90

120

150

180

Too

l Eng

agem

ent A

ngle



Tool engagement peak

Tool engagement decreased


Trochoidal Step for Complex Geometry (SprutCam v7)

0 20 40 60 80 1000

30

60

90

120

150

180

Too

l Eng

agem

ent A

ngle

0 20 40 60 80 1000

30

60

90

120

150

180

Too

l Eng

agem

ent A

ngle



Tool engagement peaks are still present!


Proposed algorithm

InputTool diameterPrescribed tool engagement angleBinary image representing the design part (2D section)Binary image representing the raw stock (2D section)The 2D sections can be obtained by thresholding a 3D depth map

Output2D milling toolpaths consisting of small linear segmentsRaw stock shape after the generated milling operation


Proposed algorithmC

ontourreached

Eng

agem

ent

exce

eded

Found point

outside contour

ContouringFound pointon the contour

No points

found

No circumference pixels engaged

All contour points visited

Constant Engagement

Start

Stop

Find Starting Point


Constant Engagement Milling

Main section of the algorithmMilling a raw stock with arbitrary shapeThe only constraint is the tool engagement angleThe raw stock shape is updated at every step

Engagement

Milling cutter

Previous trajectoryAdvancing directionAdvancing direction for 90o TEA: Previous advancing

direction:

0o360o

270o

Intersection point

90o−ref

p

90climb

90oRawstock


Constant Engagement Milling

Example: toolpath with constant engagement for arbitrary raw stock shape


Contour Milling

Contour MillingTool moves along the offset contourTool is always tangent to the design pathUsually, tool engagement angle is much smaller than theprescribed value

Stop conditionsWhen TEA exceeds the prescribed value with more than a smallthreshold, the algorithm switches to Constant Engagement modeWhen all contour points are visited, the algorithm will search for anew starting point


Contour Milling

Example: Contour milling for the test part

Direction for constant engagement

Direction for contour milling


Contour Milling

Tool engagement angle exceeded the prescribed value


Direction for contour milling


Contour Milling

The algorithm switched to constant engagement milling



Finding Starting Point

The first possibility is chosen from the following:1 Continue a contouring operation, from the point where the

algorithm switched from Contouring to Constant Engagement2 Enter the raw stock horizontally, from lateral3 Plunge the cutter into raw stock

InputCurrent cutter position: (x0,y0)Current raw stock and part shapes (2D images)

OutputStarting point for next milling operation: (x,y)Milling trajectory for moving the cutter to (x,y):

Cutter retraction moves or tangent / plunge entry toolpaths


Example movie


Results – 37◦ engagement

0 20 40 60 80 1000

30

60

90

120

150

180

Too

l Eng

agem

ent A

ngle



0 20 40 60 80 1000

30

60

90

120

150

180

Too

l Eng

agem

ent A

ngle


Sharp corners

Sharp cornersThe generated toolpath may contain external cornersThese corners do not cause an increase in the tool engagementHowever, they are not good for machine dynamicsCorners are smoothed by the machine controller, with G64

0 20 40 60 80 1000

30

60

90

120

150

180

Too

l Eng

agem

ent A

ngle


Results


Conclusions

2D toolpath generation with tool engagement controlPrescribed reference value for engagement angleMaximum allowed overshoot (default: 20◦)

Toolpath consists of small linear segmentsSuitable for arbitrarily complex part and raw stock geometryRaw stock geometry can be digitized with 3D scanningReduces lubrication requirements and increases tool lifeHigher feed rates can be used, compared to traditional toolpathsBest results are obtained using this method for roughing,combined with the method from (Uddin et al., 2006) for finishingThe algorithm is used for milling complex 3D surfaces on millingmachines with 3 or 4 axes


Automatic Generation of Milling Toolpaths with Tool Engagement … · 2010. 10. 1. · However,...

Documents

Transcript of Automatic Generation of Milling Toolpaths with Tool Engagement … · 2010. 10. 1. · However,...