Project Number: ME-HXA-MG71

Bottling Line Splitter Design

A Major Qualifying Project Report

Submitted to the Faculty of the

WORCESTER POLYTECHNIC INSTITUTE

in partial fulfillment of the requirements for the

Degree of Bachelor of Science

in Mechanical Engineering

By

William Robinson Caruso

James Saunders

___________________________________________

___________________________________________

Date: March 3rd, 2007

Approved:

____________________________________
Prof. Holly K. Ault
Keywords:
1. Bottling
2. Machine Design
3. Beverage Production
Acknowledgements
Special thanks to those persons without whom this project would not have been
possible.

Worcester Polytechnic Institute

Professor Holly K. Ault

Professor Henry Nowick
James Phelan
IGSD
Professor Robert L. Norton

E & J Gallo Wineries

Mike Delikowski
Laura Hoffman
David Booth
Loel Peters
Brandon Abell
Shawn Burns
Jim Bellins
Tim Phillipsen
Juan Nevarez

Others

RPDG San Diego

Rexell-Norcal Modesto

i
Abstract
E&J Gallo Winery in Modesto, CA has numerous high speed bottling lines where
it is necessary to split a single line of bottles into two lines. The current lane splitting
mechanism uses multiple pneumatic actuators that require costly maintenance and cause
excessive line downtime. A prototype mechanism utilizing a unique three-dimensional
camoid design and single servo motion control was designed, fabricated using rapid
prototyping methods, and tested. Preliminary tests results proved acceptable
functionality. Shape optimization and long-term tests for reliability were recommended.

ii
ACKNOWLEDGEMENTS ...........................................................................................................I

ABSTRACT .................................................................................................................................. II

LIST OF FIGURES.................................................................................................................. VIII

LIST OF TABLES.........................................................................................................................X

LIST OF EQUATIONS .............................................................................................................. XI

CHAPTER 1.0 - INTRODUCTION ............................................................................................ 1

CHAPTER 2.0 - GOAL STATEMENT ...................................................................................... 3

CHAPTER 3.0 - TASK SPECIFICATIONS .............................................................................. 4

3.1 - ASSUMPTIONS ......................................................................................................................... 4

3.2 - PERFORMANCE SPECIFICATIONS .......................................................................................... 4
3.3 - DESIGN SPECIFICATIONS ....................................................................................................... 4
3.4 - IDEAL CASES........................................................................................................................... 5

CHAPTER 4.0 - BACKGROUND............................................................................................... 6

4.1 - GALLOS HIGH SPEED BOTTLING OPERATIONS .................................................................. 6

4.2 - E&J GALLO LINE 2 OVERVIEW ............................................................................................ 6
4.3 - HEUFT REJECTION SYSTEM .................................................................................................. 7
4.3.1 - OVERVIEW ............................................................................................................................ 7
4.3.2 - MECHANISM COMPONENTS .................................................................................................. 8
4.3.3 - CONTROLS .......................................................................................................................... 10
4.4 - PLANT TOURS ....................................................................................................................... 13
4.4.1 - WACHUSETT MICRO-BREWERY ......................................................................................... 13
4.4.2 - NORTHEASTERN REGIONAL ANHEUSER-BUSCH BREWERY ............................................... 14
4.5 - OTHER HIGH SPEED LINE SPLIT SOLUTIONS .................................................................... 15
4.5.1 - PATENT RESEARCH ............................................................................................................. 15
4.5.2 - OEM PRODUCTS ................................................................................................................. 15
4.6 - RAPID PROTOTYPING........................................................................................................... 19
4.7 - SERVO MOTORS ................................................................................................................... 19

CHAPTER 5.0 - CAMOID DESIGN CONCEPTION............................................................. 20

5.1 - CAMOID GEOMETRY ............................................................................................................ 22

5.1.1 - CAMOID CONCEPT .............................................................................................................. 22
5.1.2 - CAMOID INSPIRATION ......................................................................................................... 22

iii
5.1.3 - CAMOID DESIGN ................................................................................................................. 23
5.2 - ACTUATION DESIGN ............................................................................................................. 29
5.2.1 - WHY USE A SERVO?............................................................................................................ 29
5.2.2 - CHOOSING A SERVO............................................................................................................ 30
5.2.3 - SIZING A SERVO .................................................................................................................. 30
5.2.4 - CONTROLLING THE SERVO ................................................................................................. 31
5.2.5 - ACTUATION DESIGN CONCLUSIONS ................................................................................... 32
5.3 - DESIGN CONCEPT SUMMARY ............................................................................................... 32

CHAPTER 6.0 - METHODOLOGY ......................................................................................... 33

6.1 - ASSUMPTIONS ....................................................................................................................... 33

6.1.1 - LINE 2 PROPERTIES ............................................................................................................. 33
6.1.2 - HEUFT LANER ..................................................................................................................... 33
6.2 - DATA COLLECTION AND ANALYSIS .................................................................................... 34
6.2.1 - HEUFT ANALYSIS ............................................................................................................... 34
6.2.2 - CONTOUR DESIGN AND ANALYSIS ..................................................................................... 34
6.2.3 - CHOOSING CONTOUR.......................................................................................................... 35
6.3 - CAMOID GEOMETRY DESIGN .............................................................................................. 36
6.3.1 - MANUFACTURE................................................................................................................... 36
6.4 - ACTUATION DESIGN ............................................................................................................. 38
6.4.1 - SIZING A SERVO .................................................................................................................. 38
6.5.2 - CONTROLLING THE SERVO ................................................................................................. 39
6.5.3 - STRESS ANALYSIS............................................................................................................... 41
6.5.4 - FATIGUE STRESS ANALYSIS ............................................................................................... 41
6.6 - PART GATHERING ................................................................................................................ 41
6.7 - CHASSIS DESIGN ................................................................................................................... 41
6.7.1 - MATERIAL .......................................................................................................................... 42
6.7.2 - COMPONENT LAYOUT......................................................................................................... 42
6.7.3 - IMPLEMENTATION ON LINE ................................................................................................ 43
6.7.4 - MANUFACTURABILITY ....................................................................................................... 44
6.8 - DRIVE SYSTEM ..................................................................................................................... 44
6.9 - MECHANISM ASSEMBLY ...................................................................................................... 45
6.10 - IMPLEMENTATION.............................................................................................................. 45
6.11 - TESTING .............................................................................................................................. 46

CHAPTER 7.0 - CAMOID LANER DETAILED DESIGN .................................................... 47

7.1 - CAMOID GEOMETRY DESIGN .............................................................................................. 47

7.1.1 - ITERATIVE DESIGN ............................................................................................................. 47
7.1.2 - GEOMETRY FINAL VALUES ................................................................................................ 52
7.1.3 - GEOMETRY CONSTRUCTION ............................................................................................... 52
7.2 - CAMOID PROTOTYPE ........................................................................................................... 52
7.3 - ACTUATION DESIGN ............................................................................................................. 53
7.3.1 - SERVO SIZING ..................................................................................................................... 53
7.3.2 - SERVO SYSTEM ................................................................................................................... 53
7.3.3 - CONTROL LOGIC CONCEPT ................................................................................................. 57
7.3.4 - DRIVE TRAIN SPECS ........................................................................................................... 57
7.3.5 - TIMING ................................................................................................................................ 57

iv
7.4 - MOTION ANALYSIS............................................................................................................... 60
7.4.1 - CAM EXTENSION................................................................................................................. 60
7.5 - CHASSIS DESIGN DETAILS ................................................................................................... 60
7.5.1 - MATERIAL CHOICE ............................................................................................................. 62
7.5.2 - LAYOUT .............................................................................................................................. 62
7.5.3 - TOLERANCES ...................................................................................................................... 63
7.5.4 - FASTENERS ......................................................................................................................... 64
7.5.5 - BEARINGS ........................................................................................................................... 64
7.6 - DRIVE DESIGN DETAILS ...................................................................................................... 64
7.6.1 - RECOMMENDATION ............................................................................................................ 66
7.6.2 - CLAIMED BENEFITS ............................................................................................................ 66
7.6.3 - CHASSIS LAYOUT ............................................................................................................... 66
7.6.4 - DRIVE RATIO ...................................................................................................................... 66
7.6.5 - AVAILABILITY .................................................................................................................... 66
7.6.6 - BELT TENSIONER ................................................................................................................ 66
7.7 - FULL ASSEMBLY................................................................................................................... 68
7.8 - MECHANISM ASSEMBLY ...................................................................................................... 69
7.8.1 - CAMOID POSITION ON SHAFT ............................................................................................. 69
7.8.2 - CAMOID SHAFT ALIGNMENT .............................................................................................. 69
7.8.3 - SERVO INSTALLATION ........................................................................................................ 70
7.8.4 - SPROCKET POSITION ........................................................................................................... 70
7.8.5 - BELT TENSIONING .............................................................................................................. 70
7.8.6 - INITIAL MOTION TEST ........................................................................................................ 71
7.9 - IMPLEMENTATION................................................................................................................ 71

CHAPTER 8.0 - DATA COLLECTION AND ANALYSIS .................................................... 72

8.1 - BOTTLE DYNAMIC ANALYSIS .............................................................................................. 72

8.1.1 - FRICTION TEST ................................................................................................................... 72
8.1.2 - CENTER OF GRAVITY TEST ................................................................................................. 72
8.1.3 - NECK STRENGTH TEST ....................................................................................................... 73
8.2 - CAM PROFILE DESIGN AND ANALYSIS ............................................................................... 73
8.2.1 - CONSTRAINTS ..................................................................................................................... 74
8.2.2 - KINEMATIC ANALYSIS ........................................................................................................ 75
8.2.3 - FORCE ANALYSIS................................................................................................................ 80
8.2.4 - MOMENT ANALYSIS ........................................................................................................... 82
8.3 - TRAJECTORY TEST .............................................................................................................. 83
8.4 - CHOOSING A CONTOUR ....................................................................................................... 84
8.4.1 - ACCEPTABLE FINAL VELOCITY .......................................................................................... 84
8.4.2 - MINIMIZING MAXIMUM ACCELERATION ........................................................................... 84
8.4.3 - MINIMIZING JERK ............................................................................................................... 84
8.4.4 - MINIMIZING FORCE ............................................................................................................ 84
8.4.5 - MINIMIZING MOMENT ........................................................................................................ 84
8.5 - TORQUE REQUIREMENTS .................................................................................................... 84
8.5.1 - OPERATING ROTATIONAL VELOCITY ................................................................................. 84
8.5.2 - ANGULAR ACCELERATION ................................................................................................. 85
8.5.3 - TORQUE REQUIRED............................................................................................................. 86
8.6 - STRESS ANALYSIS................................................................................................................. 87
8.7 - FATIGUE STRESS ANALYSIS ................................................................................................ 89

v
CHAPTER 9.0 - PROTOTYPE TESTING............................................................................... 91

9.1 - MATERIALS........................................................................................................................... 91
9.2 - OBJECTIVE............................................................................................................................ 91
9.3 - VARIABLES ........................................................................................................................... 91
9.4 - SETUP SAFETY PRECAUTION ............................................................................................... 91
9.5 - SETUP .................................................................................................................................... 92
9.5.1 - BOTTLE TEST BATCHES ...................................................................................................... 92
9.5.2 - TEST LOOP SPEED ............................................................................................................... 93
9.5.3 - PHOTOEYE PLACEMENT...................................................................................................... 93
9.5.4 - ELECTRONIC ATTACHMENT ............................................................................................... 93
9.6 - PROCEDURE SAFETY PRECAUTION..................................................................................... 94
9.7 - PROCEDURE .......................................................................................................................... 94
9.7.1 - DATA RECORDING .............................................................................................................. 94

CHAPTER 10.0 - TEST RESULTS........................................................................................... 95

10.1 - TESTING LIMITATIONS ...................................................................................................... 95

10.1.1 - LINE ENCODER ................................................................................................................. 96
10.1.2 - VARIABLE LINE SPEED ..................................................................................................... 96
10.1.3 - CONVEYOR CONDITION .................................................................................................... 96
10.1.4 - LINE SPEED ....................................................................................................................... 96
10.2 - PRELIMINARY TESTING ..................................................................................................... 96
10.3 - RESULTS TABLE ................................................................................................................. 98
10.3.1 - MINIMUM BOTTLE SPACING: EXTENSION ........................................................................ 99
10.3.2 - MINIMUM BOTTLE SPACING: RETRACTION ...................................................................... 99
10.3.3 - MAGNITUDE OF DISPLACEMENT ...................................................................................... 99
10.3.4 PHOTOEYE PLACEMENT ................................................................................................... 99
10.4 - MECHANISM OBSERVATIONS .......................................................................................... 100
10.5 - TEST RESULTS CONCLUSIONS ......................................................................................... 100

CHAPTER 11.0 - COST COMPARISON............................................................................... 101

CHAPTER 12.0 - RECOMMENDATIONS ........................................................................... 102

12.1 - LINE SPEED TESTING ....................................................................................................... 102

12.2 - BOTTLE TESTING ............................................................................................................. 102
12.3 - ENDURANCE TESTING ...................................................................................................... 102
12.4 - PART ACQUISITION .......................................................................................................... 102
12.5 - PROGRAM OPTIMIZATION ............................................................................................... 102
12.6 - GEOMETRY OPTIMIZATION ............................................................................................ 102

CHAPTER 13.0 - CONCLUSIONS ......................................................................................... 103

REFERENCES .......................................................................................................................... 104

APPENDICES ........................................................................................................................... 105

vi
APPENDIX A - RELEVANT PATENTS .......................................................................................... 106
APPENDIX B - RAPID PROTOTYPING QUOTES .......................................................................... 113
APPENDIX C - BILL OF MATERIALS .......................................................................................... 114
APPENDIX D - MOTION ANALYZER INPUT VALUES ................................................................. 118
APPENDIX E - SERVO DETAILS .................................................................................................. 120
APPENDIX F - SERVO DRIVER DETAILS .................................................................................... 122
APPENDIX G - CHASSIS DRAWINGS........................................................................................... 125
APPENDIX H - GOODYEAR EAGLE PD POWER TRANSMISSION .............................................. 133
APPENDIX I - BOTTLE TESTS ..................................................................................................... 141
I.1 - CENTER OF GRAVITY TEST .................................................................................................. 141
I.2 - COEFFICIENT OF STATIC FRICTION TEST ............................................................................. 145
I.3 - BOTTLE NECK FAILURE TEST .............................................................................................. 147
I.3 - BOTTLE NECK FAILURE TEST .............................................................................................. 148
APPENDIX J - RAPID PROTOTYPING METHODS ....................................................................... 150
J.1 - STEREOLITHOGRAPHY (SLA)............................................................................................... 150
J.2 - FUSED DEPOSITION MODELING (FDM) ............................................................................... 150
J.2 - FUSED DEPOSITION MODELING (FDM) ............................................................................... 151
J.3 - SELECTIVE LASER SINTERING (SLS) ................................................................................... 151
J.4 - ELECTRON BEAM MELTING (EBM) ..................................................................................... 151
APPENDIX K TEST DATA TABLE ............................................................................................ 152
APPENDIX M - DETAILED MATHEMATICS................................................................................ 153

vii
List of Figures
Figure 1: Patent 4,369,873 Concept.................................................................................... 8
Figure 2 (left): Line 2 Heuft Delta-FW On Position........................................................... 8
Figure 3 (right): Line 2 Heuft Delta-FW Off Position........................................................ 8
Figure 4: Cylinder Layout (top view) ................................................................................. 9
Figure 5 (left): Segments in Off Position (top view) .......................................................... 9
Figure 6 (right): Contour in On Position............................................................................. 9
Figure 7: Segment Depth and Bristles .............................................................................. 10
Figure 8: Segment Close-up (top view). Note bristles on end of each segment. .............. 10
Figure 9: Heuft operation off to on switch........................................................................ 12
Figure 10: Wachusett Lane Splitter .................................................................................. 14
Figure 11: Heuft Delta-K Unit.......................................................................................... 16
Figure 12: Heuft Flip Rejecter .......................................................................................... 17
Figure 14: KHS Waveform............................................................................................... 18
Figure 15: Top view of KHS Waveform .......................................................................... 18
Figure 17: Camoid Laner .................................................................................................. 20
Figure 18: Camoid Extension ........................................................................................... 21
Figure 19: Heuft Delta-K Bottle Rejecter......................................................................... 22
Figure 20: Camoid Laner Shape ....................................................................................... 23
Figure 21: Geometry Terminology ................................................................................... 24
Figure 22: Cam Segment Construction............................................................................. 25
Figure 23: Segmented Camoid.......................................................................................... 26
Figure 24: Cam segment phase shifting............................................................................ 28
Figure 25: Sections of Continuous Contour...................................................................... 28
Figure 26: Heuft Contour and Equation............................................................................ 34
Figure 27: Component Layout .......................................................................................... 43
Figure 28: Heuft Attachment to Line................................................................................ 44
Figure 29: Test Loop......................................................................................................... 46
Figure 30: Geometric Elements ........................................................................................ 48
Figure 38: Model vs. Rapid prototype .............................................................................. 53
Figure 39: Servo Control Schematic................................................................................. 54
Figure 40: Servo Motor..................................................................................................... 55
Figure 41: Timing Diagram Example ............................................................................... 58
Figure 42: Top View of Cam Extension ........................................................................... 60
Figure 43: Chassis Assembly View .................................................................................. 61
Figure 44: Chassis Exploded View................................................................................... 62
Figure 45: Important Toleranced Dimensions .................................................................. 64
Figure 46: Goodyear Eagle Pd Power Transmission ........................................................ 65
Figure 47: Lovejoy Belt Tensioner with smooth idler pulley........................................... 67
Figure 48: Full CAD Assembly ........................................................................................ 68
Figure 49: Full Assembly.................................................................................................. 69
Figure 50: Tensioning Diagram ........................................................................................ 70
Figure 54: Contour of Each Cam Program ....................................................................... 76
Figure 55: Displacement Difference, Heuft vs. Cam Programs ....................................... 77
Figure 56: Bottle Transverse Velocity.............................................................................. 78

viii
Figure 57: Bottle Transverse Acceleration ....................................................................... 78
Figure 58: Bottle Transverse Jerk ..................................................................................... 79
Figure 59: Bottle Free Body Diagram .............................................................................. 80
Figure 60: Force on Bottle ................................................................................................ 81
Figure 61: Induced Moment on Bottles ............................................................................ 82
Figure 62: Bottle Trajectories ........................................................................................... 83
Figure 63: Spin-up time vs. Buffer Angle ........................................................................ 85
Figure 64: Acceleration vs. Spin-Up Time ....................................................................... 86
Figure 65: Servo Torque Required vs. Spin-Up Time...................................................... 87
Figure 66: Test loop .......................................................................................................... 92
Figure 67: Electronics Board ............................................................................................ 93
Figure 68: Final Mechanism Cycle Sequence .................................................................. 95
Figure 69: Photoeye Placement ........................................................................................ 99
Figure 70: Patent 4,986,407 ............................................................................................ 106
Figure 71: Patent 4,643,291 ............................................................................................ 107
Figure 72: Patent 4,321,994 ............................................................................................ 108
Figure 73: Patent 4,369,873 ............................................................................................ 109
Figure 74: Patent 6588575 .............................................................................................. 110
Figure 75: Patent 6,822,181 ............................................................................................ 111
Figure 76: Patent 3,791,518 ............................................................................................ 112
Figure 77: Servo Overview............................................................................................. 120
Figure 78: Driver Details and Benefits ........................................................................... 122
Figure 79: Driver Specifications..................................................................................... 123
Figure 80: Driver Specifications..................................................................................... 124
Figure 81: Belt Nomenclature......................................................................................... 133
Figure 82: Eagle Pd Belt Product Numbers.................................................................... 134
Figure 83: Sprocket Nomenclature ................................................................................. 135
Figure 84: Eagle Pd White Sprockets ............................................................................. 136
Figure 85: Belt Nomenclature......................................................................................... 137
Figure 86: Eagle Pd Belt Product Numbers.................................................................... 138
Figure 87: Sprocket Nomenclature ................................................................................. 139
Figure 88: Eagle Pd White Sprockets ............................................................................. 140
Figure 89: Center of Gravity Experiment ....................................................................... 142
Figure 90: Free Body Diagram ....................................................................................... 143
Figure 91: Free Body Diagram ..........................................Error! Bookmark not defined.
Figure 93: Coefficient of Static Friction Experiment ..................................................... 145
Figure 94: Free Body Diagram ....................................................................................... 146
Figure 97: Bottle Neck Failure Experiment.................................................................... 148
Figure 98: Free Body Diagram ....................................................................................... 149

ix
List of Tables
Table 1: Manufacturing Method Comparison................................................................... 37
Table 2: Rapid Prototyping Method Comparison............................................................. 37
Table 3: Material Comparison .......................................................................................... 42
Table 4: Camoid Geometry Final Values ......................................................................... 52
Table 5: Servo Motor and Gearbox Overview Specs ....................................................... 56
Table 6: Driver Overview Specs....................................................................................... 56
Table 7: Tolerance Reasons .............................................................................................. 63
Table 8: Belt Used ............................................................................................................ 65
Table 9: Sprockets Used ................................................................................................... 65
Table 10: Cam Program Constraints................................................................................. 75
Table 11: Graphic Color Scheme...................................................................................... 76
Table 12: Maximum Kinematic Values............................................................................ 80
Table 13: Maximum Force Analysis Values .................................................................... 83
Table 14: Time Data ......................................................................................................... 85
Table 15: Keyway Stress Analysis ................................................................................... 88
Table 16: Shaft Stress Analysis ........................................................................................ 89
Table 17: Shaft Fatigue Analysis...................................................................................... 89
Table 18: Keyway Stress Concentration Analysis............................................................ 90
Table 19: Test Results....................................................................................................... 98
Table 20: Cost Comparison ............................................................................................ 101
Table 21: Full Bill of Materials ...................................................................................... 115
Table 22: Bill of Materials with Piggy Backed Electronics ........................................... 117
Table 23: Axis Setup Tab ............................................................................................... 118
Table 24: Cycle Profile Tab............................................................................................ 118
Table 25: Mechanism Tab .............................................................................................. 119
Table 26: Transmission Stages Tab ................................................................................ 119
Table 27: Data Recording Table ..................................................................................... 152

x
List of Equations
Equation 1: Bottle and Laner Contact Time ..................................................................... 85
Equation 2: Operational Rotational Velocity.................................................................... 85
Equation 3: Worst Case Scenario Time Between Two Bottles ........................................ 85
Equation 4: Acceleration Function ................................................................................... 86
Equation 5: Torque Function ............................................................................................ 86
Equation 6: Shear Force at Key ........................................................................................ 88
Equation 7: Keyway Average Shear Stress ...................................................................... 88
Equation 8: Safety Factor.................................................................................................. 88
Equation 9: Area of Key ................................................................................................... 88
Equation 10: Required Torque.......................................................................................... 88
Equation 11: Shaft Torsional Deflection .......................................................................... 89
Equation 12: Shaft Shear Stress due to Torsion................................................................ 89
Equation 13: Shaft Polar Moment of Inertia..................................................................... 89
Equation 14: Corrected Fatigue Function ......................................................................... 89
Equation 15: Safety Factor................................................................................................ 89
Equation 16: Fatigue Stress Concentration Factor............................................................ 90
Equation 17: Shear Stress Concentration due to Torsion ................................................. 90
Equation 18: Safety Factor................................................................................................ 90

xi
Chapter 1.0 - Introduction
E&J Gallo Winery is one of the largest winemaking operations in the world.
Founded in 1933 by Ernest and Julio Gallo, the Gallo Winery is still a family business
and since has expanded to the global market. Currently, Gallo Winery employs over 4600
people and retails wines throughout the United States and over 90 foreign countries.
Grapes used are from vineyards spanning Californias most important wine-producing
regions and Gallo runs operations in sites across Sonoma County including Modesto,
Monterey and Napa Valley. The diverse product line of Gallo Winery ranges from fine
table and sparkling wines, to distilled wine-based spirits and beverages (E&J Gallo,
2003).
This project is conducted in E&J Gallos Modesto site and focuses on 750mL
bottled wine production, specifically on Line 2. At full production Line 2 fills and
packages at up to 400 bottles per minute, running 16 hour shifts, seven days per week.
The limiting factor in the production speed of the line is the packaging process. In
attempt to maximize the efficiency of the line, Gallo employs two parallel packaging
lines resulting in the need for a lane split from single file to two independent lanes
between the bottle filler and the packing equipment.
Gallo is experiencing expensive maintenance issues with the mechanism which
splits the lanes. The current system is manufactured by Heuft. The manufacturers
recommendation for the mechanism states its function to be a single bottle rejecter.
However, in Line 2 the Heuft rejecter functions as a high speed lane splitter. The
difference is in the number of cycles per unit time. As a rejecter, the mechanism might
actuate several dozen times per shift, but as a dedicated lane splitter, it fires several
thousand times per shift. This results in a significant deterioration of performance in a
relatively short period of time, and thus requires maintenance or replacement often
enough to cause concern. Furthermore, as performance degrades, the chance of bottle
damage or bottle tipping increases which can cause costly delays in production.
In order to reduce the maintenance cost and production delays associated with the
Heuft laning mechanism, Gallo proposed a redesign of the system. Within the proposal,
Gallo specified certain criteria that the redesign must follow. The new system must
reduce maintenance costs, must fit within the existing footprint, and not damage bottles.
The new system must not be a safety hazard. And the system must ideally be replicable
on other lines with different size and shape bottles.
This project focuses specifically on the redesign of the Line 2 Heuft system.
Through problem identification, research, ideation, analysis, modeling, prototypes and
implementation, this project will propose a solution to the problems experienced with the
current system. The following report will explain the details of design, implementation,
experimentation and results of a unique prototype lane splitting mechanism.
The chapters present information in parallel structures with each successive
chapters investigation driving progressively deeper in detail. After discussing relevant
background information, the design process is explained in detail over chapters 5-7.
Chapter 5 introduces the idea explaining terminology and design basics. Chapter 6
describes the methodology used to create that design. Chapter 7 describes the detailed
design process and gives all values in the final design. Chapter 8 details all important
calculations conducted. Chapters 9 and 10 describe the experimental process; procedure

1
and results. Chapter 11 concludes the project with a cost analysis. Conclusions and
recommendations are discussed in the final chapters. This structure will tend to a variety
of audiences and allow quicker reference for those familiar with the project and detailed
step-by-step explanation for those who are not.

2
Chapter 2.0 - Goal Statement
The aim of this project is to design a means of splitting one high-speed single-file
line of bottles into two independent single-file lines. Alternatively, the existing
mechanism may be improved to reduce maintenance costs and upkeep.

3
Chapter 3.0 - Task Specifications
This chapter outlines the design specifications that must be met in order to claim a
successful solution.

3.1 - Assumptions

This specification list was compiled assuming 750mL bottles traveling at 400
bottles/min with a space of one bottle diameter between each bottle. We assume a worst
case operation of 16 hours/day, 365 days/year, and a lane switching frequency of 10
bottles per lane. A full cycle requires the mechanism to transition from off to on and
back to off, allowing 20 bottles to pass. This adds up to nearly 7 million cycles per year.

3.2 - Performance Specifications

The Laning Mechanism Must:

NOT cost more than $20,000 in parts, or more than $10,000 in installation.
Reduce maintenance costs from the current $10,000/year.
NOT damage or scuff bottles or labels.
Allow for either packer lane to be utilized independently.
o This would require the mechanism to remain in either the on or off
position for extended time if necessary.
Accommodate for variable line speeds.
Be frequency adjustable. This refers to the number of bottles to pass per cycle.
Follow food handling guidelines and regulations.
Be actuated without interfering with bottles.
o This means that any mechanism should be able to switch from on to off
without a bottle being influenced in this transition. The likely way to
accomplish this is to have the actuation take place in the space between
bottles as they pass. Our estimate of this time is 0.064 seconds.
Be able to redirect bottles moving with a kinetic energy of 0.5-1.0 Joules.
Be sustainable with routine maintenance.
Fit in the existing systems footprint.
NOT cause downed bottles.
NOT present a safety hazard to employees during operation or maintenance.

3.3 - Design Specifications

The frequency adjustability should be done by counting bottles (as opposed to a

timed action).
Routine maintenance should be accommodated in the design. Special tools or
parts should not be necessary to service and maintain the mechanism.
Tipping must be controlled, both induced (through the mechanisms workings)
and accidental.

4
3.4 - Ideal Cases

The design is applicable in various lines with various bottle types.

The design has a lifespan of 5 years (35 million cycles).

5
Chapter 4.0 - Background
Maintenance issues with a high speed lane splitter system on line 2 are causing
unneeded expenditures. This background chapter will offer insight into several of the
base issues that underlie a high speed lane splitting mechanism. Additional background
information regarding various components utilized in the design is also presented. The
topics of discussion will be:
High speed bottling process
Overview of Gallos line 2 bottle laning section
Examples of other bottling facilities
Heuft Bottle Rejection system
Rapid Prototyping
Servo Motors
Commercially Available High Speed Lane Splitters
Patent Research

4.1 - Gallos High Speed Bottling Operations

Gallo is one of the largest wineries in the world selling over 67 million cases of
wine annually, 1/3 of which is shipped to international markets. Most of Gallos bottling
operations are handled in the Modesto bottling plant.
The Modesto bottling plant is supplied by millions of gallons of stored wine
which is fed to 17 bottling lines. Each of the lines is designed to accommodate a certain
type of container. Gallos container lineup includes Bag-in-Box; 5L and 3L jugs; 1.5L,
750mL, 375mL and 187mL bottles. A variety of different products could fill any of the
bottles including table wine, sparkling wine, wine cooler beverages and ciders. The lines
run at different speeds depending on bottle size varying from 100 bottles per minute for
jugs to up to 1000 bottles per minute for the wine beverages.
The bottling lines are almost completely automated, with a small team of
operators, inspectors and mechanics for each line. Occasional problems can occur such as
broken or tipped bottles, filling mistakes, labeling mistakes, or misfeeds on caps or corks.
The line incorporates a series of inspection and rejection systems to discard any
problematic bottles; however operators are still present to ensure smooth operation.
The rejection systems can be found on almost every line. These rejection systems
use the same mechanism as the lane splitters, which also can be found on most of the
lines. The rejection systems and lane splitters are essential to the company to maintain
production quality and speed.

4.2 - E&J Gallo Line 2 Overview

Line two at E&J Gallo is a bottling line dedicated to handling 750mL bottles that
can be filled with a variety of product; from white to red table wines. Operating at a
maximum of 400 bottles per minute, line 2 is one of the fastest 750mL lines at Gallo. The
Heuft rejection system is the topic of interest for this project, however to understand the
reasons for the interest, we must explore the nature of the line.
Most lines at Gallo follow a similar procedure to fill and package bottles:

6
 1. Empty bottles arrive up-side-down in cases that are stacked on pallets
2. The rows of cases are stripped from the pallet and fed single file on a conveyor
3. The cases are tipped up-side-down to remove empty bottles
4. Cases and bottles are separated
5. Bottles are rinsed
6. Bottles are filled and corked
7. Bottles are x-ray inspected
8. Bottles are dried
9. Bottles are capped
10. Bottles are labeled
11. Bottles are cased and packaged
12. Cases are palletized and stored

The point of interest is after step 7, which is where rejection and lane splitting
occurs. The reason for the lane split is because the packaging machine is not capable of
the line speed of the filler. In order to compensate for the slower packaging machine,
Gallo employs two machines.
The method by which the bottles are divided into separate lanes is the Heuft
Rejection system. This system is designed to be used as a single bottle rejection system,
however is used on many of the Gallo lines as a dedicated lane splitter. The system
causes a significant amount of maintenance costs because of the number of cycles it must
endure. A rejecter may actuate several dozen times per day, while a dedicated laner
actuates thousands of times per day. The system is not designed for this high number of
cycles and thus experiences malfunctions and wear in a relatively short period of time.

4.3 - Heuft Rejection System

4.3.1 - Overview
The Heuft Rejection System is based on a design invented by Bernhard Heuft and
is currently manufactured and distributed by Heuft Systemtechnik Gmbh of Germany.
Founded in 1979, the company has based its business on innovation and unusual ideas
that have since become industry standards. Today Heuft is represented all over the world
with 20 subsidiary company locations on 5 continents (Heuft, 2007).
The Heuft rejecter initial concept was filed by Bernhard Heuft for U.S. patent
number 4,369,873 filed January 10, 1979 under the title Apparatus for Laterally
Deflecting Articles. This concept was later improved upon in patent number 4,321,994 on
April 21, 1980 under the title Means for Laterally Deflecting Articles from a Path of
Travel. Since conception, the system has undergone minor changes to accommodate for
modern production speeds, but the system has remained largely the same as the original
concept. This patent is documented in Appendix A.
The Heuft rejecter system used by Gallo Winery is named Heuft Delta-FW 16 and
is boasted by the Heuft Company as a robust, all around system for speed of up to
150,000 containers per hour [2500 containers/min] (Heuft, 2007). While its primary
purpose is to single out a container for rejection, Heuft states it can also be used for the
removal of fallen containers and foreign objects (Heuft, 2007).

7
4.3.2 - Mechanism Components
The system is comprised of 16 segments that are linearly actuated transversely
across the conveyor in front of the containers path of travel. Figure 1 shows a top view
of the mechanism. Notice the segments independently controlled actuation and how only
a single bottle is diverted while the others are not manipulated. The concept behind the
design ensures that the bottles are diverted in a gradual manner to reduce the risk of
tipping.

Figure 1: Patent 4,369,873 Concept 1

Figure 2 (left): Line 2 Heuft Delta-FW On Position

Figure 3 (right): Line 2 Heuft Delta-FW Off Position

The Heuft Delta-FW 16 is actuated by a means of 16 pneumatic cylinders ranging

in cylinder body size and has adjustable stroke lengths. Each of the 16 cylinders is

1
Free Patents Online. Patent Analytics and Patent Searching. Retrieved January 8th, 2007. from
www.freepatentsonline.com.

8
controlled individually by an electronic solenoid valve. Note that Gallo uses only 12 of
the 16 cylinders (Figure 2).
The 16 cylinders range from 45mm to 155mm in length and have an adjustable
stroke length via a threaded piston rod and stopper nut. The strokes are adjustable using
ordinary hexagonal sockets. The piston rod includes a rubber shock absorber between the
cylinder and the stopper nut to decrease noise as well as wear and fatigue during
actuation.

Figure 4: Cylinder Layout (top view)

Attached to each piston rod is a plastic segment by which the bottles are diverted
as seen in Figure 2. Each segment is approximately the same width. When all pistons are
in the fully off position, the bottles are not diverted and continue on a neutral default
path (Figure 2 and Figure 5). When the all pistons are actuated in the fully on position,
the segments are arranged in a curved contour. The contour of each segment is linear, as
shown in Figure 8. However the stroke lengths of each piston are such that the array
forms the non-linear contour seen in
Figure 6. Also each segment is progressively longer than the previous segment to aid in
the horizontal translation.

Figure 5 (left): Segments in Off Position (top view)

Figure 6 (right): Contour in On Position

The segments are approximately 4 inches deep and have a row of bristles that act as a
cushion for the bottle to limit impact and bottle scuffing ( Figure 7). The bristles are

9
approximately inch in length. In addition to bottle cushioning, the bristles also act as a
buffer for any irregularities in segment spacing or piston actuation length.

Figure 7: Segment Depth and Bristles

Figure 8: Segment Close-up (top view). Note bristles on end of each segment.

4.3.3 - Controls
The timing of the system is the most crucial element for the successful operation
of the system. To completely avoid the chance of tipped bottles, the basic concept of the

10
design is centered on single point guidance of a bottle. In essence, the timing of the firing
sequence is such that the contour is laid out in front of a single container so that the bottle
directly in front of the target bottle is not diverted and every bottle after the target is
diverted. For single bottle diversion, the segments are retracted immediately after the
target bottle has passed.

4.3.3.1 - Data Input

In order to start the firing sequence correctly, the position of the bottles on the line
must be recorded. This is accomplished by an electronic counting device directly before
the rejecter. In order to fire the pistons at the correct time interval after the initial piston,
the line speed must be recorded. Coupled to the rejecters control system is an encoder on
the conveyor which records the line speed and bottle position. With these data known, the
firing sequence can be timed correctly to avoid bottle tipping (Heuft USA, Inc.).

4.3.3.2 - Off to On Switch

The first piston (shortest stroke) is always the first piston to fire. This must occur
between the target bottle and the bottle directly in front of the target. Depending on the
line speed, the pistons will continue to fire between these same two bottles. This insures
that the diversion of the bottle is caused by a passive contour, i.e. the bottle will not be
punched or pushed by an actuating piston. If a punch occurs, the bottle could accelerate
laterally into the opposite side and cause an unwanted impact or tip. Once the pistons are
all actuated, they remain in this position for a predetermined number of bottles to pass.
Figure 9 shows the firing sequence of the Heufts pistons. The bottles are
highlighted for visualization. Note that the direction of travel is downward. The rejection
system is on the right side of each photo.

11
Figure 9: Heuft operation off to on switch
A. Bottle 1 passes by the off segments. Bottle 2 approaches.
B. Both Bottle 1 and Bottle 2 are passing the off segments.
C. Bottle 1 continues to Lane 1. Bottle 2 is still passing the off segments.
D-F. The segments begin to fire. Bottle 3 enters the beginning of the splitter. Bottle
3 is riding the on segments while Bottle 2 is riding the off segments
simultaneously.
G. Bottle 2 continues to Lane 1. Bottle 3 is riding the on segments.
H-I. Bottle 3 continues to Lane 2.

4.3.3.3 - On to Off Switch

Again, the first piston is the first to fire and occurs between the target bottle and
bottle directly in front of the target bottle. The sequence firing is the same as the off to
on switch; however the bottle in front of the target bottle follows the contour, while the
target bottle follows the default neutral path.

12
 The system is designed to be able to single out one bottle from a stream without
disrupting the rest of the stream by laying out a path in front of a single target bottle.
Malfunctions associated with the Heuft system occur most often with the
pneumatic actuators. The seals wear and the actuation becomes sluggish over time, which
causes timing problems. The timing problems become a liability when working at high
speeds, causing tipped bottles and broken bottles. These situations can cause the line to
slow down or shut down.

4.4 - Plant Tours

4.4.1 - Wachusett Micro-Brewery
Operations at Wachusett Brewery may not be directly applicable to the production
lines at Gallo. There is a vast difference in the volume of production between the two
companies. We felt, however, that a visit could offer valuable insight into glass bottle
handling in general.
The entire bottling system is contained in one room, with one continuous line.
The process begins as the pallet of empty bottles is unloaded to a holding table, which
then takes the 440 bottles per tray and feeds them single file into the twist washer. The
twist washer spins each bottle several times to clean it inside and out.
As the bottles leave the twist washer they get injected with a spray of liquid
nitrogen. The nitrogen evacuates the air from the bottle as it evaporates. The bottles are
then fed into the filler, which then passes the bottles to the capper, and then a washer.
From here the bottles are split to be fed to two labeling machines; the splitter simply a
thin rigid divider that holds several bottles at a time (Figure 10). Once the bottles fill the
primary lane, the remaining bottles are deflected down-line to the second labeler. At full
speed the two lanes alternate each bottle.

13
Figure 10: Wachusett Lane Splitter
The labeled bottles continue down the line to an accumulation table where they
fill four lanes and are drop packed into their cases, boxed, palletted and either stored
onsite or shipped immediately.
Witnessing the manipulation of bottles in person did help us gain some
understanding into the problem. It was particularly interesting to hear the plant workers
explain how they would separate bottles into two lines, as they have much more
experience than us dealing with bottles.
While Wachusett Brewery is orders of magnitude smaller than Gallo in
production quantities, the basic principles of dealing with glass bottles are largely the
same. The visit to Wachusett gave us a better understanding of these principles, and
demonstrated some well established methods of manipulating bottles.

4.4.2 - Northeastern Regional Anheuser-Busch Brewery

Trip Date: Wednesday, November 29th, 2006

With a bottling line rate of approximately 1400 bottles/minute, and an aluminum

can line with a rate of around 2000 cans/minute, this facility gives a lot of applicable
information to the wine bottle line we are dealing with.
Anheuser runs its main bottling line at 1400 bottles/minute, 24 hours a day, 365
days a year. Production from this brewery is remarkably well engineered and streamlined.
There is one lane split per bottling line that occurs after the bottle filler before the
entering the labeling machine. At this point in the process, the full, unlabeled bottles are
in single file and moving at maximum velocity. The bottles enter an accumulation zone,
the final result of which is 8-10 bottles filed across. This significantly lowers the forward
velocity (and thus momentum) such that the lane split is achieved by a passive wedge.

14
 This Anheuser-Busch plant does not use any form of a high speed lane splitting
mechanism.

4.5 - Other High Speed Line Split Solutions

Research was conducted on high speed lane split solutions through patent
searches and commercially available products.

4.5.1 - Patent Research

There are currently several different patent ideas on how to conduct a high speed
lane shift of a container on a conveyor. Many of these ideas are for single container
rejection, not for dedicated high speed lane shifters. The abstracts of each patent can be
found in Appendix A. All patents research was gathered from
www.freepatentsonline.com.

4.5.2 - OEM Products

As stated before there are several products on the market that have the ability for
a high speed lane shift, however, most of the products serve as rejection systems, not
dedicated lane splitters. Listed below are several companies that manufacture mechanism
to cause a diversion of a container from one path to another.

4.5.2.1 - Heuft
Heuft offers several different solutions to the lane diversion/rejection system,
most of which are patented by the founder of the company. The following list is Heufts
current line-up of commercially available products for lane splitting.

15
4.5.2.1.1 - Heuft Delta-K

Figure 11: Heuft Delta-K Unit 2

The Heuft Delta-K unit, shown in Figure 11, is similar to the Heuft Delta-FW
system, except instead of linearly actuating segments, the Delta-K employs rotating
segments that fan down in a similar timing that the Delta-FW follows.

2
Heuft USA, Inc. Container Rejection Systems. Retrieved January 10th, 2007, from www.heuft.com.

16
4.5.2.1.2 - Heuft Flip Rejecter

Figure 12: Heuft Flip Rejecter 3

The Heuft flip rejecter shown in Figure 12 is a single arm actuator that simply
pushes a single container transversely across the
conveyor. It is a simple robust option for single
container rejection.

4.5.2.1.3 - Heuft XY
The Heuft XY is a multi-segmented
linear actuation system shown in Figure 13 4
capable of multi-lane sorting of containers. It is
ideal for sorting as it can divide containers into
up to four lanes. This system is most suitable for
low speed applications

4.5.2.2 - KHS
KHS is a respected company in the
production industry. Figure 13: Heuft XY4

3
Heuft USA
4
Heuft USA

17
 KHS does not offer any dedicated lane splitters, however, they offer a unique
system that double-files a single file line of bottles. This system could potentially be used
in conjunction with a passive wedge lane divider. An example photo is shown in Figure
14.

Figure 14: KHS Waveform 5

Figure 15: Top view of KHS Waveform 6

This system is interesting as it uses two spinning belts with complementary

waveforms such that as bottles are fed through, they are laned in a staggered double file
line. This system could function as a unique laning system if the double file line was then
split with a passive wedge.

5
KHS. Container Conveying Solutions. Retrieved November 19, 2007 from
www.kisters.com/img/pool/1111_Container%20Conveying%20Systems.pdf.
6
KHS

18
4.6 - Rapid Prototyping
Rapid prototyping is the modern
production method of forming solid parts
from a CAD model without the use of
traditional fabrication techniques.
Complex geometries can be formed more
quickly for less initial investment than
other methods. There is no need for a
mold, and typically no secondary
operations are necessary. Part accuracy is
generally quite good, depending on the
process and materials used.
The general concept behind rapid
prototyping is to divide the CAD model
into many cross sections, and then to build
a physical part by accumulating these
Figure 16: Complex Geometry Realized Through
sections one on the other. There are Fused Deposition Modeling
numerous means of accomplishing this,
using stock material in the form of sheets, powder, liquid, resin or wire. The process
allows the fabrication of otherwise non-machinable geometries, as seen in Figure 16 7.
More detailed information on rapid prototyping techniques can be found in Appendix J.

4.7 - Servo Motors

Servo motors are, in basic terms, an electric motor with a brain. Servos as
described by Robert Norton in Design of Machinery, are fast-response, closed loop
controlled motors capable of providing a programmed function of acceleration or
velocity, providing position control, and of holding a fixed position against a load 8. The
control is based on the closed loop system, which means that sensors on the servo feed
back information on the motors position and velocity to the controller. The controllers
are called drivers, which is a computer that responds to the information and adjusts the
current flow to drive the motor. The drivers can be programmed to control the servo
motor dynamically to adjust for changes in load and commands. The commands can be
input in real time through user interface or can be imbedded in a cycle program.
Servos can be configured in both AC and DC, have high torque capability and
perform well in instances needing rapid acceleration and deceleration. They are capable
of providing tight toleranced constant velocities, even under dynamic loading.
Servos are gaining rapid acceptance at Gallo, making complicated operations
easier to automate. In the last five years, servos have found their way to several of the
bottling lines and continue to be implemented.

7
Figure retrieved from http://www.cs.berkeley.edu/~sequin/SCULPTS/SnowSculpt02/maquettes.html
8
Norton, Robert L. Design of Machinery. New York: McGraw Hill, 2004, page 70.

19
Chapter 5.0 - Camoid Design Conception

Figure 17: Camoid Laner

The camoid laner concept proposes to replace the Heuft Delta FW 16 segmented
rejecter system with a single rotary servo driven, three-dimensional geometric shape. The
driving force behind the concept is the need to simplify the actuation system of the
process. The Heuft system incorporates 16 independent pneumatic cylinders, which as
previously noted can cause maintenance and timing issues. The camoid laner incorporates
a single actuator which has the potential to greatly reduce the complexity and increase the
reliability of the system.
The operation of the camoid system follows the same logic as the Heuft Delta
FW; a binary system in which a contour is laid out in between two bottles such that there
is no active translation of bottles. The bottles simply follow a new static contour that
diverts them to a new lane. An overhead view of the system in operation is shown in the
photo sequence in Figure 18 and the operation sequence is guided in a step by step
manner.

20
Figure 18: Camoid Extension
A. Bottle 1 passes by the camoid. Bottle 2 approaches.
B. Both Bottle 1 and Bottle 2 are passing the camoid.
C. Bottle 1 continues to Lane 1. Bottle 2 is still in contact with the Low Dwell of
the camoid. Bottle 3 enters the beginning of the camoid.
D-F. The cam rotates, with Bottle 3 riding the High Dwell while Bottle 2 is riding
the Low Dwell simultaneously.
G. Bottle 2 continues to Lane 1. Bottle 3 is riding the High Dwell of the camoid.
H-I. Bottle 3 continues to Lane 2.

Once the contour is laid out, the camoid will remain in the diverting (on) position
until a given number of bottles passes, and then the camoid will begin to spin and the
camoid will return to the off position. The camoid will remain in the off position for a
given number of bottles, then repeat the process described above.
The operation is very similar to the Heuft Delta FW, but the design offers
improvements that will positively affect the reliability of the system. In this chapter the
camoid laner mechanism will be explained in a manner to give a basic understanding of
the concept and introduce some terminology before detailed investigation into the design
process ensues in the proceeding chapters.

21
5.1 - Camoid Geometry
5.1.1 - Camoid Concept
The heart of the camoid laner is the camoid geometric shape. By definition a
camoid is a two degree-of-freedom, three-dimensional cam. Two degrees of freedom
means that the shape can cause motion in two directions. Putting this into perspective for
a bottle laner, the two degrees of freedom can be described as follows:
First degree of freedom: The bottles traveling down-line on the conveyor, x-
direction
Second degree-of-freedom: The bottles being diverted across the conveyor,
y-direction

5.1.2 - Camoid Inspiration

Heuft currently has several rejecters on the market as explained in the background
section (Chapter 2), and one of which sparked the idea to use a camoid for the laner
design. The Heuft Delta K shown in Figure 19 is similar to the Heuft Delta
FW 16 system used on the lines currently; however, instead of linearly translating
segments, this system uses rotating fingers. The timing would be similar to the Heuft
Delta FW; just the actuation motion would be changed.

Figure 19: Heuft Delta-K Bottle Rejecter 9

As stated before, the driving force of the idea is to reduce the number of actuators.
The camoid design stems from the concept of the Heuft segments and timing wrapped
about a single axis, effectively controlling timing, translation and contour in a single
geometry provided rotation can be controlled accurately.

9
Figure retrieved from Heuft USA

22
5.1.3 - Camoid Design

Figure 20: Camoid Laner Shape

The camoid laning geometry is basically the Heuft segments physical nature and
actuation control logic wrapped about a single axis. This is easier to understand by
examining the Heuft geometry and timing in detail. In the following sections, the process
for generating the complex geometric shape shown in Figure 20 will be
explained in basic terms providing the necessary background and terminology for
understanding of the more detailed analyses in the proceeding chapters.
Figure 21 shown below is an explanation of the terminology used in this
section.

23
 Figure 21: Geometry Terminology

5.1.3.1 - Wrapping Segments Around an Axis

Its easiest to understand the shape if it is explained as a series of cross sections
acting like cams with a common rotational axis (camshaft). Each segment of the Heuft
Delta FW system can be modeled as a double dwell two-dimensional cam with the cams
high dwell to be equal to the segment displacement and low dwell to be equal to the base
circle. This is easiest to explain with the cam segment graphic shown in Figure 22.

24
Figure 22: Cam Segment Construction
The cam segment construction is actually quite simple. It consists of two
concentric circles, the base circle and cam rise circle. The base circle is representative of
the Heuft off position, i.e. all segments not actuated. The cam rise circle is representative
of the Heuft on position, its radius equal to the Heuft segment length. The cam rise circle
whose chordal length is related to a high dwell angular displacement. The importance of
this value will be discussed later in the report in Chapter 7. A tangent line, labeled in
Figure 22, is extended from the base circle to either extreme of the cam rise circle
section.
If all 16 cams are stacked along the shaft, the resultant shape long drum cam with
each successive cam rise making the contour of the Heuft segments. To do this, the base
circle of each cam segment remains constant, while the cam rise circle of each cam
section represents the length of the corresponding Heuft segment. If we were to spin this,
the contour would rise and fall; however each cams rise would occur simultaneously.
From researching the Heuft timing in the previous chapter, we know that the segments
fire in sequence between two bottles, they do not actuate simultaneously. The next step
with the camoid laner geometry is to sequence the rise such that it resembles the Heuft
sequence and allow each cams successive rise to occur between two bottles.

25
5.1.3.2 - Sequencing the Cam Rise
The main goal is to cause the rise of each successive cam segment to occur
between two bottles. If the bottles were not moving parallel to the axis of rotation, the
cams could rise simultaneously and remain rising between the two stationary bottles. As
we know, this is not the case. In order to cause the cams to rise in succession, each
successive cam is out of phase to the previous cam some angle in the direction opposite
to the rotation direction of the camshaft. This phase shift is shown in Figure 22. This
causes each successive cam to delay its rise during rotation of the camshaft, provided the
cam shaft is spinning at constant rotational velocity. Figure 23 is a representation of
what the shape would look like at this point in the explanation. Notice that the cam
segments rise wraps around the cam shaft axis in a helical manner. It is this helical shape
that allows the cam segments to sequentially rise between two moving bottles during
constant cam shaft rotation.

Figure 23: Segmented Camoid

Before the explanation goes further it is important to note that the rise profile
(tangent lines connecting high and low dwells) of the cam segments is in no way involved
in manipulating bottles. The bottles are only in contact with the cam segments high and
low dwell surfaces. It is the succession of rising cam segments which creates the profile
by which the bottles are diverted.

5.1.3.3 - Smooth Diversion

Both the basic shape of the camoid and basic functionality behind the shape have
been described thus far. However, the resolution of the contour is poor with sixteen
segments, especially if the segments are not blended together, as they are shown in
Figure 23. This resolution is crucial as the bottles will be in contact with this section of
the cam and an unblended surface could cause damage to bottles. There are two criteria
that must be met in order to create a smoothly rising contour:
a) The thickness of each cam segment must approach zero
b) The successive cam segments must rise according to a function that provides
satisfactory results in the output bottle motion (acceleration, velocity, etc.)

26
 As the cam segment thickness approaches zero, the resolution of the contour
increases, creating a smoothly blended surface that would resemble the shape shown in
Figure 20.
The rise succession of the Heuft Delta FW segments does not provide the
resolution needed for the contour design; therefore it is necessary to provide a function
that governs the successive rises of the cam segments. Since the bottles will be being
diverted by this contour, it is important to make sure that the contour will not cause any
unwanted or dangerous forces on the bottle as it is being diverted. One method to
mathematically ensure an effective contour is to use cam program design.

5.1.3.4 - Contour Design

If we treat the bottle as a follower and the contour as a cam rise profile, we can
use cam design methodology to ensure that the displacement, velocity, acceleration, jerk
and force caused by the contour are all satisfactory. For a baseline of comparison, the
Heuft contour was reversed engineered and then compared to several different cam
programs to find the must effective contour. The cam design is explained in more detail
in the following sections.

5.1.3.5 - Broad Details of Segment Phase Shifts

The segment rises are controlled in a manner such that the bottles are gently
diverted, it is crucial to time the segment rise sequence effectively. This sequential
introduction of cam segment rises is cause by the angular offset, or phase shift, of each
segment to one another. There are several things to remember when interpreting the
phase shift of the cam segments:
a) The phase shift must be large enough such that the contour rises only between
two bottles
b) It must be large enough to provide a point in camshaft rotation at which all
segments are at high dwell, i.e. a continuous strip of contour for bottle
diversion (this will be explained in the next section, Controlling the Servo).
c) It must not be too large to allow the contour to fall between two bottles.

27
 There are only 360 degrees to work within for the phase shift that will allow both
a cam segment rise and fall within one rotation. Notice in Figure 24 the phase
shifting can be clearly seen by the curve dotted in green. Note that if there were no phase
shifting, the dots would make a straight line along the camoid axis, not a curve.

Figure 24: Cam segment phase shifting

5.1.3.6 - Bottle Contact Strip

An important criterion constraining the phase shifting is the necessity of a period
in cam rotation at which all cam segments are in high dwell. The reasoning behind this
can be explained by the nature of the concept. The system must be binary and the camoid
must remain in one position while a given number of bottles are being diverted. This
requires a section of the camoid to offer a continuous section of contour at a point in its
rotation. An example of the strip is shown in Figure 25. Notice that there is a strip on
both the low dwell and the high dwell of the camoid, representing no diversion on the
low dwell and positive diversion on the high dwell of the bottle respectively.

Figure 25: Sections of Continuous Contour

28
 Correctly designing the phase shifts will result in a contour strip of satisfactory
width. However, it is a delicate process as different parts of the camoid geometry can be
drastically altered with a small change in phase shift. These variations will be discussed
in detail later in the report as there are actually several geometric variables that can be
adjusted to create a satisfactory contour strip. For the purposes of this section we will not
discuss these details.
This strip actually serves two purposes, one of which is offering the continuous
contour. In the following section we will see how the contour strip plays a significant role
in the timing of the rotational actuation.

5.2 - Actuation Design

If the camoid geometric shape is the heart of the camoid laner mechanism, then
the actuation system is the brain. As stated previously, the actuation of the camoid laner
is accomplished by a servo motor. There are several key reasons as to why the servo
motor was chosen over other actuation methods and in the following subsections of this
chapter, these reasons will be introduced. Again, greater detail and analysis behind these
decisions can be found in later chapters; however it is necessary to have a basic overview
understanding of the actuation system and terminology before any detailed explanation
could be interpreted. Topics of investigation of the actuation system in this chapter
include:
Why use a servo?
Choosing the servo
Sizing the servo
Controlling the servo

5.2.1 - Why use a Servo?

As discussed earlier in the chapter, the nature of the camoid operation calls for
several required actions during operation:
The actuation of the system is rotation
During laning process, the rotational velocity of the camoid must be constant
The camoid must remain at both high and low dwell (no rotation) for a given
number of bottles to pass

There are also several task specifications that must be adhered to:
The mechanism must be able remain in either on or off position (high or low
dwell) for extended periods of time
The mechanism must accommodate for variable line speeds
The mechanism must be frequency adjustable, i.e. change the number of
bottles to pass

From these constraints we can also deduce further requirements of the actuation system:
The system must accelerate to a full rotational operating velocity between two
bottles
The system must decelerate as quickly as possible

29
 The system must be adaptable for actuation start times based on a number of
inputs including
o Bottle velocity
o Bottle position
The operational rotational velocity must be a function of the line speed, since
the mechanism must adapt to different line speeds in real time
The system must offer high torque outputs due to high acceleration
requirements

Based on the requirements of the actuation system, it is apparent that the actuation
must be under careful control. From the discussion about servo motors in the previous
chapter we can see how a servo can fulfill the requirements of this application. Recall that
servo motors offer:
Fast response time
Position control
Capability to hold a fixed position
Velocity and acceleration program control
Rapid acceleration and deceleration capability
Tight toleranced constant velocity even under dynamic loading
High torque capability

To achieve the level of control needed by a single rotational actuator, the servo is
an excellent candidate.

5.2.2 - Choosing a Servo

Gallo currently is installing an increasing number of servo motors as more precise
automation, higher quality and easier production operation is demanded by the company.
For this reason, Gallo works closely with a well known and trusted servo automation
company, Rockwell Automation. Rockwell provides all the necessary hardware and
software to power and control the Allen-Bradley servo motors that Rockwell distributes.
Choosing a servo brand to use in this case was dependent on the sponsor company
preference.

5.2.3 - Sizing a Servo

Choosing an appropriately sized servo is an important task and care must be taken
in doing so. In this case of high accelerations, a servo motor with sufficient torque is
necessary. Choosing a servo with insufficient torque will result in the mechanism not
accelerating in time, potentially causing damaged bottles. Over-sizing the servo is
expensive and unnecessary. A large servo may also not have the acceleration necessary
because of high motor inertia.
In order to aid in the process of sizing a servo, Rockwell Automation offers
software that will provide a range of servos that would be applicable based on a number
of user defined constraints and timing criteria. This will be explained in greater detail in
later chapters.

30
5.2.4 - Controlling the Servo
The servo is controlled through a series of components that compose a closed
loop communication system. Information regarding all aspects of the motion is relayed in
real-time to adjust the output to accomplish the exact task demanded by the user.
Designing a control program for a servo requires attention to detail. After choosing the
correct servo and driver components, the general constraints on the motion of the servo
are:
Position
Operating Velocity
Acceleration
Start Time
Dwell Time

5.2.4.1 - Position
In the case of the camoid laner, the exact position of the servo is essential
information to provide the control logic as it is crucial in the successful operation of the
system. To accomplish this, the servo motor is equipped with a shaft encoder that feeds
back its position data to the control system.

5.2.4.2 - Operating Rotational Velocity

The contour must be laid out between two bottles, thus the rate at which the
contour rises is equal to the bottle velocity. This velocity is essential to successful
operation and must match the bottle velocity at all times, thus is denoted as the operating
velocity. The operating rotational velocity is calculated based on the linear velocity of the
bottles on the line.

5.2.4.3 - Acceleration
The servo needs time and distance to accelerate and decelerate the camoid up
to/down from the operating velocity. In order to accomplish this, both the geometry of the
camoid and the timing of the actuation must be adjusted. Recall that in order for
successful diversion of the bottles, the camoid geometry must incorporate a continuous
strip of contour running its entire length. Not only does this strip act to provide a
continuous contour, but also provides the buffer distance for servo acceleration and
deceleration.

5.2.4.4 - Actuation Start and Dwell Times

The start of the actuation and the camoid dwell time are both matters that require
sensing bottle positions. The actuation must initiate between two bottles. The camoid
must dwell for an amount of time that is dependent on the number of bottles that it
diverts. In order to accomplish these tasks, the servo controller must be aware of the
bottle positions and how many bottles have passed the laner. To do so, a photoeye is
placed on the line that reads when a bottle goes past it. Alone the photoeye is capable of
counting bottles as they pass. When coupled with the shaft encoder, the two instruments
are capable of sensing a bottles position as it moves down the line. With this information

31
input to the servo controller, the actuation initiation and the dwell timing are able to be
controlled.

5.2.5 - Actuation Design Conclusions

The properties of a servo make it an ideal candidate for the motion application of
the camoid laner. With the servo, all aspects of the motion can be controlled in detail and
adjusted in real-time to accommodate a wide range of scenarios that could be experienced
out on the line.

5.3 - Design concept summary

Coupling geometric design with proper motion control, the camoid laner is able to
accomplish with one actuator what the Heuft Delta FW does with sixteen actuators. Not
only is the number of actuators fractioned, but the reliability of the actuation method is
increased as well. This chapter was intended to give an overview of the concept behind
the design and to introduce some of the terminology that can be expected in the following
sections. In the next chapters the camoid design will be explained in detail with
mathematical analysis to support the decisions made in the design. Each section will
include a brief introduction to the specific aspect of design to be explained; however, if
there is any confusion before reading the analysis, this chapter should be referenced.

32
Chapter 6.0 - Methodology
This chapter outlines the steps that were taken throughout the design process and
present the process in a logical order that could be replicated. The actual methodology
involved simultaneity and iteration.

6.1 - Assumptions
A number of assumptions were made throughout the design phase of this project.
Such assumptions are necessary to prioritize different elements in both the problem
definition and the derived solution. Some of the most relevant assumptions are briefly
discussed here, along with any verification that the assumptions were valid.

6.1.1 - Line 2 Properties

The process of splitting bottles is heavily dependent upon an accurate kinematic
model of a stream of bottles, including their relative positions, velocities, and
accelerations. The relation between bottle and conveyor is also of great importance.
Considering this, an average set of conditions was taken as the design baseline, with all
analyses and decisions made in reference to these conditions. The design should allow
for a safe range of conditions around the average.
Line 2 runs at a rate of 400 bottles per minute when operating correctly. This
equates to a conveyor roughly 1 meter per second. Based on accounts from Gallo
employees and video footage, it was determined that wine bottles on Line 2 were in static
friction with the conveyor at the point of the lane split. An assumed minimum bottle
spacing (the distance between two adjacent bottles) of one bottle diameter was agreed
upon, again through the advice of Gallo engineers. These conditions describe bottles
with no relative velocity, neither to each other nor to the conveyor, and with velocity
equal to the line speed relative to the stationary splitter.

6.1.2 - Heuft Laner

The direction of the redesign was largely defined upon making one assumption in
particular; the laning system currently used by Gallo is successful at its function of
diverting bottles, but fails to achieve the desired lifetime. Prior to making this
assumption, the design team had looked at alternative methods of achieving a laning
diversion. With this assumption, however, the focus shifted to finding alternative means
of driving the existing system to improve lifetime.
The shift in focus was significant. It made analyses on tipping or breaking bottles
less important; if the Heuft Laner did not damage bottles with its current means of
actuation (pneumatic cylinders), it follows that the bottles would remain undamaged if
the actuation was replaced, but mimicked closely enough.

33
6.2 - Data Collection and Analysis
6.2.1 - Heuft Analysis
The design requirements of this project are to improve upon the existing design
for high speed lane splitting. In order for this to be accomplished, the existing system
must be analyzed and understood. The first step taken was to understand key aspects of
the operation of the Heuft laning system. The method by which the mechanism functions
is explained in the background chapter of this report.
The next step in understanding the Heuft laning system was to analyze the
contour by which the bottles are diverted. An understanding of the function of the
contour is needed as a baseline of comparison for other contours.
The Heuft contour was measured and a best fit polynomial equation was fit to the
points measured. The contour can be understood mathematically allowing kinematic and
dynamic analyses can be conducted.

Heuft Contour and Approximate Equation

3.000
2.500
Crossline Stroke (in)

3 2
y = 0.0011x + 0.0166x + 0.0362x - 0.0007
2.000 2
R = 0.9999
1.500
1.000
Heuft Contour
0.500
Approx
0.000
-0.500 0 1 2 3 4 5 6 7 8 9 10
Downline Travel (in)

Figure 26: Heuft Contour and Equation

6.2.2 - Contour Design and Analysis

The most important aspect of the design is that no bottles are tipped or damaged
during the laning process. In order to improve on the existing system, cam program
design is used to develop several contours with controlled acceleration profiles. Based on
standard cam design methodology 10, we chose three different cam programs to use as a
base for optimization of the existing system. The three cam programs of interest are as
follows:
Simple Harmonic Displacement
Modified Trapezoid Acceleration
4-5-6-7 Polynomial Function Displacement

Each of the programs is designed following some basic criteria that require
specific aspects to be constrained.

10
Norton, Robert L. Design of Machinery. New York: McGraw Hill, 2004.

34
6.2.2.1 - Simple Harmonic Displacement Constraints
Constraints are set on the final value of displacement.

6.2.2.2 - Modified Trapezoidal Acceleration Constraints

Assumptions are made of the maximum allowable acceleration.

6.2.2.3 - Polynomial 4-5-6-7 Function Displacement Constraints

By adjusting the values of the 8 constants, constraints can be set on the initial and
final position, velocity, acceleration and jerk (boundary conditions). Polynomial cam
contours offer accurate control of the dynamic conditions at the selected positions,
however the displacement function is governed by the boundary conditions and is often
difficult to control maximum values.
After specific contours are designed and plotted, they are compared against the
existing system, and analysis is conducted on each contour.

6.2.2.4 - Kinematic Analysis

Kinematic analysis includes comparison and discussion on each of the cam
displacement, velocity, acceleration and jerk profiles. The three proposed profiles are
directly compared with the current Heuft system and are also compared with each other.
Topics of investigation include:
Difference compared to Heuft
Abnormal peaks
Maximum values
Minimum values

6.2.2.5 - Force Analysis

Force analysis was conducted with the same criteria listed above in kinematic
analysis except analyzes the force and moment induced by the contour on the moving
bottle. In addition, the force and moment were compared to the friction force between the
bottle and conveyor making sure that the contour will induce enough force to overcome
static friction fast enough to minimize tipping potential.

6.2.3 - Choosing Contour

The most suitable contour is chosen based on several important criteria:
Acceptable Final Velocity
Minimizing Maximum Acceleration
Minimizing Jerk
Maximum Force
Maximum Induced Moment on Bottle

The last two criteria are most important as both are involved directly with
damaging and tipping bottles. Taking each of these criteria into consideration, a contour
is chosen that best suits high speed bottle laning.

35
6.3 - Camoid Geometry Design
The design of the camoid geometry is an iterative process. There are several
elements of the camoid that are manipulated to generate the complex geometry. The
elements are interdependent; changing one element affects the others, which in turn
affects the entire geometry of the camoid. The elements of the geometry that are
manipulated are:
Profile Rise Contour
Cam Phase Angles
High Dwell Angle Displacement
Low Dwell Angle Displacement
Buffer Angle
Base Circle Radius
Overall Length of Laner

This detailed design process can be found in detail in Chapter 7 of this report.

6.3.1 - Manufacture
It was agreed early that the material of the camoid will be some sort of plastic for
its characteristics in:
Wear resistance
Oxidation resistance
High strength to weight ratio
Low weight
Forgiveness for bottles during impact (relative to metal)
Manufacturing flexibility
Low cost

The geometry of the camoid is complex and could prove to be difficult to

manufacture for conventional methods of machining. Furthermore, the geometry must be
relatively accurate as there are tight tolerances of certain elements. There are few options
available for manufacturing complex geometries:
CNC Machining
Rapid Prototyping
Injection Molding

Each method offers advantages and disadvantages shown in Table 1.

36
 CNC Rapid Prototyping Injection Molding
Adv. Disadv. Adv. Disadv. Adv. Disadv.
Limits of
Inexpensive geometric Inexpensive for
Complex Expensive for
(relative to complexity due to Expensive large quantity
Geometries single part order
other methods) tooling and nature orders
of process
Limits of
geometric
Large Material Material Large Material
Fast turnover complexity (mold
Variety Limitations Variety
still must be
machined)
Many Availability
companies Issues
Table 1: Manufacturing Method Comparison

Due to the complex geometry of the design, the most logical choice for
manufacturing was rapid prototyping. However, there are several different methods for
rapid prototyping that offer a set of advantages and disadvantages. The rapid prototyping
comparison can be seen in Table 2.
Fused
Stereolithography Selective Laser Electron Beam
Deposition
(SLA) Sintering (SLS) Melting (EBM)
Modeling (FDM)
Adv. Disadv. Adv. Disadv. Adv. Disadv. Adv. Disadv.

Material
Most Slower than Less Less
Produces Material variety Immature
commonly other RP accuracy accuracy
robust parts Variety (same as Technology
used methods than SLA than SLA
SLS)
Non porous,
Parts not as Secondary Secondary homogenou
Material Porous
Accurate robust as curing curing s, robust Expensive
variety parts
other needed unneeded parts
(titanium)
Faster and
Secondary
more
curing
efficient
needed
than SLS
Support Support
structure structure
may be may be
needed needed Accurate
during during
production production
process process
Table 2: Rapid Prototyping Method Comparison

After researching the methods of the rapid prototyping the search was narrowed
down. For the purpose of the camoid laner, SLS is not appropriate because the part
cannot be porous due to the moist environment and the part must be smooth. EBM is too
expensive and a titanium part is unnecessary. This leaves SLA and FDM to choose from.
Stereolithography was the method of choice for several reasons:
Adequate material strength

37
 Accuracy is needed on several sections of the camoid, especially a keyway
Best price offer
Best turnover time

See Appendix B for a full list of prices and companies that were considered.

6.4 - Actuation Design

The timing and geometry criteria must be strictly adhered to for the successful
operation of this laner. Once this information is known, analysis can be conducted on the
rotational velocity, accelerations, and torques required to achieve successful timing.

6.4.1 - Sizing a Servo

Accurate sizing of the servo is important because of the precision needed in this
application. An undersized servo will not create the accelerations needed to reach
operating velocity in the buffer angle given. An oversized servo is more difficult to tune
and is more expensive to purchase and operate.
Allen-Bradley motors are Gallos trusted motor company and currently offer free
software 11 to aid in the sizing of A-B servo motors to fit specific applications. The
Rockwell Motion Analyzer software is used to accurately size the servo to this camoid
application.
The software accurately sizes a servo based on several input values including:
Voltage Supply type and nominal value
Velocity-Time cycle profile (timing diagram)
Mechanism data
o Inertia Values
o Starting angle
Transmission data
o Type (belt, spur, etc.)
o Ratio
o Efficiency
o Friction
Once these values are input, the software searches an internal database of the A-B
servo product line and produces a range of motors that fit the criteria entered. Motor data
accompanies the motor models including, but not limited to:
Peak velocity
Peak and RMS torque
Current Draw
Gearbox ratio
Relative Cost (to motors in search results)
Relative Performance (to motors in search results)

11
Rockwell Automation. Motion Analyzer Software. Retrieved February 11, 2007 from
www.ab.com/motion/software/motion_analyzer.html

38
 The motor list can be arranged according to the users desire. The software also
runs graphical torque analysis comparing the torque-speed diagram to the user required
torque.
The software offers a detailed, motor model specific comparison allowing
accuracy and confidence when attempting to size the servo to the application.

6.5.2 - Controlling the Servo

Along with camoid geometry, the control logic of the servo is the most important
part of successful operation of the camoid laner mechanism. There are several pieces of
hardware and software required to get from the control logic concept to the output motion
of the servo. These elements are:
1. Control Logic Software
2. CPU Interface Module
3. Servo Driver
4. Servo
Together these elements make a closed loop information chain to create and
adjust the servo motion exactly to specifications and real-time data feeds. These elements
will be explained in greater detail in Chapter 7
It should be noted that the control logic is designed in specific software
offered by Rockwell Automation. The logistics of the software and the programming
protocol 12 will not be explained in this report as Gallo employs specifically trained
engineers for servo controls. However, the servo actuation sequence will be explained
and how it is directly related to the camoid geometry.
The specifications for the servo control program was written by the camoid design
engineers as it is important to follow the correct criteria that Gallo demands for servo
motion control. There are several safety issues concerning servo operation and must be
included. For this reason, the actual servo control program writing was outsourced to
David Booth of Gallos Controls department.

6.5.2.1 - Assumptions
The camoid must begin spinning such that the contour rises and remains rising
between two bottles.
The camoid must be accelerated to operating rotational speed as fast as
possible.
Operating speed is directly proportional to the line speed and is such that the
camoid must complete its rotation cycle in the time that one bottle travels the
camoid length.
The cam must decelerate as fast as possible.

6.5.2.2 - Timing
1. From rest (0) position spin to high dwell at operating speed
2. Remain at high dwell for a bottle count (10 bottles)

12
Information regarding this program logic can be found at
www.rockwellautomation.com/rockwellsoftware/design

39
 3. Spin from high dwell to position 360 at operating speed
4. Remain at low dwell for a bottle count (10 bottles)

There are several aspects to keep in mind during this timing sequence:
a) The servo needs time and distance (angular distance) to accelerate the camoid
up to operating velocity
b) The servo needs time and distance to decelerate the camoid from operating
velocity
c) The operating velocity is directly proportional to the line speed (which may
change at any time)
d) The amount of time the camoid remains in dwell is dependent on a number of
bottles (which may be non-uniformly spaced)

Each of these aspects are worthy for investigation in this section to provide
baseline understanding of the program details explained in the following chapters.

6.5.2.3 - Operating Velocity

As stated before, the successful operation of the camoid laner relies on the camoid
being rotated at a constant velocity. The velocity at which it rotates is directly related to
the line speed, as the contour must rise in between two bottles on the line. However,
because the bottling line is subject to frequent shut downs and decreases in speed, the
operating velocity of the camoid laner must be able to accommodate such instances.
In order to accomplish this task, the servo controller must always be aware of
the line speed. To do so, a shaft encoder attached to the conveyor line feeds line speed
information to the servo controller. The operating rotational velocity of the camoid is
directly related to the line speed by a mathematical function. Once input into the
controller, the servo output velocity can be adjusted in real-time to accommodate the
dynamic conditions of the bottling line.

6.5.2.4 - Servo Acceleration and Deceleration

The servo needs time and distance to accelerate and decelerate the camoid up
to/down from the operating velocity. In order to accomplish this, both the geometry of the
camoid and the timing of the actuation must be adjusted.
Allowing a buffer distance for the servo to accelerate and decelerate is a matter
that concerns the camoid geometry. Recall that in order for successful diversion of the
bottles, the camoid geometry must incorporate a continuous strip of contour running its
entire length. Not only does this strip act to provide a continuous contour, but also
provides the buffer distance for servo acceleration and deceleration.
The strip acts as a buffer because of the fact that the camoid neither rises or falls,
i.e. there is no bottle diversion. This means that the camoid can rotate a certain angle
amount, the buffer angle for explicative purposes, without a bottle being manipulated.
This becomes necessary as the servo can use this buffer angle to accelerate and decelerate
after it has completed its operating cycle.
Adjusting the timing to allow for servo acceleration and deceleration is a matter
of starting the actuation somewhere within the buffer angle. The buffer angle will allow

40
the servo to accelerate up to the necessary operating velocity without any bottle
manipulation.

6.5.2.5 - Torque
Based on the acceleration necessary and the geometry of the cam, the necessary
torque to achieve the desire effects is calculated. The cam mass moment of inertia must
be known for this calculation and is obtained using 3D modeling software due to the
complex camoid geometry.

6.5.2.6 - Actuation Start and Dwell Times

The start of the actuation and the camoid dwell time are both matters that require
reading bottle positions. The actuation must initiate between two bottles. The camoid
must dwell for an amount of time that is dependent on the number of bottles that it
diverts. In order to accomplish these tasks, the servo controller must be aware of the
bottle positions and how many bottles have passed. To do so, a photoeye is placed on the
line that reads when a bottle goes past it. Alone the photoeye is capable of counting
bottles as they pass. When coupled with the shaft encoder, the two instruments are
capable of knowing a bottles position as it moves down the line. With this information
input to the servo controller, the actuation information and dwell timing are able to be
controlled.

6.5.3 - Stress Analysis

An important consideration in the camoid analysis is stress induced from the
torque. Since the torque must be transmitted to the camoid, the method used will be a key
and keyway. Several different stress analyses are conducted:
Shaft Shear Stress
Torsional Deflection
Keyway Stress Concentration
Safety Factors with different materials

6.5.4 - Fatigue Stress Analysis

The nature of the splitting operation is such that the camoid is subject to cyclic
application of torque to accelerate and decelerate to and from operating velocity. Thus, it
is important to conduct fatigue stress analysis. The fatigue strength of the shaft is
calculated to ensure adequate lifetime.

6.6 - Part Gathering

An important part in this design process was finding the correct parts needed to
build the prototype. A full bill of materials can be found in appendix C

6.7 - Chassis Design

The last step in the process is to design the frame, or chassis, that will house the
system. There are several points of interest of the design that warrant discussion:
Material

41
 General layout of components
Implementation on bottling line
Manufacturability

6.7.1 - Material
There are several criteria that the material must meet:
It must not oxidize
It must not be toxic
It must be structurally sound

The must common structural materials used in Gallos bottling lines are steel,
stainless steel and aluminum. Each material offers advantages and disadvantages which
are shown in Table 3.
Steel Stainless Steel Aluminum
Adv. Disadv. Adv. Disadv. Adv. Disadv.
Resists Resists
Inexpensive Oxidizes Expensive Expensive
Oxidation Oxidation
Requires
Widely Less Strength and
Surface Availability
Strong accepted for Toughness
Finishing Issues
food handling (relative to steel)
(paint)
Tough Strong
High
Tough
Availability
Table 3: Material Comparison

An important issue to keep in mind is that the chassis will be in a section of the
bottling line in which there are open containers of product. The environment is wet from
spilt product, line lubrication, and cleaning activities. It should be noted that Gallo
requires all steel parts to be painted with a specific grade paint to resist oxidation of the
steel and chipping.

6.7.2 - Component Layout

The layout of the components is important as it determines the overall size of the
mechanism as well as the layout of the transmission for the camoid. The layout for the
mechanism can be seen in Figure 27.

42
 Figure 27: Component Layout
There are several ways to drive the camoid; one idea was to directly couple the
camoid shaft with the servo shaft. However, the chosen layout is advantageous for
several reasons:
Minimizes overall length of the system
Fits within the footprint of the Heuft Delta W 16 system
Keeps motor away from conveyor line
Allows easy motor access for maintenance/replacement

6.7.3 - Implementation on Line

The chassis design must be designed such that the implementation on the bottling
line is accomplished easily. The nature of the design is such that it is a direct replacement
of the Heuft system. It will utilize the same method for attachment that the Heuft
currently uses. Figure 28 shows the current method by which the Heuft system is attached
to the line. Notice that the entire system is independent from the line. It is simply bolted
to the side of the conveyor on the front edge and supported in the back by a single leg.
The camoid laner chassis will be designed to be implemented on the line in a similar
fashion to the Heuft system. This implementation method is advantageous as it allows for
easy initial installment and easy replacement if needed.

43
Figure 28: Heuft Attachment to Line

6.7.4 - Manufacturability
The chassis is a simple, easily manufactured design consisting of several plates
fastened together with bolts and supported by several brackets ( Figure 27). The most
difficult part of the manufacturing process will be drilling the bearing mount holes to fit
the tolerances needed for the shaft to spin freely. Fully toleranced technical drawings
were provided to the machining company.

6.8 - Drive System

There are many different power transmission systems commercially available. In
order to narrow down the options, the design of the drive system must be constrained.
The power transmission from the servo to the camoid needs to be a robust system that
meets several criteria:
positive transmission
low backlash
low friction
low inertia
smooth/quiet operation

The easiest method for positive transmission with low backlash is to incorporate
teeth into the transmission. Spur gears and chain drives were the first options explored,
however either option could present problems and not meet the desired criteria.
The cam shaft and servo shaft are relatively far apart as the distance between must
clear the high dwell of the camoid and the servo body. This distance would require two
relatively large diameter spur gears, which would increase the inertia. Furthermore, the

44
spur gears could cause unwanted noise. A chain and sprocket drive could be used;
however this could also cause unwanted noise and high inertias.
Gallo employs a large number of transmissions on the bottling lines to drive
power from a motor to a conveyor. On newer lines, a specific type of toothed belt and
sprocket system is used, similar to an automobile engine timing belt. After consulting Mr.
Loel Peters, the benefits of this transmission became apparent; ability to handle high
torques, offer low backlash and availability in Gallos storeroom.
The Goodyear Eagle Pd Belt and sprockets incorporate a double herringbone
tooth pattern which prevents both tangential and lateral belt slip (off the side of the
sprocket). The system met all criteria set forth, provided immediate availability and was
an economic solution. Furthermore, Gallo trusts the system enough to employ it
elsewhere in the plant.

6.9 - Mechanism Assembly

The mechanism is fastened with hardware obtained from Gallos storeroom. Once
built, the mechanism was subject to an initial motion test at a simulated 600 bottles per
minute for one hour to assure the mechanism was functional. During the test,
observations were noted of any abnormalities such as:

Immediate Wear
Noise
Heat due to friction
Loose parts

Tests proved to be successful and no readjustment was necessary.

6.10 - Implementation
Implementation on the line is under the supervision of Jason Elliot of Aubry
Construction Company. The line used to test the laner is a test loop previously installed
by Gallo employees. The loop is open to allow for testing of new equipment.

45
 Figure 29: Test Loop
The test loop rails were adjusted to our design specifications. The test loop
includes all necessary hardware including photo eye and shaft encoder.
No housing for the electronics was constructed; therefore all electronics must be
portable and temporary as Gallo safety rules do not allow uncontained electronics,
especially those operating on 480V. The servo driving system was attached to a flat plate
and carted to the test loop. It was removed immediately after the testing and never left
unsupervised.
David Booth supervised the setup of the electronic systems assuring all safety
guidelines are followed.

6.11 - Testing
Testing was conducted following a detailed test protocol found in Chapter 9.

46
Chapter 7.0 - Camoid Laner Detailed Design
The detailed process of the laner design is discussed in this chapter. All final
values are either given or referred to in the appropriate appendix.

7.1 - Camoid Geometry Design

The camoid laner is based on the wrapping the Heuft Delta FW segment lengths
and actuating time around a rotating axis. If each segment of the Heuft was represented
by a cam, the segments linear translation would be the cams rise which would occur by
rotating the cam. The series of the cams are attached to a common shaft with progressive
rises that follow a contour similar to the Heuft segments. If the phase of each successive
cam is shifted it is possible to cause the similar effect of laying the contour in front of the
bottle that the Heuft accomplishes by sequentially firing the segments in front of the
bottle. Finally, if the thickness of the cams approaches zero, the contour will be a
smoothly blended surface.
In order to accomplish this task there are several elements of the geometry that
must be designed simultaneously to offer the correct output motion, provided the actuator
can be precisely controlled for angular displacement and angular velocity. In the
following section the design of the geometry will be explained in detail.
Note that the design process was conducted in PTC ProEngineer Wildfire 2.0.

7.1.1 - Iterative Design

Arguably the must crucial aspect to the successful operation of high speed laning
is the timing at which the diversion occurs; timing tolerances can be as tight as several
milliseconds. In the case of the Heuft laner, the independent linear pneumatic actuators
are fired in sequence governed by the line speed. In the case of the camoid, the timing is
based not only on the line speed, but also the rotational actuation and geometry. Since the
timing is directly related to the geometry, careful analysis is conducted to assure that the
cam behaves properly. This design process is iterative and occurs simultaneously with the
rotational timing design (explained in the next section). The elements of the geometry
that are manipulated for the design process are:
Diverting Contour
Cam Phase Angles
High Dwell Angle Displacement
Low Dwell Angle Displacement
Buffer Angle
Base Circle Radius
Overall Length of Laner

These elements are represented below by Figure 30, but will also be individually
represented in the following sections.

47
Figure 30: Geometric Elements
Each of these elements is manipulated and affects different aspects of the timing
and bottle diversion.

48
7.1.1.1 - Diverting Contour
The diverting contour is defined by
the successive rises of each cam segment.
This is the contour by which the bottles
are guided. Each cam segments rise is
governed by the cam programs explained
in the previous section.

7.1.1.2 - Cam Phase Angles

The cam phase angle is the angle
Figure 31: Diverting Contour
that each cam is offset relative to the
previous cam. This controls the rate at
which the contour is laid out in addition to
rotational velocity. This also affects the
value of rotational speed. If the phase
angles are larger, the contour will take
longer to be laid out, thus requiring a faster
rotational velocity.

7.1.1.3 - High Dwell Angular

Displacement
The high dwell angular
displacement is the angle of each segment
over which the high dwell spans. A crucial Figure 32: Phase Shift Angle
design criterion for the camoid is that when
the camoid is fully extended, every
segment of the camoid must be in high
dwell. Recall from Chapter 3 that this is the
contour strip. If the phase angles are larger,
and high dwell angular displacement is
kept the same, the strip gets thinner until it
no longer exists, creating discontinuity in
the contour. The high dwell angular
displacement is manipulated such that the
strip is kept wide enough to properly guide
a bottle during camoid dwell.

Figure 33: High Dwell Angular Displacement

49
7.1.1.4 - Low Dwell Angular Displacement
The low dwell angular displacement
is similar to the high dwell, except with the
low dwell of the segments. When the
camoid is fully retracted, the cam segments
must be in low dwell. This creates a strip of
low dwell flat contour parallel to the
camoids axis of rotation. This value is not
manipulated but rather serves as a limit to
the value of the high dwell angular
displacement.
As high dwell angular displacement
increases, low dwell angular displacement Figure 34: Low Dwell Angular Displacement
decreases. Care is taken to assure that the high dwell angle is small enough to allow
adequate low dwell angular displacement.

7.1.1.5 - Buffer Angle

The buffer angle is directly related to
the thickness of the contour strip, it is the
angle of rotation over which the strip
occupies. Because there are two contour
strips, there are two buffer angles. The buffer
angle at the high dwell is directly a function
of the cam phase angles and the high dwell
angular displacement. The buffer angle at the
low dwell is a function of cam phase angles,
high dwell angular displacement and high
dwell buffer angle. This is because the low
dwell angular displacement is dependent on
the high dwell angular displacement.
Care is taken to assure the buffer Figure 35: Buffer Angle
angle is adequate both for high and low dwell. If the buffer angle is too small, the
acceleration needed will be high and require a
significant amount of torque, increasing motor size
and thus expense.

7.1.1.6 - Base Circle Radius

The base circle radius affects the width of
the low dwell strip. If the base circle is too small,
the tangent line of the highest cam segment
approaches the tangent line of the lowest dwell
segment, effectively eliminating the low dwell. If
the base circle is too large, the cam will be too
large to fit on the bottling line. Furthermore, if the
base circle is too large, the contact point (strip)
Figure 36: Base Circle

50
will be too high on the bottle inducing a potentially high tipping force. Lastly, if the base
circle is too large, the part will be excessively massive and require higher torques to
accelerate.

7.1.1.7 - Length of Camoid

The length of the camoid affects
the acceleration and thus force caused by
the contour. If the camoid is too short, the
working contour will rise abruptly,
potentially causing high accelerations and
forces. If the camoid is too long, the
operating rotational velocity must increase
to successfully lay the contour between
two streaming bottles. This increase
causes rotational accelerations to increase,
thus increasing the required torque. With
increased torque, motor size and expense
increase as well as stress and fatigue on
the system. Figure 37: Camoid Length

51
7.1.2 - Geometry Final Values
The final values of the geometric elements are shown in Table 4. It should be
noted that the geometry of the camoid is a smooth blended surface; however it is broken
into 11 cross sections for purposes of discussion and computer modeling.
High Low
High Base
Contour Phase Dwell Dwell Camoid
Section Dwell Circle
Rise Shift Buffer Buffer Length
Number Angle Radius
(in.) (deg) Angle Angle (in.)
(deg) (in.)
(deg) (deg)
1 0.001 0 145 7.5 32.8 1.75 9.5
2 0.006856 12 140
3 0.052 12 135
4 0.159 12 130
5 0.333 12 130
6 0.574 14 130
7 0.883 14 130
8 1.259 14 130
9 1.703 15 130
10 2.21 17.5 125
11 2.758 17.5 120
Table 4: Camoid Geometry Final Values

7.1.3 - Geometry Construction

Because of the complexity of the geometry, it is not feasible to machine the part
using conventional cutting techniques such as CNC milling or lathe machines. The part is
constructed using a Stereo Lithography (SLA) rapid prototyping machine. After research
it was apparent that the process offered satisfactory tolerances and material strength.
Outsourcing the part for manufacture is a matter of emailing the correct format of the
model to the company.

7.2 - Camoid Prototype

Figure 38 is a picture of the comparison of the CAD model and the actual rapid
prototyped part.

52
Figure 38: Model vs. Rapid prototype

7.3 - Actuation Design

The servo system must provide accurate and precise actuation of the camoid
through millions of cycles. Attention to detail is important in the design of the servo
system. This detail is explained in this section.

7.3.1 - Servo Sizing

The servo is sized using Motion Analyzer software from Rockwell Automation.
For detailed tables conveying the values used for the motor sizing for this application see
Appendix D.

7.3.2 - Servo System

The servo system is set up to allow for completely automated control of the
actuation of the camoid based on line speed and bottle spacing. The system can adjust for
varying line speeds that are experienced during line start up and shut down. It can remain
in one position for extended periods of time. In order to provide the automated control,
there are several important components of the servo control system that warrant
discussion.
Servo Motor
Driver
CPU Motion Control Module
Control Program Logic

53
The basic setup for the servo control system is shown in Figure 39. The system is
adaptable allowing data inputs from a variety of different tools and also can be controlled
remotely via Ethernet. Furthermore, the system can be configured to control multiple
axes (servos) with a variety of different program profiles. The following explanation
outlines one particular setup. Note that this system meets Gallos standards and operates
on 480V 3 Phase AC electrical.

Figure 39: Servo Control Schematic

The following is a step-by-step explanation of the schematic shown in Figure 39.
Each of the numbers indicates the part of the control information transfer shown in the
figure. Note that there is a continuous loop of information and communication does not
necessarily always occur in this sequence explained.

54
 1. Control program software uploads control information to a CPU control module.
In this case the program is made with Logix 5000 software. The module
simultaneously feeds back information to the computer.
2. In addition to program logic from the software, information from sensors can be
input into the control module. In this case, a photoeye and shaft encoder will be
used to transmit bottle position and velocity to the module. It should be noted that
the control program must have the proper protocol for reading and interpreting the
data from these sensors. With this input data, the program can make adjustments
that answer to real-time dynamic variations in servo loading and/or timing
demands.
3. The CPU control interface interprets the information of the Logix 5000 software
and outputs a signal to the driver. The driver simultaneously feeds back
information to the module regarding servo statistics.
4. The driver interprets the information from the CPU module and adjusts the output
current to the servo motor. The servos internal encoder feeds back information to
the driver regarding position and velocity.

7.3.2.1 - Servo Motor

The servo motor was obtained from Gallos spare parts shelves along with the
necessary power and feedback cables. The servo motor from Gallo falls in the range of
applicable motors from the Motion Analyzer software. The motor included an attached
gearbox. Table 5 below is a basic overview of the motor and gearbox specifications. For
more detailed specifications see Appendix E.

Figure 40: Servo Motor

55
Manufacturer Allen-Bradley
Model MPL B4520 MJ22AA
Max Speed 5000 rpm
Continuous Stall Torque 6.1 N-m / 54 in.-lbs.
Power 2.5 kW
Gearbox Manufacturer Alpha
Gearbox Ratio 10:1
Table 5: Servo Motor and Gearbox Overview Specs

The servo motor includes an input and feedback cable connection. The power to
the servo is delivered by the driver. Inside the servo is an encoder that reads and outputs
the position of the servo through the feedback cable. The feedback is read by the driver
and the power is adjusted to deliver the necessary power to follow user input controls.

7.3.2.2 - Driver
The driver was obtained from Gallos control department from a decommissioned
servo driven project. The purpose of the driver is to interpret the control logic from the
computer and provide the correct electrical current to move the servo motor in the exact
manner specified by the computer control logic. The input signal can be either analog or
digital depending on the level of control demanded by the user. Analog is adequate for
systems with a few axes (less than 6). However, for more complicated systems, digital
interfacing is often used to reduce the number of connecting wires needed. Digital
systems, however, are significantly more expensive.

Manufacturer Rockwell Automation

Model 2098-DSD-HV150
Operating Voltage 480V AC 3 Phase
Peak Current (amps) 68
Continuous Current (amps) 34
Continuous Power (kW) 15
Table 6: Driver Overview Specs

7.3.2.3 - CPU Control Module

The CPU interface module is the component that interprets the control logic from
the computer software and outputs a readable signal to the driver. The input connection is
obtained via Ethernet cable. This connection is extremely versatile enabling remote
control changes over the internet. This is beneficial as it becomes possible for the control
engineer to manipulate servo controls from any location that enables internet access. The
modules used for this system were also obtained from Gallos storehouse.

7.3.2.4 - Control Program Logic

The human interface of the servo control is the Logix 5000 software, which can
be run on any PC with an Ethernet port (for data transfer). This software allows for the
ultimate control of the servo motor. Details such as timing, position, velocity,
acceleration, acceleration profiles and torque can all be controlled with fine precision. In

56
addition to servo controls, any number of safety and system check protocols can be input
into the system. For example, the servo can be told to reset itself to zero after a given
number of cycles to maintain system accuracy.

7.3.3 - Control Logic Concept

As stated before, the logistics behind creating the control program will not be
discussed in this report; however, a motion control outline is described below. There is a
detailed set of constraints and assumptions and a timing diagram.

7.3.3.1 - Assumptions
1. The actuation for contour rise must begin between bottles and be at full
operational velocity
a. The actuator is given a buffer angle, on the cam segments low dwell, in
which acceleration can occur.
2. The contour rise must occur at a rate such that it remains between bottles.
3. The actuation must stop and hold within the angle of the aforementioned high
contour strip.
4. The actuator must be held at this position for a data input number of bottles.
5. The actuation for contour fall must begin between bottles and be at full
operational velocity.
a. The actuator is given a buffer angle, on the aforementioned high dwell
strip.
6. The contour fall must occur at a rate such that it remains between bottles.
7. The actuation must stop and hold within the aforementioned low dwell contour
strip.

These timing criteria are the baselines that must be followed for the successful
operation of the camoid laner.

7.3.4 - Drive Train Specs

Final drive ratio 10:1

7.3.5 - Timing
1. From rest (0) position spin 164 at operating speed
2. Remain at high dwell for a bottle count (10 bottles)
3. Spin from 164 to position 360 at operating speed
4. Remain at low dwell for a bottle count (10 bottles)

57
Figure 41: Timing Diagram Example

Because the timing of the motor is constantly adjusted to match the line speed,
this timing diagram represents an example of the timing at a particular line speed. There
are several important characteristics of the timing diagram that should be noted.

1. The x axis is time in seconds, the y axis is the angular position

2. This diagram represents a full cycle of the camoid laner. A full cycle
includes:
a. Low camoid dwell, no bottles diverted
b. Camoid spins such that the contour rises
c. High camoid dwell, all bottles diverted
d. Camoid spins such that contour falls
3. The horizontal sections of the curve represent the camoid rotation in a
dwell period.

58
 4. The sloped line represents the camoid during it spin
a. The constant sloped section represents operating velocity
b. Notice the acceleration 13 up to operating velocity which occurs
between the horizontal and sloped linear segments.

The actual dwell and spin times are functions of the line speed and change
according to the inputs from the shaft encoder and photoeye.

13
In the actual timing diagram, the angle over which this acceleration occurs will be the same magnitude of
the buffer angle. This timing diagram is a basic theoretical representation.

59
7.4 - Motion Analysis
When the actuation design is finished, the camoid will behave as shown in Figure
42.

7.4.1 - Cam Extension

Figure 42: Top View of Cam Extension

A. Bottle 1 passes by the splitter. Bottle 2 approaches.
B. Both Bottle 1 and Bottle 2 are passing the splitter.
C. Bottle 1 continues to Lane 1. Bottle 2 is still in contact with the Low Dwell of
the cam. Bottle 3 enters the beginning of the splitter.
D-F. The cam rotates, with Bottle 3 riding the High Dwell while Bottle 2 is riding
the Low Dwell simultaneously.
G. Bottle 2 continues to Lane 1. Bottle 3 is riding the High Dwell of the cam.
H-I. Bottle 3 continues to Lane 2.

7.5 - Chassis Design Details

There are specific points of interest that are important to discuss regarding the
design of the chassis:

60
 Material Choice
Layout
Tolerances
Fasteners
Bearings
The chassis was outsourced for manufacture to Billington Steel as per
recommendation by Jim Bellins in Gallos machine shop. Fully dimensioned technical
drawings can be found in Appendix F.

Figure 43: Chassis Assembly View

61
Figure 44: Chassis Exploded View

7.5.1 - Material Choice

The material used in the chassis design is stainless steel plate. The choice follows
recommendation from Loel Peters because of the following reasons:
Provides the strength to support torque and heavy components
Resists oxidation in moist environments
Doesnt require any surface finishing
Widely used on Gallos bottling lines
Aesthetically pleasing

Referring to Figure 44, the material thicknesses chosen are as follows:

Base Plate
Shaft and Shaft and Servo Mount Plate
Brackets

The half inch plate will allow for tapped holes in the side of the plate for
fastening the other plates perpendicularly.

7.5.2 - Layout
The layout of the chassis is designed to minimize the footprint of the system such
that it fits within the Heufts footprint. In order to accomplish this, the servo must be

62
mounted next to the camoid, as opposed to inline. There are several key points that must
be kept in mind during the layout design process. Key layout design points are:
Servo and cam axis are parallel and offset enough distance to allow clearance
during camoid rotation
The axis offset is such that a nominal size belt can be used for power
transmission
There are mounting holes for a belt tensioner
The camshaft axis places the camoid low dwell at the conveyor edge
The camshaft axis places the camoid high enough to clear the bottom of the
chassis
The camshaft axis places the camoid at the correct height for bottle contact
The chassis can accommodate mounting hardware for implementation on the
line
The chassis can accommodate safety shields over drive transmission

7.5.3 - Tolerances
Because of the spinning camoid shaft and the belt drive transmission, there are
several dimensions that must be carefully toleranced to assure smooth operation. The
most important dimensions are shown in Table 7 and by Figure 45. The
table and figure are color coded for explanation.

Dimensions Reason
Bearing mount bolt pattern position on both
A Camshaft alignment
servo and shaft plates
Camshaft through holes on servo and shaft
A Camshaft alignment
plates

B Servo and shaft plate fastener holes Camshaft alignment

Bearing mount bolt pattern position and

C Transmission belt tension
servo bolt pattern position

C Servo through hole position Transmission belt tension

Table 7: Tolerance Reasons

63
 Figure 45: Important Toleranced Dimensions

For actual tolerance values, consult the drawings in Appendix G.

7.5.4 - Fasteners
All plates are fastened together using either -20 or 3/8- 16 bolts (see appendix
G for bolt types and locations) and locking nuts where nuts are needed. Bolts were
chosen over welding as welding could cause unwanted warping of the plates from the
heat.

7.5.5 - Bearings
The bearings chosen to use are SKF FY TF 4 Bolt Flange Mount Bearings and
were purchased from the Gallo storeroom. These are the standard flange mount bearings
that Gallo uses and supplies for this specific shaft size.

7.6 - Drive Design Details

The drive transmission chosen is manufactured by Goodyear 14. The system
components were readily available in Gallos part storeroom. See Appendix H for

14
Refer to Appendix H - Goodyear Eagle Pd Power Transmission for a complete list of Goodyear Eagle Pd
power transmission products

64
complete details on the Goodyear product used. The basic specifications of the belt and
sprockets can be seen in Table 8 and Table 9 respectively.

Figure 46: Goodyear Eagle Pd Power Transmission

Manufacturer Goodyear
Model Eagle Pd W-720
Length 720mm
Width 32mm
Pitch 8mm
Table 8: Belt Used

Manufacturer Goodyear
Model Eagle Pd W-28S-H
No. of Teeth 28
Hub Type Keyed Quick Disconnect (QD)
Pitch 8mm
Width 33mm
Table 9: Sprockets Used

There were several criteria that outlined the decision to use the Goodyear
products:
Recommendation
Claimed benefits of the system
Chassis layout
Drive ratio
Availability
Belt Tensioner

65
7.6.1 - Recommendation
The Goodyear products were chosen following the recommendation of Loel
Peters and Mike Delikowski.

7.6.2 - Claimed Benefits

The company claimed benefits from the website 15:
Consistent Dimensional Stability
Low Pre-tension
Low Noise
High Abrasion Resistance
Low Maintenance
High Flexibility
High Precision Linear Positioning

Each of these benefits fit the needs of the camoid laner.

7.6.3 - Chassis Layout

The sprocket size was chosen for several reasons:
a) The sprocket attached to the camoid shaft needed to be small enough such that it
would not overhang past the chassis and interfere with bottles on the line, i.e. not
be larger than the base circle diameter of the camoid.
b) Smaller sprockets have smaller inertias.
c) Smaller sprockets will need a smaller belt length

7.6.4 - Drive Ratio

The drive ratio for the transmission was chosen to be 1:1. The servo already
incorporates a satisfactory gearbox ratio of 10:1 and there is no need to adjust beyond
that. Furthermore, the parts list is simplified by using two of the same sprockets.

7.6.5 - Availability
The final criterion governing our choice for the drive transmission was the
immediate availability of most of the components in the Gallo storeroom.

7.6.6 - Belt Tensioner

In order to provide adequate tension and allow for easy initial installation, a belt tensioner
is to be attached to the chassis. Gallo currently uses Eagle Pd power transmissions that
are tensioned with a commercial tensioner arm available in the Gallo storeroom. The
tensioner is a Lovejoy SE-18 ROSTA Tensioner. It will provide a force on the outside
(smooth side) of the belt and thus requires the Lovejoy R-15/18 Roller Idler-SE15/18,
which is a non-toothed idler wheel. The orientation of the tensioner can be seen in
Figure 48: Full CAD Assembly

15
Goodyear Industrial. Eagle Pd Industrial Power Transmissions. Retrieved February 10, 2007, from
http://www.goodyearindustrialproducts.com/powertransmission/products/pdf/eagle_pd_belt.pdf.

66
.

Figure 47: Lovejoy Belt Tensioner with smooth idler pulley

67
7.7 - Full Assembly
The full assembly is shown below with all components discussed in this chapter
in the proper locations.

Figure 48: Full CAD Assembly

68
 Figure 49: Full Assembly

7.8 - Mechanism Assembly

Several important aspects about mechanism assembly must be noted to assure
proper function of the mechanism:
Camoid position on shaft
Camoid shaft trueness
Servo installation
Sprocket position of camoid and servo shaft
Belt tensioning

7.8.1 - Camoid Position on Shaft

The edge of the larger side of the camoid should line up directly with the end of
the keyway. Shaft collars should be tightened against the camoid to prevent translation
along the camoid axis.

7.8.2 - Camoid Shaft Alignment

It is important for the camoid shaft to have a true fit on the bearings, i.e. the
bearings should share the same axis. It is possible to adjust this during the installation.
Because there are tolerances on the bearing mount bolt pattern in the plate, we do not
tightened the bearing bolts until the shaft is installed and trued.

69
7.8.3 - Servo Installation
It must be assured that the servo mounting bolts are sufficiently tightened as the
servo hangs unsupported from the mounting flange.

7.8.4 - Sprocket Position

The sprockets must be positioned simultaneously on the camoid and servo shaft to
ensure the belt is centered on each sprocket, especially as the belt is specifically designed
to guard against sprocket wander. The sprockets are directional; make sure both sprockets
are oriented correctly to accommodate the belt.

7.8.5 - Belt Tensioning

The Lovejoy tensioner must be tensioned in a specific manner. See Figure 50 for
visualization.

1. A large wrench is used to apply torque to the square body into the belt
2. The angle of the arm relative to the body must read 20 degrees based on
manufacturers specifications
3. The mounting bolt is tightened
4. The safety bolt is tightened

Figure 50: Tensioning Diagram

70
7.8.6 - Initial Motion Test
The initial motion test was conducted in the controls office. The test was basic
and consisted of running the laner mechanism at a simulated speed of 600 bottles per
minute. The laner was allowed to run for 30 minutes and observations were made on the
mechanism ensuring full functionality. The initial motion testing ran smoothly. There
was no need for initial adjustments.

7.9 - Implementation
Implementation on the test loop must meet several criteria:
Photoeye positioned directly in front of (up-line) from the laner (due to
programming protocol)
Shaft encoder must be properly connected to conveyor shaft
Test loop rails must be set to accommodate a lane shift of bottles
Electronics are safely installed and connected

The implementation was completely supervised by Gallo employees to assure

proper safety and installation guidelines are followed. Jason Elliot supervised the laner
installation and resetting of rails. David Booth supervised electronic installation and
connection, including photoeye and shaft encoder.

71
Chapter 8.0 - Data Collection and Analysis
This chapter outlines the calculations that were conducted throughout the design
process. These calculations provide the reasoning behind many of the decisions made in
the final design. Calculations conducted include:
Bottle analysis
Cam profile analysis
Bottle trajectory analysis (after leaving laner)
Torque requirements
Stress analysis
Fatigue analysis

8.1 - Bottle Dynamic Analysis

To better understand the behavior of bottles on a conveyor line, several tests are
conducted to test bottle physical properties. These tests include a static friction test,
center of gravity test and a neck strength test.

8.1.1 - Friction Test

An experiment was conducted to test the coefficient of static friction of a full
wine bottle on the two materials commonly used for conveyor systems on the Gallo
bottling lines. A section of conveyor links is supported and assured to be level using an
inclinometer. The conveyor is lubricated with detergent (glycol based). With the bottle on
the conveyor section, the conveyor is inclined until the bottle begins to slide. The angle is
measured using the inclinometer.
The friction coefficient is obtained by calculating tan() 16. Two separate tests
were conducted with plastic and stainless steel conveyor materials and five trials on each
material were conducted.

Plastic Stainless Steel

Coefficient Coefficient
Angle (degrees) Angle (degrees)
Static Friction Static Friction

Average 0.125 Average 0.126

Figure 52: Plastic Friction Test Figure 51: Stainless Steel Friction Test

8.1.2 - Center of Gravity Test

An experiment was conducted to test the center of gravity of a full wine bottle. The
center of gravity is measured through the use of a pendulum, the full bottle being the
mass. The period of a pendulum is a function of the force of gravity and length of the
pendulum arm. The relationship is:

16
see Appendix I - Bottle Tests

72
 l
T = 2
g
Where T is time of one period, l is the length of pendulum from an objects center
of gravity, and g is the acceleration due to gravity. By measuring the time of one period
the effective length (from pivot to center of gravity) of the pendulum is obtained by
rearranging the above relationship:
2
T
l = *g
2
Since we know the dimensions of the bottle and length of the rope used for the
pendulum test, the location of the center of gravity in relation to the bottle is calculated.
Assumptions are also made that the bottle is symmetric about two axes such that the
center of gravity is in the center of the bottle at some height on the z axis.
Several tests were conducted with different lengths of rope. The time was
recorded for ten periods to pass. Several runs were conducted for each length of rope.
The center of gravity is calculated relative to the base of the bottle. Time was recorded
with a computer stop clock accurate to 0.05 seconds. Length was measured with a tape
measure accurate to 1/8 inch. The results are recorded in Figure 53.

Center of Gravity Pendulum Test

Height from Bottom
Height from Top (in.)
(in.)

Average 7.9 4.3

Figure 53: Pendulum Test Results

8.1.3 - Neck Strength Test

An experiment was conducted to test the failure load of the bottle neck using a
static load acting at an extreme point on the bottle neck. The bottle overhangs off a solid
surface with a rope attached to the bottle neck and a means to supply the force.
The results of the test concluded that the bottle neck was sufficiently strong to
withstand the static load of at least 30 kg. Failure was not achieved because of the lack of
equipment and the obvious dangers associated with broken glass.

8.2 - Cam Profile Design and Analysis

It is important for the bottles to be guided in a manner such that the acceleration
across the conveyor is induced gradually and with control. Several different programs
were analyzed to compare output kinematic and dynamic analysis. In this case, lower
accelerations are desirable as this leads to lower forces that the bottle will experience.
The cam programs that were analyzed included simple harmonic, modified trapezoid and
a 4-5-6-7 polynomial function. For baseline comparison the contour of the current Heuft
system follows the same analysis.

73
8.2.1 - Constraints
Several different constraints were made on each of the cam programs and are
discussed as follows. Detailed calculations can be found in Appendix M - Detailed
Mathematics.

8.2.1.1 - Simple Harmonic Displacement Constraints

The maximum final displacement is made to constrain the amplitude of the sine
function. The max displacement is based on the current Heuft system.
A second assumption is made based on the fact that the bottles are contacted only
on one side. This means that the acceleration is constrained to only being positive, since a
wall cannot provide a negative (pulling) force. This constrains the displacement curve to
be cut off at the inflection point (where concavity changes, and thus the second derivative
changes from negative to positive), meaning the effective contour is of a sine wave.

8.2.1.2 - Modified Trapezoidal Acceleration Constraints

The maximum allowable acceleration is made to set a working value to build the
function. The maximum acceleration value primarily is based on an iterative process to
provide the necessary displacement (which is solely based on the Heuft system). The
iterative process consisted of specifying maximum accelerations and finding the double
integral of the function such that the maximum displacement equaled that of the Heuft
system.
As stated above, the acceleration profile of the contour must remain positive. The
full cycle of a mod trap contour is 50% negative acceleration. Thus, in order to constrain
to positive acceleration, the acceleration profile function must be cut in half.

8.2.1.3 - Polynomial Displacement Constraints

Since the initial and final values of displacement, velocity, acceleration and jerk
can be controlled, we have many options to adjust the reaction of the bottle. In this
particular case, we want the initial values to all be zero, which will assure the bottle
reaction is brought on gradually. However, assumptions must be made on the other end of
the profile for final displacement, velocity, acceleration and jerk. These values were
adjusted to form a reasonable contour and acceleration profile and took some time to find
a satisfactory outcome. One method attempted was to input final values of another profile
into the polynomial function and observe results. Another method was to use all zero
final values.

74
 Table 10 shown below is the initial values input into the calculations to give plot
the contours of the cam programs. NS in the table signifies that the value was not
specified in the assumption.

Constraints
Simple Mod 4-5-6-7
Harmonic Trap Poly

Initial 0 NS 0
Displacement
(m) 0.07
Final 0.069 0.069
(iterative)

Initial NS NS 0
Velocity (m/s)
Final NS NS 0.6
Initial NS 0 0
Acceleration
Final NS NS 0
(m/s^2)
MAX NS 3.0 NS
Initial NS NS 0
Jerk (m/s^3)
Final NS NS 0
Table 10: Cam Program Constraints

With these assumed values and following the standard procedure for design of
each cam profile, we were able to obtain the plots of the cam contour and its derivatives.
Kinematic analysis of the contour program includes plots of the following qualities:
displacement, velocity (across conveyor, y-direction), acceleration and jerk. Also
included in the analysis is the trajectory that the bottle follows after it travels past the
contour.
Dynamic analysis of the contour program includes forces and moments that the
bottle experiences during its translation.
All analyses are conducted with the same line conditions: line speed, length of
split, bottle physical properties and friction coefficient.

8.2.2 - Kinematic Analysis

Kinematic analysis is in the form of S-V-A-J diagrams of each of the cam
contours. Each analysis is conducted as a comparison of the plots in a single chart. The
current (Heuft) system is always used as the baseline as the purpose of this project is for
improvement.

75
In the following graphs, each cam program is always represented by the same color line.
Table 11 is the graph color scheme used throughout this section.

Graphic Color Scheme

Heuft Solid Red Line
Simple Harmonic (Sinusoidal) Dotted Blue Line
Modified Trapezoidal Acceleration Dashed Green Line
4-5-6-7 Polynomial Function Dot-Dash Pink Line
Table 11: Graphic Color Scheme

When interpreting the following charts, it is essential to know the axis orientation.
For the purposes of the displacement curve we assume that the curve exemplifies what
would be seen looking directly down at the laner. Thus the x axis is down line and the y
axis is across the conveyor. The velocity, acceleration, jerk and force plots all provide
data for what the bottle experiences in the y direction, across the conveyor, which is the
direction of interest.

8.2.2.1 - Diverting Contour

The diverting contour for each cam program is shown below in Figure 54. Note
that the contours look relatively similar. However, each contour offers very different
acceleration profiles. In order to gain better understanding of the displacement profiles,
we conduct a comparison of each to the Heuft contour by plotting the difference between
the two functions shown in Figure 55.
Contour Displacement Profile
0.08

(
y heuft t split ) 0.06
Contour Profile (m)

(
swall asin , t split )
swall( atrap , t split )
0.04

y general( t split , Cout ) 0.02

hy
0

0.02
0 0.05 0.1 0.15 0.2 0.25
t split

Contour Contact Time (s)

Figure 54: Contour of Each Cam Program

The displacement difference is shown below in Figure 55. Since the standard of
the comparison is the Heuft contour, it is not shown in this plot, however the color
scheme still holds true. Notice that the modified trapezoid curve is most similar to the
Heuft system. The simple harmonic is shallower than that of the Heuft and the
polynomial function is a deeper curve.

76
 Displacement Difference: Heuft vs. Cam Programs
Displacement Difference (in)

(
diffsin t split )
in
0.2
(
difftrap t split )
in

(
diffpoly t split ) 0
in

0.2
0 0.05 0.1 0.15 0.2 0.25
t split

Contour Contact Time (s)

Figure 55: Displacement Difference, Heuft vs. Cam Programs

The shape of the contour gives limited insight into the performance of the
contour. In the following sections we will look into the motion dynamics that each
contour provides for a bottle traveling at maximum line speed of 400 bottles per minute,
which translates to roughly 1 m/s.

8.2.2.2 - Velocity Curve

Shown below in Figure 56 is the velocity of the bottle as it crosses the lane. It is
interesting to note that the Heuft system does not start at zero velocity and is always
increasing. In general, the four programs offer similar velocity values at each point along
the laner. An important note to keep in mind is that the bottles are not guided the entire
way across the conveyor. They follow a trajectory after they leave the guide of the
contour wall because there is positive transverse velocity. It is important for the bottles to
maintain this velocity to clear the division in the conveyor. Note that the simple harmonic
profile is significantly lower than the Heuft profile. A good way to accomplish this is to
match the velocity of the Heuft system, as it is already known that this system provides a
correct final transverse velocity.

77
 Bottle Transverse Velocity
0.8

( )
Bottle Velocity (m/s)
vheuft t split
0.6
(
vwall asin , t split )
vwall( atrap , t split ) 0.4

y' general( t split , Cout )

0.2

0
0 0.05 0.1 0.15 0.2 0.25
t split

Contour Contact Time (s)

Figure 56: Bottle Transverse Velocity

8.2.2.3 - Acceleration Curve

The acceleration profiles of each contour give the best insight on the behavior of
the bottles during the laning. The acceleration profile comparison shown in Figure 57
exemplifies the difference of each cam program. Notice that the Heuft system is a
linearly increasing function of acceleration. Both the mod trap and simple harmonic
functions have lower maximum accelerations, however the polynomial function has a
peak acceleration higher than all compared cam programs.
Bottle Transverse Acceleration
6

( )
Acceleration (m/s^2)

awall aheuft , t split

awall( asin , t split ) 4

awall( atrap , t split )

y'' general( t split , Cout )

2

0
0 0.05 0.1 0.15 0.2 0.25
t split

Contour Contact Time (s)

Figure 57: Bottle Transverse Acceleration

8.2.2.4 - Jerk Curve

The jerk plot (shown in Figure 58) gives good indication about the potential for
possible impacts. A good indication of impacts is a spike or cusp in the jerk profile.
Notice that the mod trap contour induces the highest maximum jerk of the contours. All
contours offer smooth jerk curves and none seem to pose any immediate dangers.

78
 Bottle Transverse Jerk
100

(
jwall aheuft , t split ) 50
Jerk (m/s^3)

jwall( asin , t split )

jwall( atrap , t split ) 0

y''' general( t split , Cout )

50

100
0 0.05 0.1 0.15 0.2 0.25
t split

Contour Contact Time (s)

Figure 58: Bottle Transverse Jerk
Table 12 shows the maximum values of the aforementioned plots. The listing
order is the same used in the previous plots; Heuft system, simple harmonic, mod trap, 4-
5-6-7 polynomial, from top to bottom. Following previously set guidelines, we are
looking for the contour to cause minimum amount of acceleration, and thus force. The
mod trap and simple harmonic contours seem to have the mildest acceleration. However,
the simple harmonic has almost 25% less final velocity than the other contours. For these
reasons, the modified trapezoid acceleration profile was chosen for the camoid contour
design.

79
 Maximum Kinematic Values

Maximum Velocity Maximum Acceleration Maximum Jerk

0.656 3.851 10.54

0.454 m 2.987 m 19.636 m
v maximum= amaximum= jmaximum=
0.592 s 3 2 78.117 3
0.6 4.709 s 57.263 s

Table 12: Maximum Kinematic Values

8.2.3 - Force Analysis

Working from the kinematic analysis data, it is possible to conduct analysis on the
forces that the bottle will be subject to during the laning process. During the laning
process, the bottle is subject to four forces; gravity Fg, normal force from the conveyor
FN, the force of the contour wall Fw and the friction force between the bottle base and
conveyor Ff (see Figure 59). Again, the primary concern of the laning device is that no
bottles are tipped or damaged. To ensure that this design specification is adhered to, we
conduct analysis to test the wall force and overturning moment which occur during the
laning process.

Figure 59: Bottle Free Body Diagram

Using Newtons physical law F=ma, we can plot the force that the bottle will see
throughout the laning process. Shown below in Figure 60 we can see the plot of force vs.
laning time to which the bottle is subject.

80
 Force on Bottle
8

(
Fwall aheuft , t split )
Fwall( asin , t split )
6
Force (Newtons)

Fwall( atrap , t split )

4
Foptimal( t split , Cout )

Ffriction 2

0
0 0.05 0.1 0.15 0.2 0.25
t split

Contour Contact Time (s)

Figure 60: Force on Bottle
Notice that the curves are shaped the same as the acceleration curve because the
mass of the bottle is constant in all four cases. Also, notice the solid horizontal cyan force
of friction line. This is the force of static friction between the bottle and conveyor (as the
bottle is assumed stationary relative to the conveyor, no slip) and represents the point at
which static friction switches to kinetic friction. The static friction force is the maximum
force that the bottle will experience opposing the force of the wall. The bottle will not
begin traversing the conveyor laterally until the force from the wall exceeds the force of
static friction. It is desirable to exceed this force as rapidly as possible, as is done in the
mod trap profile.

81
8.2.4 - Moment Analysis
Through the force analysis, the overturning moment can be analyzed, assuming a
height at which the contour contacts the bottle. Figure 61 shows the moment plots and
again are similar in shape to the acceleration plot.

Induced Moment on Bottles

0.6

(
M wall aheuft , t split , hcontact )
M wall( asin , t split , hcontact)
Moment (N-m)

0.4

M wall( atrap , t split , hcontact)

M optimal( t split , Cout , hcontact )

0.2
Mg

0
0 0.05 0.1 0.15 0.2 0.25
t split

Contour Contact Time (s)

Figure 61: Induced Moment on Bottles
The horizontal cyan line shown in the plot is the moment sufficient to cause the
bottle to begin to tip. We can understand this baseline moment calculation by examining
Figure 59. Assuming that Fg acts at the center of gravity (CG), this point force induces a
moment about the bottom corner of the bottle with the bottle radius as the moment arm.
The plots are simply the force of the wall multiplied by the height at which the wall
contacts the bottle. If this moment is greater than that caused by gravity, the bottle will
begin to tip.
Notice that the polynomial function crosses the tipping mark, meaning this
particular contour could have the potential to cause instability in the bottle. The other two
contours in question both induce lower maximum moments than the current Heuft
system.

82
 Table 13 shows the maximum force and moment of each contour. The order is the
same used in all of the plots: Heuft, Simple Harmonic, Mod Trap, and Polynomial. As we
can see from the values, the simple harmonic and mod trap yield similar values that are
smaller than the other two contours.
Force Analysis

Maximum Force Maximum Moment

4.968 0.442
3.853 0.343
Fmaximum= N M maximum= N m
3.87 0.344
6.074 0.54

Table 13: Maximum Force Analysis Values

8.3 - Trajectory Test

The camoid laner is a direct replacement of the Heuft system with no further line
or rail manipulation. In order to be certain that the laner will cause the bottles to divert
fully into the second lane; the bottle trajectories from the end of the cam contour of the
three curves are compared to the Heuft. The trajectories are shown in Figure 62.

Bottle Trajectory Curve

( )
T c_gen t traj 0.3
Displacement (m)

T c_sin( t traj)

T c_trap ( t traj)
0.2

T c_gen( t )
0.1

0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
t traj

Trajectory Time (s)

Figure 62: Bottle Trajectories

From the figure we can see that the trajectories are very similar, eliminating any
worries about the bottles following a correct diversion path.

83
8.4 - Choosing a Contour
After this analysis, it is possible to choose the contour that will most suit the
application which we need. For the purposes of the high speed bottle laner, the Modified
Trapezoid (mod trap) contour is chosen based on the following explanations.

8.4.1 - Acceptable Final Velocity

The velocity curve of the mod trap closely resembles the current system and
provides a similar final velocity, which is essential for the bottle to follow the required
trajectory for safe laning.

8.4.2 - Minimizing Maximum Acceleration

While the mod trap does not provide the lowest maximum acceleration, it comes
in at a close second to the simple harmonic. However, the simple harmonic curve is not
applicable because of an unacceptably low final velocity.

8.4.3 - Minimizing Jerk

The jerk curve of the mod trap is not the best of the group, however still provides
continuous and finite values for jerk which are still within the realms of the other
analyzed contours. The higher values for jerk are a worthy exception to make for the
satisfactory velocity and acceleration profiles.

8.4.4 - Minimizing Force

Examining Table 13, the force induced by the simple harmonic and the mod trap
are comparable in magnitude and both significantly lower than the other two contours.
Again, because of the mod traps more acceptable velocity curve, we choose it as the
better choice.

8.4.5 - Minimizing Moment

Examining Table 13 again for the moment values, we see a similar trend to the
force analysis. The mod trap is chosen over the simple harmonic for the same reasons
described above for minimizing force. The polynomial function is not an option as the
moment induced by the contour is greater than the tipping moment as seen in Figure 61.

8.5 - Torque Requirements

Once the geometry of the cam has been set, it is possible to begin to design the
second part of the mechanism which is the timing of the rotation. This timing relies
directly on the geometry

8.5.1 - Operating Rotational Velocity

If the bottle is traveling at line speed, then the time that it takes to travel the length
of the cam laner is represented by Equation 1. Equation 2 is the calculation for the
operating rotational velocity. Equation 3 shows the worst case scenario time between two
bottles with a bottle spacing of one bottle diameter center to center of bottles on
conveyor.

84
 Timing Data
Lsplit
Equation 1: Bottle and Laner tfinal := tfinal = 0.239s
v line
Contact Time

140deg
Equation 2: Operational Rotational op := op = 97.665rpm
tfinal
Velocity

d bottle
Equation 3: Worst Case Scenario t worst := t worst = 0.075s
Time Between Two Bottles v line

Table 14: Time Data

8.5.2 - Angular Acceleration

Recall that the cam must be at full operating velocity before any bottle
manipulation can occur. The angular acceleration necessary in the case is based on the
spin-up time, the time it takes for the servo to accelerate the camoid up to operating
velocity, as well as servo limitations. Recall that the geometry of the cam includes a
buffer angle that allows servo acceleration. The allowable spin-up time is directly related
to the buffer angle, the smaller the buffer angle the faster the spin-up time required. This
relationship is represented by Figure 63.

Acceleration Buffer Angle vs. Spin-up Time

60
Buffer Angle (deg)

40
(
buffer t spinup )
deg
20

0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
t spinup

Spin-Up Time (s)

Figure 63: Spin-up time vs. Buffer Angle

Notice that the spin-up time is the independent variable. This is because we can
find a lower limit on the spin-up time based on servo characteristics and design the
camoid buffer angle from there. There is more flexibility in sizing for servo torque than
there is in the design of the buffer angle on the camoid.

85
 The maximum acceleration of a servo is based on its torque, motor inertia and
inertia of the cam. For the purposes of graphical representation, we can specify a lower
extreme time of 0 seconds for spin-up. The acceleration plot is based on Equation 4.

op
Equation 4: Acceleration Function (
tspinup :=)tspinup

The acceleration plot shown in Figure 64 shows the acceleration necessary to

reach operating velocity for a range of spin-up times. The plot follows intuition as the
amount of time allowed to reach a certain velocity increases, the acceleration necessary to
reach that velocity increases, in this case, asymptotically.

Acceleration vs. Spin-up Time

4
2 .10
Acceleration (deg/s^2)

4
1.5 .10

(
t spinup ) 1 .10
4

5000

0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
t spinup

Spin-Up Time (s)

Figure 64: Acceleration vs. Spin-Up Time

8.5.3 - Torque Required

Now that the acceleration profile is plotted, we can specify the torque needed
based on the cam and motor mass inertia properties. Equation 5 shows the function used
to plot the necessary torque for a desired spin-up time.

Equation 5: Torque
Function
( ) ( )(
Tservo tspinup := tspinup Izz_cam + Izz_motor )
Cam Moment of
Where: Izz_cam + Inertia About
Rotational Axis
Motor Moment of
Izz_motor ) Inertia About
Rotational Axis

Figure 65 shows the torque vs. spin-up time plot. Notice that it is the same shape
as the acceleration curve.

86
 Servo Torque vs. Spin Up Time
100
Torque (N-m)

( )
T servo t spinup 50

0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
t spinup

Spin Up Time (s)

Figure 65: Servo Torque Required vs. Spin-Up Time

This torque graph is rendered for visualization purposes. We will demonstrate in

the following section how the motor torque is actually calculated. However, the graph
gives a range of what can be expected for the required torques.

8.6 - Stress Analysis

Once the maximum torque that will be transmitted is known, we can begin stress
analysis on the parts that will experience the most stress. The servo will drive a camshaft
which will transmit torque to the camoid by means of a key and keyway. The keyway
stress analysis is shown in Table 15. It should be noted that the required torque is based
on a spin-up time obtained from the Motion Analysis software, which is the time to
accelerate from 0 to operating velocity in the final buffer angle of 7.5 degrees. This time
was then input into Equation 5.

87
 Keyway Stress Analysis
Key Width Key Length Key Depth
3 3
W key := in Lkey := 1.5in Dkey := in
16 32
Tmax
Equation 6: Shear Fshear :=
Dshaft Fshear = 684.893N
Force at Key
2
Fshear
Equation 7: Keyway ave := ave = 2.491MPa
Average Shear Stress A sec
Sys_steel
Equation 8: Safety steel := steel = 50.177
Factor ave
Where
Equation 9: Area of Key A sec := W key Lkey
Required Torque based on Motion
Equation 10: Required Torque Tmax := Tservo ( .01367s ) Analysis Tool and MathCAD
Torque Plot
Sys_steel = 125MPa Theoretical Yield Strength of Steel
Table 15: Keyway Stress Analysis
Notice that the safety factor of 50 is adequate for this application.

Stress analysis is conducted on the shaft as shown in Table 16. The torsional
deflection and shear stress are the main points of interest for the analysis.

88
 Shaft Stress Analysis
Shaft Diameter Shaft Length Material
Dshaft := .75in Lshaft = 16in Stainless Steel
Tmax Lshaft
Equation 11: Shaft Torsional shaft := shaft = 0.036deg
Deflection Gsteel Jshaft

Dshaft
Equation 12: Shaft Shear Stress due Tmax
2 shaft = 3.172MPa
to Torsion shaft :=
Jshaft

Where
4
Equation 13: Shaft Polar Moment of Dshaft

Inertia Jshaft :=
2 2

Gsteel := 216GPa Stainless Steel Modulus of

Elasticity
Table 16: Shaft Stress Analysis

8.7 - Fatigue Stress Analysis

Because the system will be under cyclic torque loading, it is necessary to conduct
fatigue stress analysis which is shown in Table 17. The calculations are conducted for a
shaft with a ground surface finish, operating in ambient temperatures, 99% reliability and
under non-reversing torsional loading. The corrected fatigue function and safety factor
are shown in Equation 14 and Equation 15 respectively.
Shaft Fatigue
Stainless Steel
Se' = 200MPa Uncorrected Endurance
Limit
Equation 14:
Corrected Se := Cload Csize Csurface Ctemperature Creliability Se' Se = 124.049MPa
Fatigue Function
Se
Equation 15: torsion := torsion = 39.109
Safety Factor shaft

See Appendix M - Detailed Mathematics for detailed information on the mathematics of this
section
Table 17: Shaft Fatigue Analysis
Notice that the safety factor of 39 is adequate for this application.

Inevitably, keyways cause stress concentrations due to the sharp radii of the cut.
The stress concentration is based on several factors including notch radius, Neubers
Constant, and a notch sensitivity factor. The safety factor is calculated based on the yield
strength of the stainless steel.

89
 Keyway Stress Concentration
Sys_steel = 125MPa Yield Strength of Stainless
Steel

Equation 16: Fatigue Stress

Concentration Factor
( )
Kfs := 1 + q Kt 1 Kfs = 2.442

Equation 17: Shear Stress concentration := Kfs shaft concentration = 6.869MPa

Concentration due to Torsion
Sys_steel
Equation 18: Safety Factor keyway := keyway = 18.199
concentration
Table 18: Keyway Stress Concentration Analysis
Notice that the safety factor of 18 is adequate for this application.

90
Chapter 9.0 - Prototype Testing
This chapter discusses in detail the testing protocol followed for proper data
collection on the functionality of the camoid laner prototype. This section is meant to
provide the necessary information to successfully duplicate the experiment and record all
results.

9.1 - Materials
Test Loop
Camoid Laner
o Camoid
o Allen-Bradley MPL - B4520P MJ22AA servo motor
o Alpha SP-100-MF1-10 Gearbox 10:1 ratio
o Rockwell Automation Kinetix 3000 Servo Driver
o 2x SKF FY TF Flange Mount Bearings
o Lovejoy SE-18 ROSTA Tensioner w/ Hardware
o Lovejoy R-15/18 Roller Idler-SE15/18
o Stegmann DG60L WSR 5000 pulse/rev shaft encoder
o 2x Goodyear W-28S-H White Eagle Pd (QD Bushed) Sprockets
o Martin H QD Bushing
o Martin H 32mm QD Bushing
o Goodyear Pd W-720 Eagle Belt
o Custom length keyed stainless steel shaft
o Custom chassis design
30 test bottles
Data recording device (computer, pen, paper)

9.2 - Objective
This test is to ensure no bottles are tipped or damaged during operation, and
ensure the repeatability.

9.3 - Variables
There are several specific variables that will be tested in order to meet
aforementioned objective.
Bottle Spacing
Number of Bottles per Cycle

9.4 - Setup Safety Precaution

Safety is of primary concern when dealing with high voltage applications. In this
case, 480 volt electrical inputs are required. Only those qualified and experienced to
handle high voltage should do so.
Servo motors are very powerful and should be handled with care. DO NOT place
any body parts on or near any part of mechanism when power is on. DO NOT stand in
path of diverting bottles if servo should malfunction and kick bottles off line.

91
9.5 - Setup
All testing will be conducted on the specific test loop that Gallo has set up for
testing new instruments shown in Figure 66. This test setup assumes the laner has been
mounted to the line.

Figure 66: Test loop

A series of tests will be conducted testing the two variables mentioned before.
The tests will be set up to find an extreme value at which the mechanism no longer
functions correctly. Before the test is conducted, there are several initial parameters of the
test that must be discussed.
Bottle Batches
Test Loop Speed
Photoeye Placement
Electronic Attachment

9.5.1 - Bottle Test Batches

Bottle batches are small so that if any problems arise, there is not a catastrophe
with tipped bottles. Batches of six bottles will be used, each bottle labeled 1-6. Also, the
bottle spacing can be set easier with smaller batches. Once we know that the settings are
correct and tipped bottles are minimized, a stream of bottles can be placed safely on the
line.

92
9.5.2 - Test Loop Speed
The test loop has a maximum speed of 24 inches per second without overloading
the test loop system.

9.5.3 - Photoeye Placement

The photoeye is placed directly in front of camoid laner in this case for ease of
programming. The camoid is triggered as soon as the target bottle breaks the photoeye
beam. Care must be taken to assure the photoeye is triggered once per bottle. This
becomes an issue when the photoeye reads through the transparent bottle body. It is
possible to trigger twice; once on the leading edge and once on the trailing edge of the
same bottle. To avoid this, the photoeye is placed either at the bottle neck where there is
an opaque cork or at the base where the glass is sufficiently thick to block enough of the
beam. This test used the latter method.

9.5.4 - Electronic Attachment

All electronics are attached to their respective power sources by control engineer
David Booth. The electronics board is shown in Figure 67.

Figure 67: Electronics Board

The setup of the control electronics was conducted by David Booth and Brandon
Abell of Gallos Control engineering department. Information regarding the design of the
control schematics and hardware setup will not be discussed in this report. Information
this can be researched under the heading motion control.
When attaching electronics to 480V power sources, extra care should be taken.

93
9.6 - Procedure Safety Precaution
The number of bottles per cycle for the camoid laner should be NO LESS than 4
BOTTLES PER CYCLE. Doing so can result in servo misfiring, potentially kicking
bottles off the line at high speed. The first tests conducted with less than four bottles per
cycle outlines the steps taken to obtain this safety hazard information.
Again, do not stand in front of the diverting bottles at any time to avoid any
potential injury. Any tipped bottles should be removed from the line as fast as possible to
avoid misfiring.

9.7 - Procedure
The procedural steps for testing are explained below. These steps are followed
only after laner mounted, the servo is programmed, electronics are active, and laner is in
its default home position which is in the off position.

1. Set up batch of bottles to be tested with spacing according to test number 1 in the
data recording table (shown in Appendix K Test Data Table) while line is not
running
2. Set up program for bottles per cycle according to the test data table.
3. Turn on conveyors
4. Run test five times bottle batch loops five times adjust spacing each time
5. Record results
6. Repeat for each test according to test data table

9.7.1 - Data Recording

Any abnormal interactions such as impacts, punches, off timing, diversion due to
incorrect geometry, or any other potential problems will be recorded and explained in the
test data table. Each bottle is labeled in the batch and if a bottle falls, the number of the
bottle is recorded. This will give insight how the actuation and geometry of the camoid
are functioning.
In addition to the bottle interaction, a second set of observations is needed to
record the functioning of the machine itself and any abnormalities experienced with the
mechanics of the cam and actuation.
The test protocol will be set up such that any problems encountered will be
followed by a set of recommendations on how to improve and/or solve such problems, if
the problems cannot be solved by us in the allotted time.

94
Chapter 10.0 - Test Results
The results of the testing are discussed in detail in this chapter. Also explained are
safety issues and problems experienced during the testing sequence. Conclusions provide
a concise explanation of the most important findings. Picture sequence of the testing is
shown in Figure 68.

Figure 68: Final Mechanism Cycle Sequence

10.1 - Testing Limitations

There are several limitations experienced with the test loop:
1. No available line encoder

95
 2. No variable line speed
3. Conveyor Condition
4. Not full line speed

10.1.1 - Line Encoder

The test loop did not offer a reliable means to attach a line encoder. This has
implications when setting the line speed in the program logic. Since the functionality of
the design is reliant on the line speed, outside instruments were used to measure the line
speed so that it could be hard coded into the program.

10.1.2 - Variable Line Speed

Because the line speed is hard coded in the control program, the line speed must
be held constant. This made testing for variable line speed impossible without shutting
the line down and adjusting the program logic.

10.1.3 - Conveyor Condition

The conveyor condition was poor for the purposes of providing a smooth
transition across the lines. In several sections of the conveyors, remnants of adhesives
caused increased friction as bottles slide across. The test loop also does not include a
lubrication system which also results in increased friction. Furthermore, there was a
misalignment between two conveyors causing a small lip on which bottles periodically
tipped.

10.1.4 - Line Speed

We were not able to set the test loop for full line speed because of the apparent
safety hazards that could arise from the poor condition of the conveyors. The line speed
was set to 24 inches per second, which translates to roughly 250 bottles per minute.

10.2 - Preliminary Testing

The actual testing of the laner was not conducted until it was certain that the
mechanism would function safely. A series of preliminary tests were conducted to ensure
safe operation, find the minimum number of bottles per cycle and the placement of the
photoeye. Several important results were gathered from the preliminary tests using the
current programming logic.
Minimum number of bottles per cycle is 4
Photoeye must be placed such that either the base of the bottle of the corked
part of the bottles break the beam

The reason for a 4 bottle minimum per cycle lies in the geometry of the camoid
and the timing of the actuation. The camoid length is 9.5 inches, which is approximately
3 bottle diameters. If the camoid is set to actuate in 2 or less bottles, the following
sequence of events can occur:

1. Bottle 1 triggers actuation

2. Camoid begins rotating

96
 3. Bottle 2 crosses photoeye and begins to be diverted. Camoid is still
rotating
4. Bottle 3 crosses photoeye, triggering actuation during camoid rotation

This poses a problem because the second actuation is triggered during the first.
Under the current program logic, this caused erratic behavior in the servo. The minimum
number of bottles per cycle is set at four to ensure such an event will not occur under any
circumstance.
Also under the current program, the photoeye must be placed at the base or neck
of the bottle. If it is placed such that its position aligns with the body of the bottle, the
beam will be broken twice due to the transparency of some products. This will cause two
counts for every one bottle. The current program causes actuation every four counts of
the photoeye.

97
10.3 - Results Table
The results of the test are shown in Table 19.

Camoid Laner Prototype Test: Bottle Interaction

Test Administered by:
William Robinson Caruso, James Freeman
Date: February 28, 2007
Saunders,
David Booth, Shawn Burns

Bottle Type: 7305 Time: 9:00-17:00

Test Number

Line Bottle Number Bottles Down

*
Speed Spacing of per Bottle Motion Comments Recommendations
(in/sec) (inches) Bottles Cycle #

1 24 0 6 1 all ext.
2 24 1 6 1 all ext. Failure due to laning
trigger while servo
3 24 2 6 1 all ext. is in rotation. This With current timing
causes servo to program, a minimum of
4 24 3 6 1 all ext. attempt to rotate to 4 bottles per cycle is
position B in the required to avoid such
5 24 0 6 2 all ext. middle of rotating to an issue. Program
position A. This optimization could also
6 24 1 6 2 all ext. behavior is erratic avoid this issue
and causes downed
bottles.
7 24 2 6 2 all ext.

8 24 3 6 2 all ext.
The minimum
9 24 0 6 4 3 ext. spacing for this
laner is Geometry and controls
10 24 1 6 4 3 ext. approximately 2 optimization can be
inches. If the bottles conducted to attempt to
are tighter, the cause the camoid to fit
11 24 2 6 4 none ext. between tighter spaced
bottle directly before
the target bottle bottles.
12 24 3 6 4 none ext. (bottle #3) is
adversely affected.

13 24 0 6 4 none ret. The camoid doesnt

have any trouble
14 24 1 6 4 none ret. retracting even at
minimum bottle
spacing
15 24 2 6 4 none ret.
All bottles
16 24 2 24 4 none both successfully laned.
No problems.
Each test is run 5 times to assure all problems with a batch of 6 bottles looping the test line.
*ext. = extension ret. = retraction
Table 19: Test Results

98
There are several key points of interest that can be concluded from the results:
Minimum bottle spacing for reliable camoid extension: 2 inches
Minimum bottle spacing for reliable camoid retraction: no minimum
Line speed affects the magnitude of displacement of the bottle
Photoeye placement

10.3.1 - Minimum Bottle Spacing: Extension

There is a minimum bottle spacing of 2 inches that is required in order for the
camoid laner to reliably lane bottles. The potential for tipped bottles increases drastically
when the bottles are spaced closer than 2 inches. The bottle that was most often tipped
was the last bottle to continue in the default lane, bottle 3 of 6. This occurs because the
camoid contacts the bottle during rotation and actively pushed the bottle aside. If the
bottles are spaced more than 2 inches, the camoid avoids striking the bottle and rotates
successfully between the two bottles.

10.3.2 - Minimum Bottle Spacing: Retraction

There is no minimum for bottle spacing during the retraction of the camoid; the
bottles can be directly adjacent to one another. No bottles were tipped or subject to
abnormal manipulation at any point during the retraction, regardless of the spacing.

10.3.3 - Magnitude of Displacement

The line speed directly affects the magnitude of displacement. Successful laning
of the camoid relies on the trajectory of the bottle after the end of the camoid. The
trajectory is dependent on the velocity of the bottle, which is less in slower line speeds.
With slower bottle speeds, the bottle tended not to make the full diversion. This lack of
displacement is also due to the decreased magnitude of momentum the bottle possesses to
overcome frictional forces. The displacement problem could also be linked to the poor
conveyor condition and lack of lubrication.

10.3.4 Photoeye Placement

The servo program is set to trigger as soon as the photoeye is triggered. The best
placement for the photoeye for this program proved to be directly on the outside edge of
the chassis as close to the conveyor surface as possible. The placement is shown in Figure
69.

Figure 69: Photoeye Placement

99
10.4 - Mechanism Observations
Observations about the mechanism concluded no visible wear or fatigue.
Furthermore, the mechanism operated smoothly and quietly and offered no impacts due
to interference of components.

10.5 - Test Results Conclusions

The overall outcome of the testing proved to yield satisfactory results. The
original design theory based successful laning on one bottle diameter, or 3 inch, spacing.
The test results yield less than the original design concept. Also there is no minimum
bottle spacing for the retraction, which also exceeded expectation.
The problem concerning the displacement of the bottles across the line must be
addressed. It is expected that under actual line conditions (400 bottles/min, lubricated
conveyors) the displacement will act similarly to the Heuft system and fully displace the
bottles.
Further testing is essential if the camoid laner is to be implemented for
production. However, this test proved that the camoid concept is capable of high speed
bottle laning and is candidate for further more rigorous testing.

100
Chapter 11.0 - Cost Comparison
The final step in analyzing the camoid laner system is performing a cost analysis
of all parts that were used in assembling the system. The following table shows the cost
of all parts of the camoid laner compared with the Heuft Delta FW system. For detailed
cost analysis and bill of materials refer to Appendix C.
Cost Comparison
Heuft Camoid
$13,000 $15,532.32 First Unit (all electronics)
$9,649.09 Secondary Unit
Table 20: Cost Comparison

The cost of the camoid laner is analyzed for several different scenarios as shown
in Table 20. The first unit price includes all electronics used for servo motion control, and
at $15,500 it is more expensive than the Heuft system. However, the servo electronics
make up a significant portion of the total cost. As Gallo implements more servos into
production, many of the instruments needed are already on the line being used for other
motion control purposes. It is possible to piggyback the camoid laner onto the existing
control infrastructure. This reduces the cost significantly as shown by the follow-up unit
cost, which utilizes key components that are capable of controlling multiple axes.
If the laner were ever to be used in full production, several cost saving strategies
could be administered such as:
In house chassis manufacture
Camoid price reduction for bulk purchase
Minimizing servo size
Piggyback existing motion control infrastructure

If these strategies are used, the original cost could potentially be reduced further.
In addition to reducing the capital cost of the laner, a single actuation servo
system offers significant savings in maintenance costs as the number of cycles demanded
on the system is within the design specifications of a servo. Also, the camoid system is
simple enough to allow easy replacement of components as opposed to the Heuft system
which requires replacement of the entire system. For example, if a servo motor required
replacement it could be replaced immediately, on the line, provided the part is in stock.

101
Chapter 12.0 - Recommendations
Prior to implementing the camoid bottle laner in a production setting several tasks
must be completed:

12.1 - Line Speed Testing

Experimentation was hindered by the lack of encoder information on the test
conveyor. As a result, the laner was not tested under full speed line conditions (400
bottle/minute). It was designed specifically to accommodate full speed, and it is possible
that the laner will function better at 400 bottle/minute. Additionally, the laner should be
further tested reflecting ramping up and ramping down the line speed.

12.2 - Bottle Testing

Only one type of bottle was tested, the 7305 750mL bottle. The 7405 model
should theoretically perform worse than the 7305, having a higher, less-stable center of
gravity. At the very least, similar tests should be performed on 7405s. Ideally, all bottles
that could run on Line 2 would be tested as well.

12.3 - Endurance Testing

The longest single run during testing was 20 minutes. Before production
implementation the laner should run constantly for hours. There is no data that suggests
that the laner will fail from repetition.

12.4 - Part Acquisition

If additional units are desired, parts can be purchased at much lower cost. New
methods for mold fabrication using rapid prototyping now enable small batch injection
molding at reasonable cost. A mold could be purchased for future use; subsequent parts
would cost significantly less after the investment in a mold. The chassis could also be
fabricated in-house, instead of employing an outside contractor. Finally, the servo used
in the prototype was recovered from the junk rack, but was significantly larger than
necessary. In the future a smaller servo, the MPL-B230P-HJ42AA, could be used, at
roughly half the price.

12.5 - Program Optimization

Very little was done with optimization of the servos controlling logic. Certain
safeguards could be added to the logic to prevent misfires. Also, through the use of an
encoder and a programmed delay, the photoeye could be positioned elsewhere upstream.

12.6 - Geometry Optimization

The camoids geometry could be refined with more tests. Because bottles may
not impact the laner directly on the rail, the first 1-2 inches of the camoid rise contour are
frequently missed entirely. The length could therefore be shortened and the maximum
rise could be increased slightly. No geometric alterations are necessary for operation, but
could result in a smaller laner footprint.

102
Chapter 13.0 - Conclusions
Upon completion, the tests demonstrated the camoid laner to be a viable concept,
and one that should be investigated further. Several tests have been outlined, to be
completed before taking the camoid laner to a production setting. Many of the factors
explored in the construction and experimentation of the prototype suggest that it would
be a large improvement over the current system. A single-actuated bottle laner could
significantly reduce maintenance costs compared to the 12-actuator Heuft Rejecter.
Servo driven equipment is more reliable and longer lasting than pneumatic systems. In
addition, the camoid laner performs comparably to the Heuft Rejecter under similar
conditions.

103
References
E&J Gallo Winery. Who We Are. Retrieved January 10th, 2007, from
http://www.ejgallo.com/

Engineers Handbook. Rapid Prototyping: Electron Beam Molding. Retrieved January 26,
2007 from http://www.engineershandbook.com/Rapidprototyping/ebm.htm.

Free Patents Online. Patent Analytics and Patent Searching. Retrieved January 8th, 2007.
from www.freepatentsonline.com.

Goodyear Industrial. Eagle Pd Industrial Power Transmissions. Retrieved February 10,

2007, from
http://www.goodyearindustrialproducts.com/powertransmission/products/pdf/eagl
e_pd_belt.pdf.

Heuft USA, Inc. Container Rejection Systems. Retrieved January 10th, 2007, from
www.heuft.com.

Heuft USA, Inc. Heuft Operator Training. Maintenance Manual. Version 1.0. June 6,
2006. Page 18-21.

KHS. Container Conveying Solutions. Retrieved November 19, 2007 from

www.kisters.com/img/pool/1111_Container%20Conveying%20Systems.pdf.

Norton, Robert L. Machine Design: An Integrated Approach. Upper Saddle River:

Pearson Prentice Hall, 2006.

Norton, Robert L. Design of Machinery. New York: McGraw Hill, 2004.

Rockwell Automation. Motion Analyzer Software. Retrieved February 11, 2007 from
www.ab.com/motion/software/motion_analyzer.html

Rockwell Automation. Allen Bradley Servo Motors. Retrieved January 10, 2007 from
www.rockwellautomation.com.

Stratasys. Case Study. Retrieved January 19, 2007 from

http://www.stratasys.com/uploadedFiles/North_America/Media/PDF%20Beta%20pulley.pdf

Toolcraft Plastics Ltd. Explanation Of and Free Help with Stereolithography process.
Retrieved January 19, 2007 from
http://www.toolcraft.co.uk/help_stereolithography_process_sla_models.htm

104
Appendices

105
Appendix A - Relevant Patents
Title: Apparatus for controlling the path of transportation of articles
Document Type and Number: United States Patent 4986407
Link to this Page: http://www.freepatentsonline.com/4986407.html
Abstract: An apparatus for controlling the path of transportation of articles comprises a
first conveyor belt which delivers the articles to a plurality of second conveyor belt which
delivers the articles to a plurality of second conveyor belts arranged in parallel side-by-
side relationship which continue the conveyance of the articles. A plurality of deflectors
selectively deflect the articles from the first conveyor belt to one of the second conveyor
belts. In order to attain an especially compact construction the first conveyor belt extends
obliquely over and rests on the second conveyor belts and is a belt of such small
thickness that the articles can slide form the first conveyor belt extends obliquely over
and rests on the second conveyor belts and is a belt of such small thickness that the
articles can slide form the first conveyor belt onto one of the second conveyor belts
without the risk of toppling over. The deflectors are arranged on the side of the first
conveyor belt facing away form the direction of conveyance of the second conveyor
belts. The first conveyor belt can be a steel belt having a thickness between 0.1 to 0.5
mm.
Inventors: Heuft, Bernhard
Application Number: 294628
Filing Date: 1988-12-07
Publication Date: 1991-01-22

Figure 70: Patent 4,986,407

106
Title: Linear articulated pusher

Document Type and Number: United States Patent 4643291

Link to this Page: http://www.freepatentsonline.com/4643291.html

Abstract: A pusher mechanism supported on a framework for deflecting objects from a

conveyor has a retractable pusher which is mounted for linear movement transversely
across the conveyor. A paddle is pivotably attached at its midportion to an outer end of
the pusher. A movable link is secured between the framework and an end of the paddle.
Due to these paddle connections, controlled linear movement of the pusher will cause the
paddle to articulate on the pusher from a rest position to an operating position at which an
object is deflected from the conveyor and back to the rest position.

Inventors: Counter, Louis F.; Callies, Fritz A.; Lee, Phillip L.;

Application Number: 820855

Filing Date: 1986-01-21

Publication Date: 1987-02-17

Figure 71: Patent 4,643,291

107
Title: Means for laterally deflecting articles from a path of travel
Document Type and Number: United States Patent 4321994
Link to this Page: http://www.freepatentsonline.com/4321994.html
Abstract: An apparatus for laterally deflecting selected articles from a first conveyor to
one or more other conveyors comprises either extensible and retractable deflective
segments, or gas nozzles whose intensity is adjustable. The deflective segments have
tapered front faces which jointly form a smooth deflecting face whose taper increases in
the direction of conveyance. The number of segments used depends on the speed
component required, and the segments are extended by only a portion of the lateral
distance the articles are to cover, the remainder of the distance being covered by the
imparted inertia. Alternative embodiments include a deflecting wedge or flap.
Inventors: Heuft, Bernhard;
Application Number: 141847
Filing Date: 1980-04-21
Publication Date: 1982-03-30

Figure 72: Patent 4,321,994

108
Title: Apparatus for laterally deflecting articles
Document Type and Number: United States Patent 4369873
Link to this Page: http://www.freepatentsonline.com/4369873.html
Abstract: Apparatus for laterally deflecting articles, such as bottles, from the normal
path of a series of such articles, on the basis of a predetermined criterion, such as size or
shape. The apparatus may be in the form of extensors which operate transversely to the
direction of travel of the articles, in such manner that at any given moment only those
extensors are extended which contact the article then being deflected.
Inventors: Heuft, Bernhard;
Application Number: 002261
Filing Date: 1979-01-10
Publication Date: 1983-01-25

Figure 73: Patent 4,369,873

109
Title: Device for singling out articles from a flow of such articles
Document Type and Number: United States Patent 6588575
Link to this Page: http://www.freepatentsonline.com/6588575.html
Abstract: The device for diverting individual items from a stream of items which are
conveyed on a transport apparatus has an extendable and retractable diversion element ,
which is operated via a gear unit by a drive apparatus moving or swinging to and fro and
is precisely time-controllable, in order to impart a cross impulse to items to be diverted so
that they slide from the transport apparatus across the direction of transport. The
diversion element carries out a complete extension and retraction movement during a
single to-or-fro movement or swing of the drive apparatus. The drive apparatus moving to
and fro can be a pneumatic cylinder, and the gear unit which transmits the piston
movement to the diversion element can be a coulisse link mechanism or a toggle lever.
Inventors: Heuft, Bernhard; Kristandt, Gerd;
Application Number: 030746
Filing Date: 2001-11-01
Publication Date: 2003-07-08

Figure 74: Patent 6588575

110
Title: Container diverter
Document Type and Number: United States Patent 6822181
Link to this Page: http://www.freepatentsonline.com/6822181.html
Abstract: The present invention provides a device and method for selectively removing
an article from a stream or line of similar articles travelling in a pathway on a transport
system such as a conveyor. The invention utilizes a synchronous electric motor which, in
response to a signal to reject a specific article in the stream, rotates an article-contacting
member or paddle into the path of the stream of articles travelling along the pathway
whereby it contacts and smoothly removes that article from the stream. The use of a
synchronous motor to effect the rotation of the paddle is very important to the present
invention.
Inventors: Linton, Fredrick L.;
Application Number: 891616
Filing Date: 2001-06-27
Publication Date: 2004-11-23

Figure 75: Patent 6,822,181

111
Title: Side Transfer Sorting Conveyor
Document Type and Number: United States Patent 3791518
Link to this Page: http://www.freepatentsonline.com/3791518.html
Abstract: A sorting conveyor having a plurality of movable plaques on one conveyor
base arranged to sort merchandise or similar articles carried on another conveyor
positioned at one side of the sorting conveyor. The separation of articles may be
controlled by some particular property of articles such as weight, size, or color, sensed in
a known manner. The plaques on the sorting conveyor are activated to engage the articles
carried by the adjoining conveyor and change their relative position so that a series of
fences can direct the reoriented articles into predetermined paths
Inventors: Vanderhoof, Frank B.;
Application Number: 355,000
Filing Date: 1973-04-27
Publication Date: 1974-02-12

Figure 76: Patent 3,791,518

112
Appendix B - Rapid Prototyping Quotes
A number of quotes were obtained before ordering the rapid prototyped camoid.
A number of different RP techniques and companies were surveyed. The SLA option by
RPDG was eventually chosen for the fabrication of the camoid.

Company Process Material Lead Time Price Discount Final $

RPDG SLA ABS-Like 3 days 1219 20 975.2
SLS Nylon 5 days 3888 0 3888
FDM ABS-Black 3 days 1612 0 1612
FDM PC-White 3 days 1612 0 1612
Objet ABS-Like 3 days 2232 0 2232
RedEye RPM FDM PC-ABS Blend 5 days 1500 10 1350
Realize SLA Accura 25 5 days 3550 10 3195

113
Appendix C - Bill of Materials
Project Budget: Camoid Laner Single Unit
Cost
Item Description Part Number Unit Cost QTY Cost Subtotals
Camoid SLA Part Cost $1,219.00 1 $1,219.00
20% Discount -$244.00 1 -$244.00
Sales Tax $75.56 1 $75.56 $1,050.56
460V 6.1 N-m 2.3
kW MPL-B4520P-
Allen-Bradley Servo Low Inertia Servo MJ22AA $1,352.35 1 $1,352.35
A-B 2090-XXNPMP-
Servo Power Cable Power 14S05 MTR PWR $98.00 1 $98.00
A-B 2090-
Servo Feedback UXNFBMP-S05
Cable Feedback MOTOR CAB $105.53 1 $105.53
10:1 Planetary
Alpha Gearbox Gearbox $300.00 1 $300.00
PLC Modules $0.00
7 Slot ControlLogix
ControlLogix Chassis Chassis 1756-A7 $291.28 1 $291.28
CLX EtherNet/IP
10/100 Bridge
Ethernet Module Module 1756-ENBT $1,159.84 1 $1,159.84
10-31 VDC Input 16
DC Input Module Pts (20 Pin) 1756-OB16E $206.27 1 $206.27
10-31 VDC Elec
Fused Output 16
DC Output Module Pts (20 Pin) 1756-OB16E $332.14 1 $332.14
3 Axis SERCOS
Interface Servo
SERCOS Module Module 1756-M03SE $751.26 1 $751.26
19.2-32 VDC Power
Supply (5V @ 10
Module Power Supply Amp) 1756-PB72 $530.50 1 $530.50
Logix5561
ControlLogix Processor With
Processor 2Mbyte Memory 1756-L61 $3,364.20 1 $3,364.20
Module, Integrated,
400/460V, 6 kW
Conv. 9A Inv 2094-BC01-M01 $1,997.50 1 $1,997.50
SERCOS Fi-Os Cable Fiber Optic Cable $81.00 2 $162.00
Kinetix Ultra3000
Servo Driver 480V AC 3 Phase 2098-DSD-HV050 $1,863.20 1 $1,863.20
Stegmann Shaft 5000 pulse/rev 6-
Encoder 30V DG60L WSR 5000 $300.00 1 $300.00
Stegmann Encoder DOL-2312-
Cable 12-Pin Cable G03MMA3 $85.00 1 $85.00
Next-Day Shipping $40.00 1 $40.00
Banner PhotoEye PE QS18VP6LPQ5 PE QS18VP6LPQ5 $61.85 1 $61.85
Cable 9m $28.23
Bracket $16.54 1 $16.54
$13,022.9
Reflector $5.50 1 $5.50 6
Drive Sprocket Eagle PD 28-Tooth W-28S-H $52.50 2 $105.00
Eagle Pd 720mm
Drive Belt Belt Pd W-720 $35.19 1 $35.19

114
 Sprocket Bushings
Martin QD Bushing H-size 3/4 Bore H QD $9.44 1 $9.44
Martin QD Bushing H-size 32 mm Bore H32 $6.21 1 $6.21
Tax $0.46 1 $0.46
Lovejoy ROSTA
Tensioner SE18 Tension Arm SE-18 ROSTA $41.10 1 $41.10
Lovejoy ROSTA
Roller Idler Smooth Roller Idler 685144-53028 $34.32 1 $34.32
Tax $2.53 1 $2.53 $234.25

Chassis Labor $507.50 1 $507.50

Parts $576.60 1 $576.60
Tax $78.60 1 $78.60
4-Bolt Flange Mount
SKF Bearing 3/4 Shaft Bearing SKF FY TF $29.21 2 $58.42
1/4-20x1 hex head
Hardware bolt $0.04 8 $0.32
1/4 split washer $0.02 9 $0.18
1/4-20 Nylock nut $0.04 1 $0.04
5/16-18x1.25 hex
head bolt $0.04 4 $0.16
5/16 split washer $0.03 8 $0.24
3/8-16x1 hex head
bolt $0.05 8 $0.40
3/8 washer $0.02 8 $0.16
3/8-16 Nylock nut $0.04 8 $0.32
3/4 washer $0.05 4 $0.20
3/16x3/16x1 key $0.02 3 $0.06
3/4 shaft collar $2.35 1 $2.35 $1,225.55

TOTAL $15,533.32
Table 21: Full Bill of Materials

115
 Project Budget: Follow-up Cost
Item Description Part Number Unit Cost QTY Cost Subtotals
Camoid SLA Part Cost $1,219.00 1 $1,219.00
20% Discount -$244.00 1 -$244.00
Sales Tax $75.56 1 $75.56 $1,050.56
460V 6.1 N-m 2.3 kW
Allen-Bradley Servo Low Inertia Servo MPL-B4520P-MJ22AA $1,352.35 1 $1,352.35
A-B 2090-XXNPMP-
Servo Power Cable Power 14S05 MTR PWR $98.00 1 $98.00
Servo Feedback A-B 2090-UXNFBMP-
Cable Feedback S05 MOTOR CAB $105.53 1 $105.53
Alpha Gearbox 10:1 Planetary Gearbox $300.00 1 $300.00
PLC Modules $0.00

3 Axis SERCOS Interface

SERCOS Module Servo Module 1756-M03SE $751.26 1 $751.26

Module, Integrated,
400/460V, 6 kW Conv. 9A Inv 2094-BC01-M01 $1,997.50 1 $1,997.50
SERCOS Fi-Os
Cable Fiber Optic Cable $81.00 2 $162.00
Kinetix Ultra3000
Servo Driver 480V AC 3 Phase 2098-DSD-HV050 $1,863.20 1 $1,863.20
Stegmann Shaft
Encoder 5000 pulse/rev 6-30V DG60L WSR 5000 $300.00 1 $300.00
Stegmann Encoder
Cable 12-Pin Cable DOL-2312-G03MMA3 $85.00 1 $85.00
Next-Day Shipping $40.00 1 $40.00
Banner PhotoEye PE QS18VP6LPQ5 PE QS18VP6LPQ5 $61.85 1 $61.85
Cable 9m $28.23
Bracket $16.54 1 $16.54
Reflector $5.50 1 $5.50 $7,138.73
Drive Sprocket Eagle PD 28-Tooth W-28S-H $52.50 2 $105.00
Drive Belt Eagle Pd 720mm Belt Pd W-720 $35.19 1 $35.19

Sprocket Bushings

Martin QD Bushing H-size 3/4 Bore H QD $9.44 1 $9.44

Martin QD Bushing H-size 32 mm Bore H32 $6.21 1 $6.21

Tax $0.46 1 $0.46
Lovejoy ROSTA
Tensioner SE18 Tension Arm SE-18 ROSTA $41.10 1 $41.10
Lovejoy ROSTA
Roller Idler Smooth Roller Idler 685144-53028 $34.32 1 $34.32
Tax $2.53 1 $2.53 $234.25
Chassis Labor $507.50 1 $507.50
Parts $576.60 1 $576.60
Tax $78.60 1 $78.60

116
 4-Bolt Flange Mount 3/4 Shaft
SKF Bearing Bearing SKF FY TF $29.21 2 $58.42
Hardware 1/4-20x1 hex head bolt $0.04 8 $0.32
1/4 split washer $0.02 9 $0.18
1/4-20 Nylock nut $0.04 1 $0.04
5/16-18x1.25 hex head bolt $0.04 4 $0.16
5/16 split washer $0.03 8 $0.24
3/8-16x1 hex head bolt $0.05 8 $0.40
3/8 washer $0.02 8 $0.16
3/8-16 Nylock nut $0.04 8 $0.32
3/4 washer $0.05 4 $0.20
3/16x3/16x1 key $0.02 3 $0.06
3/4 shaft collar $2.35 1 $2.35 $1,225.55

TOTAL $9,649.09
Table 22: Bill of Materials with Piggy Backed Electronics

117
Appendix D - Motion Analyzer Input Values
Each table represents a tab for the motor sizing application. The software can be
downloaded for free at www.ab.com/motion/software/motion_analyzer.html.

Axis Setup
Load Type Rotary
Motor Type Rotary
Actuator Type Non-selected
Voltage Selection
Supply Type AC 3-phase
Voltage Type Single
Nominal Voltage (volts) 480
Tolerances 10%
Motor Parameters
Max ambient (C) 40
Brake NO
Table 23: Axis Setup Tab

Cycle profile
Cycle Profile mode: Multi-segment
Auto Compile: ON
Segment Number 1 2 3 4 5 6 7 8 9

Curve Type L L L L L L L L L
(Linear/S-curve)
Initial Velocity 0 0 878.6 878.6 0 0 1098 1098 1174
(deg/sec)
Final Velocity 0 878.6 878.6 0 0 1098 1098 1174 0
(deg/sec)
Distance (deg) 0 32.8 140 7.5 0 7.5 140 10 65.1

Time (sec) 0.413 0.07466 0.07466 0.01707 1.593 0.01367 0.1276 0.008805 0.1109

Acc/Dec (rpm/sec) 0 1961 1961 -8577 0 13380 0 1448 -1764

Thrust (N-m) 0 0 0 0 0 0 0 0 0
Added Inertia 0 0 0 0 0 0 0 0 0
(kg-m)
Denotes an entered value
Table 24: Cycle Profile Tab

118
Mechanism
Primary Inertia
0.0059
(kg-m)
Secondary Inertia (kg-m) 0

Secondary Mass (kg) 0

Losses (N-m) 0
Starting Angle (deg) 220

Axis Separation (mm) 0

Table 25: Mechanism Tab

Transmission Stages
Transmission Belt Drive
Ratio 10
Inertia (kg-m) 0.0005
Efficiency 98%
Friction Torque (N-m) 0
Table 26: Transmission Stages Tab

119
Appendix E - Servo Details
The following motor descriptions were gathered from Rockwell Automations
web catalog which can be found at
http://literature.rockwellautomation.com/idc/groups/literature/documents/pp/mp-pp001_-
en-p.pdf.

Figure 77: Servo Overview

120
121
Appendix F - Servo Driver Details
All information gathered from Rockwell Automation Publication
http://www.ab.com/motion/controllers/2098-BR002A-EN-P_1001.pdf

Figure 78: Driver Details and Benefits

122
Figure 79: Driver Specifications

123
Figure 80: Driver Specifications

124
Appendix G - Chassis Drawings
The following technical drawings were sent to the metalworking contractor. They detail
the construction of the chassis components, as well as its assembly. All drawings were
made in PTC Pro/ENGINEER Wildfire 2.0. The latest files of each part (PDF, .drw, and
.prt files) are included in the electronic version of this report.

125
126
127
128
129
130
131
132
Appendix H - Goodyear Eagle Pd Power Transmission

Figure 81: Belt Nomenclature

133
Figure 82: Eagle Pd Belt Product Numbers

134
Figure 83: Sprocket Nomenclature

135
Figure 84: Eagle Pd White Sprockets

Screenshots from Goodyear Eagle Pd technical data and product specifications

catalog found at:
http://www.goodyearindustrialproducts.com/powertransmission/products/pdf/eagle_pd_b
elt.pdf

136
Figure 85: Belt Nomenclature

137
Figure 86: Eagle Pd Belt Product Numbers

138
Figure 87: Sprocket Nomenclature

139
Figure 88: Eagle Pd White Sprockets

140
Appendix I - Bottle Tests
I.1 - Center of Gravity Test

An experiment was conducted to test the center of gravity of a full wine bottle.
Figure 89 demonstrates the experiment set up. The center of gravity is measured through
the use of a pendulum. The period of a pendulum is a function of the force of gravity and
length of the pendulum arm. The relationship is:
l
T = 2
g
Where T is time of one period, l is the length of pendulum from an objects center
of gravity, and g is the acceleration due to gravity. By measuring the time of one period
we are able to obtain the effective length (from pivot to center of gravity) of the
pendulum by rearranging the above relationship:
2
T
l = *g
2
Since we know the dimensions of the bottle and length of the rope we use for the
pendulum test, we are able to calculate the location of the center of gravity in relation to
the bottle. Assumptions are also made that the bottle is symmetric about two axes such
that the center of gravity is in the center of the bottle at some height on the z axis (see
Figure 90).

141
Figure 89: Center of Gravity Experiment

142
 Figure 90: Free Body Diagram

Figure 90 shows the free body diagram of the bottle during the experiment. FG is
the force of gravity acting at the center of gravity. CG is the center of gravity. LR is the
length of rope used for the experiment. Leff is the effective length of the pendulum acting
at the center of gravity. Several tests were conducted with different lengths of rope. The
time was recorded for ten periods to pass. Several runs were conducted for each length of
rope. The center of gravity is calculated relative to the base of the bottle. Time was
recorded with a computer stop clock accurate to 0.05 seconds. Length was measured with
a tape measure accurate to 1/8 inch.

143
 Center of Gravity Pendulum Test
Height
Number Total Effective Height
String Period from
Trial of Time Length from Top
Length (s) Bottom
Periods (s) (in.) (in.)
(in.)
1 32.6 10 20.3 2.0 40.5 7.9 4.4
2 32.6 10 20.3 2.0 40.5 7.9 4.4
3 32.6 10 20.3 2.0 40.5 7.9 4.4
4 27.3 10 18.8 1.9 34.5 7.2 5.0
5 27.3 10 18.8 1.9 34.5 7.2 5.0
6 16.5 10 15.9 1.6 24.7 8.2 4.0
7 16.5 10 16.0 1.6 25.0 8.5 3.7
8 16.5 10 15.9 1.6 24.7 8.2 4.0
9 42.1 10 22.7 2.3 50.4 8.3 3.9
10 42.1 10 22.6 2.3 50.0 7.8 4.4
Average 7.9 4.3

Figure 91: Pendulum Test Results

144
I.2 - Coefficient of Static Friction Test

An experiment was conducted to test the coefficient of static friction of a full

wine bottle on the two materials commonly used for conveyor systems on the Gallo
bottling lines. Figure 92 demonstrates the experiment set up. A section of conveyor links
is supported and assured to be level using an inclinometer. The conveyor is lubricated
with detergent (glycol based). With the bottle on the conveyor section, the conveyor is
inclined until the bottle begins to slide. The angle is measured using the inclinometer.

Figure 92: Coefficient of Static Friction Experiment

145
 Figure 93: Free Body Diagram

Figure 93 shows the free body diagram of the bottle during the experiment. FG is the
force of gravity acting at the center of gravity. FN is the normal force of the conveyor. FF
is the friction force of the conveyor. is the angle of the tip of the conveyor. The friction
coefficient is obtained by calculating Tan(). Two separate tests were conducted with
plastic and stainless steel conveyor materials and five trials on each material were
conducted.

146
 Plastic Stainless Steel
Angle Coefficient Angle Coefficient
Trial Trial
(degrees) Static Friction (degrees) Static Friction
1 7 0.122784561
1 7 0.122784561
2 7 0.122784561
2 7.5 0.131652497
3 7.5 0.131652497
3 7 0.122784561
4 6.5 0.113935608
4 7 0.122784561
5 7 0.122784561
5 7.5 0.131652497
Average 0.124721109
Average 0.126331735
Figure 94: Plastic Friction Test
Figure 95: Stainless Steel Friction Test

147
I.3 - Bottle Neck Failure Test
An experiment was conducted to test the failure load of the bottle neck using a
static load acting at an extreme point on the bottle neck. Figure 96 demonstrates the
experiment set up. The bottle overhangs off a solid surface with a rope attached to the
bottle neck and a means to supply the force.

Figure 96: Bottle Neck Failure Experiment

148
Figure 97: Free Body Diagram

Figure 97 shows the free body diagram of the bottle during the experiment. The
force is provided by weights hanging from a rope attached at the point FL. The bottle is
supported by a table at point FT and rotation is prevented by FS acting at the base of the
bottle. The moment is calculated at two points of interest; the point at M1 and at the
speculated weak point (highest stress concentration factor) at M2.

Bottle Neck Test

Force
Moment 1 Moment 2 Failure?
Mass (kg) (Newtons)
(N-m) (N-m) (yes/no)
FL
4 39.2 4.6 2.3 NO
8 78.5 9.1 4.6 NO
12 117.7 13.7 6.9 NO
20 196.2 22.8 11.6 NO
30 294.3 34.1 17.4 NO

The results of the test concluded that the bottle neck was sufficiently strong to
withstand the static load of at least 30 kg. Failure was not achieved because of the lack of
equipment and the obvious dangers associated with broken glass.

149
Appendix J - Rapid Prototyping Methods
Several different methods of rapid prototyping exist. Each method poses
particular benefits and constraints. All use similar basic concepts to construct complex
geometry from thin layers. A few available methods are investigated below.

J.1 - Stereolithography (SLA)

Stereolithography is the most commonly used rapid prototyping technique and
can produce complex geometries shown in Figure 16. This method uses a screened
platform in a reservoir of liquid photopolymer. This liquid hardens when exposed to
ultraviolet light. The platform is lowered into the tank of liquid so that only a thin film of
the photopolymer (from .125 to .2 mm 17) is above its surface. Then an ultraviolet laser
traces the cross section, leaving a hardened layer of plastic. The platform then lowers
one layer thickness, where a roller spreads the photopolymer to ensure the correct
thickness, and the process is repeated. Once the part is finished, the excess liquid is
washed off, and the part typically requires a curing operation to harden the photopolymer
completely.
Stereolithography is very accurate, but relatively expensive and somewhat slower
than some other RP methods. Parts may need support structures during fabrication.
Pieces formed with SLA may only be suitable for limited function, as the material is not
the most robust or durable. SLA is offered by 3D
Systems.
Figure 98 18 shows a schematic view of the
SLA fabrication process. As the platform is lowered
into the tank, the SLA model is formed out of the
photopolymer liquid resin by the suspended laser,
layer by layer. After the forming process the part
must still be finished and cured. SLS and EBM (see
below) have similar machine setups, but different
reservoir material and, in the case of EBM, an
electron beam in place of the laser.

Figure 98: SLA Schematic Diagram

17
Toolcraft Plastics Ltd. Explanation Of and Free Help with Stereolithography process. Retrieved January
19, 2007 from http://www.toolcraft.co.uk/help_stereolithography_process_sla_models.htm
18
Figure retrieved from Toolcraft Plastics Ltd.

150
J.2 - Fused Deposition Modeling (FDM)
FDM builds a solid part from a plastic wire that is heated and laid in a bead, layer
upon layer. FDM suffers slightly less accuracy than stereolithography, but offers greater
flexibility in materials. Available materials include polycarbonate and ABS plastic,
which are durable enough to be used not only as prototypes, but even in full production
settings. One manufacturer replaced a pulley on an industrial belt sander on its
production line with an FDM formed part when the aluminum piece failed. The rapid
prototyped piece lasted over one month in full production 19. This process is owned by
Stratasys, Inc.

J.3 - Selective Laser Sintering (SLS)

SLS uses a bed of powder, which may be plastic, ceramic or metal in a similar
manner as stereolithography uses photopolymer liquid. The powder is spread evenly
across a piston head and then sintered with a laser to form. The piston lowers and the
process is repeated.
SLS pieces are less accurate than those formed with stereolithography, but have
no need for support structures or secondary curing. In addition, SLS allows for a great
variety of materials to be used. Sintering does result in porous parts, and may require an
infiltration process to improve the material properties of the part. Some larger pieces
may require up to two days of cooling after the sintering process.

J.4 - Electron Beam Melting (EBM)

A newer development in rapid prototyping is Electron Beam Melting. The
process is very similar to SLS, but instead of a laser, an electron beam fully melts the
powder. The result is a fully homogenous non-porous part, typically made of titanium.
The piece requires no secondary operations, but is suitable for any conventional finish
process, including heat treatment. Build layers are quite accurate, having a thickness
from 0.05 to 0.2 mm 20.
The process is roughly five times faster than SLS and up to ten times as efficient.
This is because the laser used in SLS is largely reflected from the powder surface, while
the electron beam has much higher absorption. Stratasys offers EBM machines in North
America.

19
Stratasys. Case Study. Retrieved January 19, 2007 from
http://www.stratasys.com/uploadedFiles/North_America/Media/PDF%20Beta%20pulley.pdf
20
Engineers Handbook. Rapid Prototyping: Electron Beam Molding. Retrieved January 26, 2007 from
http://www.engineershandbook.com/Rapidprototyping/ebm.htm.

151
Appendix K Test Data Table

Table 27: Data Recording Table

152
Appendix M - Detailed Mathematics

153
Summary
This analysis serves several purposes. First of all, we are looking to find an "optimal" contour for the cam of
our divider. This requires a look at the accelerations and forces that will arise from shifting this bottle across
the lane. Secondly, we want to find the height (z-direction) at which our divider should contact the bottle, in
order to prevent tipping. Finally, a look at the forces, stresses, and impacts involved with the splitting motion is
necessary to ensure that bottles will not be broken.

There are two approaches to defining the contour of this laner. The first is the Bottom-Up method; describe the
y-value (height) of the bottle as it moves downstream, and then to perform acceleration and force analyses to
determine if the contour is acceptable. Alternately, with the Top-Down method the acceleration curve can be
described first, and the contour follows from there.

We have used both methods in this analysis. For instance, the Simple Harmonic Contour is an example of the
Bottom-Up method, while ModTrap is an example of Top-Down.

Problem Definition
Constants:

g 9.807 tan( 7 deg ) 0.123

m
2
s

Bottle Properties:
mbottle 1.29 kg Bottle Mass dbottle 2.982 in Diameter of Bottle

thickness .125 in Glass Thickness Fg mbottle g Weight of Bottle

hbottle 12.25 in Bottle Height Fg 12.651 N

hcg 4.33 in Center of Gravity Height

d
bottle 2
2
mbottle 3 hbottle
Ixx
2 Ixx 0.011 kg m
2
Iyy Ixx
12

These are very rough estimates, assuming a solid,

dbottle
2 uniform density cylinder.

Izz mbottle
2 Izz 9.251 10
4
kg m
2
2

Line Properties:
vline 1.01
m
Conveyor Velocity
s
Design Parameters
vline 39.764
in
Laner Properties: s

hcontact 3.5 in Height at which the laner contacts the bottle

h( Z) Z hcg Distance from contact point to center of gravity

Lsplit 9.5 in Total length of the laner track

hy 2.72 in Target "height" (across the conveyor)

Time Calculations:
Lsplit
tfinal Time to lane one bottle (guided by contour)
vline tfinal 0.239 s

tsplit 0 s 0.0025 s t final

X-Direction Travel:
x( t ) vline t

0.2
x t split
0.1

0
0 0.05 0.1 0.15 0.2 0.25
t split

We make the assumption that the bottle speed moving downstream remains essentially constant. The
divider will actually slow the bottle, but we argue the loss in speed will be insignificant.
Heuft Laner Acceleration
In order to gain a baseline understanding of the accelerations the bottle experiences currently, we will analyze the
contour of the Heuft laning system. The analysis is conducted by plotting points along the contour and calculating for
the best fit trend line. In this case, a cubic equation was chosen with root mean square value of .9999 (99.99%
accuracy).

yheuft( t ) ( .0011 .0254 )
x( t)
( .0166 .0254 )
x( t)
3 2
( .0362 .0254 ) ( .0007 .0254 ) m
x( t )
in in in

0.1

0.05
yheuft t split
0

0.05
0 0.05 0.1 0.15 0.2 0.25
t split

yheuft t final 2.784 in

vheuft( t )
d
yheuft( t)
dt

vheuft t split 0.5

0
0 0.05 0.1 0.15 0.2 0.25
t split
 2
aheuft( t )
d
y ( t)
2 heuft
dt

aheuft t split
2

1
0 0.05 0.1 0.15 0.2 0.25
t split

Laner Contour:
The contour of our rail can be modeled as a 4-5-6-7 cam polynomial funtion of time with a series of
coefficients. We set a series of boundary conditions to solve for the necessary contour.

ygeneral ( t C ) C C C C C C C C
2 3 4 5 6 7
t t t t t t t
0 1 s 2 s
3 s
4 s
5 s
6 s
7 s

y'general( t C ) C 2 C 3 C 4 C 5 C 6 C 7 C
2 3 4 5 6
t t t t t t
1 2 s 3 s
4 s
5 s
6 s
7 s

y''general ( t C ) 2 C 6 C 12 C 20 C 30 C 42 C
2 3 4 5
t t t t t
2 3 s 4 s
5 s
6 s

7 s

y'''general ( t C ) 6 C 24 C 60 C 120 C 210 C

2 3 4
t t t t
3 4 s 5 s
6 s
7 s

1
1

This matrix is a series of guess coefficients for the solver to

1
determine our actual coefficients.

1
C general hy
1
1
1

1
Boundary Conditions
Given These conditions are selected to minimize the risk of dangerous dynamics
in the lane splitting.

Initial Conditions

Final Conditions
0 = ygeneral 0 C general

hy = ygeneral t final C general

0 = y'general 0 C general
.6m = y'general tfinal C general

0 = y''general 0 C general
0 = y''general t final C general

0m = y'''general 0 C general
0m = y'''general tfinal C general

C out Find C general
0


0
0


0

C out 3.237 103 in This set of coefficients describes our contour as an 8 term
polynomial equation.
1.077 10
4

3
7.965 10
4
4.582 10

0.08

0.06

ygeneral t split Cout
0.04
flip

hy 0.02

0.02
0 0.05 0.1 0.15 0.2 0.25
t split
Simple Harmonic:
x( t ) h
ysin( t) hy cos y
2Lsplit

0.08

0.06
ysin t split

flip 0.04
hy

0.02

0
0 0.05 0.1 0.15 0.2 0.25
t split

Analysis
Bottle Considerations:

Ffriction mbottle g Ffriction 1.553 N Friction Force between Bottle and Conveyor

2
asin( t)
d
y (t)
2 sin
dt

Mod-Trap Acceleration

amax 3
m
The Modified Trapezoid is a piecewise acceleration
2 curve; partially sinusoidal and partially constant
s
acceleration.

atrap( t) a max sin

2t m if t
tfinal

Lsplit s 4

t final tfinal
a max if t 3
4 4

2t m 2t final
t tfinal
tfinal
a max sin if 3
Lsplit s 4 10

atrap tfinal 0.187
m
2
s
Acceleration Comparison
awall( a t ) a ( t )

Bottle Transverse Acceleration

6

Acceleration (m/s^2)

awall aheuft t split

awall asin t split 4

awall atrap t split

y''general t split Cout

2

0
0 0.05 0.1 0.15 0.2 0.25
t split

Contour Contact Time (s)

Velocity

t
vwall( a t ) a ( t ) dt

0

Bottle Transverse Velocity

0.8

Bottle Velocity (m/s)

vheuft t split

0.6
vwall asin t split

vwall atrap t split 0.4

y'general t split Cout

0.2

0
0 0.05 0.1 0.15 0.2 0.25
t split

Contour Contact Time (s)

vheuft t final

vwallasin tfinal 0.656 25.824
v 0.454 17.884
wall trap final
vfinal vfinal vfinal
m in
t
0.592 23.313
a
y'generalt final C out
s s
0.6 23.622

s
Displacement

t
t
swall( a t ) a ( t ) dt dt
0

0

Contour Displacement Profile

0.08

yheuft t split

0.06
swall asin t split
Contour Profile

swall atrap t split

0.04

ygeneral t split Cout 0.02

hy
0

0.02
0 0.05 0.1 0.15 0.2 0.25
t split

Contour Contact Time (s)

yheuft tfinal
0.071 2.784
swall asin t final 0.069 m 2.72 in
sfinal sfinal sfinal
swall atrap tfinal 0.07
0.069
2.758
2.72
general tfinal Cout
y
Displacement Comparison

diffsin t split yheuft t split swall a sin tsplit

difftrap t split yheuft tsplit swall a trap tsplit
diffpoly t split yheuft t split ygeneral t split C out

Displacement Difference: Heuft vs. Cam Programs

Displacement Difference (in)

diffsin t split
in

0.2
difftrap t split

in

diffpoly t split 0
in

0.2
0 0.05 0.1 0.15 0.2 0.25
t split

Contour Contact Time (s)

Jerk
In addition to acceleration, velocity, and displacement, we can look at the jerk over the division.

jwall( a t )
d
a( t )
dt

Bottle Transverse Jerk

100

j wall aheuft t split
j wall asin t split
50
Jerk (m/s^3)

j wall atrap t split 0

y'''general t split Cout

50

100
0 0.05 0.1 0.15 0.2 0.25
t split

Contour Contact Time (s)

Force
Fwall( a t ) a ( t ) mbottle
Foptimal t split C out y''general tsplit C out mbottle

Force on Bottle
8

F wall aheuft t split
F wall asin t split
Force (Newtons)

F wall atrap t split

F optimal t split Cout

4

F friction 2

0
0 0.05 0.1 0.15 0.2 0.25
t split

Contour Contact Time (s)

Contour Comparison Table

The following table compares the peak values of velocity, acceleration and jerk of the contour curves. This will
influence the decision when choosing the optimal contour for the cam.

Max Velocity Max Acceleration

vheuft tfinal a heuft tfinal

vwall asin t final
a sin( 0 )

v
wall trap final
vmaximum t a trap( .1s)
amaximum
a
y'general tfinal Cout t
y''general final Cout
2

s

2
s

0.656 25.824 3.851 151.633

vmaximum 0.454 m vmaximum 17.884 in amaximum 2.987 m amaximum 117.581
0.592 s 23.313 s 3 2 118.11
0.6 23.622 4.709 s 185.389

Max Jerk Max Force

j wall aheuft tfinal Fwall aheuft tfinal

jwall asin tfinal Fwall a sin 0

jwall atrap 0
jmaximum Fmaximum Fwallatrap .1s
t
y'''general final Cout Foptimal final Cout
t

4 2


3 2
s s

10.54 414.961 4.968 1.117

19.636 773.073 in 3.853 N 0.866 lbf
jmaximum jmaximum 3 Fmaximum Fmaximum
m
78.117 3
3.075 10 s3 3.87 0.87
57.263 s
2.254 103 6.074 1.366

Tip Test
Free Body Diagram
A primary concern in this application is to ensure that no bottles are
tipped over. A check is performed here that the force from the divider
never creates a moment exceeding the tipping moment of the bottle.

M wall( a t h) Fwall( a t) h Moment Induced by Wall

dbottle
M g Fg
2

M optimal( t C h) Foptimal( t C ) h
 Induced Moment on Bottles
0.6

Mwall aheuft t split hcontact
Mwall asin t split hcontact
Moment (N-m)

0.4

Mwall atrap t split hcontact

Moptimal t split Cout hcontact

0.2
Mg

0
0 0.05 0.1 0.15 0.2 0.25
t split

Contour Contact Time (s)

Maximum Moment

M wallaheuft t final hcontact

M wall a sin 0 hcontact
0.442 3.909
M maximum Mwall atrap .125s hcontact M maximum 0.343 N m M maximum 3.031 lbf in
t 0.344 3.045
M optimal final Cout hcontact 0.54 4.78
2


2
s

Bottle Trajectory
Since the contour does not guide the bottle the entire distance across the
conveyor, a check is performed to ensure the bottle's trajectory will carry the
bottle across the division.
Angle of Trajectory

t s C out
( a t) atan a ( t ) dt
0 m

gen tsplit C out atan y'general tsplit
m

Angle of Departure
40

aheuft t split
deg

30
asin t split
Angle (deg)

deg

atrap t split 20

deg

gen t split Cout
10
deg

0
0 0.05 0.1 0.15 0.2 0.25
t split

Contour Contact Time (s)

Final Angle of Departure

aheuft t final
31.772
asin t final 24.43 deg

atrap tfinal 30.632
30.964

gen t final Cout

Bottle Velocity Components

vline sin 0
0.532 20.937
vline sin 1 0.418 16.445
V x Vx Vx
m in

vline sin 2 0.515 s 20.26 s

v sin 0.52 20.458

line 3

vline cos 0
0.859 33.805
vline cos 1 0.92 36.204
V y Vy Vy
m in

vline cos 2 0.869

0.866
s 34.215
34.097
s
v cos
line 3

Bottle Friction Deceleration

Ffriction
afriction afriction 1.204
m
Acceleration due to Friction
mbottle 2
s

Line Constants

Lglide 18in
Length of Unsupported Trajectory

Lglide
tfin tfin 0.453 s
vline Time of Trajectory

ttraj 0 s 0.025 s tfin

X Displacement of Bottle of Trajectory (Downline)

xtraj_heuft ( t ) V x t xtraj_sin ( t) V x t xtraj_trap ( t) V x t

0 1 2 xtraj_gen ( t ) V x t
3

xtraj_heuft t traj
xtraj_sin t traj 0.2

xtraj_trap t traj

xtraj_gen t traj
0.1

0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
t traj

Y Displacement of Bottle Trajectory (Across Conveyor)

ytraj_heuft ( t ) V y t .5 a friction t ytraj_trap ( t) V y t .5 afriction t

2 2
0 2

ytraj_sin ( t) V y t .5 a friction t
2
ytraj_gen ( t ) V y t .5 a friction t
2
1
3

ytraj_heuft t traj

ytraj_sin t traj

0.2
ytraj_trap t traj

ytraj_heuft t traj

0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
t traj
Trajectory Curve

Tc_heuft( t ) xtraj_heuft ( t ) ytraj_heuft( t) Tc_trap ( t ) xtraj_trap ( t) ytraj_trap ( t)

2 2 2 2

Tc_sin ( t) xtraj_sin ( t) ytraj_sin ( t) Tc_gen ( t ) xtraj_gen ( t) ytraj_gen ( t )

2 2 2 2

Bottle Trajectory Curve

Tc_gen t traj 0.3
Displacement (m)

Tc_sin t traj

Tc_trap t traj
0.2

Tc_gen( t )
0.1

0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
t traj

Trajectory Time (s)

Notice that the trajectories of the proposed contours all produce similar trajectories to the Heuft model
currently in place. This reduces any concern for the proposed contours' ability to produce a bottle trajectory
that clears the dividing wedge.

Timing
The nature of the cam is such that one half revolution is a full rise cycle of the cam. The cam must complete a full
rise at the same speed of the bottle. Thus the cam must make one half revolution in a minimum t final.

Lsplit
tfinal
vline

tfinal 0.239 s

op op 97.665 rpm
140deg
Operating Angular Velocity
tfinal
 Angular Acceleration
The successful operation of this system is based upon the fact that the system must be binary, i.e. either
on or off. Binary operation is only achieved when the system is at full operating speed during the cam rise
actuation. In order to accomplish this, the cam design incorporates a buffer dwell angle to allow for servo
acceleration. The buffer dwell angle is based on a spin up time specified. The specified spinup time has a definite
range whose extremities are governed by several criteria. The lower extreme of spinup time is governed by servo
torque. The upper extreme is governed by line speed, bottle spacing and cam geometry.
There is often a spacing between the bottles on the line, however it is not unusual for bottles to be directly
adjacent to eachother. In this case, the spacing between the bottles is a bottle diameter. The worst case time
between the bottles is thus full line speed and one bottle diameter spacing.

dbottle
tworst tworst 0.075 s Worst Case Scenario Spin Up Time
vline

The fastest a servo can accelerate is based on its torque, motor inertia and inertia of the cam. We can specify an
arbitrary lower time of 0 for spinup.

op 97.665 rpm tspinup 0.001s .002s tworst

buffer t spinup t spinup op

Acceleration Buffer Angle vs. Spin-up Time

60
Buffer Angle (deg)

40
buffer t spinup

deg
20

0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
t spinup

Spin-Up Time (s)

 op

t spinup
tspinup

1.5 10
4

1 10
4

t spinup
5000

0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
t spinup

Torque

3 2
Izz_cam 18.64lb in Izz_cam 5.455 10 s m N
2
Cam Moment of Inertia

Izz_motor 3kg cm
2
Motor Moment of Inertia

Tservo tspinup t spinup Izz_cam Izz_motor

Servo Torque vs. Spin Up Time

60
Torque (N-m)

40

Tservo t spinup
20

0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
t spinup

Spin Up Time (s)

 Servo Torque vs. Spin Up Time
100
Torque (N-m)

10

Tservo t spinup
1

1 10
0.1 3
0.01 0.1
t spinup

Spin Up Time (s)

Cam Physical Properties
M cam 5.332 lb M cam 2.419 kg Mass of Cam

V cam 122.99 in V cam 2.015 10 cm

3 3 3
Volume of Cam

3
Izz_cam 18.64 lb in Izz_cam 5.455 10 kg m
2 2
Cam Moment of Ineria about Rotational Axis

Shaft Properties

steel 7.8
gm
Denisty of Steel
3
cm

Lbearing 1in Bearing Length

Wgear 2in Gear Width

Lshaft Lsplit 2Lbearing Wgear 2.5in Lshaft 16 in Shaft Length

D shaft .75in Shaft Diameter

D shaft
4
9
Izz_shaft .25 Izz_shaft 6.465 10
4
Moment of Inertia
2
m

D shaft
4
8
Jshaft Jshaft 1.293 10
4
Polar Moment of Inertia
2 2
m

D shaft
2
V shaft Lshaft V shaft 115.833 cm
3
Volume of Shaft
2

mshaft steel V shaft mshaft 0.903 kg Mass of Shaft

Keyway Analysis

Wkey
3
in Width of Key
16
Lkey 1.5in Length of Key

D key
3
in Depth of Key
32

Syt_ABS 52MPa Tensile Yield Strength of ABS Plastic

Syt_A25 38MPa Tensile Yield Strength of Accura 25 Plastic

Syt_steel 250MPa Tensile Yield Strength of Steel

Shear Stress Analysis

Tmax Tservo( .01367 s) Tmax 4.306 N m Maximum Torque Transmission

Tmax
Fshear Fshear 452.029 N Shear Force
D shaft

2
4
Asec Wkey Lkey Asec 1.815 10
2
m Cross Sectional Area of Key

Fshear
ave ave 2.491 MPa Average Torsional Shear Stress
Asec
Safety Factor

Sys_ABS 0.5 Syt_ABS Sys_ABS 26 MPa Theoretical Yield Strength

Sys_A25 0.5 Syt_A25 Sys_A25 19 MPa

Sys_steel 0.5 Syt_steel Sys_steel 125 MPa

Sys_ABS
ABS ABS 10.437 Safety Factor
ave

Sys_A25
A25 A25 7.627
ave

Sys_steel
steel steel 50.177
ave

Shaft Analysis

Gsteel 216GPa Modulus of Elasticity

Tmax Lshaft
shaft shaft 0.036 deg Shaft Torsional Deflection
Gsteel Jshaft

D shaft
Tmax
shaft shaft 3.172 MPa
2
Torsion Stress in Shaft
Jshaft
Fatigue Failure Analysis
Torsional Fatigue

Sut 400MPa Sut 400 MPa Ultimate Tensile Strength

Se' .5 Sut Se' 200 MPa Uncorrected Endurance Limit

C load 1 For Torsional Loading

.097
D shaft
C size 0.869 Size Correction
in

A 4.51 b .265

Sut
b
C surface A C surface 0.922 Surface Finish Correction
MPa

C temperature 1.0 Ambient Temperature Correction

C reliability .753 Material Reliability Correction

Se C load C size C surface C temperature C reliability Se' Corrected Fatigue Function

Se 124.049 MPa

Safety Factor

Se
torsion torsion 39.109
shaft
Keyway Stress Concentration Factors

a 0.039 a 0.039
2
Neuber's Constant

r .01 Notch Radius

q
1
Notch Sensitivity Factor
1
a
r

Kt 2.62 Static Stress Concentration Factor

Peterson stress concentration factor for flat end
mill. (furlong lecture 19, machine design 473)

Kfs 1 q Kt 1 Kfs 2.165 Fatigue Stress Concentration Factor

Shear Stress Concentration due to Torsion

concentration Kfs shaft concentration 6.869 MPa

Safety Factor

Sys_steel
keyway keyway 18.199
concentration