The Effects of blanket tension and squeeze pressure on

Rochester Institute of TechnologyRIT Scholar Works

Theses Thesis/Dissertation Collections

5-1-1991

The Effects of blanket tension and squeeze pressureon slur and doubling using conventional andcompressable blankets in sheetfed offsetlithographyAlexander M. Voutsas

Follow this and additional works at: http://scholarworks.rit.edu/theses

This Thesis is brought to you for free and open access by the Thesis/Dissertation Collections at RIT Scholar Works. It has been accepted for inclusionin Theses by an authorized administrator of RIT Scholar Works. For more information, please contact [email protected].

Recommended CitationVoutsas, Alexander M., "The Effects of blanket tension and squeeze pressure on slur and doubling using conventional andcompressable blankets in sheetfed offset lithography" (1991). Thesis. Rochester Institute of Technology. Accessed from

http://scholarworks.rit.edu?utm_source=scholarworks.rit.edu%2Ftheses%2F3776&utm_medium=PDF&utm_campaign=PDFCoverPages

http://scholarworks.rit.edu/theses?utm_source=scholarworks.rit.edu%2Ftheses%2F3776&utm_medium=PDF&utm_campaign=PDFCoverPages

http://scholarworks.rit.edu/etd_collections?utm_source=scholarworks.rit.edu%2Ftheses%2F3776&utm_medium=PDF&utm_campaign=PDFCoverPages

http://scholarworks.rit.edu/theses?utm_source=scholarworks.rit.edu%2Ftheses%2F3776&utm_medium=PDF&utm_campaign=PDFCoverPages

http://scholarworks.rit.edu/theses/3776?utm_source=scholarworks.rit.edu%2Ftheses%2F3776&utm_medium=PDF&utm_campaign=PDFCoverPages

mailto:[email protected]

The Effects of Blanket Tension and Squeeze Pressure on Slur and

Doubling Using Conventional and Compressible Blankets in Sheetfed

Offset Lithography

by

Alexander M. Voutsas Jr.

A thesis submitted in partial fulfillment of the

requirements for the degree of Master of Science in the

School of Printing Management and Sciences in the College

of Graphic Arts and Photography of the

Rochester Institute of Technology

May 1991

Thesis Advisor: Professor Clifton Frazier

School of Printing Management and SciencesRochester Institute of Technology

Rochester, New York

Certificate of Approval

Master's Thesis

This is to certify that the Master's Thesis of

Alexander M. Voutsas Jr.

With a major in Printing Technologyhas been approved by the Thesis Committee as satisfactoryfor the thesis requirements for the Master of Science degree

at the convocation of

May 1991

Thesis Committee:

Cliff FrazierThesis Advisor

Joseph L. NogaGraduate Program Coordinator

Miles SouthworthDirector or Designate

The Effects of Blanket Tension and Squeeze Pressure on Slur and

Doubling Using Conventional and Compressible Blankets in Sheetfed

Offset Lithography

I, Alexander M Voutsas Jr. hereby grant permission to the Wallace Memorial Library

of R.I.T. to reproduce my thesis in whole or in part. Any reproduction will not be for

commercial use or profit.

May 1991

I dedicate my thesis to Platon

in

Acknowledgments

I would like to thank the following people for their help, advice, and tolerance that

they showed to me from the inception to the completion of this thesis:

Wiiliam Moore, Joe Byers, Bob Andrew; Reeves Bros. Inc. Wiiliam Eisner, Clifton

Frasier, Charles Layne, Ching Chen, Dave Cohen, Sven Ahrenkilde, Franz Sigg,

Eric Sanderson, Barbara Giordano, Bob Hacker II, the R.I.T. Web Press crew,

Elfreda and John Mott, Steve Clark, Jim Manning, the R.I.T. Center for Quality and

Applied Statistics, Kelly LaForest, Katrin Eisman, Anastasia Constantacatos, John

Voutsas. I also thank the Vasil family for their support.

IV

TABLE OF CONTENTS

LIST OF FIGURES viii

LIST OF TABLES x

ABSTRACT xi

Chapters

1. INTRODUCTION 1

The Problem 1

Background 2

Conventional Blankets 4

Compressible Blankets 6

Compressible Layer 6

Surface Topography of Blankets 7

Blanket Performance Standards 8

Chapter 1 Endnotes 9

2. THEORETICAL BASES 12

Packing 12

Cylinder 12

Plate-to-Blanket Settings 1 2

True Rolling Theory 14

Equal Diameter Theory 17

Compressible Layer Construction 17

Microspheres 1 8

Blanket Tension 20

Response Variables 22

Dot Gain 22

Slur 23

Doubling 24

Murray-Davies Equation 24


3. REVIEW OF THE LITERATURE 29


TABLE OF CONTENTS (continued)

4. HYPOTHESES 32

5. METHODOLODGY 33

Blankets 35

Blanket Packing 35

Blanket Tensioning 37

Paper. 38

Ink 38

Plate 42

Targets 42

GATF Ladder Target 42

RIT Experimental Doubling Target 43

UGRAWedge Target 45

Analysis 46

Instumentation andEquipment 48

Chapters Endnotes 50

6. THE RESULTS 52

7. SUMMARY AND CONCLUSIONS 55

OtherObservations 57

8. RECOMENDATIONS FOR FURTHER STUDY. 58

9. BIBLIOGRAPHY. 60

APPENDIX A 65

Factor Breakdown and Slur Values of Obtained Data 66

Factor Breakdown and Total Dot Gain Values of Obtained Data 69

APPENDIX B 72

ANOVA for Obtained Slur Values 73

Fisher Test for Slur Values 75

ANOVA for Obtained Total Dot Gain Values 76

Fisher Test for Total Dot Gain Values 78

VI

TABLE OF CONTENTS (continued)APPENDIX C 79

Histograms of Slur Values 80

Histograms of Dot Gain Values 83

APPENDIX D 87

Table 7- Mean Values of Dot Gain 88

Table 8- Mean Values of Slur. 89

APPENDIX E 90

Press Sheet Sample

VII

LIST OF FIGURES

1. Cylinder arrangement in sheetfed offset lithography. 3

2. The 'ductorandknife'

roller 6

3. Compressible blanket construction 6

4. The parts of a blanket cylinder 13

5. Bearer contact arrangement 15

6. Bearer clearance arrangement 15

7. Relationship of two gears meshing 16

8. Blanket'bulge'

characteristics for conventional and

compressible blankets 16

9. Four methods of compressible layer construction 19

10. Packing versus squeeze pressure for conventional and


1 1 . Packing versus dot gain for conventional and


12. Dot gain characteristics 23

13. T-bar lockup system 39

14. Single reel rod lockup system 40

15. Double reel rod lockup system 41

16. GATF Ladder target 43

VIM

LIST OF FIGURES (continued)

17. R.I.T. Experimental Doubling Target 44

18. UGRAWedge target 46

IX

LIST OF TABLES

1 . Factor Breakdown and Slur Values of Obtained Data 66

2. Factor Breakdown and Total Dot Gain Values of Obtained Data 69

3. ANOVA for Obtained Slur Values 73

Fisher Test for Slur Values 75

4. ANOVA for Obtained Total Dot Gain Values 76

Fisher Test for Total Dot Gain Values 78

5. Histograms of Slur Values 80

6. Histograms of Dot Gain Values 84

7. Mean Values of Dot Gain 88

8. Mean Values of Slur. 89

Abstract

Dot gain is a printing characteristic which has an effect on printing quality, if it

is not controlled. It refers to the physical increase in size for a printed dot during the

transfer process between the film and the printed sheet. The purpose of this

experiment was to determine whether varying squeeze pressure and blanket

tension settings significantly affected directional dot gain (slur and doubling),

therefore printing quality, using conventional and compressible types of offset

lithographic blankets.

The experiment was performed on the second unit of a two-color Heidelberg

MOZP sheetfed offset lithographic press, bypassing the first unit. Directional dot

gain was evaluated by density readings taken from four RIT Experimental Doubling

Targets.

An analysis of variance (ANOVA) was performed to determine whether

statistically significant differences in dot gain values occured. It was found that

squeeze pressure was a significant factor contributing to the magnitude of slur for

both types of blankets, while blanket tension was not. The Fisher test determined

that each of the squeeze pressure settings (.006 in. and .008 in. for conventional

blankets, .006 in. and .012 in. for compressible blankets) contributed the most

towards the difference in dot gain values, while the type of blanket and the different

settings of squeeze pressure also contributed to dot gain.

Visual examination of the press sheets revealed that directional dot gain

occurred at the higher squeeze pressure settings for both conventional and

compressible blankets. When both squeeze pressure and blanket tension settings

increased, larger dot gain values were detected on conventional blankets. As

blanket tension settings increased, dot gain decreased in compressible blankets,

since the increased tension decreased the squeeze pressure, by thinning out the

blanket.This trend was also noticed on the conventional blankets at the low

squeeze pressure setting.

XII

CHAPTER 1

INTRODUCTION

The Problem

Dot gain is a printing characteristic which has an effect on printing quality if

its magnitude is not controlled. Dot gain refers to the physical increase in size for a

printed dot that occurs during the transfer process from the film to the printed sheet.

The main factors which are assumed to cause dot gain are squeeze pressure (the

pressure that exists between the plate and the blanket due to varying packing

levels of each), blanket tension (the tension applied on the blanket around its

cylinder), the type of blanket, type of ink, and the type of paper used.

The purpose of this experiment was to determine whether varying squeeze

pressure and blanket tension significantly affected directional dot gain, therefore

printing quality, using conventional and compressible offset lithographic blankets.

This experiment was performed on the second unit of a two color Heidelberg

MOZP sheetfed offset lithographic press, bypassing the first unit. Only one color ink

(black) was used in order to minimize the variables in this experiment.

The interest in this study arose from the fact that little research has been

published on offset lithographic blankets, and furthermore, the experiments that

have been done did not study the concept of tension and its effect on dot gain.

Background

Of all printing done in North America today, 47% is produced via the lithographic

process.1 The lithographic process exists in two forms, direct and offset. Direct

lithography, as its name implies, prints directly from the image carrier to the

substrate, like for instance with stone lithography. Offset lithography, on the other

hand, incorporates the use of an offset blanket which consists of two-to-four layers

or plies of cotton cloth laminated together coated with many thin layers of rubber.

The inked image is transferred to the substrate from the image carrier via the

blanket, hence the term offset lithography. Offset lithography produces a higher

volume of printing than direct lithography, because of its better quality and its

versatility in substrate use.

Before the advent of offset lithography, quality printing containing fine line

work and halftones could only be achieved through the use of premium printing

papers. All printing was done directly from the image carrier to the substrate,

hence, the need for papers with a very smooth and uniform surface. With the

introduction of the offset proocess, which incorporated a rubber blanket, quality

printing could be achieved on almost any substrate type (printable matter) because

of the pliable nature of the rubber, which allowed the blanket to mold itself onto the

paper. This allowed for longer run lengths, due to less wear on the plate by running

the blanket against the substrate (Figure. 1).

In addition to transferring ink from the image carrier to the substrate, the

offset blanket performs the function of dampening the press vibrations, which are

present in any mechanical process. Vibrations present during the transfer of ink

from plate to blanket, or blanket to substrate disrupt the uniform transfer of ink film

'

Inking (pan)

Figure 1. The cylinder arrangement in sheetfed offset lithography.

4

and cause print streaks, parallel to the lead edge (front) of thecylinder.2 Some

sources of vibration are the shock caused by the inking and dampening rollers

which are in contact with the rotating plate cylinder, the plate and blanket cylinders

rotating against one another, gears of the cylinders meshing incorrectly (the correct

fitting of the gear teeth), and ducting rollers in the inking and dampening systems

which are in contact with their respective fountains (the area from which rollers

receive ink and dampening solution). When properly set and maintained, these

sources of vibration are minimized. However, as presses age and wear out,

vibrations can become amplified and can have a negative effect on the final print.

The offset blanket can help in this respect by dampening these vibrations and

allowing the ink film to transfer evenly. There are two types of blankets available

which will perform this function, conventional and compressible.

Conventional Blankets

The conventional blanket consists of three or four layers or plies of fabric

which are laminated together with a specially formulated rubber adhesive

compound, and are commonly referred to as the carcass. The carcass provides for

a flexible, yet stable base, and enables the blanket to be tensioned around its

cylinder.3 "The main role of the fabric is to strengthen and stabilize the blanket so

that it can be mounted under tension without tearing orstretching."4 The fabric is

woven especially for this application and has a "greater concentration of the length

ways'warp'

threads than the widthways 'weft'. Since the warp yarn is spun from

selected long staple fibres, the cloth is very strong along its length."5 For this

5

reason, when blankets are cut from a roll, it is necessary to make sure that the

warp direction in the fabric is oriented in such a way that it will go around the

cylinder, when the blanket is mounted. Because of the 'overunder'

construction of

the warp and weft threads, and the degree of twist given to yarn during spinning,

woven fabrics are normally highly elastic. "Blanket fabric has this stretch removed

after weaving by a special finishing process. This ensures that the blanket will not

creep under the forces of printing or need constant retensioning."6

The face of the blanket consists of many layers of rubber, as many as fifty or

sixty individual coats, which may be as little as 0.008 mm in thickness.7 A

continuous web of carcass is coated with a rubber solution using a 'ductor knife

and roller", which controls the thickness of the coating, depending on the gap

between the knife and the roller (Figure. 2). The fabric is pulled through the

machine with controlled speed and tension under the knife, where a thin layer of

solution is applied. The blanket is heated and the solvent contained within the

rubber is removed, so that the next layer of rubber can be applied. When the

desired thickness is achieved, the rubber must be treated or'cured'

to form the

tough elastic face necessary for the lithographic process.8

The uniformity of the blanket thickness is important because it determines

the minimum amount of squeeze pressure needed to completely transfer the ink

film on the printed sheet.9

Due! knife

Duel roller

Drip tray

Figure 2. The 'ductor andknife'

roller

Cross section o/ blanket: (A) rubber print-bearing surlace; (B) compressible layer; (Qadhesive; (D) high-strength, low-stretch stabilising labnc.

Figure 3. Compressible blanket construction

Compressible Blankets

A compressible blanket shares the same construction characteristics as the

conventional blanket just described. It does, however, have an additional

compressible layer built into the carcass which has a compressible nature, hence

the term, 'compressibleblanket'

(Figure. 3). Compressibility is important, since it

allows the blanket to adjust to pressure irregularities, and also helps dampen

vibration during printing. The compressible layer also affects the way in which the

blanket deforms when under pressure, at the printing nip. Conventional blankets

flow laterally when pressure is applied, while the compressible compress vertically.

"When pressure is applied to a conventional blanket (with solid rubber

interplies), the material flows away from the pressure lines at the cylinder nips,

creating bulge formations immediately in front of them, which distorts the

blanket surfaces. The effect of this is that the blanket cylinder diameter changes,

and with it the print length. This is because solid rubber cannot be reduced in

volume and is subject to lateral flow."10

When the same pressure is applied to a blanket containing a compressible

layer, "it should compress uniformly, without distortion, with the layer containing

the millions of minute voids allowing uniform compression to take place. As

these voids beneath the area under pressure reduce in volume, they permit

vertical compression to take place at the cylindernips."11

Compressible Laver

The compressible layer can be manufactured in a variety of ways, each

imparting a unique property to the compressible characteristicsof the blanket. Most

8

commonly, "a cellular layer containing air and gas bubbles is incorporated to

increase compressibility above the conventional level."12This is known as the

microsphere process. There are four different methods in manufacturing the

compressible layer; 'microspheres', 'salt leach', 'blowingagent'

and 'impregnated

fabric'. The manufacturing process of these methods is confidential and is known

only to the corresponding manufacturers. The'microspheres'

method, can produce

a compressible layer with a 'closedcell'

structure and which is superior to the open

'saltleached'

or'blown'

structure (see chapter 2).13

Surface Topography of Blankets

The surface topography (smoothness) of the blanket is described as either

'cast'

or 'ground*. During the'curing'

or vulcanization (a process during which heat

uniformly applied, transforms the soft rubber of the blanket into a a tough, resilient,

solvent-resistant compound), a specific surface-structured paper is mounted on a

cylinder, and is brought in contact with the rubber. The pattern of this paper is

transferred onto the blanket face. This is known as a'cast'

surface. The second way

to control the surface topography of a blanket is to build the blanket to a greater

thickness than is desired, and'grind'

the blanket back down to that specified

thickness. The result is a blanket face which is generally smoother than a'cast'

blanket, and one which exhibits less caliper (thickness) variation.

The surface topography of a blanket will ultimately affect the 'printquality'

attainable with a particular blanket, and its sheet release characteristics (how fast a

sheet is released from the blanket). A rougher blanket affords better

releasecharacteristics at the expense of print quality, while a smoother surface

9

yields better dot-for-dot reproduction, yet tends to promote problems associated

with poor sheet release.14However, the influence of surface topography of the

blanket on sheet release is marginal compared to factors such as blanket

hardness, resistance to solvent penetration (used for blanket cleaning), and

construction of the carcass (discussed later), all of which have a greater bearing on

the release properties of the blanket.15

Blanket Performance Standards

The packing latitude for a compressible blanket is between 0.004 in. and 0.006 in.

above bearer, as compared to 0.002 in. and 0.004 in. above bearer for the

conventional blankets. The first compressible blanket was introduced in 1962 by

WR. Grace, and utilized a compressible paper layer to achieve compressibility.

Since then, many advances have been made in the compressible blanket

technology that revolve mainly around how the compressible layer is made, rather

than the changes in the actual design of the blanket. Some of the characteristics by

which a blanket should be evaluated are its hardness, resiliency, smoothness,

resistance to solvent penetration, quick paper release (how fast a blanket 'spitsout'

the printed sheet), resistance to stretching, and resistance toabrasions.16

Since its introduction, the compressible blanket has become in most market

segments, the blanket of choice among printers. Its ability to, withstand larger

packing latitudes, resist smashes (collapsing of the carcass, due to excessive

pressure), resist cut marks from stock edges, and minimize or even eliminate

mechanical shortcomings caused by the machinery (such as gear streaks, worn out

bearings, inadequate roller settings), makes it favorable.

10

Endnotes for Chapter 1

1 Michael H. Bruno, "Printing Process Trends", Critical Trends for the 1990's- A report to the Graphic Arts Industry. December 1989, p.55.

2GATF Staff, "Solving Sheeted OffsetProblems."

Graphic Arts Technical

Foundation. 1981, p.16.

3John Smith, Technical Aspects of OffsetBlankets."

Australasian Printer

Magazine. December 1985/January 1986, p. 10.

4E. R. Sutton, The Technology of LithoBlankets,"

The British Ink Maker.

February 1971, p.76.

5lbid

6lbid

7Anon, "BlanketCoverage,"

Printing World. May 20, 1987, p.46.

8Neville Chamberlain, "New Developments in OffsetBlankets,"

Australasian

Printer Magazine. December 1980, p.22.

9Geoffrey Leslie, "What aboutBlankets?,"

Australasian Printer Magazine.

November 1986, p11.

1 ^Neville Chamberlain, "New Developments in OffsetBlankets,"

Australasian Printer Magazine. December 1980, p.18.

11Anon, "Blankets and Rollers- Challenge to

Manufacturers."

Printing

Equipment and Materials. February 1979, p.28.

1 2E. R. Sutton, Technology of LithoBlankets,"

The British ink Maker.

February 1971, p.74.

1

1 3A. S. Porter, "Lithographic Presswork", 2d ed., Graphic Arts TechnicalFoundation. 1982, p.303.

14Bob Foley, "Blankets andRollers."

Printing World. September 22, 1982,

p.25.

1 5John Smith, Technical Aspects of OffsetBlankets,"


Magazine. December 1985/January 1986, p.9.

1 6A. S. Porter, "Lithographic Presswork", 2d ed., Graphic Arts Technical


12

CHAPTER 2

THEORETICAL BASIS

Packing

The way in which packing is distributed between the plate and blanket

cylinders depends on whether the 'true rolling'or the 'equaldiameter'

theory is

employed. These theories refer to which of the two cylinders (plate/blanket) will be

overpacked to provide the necessary squeeze pressure. That also affects the

image length.

Cylinder

The lithographic offset blanket is mounted on a cylinder. The cylinder and

the blanket together dampen press vibrations, and transfer ink from the plate to the

substrate. This cylinder is comprised of the following components: a) the cylinder

body, which is the main part, b) the bearers, which are located on the two outer

ends of the cylinder body, c) the undercut, which is a recess between the cylinder

body and the bearers. There is a gap on the cylinder body which separates the

lead (front) from the tail (back) edge of the blanket, and allows an area within which

the blanket

can be mounted (Figure.4). The bearers serve as a reference for the packing that is

required under the blanket, and which depends on the thickness (caliper) of the

blanket and the packing recommendations for each press.

Plate-to-blanket setting

There are two types of plate-to-blanket cylinder settings: a) bearer contact

13

Recess (undercut I

Figure 4. The parts of a blanket cylinder

14

(Figure.5), and b) bearer clearance ( Figure.6). Bearer contact presses are

preferred because the bearers: a) insure proper gear meshing, since the bearer

diameter is equal to the pitch circle (working) diameter of the gear (Figure.7). If the

gears are over or under meshed (how they fit with one another), then vibration,

backlash (slippage), and/or speed variation could result. This interrupts the

continuous and uniform flow of ink, and results in variations in ink film thickness,

which can create streaks on the printed sheet. These streaks are parallel to the

lead edge, and the space between them is equal to the distance between two gear

teeth: hence the term 'gear streaks'1, b) provide continuous rolling pressure

between the cylinders, particularly when the cylinder gaps face one another, c)

serve as a dependable reference for cylinder packing.2

TrueRolling'

theory

In the 'truerolling'

theory, the plate cylinder is packed to a larger diameter

than that of the blanket cylinder, which is usually packed even with the bearer. The

blanket and the plate surface speeds are, in theory, equal upon entry into the

blanket/plate nip.3 The reasoning for this is that the blanket's rubber surface, under

the pressure of the plate/blanket nip'bulges'

or flows laterally to either side of the

blanket (Figure. 8). This is due to the non-compressibility of the rubber surface of

the blanket. The blanket cylinder diameter at the entry into the nip (at the bulge) is

roughly equal to that of the packed plate cylinder diameter, therefore the cylinder

surface speeds are also equal and the image prints to length, assuming that the

sheet used is of normal thickness. Sheet thickness (caliper) and the diameter of the

impression cylinder are two factors that can affect the printlength.4

15

RECESS

XPLATE CYLINDER

BODY SETTING

BLANKET CYLINDER

-CONTACT

Figure 5. Bearer contact arrangement

RECESS

XPLATE CYLINDER

BODY SETTING

BLANKET CYLINDER

CLEARANCE

Figure 6. Bearer Clearance arrangement

16

Piicn c.rcit

Ajatnaum

Deatnaum

Figure 7. Relationship of two gears meshing

Lateral Mow resulting from pressure

(j) Compression ot conventional

blanket

Vertical compression

(6) Compression Olanket

Figure 8. Blanket'bulge'

characteristics in conventional and compressible blankets

17

'EqualDiameter'

theory

The other method, the 'equaldiameter'

theory, prints a slightly longer

impression. In fact, the plate is packed even with the bearer, and the blanket is

packed above the bearer, so that the necessary squeeze pressure is achieved

exclusively by the packing of the blanket. The two cylinders (plate and impression),

have equal diameters ar the center of the nip, and therefore have equal surface

speeds.5'Bulging'

occurs on the blanket surface and this directly affects the

diameter of the blanket cylinder. Since the diameter of the blanket cylinder is

greater than that of the plate, its surface speed becomes higher. This results in a

longer impression than the original. If the speeds are significantly different, due to

improper packing of the cylinders, and meshing of the cylinder gears is not exact,

slippage can occur. This will result in slur and gear streaks on the printed sheet.

Fortunately, blankets are resilient and elastic enough to allow for some variation in

surface speeds and packing differences between the cylinders.6 As a result,

regardless of whether the 'truerolling'

or 'equaldiameter'

theory is employed, the

quality of the resulting impression will appear equal. It is only the print length which

will be different, and yet it will not be discernible to the eye.

Compressible Layer Construction

As mentioned earlier, the two types of blankets, conventional and

compressible, differ only in the compressible layer. The carcass of both blankets

however, is made of similar fabric construction (usually cotton, because of its

resistance to stretch, its strength, and resilience). It is important and it is expected

that blankets have low stretch characteristics, so that desirable tensions could be

18

set and maintained. The compressible layer is composed of natural rubbers

(Figure 9). This layer is coated onto the fabric, and could then be expanded, by

heat, for example. The idea is to create voids containing air that will compress

under pressure. The design of the voids is either the'closed-cell'

or'open-cell'

type.7 There are four known ways to manufacture the compressible layer. In

addition to the way in which they are manufactured, they vary in the percentage of

void content. The oldest is the 'impregnated cellulosefiber'

process (70% void,

30% solid), which lacks flexibility, and special care has to be taken to avoid

wrinkles, and withstand solvent attacks. This was followed by the Porvair salt leach

process (70% void content). As the compressible material is applied onto the

carcass, heat and water force the salt out of the layer. This creates the

interconnecting, 'opencell'

construction previously mentioned. The blowing agent

process (10-20% void content), gives off gases at high temperatures, whilevoid-

size control has been proven to be problematic.8 The most recent development in

compressible layer technology is the microsphere.

Microspheres

Microspheres (40-50% void content), are simply small compressible 'beads'.

within which air has been captured. They are then mixed into the rubber solution

which is applied onto trhe carcass of the blanket. Microspheres are of the 'closed

cell'

construction. The 'closedcell'

voids are considered superior to the 'opencell'

type, because their recovery time from the compression of the blanket is quicker.9

This is due to the vertical displacement of the microsphere, within which air is

19

BLOWING LAYER

10-20 Void commit

--

_

*- --

-W- -mj~ -W~ t

SALT LEACH LAYER

70 Void content

v^

PAPER FIBRE MATRIXMICRO SPHERE VCHDS

40-50 Votd content

Figure 9. Four methods of compressible layer construction

20

contained, as opposed to the 'opencell'

structure. In the 'opencell'

design, air

circulates in and out of the interconnected cell structure in the compressible layer,

during each blanket revolution, and thus requires more time for its recovery. The

faster recovery of microspheres nevertheless, is achieved at the expense of

compressibility. As pressure on the blanket increases, air within the microsphere is

compressed, and thus absorbs this pressure, which is what the compressible layer

is trying to minimize. An advantage of using a compressible blanket is that it is

more resistant to smashing and that there is a reduced lateral flow, as opposed to

the conventional type (Figure. 10).

"Even though we've got more packing, all we're doing is closing up the

voids in the compressible layer and not having to distort the blanket

laterally in the printing nip (sic) we can have higher amounts of squeeze

with lower printing pressure and greater packing latitude with

compressible blankets than with conventional blankets."10

Packing latitudes vary according to the characteristics and thickness of the

compressible layer (Figure. 11). Excessive pressure due to severe overpacking

can, however, alter the desirable characteristics of the compressible blanket in a

negative way, by subtracting its compresibility, as well as increasing the

magnitude of dot gain on the printed sheet.

Blanket Tension

Tensioning of the blanket is the amount of torque applied on the blanket

around its cylinder. Optimum tensioning is usually achieved through trial and error

for each given press and blanket combination, since there is variation in blanket

lockup systems (mounting systems of plates and blankets), and tensioning devices.

Too little tensioning will result in blanket slack which could result in doubling. Too

much tension could result in a change in the natural characteristics of the blanket,

21

200

0 05 0 10 0 IS 0 20 0 25 0 30 1 35

Figure 10. Packing versus squeeze for conventional and compressible blankets

Figure 11. Packing versus dot gain for conventional and compressible blankets

22

especially the compressible ones, by limiting their compression and resiliency, as

well as stressing the carcass. The blanket also loses its natural crimp of the warp

yarns in the fabric plies which contributes to the overall resiliency of the carcass.11

High tension levels could also have an effect on ink transfer (the uniformity and

quantity of transfer of ink, from the plate to the blanket), smash resistance and

recovery (resistance to collapsing of the carcass under high pressures, and the

time that is needed to fully recover from such a collapse), and dot gain. The effects

of overtensioning are more obvious in compressible blankets, where the voids may

be compressed to the point where the compressible layer loses its unique

characteristics.12 It was expected that as blanket tension and squeeze preesure

settings increased, directional dot gain would increase as well, using both types of

blankets.

Response Variables

Dot Gain

Dot gain is the increase in size of the printed dot that occurs during the

transfer process from film to the final printed sheet. This increase in size is

expressed as a percentage of the original size. There is mechanical and optical dot

gain. Mechanical dot gain is the actual, physical increase in dot size, while optical

dot gain is a visual interpretation of the printed image which depends upon the

opacity or translucency of the paper surface, and the surface of the paper (ie.

glossy, matte surfaces). Optical dot gain is caused by the diffusion of light within the

paper. Only a fraction of the total light entering the paper is reflected back. The dot

appears larger because there is a shadow around that dot. Screen ruling (lines of

23

dots /inch) also affect the surface characteristics of the paper. A fine screen ruling

increases the apparent dot size more than a coarse screen ruling.13

The types of mechanical dot gain are: a) directional, which includes slur and

doubling and b) non-directional or fill-in; a uniform, circumferential increase of the

size of the dot, which is caused when the blanket'squeezes'

the ink onto the paper

(Figure. 12).

Slur

Slur is a directional increase in dot size, seen as an elongation or smearing

of the trailing edge of a halftone dot. It usually occurs in the direction of the cylinder

rotation.14 Some of the causes of slur are: differences in speed between the plate

and the blanket cylinders, excessive back cylinder pressure on coated papers,

excessive ink, low tack ink, excessive plate-to-blanket pressure, defective paper,

and insufficient blanket tension. Slur reduces shadow contrast, plate life,

resolution, and distorts the printeddot.15

A B

C..

D

A- no dot gain

B- Fill-in

C- Fill-in and Slur

D-Slur

Figure 12. Dot Gain Characteristics

24

Doubling

Doubling refers to the double image of a dot which is out of register with

itself. The direction of misregistration can occur in any direction. Doubling "varies

relative to the full strength true dot".16 Doubling is generally not considered a

blanket-related problem (except when the blanket is loose). On a multi color press,

the still wet ink transfers to the blanket of the next printing unit and transfers back to

the next sheet, but out of register. Some causes of doubling are, wavy or tight

edged paper, loose cylinders, varying positioning of plate against the blanket,

and.worn out bearings,17 Poor stability of the blanket face may also affect the

relationship between plate and blanket cylinders and cause doubling. It tends to

increase the highlight density, decreases print contrast, and decreases the

sharpness of the image. On a sheetfed press it can occur at the trail edge of a sheet

of paper due to a premature contact of the the paper with theblanket.18

The Murray-Davies Equation

The Murray-Davies Equation is used to represent the relationship of a dot

area to the density of a printed dot. The density of a halftone tint was based on the

inked dot area and the reflectivity of the ink itself. It is the reflectivity of the ink that is

considered to be the solid ink density.With the use of this equation it is possible to

determine the dot gain values. Dot gain is the difference between the value of the

original dot area and the value of the dot area on the printed sheet. Given that:

li = the intensity of the incident light and

It = the intensity of transmitted light, then

Transmittance =lt-5-li, and

Opacity =lh-lt= 1/T.

Density= log1 0 1/T

1

D= log

1-a(1-r)

where: D = Density of the tint

a = (1 -T)- % dot area of ink dot on paper

r = Reflecting power of the ink

The equation used to determine the densities of the tints is stated below:

Dt= log [ 1-a (1-10Ds)]

The equation used to solve for the dot area thus takes the form:19

1-10"Dt

a=

1-10-Ds

where: Dt= density of the tint

Ds= density of the solid

26

The above equation assumes that the reflectance of the paper is perfect. Therefore,

the densitometer should use the paper as the point of reference for calibration.

27


1 GATF Staff, "Solving Sheeted OffsetProblems,"

Graphic Arts Technical


2A. S. Porter, "LithographicPresswork," 2d

ed., Graphic Arts and Technical


3Gong Zongzhi and Wu Jinzhi, The Optimal Analysis of the Cylinder

Packing in the OffsetPress."

TAGA Proceedings.lQfifl p.240.

4 Ibid

5lbid

6John MacPhee and Larry E. Lester, "Some New Ideas on Pinpointing theCause of Horizontal Streaks", TAGA Proceedings. 1 986, p.84.

7Geoffrey Leslie, "Blankets and the Control of Dot Gain", Australasian Printer

Magazine.Vol. 38, No. 10, November/December 1987, p. 14.


Australasian

Printer Magazine. December 1980, p. 17.


Australasian

Printer Magazine. December 1980, p. 18.

10Anon, "Blankets and PrintQuality."

Graphic Arts Monthly Magazine.Vol.58.

No. 2, February 1986, p. 11 2.

11 Neville Chamberlain, "New Developments in OffsetBlankets,"

Australasian Printer Magazine. December 1980, p.22.

1 2R. C. Jenner, "Developments in the OffsetBlanket,"

The British Ink Maker.

August 1974, p.178

28

13 Miles Southworth, F. Quality Control Scanner. Vol.1 . No.2. 1981, p.4

14Milton Pearson, Irving Pobboravski and Chester Daniels, "Instrumentation

for the Measurement of Slur and Fill-in on a Lithographic WebPress,"

TAGA

Proceedings. 1979, p.163

15William H. Bureau, "What the Printer Should Know AboutPaper."

Graphic

Arts Technical Foundation. 1983, p.215.

16 Ibid p.190

1 7 Ibid

1 8 E.H. Treff,"An Engineer's View of Dot Gain on aWeb Offset Press",A PIA Annual Meeting, San Francisco, May 10, 1988 p.11

19 Miles Southworth, F. Quality Control Scanner. Vol.1 . No.2 1981, p.7

29

CHAPTER 3

REVIEW OF THE LITERATURE

Much of the blanket-related literature reviewed, has concentrated on the

physical characteristics of offset lithographic blankets, without emphasizing the

press set-up and press conditions. 'FOGRA', the German printing research

organization, provided some useful information on blankets.1 Its articles presented

information related to differences in blanket types, compression, and smash

characteristics. It was indicated that blanket smoothness and hardness played an

important role in sheet release and dot reproduction; hence the choice of a

'ground'

blanket for this experiment (see chapter 2).2Tritton (PIRA) noted that

blankets were the second

major cause for dot gain, following inks, and that overpacking amplified the small

differences that ordinarily existed between conventional and compressible

blankets.3 In the same report it was noted that lower tack inks are more susceptible

to showing changes in printing, due to squeeze pressure changes for example,

and ultimately in dot gain.4A study done by Pobboravsky, Pearson, and Daniels,

however indicated that changes in dot gain could occur without the solid ink

density beingaffected.5

In a study by DiPaoli a comparison was made between conventional and

compressible blankets in terms of solid ink density and dot gain. The results

indicated that conventional blankets showed higher dot gain values at the same

levels of solid ink density, than the compressible type (Figure.13).6

Different blanket tension settings and squeeze pressure levels have not

30

been tested simultaneously, in order to study the significance of these factors in

terms of dot gain (print quality). In a published report, Warren Rhodes reported on

the ink transfer and print quality. When density and sharpness were checked, a

relationship between density and slur could bee seen. It was noted that the amount

of ink which is transferred onto the paper is affected by the contact of the blanket

with the paper, as well as the pressure that exists between these two. These

pressures depend upon the blanket hardness, the packing, and the back cylinder

pressure. If a sheet has coarse surface characteristics, more squeeze pressure will

be needed for a more complete ink transfer, which will result in an increase in fill-in

and slur. So in order to isolate the effect of blanket tension and squeeze pressure

on print quality, it was noted that a smooth coated paper would aid the effort.7

A report, by William W. Walker, indicated that coated paper is more sensitive

to changes in pressure than uncoated paper, a quality that justified one reason why

coated paper was used for this experiment.8As mentioned in chapter 2, the

compressible blanket is able to withstand wider packing latitude, while maintaining

lower pressure levels at the nip, due to the nature of the compressible layer.

Information on the RIT Experimental Doubling target has been included in

the thesis by Amal A. Ba'adarani titled "A Study on Doubling in the Offset Printing

Process."9By using this target, dot gain can be detected in any occurring direction.

The Reeves Brothers Company, a blanket manufacturer, which supplied the

blankets for this experimentation, pointed out that randomizing tension levels on

the same blanket would be limiting for this experiment since the fabric material of

the blanket would not recover to its original length, and would thus jeopardize

tension testing at lower settings, using the same, already stretchedblanket.10

31


1 Dietmar Loibl, "Mechanisches und drucktechnisches Verhalten vonOffsetdrucktuchern."

FOGRA Mitteilunoen Nr. 93, Nov 1977.

2Warren, L. Rhodes, The Influence of Offset Blankets on Print Quality and

Ink Transfer International Bulletin June 1958, p.119.

3K.T. Tritton "Modifying the Dot Gain Characteristics of Sheeted OffsetPresses."

Pira Research Report. No. 163 c, 1981, p.29.

4lbid, p.24.

5Milton, Pearson; Irving,.Pobborovski,; and Chester, Daniels, "Application of

a dynamic Measurement Technique for the Investigation of Dot Gain in Web Offset

Lithography."

TAGA Proceedings. 1989, p.488.

8Alan, DePaoli,. The Effect of printing Conditions on Dot Gain - An Offset

NewspaperStudy."

TAGA Proceedings..1981. p.35

7Warren L. Rhodes, The Influences of Offset Blankets on Print Quality and

InkTransfer"

International Bulletin. June 1958 p.119.

8William C. Walker, The Printing Smoothness ofPaper."

Advances in

Printing Technology. Pergamon Press, Vol. 3., 1963, p.204.

9Amal A. Ba'adarani, "A Study on Doubling in the Offset PrintingProcess,"

Maser Thesis. Rochester Institute of Technology, 1989.

10Telephone conversation with Mr.William Moore, Reeves Brothers Co.,

Spartanburg S.C.

32

CHAPTER 4

HYPOTHESES

The two factors that were tested simultaneously at different levels on an offset

lithographic press were squeeze pressure and blanket tension. These factors were

tested on two types of offset lithographic blankets; conventional and compressible.

Both types were used in order to study their comparative reactions, and their effect

on the quality of the final print. There were four hypotheses to these factors, and

they are stated below:

1. There is no statistically significant difference in dot gain (doubling , slur) as

squeeze pressure changes (from 0.006 in. to 0.008 in.), when using a

conventional blanket.

2. There is no statistically significant difference in dot gain (doubling , slur) as

squeeze pressure changes (from 0.006 in. to 0.012 in.) when using a

compressible blanket.

3. There is no statistically significant difference in dot gain (doubling, slur) at

blanket tensions of (15, 30, 45 in/lbs), when using a conventional blanket.

4. There is no statistically significant difference in dot gain (doubling, slur) at

blanket tensions of (15, 30, 45 in/lbs) when using a compressible blanket.

33

CHAPTER 5

METHODOLOGY

The methodology used in this experiment was designed to tackle the following:

1. To examine how conventional and compressible blankets react to

different squeeze pressures and blanket tension settings.

2. To determine whether there is any statistically significant difference in

dot gain as squeeze pressure and blanket tension settings change

3. To identify the difference between conventional and compressible

blankets when faced with the same physical challenges such as

tension and packing changes.

4. To determine which differences in dot gain values were statistically

significant.

In order to examine the effect on directional dot gain (slur, doubling) from

tension and squeeze pressure on two different types of blankets, two packing

settings and three tension settings were tested during this experiment; a total of

twelve conditions. Each press run was 300 impressions long. This amounted to a

total of 3600 sheets. The press speed was set at 5500 impressions per hour.

Sampling for every press run was done by pulling one sheet for every 20 printed; a

34

total of fourteen samples. The Murray-Davies equation was used to calculate the

dot gain values.1 The statistical method which was used to determine the

acceptance or rejection of the hypotheses, was the analysis of variance (ANOVA).

A confidence level of 95% was used to determine whether there was any

statistically significant difference between the means. For the statistical differences

that were found to be significant, the Fisher test was used to determine where these

differences occurred. The experiment was performed on a two-color Heidelberg

MOZP sheetfed offset press, using the second unit and bypassing the first. The

targets that were used in this experiment were the RIT Experimental Doubling

Target, the Ladder target, and the UGRA Wedge.

The order of the twelve press runs was as follows:

1. CV1a 7. CM1a

2. CV2a 8. CM2a

3. CV3a 9. CM3a

4. CV1b 10. CM1b

5. CV2b 11. CM2b

6. CV3b 12. CM3b

where: CM= compressible

CV = conventional

1 , 2, 3 = blanket tension settings in ascending order (15,30,45 in/lbs)

a, b = hi/low packing levels, respectively

(0.006 in. and 0.008 in. for CV, 0.006 in. and 0.012 in. for CM)

35

Blankets

Both conventional and compressible blankets were utilized in this

experiment, in order to determine whether blanket tension and squeeze pressure

have similar effects on both types. Ground surface (vs. cast) blankets were used,

since their smooth surface was less likely to restrict the lateral flow of ink, caused

by increased pressure between the plate and blanket cylinder surfaces. The

blankets that were used for this experiment were donated by Reeves Brothers

Incorporated. The two types of blankets used were the G-44 (conventional) and the

2004 (compressible). Using a micrometer (L.S. Starrett Co. #256), these blankets

were tested for uniform thickness (caliper). When a blanket does not have a

uniform thickness (blanket + packing), pressure must be applied at the thinnest

area of the blanket, in order to print adequately, which would then result in

excessive pressure elsewhere. Blanket thickness variation could result in

scumming, poor ink transfer, fill-in, and platewear.2

Blanket Packing

The squeeze pressure settings were .006 in. and .008 in. for the

conventional blanket, and were accomplished as follows: for the lower squeeze

pressure setting, the blanket cylinder was packed even with the blanket bearer. The

plate cylinder was packed to .006 in. above the plate bearer for a squeeze

pressure of .006 in. An additional .002 in. was added to the blanket cylinder, for the

higher squeeze pressure setting of .008 in. For the compressible blankets, squeeze

pressure settings were .006 in. and .012 in., and were accomplished as follows: for

the lower squeeze pressure (.006 in.) the plate cylinder was packed to .006 in.

36

above the bearer, while the blanket cylinder was packed even with bearer. For the

higher squeeze pressure (.012 in.), .006 in. were added to the blanket cylinder. At

the end of the entire press run it was noted that the press specifications called for a

5:1 plate-to-blanket packing ratio. Due to this, three conditions were reprinted.

These conditions involved the conventional blanket, using the low squeeze

pressure setting, and all three of the blanket tension settings. The reason for

reprinting only these three conditions was because the higher level of squeeze

pressure (.008 in.) met the press specifications (.006 in. above the bearer of the

plate and .002 in. above the bearer of the blanket). No adjustments were made to

the back cylinders as the squeeze pressure increased. None of the conditions that

involved compressible blankets were reprinted, since their compressible layers are

made to withstand pressure variations, such as the ones experienced during this

experiment. The data which were drawn from the three reprinted conditions

replaced those of the original three. Thus, the data from the original conditions

were omitted in this statistical analysis. Correct packing was checked with a

packing gauge (Baldwin G81-2).3lt is an instrument whose one end rests on the

bearer of the blanket or the plate, and the other end (moving spring) rests on the

face of the blanket or the plate respectively. The difference in height between the

two, expressed in possitive or negative values indicates whether the packing level

is either above or below the bearer respectively. The compressible blankets used

were of the 'opencell'

construction design (salt leach). Once the blankets were

checked for squareness, they were mounted appropriately. Mounting procedures

were important, because tension should be allocated evenly around the blanket,

37

whose elasticity (blanket's) may prove limiting for this procedure (tensioning),

should these procedeures fail to be carried out correctly.

Blanket tensioning

The blanket tension settings were 15, 30, 45 in/lbs.4 These settings were

chosen so that they represented a significant change in printing response without

being unrealistic. They were set by using a torque wrench (Craftsman #44441).

After the blankets were run for three minutes and conformed better around the

blanket cylinder, they were retorqued to the desired setting.5The torque wrench

settings are usually affected by the press type, size, and lockup system used. When

blanket tension settings are too low, doubling (discussed in chapter 2) would

occur, while too much tension will eventually rupture the carcass, and would make

the blanket useless. The strength of a pressman, the type of blanket, and the

blanket cylinder gear ratios, affect the designated levels of tensioning (too much

tensioning may thin out the blanket and thus affect the packing set up, in terms of

squeeze pressure).

The bars that run across the tail and lead edges of the blanket, and are

clamped in the cylinder gaps, are called the clamp bars, and are part of the lockup

system (the mechanism used to securely fasten the blanket's edges within the

cylinder gap). Three examples are, the T-bar (Figure.13), the single-reel rod

(Figure.14), and the double-reel rod, used by the Heidelberg presses (Figure.15).

38

Paper

The paper which was used for this experiment was Piedmont Gloss, coated,

80lbs, C2S Text, 916960, with dimensions of 19x25 in. There were 3600 sheets

printed for this experiment - 300 sheets for each of the twelve press runs. The

coated paper was used because any variation on the press sheet would become

more apparent than using an uncoated sheet. Because of an existing curl in the

paper, there was a problem feeding the sheets through the press, and thus the

press would jam. These stoppages affected the balance of the water with the ink,

and time was needed to re-establish this balance. The curl did not cause

doubling.since the paper was printed on the second unit. It did, however, influence

the smooth flow of the press runs.

ink

The ink which was used was Marathon Hardy- Black, with a tack level of ten,

as indicated by the manufacturer. The decision to use a relatively low tack ink was

because the higher the ink tack, the'cleaner'

the ink would pull from the plate. A

lower tack ink is more susceptible to show less than optimal experimental

conditions than a higher tack ink. Pressure increases, along with the use of a

smooth blanket face, aids the ink to spread laterally, encouraging mechanical dot

gain or fill-in (see chapter 2).

The solid ink density level on the printed sheets was 1 .65 0.05 as they

came off the press (drying eventually decreased the density). The density levels

were checked with an X-Rite 408 densitometer. It was desirable to have an ink film

thick enough to make changes during the press runs more visible.

39

BLANKETCAPSCRE

packing

Figure 13. T-bar lockup system

40BLANKET

LOCKUP NUT

RATCHET PAWL-DISENGAGEDRATCHET PAWL-ENGAGED

PAWL STUD

TRAILING ECGE LEADING EDGE

3LANKET LOCKUP ROD \ / RETAINING LEDGE

VIEW BB

A <,

Jrn 1

0J1 '

OPERATING SIDE

T1 CYLINDER

LEAOING EDGE

OF BLANKET

TRAILING EDGE OF BLANKET

LEAOING ANO TRAILING

EDGES OF BLANKET

LOCKED UP ON CYLINDER

Figure 14. Single reel rod lockup system

Barring Holes

Figure 15. Double reel rod lockup system

42

pjaj

The plate which was used was a Viking G-1, an aqueous plate, made by 3M.

The term'aqueous'

refers to the use of a water-based solution for developing the

exposed plates (unexposed areas are soluble, while exposed are insoluble),

instead of using one which is solvent-based. It is a medium running plate with a life

of about 120,000 impressions before significant wear occurs. It was exposed

according to manufacturer's directions; a visual solid step at .45 density,

corresponding to a step 3 on the tone reproduction scale of the UGRA Wedge

(discussed later in this chapter).6 The plate was washed and gummed between

each treatment combination. The same plate was used for all runs so that any

potential variation from different plates would be eliminated (plate burning,

mounting, and variation in physical makeup).

Targets

In order to gather the necessary information from this experiment and

eventually evaluate the raw data , printing test targets such as the GATF Ladder

target, the RIT Experimental Doubling target, and the UGRAWedge, were used.

GATF ladder target

The ladder target is designed to show any printing variation which could

occur around the blanket cylinder, such as slur. It is comprised of three strips each

0.25 inches wide. The center strip is made of horizontal (parallel to the lead edge),

equally-spaced lines, while the two outer strips have vertical lines. The lines are

the equivalent of a 50% screen tint of 150 lines per inch. Any directional dot gain

43

will result in the darkening of the target area whose lines are perpendicular to the

direction of motion. Its limitation is its inability to show directional dot gain at a 45

degree angle, since the horizontal and vertical lines would darken equally, giving

the impression of overall dot gain or fill-in.7

Although a second ladder target mounted at 45 degrees would solve this

problem, it was impractical to do so due to space limitations. The use of the RIT

Experimental Doubling test target not only overcomes this limitation, but measures

the magnitude of doubling/slur as well as its direction. In this experiment the ladder

target was used to monitor any movement of the press that might have appeared

on the press sheet.

Figure 16. GATF Ladder Target.

RIT Experimpntal Doubling target

The RIT Experimental Doubling target is not for sale. It measures 1.25 inches

square. The center position of the target has a series of tightly drawn concentric

circles. Each of the corners of the target has a 5mm square patch: 1 ) a 3% dot

patch to observe the efffect of doubling on the small dot, 2) a solid patch to

determine ink film thickness, and thus enables it to calculate dot gain, 3) a 50% tint

to correlate between the average of light and dark area values of the concentric

44

circles and the 50% tint, and 4) a clear patch to calibrate the densitometer. The

circular patch has a directional scale of 0 to 180. This helps find the orientation of

the directional dot gain.8The advantage of this test target is a) the circular design,

which allows it to be more sensistive in all directions even at 45 degrees), b) its

large size, which makes it feasible to take numeric readings rather than quantitative

visual evaluation, by reading the lightest and darkest segments.in the lightest and

darkest segments, within the patch with a densitometer, and, c) ability to obtain

densitometric values of directional dot gain, as opposed to visual evaluation, while

small dots make it easier to differentiate between slur and doubling.9

RIT DOUBLING TARGET !

160 20 jiBLin J 140^^ ^^ 40 ,

Jkm wL\T

100 jH|

80 WU

6o^H ^LW :i

<*..40^^^H ^^^r 140

200

160 [Rocnesler Institute of Technology |

Figure 17. RIT Experimental Doubling Target 1989

In this experiment the RIT Experimental Doubling Target was utilized to

gather the necessary densitometricdata. Total dot gain was calculated by

averaging the dot gain valuesof the brightest and darkest segments. Directional

dot gain was calculated by the difference between the values of those two

45

segments. A solid ink bar positioned on the tail edge of the printed sheet measured

the solid ink density of the printed sheets. It measures three inches wide and as

long as the length of the printed sheet.

UGRA Wedge target

The UGRAWedge main function is to control the platemaking process, but is

also used to evaluate tone reproduction, slur and doubling (directional dot gain). It

measures 174mm long and 14mm wide.10 It is made up of five elements:

1. a continuous tone wedge, intended to show differences in exposure, 2. circular

to check resolution, in order to control exposure, 3. a 60 lines/cm halftone wedge

(screen angle-45 degrees), used for evaluating tone reproduction, 4. slur and

doubling patches (screen ruling of 48 lines/cm), with patches of 0, 45, and 90

degrees, which help aid the density measurements of slur and doubling, and a

small dot patch (screen ruling of 60 lines/cm), on which there are two sets of dot

patches: .5, 1, 2, 3, 4, 5, and 95, 96, 97, 98, 99, 99.5% dot patterns, which help to

identify how the small dots print from the plate to the paper, as well as to identify a

characteristiccurve.11

In this experiment, the UGRA Wedge target was used to verify proper plate making.

9 5% 99% 98% 97% 96% 95%

Figure 18. UGRAWedge 1982

46

Analysis

As previously mentioned, sampling for every press run was done by pulling

one sheet for every 20 printed, for a total of fourteen samples per press run. Once

the samples were gathered, densitometric readings were taken from each of the

twelve doubling test targets on each press sheet. The measuring instrumen was a

COSAR Autosmart system, a densitometer whose readings are immediately stored

in a computer (IBM compatible), and through which dot gain calculations were

performed, using the Murray-Davies equation. The data were used to calculate dot

gain were drawn from areas of the RIT Experimental Doubling Target. Since the

angle of directional dot gain happened to be always zero it was easy to take

automatic densitometric readings off the light and dark segments of the target.

There were five readings that were taken from each target; the 50% tint

square, the zero segment on the concentric circle area, the solid patch, the ninety

degree segment on the concentric circle area, and the solid patch once more. The

values that were calculated were the dot gain of the 50% tint, the dot gain of the

zero degree segment, and the dot gain of the ninety degree segment of the

concentric circle area of the target. Of the twelve targets on each press sheet, the

middle four were chosen to be analyzed. The decision for this selection was based

on the fact that twelve targets would have been an overkill, so the presence of the

remainder were there for reference. The four targets were picked on the basis that

variation from the lead edge (front) to the tail edge (back) of the press sheet would

reveal more interesting results than if the targets ran across the printed sheet

(mechanical variation across the printed sheet is almost always present to some

degree). The average total dot gain and the directional dot gain (slur) values of the

47

targets of all the press runs were then calculated. The total dot gain values were

calculated by averaging the dot gain values of the 0 and 90 degree areas of the

concentric circle area of the target. The slur values were the difference of these two

areas. Even though dot gain values could have been derived from both the circular

patch and the 50% patch, the first was chosen to be used for the dot gain

calculations, since densitometric readings were measured better. The resulting dot

gain values were higher than those which would have been calculated with the

50% patch. This is due to the difference in screen ruling. Higher screen ruling

results in higher dot gain values. From these mean values an analysis of variance

(ANOVA) was used to detrermine whether there was any statistically significant

difference between the mean values of the four targets. After finding that there was

no significant difference between the mean values of the four targets, the values of

one target were chosen to be used for the the final statistical analysis in this

experiment.

The data were stored in LOTUS 1-2-3 computer program. The statistical

analysis was performed using the MINITAB computer program. The analysis of

variance (ANOVA) was used to either accept or reject the fourhypotheses.12 The

value of the F ratio, a derivative of the analysis of variance, was used as the

criterion for either accepting or rejecting each of the nullhypotheses.13The level of

confidence using the ANOVA method was set at 95%. A value of F which was less

than 2.34 resulted in the acceptance of the null hypothesis (the number was found

after looking up an f-testtable).14

Any values greater than 2.34 resulted in the

rejection of the null hypothesis. The derivation of the F value is based on the

degrees of freedom, in terms of a) the number of samples per press run, b) the

48

combination of the degrees of freedom of the three dimensions that were tested

(blanket type, blanket tension, squeeze pressure), and c) the confidence level

(95% for this experiment). Since ANOVA only identifies whether or not there is any

statistically significant differences, and not, where these significant differences

occurred, the Fisher test was used, once there was an indication that significant

differences existed. The calculations were performed with the aid of the MINITAB

program using a VAX computer. Two separate sets of data were used, one for each

blanket, in order to test the four hypotheses.

Instrumentation and Equipment

The instuments and equipment which were used for this experiment are listed

below:

Densitometers:

- X-Rite 408 - for on press SID readings

-COSAR Autosmart System (LOTUS 1-2-3)

Packing Gauge- Baldwin G81-2

Torque Wrench- Craftsman # 44441

VAX computer (MINITAB)

Apple Macintosh computer (Mac Write)

Press: -Heidelberg MOZP 19x25.5 in., Two color, Type: BAE 2-0, Serial # 16689

Blankets: -Conventional- Reeves G-4

Compressible- Reeves 2004

Plate: 3M- Viking G-1 ,aqueous

Paper: Piedmont Gloss, 80 lbs, coated, Text (25x38 inches), # 916960

Ink: Marathon Hardy black, tack-10

49

RIT Experimental Doubling Target (not for sale)

Ladder Target

UGRAWedge Target

SPS 50x Magnifier

Hand held calculator

50


1Geoffrey Hyatt, "Factors Affecting the Strength and Openability of Tight-

Backed, Adhesive-Bound, Hard CoverVolumes,"

Master Thesis. Rochester Institute

of Technology, 1988, pp.7-12


Australasian

Printer Magazine. December 1980, p.22.

^Charles W., Latham,."Packing Gauge ImprovesAccuracy."

Lithographic

Technical Foundation Research Progress Report. No. 78. January 1959, pp. 1-4.

4Anon, "Precide Tensioning, Thickness Checks can Eliminate BlanketProblems."

American Printer Magazine. December 1987, p.65.

^Anon, "Precise Tightening, Thickness Checks can Eliminate BlanketProblems."

American Printer Magazine. December 1987, p.66.

6Anon.,UGRA Plate Control Wedge- Directions for use,1982, p.10

7Harry H. Hull, The GATF Ladder Target - A New Test Image,", Graphic Arts

Technical Foundation Research Progress Report No. 99. October 1973.

8Amal A. Ba'adarani, "A Study on Doubling on thge Offset PrintingProcess,"

Master Thesis. Rochester Institute of Technology, 1989, pp.23-25.

9lbid

10Anon.,UGRA Plate Control Wedge- Directions for use, 1982, p.10

11Ibid, p.10

1 2David R. Anderson, Dennis R. Sweeney ,Thomas A. Williams, "Statistics

Concepts and Applications. 2d ed., West Publishing Company, St. Paul,

51

1986, p.477-479

1 3lbid, p.563

1 4Roger C.Pfaffenberger, James H. Patterson, Statistical Methods for

Business and Economics. S^ed., Irwin Press Honewood, 111.1987, p. 11 65

52

CHAPTER 6

THE RESULTS

The data from this experiment have been gathered from measurements taken

from the four targets on each press sheet.This data were read and stored using the

COSAR Autosmart System(a combination of a densitometer connected to an IBM

compatible computer. An analysis of variance (ANOVA) was performed with the aid

of the MINITAB program.

Initially, an analysis of variance was performed in order to determine whether

any significant differences existed throughout the twelve runs, between the four

targets. It was determined that no significant difference existed in the mean values

of the four targets. Thus, the values of one target were selected to be used for the

statistical analysis of the experiment. The data of the one target were analyzed with

the use of the MINITAB program, using a VAX computer. The analysis of variance

(ANOVA) was used to determine which of the four hypotheses should be accepted

and which should be rejected, depending on the F value for each hypothesis

(Appendix B). The values of total dot gain and slur are included in the Appendix A.

Histograms are also displayed showing the distribution of the slur and total dot

gain values at every level of blanket tension, squeeze pressure, and the distinction

of the blanket type used (Appendix C)

53

First Hypothesis

Statistical analysis proved that there is a significant difference in total dot gain

and directional dot gain (in terms of slur), between the squeeze pressure settings of

0.006 in. and 0.008 in., when using conventional blankets (Tables 1, 2- Appendix

A). The F value was calculated to be 1120.06. Since this value was vastly greater

than 2.34, the null hypothesis was rejected. A visual examination indicated that

directional dot gain occurred at the higher squeeze pressure settings with an

orientation of 40, 50, and 60 degrees for blanket tension settings of 15, 30, 45 in/lbs

respectively.

Second Hvpothsis

The analysis of the data indicated that significant differences occurred, when

using two different squeeze pressure levels (.006 in. and 012 in.), on compressible

blankets. The F value was calculated at 1 26.44. This value is much greater than

2.34, and therefore the null hypothesis was rejected. The Fisher test indicated that

the significant differences of the dot gain values was primarily attributed to each of

the squeeze pressure settings, followed by the combination of the blanket type with

the higher squeeze pressure level (Table 2- Appendix-A). Directional dot gain

occurred when the squeeze pressure setting was .012 in, with an orientation of 40

degrees for all blanket tension settings of the compressible blankets.

Third Hypothesis

The statistical analysis for this hypothesis showed no significant differences in

dot gain values, when different blanket tension settings were used (15, 30, 45

54

in/lbs), on a conventional blanket. The F-value was calculated at 1.16. This value

was less than 2.34, and thus, the null hypothesis was not rejected. Directional dot

gain became more obvious as the setting of blanket tension increased, at a

squeeze pressure of .008 in.

Fourth Hypothesis

There was no significant differences in the dot gain values, when using

different blanket tension settings (15,30,45 in/lbs), on compressible blankets. The

F-value was calculated at 2.07. This was less than 2.34, and thus the null

hypothesis was not rejected. The results showed that the increase of blanket

tension increased the dot gain of the conventional blankets while it decreased for

the compressible blankets (Tables 7, 8- Appendix-D).

55

CHAPTER 7

SUMMARY AND CONCLUSIONS

In summary, four hypotheses were tested in this experiment. All four hypotheses

were stated in the form of a null hypothesis.

The first hypothesis stated that there is no statistically significant difference in

dot gain (slur, doubling) as squeeze pressure changes (.006 in. and .008 in.), when

using a conventional blanket. After analyzing the data, the F-value indicated that

squeeze pressure was a significant factor contributing to dot gain (slur). In fact, it

was the single most significant factor for these differences. Directional dot gain

occurred with an orientation of 40, 50, and 60 degrees on the RIT Experimental

Doubling target, when the squeeze pressure was set at .008 in.

The second hypothesis stated that there is no significant difference in dot

gain (slur, doubling) as squeeze pressure changes (.006 in. and .012 in.), when

using a compressible blanket. It was expected that the differences would be

smaller than those exhibited by the conventional blankets, which they were. The

results indicated that the differences were significant. The null hypothesis was thus

rejected. In the statistical analysis the values of total dot gain were also tested so

that they could be compared to the slur values (Appendix B).

When squeeze pressure was set at .006 in. for both conventional and

compressible blankets, the differences in total dot gain and slur values were small.

This was not true for the values generated with the higher squeeze pressure

settings (.008 in. and .012 in.). The values corresponding to each type of blanket

differred, and can be distinguished when shown on a histogram (Appendix C).

56

The third hypothesis stated that there is no statistically significant difference

in dot gain (slur, doubling), at blanket tension settings of 15, 30, 45 in/lbs, when

using a conventional blanket. The blanket tension, proved to contribute very little to

dot gain, compared to squeeze pressure. Directional dot gain increased as blanket

tension settings increased with an orientation of 40, 50, and 60 degrees on the RIT

Experimental Doubling target, at a squeeze pressure of .008 in.

The fourth hypothesis stated that there is no statistically significant difference

in dot gain (slur, doubling), at blanket tension settings of 15, 30, 45 in/lbs, when

using a compressible blanket. The results did show, however, that there was

statistically significant difference in the total dot gain values (F value= 20.94). A

histogram depicts a bimodal distribution of the samples, for values of slur and total

dot gain (Table 5, 6- Appendix C).

From the results it can be seen that as blanket tension and squeeze

pressure settings increased, so did the dot gain values for conventional blankets

(Tables 7, 8- Appendix-D).

With the use of compressible blankets, dot gain values increased when

squeeze pressure was changed from .006 in. to .012 in. However, at both squeeze

pressure setting, an increase in blanket tension resulted in a decrease of the dot

gain values. The average change in slur values that resulted from the difference in

the two squeeze pressure settings were 6.42 for the conventional blankets (.006 in.

and .008 in.), and 1.30 for the compressible blankets (.006 in. and .012 in.).

57

In conclusion, blanket tension did not prove to be a significant factor of dot

gain for the conditions set in this experiment. On the other hand, squeeze pressure

was clearly a very significant factor of dot gain. Compressible blankets were more

tolerant to increased blanket tension than conventional blankets. Conventional

blankets displayed a significant increase in dot gain, which can be attributed to the

lack of resiliency of these blankets at higher tension settings.

Directional dot gain occurred on both types of blankets, at the higher

squeeze pressure settings (.008 in. and .012 in.). The conventional type displayed

an increasing magnitude of doubling as blanket tension increased. The orientation

was 40, 50, and 60 degrees on the RIT Experimental Doubling target. In the

compressible blankets, doubling was consistant at every blanket tension settings,

and occurred at a 40 degree angle of the RIT Experimental Doubling target.

Other observations drawn from this experiment:

The dot gain values throughout the experiment were higher than those

recommended by Heidelberg ( approx. 20%). This increase can be attributed to the

RIT Experimental Doubling target. From studies done by Amal Ba'adarani, it was

shown that the concentric circle area generated higher values than the 50% tint on

the same target. The reason for this was not stated. It was suggested that the

screen ruling may have been the cause. For the purpose of this experiment, only

the concentric circle area of this target was used to generate and analyze the data,

and was sufficient for the tests in this experiment. The visual examination of the

small dots on the targets was done with the aid of a 50x magnifier, since it was not

easy to see the effects with anything less powerful.

58

CHAPTER 8

RECOMMENDATIONS FOR FURTHER STUDY

The following recommendations may prove to be useful in understanding

the behavior ofblanket tension on offset lithographic presses:

Press Size

The size of the press may be a contributing factor in the magnitude of the

resulting dot gain. By the same token, the size of the gears of the press cylinders

have an effect on the gear ratio, and therefore, the setting of ideal squeeze

pressure and blanket tension settings. The blanket of a small press cannot require

the same tension setting as that of a larger one.

The Physics of Tension

An interesting study can be done to determine how effective blanket tension

is around a blanket cylinder. A tension setting of 30 in/lbs may be effective in the

areas closest to the cylinder gap, but not in the area directly opposite this gap.

Identifying physical properties of blanket tension and its uniformity around a

blanket may lend more credence to dot gain patterns, and thus, print quality.

Standardization of Toroue Wrenches

A standardization and calibration technique for torque wrenches would help

establish a more scientific approach to blanket tension setting by pressmen.

Although this experiment did not show that tension was a significant factor in

directional dot gain, a larger press or in different circumstances it could, and in that

59

case it may be wise to be more careful with the use of a torque wrench, in order to

refine the method of blanket tensioning.

60

BIBLIOGRAPHY

61

BIBLIOGRARPHY

Anderson, David R.; Sweeney, Dennis, and Williams, Thomas A. Statistics

Concepts and Applications. 2d ed. St. Paul, West Publishing Company1986.

Anon. "Blankets andRollers."

Printing Eguipment and Materials. February 1979.

Anon. "Blankets and PrintQuality."

Graphic Arts Monthly. February 1986.

Anon. "BlanketCoverage."

Printing World. May 20, 1987.

Anon. "Precise Tensioning, Thickness Checks Can Eliminate BlanketProblems."

American Printer Magazine. December 1987.

Anon. The Influence of Inks and blankets on the Printing Characteristics of a

Sheetfed LithoPress."

Pack and Print. November 1982.

Arnamo, A. and Johansson, R. "Dot Density and Analysis in WebOffset."

Pira/IARIGAI International Conference on Applied Lithographic Technology.

Paper 17. London. October 7-9 1970.

Ba'adarani, Amal A. "A Study on Doubling in the Offset PrintingProcess."

Master

Thesis. Rochester Institute of Technology, 1989.

Bruno, Michael H. "Printing ProcessTrends."

Critical Trends for the 1990's - A

Report for the Graphic Arts Industry. December 1989.

Bureau, William H. "What the Printer Should Know AboutPaper."

Graphic

ArtsTechnical Foundation. 1983.

Chamberlain, Neville. "New Developments in OffsetBlankets."


Magazine. December 1980.

Clapper, F. R. and Yule J. A. C. The Effect of Multiple Internal Reflections on the

Densities of Halftone Prints onPaper."

Journal of the Optical Society of

America. No. 7. 1953.

DePaoli, Alan. The Effect of printing Conditions on Dot Gain - An Offset

NewspaperStudy."

TAGA Proceedinos.1981

62

Foley, Bob. "Blankets and Rollers - Challenge toManufacturers."

Printing World.

September 22, 1982.

GATF Staff. "Solving Sheeted OffsetProblems."

Graphic Arts Technical Foundation.

1981.

Hull, Harry H. "A Gripper-to-Tail SlurTarget."

GATF Research Report

PlQgi2S,1971.

Hull, Harry H. The GATF Ladder Target - A New TestImage."

Graphic Arts

Technical Foundation Research Progress Report. No 99. Oct. 1973.

Hyatt, Geoffrey. "Factors Affecting the Strength and Openability of Tight-Backed,

Adhesive-Bound, Hard CoverVolumes."

Master Thesis.Rochester Institute of

Technology. 1988.

Jenner, R. C. "Developments in the OffsetBlanket."

The British Ink Maker. August

1974.

Jorgensen, George W. The LTF Star Target for Ink Spread and ResolutionMeasurements."

Graphic Arts Foundation Research Progress Report. No. 52.

February 1961.

Latham, Charles. "Packing Gauge ImprovesAccuracy."

Lithographic Technical

Foundation Research Progress Report. No. 78. January 1959.

Laubman, Gerd. "Die Richtige desGummituches."

FOGRA Mitteilunoe.. Nr. 94.

December 1977.

Leslie, Geoffrey. "What AboutBlankets."

Australasian Printer magazine. November

1986.

Loibl, Dietmar. "Mechanisches und Drucktechnisches Verhalten VonOffesetrirucktuchern."

FOGRA Mitteilunoen. Nr. 93. November 1977.

MacPhee, John. "Some Ideas on Pinpointing the Cause of HorizontalStreaks."

TAGA Proceedings. 1986.

Mason, Robert L; Gunst, Richard; and Hess, James L Statistical Design and

Analysis of Experiments. John Wiley & Sons. 1989.

63

Milliken, George and Johnson, Dallas. "Analysis of MessyData."

Lifetime Learning

Publication- vol. 1 : Designed Fyperiments 1 984.

Pearson, Milton.; Pobboravski, Irving.; and Daniels, Chester. "Application of a

dynamic Measurement Technique for the Investigation of Dot Gain in WebOffset

Lithography."

TAGA Proceedings.. 1989.

Pearson, Milton.; Pobboravski, Irving.; and Daniels, Chester. "Instrumentation forThe Measurement of Slur and Fill-in on a Lithographic Web

Press."

TAGA

Proceedings. 1979.

Porter, A. S. "LithographicPresswork."

Graphic Arts Technical Foundation. 1980.

Rhodes, Warren. "Influence of Offset Blankets on Print Quality and InkTransfer."

International Bulletin.June 1958.

Sigg, Franz. "A Few Things About Microlines Most People Do NotKnow."

TAGA

Proceedings. 1988.

Sigg, Franz. "A New Densitometric Quality Control System for OffsetPrinting."

TAGA Proceedings. 1 970.

Smith, John. "Technical Aspects of OffsetBlankets."

Australasian Printer Magazine.

December 1985/January 1986.

Smith, J. M. "The Influence of Physical Properties on BlanketPerformance."

Pira

Seminar S14 - Offset Lithographic Blankets - a Technical Update.

Southworth, Miles, F. Quality Control Scanner. Vol. 1, No.2. 1981.

Sutton, E. R. Technology of LithoBlankets."

The British Ink Maker. February 1971.

Tollenaar, D. "Remarks on Ink Transfer". IARIGAI12th International Conference

Versailles, France, June 1973 1974

Tritton, K. T. "Modifying the Dot Gain Characteristics of Sheetfed OffsetPresses."

Pira Research Report PR4. November 1981.

Tyma, Louis, S. et al. "Bearers- A Necessary

Evil?"


Voglesong, William. "Standard Performance Material forDensitometry."

TAGA

Proceedings. 1989.

64

Walker, William, and Carmack, Robert F. The Printing Smoothness ofPaper."

Advances in Printing Technology. Pergamon Press, Vol. 3. 1963.

Warner, Richard D. The GATF Dot Gain Scale-II"GATF Annual Research Report.

1983/1984.

Williams, C. H. "Lithographic ProcessInks."

Inklings. No.120, 1982.

Witte, Dan. "Blankets Remain Stable in the Face ofChange."

Hioh Volume Printing.

June 1981.

Ynostroza, Roger. "A New Look at an OldFriend."

Graphic Arts Monthly. October

1984.

Zonghi, Gong, and Jinzhi Wu. The Optical Analysis of the Cylinder Packing in the

OffsetPress."


65

APPENDIX A

66

Table 1. Factor Breakdown and Slur Values.

C11 Ci12 C13 C14 C15 C16 C17 C18

1 1 1 0.49 2 3 0.26

2 1 2 6.88 2 4 2.33

3 1 1 0.55 2 3 0.95

4 1 2 7.06 2 4 2.05

5 1 1 0.16 2 3 0.00

6 1 2 5.89 2 4 2.30

7 1 1 0.58 2 3 0.53

8 1 2 7.21 2 4 2.73

9 1 1 0.32 2 3 0.69

10 1 2 7.74 2 4 2.42

11 1 1 0.47 2 3 0.49

12 1 2 7.02 2 4 2.42

13 1 1 1.25 2 3 0.65

14 1 2 7.47 2 4 1.53

15 1 1 1.28 2 3 0.65

16 1 2 7.28 2 4 0.26

17 1 1 0.45 2 3 1.22

18 1 2 8.17 2 4 1.69

19 1 1 0.85 2 3 0.17

20 1 2 7.91 2 4 2.54

21 1 1 0.29 2 3 0.09 Conventional Blankets:

22 1 2 8.10 2 4 1.85

23 1 1 0.26 2 3 0.63 C11: Blanket tension

24 1 2 7.48 2 4 2.61 C12: Conventional

25 1 1 0.61 2 3 0.23 C13: Squeeze pressure

26 1 2 7.23 2 4 2.42 C14: Slur values

27 1 1 0.36 2 3 0.43

28 I 2 7.79 2 4 3.11

29 2 1 1 0.23 2 2 3 0.40

30 2 I 2 5.08 2 2 4 0.36

31 2 I 1 0.85 2 2 3 0.63

32 2 I 2 6.05 2 2 4 1.64

33 2 I 1 0.16 2 2 3 0.81 Compressible Blankets:

34 2 I 2 4.52 2 2 4 1.18

35 2 I 1 0.29 2 2 3 0.70 C15: Blanket tension

36 2 I 2 5.94 2 2 4 1.26 C16: Compressible

37 2 I 1 0.91 2 2 3 0.44 C17: Squeeze pressure

38 2 I 2 5.59 2 2 4 1.39 C18: Slur values

39 2 I 1 0.72 2 2 3 0.33

40 2 I 2 4.54 2 2 4 1.05

67

Table 1. Factor Breakdown and Slur Va ues. [continued)C11 C12 C13 C14 C15 C16 C17 C18

41 2 1 1 0.33 2 2 3 0.57

42 2 1 2 5.00 2 2 4 0.75

43 2 1 1 0.81 2 2 3 0.37

44 2 1 2 5.53 2 2 4 2.18

45 2 1 0.49 2 2 3 0.70

46 2 2 5.34 2 2 4 0.89

47 2 I 1 0.03 2 2 3 0.16

48 2 I 2 5.53 2 2 4 1.40

49 2 I 1 0.32 2 2 3 0.13

50 2 I 2 6.32 2 2 4 1.68

51 2 I 1 0.07 2 2 3 0.13

52 2 I 2 5.73 2 2 4 1.65

53 2 I 1 0.58 2 2 3 0.67

54 2 1 2 6.40 2 2 4 2.24

55 2 1 1 0.00 2 2 3 0.44

56 2 1 2 5.13 2 2 4 1.69

57 3 1 1 0.07 3 2 3 0.66

58 3 1 2 8.55 3 2 4 2.12

59 3 1 1 0.13 3 2 3 0.09

60 3 1 2 8.29 3 2 4 0.98

61 3 1 1 2.14 3 2 3 0.83

62 3 1 2 8.07 3 2 4 1.90

63 3 1 1 0.44 3 2 3 0.54 Conventional Blankets:

64 3 1 2 7.49 3 2 4 1.94

65 3 1 1 0.55 3 2 3 0.57 C11: Blanket tension

66 3 1 2 7.80 3 2 4 1.45 C12: Conventional

67 3 1 1 0.78 3 2 3 0.31 C13: Squeeze pressure

68 3 1 2 7.65 3 2 4 1.68 C14: Slur values

69 3 1 1 0.10 3 2 3 0.57

70 3 1 2 7.73 3 2 4 1.38

71 3 1 1 0.54 3 2 3 0.57

72 3 1 2 7.81 3 2 4 0.78

73 3 1 1 0.88 3 2 3 0.23 Compressible Blankets:

74 3 1 2 7.46 3 2 4 1.12


76 3 1 2 7.35 3 2 4 1.22 C16: Compressible


78 3 1 2 7.87 3 2 4 1.55 C18: Slur values

79 3 1 1 0.81 3 2 3 0.78

80 3 1 2 7.66 3 2 4 1.94

68

Table 1. Factor Breakdown and Slur Values, (continued)C11 C12 C13 C14 C15 C16 C17 C18

81 3 1 1 0.00 3 2 3 0.14

82 3 1 2 8.12 3 2 4 2.16

83 3 1 1 0.84 3 2 3 0.14

84 3 1 2 8.51 3 2 4 1.73

69

Table 2. Factor Breakdown and Total Dot Gain'

Values.

C11 C12 C13 C14 C15 C16 C17 C18

1 1 1 30.42 2 3 34.58

2 1 2 39.07 2 4 36.50

3 1 1 30.75 2 3 35.28

4 1 2 40.00 2 4 35.87

5 1 1 30.71 2 3 34.54

6 1 2 39.08 2 4 35.97

7 1 1 31.15 2 3 34.81

8 1 2 39.52 2 4 35.93

9 1 1 31.47 2 3 34.56

10 1 2 39.56 2 4 36.06

11 1 1 31.87 2 3 35.02

12 1 2 39.80 2 4 35.63

13 1 1 30.05 2 3 34.60

14 1 2 39.53 2 4 36.04

15 1 1 31.29 2 3 33.96

16 1 2 39.59 2 4 36.13

17 1 1 31.10 2 3 35.04

18 1 2 39.08 2 4 36.40

19 1 1 31.24 2 3 34.50

20 1 2 39.67 2 4 36.36

21 1 1 30.88 2 3 34.27 Conventional Blanket

22 1 2 40.00 2 4 36.18

23 1 1 31.36 2 3 34.28 C11: Blanket tension

24 1 2 39.76 2 4 35.85 C12: Conventional

25 1 1 31.33 2 3 34.58 C13: Squeeze level

26 1 2 39.14 2 4 36.31 C14: Dot gain values

27 1 1 31.07 2 3 34.90

28 1 2 39.66 2 4 35.54

29 2 1 1 31.52 2 2 3 31.31

30 2 1 2 42.31 2 2 4 34.87

31 2 1 1 30.78 2 2 3 32.45

32 2 1 2 40.65 2 2 4 34.87

33 2 1 1 31.59 2 2 3 31.67 Compressible Blankets:

34 2 1 2 40.39 2 2 4 34.90


36 2 1 2 40.92 2 2 4 34.80 C16: Compresssible


38 2 1 2 40.92 2 2 4 34.50 C18: Dot gain values

39 2 1 1 31.53 2 2 3 31.46

40 2 1 2 42.37 2 2 4 34.93

70

abUi 2. Factor Breakdown and Total Dot Gain

C11 C12 C13 C14 C15 C16 C17'

C18

41 2 1 1 31.03 2 2 3 31.19

42 2 1 2 41.19 2 2 4 35.03

43 2 1 1 31.92 2 2 3 31.18

44 2 1 2 41.71 2 2 4 34.95

45 2 1 1 30.98 2 2 3 31.71

46 2 1 2 41.86 2 2 4 34.73

47 2 1 1 31.50 2 2 3 31.46

48 2 1 2 41.81 2 2 4 34.75

49 2 1 1 32.89 2 2 3 31.54

50 2 1 2 41.86 2 2 4 34.61

51 2 1 1 31.83 2 2 3 32.33

52 2 1 2 42.15 2 2 4 34.67

53 2 1 1 31.49 2 2 3 31.87

54 2 1 2 42.14 2 2 4 34.55

55 2 1 1 31.95 2 2 3 31.73

56 2 1 2 41.59 2 2 4 34.36

57 3 1 1 32.44 3 2 3 30.71

58 3 1 2 40.85 3 2 4 35.09

59 3 1 1 31.55 3 2 3 31.04

60 3 1 2 40.85 3 2 4 35.07

61 3 1 1 31.56 3 2 3 30.82

62 3 1 2 41.03 3 2 4 34.98

63 3 1 1 30.93 3 2 3 30.31

64 3 1 2 41.44 3 2 4 34.80

65 3 1 1 30.72 3 2 3 31.02

66 3 I 2 41.07 3 2 4 35.18

67 3 I 1 30.22 3 2 3 31.35

68 3 I 2 41.24 3 2 4 34.10

69 3 I 1 30.78 3 2 3 30.86

70 3 I 2 41.29 3 2 4 34.25

71 3 I 1 30.98 3 2 3 30.31

72 3 2 41.08 3 2 4 34.83

73 3 1 30.63 3 2 3 30.79

74 3 2 41.14 3 2 4 34.91

75 3 1 1 30.33 3 2 3 30.76

76 3 1 2 40.88 3 2 4 34.40

77 3 1 1 30.75 3 2 3 31.36

78 3 1 2 41.19 3 2 4 34.62

79 3 1 1 30.65 3 2 3 30.98

80 3 1 2 41.23 3 2 4 34.83

Values, (continued)

Conventional Blankets:

C11: Blanket tension

C12: Conventional

C13: Squeeze pressure

C14: Dot gain values

Compressible Blankets:

C15: Blanket tension

C16: Compressible

C17: Squeeze pressure

C18: Dot gain values

. 71

Table 2. Factor Breakdown and Total Dot Gain Values. (continued)C11 C12 C13 C14 C15 C16 C17 C18

81 3 1 1 30.01 3 2 3 30.07

82 3 1 2 41.33 3 2 4 35.41

83 3 1 1 30.70 3 2 3 30.5684 3 1 2 41.12 3 2 4 34.89

72

APPENDIX B

73

Table 3. Analysis of Variance for Slur Values

Analysis of Variance of Tension on Conventional Blankets

SOURCE DF SS MS F ptension 2 25.6 12.8 1.16 0.319

ERROR 81 896.5 11.1

TOTAL 83 922.2

LEVEL N MEAN STDEV + + + +

1 28 3.970 3.498 (*

)2 28 2.946 2.620 (

*

)3 28 4.224 3.755 (

*

)+-

+ + +

POOLED STDEV = 3.327 2.4 3.6 4.8 6.0 (slur)

Analysis of Variance of Tension on Compressible Blankets

SOURCE DF SS MS F P

tensions 2 2.531 1.266 2.07 0.132

ERROR 81 49.419 0.610

TOTAL 83 51.951

LEVEL N MEAN STDEV + + +

1 28 1.3304 0.9997 (*

)2 28 0.9229 0.6126 (

*

)3 28 1.0214 0.6751 (

*

)+ + +

POOLED STDEV = 0.7811 0.90 1.20 1.50 (slur)

Analysis of Variance of Squeeze Pressure on Conventional Blankets

SOURCE DF SS MS F P

squeeze 1 859.264 859.264 1120.06 0.000

ERROR 82 62.907 0.767

TOTAL 83 922.171

74


1 42 0.5150 0.4162 (-*)2 42 6.9117 1.1667 (-*)

+ + +

POOLED STDEV = 0.8759 2.0 4.0 6.0 (slur)

Analysis of Variance of Squeeze Pressure on Compressible Blankets

SOURCE DF SS MS F psqueezes 1 31.513 31.513 126.44 0.000

ERROR 82 20.438 0.249

TOTAL 83 51 .951

LEVEL N MEAN STDEV - + +-+

+--

3 42 0.4790 0.2995 (--*--)4 42 1.7040 0.6394 (--*--)

+ + ++..

POOLED STDEV = 0.4992 0.50 1.00 1.50 2.00 (slur)

75

FISHER TESTS ON SLUR.

Fisher Test of Squeeze Pressure on Conventional Blankets

Nominal level = 0.0500

Family error rate = 0.0500Individual error rate = 0.0500

Critical value = 1 .989

Intervals for (mean of column group)- (mean of row group)

1

2 -6.7768

-6.0165

Fisher Test of Squeeze Pressure on Compressible Blankets


Family error rate = 0.0500

Individual error rate = 0.0500



3

4 -1.4417

-1.008

76

Table 4. Analysis of Variance for Total Dot Gain Values.

Analysis of Variance on Tension-Conventional Blankets

SOURCE DF SS MS F ptension 2 21.5 0.7 0.45 0.638

ERROR 81 1921.2 23.7

TOTAL 83 1942.7

LEVEL N MEAN STDEV --

+ + +

1 28 35.146 4.249 (*

)2 28 36.350 5.066 (

*

)3 28 36.000 5.238 (

*

)+ + +

STDEV = 4.870 34.5 36.0 37.5 (total dot gain)

Analysis of Variance on Tension-Compressible Blankets

SOURCE DF SS MS F P

tensions 2 104.78 52.39 20.94 0.000

ERROR 81 202.66 2.50

TOTAL 83 307.43

LEVEL N MEAN STDEV --+ ++- +----

1 28 35.346 0.786 (.-*.-)2 28 33.213 1.595 (.*-)

3 28 32.796 2.084 (-*-)+ + +

+.

POOLED STDEV = 1.582 32.4 33.6 34.8 36.0 (total dot gain)

77

Analysis of Variance on Squeeze Pressure-Conventional Blankets

SOURCE DF SS MS F psqueeze 1 1836.62 1836.62 1420.04 0.000

ERROR 82 1942.68


1 42 31.156 0.599 (*)2 42 40.508 1.493 (*)

+ + +

POOLED STDEV = 1.137 33.0 36.0 39.0 (total dot gain)

Analysis of Variance Squeeze Pressure-Compressible Blankets

SOURCE DF SS MS F P

squeezes 1 169.60 169.60 100.90 0.000

ERROR 82 137.83 1.68

TOTAL 83 307.48

LEVEL N MEAN STDEV -+ + ++-

3 42 32.364 1.707 (*)4 42 35.206 0.670 (---*)

-+ + + +

POOLED STDEV = 1.296 32.0 33.0 34.0 35.0 (total dot gain)

78

FISHER TESTS ON DOT GAIN.

Fisher Test for Squeeze Pressure on Conventional Blankets


Family error rate = 0.0500Individual error rate = 0.0500



1

2 -9.846

-8.858

Fisher Test for Squeeze Pressure on Compressible Blankets


Family error rate = 0.0500

Individual error rate = 0.0500



3

4 -3.405

-2.279

79

APPENDIX C

Table 5. HISTOGRAMS OF SLUR VALUES.

Histogram of Slur Values for Conventional Blankets (CV)

Midpoint Frequency(A% slur)

0 23

1 18

2 1

3 0

4 0

5 6

**********************

*****************

6 9

7 11

8 14

9 2

******

*********

***********

**************

**

Histogram of Slur for Compressibe Blankets (CM)

FrequencyMidpoint

(A% slur)

0

Fr(

23

1 34

2 23

3 4

4 0

5 0

6 0

7 0

8 0

9 0

**********************

********************************

**********************

****

81

Histogram of Slur for Tensions 15 in/lbs (for CV and CM combined)Midpoint Frequency(A% slur)

0 16****************

1 13*************

2 9*********

3 4****

4 0

5 0

6 1*

7 8********

8 5*****

******************

******************

******

Histogram of Slur for Tensions 30 in/lbs (for CV and CM combined)


0 18

1 18

2 6

3 0

4 0

5 6

6 8

7 0

8 0

9 0

Histogram of Slur for Tensions 45 in/lbs (for CV and CM combined)


0 12

******

********

1 21

2 9

3 0

4 0

5 0

6 0

7 3

8 9

9 2

************

*********************

*********

***

*********

*

82

Histogram of Slur for Squeeze Pressure (CV=.006 in. CM=.006 in.)

FrequencyMidpoint Frc-

(A% slur)

0 44

1 39

2 1

3 0

4 0

5 0

6 0

7 0

8 0

9 0

*********************************************

***************************************

Histogram of Slur for Squeeze Pressure (CV=.008 in. CM=.012 in.)


0 2

1 13

2 23

3 4

4 0

5 6

6 9

7 11

8 14

9 2

**

**************

**********************

****

******

*********

***********

**************

**

Table 6. HISTOGRAMS OF TOTAL DOT GAIN VALUES.

Histogram of Dot Gain for Conventional Blankets (CV)

Midpoint Frequency(A% d. gain)

30 5*****

31 25*************************

32 11***********

33 1*

34 0

35 1*

36 1*

37 0

39 4****

40 10**********

41 17*****************

42 9*********

Histogram of Dot Gain for Compressible Blankets (CM)

Midpoint Freqilency

(A% d. gain)

30 1*

30.5 4****

31 9*********

31.5 10**********

32 2**

32.5 2**

33 0

33.5 0

34 2**

34.5 17*****************

35 21*********************

35.5 4*****

36 8********

36.5 4****

84

Histogram of Dot Gain for Tensions 15 in/lbs (CV and CM combined)


30 2**

31 11*

32 1*

33 0

34 3***

35 11***********

36 14**************

37 1*

38 0

39 4****

40 9*********

41 0

42 0


Midpoiint Frequency(A% d. gain)

30 0

31 11***********

32 16****************

33 1*

34 1*

35 1**************

36 0

37 0

38 0

39 0

40 1*

41 3***

42 9*********

85



30 6******

31 19*******************

32 3***

33 0

34 3***

35 11***********

36 0

37 0

38 0

39 0

40 0

41 14**************

42

86

Histogram of Dot Gain for Squeeze Pressure (CV and CM= .006 in)

Midpoint

(A% d. gain)30

Frequency

p********

31 41*****************************************

32 20********************

33 1*

34 3***

35 11***********

36 0

37 0

38 0

39 0

40 0

41 0

42 0

Histogram of Dot Gain for Squeeze Pressure (CVs.008 in. CM=.012 in)

Midpoint

(A% d. gain)

30

Frequency

0

31 0

32 0

33

34

35

0

4

25

****

*************************

36 14**************

37 1*

38

39

40

41

0

4

10

17

****

**********

*****************

42 9*********

87

APPENDIX D

TABLE 7. MEAN VALUES OF DOT GAIN

88

Squeeze_Pressure"

o

CO

LU

ir>

.O

oCO

V)

in

CONVENTIONAL

BLANKET

COMPRESSIBLE

BLANKET

.006 IN.008.IN

.006 IN.012 IN

in

in

31.05 39.24 34.64 36.06

35.15 35.35

CO

S

31.55 41.16 31.69 34.75

36.35 33.22

o

: %::::.

o

30.88 41.12 30.78 34.81

36.00 32.80

31.16 40.51 32.37 35.21

^^^?^^|^| 33,79

TABLE 8. MEAN VALUES OF SLUR

89

Squeeze_Pressure"

o

c/)

zLUr-

.Q

CO

oCO

to

X)

in

CONVENTIONAL

BLANKET

COMPRESSIBLE

BLANKET

.006 IN .008.IN.006 IN

.012 IN

is

o

.58 7.37 .50 2.33

3.97 1.42-: ':;-:

-..

<0

0

;i;1r~;?::-:r

.41 5.48 .46 1.38

2.80.92

to

| .49 j 7.88 .47 1.64

4. 39 1.06

.49 6.91 .48 1.78

3.72 113

90

APPENDIX E

Documents

The Effects of blanket tension and squeeze pressure on