published by smithers rapra technology ltd, 2012 rubber...

Shawbury, Shrewsbury, Shropshire, SY4 4NR, UK Telephone: +44 (0)1939 250383 Fax: +44 (0)1939 251118 Web: www.polymer-books.com

Practical Guide to H

ydrogenated Nitrile B

utadiene Rubber TechnologyRobert Keller

High performance engineering plastics are used in an increasingly wide range of applications and environments. Their growth in importance is a response to the ever-increasing demand for more reliable, high performance components.

This book is the product of the author’s first-hand experience and understanding of high performance engineering plastics; specifically hydrogenated nitrile rubbers, which are progressively supplanting the simpler non-hydrogenated varieties thanks to their superior properties. A practical overview of their key properties and formulation principles is provided, based on the author’s own background and practical experience. Each chapter contains information on their product forms, properties, processing and applications, with the emphasis on materials and concepts shown to work in practice.

Readers will learn why hydrogenated nitrile rubbers are now the first choice for a range of demanding applications, how their characteristics arise and how their properties can be adapted. Many readers will welcome the practical nature of the examples given and the way in which problems can be resolved, for example by employing statistical experimental design. Not only is this concept valuable in overcoming production issues in a logical and cost-effective manner, it is also of help in communicating with raw material suppliers and those equipment manufacturers who have become dependent on nitrile rubbers.

Published by Smithers Rapra Technology Ltd, 2012

Practical Guide to Hydrogenated Nitrile Butadiene Rubber Technology

Robert Keller

Practical Guide to Hydrogenated Nitrile

Butadiene Rubber Technology

Robert Keller

A Smithers Group Company

Shawbury, Shrewsbury, Shropshire, SY4 4NR, United Kingdom Telephone: +44 (0)1939 250383 Fax: +44 (0)1939 251118

http://www.polymer-books.com

First Published in 2012 by

Typeset by Argil Services

ISBN: 978-1-84735-521-8 (softback)

978-1-84735-522-5 (ebook)

Every effort has been made to contact copyright holders of any material reproduced within the text and the authors and publishers apologise if

any have been overlooked.

A catalogue record for this book is available from the British Library.

All rights reserved. Except as permitted under current legislation no part of this publication may be photocopied, reproduced or distributed in any form or by any means or stored in a database or retrieval system, without

the prior permission from the copyright holder.

©2012, Smithers Rapra Technology Ltd

Smithers Rapra Technology LtdShawbury, Shrewsbury, Shropshire, SY4 4NR, UK

iii

Contents

1 Overview of the Chemistry and Manufacture of Hydrogenated Nitrile Butadiene Rubber Polymers ...................... 1

1.1 Introduction ..................................................................... 1

1.2 The Starting Polymer: Nitrile Butadiene Rubber ................ 1

1.3 Hydrogenation of Nitrile Butadiene Rubber to Produce the Hydrogenated Polymer ................................................ 8

1.4 Summary ......................................................................... 12

2 Types of Hydrogenated Nitrile Rubber Polymers Available ....... 13

2.1 Introduction .................................................................... 13

2.2 Summary of Grades Available .......................................... 13

2.3 HNBR Grades and Technology ........................................ 17

2.4 Summary ......................................................................... 20

3 The Flow and Processing Behaviour of Hydrogenated Nitrile Rubber ........................................................................... 21

3.1 Introduction .................................................................... 21

3.2 The Basics of Viscoelastic Flow in Rubber ....................... 21

3.2.1 Viscosity ............................................................... 21

3.2.2 Elasticity .............................................................. 24

3.3 Effect of Process Variables ............................................... 26

3.3.1 Effect of Temperature on Viscosity ....................... 26

3.3.2 Heat Build-up During Flow .................................. 27

iv


3.4 Effect of Compounding Variables on Viscosity, Elasticity and Flow .......................................................... 29

3.4.1 Effect of Polymer Molecular Weight ..................... 29

3.4.2 Effect of Fillers ..................................................... 32

3.4.3 Effect of Plasticisers .............................................. 35

3.4.4 Effect of Process Aids ........................................... 36

3.5 Flow through Sprues and Runners: Mould and Machine Parameters Influencing Elastomer Flow ............. 37

4 Properties of Hydrogenated Nitrile Rubber and Comparison with Other Elastomers ........................................... 43

4.1 Introduction ................................................................... 43

4.2 Relationship between Hydrogenated Nitrile Rubbers and Other Speciality Elastomers ...................................... 43

4.2 Specific Comparison of HNBR and NBR ......................... 46

4.3 Comparison of HNBR with HNB and Other Speciality Elastomers ....................................................... 49

5 Typical Applications of Hydrogenated Nitrile Rubber ............... 55

5.1 A Brief Recap .................................................................. 55

5.2 Abrasion-resistant Belts and Conveyor Components........ 56

5.3 Flexible Boots for Power Transmission Joints .................. 57

5.4 Static Seals for Power Transmission Fluids....................... 58

5.5 Dynamic Fluid Seals for Power Transmission................... 59

5.6 Hydraulic Fluid and Lubricant Hoses .............................. 59

5.7 Seals for High-volume Air Conditioning ......................... 60

5.8 Seals used in Petroleum Exploration and Drilling ............ 60

5.9 Other applications ........................................................... 60

6 Formulating Guidelines for Hydrogenated Nitrile Rubber ......... 63

6.1 Introduction .................................................................... 63

Contents

v

6.2 Vulcanisation Systems ...................................................... 63

6.3 Carbon Black Fillers ........................................................ 70

6.4 Non-black Fillers ............................................................. 73

6.5 Plasticisers ....................................................................... 75

6.6 Antioxidants and Antiozonants ....................................... 78

6.7 Other Ingredients ............................................................ 78

6.8 Blends with Ethylene–Propylene Diene Rubber ................ 79

6.9 Examples ......................................................................... 79

6.9.1 Black HNBR Compound for a Reciprocating Lip Seal on a Power Steering Rack End ................ 79

6.9.2 Light Brown HNBR Compound for a Hydraulic System O-ring ...................................... 80

7 Examples of Hydrogenated Nitrile Rubber Formulae for Specific Applications .................................................................. 83

7.1 Introduction .................................................................... 83

7.2 Statistical Experimental Design and Desirability Functions ...................................................... 83

7.3 Theoretical Example: High-temperature HNBR Joint Boot ........................................................................ 85

7.4 Compound Example: High Temperature Oil Cooler O-ring Seal ...................................................................... 93

7.5 Compound Example: Oil Field High-pressure Well Packer ...................................................................... 95

7.6 Compound Example: High temperature Long-life Automotive Serpentine Belt ............................................. 97

7.7 Compound Example: Orange Water Pump Mechanical Seal Protective Bellows ................................. 99

7.8 Compound Example: High-temperature Differential Shaft Seal .................................................... 101

7.9 Compound Example: Short Steering System Hose ......... 102

vi


7.10 Compound Example: Chemically Resistant, Low Hysteresis Roller for Paper Mills ........................... 104

7.11 Concluding Comments .................................................. 105

8 Solving Hydrogenated Nitrile Rubber Processing Issues .......... 107

8.1 Introduction .................................................................. 107

8.2 Example 1: Poor Flow, Knit Lines and Non-fills in Injection Moulding ........................................................ 107

8.2.1 Possible Factor: Temperature of the Unvulcanised Rubber During Injection ............... 107

8.2.2 Possible Factor: Excessive Shear During Injection ............................................................. 108

8.2.3 Possible Factor: Sprue and Runner Design ......... 111

8.2.4 Solving the Injection Flow Problem using the Above Factors .............................................. 112

8.3 Example 2: Sporadic Dimensional Shift in Injection Moulded Parts ................................................ 113

8.3.1 Possible Factor: Mould Temperature .................. 113

8.3.2 Possible Factor: Injection Pressures and Holding Pressures and Times .............................. 114

8.3.3 Solving the Injection Moulding Dimensional Problem using the Above Factors ....................... 115

8.4 Example 3: Variation in Hose Tube Thickness during Extrusion ............................................................ 116

8.5 Example 4: Variation in Dimensions of High-volume Compression Moulded O-ring ....................................... 118

8.5 Conclusions ................................................................... 122

Abbreviations .................................................................................... 123

Index ............................................................................................... 127

vii

Preface

The goal of this book has been to give the reader a short but comprehensive view of the technology of hydrogenated nitrile rubber (HNBR). It is not intended to provide an exhaustive literature search, but more to act as a guide for those new to HNBR and seeking more information. My purpose has been to give an overview of:

• ThemanufactureandchemistryofHNBRelastomers.

• Thecommercialgradescurrentlyavailable.

• HowHNBRfitsintothecurrentworldofspecialityelastomers.

• ThebasicsofcompoundingandformulatingHNBR.

• Examplesofcompoundingforspecificapplications.

• Itsrheologicalpropertiesandhowthesecanbeapplied.

• Problemsolvingtechniquesusingspecificexamples.

Much of the information will be found applicable not only to HNBR but also to other elastomers.

viii


1

1.1 Introduction

This introductory chapter reviews the manufacture of hydrogenated nitrile butadiene rubber (HNBR) polymers. It is not intended to cover every facet of the production process, but the information provided will give the reader a background to the manufacture of these industrially significant rubber polymers. It is also not the intention to cover the proprietary specifics of polymer manufacturers.

As the name suggests, hydrogenated nitrile polymers are manufactured by the catalytic hydrogenation of conventional nitrile butadiene rubber (NBR) polymers. Firstly, the manufacture of the basic NBR polymers will be discussed. Secondly, the process flow and the chemistry of the catalytic hydrogenation process will be outlined. Finally, the challenges presented by the manufacture of the HNBR polymers with lower acrylonitrile (ACN) content will be briefly discussed. These HNBR polymers are important for achieving low temperature flexibility in the resulting moulded goods.

1.2 The Starting Polymer: Nitrile Butadiene Rubber

Nitrile rubber, ASTM D1418 designation NBR, is a copolymer of ACN and 1,3-butadiene. It is formed by free-radical polymerisation, shown in Figure 1.1.

The polymerisation is a typical emulsion polymerisation, which has the following characteristics:

• Themonomersareusuallycontained inemulsifieddroplets in thecontinuousphase. In the polymerisation of NBR the continuous phase is water, and the butadiene and ACN monomers and resulting polymers are contained in the emulsified droplets in the ‘oil’ or emulsified phase.

• Theinitiator,inthiscaseafree-radicalspecies,isgenerallyquitesolubleinthecontinuous phase. The majority of NBR polymers are described as ‘cold’ polymers, since a free radical is formed at sub-ambient temperature (typically 5 °C) in an oxidation–reduction reaction. Other NBR polymers are formed at elevated

1 Overview of the Chemistry and Manufacture of Hydrogenated Nitrile Butadiene Rubber Polymers

2


temperatures (typically around 95 °C) and are known as ‘hot’ polymers. Since the free-radical initiator is soluble in the continuous phase but only partially soluble in the emulsified phase, free radicals diffuse slowly into the emulsified phase. In contrast to bulk or solution free-radical polymerisation, in which a relatively high level of free radicals occur with the chain-growth polymer and cause early termination of polymerisation, emulsion polymerisation is diffusion-controlled and chain growth gives very high molecular weights. Chain transfer agents such as organo-mercaptans may be added to the emulsified phase to control the molecular weight of the polymer to give the desired bulk viscosity.

• Anemulsifier,whichmaybesimplyasodiumstearatesoap,assuresfinedispersionand emulsification of the monomer/polymer-rich phase.

• Thecontinuousphaseactsasanexcellentheattransfermediumintheexothermicpolymerisation reaction. In some cases bulk free-radical polymerisation generates so much heat that the process becomes impractical. Due to heat transfer in the continuous phase in emulsion polymerisation very tiny droplets of bulk polymer are generated, eliminating the need for expensive solvents and solvent recovery.

• Theviscositydoesnotincreaseduringtheconversion,asisthecaseinbulkorsolution polymerisation. The viscosity remains essentially constant at the viscosity of the continuous phase.

R*

Initiator = free radicalOnly partially soluble in ‘oil’ phase

Monomers in emulsified ‘oil’ phasePolymer forms here

Typical size of emulsified phase = 100 nm

CH2=CH---CH=CH2 CH 2=CH >>>>> ---(CH2 ----CH=CH----CH2)x-(CH2----CH)y-(CH 2=CH )Z-| | |CΞ 2 CΞN

Typical composition of resulting NBR polymer = X = 84-50%, Y = 5% of butadiene incorporated, Z = 16-50%

+

N CH=CH

Figure 1.1 Emulsion free-radical polymerisation of NBR

3

Overview of the Chemistry and Manufacture of Hydrogenated Nitrile Butadiene Rubber Polymers

In general, NBR polymers contain between 18 and 50 mole% ACN content, the remainder being butadiene. A crude schematic of the emulsion polymerisation is shown in Figure 1.2.

Polymerisation Coagulation

Emulsifier

Acrylonitrile

Butadiene Initiator Water

Drying and dewateringFinal polymer bales

Figure 1.2 Schematic showing NBR polymerisation

NBRarecopolymersandfollowtheMayo–Lewiscopolymerequation[1].Sincetwomonomers, ACN and butadiene, are involved the active terminal grouping in the growing chain can undergo the following possible reactions:

M1* + M1 ----- M1–M1*, reaction rate = k11 (1.1)




4


The relative reactivity ratios of the monomers can then be defined as follows:

r1 = k11/k12 = reactivity ratio for monomer 1 (1.5)

r2 = k22/k21 = reactivity ratio for monomer 2 (1.6)

The molar concentration of monomers at any instant during the polymerisation is givenbyEquation(1.7):

d[M1]

d[M2]

[M1]*(r1[M1]+[M2])

[M2]*([M1]+r2[M2])= (1.7)

where[M1]and[M2]arethemolarconcentrationsofmonomer1andmonomer2,respectively.

ThelimitingcasesinEquation(1.1)areasfollows:

• r1=r2>>1.Thetwomonomerreactivityratiosarehigh,indicatingthattheactivated monomers have little tendency to react with one another. Each activated monomer would more probably react with the same monomer, leading to a mixture of homopolymers in the resulting product (1-1-1-1-1-1-1-….. + 2-2-2-2-2-2-…..).

• r1=r2>1.Whenmonomer1formsthereactivechainenditwillbemorelikelyto react with further monomer 1. In the instance where cross-polymerisation occurs and monomer 2 becomes the reactive chain end, activated monomer 2 will probably react with further monomer 2. The resulting polymer will then be a block copolymer containing long segments of monomer 1 attached to long segments of monomer 2 (1-1-1-1-1-…….2-2-2-2-2-2-2….).

• r1= r2= approximately 1.The reactive chain ends,whethermonomer1ormonomer 2, have no preference for either monomer 1 or monomer 2 and the result is a random copolymer (e.g., 1-2-1-1-2-2-2-1-2-1-1-1-1-2-2-1-2-2-2-2-2-1-2-2-2….).

• r1=r2=approaching0.Ifmonomer1formsthereactivechainendithaslittlelikelihood of reacting with further monomer 1, and similarly, if monomer 2 is the reactive chain end it has little likelihood of reacting with monomer 2. In this case, a perfectly alternating copolymer will result (1-2-1-2-1-2-1-2….).

5


• r1>>1>>r2.Atthebeginningofthepolymerisationmonomer1hasamorereactive chain end than monomer 2, and the initial copolymer is therefore rich in monomer 1. Later, as the monomer 1 concentration is reduced, the copolymer becomes rich in monomer 2. This is known as compositional drift (early in the reaction: 1-1-1-1-1-2-1-1-2-1-1-1-….., followed later by: 1-2-2-2-2-2-2-1-2-2-2-1-2-2-2….).

• 0<r1<1and/or0<r2<1.Thereactiontendstowardshomopolymerisation,and this tendency increases with decreasing rX.

WecanalsorearrangetheMayo–Lewisequationtoexpresstheratioofmonomer1 in the instantaneous polymer with respect to the monomer concentrations and reactivity ratios:

F1[r1f12 + f1f2]

[r1f12 + 2f1f2 + r2f22]= (1.8)

where:

F1 = instantaneous mole fraction of monomer 1 in the instantaneous polymer formed,

f1 = mole fraction of monomer 1 in the instantaneous monomer feed, and

f2 = mole fraction of monomer 2 in the instantaneous monomer feed.

Againstthisbackground,thequestionarisesastothevaluesofr1andr2fortheACNand butadiene monomers used in the emulsion polymerisation of NBR. From this we can gain a perspective on some of the challenges presented by the polymerisation reaction.

A paper describing the reactivity ratios of ACN and butadiene has been published byEmbreeandco-workers[2].TheemulsionpolymerisationofNBRpolymershasa number of complicating factors:

• ACNhas reasonably high solubility in the aqueous continuous phase.ACNmonomer will therefore be present in both the emulsified and the continuous phase.

• TheresultingNBRpolymerisnotsolubleinbutadiene,andthismayresultinanon-homogeneous emulsified phase.

6


In their study of polymer composition and molecular weight, Embree and co-workers used previously published data in addition to a number of their own experiments. Their analyses indicated that r1 (butadiene) was 0.18 and r2 (ACN) was 0.03. In this case,r1>>r2,andbothr1andr2weremuchlowerthan1.0.Ifthereactivechainendwas a butadiene monomer, the reactivity ratios would indicate a greater probability that the reactive chain end would react with a further butadiene monomer unit. On the other hand, in the case where the reactive chain end was an ACN monomer the reactivity ratios would indicate a greater likelihood that the reactive chain end would also react with another butadiene monomer. From these reactivity ratios we would expect the resulting NBR copolymer to contain small segments of repeating butadiene and relatively isolated ACN units.

Wecanalsouse thedataofEmbreeandco-workers [2] tounderstandtheeffectof conversion on the composition of the polymer (i.e., how much of the initial monomer charge is converted to polymer) for a given initial charge of monomers into the emulsion polymerisation system. For example, if we take their example of a 75/25molarchargeofbutadiene/ACNmonomersinthemixture,theirdatamaybereplotted as shown in Figure 1.3.

40

30

20

10

00 20 40

% Conversion

% ACN in polymer, 75/25Butadiene/ACN Charge mixture

% ACN incrementalpolymer

% ACN in polymer mix

% A

CN

60 80 100

Figure 1.3 Example of polymer ACN versus conversion

In this case, the incremental (instantaneous) polymer composition early in the polymerisation is relatively constant at 30–35 mole% ACN. At higher conversions later in the reaction the incremental polymer becomes richer in butadiene, with a lower mole% of ACN. It is obvious that very low conversion values are impractical,

7


since the remaining monomer will have to be stripped out of the polymer. From Figure 1.3,apracticalbalanceforthisblendofmonomersmightgive70–80%conversion,when the overall polymer produced would contain roughly 33 mole% ACN, with reasonably high conversion of monomer to polymer, and a reduction in the energy andresourcerequiredtostripresidualmonomerfromtheresultingemulsionmixture.WecanalsouseEquation(1.3)todeterminethemole%ACNintheinstantaneouspolymer formed as a function of the mole% composition of monomer (Figure 1.4):

0%10%20%30%40%50%60%70%80%90%

100%

0% 20% 40% 60% 80% 100%

poly

mer

NC

A%

%ACN in monomer feed

%ACN in instantaneous polymer

%ACN in polymer

Ideal

Figure 1.4 ACN content of polymer versus monomer feed

The ideal line in Figure 1.4 would be one at which the reactivity ratios are 1.00 and the monomer feed composition results in the corresponding instantaneous polymer composition.

This understanding of the NBR polymer formed by emulsion polymerisation is important in appreciating the properties of the HNBR polymer formed by the catalytic hydrogenation of the NBR polymer, and the challenges presented by the process. The key points in the emulsion polymerisation yielding NBR polymers are as follows:

• Thereactivityratiosofthetwomonomers,butadieneandACN,indicatethatthe NBR polymers formed will have at least short segments of repeat butadiene monomer and relatively isolated ACN monomer units. This is particularly true in the case of NBR polymers of very low ACN content.

8


• AthighconversionvaluesthecompositionoftheNBRpolymerchanges,andthe resulting bulk NBR polymer may well be a blend of a range of polymers of different ACN content.

• The reactivity ratios ofACNand butadiene tend to favour the presence ofalternating monomers in the resulting polymer.

1.3 Hydrogenation of Nitrile Butadiene Rubber to Produce the Hydrogenated Polymer

The hydrogenation process follows the scheme shown in Figure 1.5. As discussed previously, the first significant step is emulsion polymerisation to give the initial NBR polymer. The NBR polymer is then coagulated, dried and stripped of residual monomer. The hydrogenation reaction is completed by dissolving the NBR polymer in a suitable solvent, followed by catalytic hydrogenation to give HNBR.

NBR Polymer

Solvent

Dissolution

Hydrogen

High pressurecatalytic hydrogenation

Separation fromsolvent and solventremoval

DryingBales of HNBR polymer

Figure 1.5 Schematic process for the catalytic hydrogenation of NBR to form HNBR

A catalyst which facilitates hydrogenation is normally employed. A typical combination is monochlorobenzene as solvent and OsHCl(CO)(O2)(PCy3)2 as the

9


catalyst[3],inwhichCydenotesacyclohexylresidue. The hydrogenation reaction is illustrated in Figure 1.6.

----{CH2—CH}x----{CH2—CH=CH—CH2}y---{CH2—CH}z----

CN CH=CH2

acrylonitrile 1,4-butadiene addition

1,2-butadiene addition, vinyl residue

NBR polymer

Metal catalysts, high pressure

+ (2x+2y) H2 (gas)

----{CH2—CH}x----{CH2—CH2 –CH2 —CH2}y ---{CH2—CH=CH—CH2}a ---{CH2—CH}z----

CN CH2 – CH3

acrylonitrile ethylene propylene

HNBR polymer

residual unsaturationfor vulcanisation

Figure 1.6 Catalytic hydrogenation of NBR

In effect, the resulting HNBR is a tetrapolymer of ACN, ethylene and propylene, containing a small amount of residual unsaturation which allows rapid vulcanisation by means either of sulfur or a peroxide cure.

The emulsion polymerisation process to give the NBR starting polymer may be summarised as follows:

• Thereactivityratiosofthetwomonomers,butadieneandACN,indicatethatthe NBR polymer formed will comprise short segments of repeating butadiene and relatively isolated ACN monomer units. This is particularly true for NBR polymers of very low ACN content. The resulting HNBR polymer would thus contain segments of an ethylene–propylene polymer of high ethylene content.

• Athighconversionvalues,thecompositionoftheNBRpolymerchanges,anditmay well comprise a blend of polymers of different ACN content.

These segments of ethylene–propylene polymer of high ethylene content create major challenges inthemanufactureofHNBRpolymersof lowACNcontent,required

10


when the end application demands improved low temperature flexibility. However, HNBR polymers of low ACN content produced from conventional NBR polymers actually have inferior low temperature flexibility to HNBR polymers of medium or high ACN content. The tendency for ethylene–propylene segments of high ethylene content to crystallise is an example of this. In the author’s laboratories the glass transition temperatures (Tg) of ethylene–propylene rubber polymers over a range of ethylene levels and low residual unsaturation have been studied using differential scanning calorimetry (DSC). The results are shown in Figure 1.7.

040 45 50 55 60 65 70 75 80

Tg (ºC)

linear Tg (ºC)

y = 1.2688 x – 102.79R2 = 0.9679

–10

–20

–30Tg

Mole% ethylene

Tg (ºC) of low vinyl EPDM

–40

–50

–60

Figure 1.7 Tg of ethylene–propylene copolymers versus ethylene content

As can be seen from Figure 1.7, low temperature flexibility as measured by Tg deteriorates rapidly at higher ethylene content due to the tendency of the ethylene segments to crystallise, in effect forming low-density polyethylene segments. In the case of a typical cold-polymerised NBR base polymer, the ethylene–propylene segments in the resulting HNBR polymer will comprise roughly 95 mole% ethylene, with very poor low temperature flexibility, using Figure 1.7 to give an estimated value of Tg. Put simply, if Figure 1.7 can be used to estimate the Tg of a HNBR of low molar ACN content, and assuming that the ethylene content of the short ethylene–propylene segments to be 95 mole% (typical for a cold-polymerised NBR starting polymer), then an estimate of the Tg of the resulting HNBR polymer would be +18 °C.

11


The challenge for manufacturers of HNBR polymers is thus to disrupt the regularity of the high ethylene content segments of the resulting HNBR polymer. This can be achieved by introducing a third diene monomer into the initial NBR polymerisation, since this inhibits the tendency of the ethylene segments to crystallise. An example is the emulsion copolymerisation of ACN, 1,3-butadiene and isoprene, with 1,3-butadiene/isopreneratioslyingwithintherange0.75/1.0to1.0/0.75[4].Thependantmethylgroup on the isoprene monomer then disrupts the regularity and crystallisation of the resultinghydrogenatedpolymer.Thereactionsequenceandfinalpolymerstructureare indicated in Figure 1.8.

----{CH2—CH}x----{CH2—CH=CH—CH2}y----{CH2—CH=CH—CH2}y--{CH2—CH}z----

CN CH3 CH=CH2

arylonitrile 1,4-butadiene addition

1,2-butadiene addition, vinyl residue

NBR Polymer

Metal catalysts, high pressure+ (2x+2y) H2 (gas)

----{CH2—CH}x----{CH2—CH2 –CH2 —CH2}y ---{CH2—CH=CH—CH2}a ---{CH2—CH}z----

CN CH2 –CH3

acrylonitrile ethylene propylene

HNBR Polymer

residual unsaturationfor vulcanisation

1,4-isopreneaddition

higher propylene content due to Isoprene

Figure 1.8 Hydrogenation reaction of ACN/1,3-butadiene/isoprene terpolymer

The incorporation, for example, of a 1/1 ratio of 1,3-butadiene/isoprene into the startingpolymerwouldresultinaroughly73/27molarratioofethylene/propylenein the final HNBR polymer (1 mole of 1,3-butadiene results in 0.95 moles of ethylene and 0.05 moles of propylene in the final HNBR polymer, and 1 mole of isoprene results in 0.5 moles of ethylene and 0.5 moles of propylene):

12


• 1moleof1,3-butadiene=0.95molesethylene+0.05molespropyleneinHNBRpolymer.

• 1moleofisoprene=0.5molesethylene+0.5molespropyleneinHNBRpolymer.

• Totalmolesofethylene=0.95+0.5=1.45.

• Totalmolesofpropylene=0.05+0.5=0.55.

• Ethylene/propyleneratiointhefinalHNBRpolymer=1.45/(1.45+0.55)=0.73.

• Ethylene/propylene ratio in thefinalHNBRpolymerwithno isoprene in thestarting polymer = 0.95.

The reduced ethylene content in the ethylene–propylene segment gives significantly improved low temperature flexibility.

1.4 Summary

The background to the technology of the HNBR polymer manufacturing process and a brief review of some of the challenges involved in manufacturing these polymers have been presented.

References

1. F.R. Mayo and F.M. Lewis, Journal of the American Chemical Society, 1944, 66, 9, 1594.

2. W.H.Embree,J.M.MitchellandH.L.Williams,Canadian Journal of Chemistry, 1951, 29, 3, 253.

3. C. Mouli, R. Madhuranthakam, Q. Pan and G.L. Rempel, AIChE Journal, 2009, 55, 11, 2934.

4. H.Bender,R.Casper,H.R.Winkelbach,H.C.Strauch,P.Nguyen,S.X.GuoandJ.Gamlin,inventors;BayerAG,assignee,USPatent7,091,284,2006.

13

2.1 Introduction

This chapter provides an overview of the types and grades of HNBR polymers commercially available and their performance characteristics in the context of other oil-resistant polymers.

2.2 Summary of Grades Available

Table 2.1 summarises the grades of HNBR polymers available from two manufacturers, Zeon Chemicals and Lanxess. The primary sort was made by mole% ACN, a secondary sort by approximate residual unsaturation and a tertiary sort by nominal Mooney viscosity. It is seen that HNBR polymers are available with mole% ACN within the range 17 to 50%, at a variety of residual unsaturation levels and bulk viscosities. For a given mole% ACN content, a higher residual unsaturation figure is useful for providing more rapid sulfur vulcanisation or when greater crosslink density is required in the finished product. A lower bulk Mooney viscosity results in improved high-shear flow in transfer or injection moulding.

2 Types of Hydrogenated Nitrile Rubber Polymers Available

14


Table 2.1 HNBR grades commercially available, December 2010

Grade Manufacturer Mole% ACN

(nominal)

Mooney viscosity ML(1+4) at 100°C (nominal)

Residual % unsaturation

(approximate)

Zetpol 4300EP Zeon 17 30 0.5

Zetpol 4300 Zeon 17 75 0.5

Zetpol 4310EP Zeon 17 30 5

Zetpol 4310 Zeon 17 62 5

Therban AT LT 2004 VP Lanxess 21 39 0.9

Therban LT 2007 Lanxess 21 74 0.9





Therban AP A 3404 Lanxess 34 39 0.9

Therban 3406 Lanxess 34 63 0.9


Therban AT C 3443 VP Lanxess 34 39 4

Therban 3446 Lanxess 34 61 4


Therban VP KA 8837 Lanxess 34 55 18

Zetpol 2000EP Zeon 36 30 0.5

Zetpol 2000L Zeon 36 65 0.5

Zetpol 2000 Zeon 36 85 0.5

15

Types of Hydrogenated Nitrile Rubber Polymers Available


Therban 3627 Lanxess 36 87 2

Zetpol2010L Zeon 36 58 4


Zetpol 2010H Zeon 36 135 4

Therban AT A 3904 VP Lanxess 39 39 0.9


Therban AT 4304 VP Lanxess 43 39 0.9






Zetpol 1000L Zeon 44 65 2




Zetpol 1020L Zeon 44 57 9



Therban 5008 VP Lanxess 49 80 0.9

Therban AT 5065 VP Lanxess 49 55 6



Therban® is a registered trademark of Lanxess Zetpol® is a registered trademark of Zeon Chemicals

16


In addition to these grades, speciality polymers are available for specific applications and end-products. These are summarised in Table 2.2.

Table 2.2 HNBR speciality grades available

Grade Manufacturer Mole% ACN

(nominal)

Mooney viscosity ML(1+4) at 100 °C (nominal)

Residual % unsaturation

(approximate)

Modification

Zeoforte ZSC 2295CX

Zeon 36 95 9 Zinc methacrylate

Zeoforte ZSC 2295L


Zeoforte ZSC 2385


Therban XT VP KA 8889

Lanxess 33 77 3.5 Carboxylated (XHNBR)

Therban VP KA 8796

Lanxess 34 22 5.5 Acrylate

Therban® is a registered trademark of Lanxess Zetpol® and Zeoforte® are registered trademarks of Zeon Chemicals

Modification with acrylates such as zinc methacrylate gives tough and mechanically durable materials for use in power transmission belts and conveyor belts, and in similar products requiring a high level of resistance to abrasion, cutting and wear.

17


2.3 HNBR Grades and Technology

One way to illustrate the position of HNBR within the elastomer market is to employ the ASTM D2000 classification system [1]. This is illustrated in Figure 2.1, along with a selection of other elastomers for comparison. Not only does the HNBR family fill a gap in the properties identified, HNBR also offer tear resistance, abrasion resistance and overall toughness that polymers such as acrylic elastomers (ACM) and ethylene acrylic copolymer elastomers (AEM) cannot match.

ASTM D2000 Classification

0

50

100

150

200

250

300

050100150200

% Swell in ASTM IRM–903 Oil

Tem

per

atur

e re

sist

ance

(ºC

)

NRCR

AEM

VMQ FVMQ

FKM

NBR

ACM

Class by Oil Swell

EPDMHNBR

Figure 2.1 Graphical illustration of speciality elastomers, showing the position of HNBR within the ASTM D2000 system

NBR are among the oldest and most widely used oil-resistant polymers and provide an excellent combination of properties and durability, but due to their high level of unsaturation they are prone to oxidation and to attack by sulfur. HNBR materials are much more resistant to oxidation and sulfur attack and combine the compounding flexibility and toughness of NBR with improved temperature and chemical resistance. This is illustrated in Figures 2.2 and 2.3 in the case of resistance to engine oil and automatic transmission fluids, respectively. The loss of elongation after 1008 hours at 150 ºC is shown for HNBR, NBR, ACM and AEM materials [2]. An estimate of the life of an elastomer is the time required to reduce the elongation to 50% of its original value. Figures 2.2 and 2.3 show that HNBR materials still retain some useful life after long-term ageing, whereas NBR and CSM materials had reached the

18


end of their useful life before the end of the ageing test. The ACM material showed excellent long-term ageing performance in these fluids, but did not have the tear and abrasion resistance of HNBR materials.

–60–50–40–30–20–10

010203040

35% ACN HNBR

25% ACN HNBR

ACM AEM

% E

lon

gati

on

Lo

ss

Material

Comparison of Oil-ResistantElastomers, 10W30SG engine oil at 150 °C

% Elongation Loss, 504 h


Figure 2.2 Comparison of engine oil resistance of HNBR, ACM and AEM elastomers

–30–25–20–15–10

–505

10152025

35% ACN HNBR

25% ACN HNBR

ACM AEM

% E

lon

gati

on

Lo

ss

Material

Comparison of Oil-resistant

Elastomers, Dexron III ATF at 150 °C



Figure 2.3 Comparison of resistance to automatic transmission fluid (ATF) of HNBR, ACM and AEM elastomers

19


The relative abrasion resistance of HNBR materials compared to other oil-resistant elastomers, such as NBR, epichlorohydrin–ethylene oxide copolymers (ECO), fluorocarbons (FKM) and ACM, is shown in Figure 2.4 [3]. Traditional NBR materials are given a relative rating of 100.

Akron-type abrasion resistance rating

0

50

100

150

200

45% ACNHNBR

NBR ECO FKM ACM

Material

Rel

ativ

e re

sist

ance

to

ab

rasi

on

Figure 2.4 Comparison of the abrasion resistance of various elastomers

The effects of abrasion, wear and friction are not well understood, but this type of testing is valuable in providing a relative rating for a range of materials. The combination of increased high temperature performance and improved wear resistance emphasises the niche occupied by HNBR among the available oil-resistant elastomers.

In the remainder of this chapter the effect of HNBR and compounding ingredients on the final properties of the compound are discussed. This is not intended to be an exhaustive account of every compounding possibility, but the use of statistical experimental design will be emphasised and illustrated as a cost-effective way of studying the compounding variables.

Finally, it is important to compare the relative cost of the various oil-resistant elastomers. The pound–volume cost is used, which is simply the cost of the compound multiplied by its specific gravity. This gives the relative cost of the compound required to fill a given volume such as a mould cavity. This has been shown earlier in Table 2.2, along with the upper temperature limits of the various materials. While acrylic elastomers such as ACM and AEM are cost-effective up to 150 ºC, HNBR are much tougher and more abrasion-resistant. FKM and fluorosilicones (FVMQ) have excellent upper temperature limits, but their pound–volume cost is heavily influenced by their relatively high specific gravity.

20


Table 2.3 gives a cost comparison of oil-resistant elastomers, and a combination of Table 2.3 and Figures 2.2 to 2.4 puts the niche filled by HNBR materials in perspective. HNBR combine toughness, abrasion resistance, fluid resistance and good cost effectiveness at temperatures up to 150 ºC.

Table 2.3 Commercial comparison of oil-resistant elastomers

Polymer SG Cost of compound ($)

lb–vol cost ($) Operating temperature, upper limit (ºC)

NBR 1.22 $1.00 $1.22 100

ACM 1.32 $2.30 $3.04 150

AEM 1.32 $2.40 $3.17 150

HNBR 1.22 $10.40 $12.69 150

FVMQ 1.53 $23.00 $35.19 200

FKM 1.86 $16.00 $29.76 250

Note: 75 Shore A, assumes carbon black filler, except for FVMQ

2.4 Summary

This chapter has given a brief overview of the polymers currently available and how these fit into the overall range of oil-resistant elastomers. Formulation guidelines and examples of formulations for specific applications will be given in detail later.

References

1. Rubber, American Society of Testing and Materials, Washington, DC, USA, 2006, 9, 2, D2000-06.

2. A Comparison of Oil Resistant Elastomers in Engine Oil and ATF, Z7.3.16, Zeon Chemicals LP, Louisville, KY, USA, 1999, p.1.

3. Zetpol Product Guide, Zeon Chemicals LP, Louisville, KY, USA, 1999.

21

3.1 Introduction

In this chapter the principles of polymer viscoelastic flow behaviour are discussed. There are obviously a number of variables, such as compounding, and mould and machinery design, which will influence the flow behaviour of fully compounded HNBR materials. This chapter is not intended to discuss the finer details but will offer some insight into the many factors affecting flow and processing behaviour. This will provide guidelines for studying flow and processing behaviour, as well as indicating some of the useful tools available for solving production processing problems.

3.2 The Basics of Viscoelastic Flow in Rubber

3.2.1 Viscosity

Rubber and plastic melts can be regarded as extremely high viscosity fluids. This is only an approximation, however, and polymers generally show viscoelastic properties, a combination of viscous flow and elastic recovery. In turn viscosity is a measure of resistance to flow under a given set of circumstances. Viscosity is designated by the Greek letter h. For an ideal or Newtonian fluid, viscosity is simply the ratio between the shear stress (t), the pressure applied to the fluid to create flow, and the shear rate (g), the flow occurring over a given time:

h = t/g, (3.1)

where h = viscosity expressed as Pa. s, t = shear stress in Pa, and g = shear rate in s–1.

For an ideal fluid the viscosity is thus constant at a given shear stress over the entire range of shear rates. Life would certainly be much easier if viscosity were as simple

3 The Flow and Processing Behaviour of Hydrogenated Nitrile Rubber

22


as this. However, polymer melts such as rubbers do not in general exhibit uniform viscosity across a range of shear rates, in fact they are described as pseudoplastic, which means that their viscosity is reduced as the shear rate increases.

An easy way to visualise pseudoplastic behaviour is to imagine a bottle of thick tomato ketchup. If the bottle is allowed to stand and then upended the ketchup will remain in the bottle, but if the bottle is shaken vigorously the ketchup will then flow freely. At zero shear rate the viscosity (resistance to flow) is extremely high and flow does not take place, but on agitation the shear rate increases dramatically, the viscosity falls and the product can be poured without difficulty.

A typical viscosity curve for a rubbery material as a function of shear rate is shown in Figure 3.1. At low shear rates, the material shows Newtonian behaviour, which means that its viscosity remains constant regardless of shear rate. As the shear rate increases, the material is transformed from Newtonian to pseudoplastic, and its viscosity then decreases with increasing shear rate. This is a curve that we will encounter many times later in the chapter and we will see how it can help in understanding flow during moulding and in overcoming flow-related problems. A mental picture of this curve and its shape will help in practical situations.

1000

10000

100000

1000000

1 10 100 1000

Shear rate (s–1)

Vis

cosi

ty (

poi

se)

Newtonian region

Figure 3.1 Typical profile of viscosity versus shear rate for a rubber polymer

Once we appreciate that rubber materials are not ideal Newtonian fluids, and that their viscosity decreases with increasing shear rate, we find that these variables have a

23

The Flow and Processing Behaviour of Hydrogenated Nitrile Rubber

mathematical relationship. A great deal of scientific analysis has been devoted to the mathematics underlying the flow of rubber and other polymers [1–11], and Equations (3.2) and (3.3) provide a starting point:

h = Kgn–1, (3.2)

where h is the viscosity, g is shear rate in s–1 and K and n are a curve-fitting parameters determined from the experimental data.

h = ho/(1 + Ag1–n), (3.3)

where ho is the viscosity at zero shear rate and A, n and K are curve-fitting parameters determined from the experimental data.

Equation (3.2) is known as the Power Law and Equation (3.3) is the Cross Model. Plots of these two models are shown in Figure 3.2. In general, the Cross Model is the more successful in describing the actual viscosity versus shear rate data observed for commercial polymers.

Vis

cosi

ty

Shear rate

100000

10000

10001 10 100 1000

Figure 3.2 Comparison of theoretical models describing the pseudoplastic behaviour of polymers

24


The shear rate in compression moulding is normally within the range 1–10 s–1, for transfer moulding in the range 10–1,000 s–1, and for injection moulding in the range 500–10,000 s–1. Combinations and variations of these techniques give intermediate shear rates – and a number of fresh problems.

3.2.2 Elasticity

Unfortunately the elastic response of rubber to flow is probably the least well-understood feature of its viscoelastic nature. A good illustration of elastic flow is provided by chewing gum when pulled to breaking point. After chewing and softening, if one pulls the chewing gum to fracture the broken ends of the gum are seen to retract. The flow of the gum is the extension achieved on pulling, and its elasticity is represented by the recovery of the broken ends. In compression, transfer or injection moulding the flow of the rubber compound is similarly accompanied by elastic rebound.

If the elastic component of rubber flow is not allowed for, the results can sometimes be quite expensive. One of the best ways to study this effect has been to study the extrudate swelling (die swell) of polymers forced through a die. As the polymer exits the die it swells in diameter and decreases in length. Unfortunately, at the high shear rates encountered in injection moulding and with lightly filled elastomers, the extrudate swelling can be anything but smooth and predictable. The familiar knotted and twisted appearance of unfilled or lightly loaded elastomers on exiting a die at moderate to high shear rates is very difficult to quantify and study. This may also account for the knotted, twisted or dimensionally non-compliant nature of moulded rubber parts when the compound is lightly loaded or when shear rates are particularly high.

In an attempt to quantify and describe extrudate swelling Tanner [12] has found that the extrudate swell, B, can be estimated using Equation (3.4):

B = C + [1 + ½(tgw)2]1/6, (3.4)

where C is a constant, approximately 0.12, t is shear stress and gw is the shear rate at the wall of the die.

Equation (3.4) applies to molten polymers but does not properly describe fully filled or compounded elastomers. It is however useful for quantifying the change in elastic

25


rebound induced by increasing shear rate and shear stress. Using Equation (3.4) we can establish the proportionality between elastic recovery and shear stress or shear rate. Figure 3.3 shows the effect of a percentage increase in shear rate on the extrudate swell, B, or elastic recovery.

% D

ie s

wel

l

Shear rate (s–1)

454035302520151050

1 10 100 1000

Figure 3.3 Die swelling (elastic recovery) versus shear rate

This type of relationship is important in describing dimensional problems at the high shear rates encountered in transfer or injection moulding. For example, a rubber part may have originally been formed by injection moulding, with a 20 s injection time followed by 60 s vulcanisation. As a result of cost pressures, reductions in cycle time are required and the manufacturer then increases the injection pressure to bring the injection time down to 10 s. Unfortunately this increases the shear rate by a factor of 2, which increases the elastic rebound shown as B in Equation (3.4). The rebound is manifested by a sudden shift in the diameter of the finished part, making it undersized. Very soon the manufacturer’s injection press becomes an expensive scrap machine. As mentioned above, this is the least understood facet of the viscoelastic flow of rubber – but it is important in understanding some of the unwelcome consequences of process changes.

26


3.3 Effect of Process Variables

3.3.1 Effect of Temperature on Viscosity

Like engine oil, polymer melts show reduced viscosity with increasing temperature. In the absence of experimental data defining the curves, which is normally the case, this can be determined using the Williams–Landel–Ferry relationship for time–temperature superposition [13]. This relationship can be applied to values of modulus and also viscosity. Applying the relationship to viscosity, we obtain Equation (3.5):

log(hT1/T1) – log(hTg/Tg) = log(aT1), (3.5)

where hT1 is the viscosity at absolute temperature T1, hTg is the viscosity at the glass transition temperature and aT is a shift factor derived from the data or by estimation.

Obviously, the determination of viscosity at Tg would be difficult or almost impracticable. However, if we are only interested in comparing the viscosities at two different process temperatures, T2 and T1, we can simplify matters and use a relative comparison:

log(µT2/T2) – log(µT1/T1) = log(aT2) – log(aT1), (3.6)

in which the values of aTx can be estimated using Equation (3.7):

log(aTx) = [–17.44(Tx – Tg)] / [51.6 + (Tx – Tg)]. (3.7)

This is shown in Figure 3.4 for the theoretical polymer shown in Figure 3.1, assuming the Tg to be –20 ºC. As materials move further away from the Tg the reduction in viscosity with temperature becomes less apparent, and for a material with a very low Tg, such as a silicone, the reduction in viscosity with increasing temperature is similarly less dramatic.

27


1000000

100000

10000

1000

100100

100 °C

125 °C

150 °C

1000101

Vis

cosi

ty (

Pa. s

)

Shear rate (s–1)

Figure 3.4 Temperature effect on flow profile of a typical rubber material

3.3.2 Heat Build-up During Flow

One of the consequences of the viscoelastic behaviour of rubber is heat build-up during high shear flow. During relatively low shear compression moulding this is probably not a significant factor, but on the other hand during higher shear moulding, such as in transfer and injection moulding, heat build-up can be the cause of premature scorching and other quality problems. An exact relationship has not been established to describe heat build-up during high shear flow, but Equation (3.8) can predict the effect of an increase in viscosity and shear rate.

DT = 0∫t (µ/rc)g2 dt, (3.8)

where DT is the temperature change in ºC s–1, µ is the viscosity in Pa. s and g is shear rate in s–1.

Using Equation 3.8 we can establish the following proportionalities in order to estimate the effect of viscosity and shear rate on heat build-up:

DT ∝ h, and DT ∝ g2 (3.9)

28


Using these proportionalities we can plot the effect of viscosity and shear rate on heat build-up, as illustrated in Figure 3.5.

350%300%250%200%150%100%50%0%

–50%–100%

–100% –50%

% change in parameter

0% 50% 100% 150%

% c

hang

e in

hea

t bu

ild-u

p

Viscosity effect Shear rate effect

Figure 3.5 Effect of viscosity and shear rate on heat build-up during flow

Using the example from the previous section we have studied the decrease in injection cycle time resulting from increasing the injection pressure and by reducing the injection time from 20 s to 10 s. Since the configuration of the runner system and the mould cavities remained unchanged, a reduction in injection time by half had the effect of doubling the shear rate. In turn, doubling the shear rate caused a four-fold increase in heat build-up during flow. If this increase in heat build-up was sufficient for the critical scorch temperature of the vulcanisation system to be reached, we would encounter premature scorch problems, in addition to dimensional variances. Obviously, as we increase the shear rate the viscosity decreases, causing a linear decrease in heat build-up. However, the dependence of heat build-up on the square of the shear rate can easily override the viscosity drop and give rise to scorching.

29


3.4 Effect of Compounding Variables on Viscosity, Elasticity and Flow

3.4.1 Effect of Polymer Molecular Weight

This is the starting point for practical rubber compounds and will naturally be the first item to consider. The molecular weights of commercial polymers, including rubbers, tend to be skewed towards higher figures. A variety of averages are used to describe the molecular weight distribution and its width. The general formula for the j-average molecular weight (Mj) is given by Equation (3.10):

Mj = (ΣNiMij)/(ΣNiMi

j–1), (3.10)

where Ni is the number of molecules of molecular weight Mi and is a weighting average.

If j = 1, then it is the number average or simple arithmetic mean, Mn.

If j = 2, then this is the weight-average molecular weight, Mw, which is sensitive to higher molecular weight fractions.

If j = 3, this is known as the z-average molecular weight, Mz.

If j = 4, then this is described as the z + 1 average molecular weight, Mz+1.

A typical distribution of molecular weights, with various averages, is shown in Figure 3.6. The breadth of the distribution may be expressed as the weight-average molecular weight, Mw, divided by the number-average molecular weight, Mn, and is described as the polydispersity. For a typical emulsion polymer such as styrene–butadiene rubber (SBR) or nitrile rubber (NBR), polydispersity values are typically around 2. The polydispersity number can be larger than 2 for broad molecular weight distribution polymers such as Ziegler–Natta-catalysed polyethylene or ethylene–propylene diene rubber (EPDM).

30


0

0.1

0.2

0.3

0.4

0.5

0 100000 200000 300000 400000 500000 600000 700000

Molecular weight

Fra

ctio

n

Mn Mw MzMv

Figure 3.6 Example of molecular weight distribution of polymers

A great deal of experimental work has been carried out in the past to understand the influence of molecular weight on viscosity [14–22]. In linear polymers the viscosity varies linearly with molecular weight until a critical value of molecular weight is reached, at which entanglement of the polymer chains occurs, and the viscosity then varies as the power to 3.4 of the molecular weight, as in Equation (3.11):

h ∝ M when M < Mc, and h ∝ M3.4 when M > Mc, (3.11)

where h is viscosity, M is molecular weight and Mc is the critical molecular weight for entanglement.

The value of Mc varies with the type of polymer, and in vinyl (carbon–carbon backbone) polymers it is dependent on the bulk of the pendant groups. The value of Mc for polyethylene (a carbon–carbon backbone with no bulky pendant groups) is approximately 4,000, while that for polystyrene (carbon–carbon backbone with a bulky phenyl pendant group) is approximately 38,000. In general, rubber polymers have molecular weights well above Mc and the viscosity shows a power dependency of 3.4 on molecular weight.

Since rubber polymers contain a distribution of molecular weights, the question then arises as to which average molecular weight should be used to establish the dependency of viscosity. Fox and co-workers found that the weight-average molecular weight, Mw, was best able to describe the dependence of viscosity on molecular weight, Equation (3.12):

31


ho = kMw3.4, (3.12)

where ho is zero shear rate viscosity and k is a constant for a particular polymer system.

An interesting analysis carried out on polyethylene with a very broad molecular weight distribution [24] has indicated that viscosity-average molecular weight gives a better correlation with the experimental data. This may be relevant in the case of EPDM polymers produced with similar polymerisation catalysts, giving a very broad range of molecular weights.

A convenient way of estimating viscosity and molecular weight is by the use of the traditional Mooney viscometer. Since this operates at a shear rate around 2 s–1, and this is very close to the zero shear rate viscosity, it should show proportionality to Mw

3.4. The data reported by polymer suppliers as Mooney viscosity can thus give an estimate of the relative average molecular weight. As a first approximation we can assume that two polymers varying in Mooney viscosity will have the same degree of branching or linearity, and the viscosity shear rate curves will be parallel. Therefore, if we know the viscosity difference between the two polymers at low shear rates and we also have the viscosity shear rate curve for one of the polymers, we can estimate the viscosity at higher shear rates. For example, if two polymers from the same supplier and made by a similar polymerisation method have Mooney viscosities of 35 and 50, and we know the relative viscosities of the two polymers at high shear rates, these should follow the ratio of Mooney viscosities. In other words, the polymer of Mooney viscosity 50 s–1 has roughly 43% higher viscosity at low shear rate, and should also have roughly 43% higher viscosity at higher shear rates, since the curves are parallel. This example is illustrated in Figure 3.7.

10

100

1000

10000

100000

1000000

1 10 100 1000

Shear rate (s–1)

Vis

cosi

ty (

pois

e)

Molecular weight = 80,000Mooney viscosity = 23



Figure 3.7 Effect of molecular weight on flow behaviour

32


3.4.2 Effect of Fillers

Unfortunately this area is not well understood. The basic theory of particle reinforcement of liquids was developed as a refinement to the Einstein equation [25, 26]:

h1 = h0 (1 + Af + Bf2), (3.13)

where h0 is the unfilled viscosity, h1 is the viscosity of filled material, f is the volume fraction of filler in the mixture, A is a constant (2.5 for spherical particles) and B is a different constant (within the range 10 to 14 for spherical particles).

G1 = G0 (1 + Af + Bf2), (3.14)

where G0 is the shear modulus of unfilled material, G1 the shear modulus of filled material and f is the volume fraction of filler in the mixture.

As can be deduced from Equation (3.13), the type of filler should have neither an impact on viscosity nor on modulus of elasticity. In the practical world that we live in, we know that the loading of N110 SAF carbon black at the same volume or phr (parts per hundred rubber polymer) gives much higher viscosities and much higher modulus values compared to N990 MT carbon black. Thus, Equation (3.13) is not very effective in predicting the effect of fillers. From this point onwards the situation becomes unclear, since comprehensive studies of various fillers and their effect on viscosity and modulus are relatively rare.

White and Crowder [27] studied the effect on viscosity of various low-structure carbon blacks and found that those of smaller particle size caused an increase in viscosity and shifted the viscosity relative to the shear rate curves. The use of low-structure carbon blacks limited the structural effect caused by the agglomeration of particles. Unfortunately, many useful commercial elastomer compounds employ high-structure carbon blacks, such as furnace black N550 FEF.

In addition, the author has studied the combined effect of structure and particle size of carbon black on the properties of an NBR compound. Obviously the particle size on its own does not determine the results, but fortunately the DBP number, given in

33


ASTM D2414, combines the effect of particle size and structure. The DBP number represents the quantity of dibutyl phthalate (DBP) that the carbon black can absorb, measured by titration to a mechanically sensed end-point; when the carbon black surface can absorb no more DBP the conversion from free-flowing powder to paste is sensed as a change in torque. The experimental design programme is summarised in Table 3.1 and the results for compression modulus and low-shear rate viscosity are shown in Figure 3.8.

Table 3.1 Experimental design formulations, NBR compound, triplicate data points for each formula basic recipeIngredient Phr

33% Acrylonitrile, 35 ML(1+4)@100C polymer 100.00

Stearic acid 2.00

Agerite resin D 1.00

Vanox ZMTI 1.00

Carbon black (see below) 30.00 to 50.00

Zinc oxide 2.00

Varox DCP-40C 3.00

Compound Phr carbon black DBP number of carbon black

1 30.00 43 (N-990)

2 50.00 43

3 70.00 43

4 30.00 65 (N-762)

5 50.00 65

6 70.00 65

7 30.00 90 (N-660)

8 50.00 90

9 70.00 90

10 30.00 121 (N-550)

11 50.00 121

12 70.00 121

34


Response surface for compression modulus in Pa

DBP number

Com

pres

sion

mod

ulus

phr Carbon black

15.5

13.5

11.5

9.5

7.5

5.5

30 40 50 60 700

3060

90120

150

Figure 3.8 Effect of carbon black on compression modulus

The correlation between the experimental values and the plotted response surfaces was greater than 0.900, confirming that the DBP number gives a good prediction of the reinforcing nature of carbon blacks. The study was not extensive, however, and will need to be developed further to include a comprehensive range of polymer families, carbon blacks and other fillers. Prior to formulating compounds for compression, transfer or injection moulding, the compounder should conduct a similar series of experiments to optimise the fillers that will give a combination of properties such as compression set, strength of vulcanisate and flow in the mould.

In the study, elasticity was seen to improve with increasing filler loading and DBP value of carbon black, confirming earlier observations [28, 29]. Loss of elasticity was measured in terms of the hysteresis observed in dynamic compression testing, and is illustrated in Figure 3.9. It is of course well known in the rubber industry that high structure carbon blacks of small particle size give smooth extrusion, which delivers extrusions with smooth or glossy surfaces.

35


Hysteresis as a function of filler and loading

phr Carbon Black

DBP Number

% H

yste

resi

s, C

ompr

essi

on

30 40 50 60 70 30507090

11013023

33

43

53

63

R-Squared adjusted for degrees of freedom = 96%

Figure 3.9 Effect of carbon black on rubber material hysteresis

3.4.3 Effect of Plasticisers

This is a further area where comprehensive studies need to be undertaken, although good work has been conducted to pave the way. Kraus and Gruver [30] have studied the effects of various plasticisers on the viscosity of rubber compounds and have proposed the following equation to describe their influence on viscosity:

h = f23.4 F (gf2

1.4), (3.14)

where f2 is the volume fraction of polymer in the compound, F is a constant for the polymer and plasticiser and g is shear rate.

From Equation (3.16) we can establish the proportionality between plasticised and unplasticised material, as follows:

h1/h0 = (f13.4F(gf1

1.4)/ (f03.4F(gf0

1.4) = f14.8/f0

4.8 (3.15)

36


As more plasticiser is added, viscosity at a given shear rate falls. This is plotted in Figure 3.10, using our example from Figure 3.1. For a particular polymer and plasticiser system it should be straightforward to run the experiments and adapt the parameters of Equation (3.15) to the design of other compounds.

1000

10000

100000

1000000

1 10 100 1000

Vis

cosi

ty,

Pa. s

Shear rate (s–1)

Phi = 1

Phi = .9

Phi = .8

Figure 3.10 Effect of plasticiser on viscosity

3.4.4 Effect of Process Aids

Most process aids are designed to assist mould flow and mould release, and function by having only limited solubility in the polymer matrix. For example, microcrystalline polyethylene waxes make good process aids for highly polar polymers such as fluorocarbon elastomers (FKM) since they are deposited at the surface of the shear flow front and on the moulded parts, assisting lubrication and release. Similarly, specialised FKM polymers are available as process aids and slip agents for use in polyethylene manufacture. In general, effective process aids for mould flow and mould release are used at very low and tightly controlled levels in the polymer. Using our example of microcrystalline wax in FKM materials as a process aid in transfer or injection moulding, while 0.500 phr is very effective, 1.00 phr is also effective but may lead to mould deposits during extended operation involving several hundred cycles, and levels of 1.50 phr or above will give knit lines and parts that are ‘wet’ with excess wax as they emerge from the mould.

37


Techniques are available for studying the effect of process aids on mould flow, for example a capillary rheometer may be useful for this purpose. The pressure drop through the die of a capillary rheometer is made up of the total pressure drop at the ends plus the pressure drop within the die:

PT = Dpe + Dp, (3.17)

where PT is the total pressure drop through the die, Dpe the pressure drop at the ends of the die and Dp is the pressure drop within the die.

The shear stress at the wall of the capillary is governed by the total pressure drop and the length and diameter of the die:

PT = Dpe + 4(s12)w L/D, (3.18)

where (s12)w is the shear stress at the wall, L the die length and D the die diameter.

By running experiments at constant die diameter and total volumetric flow rate but with different L/D ratios, the slope of the plot of PT versus L/D allows the shear stress at the wall to be ascertained. From this, the effectiveness of process aids in improving flow can be measured in terms of the reduction in shear stress at the wall caused by their lubricating effect.

Assessing the effectiveness of a process aid in terms of mould cleanliness and release can be challenging. In many cases the only accurate measure is a carefully monitored extended production run, comparing long-term release and mould cleanliness with and without the process aid, or by comparing different process aids.

3.5 Flow through Sprues and Runners: Mould and Machine Parameters Influencing Elastomer Flow

This is an area that seems to lack a basic understanding at the plant and production level in the speciality elastomer industry. A typical problem might involve a multi-deck injection mould where there is insufficient flow to one of the decks. For a variety of

38


reasons the default diagnosis in elastomer production needs to be to fix this type of problem by changing or reformulating the rubber material. However, when we look at the underlying science this would seem to be the least effective solution to flow problems in production processing.

Viscosity is simply the ratio of shear stress and shear rate. For flow through a round capillary this can be expressed as follows:

Shear stress = DP . R / 2L, (3.19)

where DP is pressure drop and R and L are the radius and length of the capillary, respectively.

Shear rate = 4Q / pR3, (3.20)

where Q and R is the radius of the capillary.

Apparent viscosity happ = shear stress / shear rate

= DP . R4 . p / 4 . Q . 2L, and

the volumetric flow rate Q = DP . R4p / 4happ . 2L. (3.21)

By only a small increase in the radius of the flow channel the volumetric flow rate will thus increase by the fourth power of the radius. Decreasing the apparent viscosity of the compound, happ, will therefore have not nearly as profound an effect on improving flow rate as a small change in the radius or gap of the flow channel. The relative impact of these various factors is illustrated in Figure 3.11.

From Equation (3.21) and Figure 3.11, it is clear that a minor change in a sprue or runner system will have a disproportionate effect on rubber flow. In the case of a double-deck injection mould in which one deck is not filling completely, it is obvious that a small increase in the diameter of the sprues or runners leading to the problem deck would provide the most effective solution to the problem. It is strange that the immediate response often seems to be to reduce the viscosity by changing the formulation.

39


Radius of runner Length of runner Viscosity of material

% Change in factor

Net

% c

hang

e on

flo

w r

ate 120.0%

100.0%80.0%60.0%40.0%20.0%0.0%

–20.0%

–20.0% –10.0% 0.0% 10.0% 20.0% 30.0%

–40.0%–60.0%

Figure 3.11 Effect of various parameters on the flow rate of an elastomer

References

1. E.A. Collins and J.T. Oetzel, Rubber Age, 1970, 102, 64.

2. E.A. Collins and J.T. Oetzel, Rubber Age, 1971, 103, 3.

3. S. Einhorn and S.B. Turetzky, Journal of Applied Polymer Science, 1964, 8, 1257.

4. J.R. Hopper, Rubber Chemistry and Technology, 1967, 40, 462.

5. M. Mooney, Physics, 1936, 7, 413.

6. M. Mooney in Rheology: Theory and Applications, Ed., F.R. Eirich, Academic Press, New York, USA, 1958, p.196.

7. N. Nakajima and E.A. Collins, Polymer Engineering Science, 1974, 14, 137.

8. F.C. Weissert and B.L. Johnson, Rubber Chemistry and Technology, 1967, 40, 590.

9. J.L. White and J.W. Crowder, Journal of Applied Polymer Science, 1974, 18, 1013.

40


10. J.L. White and N. Tokita, Journal of Applied Polymer Science, 1974, 9, 1929.

11. W.E. Wolstenholme, Rubber Chemistry and Technology, 1965, 38, 769.

12. R.I. Tanner, Journal of Polymer Science, Part A2, 1970, 8, 2067.

13. M.L. Williams, R.F. Landel and J.D. Ferry, Journal of the American Chemical Society, 1955, 77, 3701.

14. P.J. Flory, Journal of the American Chemical Society, 1940, 62, 1057.

15. J.R. Shaefgen and P.J. Flory, Journal of the American Chemical Society, 1948, 70, 3709.

16. T.G. Fox and P.J. Flory, Journal of the American Chemical Society, 1949, 70, 2384.

17. T.G. Fox and V.R. Allen, Journal of Chemical Physics, 1964, 41, 344.

18. T.G. Fox and S. Loschaek, Journal of Applied Physics, 1955, 26, 1082.

19. W.F. Busse and R. Longworth, Transactions of the Society of Rheology, 1962, 6, 179.

20. T. Masuda, K. Kitagawa, T. Inoue and S Onogi, Macromolecules, 1970, 3, 116.

21. R.S. Porter and J.F. Johnson, Proceedings of the 4th International Rheology Congress, Providence, RI, USA, 1965, 2, 467.

22. R.S. Porter and J. F. Johnson, Chemical Reviews, 1966, 66, 1.

23. L.A. Belfiore, Physical Properties of Macromolecules, John Wiley and Sons, Hoboken, New Jersey, 2010, p.403.

24. W.F. Busse and R. Longworth, Transactions of the Society of Rheology, 1962, 6, 179.

25. E. Guth and R. Simha, Kolloid-Zeitschrift, 1936, 74, 266.

26. D. Barthes-Biesel and A. Acrivos, International Journal of Muliphase Flow, 1973, 1, 1.

27. J.L. White and J.W. Crowder, Journal of Applied Polymer Science, 1974, 18, 1013.

41


28. C.D. Han, Journal of Applied Polymer Science, 1974, 18, 821.

29. N. Minagawa and J.L. White, Applied Polymer Science, 1975, 20, 501.

30. G. Kraus and J.T. Gruver, Transactions of the Society of Rheology, 1965, 9, 2, 17.

42


43

4.1 Introduction

In this chapter the properties of hydrogenated nitrile rubber (HNBR) are discussed and compared with those of other speciality elastomers. This background will provide the perspective for later chapters in which the applications of HNBR materials are described.

4.2 Relationship between Hydrogenated Nitrile Rubbers and Other Speciality Elastomers

A good starting point for this comparison is to show where HNBR materials fit into the classification used by ASTM D2000 [1], which classifies elastomers by type and class. The ‘type’ classification gives an approximate upper temperature limit for the material and the ‘class’ characterises its degree of swelling in hydrocarbon-based IRM–903 oil, which has a high aromatic content. Table 4.1 gives a tabulation of type and class in ASTM D2000 [1].

Since HNBR is a hydrogenated derivative of NBR, a useful first step might be to compare the two. NBR polymers are generally characterised as BG (100 ºC test temperature, 40% or less volume swelling in IRM–903 oil). NBR materials can however be formulated with a high ACN content to give very low swelling in hydrocarbons, and are then characterised as BK (100 °C test temperature, 10% or less volume swelling in IRM–903). On the other hand, NBR with very low ACN content gives greater swelling in hydrocarbons but much improved low temperature flexibility (ASTM D2000 BF). Using peroxide curing and suitable antioxidants, NBR can also be made to have a higher upper temperature limit (CH). NBR elastomers generally lie within the range: BF, BG, BK or CH, as illustrated in Figure 4.1.

4 Properties of Hydrogenated Nitrile Rubber and Comparison with Other Elastomers

44


Table 4.1 ASTM D2000 type and class system

Type Test temperature (°C)

Class %Volume swelling in IRM–903 oil after 70 h immersion

A 70 A No requirement

B 100 B 140

C 125 C 120

D 150 D 100

E 175 E 80

F 200 F 60

G 225 G 40

H 250 H 30

J 275 J 20

K 300 K 10

0

50

100

150

200

250

300

050100150200

Tem

pera

ture

res

ista

nce

(°C

)

% Swell in IRM–903 Oil


NBR

A B C D E F G H J K

Class by Oil Swell

ABCDEF

H

Typ

e by

tes

t te

mpe

ratu

re

Figure 4.1 Graphical representation of NBR in the ASTM D2000 system

45

Properties of Hydrogenated Nitrile Rubber and Comparison with Other Elastomers

Following hydrogenation the degree of unsaturation in NBR is significantly reduced to give HNBR, as indicated in Chapter 1. This reduces the tendency for high temperature oxidation and gives HNBR improved thermal stability in air and other oxygen-containing environments. Due to the wide range of ACN contents used in HNBR (from 17 to 50 mole% ACN) and its higher upper temperature limit, it is not surprising that HNBR falls within the range CF, CH, DF, DH, CJ or DJ. In Figure 4.2 HNBR has been included with NBR in a graphical representation of the ASTM D2000 system.

0

50

100

150

200

250

300

050100150200

Tem

pera

ture

res

ista

nce

(°C

)

% Swell in IRM–903 Oil


NBR

A B C D E F G H J K

Class by Oil Swell

ABCDEF

H

Typ

e by

tes

t te

mpe

ratu

re

HNBR

Figure 4.2 Graphical representation of NBR and HNBR in the ASTM D2000 system

Finally, Figure 4.3 places HNBR within the ASTM D2000 system in the perspective of other speciality elastomers.

46


0

50

100

150

200

250

300

050100150200

Tem

per

atur

e R

esis

tan

ce (

°C)



NRCR

AEM

VMQ FVMQ

FKM

NBR

ACM

A B E F G H J K

Class by Oil Swell

EPDM

A

BCD

EF

H

HNBR

Note : NR = natural rubber, EPDM = ethylene–propylene diene rubber, CR = chloroprene rubber,

VMQ = vinyl–methyl silicone, ACM = acrylic, AEM = ethylene–acrylic, FVMQ = fluorinated vinyl–

methyl silicone, FKM = fluorocarbon

Figure 4.3 Graphical representation of a variety of elastomers within the ASTM D2000 system

4.2 Specific Comparison of HNBR and NBR

The effect of catalytic hydrogenation in converting NBR to HNBR is interesting, and previously unpublished studies by the author may amplify this comparison. A simple, basic recipe was used to compare NBR and HNBR filled with N762 carbon black at 50 parts per hundred rubber (phr). The recipe, vulcanisation conditions and test results are summarised in Table 4.2.

47


Table 4.2 Comparison of NBR and HNBR

Polymer A: HNBR polymer; ACN: 36%; nominal unsaturation: 4%; viscosity ML(1+4) at

120 ºC: 85 Mooney units

Polymer B: NBR polymer, cold polymerised; ACN: 33%; viscosity ML(1+4) at 100 ºC: 50

Mooney units

Ingredient Recipe A (phr) Recipe B (phr)

Polymer A (HNBR) 100 –

Polymer B (NBR) – 100

N762 carbon black 50 50

Polymerised 1,2-dihydro-2,2,4-trimethyl quinoline 1 1

Stearic acid 1 1

Dicumyl peroxide, 40% on inert mineral carrier 7.5 3.5

Triallyl isocyanurate, 72% on inert mineral carrier 4.5 –

Zinc oxide 5 5

Note: phr = parts per hundred rubber polymer

Properties compared Results: Recipe A (NBR)

Results: Recipe B (HNBR)

Shore A hardness, ASTM D2240 74 72

Tensile strength (mPa), ASTM D412, die C 15.8 16.5

Ultimate elongation (%), ASTM D412, die C 145 241

Compression set, ASTM D395, Method B: piled discs, 22 h at 150 ºC 26.5 9.8



% Volume change, IRM–901 oil, 70 h at 150 ºC, ASTM D471 –0.2 +0.3

% Volume change, IRM–902 oil, 70 h at 150 ºC, ASTM D471 +3.8 +10.1

% Volume change, IRM–903 oil, 70 h at 150 ºC, ASTM D471 +14.8 +16.8

% Volume change, ASTM Reference Fuel A, 70 h at 23 ºC, ASTM

D471

+1.5 +4.8

% Volume change, ASTM Reference Fuel B, 70 h at 23 ºC, ASTM

D471

+25.6 +45.2

% Volume change, ASTM Reference Fuel C, 70 h at 23 ºC, ASTM

D471

+32.1 +54.8

48


Since the HNBR recipe contained a slightly higher molar percentage of ACN this is not an exact comparison, but it is interesting that HNBR showed noticeably greater fluid swelling than NBR. From these data we can see that the hydrogenation of NBR to form HNBR results in:

• Betteroxidativethermalstability.

• Increasedswellinginhydrocarbons.

Since both NBR and HNBR are now available with a wide range of molar percentages of ACN, the author conducted an extended comparison using the recipes shown in Table 4.3. The results in terms of swelling after immersion in ASTM Reference Fuel C for 70 hours at 23 ºC are shown in Figure 4.4.

Table 4.3: Recipes used for comparison of swelling in ASTM Reference Fuel C, HNBR versus NBR

Ingredient Recipe A (phr)

Recipe B (phr)

HNBR containing required mole% ACN 100 –

NBR containing required mole% ACN – 100

N762 carbon black 50 50

Polymerised 1,2-dihydro-2,2,4-trimethyl quinoline antioxidant

1 1

Stearic acid 1 1

Zinc oxide (French process) 5 5

Dicumyl peroxide, 40% on inert mineral carrier 7.5 3.5

Triallyl isocyanurate, 72% on inert mineral carrier 4.5 –

Test slabs vulcanised 5 min at 177 ºC in each case

Linear regressions are provided in Figure 4.4 since this may assist readers in designing further recipes to give resistance to hydrocarbon fuels or other types of fluid. Detailed information on the formulation of HNBR compounds is given in Chapter 6.

49


Mole % ACN

90

80

70

60

50

40

30

20

10

0

0 10 20 30 40 50 60–10

NBR

HNBR

Linear (HNBR)

Linear (NBR)

% V

olum

e ch

ange

NBRy=–1.9652x+97.386

R2=0.9995

HNBRy=–1.7361x+116.54

R2=0.9996

Figure 4.4 Swelling of NHBR and NBR in Reference Fuel C

4.3 Comparison of HNBR with HNB and Other Speciality Elastomers

As has been demonstrated earlier, HNBR combines the compounding flexibility and toughness of NBR with improved temperature and chemical resistance. This gives much better resistance to oxidation and sulfur attack at higher temperatures, although at the cost of slightly greater swelling in hydrocarbons. A clear illustration of this is given in Figures 4.5 and 4.6 for engine oil and automatic transmission fluid, respectively. The loss of elongation after 1008 hours at 150 ºC is shown for two HNBR, ACM and AEM materials [2]. An estimate of the life of an elastomer is expressed as the time over which elongation is reduced to 50% of its original value. Figures 4.5 and 4.6 confirm that HNBR materials retain their useful life after long-term ageing. ACM and AEM materials show excellent long-term resistance to hydrocarbon fluids, but in general do not offer the tear and abrasion resistance of HNBR materials.

50


-60

-40

-20

0

20

40

35% ACN HNBR

25% ACN HNBR

ACM AEM% E

lon

gati

on

loss

Material

Comparison of oil-resistantelastomers, 10W30SG engine oil at 150 °C



Figure 4.5 Comparison of oil-resistant elastomers in loss of elongation in automotive engine oil at high temperature

-30-20-10

0102030

35% ACN HNBR

25% ACN HNBR

ACM AEM% E

long

atio

n lo

ss

Material

Comparison of oil-resistant

elastomers, Dexron III ATF at 150 °C

% Elongation loss, 504 h

% Elongation loss, 1008 h

Figure 4.6 Comparison of oil-resistant elastomers in loss of elongation in automatic transmission fluid (ATF)

51


An area in which HNBR excels is in its wear and abrasion resistance. This is shown graphically in Figure 4.7 by comparison with other oil-resistant speciality elastomers.

Akron-type abrasion resistance rating

0

50

100

150

200

45% ACNHNBR

NBR ECO FKM ACM

Material

Rel

ativ

e re

sist

ance

to

abra

sio

n

Figure 4.7 Comparison of oil-resistant elastomers in relative abrasion resistance

One of the more demanding areas for elastomeric products is in the petroleum exploration and drilling environment. As exploration and drilling become deeper increased temperatures are encountered, making NBR no longer viable for these applications. In addition, many new sources of crude petroleum contain significant levels of hydrogen sulfide, which is reactive to a number of traditional elastomer components. Moreover, exploration and drilling systems often require elastomeric components having maximum ability to withstand high pressure and abrasion.

Zeon Chemicals have tested a range of elastomers as inhibitors against aggressive corrosion in petroleum applications and in mixtures of carbon dioxide, methane and hydrogen sulfide [4]. The results for amine exposure are replotted in Table 4.4.

52


Table 4.4 Resistance of various elastomers for loss of elongation in reference oil with amine corrosion inhibitor

Rubber material % Elongation change

36% ACN HNBR, peroxide cured 9

45% ACN HNBR, peroxide cured 32

FEPM, peroxide cured 15

Type 1 FKM (66 mole% fluorine), bisphenol cured –40

Type 2 FKM (70 mole% fluorine), peroxide cured –37

NBR (41 mole% ACN), Semi-EV sulfur cured –26

Note: Test fluid: IRM–902 oil containing 1% NACE Amine B corrosion inhibitorTest immersion conditions: 168 h at 150 °C

It is seen from Table 4.4 that HNBR materials showed excellent retention of flexibility and elasticity and an increase in percentage elongation. FEPM materials had similar properties, but in contrast FKM and conventional NBR showed a noticeable decrease in elongation, indicating amine attack and hardening.

The results of ‘sour gas’ testing (natural gas containing 5% hydrogen sulfide) are shown in Figure 4.8. The inclusion of 36 mole% ACN in HNBR gave excellent resistance to sour gas compared to other oil-resistant elastomers. Table 4.4 and Figure 4.8 also confirm the relative resistance of HNBR in petroleum exploration and drilling environments. These properties, combined with their toughness and abrasion resistance, make HNBR materials particularly useful in petroleum exploration and drilling.

Finally, it is useful to compare the relative costs of various oil-resistant elastomers. These are summarised in Table 4.5. The pound–volume cost is the cost of the compound multiplied by the specific gravity, and gives a relative idea of the cost of the parts produced. These values are based on the published prices of raw materials in late 2010.

53


120

100

80

60

40

20

00 50 100 150 200

36%ACN HNBR, peroxidecured

NBR (41 mole%ACN),Semi-EVSulfur cured

Type 2 FKM (70 mole%Fluorine), peroxide cured

Type 1 FKM (66 mole%Fluorine), Bisphenol cured

FEPM, peroxide cured

45%ACN HNBR, peroxidecured

Ageing time (h)

% E

long

atio

n re

tain

ed

Figure 4.8 Comparison of various elastomers for resistance to sour gas at 23 ºC

Table 4.5 Cost comparison of oil-resistant elastomers

PolymerSpecific gravity of compound

Cost of compound

lb–volume cost

Upper operating temperature limit (ºC)

NBR 1.22 $1.00 $1.22 100

ACM 1.32 $2.30 $3.04 150

AEM 1.32 $2.40 $3.17 150

HNBR 1.22 $11.10 $13.54 150

FVMQ 1.53 $23.00 $35.19 200

FKM 1.86 $15.64 $29.09 250

To summarise, while HNBR materials are similar in performance to ACM materials, HNBR becomes the choice when toughness and abrasion resistance are required, along with resistance to oils and high temperatures.

54


References

1. American Society for Testing and Materials, D2000–08 in Rubber, 9.02, American Society of Testing and Materials, Washington, DC, 2009, p.1.

2. Zeon Chemicals LP, A Comparison of Oil Resistant Elastomers in Engine Oil and ATF (Z7.3.16), Louisville, KY, USA, 1999, p.1.

3. Zeon Chemicals LP, Zetpol Product Guide, Louisville, KY, USA, 1999.

4. Zeon Chemicals LP, Zetpol HNBR Elastomers: Technical Bulletin SA–13, Applications for Oil Field Drilling, Recovery and Processing, Louisville, KY, USA, 2001.

55

Introduction

Previous chapters have described the manufacture, chemistry, the grades available and basic properties of HNBR compared to other speciality elastomers. The present chapter provides examples of its applications, for which formulations are given in Chapter 7.

5.1 A Brief Recap

Figure 5.1 illustrates the location of HNBR within the ASTM D2000 classification system for rubbers. This provides the context for the applications for which HNBR is important and shows that it has good resistance to hydrocarbon fluids, similar to that of NBR but with improved high temperature performance.

Information has been presented in Chapter 4 demonstrating the excellent abrasion resistance and enhanced thermal stability of HNBR compared to other oil-resistant elastomers. This feature is important in understanding that in many applications in which NBR has historically been used, but for which higher application temperatures and the demand for longer component life now require it to be replaced with HNBR.

Finally, the hydrogenation of NBR to form HNBR virtually eliminates the carbon–carbon unsaturation which is the primary site for oxidation [1]. Oxidative attack is a feature of a number of high temperature applications, and resistance to oxidation or ozonation is critical in extending the life of flexible dynamic rubber components.

5 Typical Applications of Hydrogenated Nitrile Rubber

56


0

50

100

150

200

250

300

050100150200

Tem

per

atu

re r

esis

tan

ce (

°C)



NRCR

AEM

VMQ FVMQ

FKM

NBR

ACM

A B E F G H J K

Class by Oil Swell

EPDM

A

BCD

EF

H

HNBR

Note : NR = natural rubber, EPDM = ethylene–propylene diene rubber, CR = chloroprene rubber,

VMQ = vinyl–methyl silicone, ACM = acrylic, AEM = ethylene–acrylic, FVMQ = fluorinated vinyl–

methyl silicone, FKM = fluorocarbon

Figure 5.1 Graphical representation of a variety of elastomers within the ASTM D2000 system

5.2 Abrasion-resistant Belts and Conveyor Components

In Chapter 4 we discussed the relative abrasion resistance of HNBR compared to other elastomers. HNBR materials also have excellent flex life in belts and other conveying equipment, which results in the longer life of components. The removal of most of the unsaturation in NBR is the key to the excellent flex life of HNBR. Similarly, HNBR is much less susceptible to oxygen and ozone attack, and this also greatly extends its flex life.

Acrylate modification of HNBR (Chapter 2) further improves its toughness, abrasion resistance and flex life. The enhanced high temperature resistance of HNBR materials also helps to extend the life of components in which long-term exposure to high temperatures is important for improving the durability of elastomeric components.

57

Typical Applications of Hydrogenated Nitrile Rubber

This applies for example in automotive engine compartments, where elastomers such as NBR and CR do not have sufficient thermal and oxidative resistance to fulfil the target life of the components.

Typical belt and conveyor applications for HNBR and acrylate-modified HNBR vulcanisates include the following:

• Fabric-reinforcedbeltsforautomotivepoweraccessories:

– Timing belts.

– Drive belts for air conditioning compressors, alternators and hydraulic steering pumps.

– Belt-driven power take-off pumps.

• Conveyorbelting,eitherhomogeneousrubberorfabric-reinforced:

– Acrylate-modified HNBR for conveying abrasive powders, including minerals and grain.

• Rubberrollers:

– In paper printing, especially acrylate-modified HNBR.

– Rollers for conveyor systems handling abrasive minerals, grain or grain by-products.

5.3 Flexible Boots for Power Transmission Joints

As mentioned earlier, HNBR materials offer excellent abrasion resistance and flex life. In addition, the hydrogenation of NBR gives it enhanced thermal resistance. As can be seen from Figure 4.3, HNBR is also reasonably resistant to hydrocarbon fluids. This combination makes it a good choice for flexible boots and bellows for sealing power transmission joints, keeping dust, dirt and water out of the joint and retaining the hydrocarbon lubricant in place.

HNBR is relatively expensive, as shown in Chapter 4, but it may be the only choice where boots or bellows are exposed to temperatures at which elastomers such as thermoplastic esters, olefin/rubber blends and urethanes run the risk of partial melting. One could, for example, envisage a protective power take-off boot on an all-wheel-drive vehicle where the boot is close to the exhaust pipe or a catalytic converter. In such cases the temperature near the boot might easily reach 120 °C, which would

58


be too high for the thermoplastic elastomers traditionally used. HNBR materials offer a good balance of properties for high temperature protective boot and bellows applications, including:

• Drivetrainbootsinautomotiveandtruckapplications:

– Constant-velocity joints near exhaust headers.

– Axle joints close to exhaust pipes or catalytic converters.

• Steeringboots:

– Rack and pinion ends near exhaust headers.

– Input shafts close to the engine.

• Suspensioncomponents:

– Tie-rod end boots and strut end covers near exhaust systems.

5.4 Static Seals for Power Transmission Fluids

From Figure 4.3 we can see that the properties of NBR are supplemented in HNBR by good resistance to hydrocarbon fluids together with enhanced performance at high temperatures. Many of the static seals used in power transmission, such as in power-assisted steering, automatic transmission and differentials, have traditionally been based on NBR due to its excellent hydrocarbon fluid resistance and relatively low cost. NBR static seals are however no longer adequate now that component temperatures are higher, and longer warranty periods have increased the life expectancy required. Acrylic elastomers such as ACM or AEM may be more cost effective than HNBR, but acrylic elastomers do not provide adequate fluid pressure resistance, especially in hydraulically assisted steering and power transmission systems.

In addition, the fluids used in many power transmission applications have been upgraded by the inclusion of antioxidants, sludge dispersants and metal protective additives. The reactive organic sulfur or phosphorus compounds included as metal protective additives may themselves act as vulcanisation chemicals in unsaturated elastomers, and the saturation of HNBR makes it much more resistant to the hardening effect of such additives.

59


Examples of static seals used in power transmission include:

• O-rings and square cross-section seals for long-life hydraulic power steeringgaskets, pumps and reservoirs:

– Pump seals.

– Fluid tube fitting seals.

– Fluid reservoir seals.

– Rack and pinion fluid seals.

• O-rings,gasketsandcross-sectionsealsforautomaticandmanualtransmissionfluids, differential joints and pan seals.

5.5 Dynamic Fluid Seals for Power Transmission

Due to its resistance to hydrocarbon fluids and additives, combined with its enhanced thermal resistance and excellent abrasion resistance, HNBR is an ideal alternative to NBR for radial shaft seals in power transmission applications. In this application dust exclusion and abrasion resistance are vital, and HNBR has found use in the following power transmission applications:

• Inputandoutputradialshaftsinautomaticandmanualtransmissionsystems.

• Inputshaftsandreciprocatingracksinhydraulicpower-assistedrackandpinionsteering systems.

• Radialdifferentialshafts.

• Hydraulicpower-assistedsteeringpumps.

5.6 Hydraulic Fluid and Lubricant Hoses

Good resistance to hydrocarbon fluids, oxygen and ozone, combined with thermal stability and abrasion resistance make HNBR valuable for hoses containing hydrocarbon hydraulic fluids and lubricants, for which NBR is prone to fluid additive attack at high temperatures. HNBR can be used in the liner, the reinforcing layer or the cover for such hoses.

60


5.7 Seals for High-volume Air Conditioning

HNBR provides an excellent combination of thermal resistance and resistance to commercial refrigerants. Automotive systems include the older CFC–12/mineral oil lubricants or the newer HFC–134a/polyalkylene glycol systems. HNBR also has good resistance in seals for high-volume air conditioning (HVAC) in buildings, such as HFC–22/mineral oil lubricant systems, making it an excellent choice for these applications:

• Staticseals:

– O-rings.

– Square cross-section seals.

– Gaskets (homogeneous rubber or composite-reinforced).

• Dynamicseals:

– Radial shaft seals for rotation or reciprocation.

5.8 Seals used in Petroleum Exploration and Drilling

In common with other applications, a combination of thermal stability, resistance to aggressive chemicals and abrasion resistance make HNBR an excellent choice in petroleum handling, for example:

• Packerelementsandblow-outpreventers.

• O-rings:

– Packer element static seals.

– Tubing seals.

– Pump static seals.

– Drill bit seals.

• Pumpradialshaftseals.

5.9 Other applications

These include:

61


• Static seals for hydrocarbon fluid cooler systems (e.g., engine oil, or powertransmission fluids):

– O-rings.

– Square cross-section seals.

– Gaskets.

• Flexibleprotectivebellowsonmechanicalfaceseals.

• Enginefuelstaticseals(awayfromtheheatoftheengine):

– O-rings for fuel injector return lines.

References

1. R.W. Keller, Rubber Chemistry and Technology, 1985, 58, 3, 637.

62


63

6.1 Introduction

In this chapter the general guidelines for HNBR elastomer formulations are discussed, including those for vulcanisation, filler and plasticiser systems.

6.2 Vulcanisation Systems

In Chapters 1 and 2 the manufacture, chemistry and the available grades of HNBR polymers were discussed. Most commercial HNBR polymers contain a small degree of residual unsaturation in the polymer backbone. This provides the opportunity for vulcanisation, either using accelerated sulphur- or peroxide-based systems. The most reactive areas of unsaturation in NBR polymers naturally provide preferential sites for catalytic hydrogenation. It follows that any residual unsaturation in a HNBR polymer is likely to be of a less reactive nature. This means that in the case of either accelerated sulfur or peroxide vulcanisation a much higher level of crosslinking agent and accelerator will normally be required for HNBR than for NBR.

A general study of sulfur-cured systems has been conducted by Zeon Chemicals, and the results are summarised in Table 6.1 [1]. The Zeon study was carried out using a base polymer with approximately 9% residual unsaturation. From the available polymer grades listed in Chapter 2 it is seen that very low residual unsaturation may not be adequate for accelerated sulfur vulcanisation.

Zeon has also conducted an interesting study on the peroxide plus coagent vulcanisation of a polymer of medium ACN level and 4% residual unsaturation [2]. The results of this study are summarised in Table 6.2 and are plotted in Figures 6.1 to 6.4.

6 Formulating Guidelines for Hydrogenated Nitrile Rubber

64


Table 6.1 Sulfur cure study by Zeon: Recipes and properties

Ingredient Recipe

1 2 8 9 10 11

Zetpol 1020 100.0 100.0 100.0 100.0 100.0 100.0

Zinc oxide 5.0 5.0 5.0 5.0 5.0 5.0

Stearic acid 1.0 1.0 1.0 1.0 1.0 1.0

N550 carbon black 40.0 40.0 40.0 40.0 40.0 40.0

Permanox OD 2.0 2.0 2.0 2.0 2.0 2.0

Sulfur (325 mesh) 0.5 0.3

4,4′-Dithiomorpholine 0.50 0.50 2.0 2.0 1.0 1.0

Dipentamethylene thiuram

hexasulfide

0.7 0.7 0.7

Tetramethylthiuram disulfide 2.0 2.0 0.7 0.7 0.7 0.7

N-cyclohexyl-2-benzothiazole

sulfenamide

1.0

2-Mercaptobenzothiazole 0.5

Zinc dimethyl dithiocarbamate 0.5 0.5 0.5

Zinc dibutyl dithiocarbamate 0.5 0.5 0.5 0.5

Tellurium diethyl

dithiocarbamate

0.5

Properties after vulcanisation, 20 min at 160 ºC

Property Recipe from above

1 2 8 9 10 11

Shore A hardness 74 74 75 75 75 75

Tensile strength, MPa 23.8 24.4 23 23.1 23.1 22.5

Ultimate elongation, % 460 450 410 400 420 390

100% Modulus, MPa 4.1 4.4 4.6 4.8 4.3 4.9

300% Modulus, MPa 16.6 16.9 18.3 18.6 17.8 18.6

Compression set, ASTM D395 Method B, 25% compression

Recipe from above

1 2 8 9 10 11

70 h at 100 ºC 35 34 37 41 42 41

70 h at 120 ºC 54 46 46 56 57 57

65

Formulating Guidelines for Hydrogenated Nitrile Rubber

Table 6.2 Peroxide + coagent: A study by Zeon [2], recipes and properties

Ingredient phr in recipe

1 2 3 4 5 6

Zetpol 2010 100.0 100.0 100.0 100.0 100.0 100.0

Zinc oxide 5.0 5.0 5.0 5.0 5.0 5.0

Stearic acid 0.5 0.5 0.5 0.5 0.5 0.5

N330 carbon black 40.0 40.0 40.0 40.0 40.0 40.0

Naugard 445 1.5 1.5 1.5 1.5 1.5 1.5

zinc salt of 4- and 5-methyl mercaptobenzimidazole

1.5 1.5 1.5 1.5 1.5 1.5

α,α-bis(t-butylperoxy)diisopropylbenzene

5.0 5.0 5.0 5.0 7.5 7.5

N,N′-m-phenylene dimaleimide 0.0 2.5 5.0 7.5 0.0 2.5


7 8 9 10 11 12

Zetpol 2010 100.0 100.0 100.0 100.0 100.0 100.0

Zinc oxide 5.0 5.0 5.0 5.0 5.0 5.0

Stearic acid 0.5 0.5 0.5 0.5 0.5 0.5

N330 carbon black 40.0 40.0 40.0 40.0 40.0 40.0

Naugard 445 1.5 1.5 1.5 1.5 1.5 1.5

Zinc salt of 4- and 5- methylmercaptobenzimidazole

1.5 1.5 1.5 1.5 1.5 1.5

α,α-bis(t-butylperoxy)diisopropylbenzene

7.5 7.5 10.0 10.0 10.0 10.0

N,N′-m-phenylene dimaleimide 5.0 7.5 0.0 2.5 5.0 7.5

Properties after vulcanisation 15 min at 170 ºC

Property Recipe

1 2 3 4 5 6


Tensile strength, MPa 28.0 30.1 28.1 28.0 32.4 30.2


50% Modulus, MPa 1.6 2.0 2.5 3.3 1.7 2.2

66


100% Modulus, MPa 2.2 4.1 6.4 9.0 2.9 5.0

Compression set, ASTM D395 Method B, 70 h at 150 ºC

61 32 26 22 42 29

Property Recipe

7 8 9 10 11 12


Tensile strength, MPa 29.3 27.1 32.0 28.2 28.6 27.7


50% Modulus, MPa 2.8 3.8 1.8 2.5 3.2 4.4

100% Modulus, MPa 7.7 11.0 3.7 6.6 9.9 13.4


26 23 34 25 24 22

The response surface analysis of the data gives the following predictive equations for the properties as a function of the peroxide and coagent levels:

• ShoreAhardness=74.70+0.75P+4.0C+0.5P2 – 0.375C2+0.15CP, (6.1)

where P=phrα,α-bis(t-butylperoxy)diisopropylbenzene peroxide, and

C=phrN,N′-m-phenylene dimaleimide coagent.

R-squared,adjustedfordegreesoffreedom=97.74%.

• 100%modulus=1.774–0.052P – 0.034C+0.006P2

+0.0187C2+0.0232PC (6.2)


• %elongation=711.6–49.6P – 76.9C+1.0P2+3.73C2+2.96PC (6.3)


• %compressionset=84.1–6.18P – 12.3C+0.12P2+0.56C2+0.688PC (6.4)


67


78

76

74

72

Peroxide phr

Coa

gent

phr

1098765

7

6

5

4

3

2

1

0

Figure 6.1 Contour plot of Shore A hardness as a function of peroxide and coagent level

4.0

3.5

3.0

2.5

2.0

Peroxide phr

Coa

gent

phr

1098765

7

6

5

4

3

2

1

0

Figure 6.2 Contour plot of modulus at 100% elongation as a function of peroxide and coagent level

68


450400

350

300

250

200

Peroxide phr

Coa

gent

phr

1098765

7

6

5

4

3

2

1

0

Figure 6.3 Contour plot of ultimate elongation as a function of peroxide and coagent level

5045

40

35

30

25

1098765

7

6

5

4

3

2

1

0

Peroxide phr

Coa

gent

phr

Figure 6.4 Contour plot of compression set as a function of peroxide and coagent level

69


Oven post-cure processes are also effective with HNBR compounds, as illustrated in a further study conducted by Zeon (Table 6.3 and Figure 6.5) [3].

Table 6.3 Zeon study on the effect of post-cure: Recipes and properties [3]


1 2 3 4 5 6 7 8

Zetpol 1020 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Zinc oxide 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0

Stearic acid 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0

N770 carbon black 40.0 40.0 40.0 40.0 40.0 40.0 40.0 40.0

Dibutoxyethoxyethyl adipate 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0

Sulfur (325 mesh) 0.5 0.5 0.5 0.5

Tetramethyl thiuram disulfide 2.5 2.5 2.5 2.5

2-Mercapto benzothiazole 0.5 0.5 0.5 0.5

40% Dicumyl peroxide on an inert carrier

8.0 8.0 8.0 8.0

Conditions Recipe

1 2 3 4 5 6 7 8

Press cure conditions 30 min at 160 ºC

30 min at 160 ºC

30 min at 160 ºC

30 min at 160 ºC

30 min at 160 ºC

30 min at 160 ºC

30 min at 160 ºC

30 min at 160 ºC

Oven post-cure conditions, h at 150 ºC

0 2 4 8 0 2 4 8

Property Recipe

1 2 3 4 5 6 7 8

Shore A hardness 65 66 66 67 65 67 67 67

Tensile strength, Mpa 27.5 23.5 25.1 25.6 24.3 26.8 26.4 25.6

Ultimate elongation, % 510 460 480 460 450 440 430 410

100% Modulus, Mpa 2.4 2.7 2.7 2.7 2.4 2.6 2.6 2.9

200% Modulus, Mpa 6.4 7.7 7.7 8.1 7.8 8.4 8.7 10


50 36 31 24 27 15 14 13


75 67 60 50 35 25 24 23

70


0

10

20

30

40

50

60

0 2 4 6 8 10

Com

pres

sion

Set

(70

h a

t 12

0 °C

)

Post-cure time (h at 150 °C)

Sulfur

Peroxide

Figure 6.5 Effect of post-cure on compression set

6.3 Carbon Black Fillers

Carbon black is one of the most universally used fillers in elastomers, not only for producing a black coloration but also for reinforcing and extending them. Zeon has conducted an extensive study of the effect of various grades of carbon black on the properties of HNBR vulcanisates [4]. In addition, work carried out by the author also suggests that the reinforcing effect of carbon black can be predicted from its phr and its dibutyl phthalate (DBP) absorption number [5]. By combining the results of the Zeon study and those of the author, predictive equations can be derived from the response surface analyses. The test recipe used by Zeon is given in Table 6.4 [4].

Table 6.4 Zeon study of carbon black in HNBR, recipes

Test recipe ingredient phr

Zetpol 2010 100.0

Carbon black (various grades) 0.0–75.0

Zinc oxide, Kadox 911C 5.0

Stearic acid 0.5

TOTM, Plasthall 5.0

Naugard 445 1.5

Vanox ZMTI 1.0

Vul-Cup 40KE 8.0

71


The results for various carbon black levels and grades, along with nominal DBP absorption values, are illustrated in Figures 6.6 to 6.8.

1

125

0050

60

75

70

0

80

20 5040

60

Shor

e A

DBP

phr

Figure 6.6 Surface plot showing effect of carbon black on Shore A hardness

1

125

000

4

75

8

0

12

20 5040

60

10

0%

mod

ulus

DBP

phr

Figure 6.7 Surface plot showing effect of carbon black on modulus at 100% elongation

72


1

125

0050

100

750

150

20 5040

60

ML

(1+4

) at

10

0 °

C

DBP

phr

Figure 6.8 Surface plot showing the effect of carbon black on Mooney viscosity

The predictive equations for these properties as a function of the phr and DBP number of the carbon black are as follows:

• MV=Mooneyviscosity,ML(1+4)at100°C

=60.16–0.929p+0.746D+0.00986p2 – 0.000540D2+0.0120pD, (6.5)

where p=phrcarbonblackand

D=DBPnumberofcarbonblack.

R-squared,adjustedfordegreesoffreedom=89.66%

• H=ShoreAdurometerhardness

=34.36+0.264p+0.372D – 0.00071p2 – 0.00202D2+0.00207pD (6.6)


• M1=stressat100%elongation(100%modulus)inMPa

=–0.0804p +0.0593D+0.00149p2 – 0.000360D2+0.000808pD (6.7)


73


It is appreciated that these equations bear little resemblance to the theoretical equations for filler reinforcement, and they have no theoretical basis. In the real world, however, theoretical equations are rarely useful for predicting practical situations, and empirical relationships derived from actual studies such as these are often more valuable. From the standpoint of practical rubber compounding these equations are helpful for predicting the effect of carbon black reinforcement (high values of R-squared) and are therefore useful in formulating. In discussing the effects of fillers on viscosity and modulus (Chapter 3), the following theoretical equation was presented:

G1=G0(1+Af+Bf2) (6.8)

where G0 =shearmodulusofunfilledmaterial,

G1=shearmodulusoffilledmaterial,and

f=volumefractionoffillerinthemixture.

Equation (6.6) might suggest that the level of filler was the only important factor in reinforcement, and not the nature of the filler. If that were the case there would really be no requirement for the wide variety of carbon black grades available, each with its own particle size and structure.

6.4 Non-black Fillers

It is sometimes unacceptable for the finished rubber article to be black. An example of this might be when a black NBR component was being replaced by a HNBR component of higher temperature capability, giving longer life to the final assembly. If the NBR and HNBR rubber articles were both black it would be difficult for the customer to monitor the changeover from NBR to HNBR. It might therefore be desirable to make the finished HNBR rubber article a different colour, say green, to avoid mistakes in assembly and warranty situations. Mineral fillers are important for producing colours other than black in the finished rubber articles.

Here again Zeon have provided a comprehensive study of non-black fillers in a wide range of particle sizes and reinforcing ability [6]. The test recipe and results are summarised in Table 6.5.

74


Table 6.5 Zeon study of non-black fillers

Ingredient phr Relative particle

size

Shore A

100% modulus

Δ Shore A per phr

Δ Modulus at 100% per phr

None 0 50 1.1

Ultra-Pflex 90 small 65 1.6 0.167 0.0056

Roy-CalL 90 medium 65 1.8 0.167 0.0078

Zeeospheres 200 90 medium 66 2.4 0.178 0.0144

Imsil A–8 90 medium 68 3.4 0.200 0.0256

Nulok 321 90 medium 73 5.0 0.256 0.0433

Nulok 390 90 medium 73 7.0 0.256 0.0656

Pyrax B 90 large 74 7.5 0.267 0.0711

Mistron

Cyprubond

90 medium 76 8.1 0.289 0.0778

Translink 77 90 medium 77 9.2 0.300 0.0900

Celite 350 90 large 80 9.2 0.333 0.0900

Silene 732D 90 small 83 10.2 0.367 0.1011

Zeolex 23 40 medium 65 2.7 0.375 0.0400

HiSil 532EP 40 small 68 3.3 0.450 0.0550

Cab-O-Sil TS–720 40 small 71 2.5 0.525 0.0350

HiSil 233 40 small 73 2.9 0.575 0.0450

Ultrasil VN–3 SP 40 small 74 3.0 0.600 0.0475

Aerosil 300 40 small 76 2.8 0.650 0.0425

Cab-O-Sil M–7D 40 small 78 3.1 0.700 0.0500

Base formula without filler

Ingredient phr

Zetpol 2010 100.0

TOTM, Plasthall 5.0

Titanium dioxide 5.0

75


Zinc oxide 5.0

Stearic acid 0.5

Naugard 445 1.5

Vanox ZMTI 1.0

Carbowax 3350 2.0

A–172 Silane 1.0

Vul-Cup 40KE 8.0

The author has calculated the reinforcing effect of various fillers in terms of the change in hardness and 100% modulus per phr given by each. This information will be of use in selecting the appropriate mineral filler. A typical example might be in the production of a green article of 70 nominal Shore A hardness. A mineral filler with a relatively low reinforcing factor is needed, and a higher level of filler might be used to keep down the cost. We know from Table 6.5 that our base recipe without filler would give 50 Shore A hardness. By selecting a couple of mineral fillers from Table 6.5 the reinforcing factor from the table can be used as follows:

PhrNulok390=(70–50)/0.256=78phr

PhrHiSil233=(70–50)/0.450=44phr.

If we needed to maximise the filler loading to keep the cost down, then Nulok 390 would be a better choice than HiSil 233 as a non-black filler, assuming the other properties were satisfied.

6.5 Plasticisers

Since HNBR polymers are chemically related to NBR, the plasticisers commonly used with NBR compounds are in general equally suitable for HNBR compounds. In the case of HNBR, however, the polymer has a significantly higher temperature limit and this means that the more volatile plasticisers suitable for NBR may not be practical for HNBR. Zeon has conducted a study showing the effect of a variety of plasticisers used in a simple HNBR recipe [7]. Some of the data are given in Table 6.6, together with the softening factors calculated by the author for each plasticiser.

76


Table 6.6 Plasticisers for HNBR

Plasticiser at 15 phr in base recipe

Property No plasticiser

Dioctyl phthalate

Dibutoxy ethoxyethyl

adipate

Ether ester

Trioctyl trimellitate

Triisononyl trimellitate

Compound

Mooney

viscosity, ML1+4(100

ºC)

133 86 78 81 91 87

Oscillating

disc rheometer

at 170 ºC,

Vmin

11.4 6.5 7.1 7.4 8.0 8.3

Oscillating

disc rheometer

at 170 ºC,

Vmax

51.6 40.4 46.2 41.4 53.1 56.3

Oscillating

disc rheometer

at 170 ºC,

T95, min.

19.5 18.3 15.1 16.5 17.2 17.0

Shore A

hardness

76 70 69 69 70 69

Ultimate

elongation, %

310 360 450 430 380 370

100%

Modulus,

MPa

6.2 4.1 3.4 3.3 4.2 4.3

Weight loss

after 168 h

at 150 ºC air

ageing, %

–1.4 –9.7 –8.6 –6.1 –3.4 –1.7

Temperature

retraction per

ASTM D1329

77


TR-10, ºC –22.6 –28.7 –25.1 –25.0 –28.1 –27.9

TR-30, ºC –16.5 –20.6 –12.7 –12.5 –20.0 –19.6

TR-70, ºC –4.5 –8.5 –3.3 –3.8 –6.8 –6.8

Heat

resistance

weight loss,

168 h at 150

ºC

–1.4 –9.7 –8.6 –6.1 –3.4 –1.7

Softening Effects

Δ Shore A per

phr plasticiser

–0.4 –0.5 –0.5 –0.4 –0.5

Δ Mooney

viscosity per

phr plasticiser

–3.1 –3.7 –3.5 –2.8 –3.1

Δ 100%

Modulus per

phr plasticiser,

MPa

–0.140 –0.187 –0.193 –0.133 –0.127

Δ TR-10 per

phr plasticiser

(ºC)

–0.407 –0.167 –0.160 –0.367 –0.353

Base recipe ingredient

phr

Zetpol 2010 100

N550 carbon

black

50

α,α′-bis(t-

butylperoxy)

diisopropyl-

benzene,

40% on inert

carrier

6

78


Trioctyl trimellitate (TOTM) is one of the most common plasticisers used in HNBR formulations and the reasons for this are clear from Table 6.6, which shows that TOTM offers a good balance of raw compound viscosity reduction, reduction of low temperature properties as measured by TR–10, and excellent thermal stability in terms oflowweightlossat150°C.ThesofteningeffectlistedinTable 6.6 should help in selecting the correct plasticiser and quantity, as well as determining the amount of additional filler required to bring the hardness back to the required level.

6.6 Antioxidants and Antiozonants

While HNBR has much better stability than NBR to attack by oxygen or ozone, it still needs to be stabilised against oxygen and, depending on the application, attack by ozone. Studies by Zeon Chemicals [8] have shown that polymerised 1,2-dihydro-2,2,4-trimethyl quinoline (available as Agerite Resin D, Flectol H or Naugard Q) at levels between 1.5 and 3.0 phr are very effective for the protection of peroxide-vulcanised HNBR against oxygen. The addition of 0.5 phr of a protective wax to the polymerised 1,2-dihydro-2,2,4-trimethyl quinoline further gives excellent dynamic ozone resistance. These studies also demonstrate that an effective long-term antioxidant against air ageing is a combination of 4,4′-bis(α,α′-dimethylbenzyl) diphenylamine, available as Naugard 445, and a zinc salt blend of 4- or 5-methylmercaptobenzimidazole, under the trade name Vanox ZMTI, each at 1.5 phr. It is important to note that some of these systems may inhibit the peroxide curing mechanism and a higher proportion of coagent may therefore be needed for effective vulcanisation [8].

6.7 Other Ingredients

Tackifiers may be required for combining plies of HNBR compounds with other HNBR compounds or different polymer formulations for the construction of composites. Coumarone–indene resins, phenol–acetylene resins, aromatic hydrocarbon resins or styrene–vinyl toluene resins work well at levels up to15 phr to tackify the resulting HNBR compound.

In some instances, for example in O-rings and seals for valves, internal lubrication is desirable to prevent valve sticking or squeaking. Oleamides are by far the most efficient internal lubricants in HNBR for this purpose. Commercial varieties of such oleamides include Armoslip CP and Kenamide E.

79


6.8 Blends with Ethylene–Propylene Diene Rubber

NBR can be blended with ethylene–propylene diene rubber (EPDM) to improve its heat and ozone resistance, and this technique similarly works with HNBR. As might be expected, EPDM improves low temperature flexibility and weather resistance, with some sacrifice in resistance to swelling in hydrocarbon fluids. Generally, 10–20 phr of EPDM elastomer, with medium to high third monomer content to assist vulcanisation, will give improvements in low temperature flexibility and weather resistance without loss of hydrocarbon fluid resistance. The ethylene segments in HNBR resulting from the hydrogenation process assist blending with EPDM.

6.9 Examples

6.9.1 Black HNBR Compound for a Reciprocating Lip Seal on a Power Steering Rack End

In this example we know from our previous experience with NBR seals that for this application we need a target of 80–85 Shore A hardness. We also require good wear resistance, and previous experience suggests that N550 carbon black will provide a good balance of properties and processing performance. The seal has a rubber-covered outer diameter (OD), which demands good high temperature compression set resistance to prevent the seal from spinning in the bore after extended use. The customer also requires the best low temperature performance available. The use of plasticisers is not permissible since the finished seal cannot be allowed to shrink due to plasticiser leaching into the power steering hydraulic fluid. With this knowledge in hand we are able to design a starting compound as follows:

• Polymer: Zetpol 4310 EP (17mole% acrylonitrile), to give the best lowtemperature flexibility, selecting the low-viscosity grade for injection moulding.

• Curesystem:Peroxidepluscoagentsystemforcompressionsetresistance:10.0phr α,α-bis(t-butylperoxy)diisopropylbenzene peroxide with 7.50 phr N,N′-m-phenylene dimaleimide.

• Carbonblack:N550(basedonpastexperiencewithNBRsealsinthisapplication).Shore A hardness H is provided by Equation (6.6) above:

H=34.36+0.264p+0.372D – 0.00071p2 – 0.00202D2+0.00207pD.

The value of D for N550 carbon black is 121, so we can solve for phr knowing

80


that our target for H is 82. We can also use a surface or contour plot such as that in Figure 6.6, giving 75.00 phr for N550.

• 1.5phr1,2-dihydro-2,2,4-trimethylquinolineantioxidant.

• PastexperiencewithHNBRcompoundshasshownthat5.00phrFrenchprocesszinc oxide and 1.0 phr stearic acid provide excellent cure activation and stability.

This gives the starting recipe in Table 6.7

Table 6.7 Possible recipe for reciprocating lip seal

Ingredient phr Comments

Zetpol 4310 EP 100.0 Very good low temperature

flexibility polymer

French process zinc oxide 5.0 Cure activator

Stearic acid 1.0 Cure activator, mould lubricant

1,2-Dihydro-2,2,4-trimethyl quinoline

antioxidant

1.5 Effective antioxidant

N550 Carbon black 75.0 Carbon black filler

α,α-Bis(t-butylperoxy)-

diisopropylbenzene peroxide

10.0 Efficient crosslinking, low

compression set

N,N′-m-Phenylene dimaleimide coagent 7.5 Efficient crosslinking, low

compression set

6.9.2 Light Brown HNBR Compound for a Hydraulic System O-ring

In this example the customer needs to replace an NBR O-ring, taking advantage of the high temperature resistance of HNBR. The NBR O-ring is black, and the customer needs a different colour for the HNBR replacement to be able to trace warranty returns and avoid errors at the component service locations – the instructions might be no more than to replace the black O-ring with the light brown O-ring. Since we are replacing a NBR compound of medium ACN a HNBR of medium ACN content is suitable, and low temperature flexibility is not required in this application. On the other hand, since the article is an O-ring compression set is a primary functional

81


requirement. We also need to keep the cost of the compound as low as possible to lessen the impact of the change from NBR to HNBR. We wish to select a mineral filler capable of high loadings, and will therefore use 10 phr of TOTM plasticiser to allow a higher filler content without seriously sacrificing high temperature compression set. The target Shore A durometer reading for this compound is 75.

• Polymer:Therban3446(34mole%ACN,4%residualunsaturationforhighcrosslink density and good compression set resistance.

• Heat-stablepigmentarysystem:1.00phrN990carbonblackplus4.00phrrediron oxide powder.

• Filler:Zeeospheres200toallowhighfillerloading:

We require 25 points Shore A from the polymer and a further 4 points Shore A from the filler to offset the effect of the plasticiser:

Totalfillerloading=(25+4)/0.178=163phr(Table 6.5 and Table 6.6).

The resulting starting recipe is given in Table 6.8.

Table 6.8 Possible compound for light brown O-ring for a hydraulic system

Ingredient phr Comments

Therban 3446 100

French process zinc oxide 5 Cure activator

Stearic acid 1 Cure activator, mould lubricant

1,2-dihydro-2,2,4-trimethyl quinoline

antioxidant

1.5 Effective antioxidant

N990 carbon black 1 Pigment system

Red iron oxide powder 4 Pigment system

Zeeospheres 200 163 Filler

Trioctyl trimellitate plasticiser 10 Plasticiser

α,α-bis(t-butylperoxy)

diisopropylbenzene peroxide


compression set

N,N′-m-phenylene dimaleimide

coagent


compression set

82


These examples will give the reader some idea how to use this information to design formulations. In the chapter following we will discuss specimen formulations for specific applications of HNBR. We shall also discuss the value of statistical experimental design and desirability functions for studying formulation variables and selecting the optimal formula for the application concerned.

References

1. ZeonChemicalsLP,Semi-EV and EV Curing Systems for Zetpol 1020 (Z5.1.2),Louisville,KY,USA,1999.

2. ZeonChemicalsLP,Study of Level and Ratio of Peroxide and Co-Agent in Zetpol 2010 (Z5.1),Louisville,KY,USA,1999,p.1.

3. ZeonChemicalsLP,Effects of Post Curing on Zetpol 1020 (Z5.1), Louisville,KY,USA,1999,p.1.

4. ZeonChemicalsLP,Zetpol HNBR Elastomers: Carbon Black Study (SA–26), Louisville,KY,USA,April2001,p.1.

5. R.W. Keller, Hydrogenated Nitrile Rubber, in Handbook of Specialty Elastomers,Ed.,R.A.Klingender,CRCPress,BocaRaton,FL,USA,2008.

6. ZeonChemicalsLP, Evaluation of Larger Particle Non-Black Fillers in Peroxide Cured Zetpol 2010 (Z5.3.3),Louisville,KY,USA,1999,p.1.

7. ZeonChemicalsLP,Low Temperature Properties of Zetpol and Influence of Plasticizers and Tackifiers (Z5.4),Louisville,KY,USA,1999,p.1.

8. ZeonChemicalsLP,Protection Systems for Zetpol (Z5.5),Louisville,KY,USA, 1999.

83

7.1 Introduction

This discussion opens with an overview of statistical experimental design and the use of the desirability function. If there was ever an area where the use of statistical experimental design techniques was desperately needed, it is in the formulation of speciality elastomers. The author has personally encountered a situation in which variant 61 of an experimental formula was the one that finally went into production after many months of trying one variable at a time. Rubber formulations are prone to interaction between the ingredients, and a full factorial experimental design can detect and quantify these interactions. A fictitious example will follow in which statistical experimental design is used together with the desirability function to facilitate the development of the process. Subsequent examples will involve recipes containing a number of variables, and the only possible way to accommodate these effectively is by statistical experimental design techniques together with desirability functions.

7.2 Statistical Experimental Design and Desirability Functions

For background information the reader is encouraged to refer to two excellent books on the subject of statistical experimental design [1, 2]. The first of these contains a relatively brief but very adequate description of the techniques and calculations and will give the reader a quick start in these techniques. The second book provides a more comprehensive background to statistical experimental design.

As mentioned above, the variables in a rubber formulation may interact, and statistical experimental design can detect and quantify these interactions, particularly when full factorial design and response surface designs are employed. These techniques are illustrated below using a theoretical example. Starting compounds and full factorial design are discussed and employed in other examples in this chapter.

In addition to statistical experimental design, the desirability function concept is also useful in evaluating rubber formulations [3, 4]. The basics of the desirability function are given by Equation (7.1).

7 Examples of Hydrogenated Nitrile Rubber Formulae for Specific Applications

84


D = (d1, d2, d3, ... dn)

d=1∏nn

, (7.1)

where D is a composite desirability number between 0 and 1.00 (0 = completely undesirable and 1.00 = completely desirable),

d1 is the desirability of property 1 in the range 0 to 1.00,

d2 is the desirability of property 2 in the range 0 to 1.00 and

dn is the desirability of property n in the range 0 to 1.00.

The individual desirabilities can have pass/fail, linear or curvilinear functions. An example of a pass/fail property might be ozone resistance (ASTM D1171) above 168 h, any cracking being considered undesirable (0) and total absence of cracking being completely desirable (1.00). The dn relationships are usually arrived at after full involvement of the customer. The remainder is a straightforward calculation.

As an example, let us imagine that we are measuring five properties (d1 to d5) and we have considered the desirability of each property. The properties and the individual desirability of each have been carefully considered and agreed with the customer, with full regard to the anticipated application of the product. Table 7.1 gives the individual desirability of two formulae and the composite desirability values for each formula. Note that a zero value for any individual desirability value (dn = 0) would make the overall desirability for the formula zero, since it is a product function.

Table 7.1 Example of composite desirability calculations

Formulation d1 d2 d3 d4 d5 D

Rubber Formula 1 0.1 0.8 0.9 0.3 0.2 0.336587

Rubber Formula 2 0.6 0.9 0.9 0.9 0 0

This results in Equation (7.2):

D = (d1 * d2 * d3 * d4 * d5)5 (7.2)

85

Examples of Hydrogenated Nitrile Rubber Formulae for Specific Applications

For properties 1–4 (d1 to d4), Formula 2 (Equation (7.2) is suitable. However, the desirability of property 5 (d5) is zero, rendering the composite desirability given by Formula 2 also zero.

Rubber formulations often need to balance a number of properties, making the desirability function approach ideal for providing a composite numerical rating for a number of formulations and pointing the way to the most appropriate overall formula. In Section 7.3 both statistical experimental design and desirability function are employed in a theoretical example of a HNBR formula for a specific customer application.

7.3 Theoretical Example: High-temperature HNBR Joint Boot

In this example we will imagine that we have been approached by a customer, a manufacturer of drive shaft components, who has a specific requirement. On an all-wheel drive vehicle the joint boot must exclude dirt and moisture as well as retaining the joint grease, but the joint happens to be close to the exhaust pipe and traditional materials such as CR rubber or thermoplastic elastomers are unable to withstand the temperature and satisfy long-term warranty requirements. The customer has estimated a typical operating temperature at the joint boot of 125 °C. A material is required that gives good flex life, resistance to the joint grease, and wear and abrasion resistance, all at this operating temperature. From the ASTM D2000 classification chart in Chapter 4, shown in Figure 7.1, it is clear that HNBR is the correct choice.

0

50

100

150

200

250

300

050100150200

Tem

pera

ture

res

ista

nce

(°C

)

% Swell in Irm-903 OilASTM D2000 Classification

NBR

A B C D E F G H J K

Class by Oil Swell

A

BC

DEF

H

Typ

e by

tes

t te

mpe

ratu

re

NRCR

EPDMAEM

VMQFVMQ

FKM

ACM

HNBR

Figure 7.1 Graphical presentation of ASTM D2000

86


Working with the customer, we have identified the individual desirabilities shown in Table 7.2.

Table 7.2 Definition of desirability of individual properties, theoretical joint boot compound

d1 Definition: d1 = 1.00 if weight loss after 1,000 hours at 125 °C in dry heat is 0–5%

d2 Definition: d2 = 1.00 if volume change in customer axle grease is > 0% after 1,008 hours at 125°C (ASTM D471); d2 = 0.00 if volume change is negative in the grease exposure test

d3 Definition: d3 = 0.00 if TR–10 temperature is above –40 °C; if TR–10 temperature is below –40 °C, then linear scale applies from 0.00 at –40 °C to 1.00 at –50 °C (A, below)

d4 Definition: d4 = 1.00 if dynamic flex life rating is < 3 after 1,000,000 cycles; d4 = 0.00 if dynamic flex life rating is ≥ 4 after 1,000,000 cycles

d5 Definition: d5 = 1.00 of static ozone quality retention rating is 100 after 168 h (ASTM D1171); d5 = 0.00 if quality retention rating is < 100

d6 Definition: d6 = 1.00 if plied disc compression set in customer axle grease is 0% after 1,008 h at 125°C; d6 = 0.00 if plied disc compression set is 100% or greater; linear between 0 and 100% compression set (B, below)

A

–52

d3

–47

TR-10, degrees C

0.8

0.6

0.4

0.2

0

1

–42

B

d6

2000

0.2

0.4

0.6

0.8

1

40 60 80 100

% Comp. set, 1,008 hrs. @ 125C in grease

87


A number of key desirabilities, d1, d2, d4 and d5, are pass/fail properties, for which our recipe must satisfy the minimum requirement. Two other properties, d3 and d6, are linear functions. Based on the information in Chapter 6, we have developed the basic recipe shown in Table 7.3, and the experimental design can be constructed around this recipe.

Table 7.3 Experimental design starting recipe for HNBR joint boot compound

Ingredient phr Requirements Experimental design optimisation

17 mol% HNBR polymer, 4% residual unsaturation

80–100

Good combination of freeze resistance and low–temperature flex properties

Vary level from 80 to 100 phr blending with EPDM polymer to achieve total 100 phr polymer

EPDM polymer – low ethylene, low viscosity, 4.5% third monomer unsaturation level

0–20 Low viscosity for injection mould flow; medium to high third monomer level for efficient vulcanisation

Vary level from 20 to 0 phr with blending of HNBR polymer to achieve total 100 phr polymer

Zinc oxide 2.0 Cure activator

Stearic acid 1.0 Cure activator

Polymerised 1,2-dihydro-2,2,4-trimethyl quinoline antioxidant

1.0 Optimal antioxidant and antiozonant balance

Paraffin wax 2.0 Optimal antioxidant and antiozonant balance

N762 Carbon black 55.0 Reasonable reinforcement while maintaining high elongation for flex resistance

Trioctyl trimellitate plasticiser

10.0 Reduction of compound viscosity for injection moulding; improvement of low temperature flexibility; good heat stability

Dicumyl peroxide, 40% on inert mineral carrier

3.0–5.0

Fast and efficient curing in injection moulding where cycle time is critical

Optimise level of peroxide for best balance of properties

88


The reader may be daunted at the prospect of making up and testing as many as nine formulae. On the other hand, guessing a formula and hoping it will perform well in all respects will almost certainly involve making up and evaluating many more formulations than this. The experiments are illustrated graphically in Figure 7.2.

3.003.203.403.603.804.004.204.404.604.805.00

00.0200.0100.0

phr EPDM in polymer blend

phr

pero

xide

Figure 7.2 Graphical depiction of HNBR joint boot compound experimental design

It should be noted that one of the key principles is to randomise the order of the experiments to eliminate bias, in this case in either the mixing or the testing. We therefore submit the formulae to the mixing and testing laboratories in a random order, using computer software to plot the results for each of the key properties involved, as illustrated in Figures 7.3 to 7.6. All formulae are found to pass the dynamic flex life requirement and they all give a 100% quality retention rating on accelerated ozone exposure testing.

89


–2.4

–2.6

–2.8

–3.0

–3.2

–3.4

phr EPDM

phr

pero

xide

20151050

5.0

4.5

4.0

3.5

3.0

Figure 7.3 Experimental design of HNBR joint boot compound: Surface plot showing the effect of peroxide and EDPM on weight loss at 125 ºC

1.0

0.50.5

0.0

–0.5

–1.0

–1.5

–2.0

phr EPDM

phr

pero

xide

20151050

5.0

4.5

4.0

3.5

3.0

Figure 7.4 Experimental design of HNBR joint boot compound: Contour plot showing the effect of peroxide and EDPM on grease volume at 125 ºC

90


–42.0

–43.5

–45.0

–46.5 –48.0

phr EPDM

phr

pero

xide

20151050

5.0

4.5

4.0

3.5

3.0

Figure 7.5 Experimental design of HNBR joint boot compound: Results for low temperature retraction, TR–10

80

70

60

50

phr EPDM

phr

pero

xide

20151050

5.0

4.5

4.0

3.5

3.0

Figure 7.6 Experimental design of HNBR joint boot compound: Effect of peroxide and EDPM on compression set (Cset)

91


When we analyse the nine formulae in terms of our desirability function and calculate a composite desirability value for each formula, this gives the results summarised in Table 7.4.

Table 7.4 Experimental results for HNBR flex boot experimental designFactors Responses

phr EPDM phr peroxide d1 d2 d3 d4 d5 d6 D

0.00 3.00 1.00 0.00 0.05 1.00 1.00 0.15 0.00

0.00 3.00 1.00 0.00 0.11 1.00 1.00 0.14 0.00

0.00 3.00 1.00 0.00 0.08 1.00 1.00 0.13 0.00

10.00 3.00 1.00 1.00 0.35 1.00 1.00 0.32 0.69

10.00 3.00 1.00 1.00 0.30 1.00 1.00 0.32 0.68

10.00 3.00 1.00 1.00 0.32 1.00 1.00 0.31 0.68

20.00 3.00 1.00 1.00 0.55 1.00 1.00 0.45 0.79

20.00 3.00 1.00 1.00 0.56 1.00 1.00 0.44 0.79

20.00 3.00 1.00 1.00 0.61 1.00 1.00 0.45 0.81

0.00 4.00 1.00 0.00 0.12 1.00 1.00 0.25 0.00

0.00 4.00 1.00 0.00 0.15 1.00 1.00 0.24 0.00

0.00 4.00 1.00 0.00 0.16 1.00 1.00 0.24 0.00

10.00 4.00 1.00 1.00 0.46 1.00 1.00 0.40 0.75

10.00 4.00 1.00 1.00 0.50 1.00 1.00 0.40 0.76

10.00 4.00 1.00 1.00 0.50 1.00 1.00 0.40 0.77

20.00 4.00 1.00 1.00 0.72 1.00 1.00 0.52 0.85

20.00 4.00 1.00 1.00 0.72 1.00 1.00 0.52 0.85

20.00 4.00 1.00 1.00 0.75 1.00 1.00 0.52 0.86

0.00 5.00 1.00 0.00 0.20 1.00 1.00 0.40 0.00

0.00 5.00 1.00 0.00 0.25 1.00 1.00 0.40 0.00

0.00 5.00 1.00 0.00 0.21 1.00 1.00 0.39 0.00

10.00 5.00 1.00 1.00 0.58 1.00 1.00 0.50 0.81

10.00 5.00 1.00 1.00 0.58 1.00 1.00 0.49 0.81

10.00 5.00 1.00 1.00 0.60 1.00 1.00 0.50 0.82

20.00 5.00 1.00 1.00 0.81 1.00 1.00 0.61 0.89

20.00 5.00 1.00 1.00 0.81 1.00 1.00 0.60 0.89

20.00 5.00 1.00 1.00 0.85 1.00 1.00 0.61 0.90

92


Table 7.4 shows that the formulae with no EPDM blended with the HNBR polymer has zero composite desirability, since these all show shrinkage on high temperature exposure to the customer’s axle grease, and shrinkage in this test is defined as a zero individual desirability. Finally, we can plot the composite desirability (D) values to find the ideal formula from the nine under examination.

0.00

0.25

0.50

010

0.75

4

5

320

com

posi

te D

phr peroxide

phr EPDM

Figure 7.7 Surface plot showing the effect of peroxide and EDPM on composite desirability (D) in the experimental design of a HNBR compound for a joint boot

We can see from Figure 7.7 that the most desirable formula contains at least 15 phr EPDM polymer, but Table 7.4 indicates that it should have 20 phr EPDM polymer and 5.0 phr of peroxide curative. Use of statistical experimental design and desirability function calculations has now provided us with an optimal formulation over the range studied and a numerical method of ranking the formulations based on the customer’s input of key properties.

93


The following sections illustrate further examples of formulae for HNBR in specific applications in terms of the variables studied and give suggestions for statistical experimental design.

7.4 Compound Example: High Temperature Oil Cooler O-ring Seal

In this example the customer is a diesel engine manufacturer who employs circulating water plus an ethylene glycol coolant in a small heat exchanger to lower the temperature and extend the life of the engine lubricating oil. The O-ring seal will be exposed to both engine oil and coolant, normally within the range 100–120 °C, but with a maximum operating temperature of 125 °C. In this application conventional NBR would have insufficient durability at 125 °C, EPDM have good coolant resistance but poor resistance to engine oils, and FKM are expensive and also have poor resistance to high temperature water-based systems. Based on these factors a material based on HNBR is clearly the optimal choice.

Being an O-ring in a mainly static seal, its critical properties are resistance to the fluids concerned and the application temperature, combined with excellent compression set resistance. O-rings universally require a Shore A hardness of 75, and this will be our target. Standard compression moulding will be used, and low unvulcanised viscosity is therefore not a requirement. A starting recipe and recommendations for optimisation are shown in Table 7.5. Based on the compounding parameters described earlier, this compound will give the best overall compression set and fluid resistance. The optimisation factors can then be combined in an experimental design with response surface techniques for final development of the compound.

Responses (properties) to be measured at 125–150 ºC include:

• Resistancetodieselengineoil.

• Resistancetowater/glycolcoolant.

• O-ringcompressionsetresistance.

94


Table 7.5 Oil cooler O-ring compound experimental design

Ingredient phr Logic/comments Experimental design

optimisation

HNBR polymer

with 34–36% ACN;

4–5.5% residual

unsaturation

100.0 Best overall balance of fluid resistance

and low temperature properties; the

higher levels of unsaturation give

the most efficient cure and balance

between temperature resistance and

compression set resistance

Zinc oxide 2.0 Activator

Stearic acid 1.0 Activator

Naugard 445

antioxidant

0.7 Best overall heat resistance stabiliser

package

Vanox ZMTI

antioxidant

0.7 Best overall heat resistance stabiliser

package

Carbon black

(VARIABLE)

50–75 N762 black as low level in design at

75 phr, N550 black as high level in

design at 50 phr

Evaluate two

carbon black

types for best

overall balance of

properties

Varox VC40KE 10.0 Peroxide for compression moulding at

180 ºC

Vanax MBM

(VARIABLE)

4.5–7.5 Coagent for increased efficiency

of cure and best compression set

resistance

Optimise coagent

levels for best

balance of

properties

Optimisation:

2-factor 3-level full

factorial design; 8

compounds and

experiments

95


7.5 Compound Example: Oil Field High-pressure Well Packer

In this case our customer is a manufacturer of oil well tools and requires a rubber packer element with excellent resistance to crude oil, hydrogen sulfide, aggressive amines and water. Fluorocarbon elastomers have outstanding high temperature resistance but their inability to withstand aggressive amines rules them out. Conventional NBR have the required oil, amine and water resistance, but cannot operate at 125 °C with excursions to 150 °C. When the packer element seals the well very high pressure differentials may be encountered, and we need to formulate a tough material of high hardness.

The packer element has a relatively simple cylindrical shape, and compression moulding is therefore suitable. Its critical properties are heat resistance and pressure, as indicated above, plus resistance to crude oil, hydrogen sulfide, steam and aggressive amines. HNBR show a smaller drop in modulus of elasticity and stiffness at increasing temperatures, and a HNBR packer retains its pressure resistance better than alternative materials at high temperature. A starting compound is shown in Table 7.6, with variables listed for further study and optimisation using experimental design, desirability functions and response surface methods.

Responses (properties) to be measured include:

• Heatresistance,125–150ºC.

• Crudeoilresistance.

• Hydrogensulfideresistance.

• Steamresistance.

• Aggressiveamineresistance.

• Pressureextrusionresistance.

96


Table 7.6 Oil well packer compound experimental design

Ingredient phr Logic/comments Experimental design optimisation

HNBR polymer with 34-36% ACN; less than 2% residual unsaturation

100.0 Best overall balance of heat and chemical resistance



Naugard 445 1.5 Best antioxidant package for long-term heat resistance

Vanox ZMTI 1.5 Best antioxidant package for long-term heat resistance

N330 or N550 carbon black (VARIABLE)

55.0 To obtain > 90 Shore A durometer for pressure resistance

Evaluate two different carbon blacks for best overall balance of properties

Varox VC40KE 10.0 High temperature peroxide for compression moulding at 180 ºC

Coagent: Vanax MBM at 7.5 phr or TAIC at 5.0 phr (VARIABLE)

VARIABLE High level of coagent for efficient vulcanisation

Optimise compound based on evaluation of two chemical types of coagent

Optimisation: 2-factor 2-level full factorial design with centre point; 5 compounds and experiments

97


7.6 Compound Example: High temperature Long-life Automotive Serpentine Belt

In this example our customer is a manufacturer of automotive engines who has encountered unacceptable warranty experience with chloroprene rubber serpentine belts. While chloroprene rubber has excellent flex and weather resistance, the reduced air flow through the engine compartment resulting from reduced drag on the car has resulted in the serpentine belt operating continuously in the range 100–125 °C. Previously, the serpentine belt operated at 70–90 °C and reduced life due to heat ageing did not give rise to warranty claims. The belt is fabric reinforced to prevent creep and elongation, so compression set is not critical. We require a compound of relatively high elongation that gives excellent dynamic and static flex life.

Critical properties are therefore heat resistance over the range 100–125 °C, static and dynamic ozone resistance, and dynamic flex life. The better retention of stiffness at high temperatures offered by HNBR compounds will ensure meshing of the belt at high temperatures. A basic recipe for this application is shown in Table 7.7, including compounding variables for experimental design and optimisation.

Responses (properties) to be measured include:

• Heatresistance,100–125ºC.

• Staticanddynamicozoneresistance.

98


Table 7.7 Serpentine belt compound: Experimental design


HNBR polymer with

34-36% ACN; 1-4%

residual unsaturation

(VARIABLE)

70-90 Best overall balance

between heat

resistance and efficient

vulcanisation

Vary levels of zinc

methacrylate modifier to

balance properties; total

polymer phr = 100

HNBR modified with

zinc methacrylate

30-10 Added tear and tensile

strength

Vary levels of zinc

methacrylate modifier to

balance properties; total

polymer phr = 100



Agerite Resin D 1.0 With wax, gives best

overall heat and ozone

resistance

Protective wax 2.0 With Agerite Resin D,

gives best overall heat

and ozone resistance

N762 carbon black

(VARIABLE)

10-20 Reinforcement without

loss of flex resistance

Vary level between 10 and

20 phr to obtain best balance

of properties

Varox VC40KE

(VARIABLE)

4-8 High temperature

vulcanisation system

Vary level between 4 and 8

phr to obtain best balance of

properties

Optimisation:

3-factor, 3-level

full factorial design

with centre point;

27 compounds and

experiments

99


7.7 Compound Example: Orange Water Pump Mechanical Seal Protective Bellows

Our customer is a manufacturer of mechanical face seals for automotive water pumps. He has been using conventional NBR for the protective bellows around the water pump seal but the engine compartment temperature has now risen to 125 °C with excursions up to 150 °C, and NBR is unable to meet the 100,000-mile warranty. Due to its combination of toughness, weather resistance, splash fluid resistance and cost HNBR is a good alternative. The customer requires the bellows to be orange in colour for the mechanics and assemblers to be able to distinguish them from the previous black NBR version. There is some flexing of the bellows, demanding flex fatigue resistance, but the flexing is relatively small.

The initial target hardness for the material is 65 Shore A, and other key properties include colour, heat resistance continuously at 125 °C with excursions to 150 °C, weather resistance, water and coolant resistance, low temperature flexibility for cold starting and dynamic flex resistance. For high volume production of this fairly complex shape injection moulding is the most suitable production process. A starting recipe is shown in Table 7.8, with additional variables for statistical experimental design and final compound optimisation. Organic pigments are not suitable for peroxide-cured materials since the pigments can act as free-radical scavengers and may interfere with the vulcanisation reaction. In addition, the oxidising effect of peroxides may destroy the colour.

Responses (properties) to be measured:

• Heatresistance,1008hat125ºC.


• Ozoneresistance.

• Coolantresistance,1008hat125ºC.

• Lowtemperatureflexibility.

• Dynamicflexfatigueresistance.

• Mouldingperformance,flowandcurerates.

100


Table 7.8 Compound for water pump bellows experimental design


21–25% Acrylonitrile

HNBR polymer, 1–4%

residual unsaturation

100.0 Best low temp. flexibility for

cold starts



Agerite Resin D 1.0 With wax, gives best overall

heat and ozone resistance

Protective wax 2.0 With Agerite Resin D, gives

best overall heat and ozone

resistance

Plasthall TOTM 10.0 Reduction of compound

viscosity for injection

moulding; improvement of low

temperature flex; good heat

stability

Nulok 321 50.0 Extending clay filler

HiSil 532EP (VARIABLE) 0–20 Reinforcing filler Level 1 will be this

grade

HiSil 233 (VARIABLE) 0–20 Reinforcing filler Level 2 will be this

grade

Vinyl-tris(ß-

methoxyethoxy) silane

(Silane A–172)

1.0 Coupling agent for filler

dispersion

Red iron oxide 4.0 Stable pigment

Varox DCP–40C

(VARIABLE)

4–8 Rapid vulcanisation on

injection moulding at 170 ºC

Three levels to

optimise properties

Optimisation:

2-factor and 3-level full

factorial design;

6 compounds and

experiments

101


7.8 Compound Example: High-temperature Differential Shaft Seal

In this application the customer requires a high temperature seal for the output shaft of the drive differential joint of an automobile. He has been using NBR-based shaft seals for many years, but relocation of the exhaust and catalytic converter system has now created an environment near the shaft seal in which the seal elastomer lip encounters continuous temperatures in the range 100–125 °C; this has caused premature hardening of the NBR seal, shortened life and warranty issues. The extreme pressure lubricant oils used in the differential contain sulfur and phosphorus compounds which tend to induce vulcanisation of the highly unsaturated NBR material. The saturated nature of HNBR should offer resistance to differential gear lubricants at the temperatures involved, and its toughness and abrasion resistance will provide long-term life in the abrasive environment caused by dust and mud. For the high volumes required transfer moulding is the preferred process, and in addition the HNBR material must bond well to the metal casing of the shaft seal. Optimal performance in this application dictates a Shore A hardness of 80. Other critical properties include long-term resistance to gear lubricants at 125 °C, moulding performance in terms of flow and cleanliness, bonding to the steel case, and shaft seal life in the simulated application extremes test programme. The starting recipe and variables for experimental design and optimisation are shown in Table 7.9.

Responses (properties) to be measured are:

• Resistancetogearlubricantat125ºCfor1008h.

• Mouldingevaluations:flow,cleanlinessandbondingtothesteelcase.

• Shaftseallifetestingatextremesofoperatingenvironment.

102


Table 7.9 Differential shaft seal compound experimental design


HNBR with 34-36% ACN;

1-4% residual unsaturation

100.0 Overall balance of oil

resistance and low

temperature properties



Naugard 445 0.7 With ZMTI, gives best

antioxidant performance

Vanox ZMTI 0.7 With Naugard 445,

gives best antioxidant

performance

Celite 350 or Mistron

Cyprubond (VARIABLE)

75.0 Mineral fillers show

best rotating shaft seal

performance

Evaluate two different

fillers for best overall

properties

N550 Carbon black 10.0 Black pigment

Varox VC40KE 5-10 Efficient vulcanisation Evaluate three levels of

peroxide for best overall

properties

Optimisation:

3-factor, 2,2- and 3- level

full factorial design;

12 compounds and

experiments

7.9 Compound Example: Short Steering System Hose

For this application the customer needs a short hose segment for power steering hydraulic fluid. Due to restricted space, the fabric-reinforced hose needs to be based on a compound that will resist heat, weathering and ozone on the outside of the hose and mineral-based hydraulic fluid on the inside. The hose will be plied around the

103


braided fabric reinforcement and the plies will be formed by extrusion. Due to the location of an exhaust header near the short section of hose the operating temperature of the hose will be normally at 125 °C, with an occasional excursions to 150 °C, for example when idling in heavy traffic. These requirements dictate the use of HNBR to provide temperature resistance, overall toughness, ozone resistance and retention of modulus at high temperatures. The critical properties of the compound and hose assembly include static and dynamic ozone resistance, flex fatigue resistance, smooth processing by extrusion, adhesion of the plies and the ability to withstand the customer’s pressure pulse test on the finished hose section. A compound giving 75 Shore A hardness offers the best balance of pressure resistance and retention of conformity. A starting recipe and variables for statistical experimental design optimisation are given in Table 7.10.

Table 7.10 Steering system hose compound experimental design


HNBR polymer with

34–36% ACN; 1% and 5%

unsaturation

100.0 Best overall balance

of fluid resistance

Low level at 1%

unsaturation, high level at

5% unsaturation

Zinc oxide 2.0

Stearic acid 1.0

N550 Carbon black 60.0

Plasthall TOTM 10.0 Viscosity reduction

without sacrifice of

heat ageing

Agerite Resin D 1.0 With protective wax,

gives best ozone

resistance

Protective Wax 2.0 With Agerite Resin

D gives best ozone

resistance

Varox DCP–40C 3.0–

5.0

Fast and efficient

curing

Variable level to optimise

properties

Optimisation:

2-factor, 2- and 3-level full

factorial design; 12 compounds

and experiments

104


Responses (properties) to be measured:

• Staticozoneresistance.

• Dynamicozoneresistance.

• Flex-fatigueresistance.

• Extrusionprocessing.

• High-temperaturepressurepulsetest,simulatingapplicationconditions.

7.10 Compound Example: Chemically Resistant, Low Hysteresis Roller for Paper Mills

The customer for this application is a paper mill with a requirement for long-running gloss calender rolls operating at temperatures slightly above 120 ºC. The chemicals employed dictate that the roll compound must have good resistance to water and mineral-based fluids. The current polyurethane rolls are not sufficiently long-lasting due to the combination of the chemicals and temperature. Low hysteresis is required to prevent excessive heat build-up in the rollers which will upset the process. The desired hardness of the roller compound is 60 Shore D. Because the rolls will be used with brightly coloured papers a black compound is not considered acceptable. The abrasion resistance, temperature resistance, chemical resistance and retention of modulus at high temperatures make zinc methacrylate-modified HNBR a suitable choice for this application. The rolls will be built from smooth calendered stock moulded on steel cores and they will be ground to size after moulding. Critical properties for the roll compound are low hysteresis, temperature resistance to 120 ºC, resistance to smooth calendering and to the application chemicals. A starting compound for experimental design optimisation is shown in Table 7.11.

Responses (properties) to be measured are:

• Dynamichysteresis.


105


Table 7.11 Low hysteresis paper mill roller compound experimental design


HNBR polymer with 34–36%

ACN; 5% unsaturation

50.0 Basic HNBR for

chemical resistance

Level 1 evaluation of

HNBR

HNBR polymer with 34–36%

ACN; 15% unsaturation

50.0 Higher unsaturation

gives less hysteresis

Level 2 evaluation of

HNBR

HNBR polymer modified with

zinc methacrylate

50.0 Adds toughness,

abrasion resistance and

good hardness

Silane-treated calcined clay 70.0 Mineral filler Level 1 filler evaluation

Mistron Cyprubond 70.0 Mineral filler Level 2 filler evaluation

Varox VC40KE 7.0

Vanax MBM 4.5–7.5 Efficient curing Variable levels to

optimise properties

Optimisation:

3-factor, 2,2- and 2-level full

factorial design; 8 compounds

and experiments

7.11 Concluding Comments

This chapter has reviewed the formulating principles for HNBR polymers covering a range of specific examples. It has also introduced the reader to statistical experimental design and desirability functions for optimising HNBR formulae to give the best overall balance of properties. Using these techniques it is possible to formulate compounds to give the best overall capability for meeting customer requirements.

106


References

1. L.B. Barrentine, in An Introduction to Design of Experiments: A Simplified Approach, American Society for Quality, Milwaukee, WI, USA, 1999.

2. G.E.P. Box, J.S. Hunter and W.G. Hunter, Statistics for experimenters: Design, discovery and innovation, John Wiley and Sons, Hoboken, NJ, USA, 2005.

3. E.C. Harrington, Industrial Quality Control, 1965, 21, 494.

4. G. Derringer and R. Suich, Journal of Quality Technology, 1980, 12, 214.

107

8.1 Introduction

In this final chapter examples are discussed of typical production processing issues that are likely to be encountered when processing speciality elastomers such as HNBR. When a problem arises there is an unfortunate tendency to focus on a single solution, and this is usually to change the compound. It must however be borne in mind that even a minor change in the compound may require its complete requalification, and particularly its effect on the nature of the parts produced. In any case simply changing the compound without examining the other process variables is unlikely to lead to a permanent solution, and the examples described below and the solutions offered will provide some pointers to solving the issues that may arise.

8.2 Example 1: Poor Flow, Knit Lines and Non-fills in Injection Moulding

In this example a vulcanised HNBR article is being manufactured by multi-cavity or multi-deck injection moulding. A sporadic problem has been encountered involving poor flow and possible scorch (premature curing), causing unacceptable knit lines and non-fills in the finished parts. A number of possible steps that may solve the problem are described, and we will see how statistical experimental design can help to determine the effect of each variable and possible interactions between them.

8.2.1 Possible Factor: Temperature of the Unvulcanised Rubber During Injection

At the beginning of the injection moulding process the first step is usually to preheat the unvulcanised rubber and feed it into the injection mechanism, which can be either a reciprocating screw or a screw and ram. Uniform preheating is achieved by use of a circulating liquid. A tendency the author has observed is the general belief that a lower injection feed temperature will result in less premature scorching and a reduced tendency for non-fills. However, if we bear in mind the nature of elastomeric flow we

8 Solving Hydrogenated Nitrile Rubber Processing Issues

108


can see that this is unlikely to happen. From the discussion in Chapter 3, the typical effect of temperature on viscosity is shown in Figure 8.1.

Shear rate (s–1)

Temperature effect

1000000

100000

10000

1000

100100

100 ºC

125 ºC

150 ºC

1000101

Vis

cosi

ty (

pois

e)

Figure 8.1 Effect of temperature on pseudoplastic polymer viscosity

If the injection process is operating at a shear rate of 1,000 s–1 and the injection feed mechanism is controlled at 100 °C, the resulting viscosity of the material under these conditions will be roughly 1,000 poise. If we raise the injection feed temperature to 125 °C the viscosity drops to 200 poise, or one-fifth of the viscosity at 100 °C. We obviously need to keep the injection feed mechanism temperature below the temperature at which the material will begin to vulcanise prematurely and scorch before it completely fills the mould cavity. A higher temperature in the feed zone of the injection moulding machine may solve the issue, or it could be just one of a number of factors to be considered in making the process more robust and avoiding sporadic flow and non-fills.

8.2.2 Possible Factor: Excessive Shear During Injection

After feeding and preheating the rubber prior to injection, the next step is its injection into the mould. A possible factor at this point is the shear rate during injection. In production processes time is always a pressing issue and the tendency is always to inject the rubber as rapidly as possible. By elevating the injection pressure the rubber

109

Solving Hydrogenated Nitrile Rubber Processing Issues

enters the mould more quickly and in turn the shear rate is increased. Again, from the discussions in Chapter 3, and noting that the shear will cause the HNBR compound to become hot, too high a shear rate may be a major factor in the flow and non-fill issues observed. A reduction in injection pressure will certainly reduce the flow rate of the rubber and increase injection time, but this may be a small price to pay in preventing the non-fill and flow issues caused by excessive heat build-up during injection.

–100%–100%–50%

0%50%

100%150%200%250%300%350%

–50% 0% 50% 100% 150%

% c

hang

e in

hea

t bu

ild-u

p

% change in parameter

Viscosity effect Shear rate effect

Figure 8.2 Typical heat build-up during injection

As discussed in Chapter 3, Equation (8.1) can be used to establish correlation and predict the effect of viscosity and shear rate increase:

DT = 0∫t (h/rc)g2 dt, (8.1)

where DT = temperature change in ºC s–1,

h = viscosity in Pa. s and

g = shear rate in s–1.

110


Using Equation (8.1) we can establish the following proportionalities to estimate the effect of viscosity and shear rate on heat build-up, as shown in Figure 8.2:

DT ∝ h, and DT ∝ g2 (8.2)

Reducing the viscosity, h, of the unvulcanised rubber would certainly help, and we may be able to accomplish this by increasing the temperature of the rubber feed, as discussed in Section 8.2.1, but it can be seen from Equation (8.2) that increasing the shear rate has a much greater effect on temperature build-up during injection. In the particular case being considered, as a factor in solving the problem we may decide to reduce the effective shear rate during injection by reducing the injection pressure from 150 to 100 bar. Again, this is just one factor involved in solving the overall problem, and the reader may begin to suspect that an experimental problem-solving design is on the horizon as we attempt to resolve the flow and non-fill issues:

Shear stress = (R . DP) / (2L), (8.3)

where DP = pressure drop,

R = radius of capillary and

L = length of capillary.

Shear rate = (4Q) / (pR3), (8.4)

where Q = volumetric flow rate,

R = radius of capillary and

happ = apparent viscosity

= (shear stress) / (shear rate)

= (DP . R4 . p) / (4Q . 2L)

111


Volumetric flow rate Q = (DP . R4 . p) / ( 2L . 4happ) (8.5)

8.2.3 Possible Factor: Sprue and Runner Design

This issue may be a factor, particularly if the sporadic flow and non-fill issues are isolated to the same cavities or decks of the injection mould. If we look at Equation (8.5) and plot the effect of changing viscosity (happ) of the material against the radius of the runner or sprue system, it is clear that the latter has a dramatic effect on the volume flowing through, since the flow volume, Q, is proportional to the fourth power of the capillary radius. These effects are illustrated in Figure 8.3.

% Change in factor

120%100%80%60%40%20%0%

–20%

–20% –10% 0% 10% 20% 30%

–40%–60%

Radius of runner Length of runner Viscosity of material

Net

% c

hang

e in

flo

w r

ate

Figure 8.3 Effect of moulding parameters on flow rate

To improve the overall flow and to eradicate the sporadic flow and non-fill issues, we could of course change the compound to reduce its viscosity, but it would seem more sensible to make small increases in the size of the sprues and runners in the injection mould. Again, the sprue and runner size may be just one of the factors to be studied in an experimental design to solve the sporadic flow and non-fill problem.

112


8.2.4 Solving the Injection Flow Problem using the Above Factors

From Chapter 7 the reader may already have concluded that a statistical experimental design would be appropriate using the factors discussed above. Listed in Table 8.1 and illustrated in Figure 8.4 is an experimental programme designed to resolve the flow and non-fill issue.

Table 8.1 Experimental design for sporadic flow and non-fill injection moulding issues

Mould temperature

(°C)

Injection time (s)

Injection pressure holding time (s)

Responses (dimensions)

parts/100 with non-fills

Comments

180 20 5 –1, +1, –1

180 20 15 –1, +1, +1

180 10 5 –1, –1, –1

180 10 15 –1, –1, +1

190 20 5 +1, +1, –1

190 20 15 +1, +1, +1

190 10 5 +1, –1, –1

190 10 15 +1, –1, +1

185 15 10 Centre point

Again, the order would need to be randomised and a number of runs are required to determine the significance of the response. Multiple runs would for example help us determine whether the difference between, say, 4 parts/100 and 6 parts/100 would be statistically different in terms of non-fills. Obviously, we would require one of these combinations to consistently produce zero non-fills in successive multiple runs.

113


Injection feed temperature, °C

Run

ner

size

, mm

521001

100

1505.0

6.0

Figure 8.4 Injection moulding flow: Graphical depiction of experimental design

8.3 Example 2: Sporadic Dimensional Shift in Injection Moulded Parts

This problem arises in many production moulding operations, especially in the case of parts with linear or diametric dimensions greater than about 7.5 cm. The author has experienced this on many occasions, the initial reaction being that something must be wrong with a particular batch or mix of rubber compound. If we look closely at the physics involved, however, the problem could have little or nothing to do with batch-to-batch variation in the rubber.

8.3.1 Possible Factor: Mould Temperature

The first step is to examine the mould or platen temperature, since this is by far the most critical factor in determining the dimensions of the finished part. Certainly, production personnel and the tooling source will claim that it must be something that has happened to the compound that is causing the change – but temperature is still the most likely factor. For example, a typical coefficient of linear expansion for an HNBR compound is 0.00015 mm/mm/ºC. If room temperature is 25 ºC and the moulding temperature is 180 ºC, the total shrinkage after moulding would be 0.023 mm/mm. If the mould were cut allowing for 0.023 mm/mm shrinkage at a moulding

114


temperature of 180 ºC on a 76 mm part and the moulding temperature drifted to 190 ºC, the shrinkage would increase to 0.025 mm/mm. While this might not seem a major change in a 76 mm part, the change would reduce its finished dimension from 76.00 mm to 75.84 mm. Similarly, on a 127.00 mm target dimension the finished part would be reduced in size from 127.00 mm to 126.62 mm.

This effect is amplified in the case of lightly filled elastomers with a high coefficient of linear expansion, such as a HNBR elastomer of lower target hardness. A carbon black-filled HNBR compound of this type might have a coefficient of linear expansion of 0.00025 mm/mm/ºC. If the mould was designed for a vulcanisation temperature of 180 ºC but the temperature was actually 190 ºC, this would reduce a nominally 76.20 mm dimension to 76.02 mm.

8.3.2 Possible Factor: Injection Pressures and Holding Pressures and Times

Step 2 should be to examine the flow conditions and any shifts in moulding. As mentioned in Chapter 3 and confirmed by Figure 8.4, shear rate has a major impact on the elastic recovery of the material. As shown in Figure 8.5, elastic recovery can be pronounced after demoulding, when the part acts in the same way as stretched chewing gum and retracts in the direction of flow. In addition, from Figure 8.2, the effect of increasing shear rate on heat build-up in the stock can be pronounced. A decrease in injection time from 20 s to 10 s would double the shear rate, and from Figure 8.2 this would in turn cause a three-fold increase in the build-up of heat in the compound during flow, ignoring any reduction in viscosity at higher shear rate. In effect this increases the moulding temperature of the material without affecting the mould or platen temperature – and the moulding temperature is of course the biggest factor in determining the dimensions of the finished part.

Perhaps the most surprising influence on the dimensions of an injection moulded part is variation in the injection pressure holding time. Injection pressure is applied, and then released after a specified time to allow the injector to recharge for the next shot. Releasing the injection pressure early can have a profound effect on the dimensions of the finished part. This has been frequently observed but has never been thoroughly studied or quantified. However, in the case of large diameter parts such as O-rings this factor can easily make the difference between dimensionally compliant and rejected parts. In the initial set-up of the injection process, injection pressure holding time must be considered for its effect on the finished parts. If the relationship between injection holding time and final dimensions is established during the initial production approval process, it is then more likely to be monitored and controlled during routine production.

115


4540353025201510

10 100 10001

50

% D

ie s

wel

l

Shear rate (s–1)

Figure 8.5 Elastic recovery (% die swell) as a function of shear rate

8.3.3 Solving the Injection Moulding Dimensional Problem using the Above Factors

Here again a statistical experimental design including the above factors will be helpful in solving the problem. Table 8.2 and Figure 8.6 summarise an experimental design appropriate for addressing the dimensional issue.

Table 8.2 Experimental design for rectifying dimensional defects

Mould temperature (°C)

Injection time (s)

Injection pressure hold

time (s)

Responses (dimensions)

Comments

180 20 5 –1, +1, –1

180 20 15 –1, +1, +1

180 10 5 –1, –1, –1

180 10 15 –1, –1, +1

190 20 5 +1, +1, –1

190 20 15 +1, +1, +1

190 10 5 +1, –1, –1

190 10 15 +1, –1, +1

185 15 10 Centre point

116


Mould temperature, °C

Pres

s ho

ld t

ime,

s

Inje

ctio

n ti

me,

s

091081

10

205

15

Figure 8.6 Injection dimensional shift: Graphical depiction of experimental design

Again, the order of the experiments would need to be randomised and multiple runs required to determine the relative significance of the factors on the dimensions of the moulded parts.

8.4 Example 3: Variation in Hose Tube Thickness during Extrusion

In this example, the product was a short non-reinforced HNBR hose. The hose was manufactured by extruding the tube, followed by cutting to the proper length and vulcanisation in an autoclave. A challenge in the manufacture of a short non-reinforced HNBR hose is variation in wall thickness during extrusion. As part of a variability reduction programme a team was formed to identify the variables affecting the thickness of the extruded tube, comprising representatives of the technical disciplines as well as production staff. Based on mutually exclusive and collectively exhaustive (MECE) discussions, the following variables were identified for further investigation:

117


• Slipagentusedinthecompound:

– Low level: paraffin wax at 1.00 phr.

– High level: fatty amide at 1.00 phr.

• Extruderbarreltemperature:

– Low level: 80 °C.

– High level: 100 °C.

• Extruderbreakerplateholesize:

– Low level: 7.00 mm hole size.

– High level: 8.00 mm hole size.

• Taperangleatentrancetothetubedie:

– Low level: 30º angle.

– High level: 45º angle.

• Extruderspeed:

– Low level: 30 rpm.

– High level: 40 rpm.

• Extruderdietemperature:

– Low level: 100 °C.

– High level: 120 °C.

A full factorial design would have entailed 64 experimental runs, which was clearly prohibitive. Initially we needed to determine which of the variables were the most significant, and interactions between the variables would be ignored. If we discovered two or three significant variables we could then go back and study these using a full factorial experimental design to resolve interactions. In this case a Plackett–Burman screening design would be appropriate, as illustrated in Table 8.3:

118


Table 8.3 Hose tube extrusion, 8-run Plackett–Burman screening design

Factors

Run A B C D E F G(dummy)

1 1.0 1.0 1.0 –1.0 1.0 –1.0 –1.0

2 –1.0 1.0 1.0 1.0 –1.0 1.0 –1.0

3 –1.0 –1.0 1.0 1.0 1.0 –1.0 1.0

4 1.0 –1.0 –1.0 1.0 1.0 1.0 –1.0

5 –1.0 1.0 –1.0 –1.0 1.0 1.0 1.0

6 1.0 –1.0 1.0 –1.0 –1.0 1.0 1.0

7 1.0 1.0 –1.0 1.0 –1.0 –1.0 1.0

8 –1.0 –1.0 –1.0 –1.0 –1.0 –1.0 –1.0

Significant variables (factors):A = Slip agent in compound: –1 = paraffin wax at 1.00; +1 = fatty amide at 1.00B = Extruder barrel temperature: –1 = 80 ºC; +1 = 100 ºCC = Extruder breaker plate hole size: –1 = 7.00 mm; +1 = 8.00 mmD = Taper angle at entrance to die: –1 = 30º; +1 = 45ºE = Extruder speed: –1 = 30 rpm; +1 = 40 rpmF = Extruder die temperature: –1 = 100 ºC; +1 = 120 ºC

As mentioned earlier, this is a screening design and interactions between the variables are ignored, but the design will determine which variables are significant, and the next step is a full factorial experimental design to identify and resolve any interactions.

8.5 Example 4: Variation in Dimensions of High-volume Compression Moulded O-ring

This is an actual case in which a multi-disciplinary approach was used with experimental design to solve a factory processing issue. A fairly large O-ring, nominal ID 100 mm, was being produced by compression moulding in large moulds for a high-volume application. The biggest issue was day-to-day and run-to-run variation in the ID of the finished O-rings. Rather than attempting to study one variable at a time, key personnel from technical and production areas were brought together for a MECE discussion of possible variables. This discussion was conducted leaving every possible factor open for discussion, eliminating any where the data supported

119


this action, and focusing on variables that might possibly be contributing to the dimensional variation, including the following:

• Compound variable: carbon black filler levels at the extremes ofweighingtolerance.

• Compoundvariable:vulcanisingagent(peroxide)attheextremesofweighingtolerance.

• Processvariable:rawrubberweightsattheextremesoftolerance.

• Processvariable:variationsinmouldsurfacetemperature.Theplatentemperaturewas set at 182 °C but the actual temperature at the mould surface was found to vary between 177 °C at the corners and 188 °C in the centre.

• Processvariable:mouldclosure.Duetotheplatenandmouldbeingmarginallyout of parallel and flatness, the mould could remain open by as much as 0.05 mm.

• Processvariable:ovenpost-curetemperature.Theproductunderwenta2hpost-cure with the oven set at 205 °C, but measurements within the oven revealed that the actual temperature could lie anywhere between 193 and 215 °C.

With six possible factors, even a fractional factorial design would obviously be prohibitive in terms of time and cost. To determine which of the factors might warrant further study an 8-run Plackett–Burman screening design was selected. As mentioned earlier, fractional factorials and screening designs tend to confound and obscure interactions between factors but can indicate which of these deserve further study. In the present case it was believed that Plackett–Burman screening design would identify the key factors for more detailed study using full factorial experimental design.

The rubber compound was carefully weighed and mixed to the extremes of the compound variable tolerances and the raw rubber preforms were segregated by weight and sorted to provide the extremes of performance weight. To avoid interrupting production, moulds were constructed in the prototype laboratory. Overall mould closure was controlled by the use of shims, zero opening requiring no shims and 0.05 mm opening 0.05 mm shims. To provide better resolution of the significant factors five replicate runs of each experimental design were carried out.

The 8-run Plackett–Burman experimental design, with results and analysis, is shown in Table 8.4 and Figure 8.7. The values given are ID measurements in mm.

120


Table 8.4 Screening design for dimensional resolution: 8-run Plackett–Burman screening design, compression moulded O-ring, 100 mm nominal ID

Run A B C D E F G (dummy)1 96.3 96.3 96.3 –96.3 96.3 –96.3 –96.3

2 –96.6 96.6 96.6 96.6 –96.6 96.6 –96.63 –92.3 –92.3 92.3 92.3 92.3 –92.3 92.34 92.7 –92.7 –92.7 92.7 92.7 92.7 –92.75 –96.6 96.6 –96.6 –96.6 96.6 96.6 96.66 99.9 –99.9 99.9 –99.9 –99.9 99.9 99.97 96.3 96.3 –96.3 96.3 –96.3 –96.3 96.38 –100.7 –100.7 –100.7 –100.7 –100.7 –100.7 –100.7Sum + 385.2 385.9 385.2 377.9 377.9 385.9 385.2Sum – –386.3 –385.6 –386.3 –393.6 –393.6 –385.6 –386.3Average E+ 96.3 96.5 96.3 94.5 94.5 96.5 96.3Average E– –96.6 –96.4 –96.6 –98.4 –98.4 –96.4 –96.6Effect –0.3 0.1 –0.3 –3.9 –3.9 0.1 –0.3Significant? no no no YES YES no noRESULTSRun y1 y2 y3 y4 y5 Y (avg.) Si2 Replicates (n)

1 96.2 96.3 96.5 96.1 96.5 96.3 0.0632 52 96.8 96.5 96.2 96.8 96.9 96.6 0.1968 53 92.1 92.0 91.9 92.8 92.8 92.3 0.3272 54 92.6 93.0 93.1 92.1 92.5 92.7 0.6008 55 96.2 96.8 97.1 96.5 96.5 96.6 0.4392 56 100.2 100.1 99.5 99.6 100.3 99.9 0.2728 57 96.3 96.0 95.8 96.8 96.5 96.3 0.5312 58 101.0 100.8 100.2 101.1 100.5 100.7 0.4512 5

Si2 = sum of squares = Σ(yi–y(avg.))2 Sum Si2 2.882

Avg. Si2 0.36

Se = √Si 0.6

Seff 0.19 = Se × √(4/N)

DF = 32 (n–1) × number of runs

N = total number of experimentsA = carbon black filler level at maximum of weigh-up toleranceB = vulcanising agent levels at maximum of weigh-up toleranceC = preform weight levels at extremes of tolerancesD = mould surface temperature: – = 175 ºC, + = 190 ºCE = mould closure: – = 0.00 mm, + = 0.05 mmF = Post-cure temperature: – = 190 ºC, + = 215 ºCMSFE: 0.387 t × S(eff)t (0.950, df) 2.04 (n–1) × (number of runs) = degrees of freedom

121


Mou

ld te

mpe

ratu

re

Mou

ld cl

osur

e

Carbo

n bl

ack

level

Prefo

rm w

eight

Curat

ive le

vel

Post-

cure

tem

pera

ture

4.5

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5

0.0

Absolute factor effects MSFE

Figure 8.7 Bar chart rating magnitude of compression moulding dimensional effects

The bold line in Figure 8.7 is the minimum significant factor effect (MSFE) based on a 95% confidence limit. The only two significant factors in this screening design were the surface temperature of the mould and its actual degree of closure, none of the other factors having a significant influence on the ID of the moulded O-rings. Mould temperature and closure were therefore selected for more detailed study using full factorial experimental design.

The outcome was the provision of multi-zone platen heating to narrow the temperature variation across the mould, and the introduction of mould and platen machining and maintenance schedules to reduce variation in mould closure. Had one variable at a time been studied this would have taken considerably longer – and other significant factors might have been missed. Following capital investment to reduce variations in mould surface temperature and mould closure, variation in ID of the moulded O-rings was virtually eliminated.

122


8.5 Conclusions

Techniques have been demonstrated which will help in resolving production issues. In cases where multiple variables are involved, emphasis has been placed on statistical experimental design techniques in order to determine which are significant. This is intended to guard against the knee-jerk response that a change in formulation is the way to solve a production processing issue, and also attempting to examine the variables one at a time and missing the possibility of interaction between them. Statistical experimental design is a technique that can help to reduce variation in manufacture. Unfortunately many process and product improvement projects do not in practice employ these techniques properly, or fail to plan the experimental design effectively. The involvement of a wide range of personnel and disciplines is critical in planning stage and developing proper experimental design.

123

ACM Acrylic elastomer (per D1418)

ACN Acrylonitrile

AEM Ethylene–acrylic elastomer

ASTM American Society for Testing and Materials

ATF Automatic transmission fluid

CFC Chlorofluorocarbon

CR Polychloroprene

D Composite desirability

DBP Dibutyl phthalate

d Desirability of an individual property

DSC Differential scanning calorimeter

ECO Epichlorohydrin–ethylene oxide copolymer

EDPM Ethylene–propylene–diene terpolymer

FEF Fast extrusion furnace black

FEPM Polytetrafluoroethylene–propylene copolymer

FKM Fluorinated hydrocarbon rubber

FVMQ Fluorinated vinyl–methyl silicone elastomer

HFC Hydrofluorocarbon

Abbreviations

124


HNBR Hydrogenated nitrile rubber

HVAC High-volume air conditioning

ID Internal diameter

L Length of capillary flow

Mc Critical molecular weight for entanglement

MECE Mutually exclusive and collectively exhaustive

Mn Number average molecular weight

Mw Weight average molecular weight

MSFE Minimum significant factor effect

NBR Nitrile rubber

NR Natural rubber

OD Outer diameter

phr Parts per hundred rubber polymer

Pt Total pressure drop in capillary

Q Volumetric flow through capillary

R Radius of flow in capillary

SAF Super-abrasion furnace black

SBR Styrene–butadiene rubber

Tg Glass transition temperature

TOTM Trioctyl trimellitate

VMQ Vinyl–methyl silicone elastomer

XHNBC Cross-carboxylated HNBR

γ Shear rate

125

Abbreviations

ΔP Pressure drop

Ø Volume fraction in a mixture

Ø2 Volume fraction of polymer

η Viscosity

τ Shear stress

126


127

2-Mercaptobenzothiazole 64100% Modulus 64, 66, 69, 72, 75200% Modulus 69α,α-bis(t-butylperoxy)diisopropylbenzene 65-66, 77, 79-81

A

ACM 17-20, 46, 49-51, 53, 56, 58, 85ACN 1, 3, 5-11, 13-14, 16, 18-19, 43, 45, 47-53, 63, 80-81, 94, 96, 98, 102-103,

105Acrylonitrile 1, 3, 9, 11, 33, 79, 100Activator 80-81, 87, 94, 96, 98, 100, 102AEM 17-20, 46, 49-50, 53, 56, 58, 85Aerosil 300 74Air conditioning 57, 60ASTM D2000 17, 43-46, 55-56, 85Automatic transmission 17-18, 49-50, 58

B

Butadiene 1-12, 14, 16, 18, 20, 22, 24, 26, 28-30, 32, 34, 36, 38, 40, 42, 44, 46, 48, 50, 52, 54, 56, 58, 60, 62, 64, 66, 68, 70, 72, 74, 76, 78, 80, 82, 84, 86, 88, 90, 92, 94, 96, 98, 100, 102, 104, 106, 108, 110, 112, 114, 116, 118, 120, 122

C

Cab-O-Sil M-7D 74Cab-O-Sil TS-720 74Capillary 37-38

length 110radius 111reheometer 37

Carboxylated 16Celite 350 74, 102Coagent 63, 65-68, 78-81, 94, 96

Index

128


Compression modulus 33-34Compression moulding 24, 27, 93-96, 118, 121Compression set 34, 47, 64, 66, 68-70, 79-81, 86, 90, 93-94, 97Contour plot 67-68, 80, 89Conveyor 56-57

belt 16, 57component 56

Copolymerisation 11

D

DBP number 32-34, 72Desirability function 83, 85, 91-92Dibutoxyethoxyethyl adipate 69Dicumyl peroxide 47-48, 69, 87Die swell 24-25, 115Dimensional shift 113, 116Dioctyl phthalate 76Dynamic ozone resistance 78, 97, 103-104

E

Einstein equation 32Elasticity 24, 29, 32, 34, 52, 95Emulsion polymerisation 1-3, 5-9EPDM 10, 17, 29, 31, 46, 56, 79, 85, 87-93Ether ester 76Ethylene 9-12, 17, 19, 29, 46, 56, 79, 87, 93Extrusion 34, 95, 103-104, 116, 118

F

Fatty amide 117-118Filler 20, 32, 34-35, 63, 73-75, 78, 80-81, 100, 105, 119-120FKM 17, 19-20, 36, 46, 51-53, 56, 85, 93Flexible boots 57Full factorial design 83, 94, 96, 98, 100, 102-103, 105, 117FVMQ 17, 19-20, 46, 53, 56, 85

G

Glass transition temperature 10, 26

129

Index

H

Heat build-up 27-28, 104, 109-110, 114HiSil 233 74-75, 100HiSil 532EP 74, 100HNBR 1, 7-14, 16-21, 43, 45-60, 63, 69-70, 73, 75-76, 78-82, 85, 87-105, 107,

109, 113-114, 116Hydrogenation 1, 7-9, 11, 45-46, 48, 55, 57, 63, 79Hysteresis 34-35, 104-105

I

Imsil A-8 74Initiator 1-3Injection moulding 13, 24-25, 27, 34, 36, 79, 87, 99-100, 107-108, 112-113, 115Isoprene 11-12

K

Knit lines 36, 107

L

Lanxess 13-16

M

Manual transmission 59Mayo-Lewis copolymer equation 3Minimum significant factor effect (MSFE) 121Mistron Cyprubond 105Mn 30Modulus 26, 32-34, 64-67, 69, 71-77, 95, 103-104Mole % ACN 49Mole fraction of monomer 5Molecular weight 2, 6, 29-31Molecular weight distribution 29-31Mooney viscosity 13, 31, 72, 76Mould flow 36-37, 87Mz 30

N

N110 32N330 65, 96N550 32, 64, 77, 79-80, 94, 96, 102-103

130


N762 46-48, 87, 94, 98N990 32, 81NBR 1-3, 5-11, 17, 19-20, 29, 32-33, 43-49, 51-53, 55-59, 63, 73, 75, 78-81, 85,

93, 95, 99, 101Newtonian fluid 21Non-black fillers 73-74, 82Non-fills 107-108, 112NR 17, 46, 56, 85Nulok 321 74, 100Nulok 390 74-75

O

O-rings 59-61, 78, 93, 114, 118, 121Oil cooler O-ring 93-94Oil well packer 96Oxidation 17, 45, 49, 55

-reduction 1Ozone resistance 78-79, 84, 97-100, 103-104

P

Packer elements 60Paper mill roller 105Paraffin wax 87, 117-118Peroxide 9, 43, 47-48, 52-53, 63, 65-70, 78-82, 87-92, 94, 96, 99, 102, 119Petroleum exploration 51-52, 60phr 32-36, 46-48, 65-70, 72, 74-81, 87-92, 94, 96, 98, 100, 102-103, 105, 117Plackett-Burman screening design 117-120Plasticiser 35-36, 63, 75-79, 81, 87Polymerisation 1-9, 11, 31Post-cure 69-70, 119-121Pressure drop 37-38, 110Process aids 36-37Propylene 9-12, 29, 46, 56, 79Pseudoplastic 22-23, 108Pyrax B 74

R

Radial shaft seals 59-60Reactivity ratios 4-9Recipe 33, 46-48, 64-66, 69-70, 73, 75-77, 80-81, 87, 93, 97, 99, 101, 103

131

Index

Residual unsaturation 9-10, 13, 63, 81, 98, 100, 102Response surface 34, 66, 70, 83, 93, 95Rollers 57, 104Roy-Cal L 74Runners 37-38, 111

S

Serpentine belt 97-98Shaft seals 59-60, 101Shear 13, 21-25, 27-28, 31-33, 35-38, 73, 108-110, 114-115Shear rate 21-25, 27-28, 31-33, 35-36, 38, 108-110, 114-115Shear stress 21, 24-25, 37-38, 110Silene 732D 74Sprues 37-38, 111Static ozone resistance 104Static seals 58-61Statistical experimental design 19, 82-83, 85, 92-93, 99, 103, 105, 107, 112, 115,

122Stearic acid 33, 47-48, 64-65, 69-70, 75, 80-81, 87, 94, 96, 98, 100, 102-103Steering system hose 102-103Sulfur 9, 13, 17, 49, 52, 58, 63-64, 69-70, 101

T

Tensile strength 47, 64-66, 69Tetramethyl thiuram disulfide 69Tg 10, 24, 26Therban® 15-16Time-temperature superposition 26Transfer moulding 24, 101Translink 77 74Trioctyl trimellitate 78, 81, 87

U

Ultimate elongation 47, 64-66, 68-69Ultra-Pflex 74Ultrasil VN-3 SP 74

V

Viscosity 2, 13-14, 16, 21-23, 26-33, 35-36, 38-39, 47, 72-73, 76-79, 87, 93, 100, 103, 108-111, 114

132


VMQ 17, 46, 56, 85Volume fraction 32, 35, 73Volumetric flow 37-38, 110-111

W

Williams-Landel-Ferry relationship 26

Z

Zeeospheres 200 74, 81Zeolex 23 74Zeon Chemicals 13, 15-16, 20, 51, 54, 63, 78, 82Zetpol® 15-16Ziegler-Natta-catalyst 29Zinc methacrylate 16, 98, 104-105Zinc oxide 33, 47-48, 64-65, 69-70, 75, 80-81, 87, 94, 96, 98, 100, 102-103

a b c d

a b c d

a b c d

a b c d

a b c d

a b c d

a

a b c d

a

Shawbury, Shrewsbury, Shropshire, SY4 4NR, UK Telephone: +44 (0)1939 250383 Fax: +44 (0)1939 251118 Web: www.polymer-books.com

Practical Guide to H

ydrogenated Nitrile B

utadiene Rubber TechnologyRobert Keller

High performance engineering plastics are used in an increasingly wide range of applications and environments. Their growth in importance is a response to the ever-increasing demand for more reliable, high performance components.

This book is the product of the author’s first-hand experience and understanding of high performance engineering plastics; specifically hydrogenated nitrile rubbers, which are progressively supplanting the simpler non-hydrogenated varieties thanks to their superior properties. A practical overview of their key properties and formulation principles is provided, based on the author’s own background and practical experience. Each chapter contains information on their product forms, properties, processing and applications, with the emphasis on materials and concepts shown to work in practice.

Readers will learn why hydrogenated nitrile rubbers are now the first choice for a range of demanding applications, how their characteristics arise and how their properties can be adapted. Many readers will welcome the practical nature of the examples given and the way in which problems can be resolved, for example by employing statistical experimental design. Not only is this concept valuable in overcoming production issues in a logical and cost-effective manner, it is also of help in communicating with raw material suppliers and those equipment manufacturers who have become dependent on nitrile rubbers.

Published by Smithers Rapra Technology Ltd, 2012


Robert Keller

published by smithers rapra technology ltd, 2012 rubber...

Documents