tribological characteristics of acrylonitrile … · 2015. 4. 10. · 361 int. j. mech. eng. &...

361

Int. J. Mech. Eng. & Rob. Res. 2013 Praveen R N et al., 2013

TRIBOLOGICAL CHARACTERISTICS OFACRYLONITRILE-BUTADIENE-STYRENE (ABS)

THERMOPLASTIC COMPOSITES

Praveen R N1*, Muruli D1 and Sandesh S Nayak1

*Corresponding Author: Praveen R N,[email protected]

The selection of material and methods for wear applications is an important part for bothtechnological advancement and manufacturing activities. Materials application, performanceand manufacturability are all key parts for the selection of wear resistance applications. Wearmaterials are used to reduce dimensional changes due to unwanted material removal, reducefrictional losses, to tailor the physical performance of a component and to provide physicallystable working surface. Acrylonitrile-Butadiene-Styrene (ABS) is a widely used thermoplastic.ABS polymers have a high toughness, satisfactory rigidity, and good resistance to heat, chemical,and environmental stress cracking. Molded articles with high dimensional stability and goodsurface quality can be produced by simple processing technique. It is essential to carryoutinvestigations to characterize ABS thermoplastic composites for its wear behavior for tribologicalapplications.

Keywords: ABS, Friction and wear, Design of Experiments, Central Composite Design, Analysisof Variance

INTRODUCTIONComposite materials have generallyconsiderable research interest during recenttimes. They are replacing many conventionalengineering materials due to their specificproperties of strength and stiffness. Theapplications are either weight critical orperformance critical as seen in automobile andaerospace industries. The concept of

ISSN 2278 – 0149 www.ijmerr.comVol. 2, No. 4, October 2013

© 2013 IJMERR. All Rights Reserved

Int. J. Mech. Eng. & Rob. Res. 2013

1 Department of Mechanical Engineering, Ekalavya Institute of Technology, Chamarajanagar, Karnataka, India.

combining two or more constituent materialsto form a composite and to make the best useof the more desirable properties of theconstituents has opened up several avenuesfor intelligent exploitation of compositematerials.

A composite material is made by combiningtwo or more materials to give a uniquecombination of properties. The above

Research Paper

362


definition is more general and can includemetals alloys, plastic co-polymers, minerals,and wood. Fiber-reinforced compositematerials differ from the above materials in thatthe constituent materials are different at themolecular level and are mechanicallyseparable. In bulk form, the constituentmaterials work together but remain in theiroriginal forms. The final properties ofcomposite materials are better than constituentmaterial properties (Composite MaterialHandbook).

Based on the nature of matrix, thecomposites can be Ceramic MatrixComposites (CMCs), Metal MatrixComposites (MMCs) or Polymer MatrixComposites (PMCs). Polymer MatrixComposites have matrices of thermoplasticor thermosetting polymers. A large number ofresin formulations, curing agents and fillersprovide an extensive range of possibleproperties. PMCs are increasingly beingemployed as structural materials in civil,defense, aerospace, automotive and industrialapplications where the weight reduction cancontribute toward saving in fuel and othermaterials (Composite Material Handbook).Many common thermoplastics are too stiff bythemselves, limiting their commercialapplications. ABS represents industrially themost important thermoplastic two-phasesystem with an amorphous structure. The ABSpolymers are based on three monomers:Acrylonitrile, Butadiene, and Styrene.Acrylonitrile-Butadiene-Styrene (ABS) is awidely used thermoplastic. In ABS, acrylonitrilecauses an improvement in chemical resistanceand weatherability, butadiene has thecharacter of rubber toughness, and styrene

offers glossiness and processability. Thecompositions of the various components canbe controlled to meet the requirements of avariety of applications. Acrylonitrile ButadieneStyrene is a common thermoplastic. It is acopolymer made by polymerizing styrene andacrylonitrile in the presence of polybutadiene.ABS polymers have a high toughness,satisfactory rigidity, and good resistance toheat, chemical, and environmental stresscracking. Molded articles with highdimensional stability and good surface qualitycan be produced by simple processingtechnique.

FRICTION AND WEARFriction and wear are terms that most peopleuse in their daily lives. Most people accept thecost of sport shoes wearing out after 4 monthsof use; people accept wear of roadways andflooring; people accept 30,000 miles as thelimiting use of an automobile before fan belts,brakes, and other components start to wearout. This guide reviews current friction andwears fundamentals and describes the benchtests that are most often used to study andsolve tribology problems. Tests are comparedand critiqued. Information is presented to helpthe reader select a test that he or she mightuse to address a tribology concern that theyare responsible for solving. The overallobjective of the guide is to lower the annualcost of wear, and unwanted friction throughappropriate tribo testing.

FrictionFriction is an energy dissipation process thataccompanies a body or substance in relativemotion in contact with another body orsubstance. When a solid body slides on

363


another solid, it takes energy to put the bodyin motion and keep it moving. Friction is theresistance to movement of one body overanother body. The word comes from Latin verbfricare, which means to rub. This energy isdissipated at the rubbing surfaces, usually inthe form of heat or deformation. When a solidbody is moved through a fluid, the energyrequired to produce motion is dissipated bymovement of the fluid as in waves or evenresults in heating.

A problem in dealing with friction is thatpeople often say that a material is low friction.A material cannot have a co-efficient of friction.It is a system effect. This co-efficient is uniqueto a system. It takes into consideration factorssuch as:

• Surface texture,

• Sliding speed,

• Contact geometry,

• Type of motion (reciprocity, continuous,etc.),

• Environment,

• Mating materials,

• Mechanical properties of mating materials,

• Separating films/particles and,

• Contaminants.

WearWear refers to the progressive removal ofmaterial from a surface and plastic deformationof material on a surface due to the mechanicalaction of the other surface (Kenneth, 2007).The necessity for relative motion between theoperating surfaces and initial mechanicalcontact between asperities leads an important

role between mechanical wear compared toother processes with similar outcomes(Kenneth, 2007).

The Organization for EconomicCooperation and Development (OECD)defines wear as “the progressive loss ofsubstance from the operating surface of a bodyoccurring as a result of relative motion at thesurface”.

Wear MechanismWear is a surface damage that results inremoval of materials from the surfaces whichare in relative motion. Surface damage in thiscontext is defined as topographical or microstructural changes, or both, in a surface layer.Material can be removed from a solid surfacein only three ways: by melting, by chemicaldissolution, or by the physical separation ofatoms from the surface. The last method canbe accomplished either by the one-timeapplication of a high strain or by cyclicstrainingat lower magnitudes. Perhapsbiggest challenge in solving wear problems inthat of anticipating the type(s) of wear to whichcomponent will be subjected (Williams, 2005).

In principle, a tribosystem may exhibitdamage of single type, but generally thepattern is a combination of two or more types.The types of damage that will in practiceinfluence a layer are:

• Surface damage without exchange ofmaterial: structural changes, plasticdeformation and surface cracking.

• Surface damage involving loss of material:shear fracture, extrusion, chip formulation,chip formation, tearing, brittle fracture,fatigue fracture, chemical dissolution anddiffusion.

364


• Surface damage involving gain of material:pickup of loose particles, material transferfrom counter surface and corrosion.

Friction and Wear TestsThe Tribometer uses a pin-on-disk system tomeasure wear. It consists of a pin on disc,loading panel and friction and wear monitor.The sliding wear of pure ABS and carbonpowder filled ABS are carried out with differentloads by varying time and sliding distances.To evaluate the performance of thesecomposites under dry sliding and lubricationconditions, wear tests will be carried out in apin-on-disc type friction and wear monitoringtest rig as per ASTM G 99-95a. The counterbody is usually a disc made of hardenedground steel. The specimen is held stationaryand disc is rotated while a normal force isapplied through a lever mechanism.

Standard MethodologyDesign of ExperimentsDesign of experiments can be defined as “Thesimultaneous evaluation of two or more factorsfor their ability to affect the resultant averageor the variability of the particular product orprocess characteristics”. To accomplish thisin an effective and statistically proper fashions,the level of the factors are varied in a strategicmanner, the result of particular testcombinations are observed, and the completeset of the results are analyzed to determinethe influential factors and preferred levels, andweather increase or decrease of those levelswill potentially lead to further improvement. Itis important to note that this is an iterativeprocess; the first round through DOE processwill many times leads to subsequent rounds ofexperimentation. The beginning round oftenreferred to as a screening experiment is usedto find the few important influential factors outof many possible factors involved with aproduct or process design (Douglas, 2010).

To use the statistical approach in designingand analyzing an experiments, it is necessaryfor everyone involved in the experiment to havea clear idea in advance of exactly what is tobe studied, how the data are to be collected,at least a qualitative understanding of howthese data are to be analyzed. An outline ofthe recommended procedure is shown below(Douglas, 2010).

• Recognition of and statement of theproblem.

• Choice of factors, levels, and ranges.

• Selection of the response variable.

• Choice of experimental design.

Figure 1: Set Up to Perform an AbrasiveWear Test on Pin-on-Disc

365


• Performing the experiment.

• Statistical analysis of the data.

• Conclusions and recommendations.

Central Composite Design (CCD)Design of experiment is a powerful analysistool for modeling and analyzing the influenceof control factors on performance output. Themost important stage in the design ofexperiment lies in the selection of the controlfactors. Therefore, a number of factors areincluded so that non-significant variables canbe identified at earliest opportunity. Twoparameters viz., normal load and time each attwo levels are considered in this study inaccordance with Central Composite Design(CCD) (Philip, 1995).

Central Composite Design (CCD) is oneof the methods of response surface design.CCD is very effective experimental techniquein studies involving large number of factors. Aset of experimental design that can look at kfactor an n observation with each factor at twolevels is called two level factorial designs,which can prove good and efficient when alinear relationship prevails between the factorsand response.

Figure 2 shows the geometric form of thedesign when it is rotated by 450. It can benoticed that the levels of each factor have beenincreased from three levels without increasingthe total number of Test combination (Tc) whichis still equal to nine. The design as representedin geometric form in Figure 3 is called CentralComposite Design (CCD) for it is made up ofa central point and optimal joining of the 22

factorial designs with one factor-at-a-timedesign.

The experimental design based on CCDallows investigation of factor up to fourth orderquartics relationship due to the existence offive levels of each factor. The real value of thedesign is the increased scope of theexperiment central composite design are easyto construct since they are based on two levelfactorials (2k) and one-factor-at-a-timetechniques. The general representation ofCCD is:

No. of test combinations (Tc) = 2k–p + 2*k + 1

Here, k = no. of factors and p = thefractionalization element

Figure 2: Geometric Form of CCD

Figure 3: 22 Factorial Designs with OneFactor-at-a-Time Design

366


Table 1 provides details of factors and theirrange considered in the present investigation.

Analysis of Variance (ANOVA)ANOVA is performed to evaluate the relativeimportance of the factors and the error variance(Madhav, 1989). The results of the ANOVA areshown in Table 4 below. The P-value for themodel and F statistic in the ANOVA indicatesthat there is a difference in the meanpercentage increase in wear between the twofactors. The P-value is the probability that thetest statistic will take on a value that is at leastas extreme as the observed value of thestatistic. Once the P-value is known, it is easyto determine how significant the data arewithout the data analyst formally imposing apreselected level of significance.

Statistically, there is a tool which providesa decision at some confidence level as towhether these estimates are significantlydifferent. This tool is called an F-test, named

Factors Range

Load, N X1 10 to 80

Time, Sec X2 500 to 1000

Table 1: Details of Factor and Their Range

The distance () of the star (*) points fromthe centre is determined using equation

= 2k/4 = (22)1/4 = 1.414

These factors and their levels in CCDexperimental plan are summarized on Table 2.

RESULTS AND ANALYSISResults and Analysis of ABS UnderDry Sliding ConditionFrom Table 3, CCD experimental plan and thewear test results of ABS thermoplastics areshown. Analyses are made using popularsoftware specifically used for design ofexperiment applications known as Minitab.Before any attempt is made to use this simplemodel as a predictor for the measure ofperformance, the Regression models,Analysis of variance (ANOVA), main effectsand the possible interactions between thecontrol factors must be considered. Thusfactorial design incorporates a simple meansof testing for the presence of the interactioneffects.

Load,N X1 10 20 45 70 80

Time,sec X2 500 573.2 750 926.8 1000

Table 2: Factors and Their Levelsin CCD Experimental Plan

LevelsFactors–1.414 –1 0 1 +1.414

Note: No. of test combinations (tc) = 2k-p + 2*k + 5tc = 22-0 + 2*2 + 5 = 13.

Table 3: CCD Experimental Planand the Wear Test Results of ABS Under

Dry Sliding Condition

1 –1 –1 20 573.2 0.0003 0.502

a +1 –1 70 573.2 0.0008 0.451

b –1 +1 20 926.8 0.0009 0.406

ab +1 +1 70 926.8 0.0014 0.441

–a –1.414 0 10 750 0.0002 0.497

+a +1.414 0 80 750 0.0019 0.447

–b 0 –1.414 45 500 0.0003 0.474

+b 0 +1.414 45 1000 0.0014 0.490

zero 0 0 45 750 0.0011 0.528

zero 0 0 45 750 0.0012 0.475

zero 0 0 45 750 0.0012 0.586

zero 0 0 45 750 0.0011 0.552

zero 0 0 45 750 0.0010 0.496

CodedVariables

NaturalVariables Responses

tcA B Load,

NTime,sec

Wear,g COF

367


after Sir Roland Fisher. Each combination ofconfidence, numerator degrees of freedomand denominator degrees of freedom has anF-ratio associated with it. F, v1, v2 is theformat for determining an explicit F-value(Philip, 1995). Using = 0.05, the critical valueof F for load and time is F0.05, 1, 9 = 5.12.Because 26.11>5.12 and 17.11>5.12, itconcludes that the factors load and timesignificantly affects the wear. The P-value isconsidered for determining the significantfactors. The values of P less than 0.050indicate model terms are significant. In thiscase factors X1 (load) and X2 (Time) aresignificant model terms, the AB (load-time)interaction value greater than 0.050 indicatesinsignificant model terms.

In addition to the basic analysis of variance,the program displays some additional usefulinformation. The quantity “R-squared” explainsabout 82.76% of the variability in wear. The“adjusted” R2 is a variation of the ordinary R2statistic that reflects the number of factors in

the model. The “predicted R-squared” of0.8276 is in reasonable agreement with the“adjusted R-squared” of 0.7702. Clearly, it musthave 0 < R2 < 1, with larger values being moredesirable (Douglas, 2010).

The normal probability plot of these effectsis shown in Figure 4. All of the effects that liealong the line are negligible, whereas the largeeffects are far from the line. The significanteffects that emerge from this analysis are themain effects A (load) and B (time) and theinteraction AB.

Table 4: ANOVA-General Linear Model

Figure 4: Normal Probability Plotof Standardized Effects

The effects of the factors are defined to bethe change in response produced by thechange in the level of factor. This is frequentlycalled a main effect because it refers to theprimary factors of interest in the experiment.

To assist practical interpretation of thisexperiment, Figure 5a presents plots of the twomain effects load and time and the AB (load-time) interactions shown in Figure 5b. Themain effect plots are just graphs of the marginalresponse averages at the levels of the twofactors. Notice that all two variables havepositive main effects; that is, increasing the

368


variable moves the average deviation from thefill target upwards.

In Figure 6a shows a plot of the plane ofwear generated by the various combinationsof load and time. This three dimensional graphis called a response surface plot. Figure 6bshows the contour lines of the constantresponse wear in the load and time plane.Notice that because the response surface isa plane, the contour plot contains parallelstraight lines.

Frequently, the initial estimates of theoptimum operating conditions for the systemwill be far from the actual optimum. In such

circumstances, the objective is to move rapidlyto the general vicinity of the optimum. It isusually assume that a first order model is anadequate approximation to the true surface ina small region of the wear. The method ofsteepest ascent is a procedure for movingsequentially in the direction of the maximumincrease in the response.

Results and Analysis of ABS UnderLubrication ConditionThe tests were carried out under lubricatingcondition by using SAE 10 W oil. Sufficientamount of oil was maintained during the testfor oil to remain continuously in contact with

Figure 5a: Main Effect Plot

Figure 5b: Interaction Plot

Figure 6a: Surface Plot

Figure 6b: Contour Plot

369


information. The quantity “R-squared” explainsabout 79.70% of the variability in wear. Clearly,it must have 0 < R2 < 1, with larger values beingmore desirable. The “adjusted” R2 is avariation of the ordinary R2 statistic that reflectsthe number of factors in the model. The“predicted R-Squared” of 0.7970 is inreasonable agreement with the “adjustedR-squared” of 0.7293.

The normal probability plot of these effectsis shown in Figure 7. All of the effects that liealong the line are negligible, whereas the largeeffects are far from the line. The importanteffects that emerge from this analysis are themain effects A (load) and B (time) and theinteraction AB.

Figure 8a presents plots of two main effectsand the (load-time) interaction. The main effectplots are just graphs of the marginal responseaverages at the levels of the two factors.

Notice that all two variables have positivemain effects; that is, increasing the variables

the specimen and the disc surface (George,2005; and Scientific Journal of Riga TechnicalUniversity, 2010). CCD experimental plan andthe wear test results of pure ABS underlubricated conditions are provided in Table 5.

Table 5: CCD Experimental Planand the Wear Test Results of ABS Under

Lubrication Condition

1 –1 –1 20 573.2 0.0001 0.169

a +1 –1 70 573.2 0.0005 0.198

b –1 +1 20 926.8 0.0004 0.202

ab +1 +1 70 926.8 0.0007 0.166

–a –1.414 0 10 750 0.0001 0.153

+a +1.414 0 80 750 0.0011 0.180

–b 0 –1.414 45 500 0.0002 0.156

+b 0 +1.414 45 1000 0.0009 0.199

zero 0 0 45 750 0.0004 0.152

zero 0 0 45 750 0.0003 0.211

zero 0 0 45 750 0.0004 0.174

zero 0 0 45 750 0.0005 0.180

zero 0 0 45 750 0.0003 0.151

CodedVariables

NaturalVariables Responses

tcA B Load,

NTime,sec

Wear,g COF

Analysis of Variance (ANOVA)The ANOVA are summarized in Table 6. It canbe seen that the percentage of load and timesignificantly affect the wear. Using = 0.05,the critical value of F for load and time is F0.05,1, 9 = 5.12. Because 23.54 > 5.12 and11.69>5.12. The values of P less than 0.050indicate model terms are significant. In thiscase factors A (load) and B (Time) aresignificant model terms, but the AB (load-time)interaction value greater than 0.050 indicatesinsignificant model terms.

In addition to the basic analysis of variance,the program displays some additional useful

Table 6: ANOVA-General Linear Model

370


Figure 7: Normal Probability Plotof Standardized Effects

Figure 8a: Main Effect Plot

Figure 8b: Interaction Plot

moves the average deviation from the mean.The interaction between the load and time is

fairly small, as shown the shape of the threecurves in below Figure 8b.

In Figure 9a, plots of the plane of wearvalues generated by the various combinationsof load and time. This three dimensional graphis called a response surface plot. Figure 9bshows the contour lines of constant responsewear in the load and time plane. Notice thatbecause the response surface is a plane, thecontour plot contains parallel straight lines.Examination of this response surface indicatesthat maximum wear of 0.0011 gm is achievedaround 80 N loads and time of 750 sec.

Figure 9a: Surface Plot

Figure 9b: Contour Plot

371


COMPARATIVE STUDYA comparative study of ABS has been takenfor both dry and lubricating conditions. Wearfailure of polymers is controlled by a numberof factors, which include sliding speed, normalload co-efficient of friction, adhesive propertyof the transfer film and thermal properties ofpolymers. Comparative study of pure ABSunder dry sliding and lubricating conditions arediscussed here.

Effect of Friction and Wear of PureABS by Varying LoadEffect of friction and wear of pure ABS byvarying load under dry sliding and lubricatingconditions are illustrated in Figures 10a and10b. Polymers in the presence of lubricantmolecules, plasticize, which reduce friction, butwear can be high because of the decrease inthe mechanical strength of the polymer due toplasticization (Sinha, 2002). The impact ofthese interactions on the mechanicalproperties and failure of an affected polymerare many. One effect can be plasticization ofthe polymer by the adsorbed chemical.Plasticization of a polymer can result in thepolymer being transformed from a rigid, glassymaterial to a soft flexible material (Donald,2002).

Figure 10a gives an idea about the wearbehavior of pure ABS under dry sliding andlubricating conditions. From the above figureit is obvious that the wear under lubricating isless as compared to dry sliding condition. PureABS exhibits superior wear properties inlubricating condition than in dry slidingcondition.

Figure 10b confirms the variation of the co-efficient of friction with load at constant velocity.

From the above graph it is clear that, pureABS has the minimum value of co-efficient offriction is observed in oil lubricated condition.This is due to lubricant interacts with polymermolecules and forms a chemically bondedmonolayer on the outer surface of the polymer.This can drastically reduce the co-efficient offriction when the polymer slides against a hardsurface.

CONCLUSIONThe experimental and statistical analysis arecarried out for the friction and wear behavior

Figure 10a: Wear of ABS Under Dry Slidingand Lubricating Conditions

Figure 10b: Co-efficient of Friction of ABSUnder Dry Sliding and Lubricating

Conditions

372


of under dry sliding and lubricating conditionsleads to the following conclusions.

Pure ABS exhibits superior wearproperties in lubricating condition than in drysliding condition and the co-efficient of frictionof pure ABS has minimum under lubricatingcondition and it remains constant at higherloads.

Besides this study concerning the influenceof characteristic properties of ABS, theseinvestigations will not only help inunderstanding friction and wear behavior butalso assist in design and fabrication ofmaterials for tribological applications basedon ABS.

REFERENCES1. Composite Material Handbook, Vol. 3,

Polymer Matrix Composite MaterialsUsage, Design and Analysis, USDepartment of Defense.

2. Donald E Duvall (2002), “Effect ofEnvironment on the Performance of

Plastics”, ASM International, Vol. II,pp. 796-799.

3. Douglas C Montgomery (2010), Designand Analysis of Experiments, 7th Edition.

4. George Plint (2005), Guidance Notes onLubricated Wear Testing.

5. Kenneth G Budinski (2007), “Guide toFriction, Wear and Erosion Testing”,ASTM Stock Number: MNL 56.

6. Madhav S Phadke (1989), QualityEngineering Using Robust Design ,AT&T Laboratories.

7. Philip J Ross (1995), TaguchiTechniques for Quality Engineering, 2nd

Edition, McGraw-Hill.

8. Sinha K (2002), “Wear Failure ofPlastics”, ASM International, Vol. 11,pp. 1019-1027.

9. Williams J A (2005), TribologyInternational, Vol. 38, p. 863.

tribological characteristics of acrylonitrile … · 2015. 4. 10. · 361 int. j. mech. eng. &...

Documents