Materials and Design 37 (2012) 429–434

Contents lists available at SciVerse ScienceDirect

Materials and Design

journal homepage: www.elsevier .com/locate /matdes

Materials selection for cleaner production: An environmental evaluation approach

Rui Zhao a,⇑, Gareth Neighbour b, Pauline Deutz c, Michael McGuire a

a Faculty of Science, Department of Engineering, University of Hull, East Yorkshire HU6 7RX, UKb Faculty of Technology, Design and Environment, Oxford Brookes University, Wheatley, Oxford OX33 1HX, UKc Faculty of Science, Department of Geography, University of Hull, East Yorkshire HU6 7RX, UK

a r t i c l e i n f o a b s t r a c t

Article history:Received 8 November 2011Accepted 6 January 2012Available online 15 January 2012

Keywords:H. Selection of materialsE. Environmental performanceH. Weighting and ranking factors

0261-3069/$ - see front matter � 2012 Elsevier Ltd. Adoi:10.1016/j.matdes.2012.01.014

⇑ Corresponding author. Tel.: +44 1482 465121.E-mail address: Rui.Zhao@2009.hull.ac.uk (R. Zhao

This paper demonstrates an integrated approach to the selection of materials for cleaner production,using binary dominance matrix and grey relational analysis to aid decision-making by engineers. Thestarting point of this study is to determine the environmental evaluation criteria and to assign weightsreflecting their relative importance. In order to assist engineers in the application of the grey evaluationprocess, and to simplify the computational complexity, a calculating tool based on spreadsheet applica-tion has been developed. A case example is provided, in which five poly(vinyl chloride) (PVC) materialsare subjected to environmental evaluation to determine their sustainability for handbag manufacture.The result of the grey relational analysis and the limitations of the approach are discussed, laying a foun-dation for further work.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Materials selection can be done in different ways, but the prin-ciples are quite similar. Generally, selection of materials for prod-ucts is either ‘design oriented’, driven by functional demands,requiring substantial consideration of mechanical, physical andchemical properties [1–3], ‘product oriented’ by market or publicdemand [4], or ‘cost oriented’ aiming at manufacturing cost reduc-tion [3,5]. As a measure against climate change, cleaner productionhas been introduced into product life cycle management [6,7].Simultaneously, the concept of ‘‘environmental conscious design’’has gradually emerged in manufacturing, in particular focusingon ‘‘design for environment’’ [8,9]. The mitigation of the adverseimpact of products on the environment and human health can beachieved in part by the careful selection of raw materials, thusimproving product life cycle (including reusability and recyclabil-ity), reforming the quality of products, minimising the amountand toxicity of waste generated during manufacturing processetc. [10,11]. In this context, environmental consideration is anincreasingly important component in materials selection, in orderto adapt product design to promote cleaner production, providesustainable products for public consumption and achieve longterm commercial success.

A number of approaches to environmental material selectionhave been proposed. Holloway [12] has extended the Ashby flowdiagram of materials selection by taking environmental factorsinto account for mechanical design, in which air and water

ll rights reserved.

).

pollution are calculated and plotted on the Ashby’s energy contentchart to help engineers or designers better understand the optimalenvironmental design criteria. Ermolaeva et al. [13] have used lifecycle assessment (LCA) to aid materials selection in an automotivestructure, based upon structural optimisation. Three importantindicators, ‘Disability-Adjusted Life Years (DALYs)’, ‘Potentially Dis-appeared Fraction (PDF)’, ‘Mega Joule Surplus (MJS)’, are proposedfor the environmental impact assessment, corresponding to areasof ‘Human Health’, ‘Ecosystem Quality’ and ‘resource depletion’.It is suggested that low-density phenolic foam can be used as acore material for an automotive structure. In a similar study, Boveaand Vidal [14] have applied LCA methodology to selecting materi-als with low environmental impact during the process of woodbased furniture design. Furthermore, Giudice et al. [15] have ap-plied a product life-cycle design approach to the selection of mate-rials, not only to meet the product requirements of functionalitiesand performances, but also to minimise the environmental impactfrom ‘Cradle to Grave’. The brake disc is cited as a case example,and shows how this approach can be integrated with conventionaldesign methods.

To mitigate the effect of ‘climate change’, products are increas-ingly being developed that are less damaging to the environmentand more ‘eco-friendly’, gradually moving towards the concept of‘sustainability’. Thus, Ljungberg [16] has suggested that morerenewable and easily recyclable materials be selected for sustain-able product design, aimed at environmental impact reduction,usually referred to as ‘design for environment’. Further, Zhouet al. [17] have built a multi-objective optimisation model of mate-rials selection for sustainable products, in which life cycle assess-ment is integrated with artificial neural networks.

Table 1Selection of evaluation criteria.

Indicators

Energy consumptionHuman health risk/toxicityMaterials recyclabilityMaterials reusabilityUse of recycled materialsManufacture defectsDefects recyclabilityBy-productsBy-products recyclabilityWaste productionManufacturing costSales profit

430 R. Zhao et al. / Materials and Design 37 (2012) 429–434

While these previous studies are very useful in informing ourapproach, most of them have used life cycle assessment (LCA) ap-proach to the materials selection in terms of environmental con-sideration, which is still a complex process involving a largeamount of available data to determine the specific scenarios, thuslimiting its wider application [17,18]. There is thus a need for anapproach that is simple to use. In addition, most existing studiespresent methods for selecting between widely different materials[13–15]. However, while choosing between materials with sub-stantially different properties is a relatively straightforward task,selection between similar materials, or different varieties of thesame material would require a highly precise data.

In this study, the authors select some typical indicators by tak-ing generic environmental criteria into account, and use a ‘binary-dominance matrix’ integrated with ‘grey relational analysis’ todevelop an evaluation methodology for application by engineers toaid decision-making on materials selection. A computational tool,based on a spreadsheet calculation, has also been devised to helpengineers and designers better understand and employ this ap-proach. Moreover, this work mainly focuses on selection betweensimilar materials (including type and composition) to further im-prove the environmental performance of the product. A study toselect a range of poly(vinyl chloride) (PVC) materials for handbagmanufacturing is used as a case example.

2. Methodology

2.1. Evaluation indicators selection

The evaluation approach starts from the selection of necessaryindicators for cleaner production. The use of environmental andeconomic evaluation indicators is an efficient way to help improveand monitor the processes of product design. According to the ba-sic principles and guidelines of ‘design for environment’ [9,16,19],this study has identified a number of typical indicators to representenvironmental sustainability as a basis for materials selection, e.g.human health risk/toxicity, economics and resource efficiency, inTable 1.

2.2. Determination of weighting factors

As the selected evaluation criteria are not equally important rel-ative to each other and highly dependent on the product, it is nec-essary to introduce some form of weighting as part of theevaluation process. Generally, there are two common approachesfor determining the extent of weighting required objective andsubjective. The former is mainly based upon the data of the attri-butes, e.g. entropy approach, while the later one is based on thedecision maker’s preferences on the attributes, e.g. Analytic Hierar-chy Process (AHP) [20–22]. The binary dominance matrix is a sim-plified way of applying the subjective weighting approach, and hasbeen employed in this study, in order to reduce the computationaltime and complexity for weights determination [23].

In this approach, all the criteria are listed on both vertical andhorizontal axes of a matrix, and then the numerical value 1, 0.5 or0 is input into each box generated by the intersection of the axesaccording to their relative importance. Each individual criterion isevaluated separately against every other criterion. If a pair of indi-cators is judged as equal in importance, a value of 0.5 is assignedto each. Once all the indicators have been scored by 0, 1 or 0.5,each row of the matrix should be increased by 1 to ensure thearithmetic validity. Consequently, the weight can be derived asthe result of the aggregated score of each indicator divided bythe total score of all the indicators, as shown in the followingequation.

W ¼ xi þ 1Pni¼1xi þ 1

ð1Þ

where W is the weight, xi is the score of the individual criterion.

2.3. Grey relational analysis

The grey decision approach derives from grey systems theoryand is applied to those uncertain systems with poor or partiallyknown information. General decision models with grey elements,or integrated with grey system models, can be solved by meansof grey analysis [24,25]. Grey relational analysis is a quantitativemethod used to aid decision making, by examining how closelythe value of each attribute approaches to value of the ideal param-eter. The greater the number of attributes that satisfy this condi-tion, the greater is the grey relational correlation [26]. In thisstudy, the grey relational analysis is a preferred method for select-ing ‘environmentally friendly’ materials for wider industrialapplication.

The basic model for grey evaluation can be described as theproduct of the weight and the result of grey relational analysis[27], as follows:

R ¼W � E ð2Þ

where R = [r1, r2, � � �, rm]T is the result vector of m evaluation objects,and W is the weighting value for each evaluation indicator. E is thematrix composed of all the evaluation indicators, and di(k) is thegrey relational coefficient between the kth indicator and the opti-mal indicator:

E ¼

d1ð1Þ d1ð2Þ � � � d1ðnÞd2ð1Þ d2ð2Þ � � � d2ðnÞ

..

. ... ..

.

dmð1Þ dmð2Þ � � � dmðnÞ

266664

377775

ð3Þ

In this study, all the evaluation indicators have been normalisedin terms of percentage. We further assume that the ideal index ofevaluation indicators is fa�g ¼ ½a�1;a�2; � � �a�n�; in which a�k is theoptimal value for the kth evaluation indicator. The ideal valuesare set depending target requirements and justification of themanufacturer. Some criteria need to be maximised, such as mate-rials recyclability, reusability, and sales profit, and while some areto be minimised, such as energy consumption, health risk, andmanufacturing cost. However, it is important to realise that theideal value for each criterion should be set at an appropriate level,according to the technical and economic feasibilities relevant to aspecific manufacturer.

Once a⁄ has been determined, a new grey relational matrix acan be constructed, where ai

k is the original numerical value ofthe kth indicator:

Fig. 1. Five PVC materials with different compositions.

Table 2Definition of measurements for evaluation criteria.

Indicators Measurements

Energy consumption (%) Ec ¼ Ec ðiÞEt

where Ec(i) is the energy consumed by using the ith material for industrial manufacture, Et is the total energy consumed by

the production, i = A–E. The energy used for PVC materials manufacturing is specified as electricity. For instance, every 100 Kw helectricity, 1 tonne material A will use 97 Kw h for industrial processing, B 93.6 Kw h, C 89.5 Kw h, D 83.0 Kw h and E 80.0 Kw h(values from Table 3)

Human health risk (%) Rh ¼ RhðiÞRt

where Rh(i) is the health risk resulted from using the ith PVC material for industrial manufacture, Rt is the total health risk

caused by using all kinds of the PVC materials for handbags production, i = A–E. Rh(i) = Pe(i) � Ce(i) where Pe(i) is the probability ofenvironmental hazards during selection of the ith material for production, Ce(i) is the consequence of the hazard scenario in terms ofthe number of fatalities

Materials recyclability (%) Rec ¼ Qrec ðiÞQt ðiÞ where Qrec(i) is the quantity of recycled materials while using the ith PVC material for industrial processing, Qt(i)

is the total quantity of the ith PVC material consuming, i = A–EMaterials Reusability (%) Rus ¼ Qus ðiÞ

Qt ðiÞ where Qus(i) is the quantity of reused materials while using the ith PVC material for industrial processing, Qt(i) is the total

quantity of the ith PVC material consuming, i = A–EUse of recycled materials (%) Rur ¼ Qur ðiÞ

Qrec ðiÞ where Qur(i) is the usable quantity of the recycled materials, Qrec(i) is the quantity of recycled material while using the ith

PVC material for industrial processing, i = A–EManufacturing defects (%) Md ¼ QdðiÞ

QM ðiÞ where Qd(i) is the quantity of the manufacture defects, QM(i) is the quantity of products (handbags) while using the ith

material for industrial manufacturing, i = A–EDefects recyclability (%) Drec ¼ QrdðiÞ

QdðiÞ where Qrd(i) is the quantity of the recycled manufacture defects, Qd(i) is the quantity of the manufacture defects while

using the ith material for industrial manufacturing, i = A–EBy-products (%) Bp ¼ QbpðiÞ

QM ðiÞ where Qbp(i) is the quantity of the manufactured by-products, QM(i) is the quantity of products (handbags) while using the

ith material for industrial manufacturing, i = A–EBy-products recyclability (%) Brec ¼ Qbrec ðiÞ

QbpðiÞ where Qbrec(i) is the quantity of the recycled by-products, Qbp(i) is the quantity of the by-products while using the ith

material for industrial manufacturing, i = A–EWaste production (%) Wp ¼ QwpðiÞ

Qt ðiÞ where Qwp(i) is the quantity of waste while using the ith material for industrial processing, Qt(i) is the total quantity of the

ith material consuming, i = A–EManufacturing cost (%) CM ¼ CM ðiÞ

Ctwhere CM(i) is the cost by using the ith material for industrial production, Ct is the total cost paid by using all kinds of the

PVC materials for handbags production, i = A–ESales profit (%) Sp ¼ SpðiÞ

Stwhere Sp(i) is the expected sales profit by using the ith material for industrial production, St is the total expected sales profit

gained by using all kinds of PVC materials for handbags production, i = A–E

R. Zhao et al. / Materials and Design 37 (2012) 429–434 431

a ¼

a�1 a�2 � � � a�na1

1 a12 � � � a1

n

..

. ... ..

.

am1 am

2 � � � amn

266664

377775

ð4Þ

Thus, the grey relational coefficient can be calculated asfollows:

diðkÞ ¼miniminkja�k � ai

kj þ emaximaxkja�k � aikj

ja�k � aikj þ emaximaxkja�k � ai

kjð5Þ

Table 3Data for different indicators provided by a handbag manufacturer.

Indicators (%) Material A Material B Material C Material D Material E Ideal value

Energy consumption 97.00 93.60 89.50 83.00 80.00 70.00Human health risk 0.0015 0.0013 0.0005 0.0007 0.0002 0.0001Materials recyclability 94.70 95.30 94.00 96.30 98.00 98.00Materials Reusability 0.13 0.16 0.22 0.11 0.23 0.23Use of recycled materials 90.80 91.00 91.90 92.70 92.00 93.00Manufacturing defects 8.10 7.30 5.80 6.00 5.00 2.00Defects recyclability 90.00 93.00 93.60 89.00 94.00 95.00By-products 0.020 0.017 0.013 0.013 0.011 0.011By-products recyclability 0.12 0.14 0.16 0.10 0.20 0.20Waste production 4.90 4.50 3.60 2.90 2.50 2.00Manufacturing cost 42.00 38.00 32.00 34.00 28.00 30.00Sales profit 40.00 39.80 38.60 35.70 31.00 45.00

Cost

Human health risk

Sales Profit

Waste production Materials recyclability Materials reusability

Energy consumption

Manufacturing defects By-products

Defects recyclability By-products recyclability Use of recycled materials

Fig. 2. Levels of evaluation criteria ranked by the cited handbag manufacturer.

432 R. Zhao et al. / Materials and Design 37 (2012) 429–434

where e is known as the distinguishing coefficient and e e (0, 1).Empirically, e is set as 0.5 [24,25], and this value is adopted in thisstudy.

From the Eqs. (2) and (5), the grey decision-making model canbe transformed as follows:

Table 4Binary dominance matrix of the case example.

Indicators 1 2 3 4 5 6

1 Energy consumption � 1 0 1 1 12 Human health risk 1 � 1 1 1 13 Materials recyclability 1 0 � 0 1 14 Materials Reusability 1 0 0.5 � 1 15 Use of recycled materials 0 0 0 1 � 06 Manufacture defects 0 0 0 1 1 �7 Defects recyclability 0 0 0 0 0.5 08 By-products 0 0 0 0 1 0.59 By-products recyclability 0 0 0 0 0.5 0

10 Waste production 1 0 0.5 0.5 1 111 Manufacturing cost 1 1 1 1 1 112 Sales profit 1 0 1 1 1 1

Ri ¼Xn

k¼1

WðkÞ � diðkÞ ð6Þ

The evaluation result Ri can be interpreted as the ‘degree of greyincidence’, which is used to express the rates of change of themathematical progressions {ai} relative to an ideal progression{a⁄}. In addition, the evaluation object corresponding to the maxi-mum grey correlative degree is deemed as the optimal object, andvice versa.

In order to simplify computational complexity and to provideconvenient employment of the grey analysis for engineers anddesigners, this approach has been embodied into a computationaltool based on a spreadsheet application (Microsoft Excel 2007).This grey evaluation tool has been divided into two modules, onefor the input parameters and another for the output of the analysisresults. An example of the application of this tool is presented inthe following section.

3. Case example

A leading handbag manufacturer in Guangdong Province, China,has kindly consented to their data being used for this case study.The material commonly used for handbags is PVC and five differentgrades with slight, but significant, difference in composition areinvestigated. The designations A–E are used to distinguish the fivematerials, shown in Fig. 1.

The purpose is to evaluate these five materials by means of thegrey evaluation, and ultimately determine which material is themost sustainable. The generated evaluation results will be ana-lysed in comparison with the current selection of materials usedby the cited company. Table 2 shows how the selected criteriaare measured by using the PVC materials.

According to the measurements, the data collected from thehandbag manufacturer have been normalised by percentage, asshown in Table 3. As for the ideal values, some are set by the

7 8 9 10 11 12 Score Weights

1 1 1 0 0 0 7 0.0891 1 1 1 0 1 11 0.1391 1 1 0.5 0 0 8 0.1011 1 1 0.5 0 0 8 0.1010.5 0 0.5 0 0 0 2 0.0251 0.5 1 0 0 0 4.5 0.057� 0 0.5 0 0 0 2 0.0251 � 1 0 0 0 4.5 0.0570.5 0 � 0 0 0 2 0.0251 1 1 � 0 0 8 0.1011 1 1 1 � 1 12 0.1521 1 1 1 0 � 10 0.127

Fig. 3. Screen shot of the computational tool.

Fig. 4. Screen shot of the grey evaluation result.

R. Zhao et al. / Materials and Design 37 (2012) 429–434 433

production strategies by the cited manufacturer. For example, themanufacturer expects to reduce the electricity consumed by thePVC material processing to 70%, as well as to reduce the manufactur-ing defects to 2%. Similarly, the manufacturing cost should beminimised to 30%, whilst raising sales profit to 45%.

In this case example, the product manager in the cited handbagmanufacturer was requested to rank the importance of each crite-rion, which can lead to a more realistic evaluation. The productmanager has classified the evaluation criteria as seven levels interms of their relative importance, as shown in Fig. 2.

Since some indicators are incorporated into the same level,these are scored 0.5 for each. Based on the importance level, thebinary dominance matrix is constructed, shown in Table 4.

Fig. 3 depicts the input module of the developed tool, startingfrom the weights determination, by marking 0, 0.5 or 1 in the de-signed binary-dominance matrix table (See Fig. 3a first step), fol-lowed by entering the numerical value for each indicator derivedfrom the case example (See Fig. 3b second step).

When all the information has been submitted, the evaluationresult can be generated automatically by the spreadsheet program-ming, and a sequence of the evaluated materials can be rankedfrom the highest to the lowest grade, shown in Fig. 4. Accordingto the grey evaluation result (RE > RC > RD > RB > RA), material E is

the optimum material from an ‘environmentally friendly’ point ofview, material A being the worst choice.

4. Results and discussions

The degree of grey incidence ‘Ri’ reflects how closely the datacolumn of each evaluated material approximates the selected idealvalue (see Table 3), and the greater ‘Ri’ is, the closer the evaluatedmaterial approaches the ideal value set [24–26]. According to theresult from grey relational analysis (RE > RC > RD > RB > RA), materialE is the optimal choice in term of sustainability. This can be de-duced by inspection in Table 3, where the corresponding data showthe highest similarity to the ideal values, especially for the criteriaof materials recyclability, materials reusability, by-products andby-products recyclability. On the other hand, material A is shownto be the least sustainable material among them.

We have also examined the actual consumption of these fivePVC materials used by the cited manufacturer. Material C has thehighest consumption rate for handbag production, followed bymaterials E, D, B, and A, i.e. MC > ME > MD > MB > MA, which reflectsthat the material selection for handbag production is, in reality,strongly influenced by criteria such as ‘market demand’, ‘materials

434 R. Zhao et al. / Materials and Design 37 (2012) 429–434

performance’, and ‘cost-benefit’ [1–5]. In other words, apart fromproduct functionality and company profitability, popular fashionis also a fundamental driver in product design. In terms of ‘cost-benefit’, Table 3 shows that material A has the highest manufactur-ing cost, whilst it also contributes the highest proportion of profits,both about 40%. Comparatively speaking, the processing cost ofmaterial C (about 32%) is less than material A (about 42%), but ityields higher profit (38.6%). ‘Materials performance’ is mainly re-lated to the properties of materials, which affects the quality ofproducts [2,3]. For the PVC materials used in the present case,the laboratory performance tests include a mechanical fatigue test,chemical corrosion test by means of salt spray, oven test for ther-mal expansion and a colour fastness test [28–30]. According to thetest results, material C shows the best abrasive resistance, decay,heat resistance, and colour fastness. However, material A showspoor performances, for example, the surface is easily rubbed offduring the process of manufacture.

In addition, we can identify that the sequence of actual materi-als consumption (MC > ME > MD > MB > MA) is quite similar to theresult from grey relational analysis (RE > RC > RD > RB > RA) in thecase example. Material C shows not only the highest consumptionrate for handbag production, but also the second highest degree ofgrey incidence. Although material E is superior to C in terms of greyevaluation result, the actual consumption of material E is less thanC. This is because material C is more favourable as the main mate-rial, while material E is more commonly used as an auxiliary mate-rial for handbag design. Thus, material C can be determined as thenear optimal choice. It is suggested, therefore, that the actual con-sumption of material could be integrated in the environmentalevaluation to aid decision making on materials selection. In thiscase study, both material E and C emerge as ‘environmentallyfriendly’ for cleaner production, while the use of material A shouldbe reduced or eliminated all together from the manufacturingprocess.

5. Conclusion and further study

This study has introduced a convenient approach for applicationby designers to aid decision-making on materials selection whiletaking environmental evaluation into account. By integrating bin-ary-dominance matrix with grey relational analysis, a computa-tional tool has been developed using a spreadsheet application.This simplified approach allows reduced computational cost andease of use, aimed at encouraging designers to incorporate the con-cept of sustainability into the product design and manufacturingprocesses. The proposed evaluation approach is flexible, and canprovide an environmental index to judge a range of materials. Thisis reflected in the case example in which the most sustainablematerial for handbag manufacturing has been identified among aseries of similar PVC materials. We also suggest that the actualconsumption of material could be incorporated into the evaluationprocess for better justification of the material selection in realsituation.

There is still scope for improvement in our approach. First of all,the evaluation criteria are generic in the context of ‘sustainabledevelopment’. These can be improved further by integrating otherparameters considered during the process of materials selection,e.g. materials performance for the specific industries, thus to makethe evaluation results better fit in various application scenarios.Secondly, the binary dominance matrix technique can be com-pared with other methods for determining weighting factors e.g.Analytic Hierarchy Process (AHP), entropy method etc. to justifyits effectiveness. Moreover, the computational tool can be furtherimproved as a decision support tool for materials selection byincluding other important aspects such as market demand and

materials performance. Although the proposed approach showspromise it is necessary to test the validity and sensitivity of thegrey evaluation model further by applying it to more variety ofcase scenarios.

Acknowledgement

The authors would express sincere thanks to Ms. Alina Lin forthe data collection and Dr. Boon Hon Hong for the usefuldiscussion.

References

[1] Edwards KL. Towards more effective decision support in materials and designengineering. Mater Des 1994;15(52):251–8.

[2] Ashby MF. Materials selection in mechanical design. 3rd ed. Oxford:Butterworth-Heinemann; 2005.

[3] Kalpakjian S, Schmid SR. Manufacturing engineering and technology. 5thed. Singapore: Prentice Hall; 2006.

[4] Ljungberg LY, Edwards KL. Design, materials selection and marketing ofsuccessful products. Mater Des 2003;24(7):519–29.

[5] Farag MM, Magd EEI. An integrated approach to product design, materialsselection and cost estimation. Mater Des 1992;13:323–7.

[6] DEFRA (Department for Environment, Food and Rural Affairs); 2007. Wastestrategy for England 2007. <http://www.defra.gov.uk/environment/waste/strategy/strategy07/index.htm>. [accessed 16.11.09].

[7] UNEP (United Nations Environment Program). Environmental agreements andcleaner production: questions and answers. 2006. <http://www.unep.fr/shared/publications/pdf/DTIx0833xPA-EnvAgreementsEN.pdf>. [accessed27.07.10].

[8] Zhang HC, Kuo TC, Lu HT, Huang SH. Environmentally conscious design andmanufacturing: a state of the art survey. J Manuf Syst 1997;16(5):352–71.

[9] Fiksel J. Design for environment: a guide to sustainable product development.2nd ed. New York: McGraw-Hill; 2009. p. 83–115.

[10] Tchobanoglous G, Kreith F. Handbook of solid waste management. 2nded. New York: McGraw-Hill; 2002.

[11] Williams PT. Waste treatment and disposal. 2nd ed. Chichester: John Wiley &Sons; 2005. pp. 127–130.

[12] Hollway L. Materials selection for optimal environmental impact inmechanical design. Mater Des 1998;19:133–43.

[13] Ermolaeva NS, Castro MBG, Kandachar PV. Materials selection for anautomotive structure by integrating structural optimization withenvironmental impact assessment. Mater Des 2004;25:689–98.

[14] Bovea MD, Vidal R. Materials selection for sustainable product design: a casestudy of wood based furniture eco-design. Mater Des 2004;25:111–6.

[15] Giudice F, Rosa GL, Risitano A. Materials selection in the life-cycle designprocess: a method to integrate mechanical and environmental performancesin optimal choice. Mater Des 2005;26:9–20.

[16] Ljungberg LY. Materials selection and design for development of sustainableproducts. Mater Des 2007;28:466–79.

[17] Zhou CC, Yin GF, Hu XB. Multi-objective optimization of material selection forsustainable products: artificial neural networks and genetic algorithmapproach. Mater Des 2009;30:1209–15.

[18] Yang QZ. Life cycle assessment in sustainable product design. SIM TechTechnical 2007;8(1):57–64.

[19] Tscoulfas TG, Pappis CP. A model for supply chains environmentalperformance analysis and decision making. J Clean Prod 2008;16:1647–57.

[20] Rao RV, Davim JP. A decision-making framework model for material selectionusing a combined multiple attribute decision-making method. Int J Adv ManufTechnol 2008;35:751–60.

[21] Shanian A, Savadogo O. A material selection model based on the concept ofmultiple attribute decision making. Mater Des 2006;27:329–37.

[22] Rao RV, Patel BK. A subjective and objective integrated multiple attributedecision making method for material selection. Mater Des 2010;31:4738–47.

[23] Hurst K. Engineering design principles. New York: John Wiley & Sons Inc;1999. p. 54–56.

[24] Liu SF, Lin Y. Grey information: theory and practical application. London:Springer; 2006.

[25] Liu SF, Lin Y. Grey systems: theory and application. Heidelberg: Springer;2010.

[26] Zhai L, Khoo L, Zhong Z. Design concept evaluation in product developmentusing rough sets and grey relational analysis. Expert Syst Appl2008;3(36):7072–9.

[27] Du D, Pang QH, Wu Y. A modern comprehensive evaluation method and caseselection. Beijing: Qinghua University Press; 2008. p. 115–116 [In Chinese].

[28] SATRA Technology; Leather goods & Luggage Testing. <http://www.satra.co.uk/portal/page.php?id=46>. [accessed 27.12.11].

[29] Allen P. SATRA Spotlight-colour fastness and rub testing. <http://www.satra.co.uk/bulletin/article_view.php?id=740>. [accessed 27.12.11].

[30] Winckle S. SATRA Spotlight-colour transfer from handbags. <http://www.satra.co.uk/spotlight/article_view.php?id=129>. [accessed 27.12.11].