Economic Impact of AluminumIntensive ... - Wiley Online Library

❙

R E S E A R C H A N D A N A LY S I S

Economic Impact of Aluminum-Intensive Vehicles on the U.S. Automotive Recycling Infrastructure Jane E. Boon Jacqueline A. Isaacs Surendra M. Gupta Department of Mechanical, Industrial, and Manufacturing Engineering Northeastern University Boston, Massachusetts, USA

Keywords aluminum automobile disassembly goal programming recycling scrap

Address correspondence to: Prof. Jacqueline Isaacs Northeastern University Department of Mechanical, Industrial, and Manufacturing Engineering 334 Snell Engineering Center 360 Huntington Avenue Boston, MA 02115 USA [email protected]

© Copyright 2001 by the Massachusetts Institute of Technology and Yale University

❙

Summary The use of aluminum alloys in automobile production is growing as automakers strive to lower vehicle fuel consumption and reduce emissions by substituting aluminum for steel. The current recycling infrastructure for end-of-life vehicles is mature, profitable, and well suited to steel-intensive vehicles; increased use of cast and wrought aluminum, however, will present new challenges and opportunities to the disassembler and shredder, who now comprise the first stages of the vehicle recycling infrastructure. Using goal programming techniques, a model of the auto recycling infrastructure is used to assess the materials streams and process profitabilities for several different aluminum-intensive vehicle (AIV) processing scenarios.The first case simulates the processing of an AIV in the current recycling infrastructure. Various changes to the initial case demonstrate the consequences to the disassembler and shredder profitabilities whenever the price of nonferrous metals changes; greater fractions of the vehicle are removed as parts; the parts removed by the disassembler have increased aluminum content; the quantity of polymer removed by the disassembler is increased; the disassembly costs increase; the disposal costs for shredder residue and hazardous materials increase; the shredder processing costs increase; and different AIV designs are considered. These profits are also compared to those achieved for a steel unibody vehicle to highlight the impact of introducing AIVs into the existing infrastructure. Results indicate that the existing infrastructure will be able to accommodate AIVs without economic detriment.

Volume 4, Number 2

❙ Journal of Industrial Ecology

117

❙


Introduction The primary driver encouraging the substitution of aluminum alloys for steel is the desire of automakers to improve the fuel economy of their products. Although aluminum costs considerably more than steel, its use can promote vehicle curbweight reduction, which in turn has potential to reduce the overall vehicle emissions. One study identified that an aluminum-intensive vehicle (AIV) with a curb weight of 1,176 kg 1 would weigh 10.6% less than the curb weight of a conventional steel unibody at 1,316 kg (Field et al. 1994). In 1991, U.S. passenger vehicles contained an average of 86.7 kg of aluminum, whereas in 1999 passenger vehicles were expected to contain approximately 109.4 kg of aluminum. Since 1977, the aluminum content in vehicles has increased by 68 kg, for an increase in usage of 4.3% each year (Anonymous 1998a). By the year 2006, the average usage of aluminum could reach over 180 kg per vehicle. For automobiles redesigned to showcase and exploit aluminum (i.e., AIVs), the aluminum content could reach over 400 kg per vehicle (Litalien et al. 1997). The reduction in curb weight enables an AIV to more easily satisfy corporate average fuel economy (CAFE2) requirements (Naggar 1992). The Aluminum Association (2000) notes an accompanying secondary weight reduction, where “a 10% reduction in vehicle weight improves fuel economy approximately 6% to 8%.” Primary weight reduction is attributed to a reduction in mass of an automobile body, whereas secondary weight reduction results when lighter engines and transmissions can then propel the vehicle. To further promote the use of aluminum in vehicles, the Auto Aluminum Alliance was formed in 1999 as a working group within the United States Council for Automotive Research (USCAR). This group includes Ford Motor Co., DaimlerChrysler, General Motors, and ten aluminum suppliers. Its mandate is to help automakers cut overall vehicle weight by 40% (Robinson 1999a). This goal to reduce vehicle weight will become increasingly acute when CAFE standards are increased in the future (Stoffer 1999). As these AIVs reach the end-of-life (EOL), they will be processed and recycled. Currently, in the United States, an extensive recycling in118

Journal of Industrial Ecology

frastructure exists. Its primary constituents are the disassembler and the shredder, who recover and recycle over 75% of the materials in the EOL vehicle. The profitability of removing parts and other materials determines the extent to which a given vehicle is disassembled prior to shredding. The dismantled vehicle is then sold to the shredder, who processes hulks for their metallic scrap. At present, most of the aluminum used in vehicles is in the form of castings, which are often large, easily removable parts (e.g., the engine block). There is a mature market for the castaluminum scrap, where the material is sold to the secondary metals market and ultimately recycled into new castings. In the future, however, AIVs will contain sheet and extruded wrought alloys to form parts such as body panels or the spaceframe. AIVs utilize a spaceframe design, where a cage-like structure made from extruded alloys is designed to withstand all mechanical loads and body panels are attached to this frame. This recent design differs from the unibody design, where the body panels are an integral part of the structural design and contribute to the integrity of the welded unibody structure. Wrought aluminum (i.e., mechanically deformed and not cast to shape) is more expensive to use and potentially more valuable at EOL than cast aluminum. The value of aluminum scrap remains uncertain, as the technologies to identify, process, and recycle aluminum alloys are in their infancy. Hence, the future value of aluminum-intensive vehicles is strongly linked to improvements in sorting technologies.

Current Recycling Infrastructure The final owner of a vehicle is paid $50 or more for the vehicle by the dismantler/ disassembler, although in some cases the last owner must pay to have the vehicle reclaimed. This price depends on the condition of the vehicle, and the perceived value of its components and materials. The dismantler removes reusable components (e.g., electronics and selected parts) and any particularly valuable materials fractions (e.g., large castings, batteries, and catalytic converters). The disassembler also re-


moves tires and fluids to insure that the hulk is acceptable to the shredder (Field et al. 1994). The shredder buys the hulk for around $50 (Chen 1995). These hulks are processed by a rotating hammer mill into fist-sized pieces, which are then separated into ferrous, nonferrous, and nonmetallic automobile shredder residue (ASR) fractions by the shredder, which uses eddy current separation techniques and, occasionally, media separation techniques (Taylor 1999b). The ferrous fraction is sold to the minimill industry for steelmaking. The nonferrous fraction is sold to specialized shops, where it is further sorted through the use of sink-float separation technologies and eddy current processing. At present, this process produces an aluminum recovery rate of over 80% (Litalien et al. 1997) and the profits have been estimated at over $44/hulk (Chen and Field 1993). ASR is not recycled; it is sent to landfill, although numerous groups are seeking ways of reclaiming and reusing it, or deriving energy from it (Anonymous 1995; Field et al. 1994; Jody et al. 1996; Winter 1996).

Recycling Aluminum Because the recovery of the nonferrous fraction drives shredder profitability (Taylor 1999a), it is worthwhile to separate the various metals present, despite the difficulty of separating nonferrous metals from each other and from stainless steels. If the secondary aluminum mixture is melted with diverse aluminum alloys, the only application for this material would be as a casting alloy (Nijhof and Rem 1999). It is worthwhile to recover the secondary aluminum, however, because the energy required to melt and recover the aluminum is approximately 5% of that required to extract aluminum from bauxite ore (Oye et al. 1999). If the recovered aluminum has too many impurities or alloying additions, it would be necessary to put the scrap aluminum mixture back through the same process by which the aluminum was originally extracted from the bauxite ore. This energy-intensive process is the only method for removing impurities, but it removes all impurities; the resulting pure aluminum must subsequently be realloyed. This electrolytic purification process, for example, is unnecessary when aluminum beverage cans are recycled because the lid and body alloys

❙

are engineered so that when the used beverage cans (UBCs) are melted down, the resulting alloy may be used in future beverage cans. Given the many different applications for aluminum alloys within an automobile, this kind of materials engineering would prove extremely difficult. Although it may be possible to remove and segregate some of the closure panels and the major castings prior to shredding, the main body structure of an AIV will need to be shredded to facilitate the separation of the various materials. Furthermore, the fact that the aluminum structural components are buried deep within the AIV will discourage the greater segmentation and disassembly of the aluminum frame (Kirchain 1999) and will reinforce the importance of shredding as a means of capturing the aluminum. Shredding also facilitates the ultimate decoating of the aluminum. A decoating step is not required in steel production, because the steelmaking process is not as sensitive to the presence of organic compounds or other metals. Without a decoating step in secondary aluminum processing, incomplete pyrolysis of the coatings and excessive dross losses would occur when the painted scrap was remelted. Decoating is a thermal process that removes organic compounds and other contaminants attached to the aluminum by decomposing and oxidizing them at high temperatures. By-products from this process are gas, tar, and char (black carbon). The formation of gases means that the decoater must be equipped with an exhaust gas control system to neutralize emissions and to avoid any secondary reactions. The formation of particulate waste requires the filtering of the gas stream. Decoaters have been used successfully in the recycling of UBCs where the level of organic material does not exceed 4%. Heavily painted sheet or laminated foils may have a level of up to 50% organic material. Once the aluminum is decoated, it must then be sorted so that the various alloys may be segregated and recycled into their original form. The growth in aluminum usage implies a corresponding increase in demand for sheet products. Although castings can continue to absorb aluminum scrap into the next decade, it is preferable to return wrought material to its original application, because wrought alloys have higher value than

Boon, Isaacs, and Gupta, Recycling of Aluminum-Intensive Vehicles

119

❙


cast alloys. Furthermore, greater demands are being placed on the metallurgical specifications for automotive castings, which further emphasizes the importance of accurate aluminum alloy sorting techniques (Litalien et al. 1997; Oye et al. 1999). Various methods for sorting aluminum alloys have been proposed. One possibility is a hot crush and screening method that capitalizes on the difference in embrittling temperatures for different alloys. This method is only appropriate for recycling the cast fraction of the shred into foundry alloys, however, and would not permit the wrought fraction to be used in its original form (Brown et al. 1985). Eddy current techniques are currently employed to sort out the different metals in the nonferrous fraction of the auto shred. This technique could be used for sorting the different aluminum alloys, but its usefulness is constrained by the fact that the pieces to be separated must be close in size, and there must be conductivity differences among the various alloys. The method most equipped to deal with the broad range of alloys present in a stream of scrap aluminum would be one that assesses the direct composition of an individual piece of scrap. Optical emission spectroscopy (OES) is one such technique. OES is currently used in casting centers and foundries, but a system that could accommodate high speeds, different-shaped pieces of scrap, and a wide variety of alloys is still being developed. Alcan has assessed the basic feasibility of such a system by employing a pulsed laser that was focused onto successive aluminum alloy samples arrayed on a turntable. By first establishing baseline spectra from alloy disks, Alcan was able to sort and identify a wide variety of aluminum alloys based on their silicon, copper, and zinc content (Litalien et al. 1997). Once sorted, the aluminum alloy mixture is melted in reverberatory furnaces. Various alloying elements will be present, due to the variety of aluminum alloys used in automobiles. The levels of unwanted alloy additions are generally controlled by chlorination and scrap blending. The most problematic alloying element found in the melt is magnesium because the limit for magnesium in die-casting alloys is typically 0.1%. As a result, it is necessary to closely control the level of magnesium. Chlorination is the 120


most common means of controlling the magnesium content of the melt; the corrosive nature of chlorine, however, and its potential harm to the ozone layer may result in the future regulation of its use. As a result, a variety of alternatives to chlorination have been explored, including the use of fluxes, molten salt electrorefining, and vacuum refining (Bhakta 1994). These processes are all more expensive and not competitive with chlorination; environmental issues, however, may propel their use, and make the recycling of aluminum more costly.

Aluminum-Intensive Vehicles AIVs, such as the Audi A8 (Anonymous 1998c), are composed of considerable amounts of aluminum sheet and extruded wrought alloys. The spaceframe structure, made from wrought aluminum alloys, is at the heart of an AIV. Aluminum is also employed in a variety of other applications. Forged products like wheels; sheet products like outer body panels, radiators, and condensers; and extruded parts such as bumper reinforcements and door sashes (Kurihara 1994) are often made of aluminum. The Saturn car company uses an aluminum engine block that weighs 21.3 kg, and engineers believe that a comparable cast-iron engine block would weigh 39.5 kg (Hogan 1998). When aluminum is used extensively in a car structure, Audi engineers have found that the structure may be up to 67% lighter than its steel counterpart. In a concession to crash safety, however, the engineers chose to reduce the frame weight by only 40% (Ashley 1994). Research is now being undertaken to evaluate the use of aluminum in a variety of additional applications, such as door beams (Ashley 1994; Yamashita and Hirano 1998); fenders (Mascarin and Dieffenbach 1992); and stressed body panels, engine cradles, radiator brackets, and suspension beams (Robinson 1999b). Furthermore, aluminum metal matrix composites (MMCs) are currently being tested in applications such as drive shafts and brake rotors. Aluminum MMCs, however, are very costly (typically $2/pound, versus $0.60 to $0.80/ pound for aluminum) (Mangin et al. 1996) and, therefore, these composites will not be intro-


duced until cost and processing difficulties are remedied. There is also a concern that the ceramic particles in MMCs may contaminate the aluminum recycling operations (Hogan 1998). A further expansion of the use of aluminum is as a decorative element in some vehicles. For instance, aluminum is currently being used throughout the Audi TT Coupe interior to convey a high-tech and high-performance look. The TT has its gear shifter, air vents, and cupholders all made of aluminum. It also has an aluminum fuel filler cap (Keebler 1999). Toyota is employing the metal in a similar manner by making the pedals of its Celica out of aluminum (Rechtin 1999). Another trend in aluminum usage is to offset the weight of luxury features through the use of aluminum, as in the Audi A8, where the extensive use of aluminum permits competitive performance levels despite many additional features (Anonymous 1998c). Despite aluminum’s weight advantages over steel, there are factors that impede the widespread use of this material beyond its greater cost. The long-standing use of steel means that the auto industry lacks the expertise to produce large volumes of vehicles with all-aluminum structures. As a result, the aluminum companies have had to assist the automakers with engineering related to the use of aluminum in automotive parts. Beyond the unique formability issues of aluminum, the automakers must contend with issues like materials handling, production line rates, and joining when employing aluminum in large production volumes of vehicles. To accommodate the widespread use of aluminum in a typical passenger car composed of approximately 300 stamped parts would require special equipment that would add $1,000 to $2,000 to the price of the vehicle, beyond the additional cost of the raw materials (Robinson 1999b). As a result of this increased cost, the extensive use of aluminum may be found more frequently in upscale products, such as the Audi A8, and less frequently in lower-cost models with high sales volumes. Furthermore, in a study from the Materials Systems Laboratory at the Massachusetts Institute of Technology, it is projected that an AIV fleet will require 15 to 17 years before its lifetime release of CO2 declines below the lifetime release of CO2 by a conventional fleet when the emissions

❙

involved in the production are included with the emissions released during driving (Anonymous 1999). This study also indicates that when an aluminum vehicle is compared with a steel vehicle, the release of CO2 by the AIV would be less than that of a steel vehicle after about six years. While these conclusions may suggest a longer delay for emission reductions, a reason still exists to continue to analyze the AIV and the recycling infrastructure: The immediate benefits of aluminum to the automakers in terms of improved vehicle fuel economy will encourage its substitution for many years.

Goal Programming Model To analyze the effects of changes to the current recycling infrastructure, a model using goal programming is developed. Goal programming seeks to establish a level of achievement for prescribed goals or targets (Steuer 1986). To this end, equations are defined for the economic relationships of the disassembler and shredder that model the removal and sale of materials and parts from the AIV. The conflicting goals for the disassembler and the recycler are formulated based on assumptions, constraints, and profitability targets for the disassembler and the shredder that build upon the work of Isaacs and Gupta (1997). Assumptions 1. Polymers can be hazardous or nonhazardous. 2. No market exists for the sale of the polymers removed. 3. All hazardous materials removed are disposed of in a landfill. 4. There is sufficient demand for parts and materials removed by the disassembler and the shredder. 5. The parts removed by the disassembler are composed of weights of ferrous and nonferrous materials in the same proportion as these materials are found in the AIVs. 6. The disassembler will seek to maximize profits unless constrained by the shredder. 7. The shredder will, by assumption, aspire to receive profits that are at least $35/vehicle (80%) of the prevailing profits of $44/vehicle (Chen 1995).


121

❙


Constraints

Goals

The definitions for the variables and constants found in the equations defining the goal programming model are found in tables 1 through 4. The variable Xi (i = 1, 2, 3, 4, 5, 6) represents the masses of ferrous, nonferrous, polymer, nonhazardous, hazardous materials, and parts removed by the disassembler, respectively. 1. If the variable mi (i = 1, 2, 3, 4, 5) represents the masses of ferrous, nonferrous, polymer, nonhazardous, and hazardous materials in the automobile respectively, then the curb weight, W, is given by:

The following equations are the profit functions for the disassembler and the shredder, and represent the differences in the costs of processing and disassembling a vehicle and the revenues generated from the sale of its parts and materials. 1. Disassembler Profit The disassembler generates revenues from the sale of the ferrous materials (R1 · X1), the nonferrous materials (R2 · X2), polymer materials (R3 · a6 · (1 – a7) · X3), nonhazardous materials (R4 · (1 – a8) · X ), hazardous materials (R · (1 – a ) · (1 – a ) · 4 5 6 9 X3), parts (R6 · X6), and the sale of the hulk to the 



6

W = m1 + m2 + m 3 + m4 + m5

shredder R7 ⋅  W − ∑ X1  . The costs incurred are   i =1 from the acquisition of the EOL vehicle (C1 · W), the disassembling and removal of various materi-

2. The amounts of a material removed by the disassembler must exceed a nominal mass:

als and parts C2 ⋅ ∑ X1 , the disposal of polymer

X i ≥ w i ∀i = 1,...6

materials (a6 · a7 · C3 · X3) and other nonhazardous materials (a8 · C3 · X4), and the disposal of contaminated polymer materials ((1 – a6) a9 · C4 · X3) and other hazardous materials (C4 · X5). Thus the disassembler profit can be expressed as:

3. The disassembler removes parts composed of ferrous, nonferrous, and polymer fractions. These fractions may vary between a steel unibody vehicle (a1s, a2s) and an AIV (a1a, a2a):

a1 + a2 + a 3 = 1

6

i =1

(

)

(

)

R1 ⋅ X1 + R2 ⋅ X2 + R 3 ⋅ a 6 1 − a 7 X 3 + R4 1 − a 8 X4 6   + R5 1 − a 6 1 − a 9 X 3 + R6 ⋅ X6 + R7 ⋅ W − Xi    i =1

(

)(

)

∑

6

4. The disassembler cannot remove more material as secondary material or as parts than what exists in the EOL vehicle. The fraction, a4, is associated with the portion of the total nonhazardous materials removed by the disassembler: X1 + a1 ⋅ X6 ≤ m1 X2 + a2 ⋅ X6 ≤ m2 X 3 + a 3 ⋅ X6 ≤ m 3 X4 ≤ a 4 ⋅ m 4 X5 ≤ m 5 5. The mass of parts removed by the disassembler must not exceed a certain fraction of the curb weight, where a5 is the fraction of curb weight removed by the disassembler as parts:

∑X − a ⋅a ⋅C ⋅X − a −(1 − a )a ⋅ C ⋅ X − C ⋅ X

− C1 ⋅ W − C2 ⋅

6

X i ≥ 0 ∀i = 1,...6 122


7

3

3

8

⋅ C3 ⋅ X4

9

4

3

4

5

2. Shredder Profit The shredder generates revenues from the sale of the ferrous materials extracted from the hulk after the EOL vehicle has been processed by the disassembler (R1 · a10(m1 – X1 – a1 · X6)), and from the sale of the nonferrous materials (R2 · a11(m2 – X2 – a2 · X6)). The shredder’s costs are related to the buying and processing of the 



6

hulk (C5 + C6 ) ⋅  W − ∑ X1  and to the disposal of the ASR  C7 W − 

i =1

6

∑X

1

i =1

 − a10 m1 − X1 − a1 ⋅ X6 − a11 m2 − X2 − a2 ⋅ X6  . 

(

)

(

)

Thus, the shredder profit can be expressed as:

( (

(

 − C7 W − 

6

∑ X − a (m i

i =1

)

R1 ⋅ a10 m1 − X1 − a1 ⋅ X6 + R2 ⋅ a11 m2 − X2 − a2 ⋅ X6 6   − C5 + C6 ⋅ W − Xi    i =1

X6 ≤ a 5 ⋅ W 6. The mass of materials removed by the disassembler cannot be negative:

6

i

i =1

10

1

)

)

∑

 − X1 − a1 ⋅ X6 − a11 m2 − X2 − a2 ⋅ X6  

)

(

)


Table 1 Inventory list for vehicle materials content Materials (kg)

AIV

Ferrous content (m1) 612.27 Nonferrous content (m2) 240.23 Polymer content (m3) 100.91 Nonhazardous content (m4) 146.14 Hazardous content (m5) 75.91 Curb Weight (W) 1,175.45 +

Steel unibody+ 891.82 101.59 100.91 146.14 75.91 1,316.37

Source: Isaacs and Gupta (1997).

Table 2 Nominal materials to be removed by disassembler Materials (kg)

AIV

Ferrous content (w1) Nonferrous content (w2) Polymer content (w3) Nonhazardous content (w4) Hazardous content (w5) Parts (w6)

56.16 49.76 0.00 47.84 75.91 135.26

+

Model Parameters The materials content of an AIV automobile design and the resulting profits for stakeholders who process, sell, or dispose of the materials are investigated in these analyses. The materials are segregated into five categories: ferrous, nonferrous, polymer, nonhazardous, and hazardous. The sum of these materials equals the curb weight of the car. The nonferrous mix is assumed to be 87.7% aluminum alloy, 8.7% copper, and 3.6% zinc. The AIV was assumed to be a unibody design, because this particular design would contain less aluminum than an aluminum spaceframe, thus mitigating the magnitude of the revenues derived from the secondary aluminum and further challenging the profitability of the recycling infrastructure. The vehicle content assumptions are listed in table 1. A steel unibody is included in the table to serve as a comparative reference for the AIV. Within the existing infrastructure, the dismantler removes a given quantity of materials (as parts or as secondary materials) based on regulatory requirements, expected profitability, or

❙

Steel unibody+ 67.96 18.59 0.00 47.84 75.91 113.28


shredder need. To capture this phenomenon, in the model the disassembler is required to remove at least a nominal mass of materials (w i), and disassemblers may remove more of a given material if it is in their best interest to do so. Table 2 shows the values for these nominal materials. In addition to the required nominal materials, the disassembler may remove other materials as well. The mass of these materials (X i) removed by the disassembler represents the decision variables for the multicriteria optimization formulation. Table 3 shows the assumptions for the cost and revenue parameters resulting from this processing as well as from subsequent processing. Table 4 identifies assumptions for additional input parameters.

Results The results from eight AIV case studies are highlighted here. The first case investigates how profitabilities of the shredder and the disassembler change with increasing prices for nonferrous scrap. The next three cases explore the effect of increases in quantities of various materials that

Table 3 Assumptions for cost and revenue input parameters+ Costs C1 C2 C3 C4 C5 C6 C7 +

Buying EOL vehicle Disassembling EOL vehicle Disposing of nonhazardous materials Disposing of hazardous materials Buying hulk Shredding and separating hulk Disposing of ASR

($/kg) 0.035 1.5 0.033 0.20 0.05 0.0188 0.033

Revenues R1 R2 R3 R4 R5 R6 R7

Ferrous materials Nonferrous materials Polymer composites Nonhazardous materials Hazardous materials Parts Selling hulk

($/kg) 0.11 0.88 0 0 0 4.1 0.05



123

❙


Table 4 Definitions and values for additional input parameters Fraction of ferrous materials removed by disassembler as parts in AIV Fraction of ferrous materials removed by disassembler as parts in steel unibody Fraction of nonferrous materials removed by disassembler as parts in AIV Fraction of nonferrous materials removed by disassembler as parts in steel unibody Fraction of polymer removed by disassembler as parts Fraction of total nonhazardous materials removed by disassembler Fraction of curb weight removed by disassembler as parts Fraction of polymer removed by disassembler as nonhazardous Fraction of nonhazardous polymer disposed Fraction of nonhazardous materials disposed Fraction of hazardous polymer disposed Extraction efficiency of the shredder on ferrous materials Extraction efficiency of the shredder on nonferrous materials

a1a a1s a2a a2s a3 a4 a5 a6 a7 a8 a9 a10 a11 +

0.7 0.8+ 0.3 0.2+ 0 0.33+ 0.15+ 1+ 1+ 0.6+ 0+ 0.9+ 0.9+


the disassembler removes from the hulk. These are subsequently followed by three cases that examine the changes in various processing costs. The last case focuses on the effect of AIV design changes on recycling profitabilities. Increased Price of the Nonferrous Mix As the use of advanced materials is expanded, there will be repercussions to the disassembler and the shredder. The price of the nonferrous

mix could vary considerably, depending on the nature of the materials, and on the ability of these materials to be sorted and identified. For instance, if the nonferrous fraction is not readily sortable, the mix would be devalued due to the presence of unwanted additions. If sorting technologies are such that the different alloys may be readily separated and categorized, however, the value of the nonferrous mix may be substantially higher. Furthermore, the market price for secondary metals does vary considerably, and this

Figure 1 Expected profits with increasing market price of secondary nonferrous metals (R2).

124



❙

Figure 2 Expected profits as the fraction of curb weight removed as parts (a5) increases for AIV.

variability will affect the revenues for the disassembler and the shredder (Taylor 1999a). Figure 1 shows the relationship between the maximum disassembler and shredder profits as the revenue for nonferrous materials varies from $0.44/kg to $1.32/kg (i.e., when R 2 is varied ±50% from its current market price of $0.88/kg). Shredder profits are more closely linked to the market price of nonferrous materials because the only sources of revenue for the shredder are from the sale of the ferrous and nonferrous scrap metals, unlike the disassembler, who may also sell used parts. As a result, if the market value of the nonferrous materials declines below $0.44/kg, the shredder will be unable to reach his typical profitability of $35/vehicle, and thus the disassembler may be forced to accept a lower price for the hulk from the shredder. At nonferrous prices above that level, however, the shredder will be able to meet and exceed this profitability goal. Not surprisingly, the shredder generates more revenue with AIVs than with steel unibody vehicles. For instance, when the price of nonferrous metals is $0.66/kg, the AIV generates profits for the shredder of $63.16/AIV, while for similar modeling conditions, the steel unibody generates profits for the shredder of $20.34/vehicle. The price of non-

ferrous metals still exceeds the price of ferrous metals, and the greater mass of the high-value nonferrous metals in the AIV makes for greater revenues for the shredder. Increased Quantities of Parts Removed by the Disassembler Figure 2 shows the profitabilities of the disassembler and the shredder when the disassembler removes and sells different percentages of the EOL AIV as parts (i.e., when a5 is varied). It is assumed that the parts comprise 70% ferrous materials and 30% nonferrous materials, as these are the relative proportions of the materials in the modeled AIV. When only a small percentage of the vehicle is sold as parts, it considerably limits the disassembler’s profitability. For instance, the highest profit that a disassembler may achieve when parts comprise 11% of the curb weight is $25.20/AIV, whereas the disassembler may achieve maximum profits of $864.48/AIV when the fraction is 39% of the curb weight. If there is a limited market for the vehicle parts, disassemblers fail to achieve their typical profits, and there may be reluctance to undertake the re-


125

❙


Figure 3 Expected profits as the fraction of parts composed of nonferrous metals removed by the disassembler (a2) increases.

cycling of the vehicle. Conversely, the shredder benefits when there is only a limited market for the used parts. As more parts are removed from the vehicle, there is less material for the shredder to sell. When the disassembler removes only 11% of the vehicle as parts, the shredder receives profits of $101.44/AIV. In comparison, a steel unibody would generate profits of $70.73/ vehicle for the disassembler and $37.67/vehicle for the shredder when a5 = 11%. The disassembler earns more with the steel unibody vehicle than with the AIV because the steel vehicle has a greater mass, and would therefore generate a greater mass of parts to be sold—although parts are normally sold for functionality. (This assumption is not expected to skew the results.) The shredder earns substantially less with the steel unibody vehicle because there is a reduced mass of profitable nonferrous metals that may be harvested and sold. As the disassembler removes more and more parts, the shredder’s profits decline. It is only once the disassembler has removed approximately 35% of the curb weight of the AIV as parts that the shredder is not able to achieve the minimum profitability goal of $35/vehicle. Under current conditions, the disassembler will not have to take any extraordinary measures to accommodate the profitability needs of the shredder. If there is little demand for the used parts, however, the disassembler may be hesitant to purchase the EOL vehicle. This possibility may be offset by paying less for the vehicle, or by requiring the shredder to pay a premium for the aluminum-rich hulk. 126


Increased Aluminum Composition in Parts Removed by the Disassembler It is difficult to anticipate the materials content of the parts removed by the disassembler. If 15% of the curb weight of the vehicle is assumed to be removed as parts, then the materials content of the parts will have a significant effect on the profits earned by the shredder, because ferrous materials may be sold for $0.11/kg, and nonferrous materials may be sold for $0.88/kg. If the parts are made up primarily of the more valuable materials, the shredder will have less of these materials to sell to the secondary metals market. Figure 3 shows the relationship among the materials content of the parts and disassembler and shredder profitabilities (i.e., when the fraction of nonferrous metals removed as parts (–a2) is varied from 0.2 to 0.8). For instance, if only 20% of the mass of the parts is nonferrous materials and 80% is ferrous, the shredder will receive profits of $102.62/ AIV. If the percentages are reversed, however, the shredder will receive profits of $29.30/AIV—an amount that is less than the shredder’s minimum acceptable profits of $35/vehicle. The profits received by the disassembler are unaffected by changes in the composition, because the market value of a used part is determined not by its composition, but instead by its utility. Increased Polymer Removal by Disassembler When the disassembler is not required to remove and dispose of any polymer, the disassembler is able to achieve the greatest profits. The cost of


❙

Figure 4 Expected profits as more polymer is removed and disposed of by the disassembler.

disassembling and disposing of the polymer is quite significant, resulting in a decline in disassembler profits of approximately $1.58/kg polymer removed, as seen in figure 4 (where w3 is varied from 0 to 30 kg). The disassembler must incur the cost to remove and dispose of the polymer, as well as endure the reduction in hulk revenues from the shredder, because the shredder typically purchases the EOL vehicles by mass. As more polymer is removed by the disassembler, however, the shredder benefits because hulk costs are reduced and there is less ASR for disposal. The benefit to the shredder is relatively small, because the profits increase by only $0.10/kg polymer removed by the disassembler. For example—when the disassembler removes 20 kg of polymer from an AIV instead of no polymer, the disassembler’s profits decrease from $145.09/AIV to $113.43/AIV, whereas the shredder’s profits increase from $90.40/AIV to $92.44/AIV. In comparison to a steel unibody vehicle under the same conditions, the disassembler’s profits decrease from $205.00/vehicle to $173.34/vehicle, and the shredder’s profits increase from $28.96/vehicle to $30.99/vehicle. Given the negligible benefit to the shredder—whose profits are already substantial for the AIV—disassemblers are unlikely to remove any

polymer from an AIV that they are not mandated to remove or that cannot be sold as parts. If the cost to dispose of ASR increases, however, a greater beneficence may be required of the disassembler. Increased Cost of Disassembly When the disassembler removes parts, some parts will be readily removed, whereas others will be more difficult. Ease of removal directly influences the disassembly cost. To simulate this phenomenon using the model, when the fraction of the vehicle curb weight that is removed as parts exceeds 15%, the average cost (C2) to remove all of the parts is assumed to increase by 10%, from $1.50/kg to $1.65/kg. When this occurs, the disassembler must have a significant market for the used parts before it becomes economical to absorb the increased disassembly costs. The results in figure 5 indicate that not until over 18% of the vehicle curb weight can be sold as parts by the disassembler does it become worthwhile for the disassembler to remove parts beyond the threshold of 15%, where the profits are $145.09/AIV. The additional disassembly costs exceed the additional profits until there is a market for 18% of the vehicle, where the


127

❙


Figure 5 Expected profits for increases in the average cost of disassembly (C2) with a threshold of 15%.

Figure 6 Expected profits as the costs of ASR and hazardous waste disposal (C4 and C7) increase from $0.20/kg.

128



❙

Figure 7 Expected profits as hulk processing costs increase.

disassembler would earn profits of $168.83/AIV. The shredder’s profits are related to the mass of parts removed by the disassembler, and as that mass increases, the shredder’s profits decline. When the disassembler is discouraged by the increased disassembly costs from removing all of the parts for which there is a market, the shredder is able to capture profits that they might not have enjoyed otherwise. Increased Hazardous Waste and ASR Disposal Costs The disposal of hazardous materials and of ASR is relatively inexpensive in the United States. At present, hazardous materials may be landfilled for approximately $0.20/kg, and nonhazardous materials may be disposed of for approximately $0.033/kg. If ASR is classified as hazardous waste, this substantially affects the profitability of the shredder. Figure 6 shows the relationship between disassembler and shredder profits, and the cost to dispose of hazardous wastes, including ASR (i.e., when C7 = C4 and C4 increases from $0.20/kg to $0.34/kg). As the cost to dispose of these wastes increases, both the disassembler and the shredder experience declining profits. When the disposal cost is $0.26/kg, the disassembler receives profits of $140.54/AIV and the shredder receives $32.23/ AIV, slightly less than the goal of $35/vehicle. This is less than the profits of $47.61/AIV received by the shredder when hazardous wastes are only $0.20/kg and ASR is classified as nonhazardous waste. The AIV profits compare fa-

vorably with the profits derived from recycling a steel unibody when waste disposal costs are $0.26/kg, where disassembler’s profits are limited to $44.61/steel unibody, because the model imposes the removal and disposal of 98.06 kg of polymer on the disassembler to prevent the shredder from losing money. Once the disposal cost exceeds approximately $0.25/kg, a shredder would generate profits less than the average of $35/vehicle. If the waste disposal costs exceed that threshold, the shredder would need some form of subsidy from the disassembler. This could result in the disassembler removing and disposing of some of the polymer materials present in the EOL vehicle so that these costs are not borne entirely by the shredder. The shredder could also pay less for the hulk, or charge a fee to process the hulk. Increased Shredding Costs The cost of separating and shredding a hulk is related to the cost of the energy used to operate the facility, which has a very direct impact on the profitability of the shredder. Under present conditions, the cost to process a hulk is quite low—only $0.0188/kg. Figure 7 shows the impact of increased processing costs on the profitability of the shredder (i.e., where C6 varies from $0.0188/kg to $0.0888/kg). As processing costs increase, there is no effect on the disassembler—this cost increase is borne entirely by the shredder. When the cost to shred reaches $0.088/kg, over four times greater than found currently, the shredder’s profits fall to


129

❙


Figure 8 Expected profits with changes in AIV design: increasing the ratio of ferrous materials (m1) to nonferrous materials (m2) for a constant curb weight of 1,143.4 kg.

$36.54/AIV. This is less than the typical profit of $44/vehicle that a shredder earns, but it just exceeds the minimum acceptable profit of $35/vehicle. As a result, the shredder may require an incentive to undertake the processing of these vehicles when the costs to process are high. When compared, however, to the losses of $15.57/vehicle that would be incurred by a shredder processing a steel unibody vehicle at these cost levels, processing the AIV is extremely attractive. Different AIV Designs Future AIVs are likely to feature spaceframe design. Although AIV design data are proprietary, an approximate curb weight can be determined by subtracting 381 lb (Phillips 1996) from the curb weight of the steel unibody (W = 1,316.37 kg). Because the individual masses for the ferrous and nonferrous materials are unknown for the AIV spaceframe design, these masses were approximated from the ratio of ferrous materials (m1) to nonferrous materials (m2) found in the unibody AIV (m1/m2 = 2.55). By varying the ratio of ferrous to nonferrous materials from 2.30 to 3.55 and holding all other variables and the curb weight constant, different materials contents of the AIV can be evaluated. Figure 8 shows the ramifications of the different AIV designs on the profits earned by the disassembler and shredder. The relative content of ferrous to nonferrous materials does not impact the disassembler, whose profits remain constant at $132.35/vehicle. The shredder, however, receives declining profits as the mass of 130


ferrous materials in the AIV increases at the expense of the more valuable nonferrous materials. As long as the ratio of ferrous to nonferrous materials present in the spaceframe AIV meets or exceeds the ratio present in the unibody AIV (2.55, where m2 = 231 kg), the shredder will exceed his minimum goal of earning profits of $35/ vehicle. When the nonferrous content declines such that the ratio of ferrous to nonferrous materials exceeds 3.55 (and m2 is less than 180 kg), the shredder will begin to lose money. If the relative nonferrous content of an AIV is low, the shredder may be unwilling to pay as much for a given hulk. The profits earned by the disassembler, however, are significant, and may offset any loss in revenues due to the sale of the hulk to the shredder. As the AIV curb weight declines further, the profits earned by the disassembler and the shredder declines as well. For instance, when the steel unibody curb weight is reduced by 688 lb (Phillips 1996) to model a compact AIV, and the ratio of ferrous to nonferrous materials is the same as for the unibody AIV (m1/m2 = 2.55), the disassembler receives profits of $76.95/vehicle and the shredder receives profits of $16.60/vehicle. These results indicate that the EOL infrastructure needs to be able to accommodate a broad range of AIV lightweight designs containing differing levels of nonferrous materials.

Discussion To reduce air emissions and fuel consumption, the substitution of lighter materials for steel is expanding. Aluminum is a favorite material be-


cause of its lower density. As AIVs reach their EOL, the current recycling infrastructure can accommodate them profitably and readily. Several scenarios exist, however, that could undermine this successful recycling. Because most of the shredder’s revenues from an AIV would be generated by the sale of a mixture of nonferrous metals, the shredder’s profits are particularly sensitive to any fluctuations in the price. For instance, if the presence of other nonferrous materials results in a low-grade, contaminated aluminum alloy mixture that could not be easily sorted, the shredder would receive less for the materials, and this could jeopardize the shredder’s profitability goals. If new sorting technologies are introduced that can separate the different aluminum alloys, this would result in higher values for the aluminum scrap, but would also represent a new cost to be borne by the shredder or by the nonferrous separator. Although the technologies for sorting, decoating, and processing secondary aluminum are advancing, it is unclear if the increased revenues from the sale of the separated alloys will offset the costs of the additional steps. Because aluminum body panels are more difficult to repair than steel, there is a substantial market for used aluminum parts for the collision industry. As a result, this could lead to the disassembler removing as many aluminum components as possible. One dismantler has a policy of “removing all the aluminum that can be removed” and then selling these parts first to the collision industry if they are in good shape, or selling the parts for the value of their metal (Phillips 1996). If the dismantler is receiving most of the revenues from the aluminum, the shredder’s profits will be reduced. For instance, as demonstrated in figure 3, when the composition of the parts removed by the disassembler exceeds 70% nonferrous materials, the shredder’s profits decline substantially: from $90.40/AIV when the composition of the parts removed is the same as the composition of the vehicle (30% nonferrous, 70% ferrous), to $41.52/AIV when the parts are 70% nonferrous and 30% ferrous. These numbers are based on the assumption that only 15% of the vehicle will be removed as parts by the disassembler. With an AIV, this assumption may be too low, given the desirability of aluminum parts and the value of

❙

the material. Figure 2 explores the relationship between disassembler and shredder profits when increasing percentages of parts are removed from the EOL AIV. When 35% of the curb weight is removed as parts (whose composition is 30% nonferrous, 70% ferrous), the shredder earns profits of $35.20/AIV, a minimally acceptable level for the shredder. If the composition of the parts removed is changed to reflect the greater market for aluminum, the shredder is no longer able to achieve these profits. As a result, the shredder is particularly sensitive to the disassembler’s aggressiveness in removing aluminum parts and materials. One of the problems facing the recycling of aluminum at present is the reluctance of the primary smelters to accept aluminum scrap. The secondary smelters, where ingots for castings are produced, remain the principal market for the scrap. The presence of zinc, copper, or other impurities makes the scrap unacceptable to the primary smelters where wrought aluminum is produced (Phillips 1996). This limitation on the market for the scrap implies that the current price of aluminum scrap may not remain constant if the primary smelters begin to accept scrap, or if sorting techniques are employed that segregate the wrought alloys from the cast alloys. Because body panels made of high-value wrought aluminum may be readily removed by the disassembler, the resulting aluminum shredder scrap, which contains a higher fraction of cast alloys, may be less desirable in the market, and the shredder may receive fewer revenues with a higher fraction of this lower-value material. Another trend that could affect the price of the aluminum scrap is the increasing use of a variety of other nonferrous metals. Magnesium, titanium, and other advanced materials are currently being introduced for automotive applications (Anonymous 1998b; Froes et al. 1998; Hartman et al. 1998), and when a vehicle containing magnesium or some other lighter-weight material is shredded, these nonferrous metals would be mixed in with the aluminum scrap. Unless satisfactory sorting technologies exist, the presence of these other nonferrous materials could adversely affect the price of the nonferrous scrap. As a result, there is considerable uncertainty regarding the future price of scrap alumi-


131

❙


num. Figure 1 demonstrates the impact of various aluminum scrap prices on the profitability of the shredder and the disassembler. When the value of the scrap is $0.44/kg, half the assumed rate, the shredder is just able to match his minimum acceptable profits of $35/vehicle. Any further decline in the value of the scrap, and the shredder may begin to lose money, hence requiring some consideration (e.g., reduced hulk price) from the disassembler, whose profits are affected to a much lesser degree, because they are derived more from the sale of parts, rather than raw materials. If waste disposal costs become higher, this would represent a substantial burden on the shredder, particularly if ASR is classified as a hazardous waste. As seen in figure 6, beyond certain thresholds, the shredder cannot achieve desired profits and may require some kind of incentive to participate in the recycling. The revenue derived from the sale of the aluminum scrap does offset the problem of increased hazardous waste disposal costs, and the shredder is not faced with losses. When compared to a conventional vehicle or a polymer-intensive vehicle, however, this is not the case. In vehicles where steel is used primarily, when ASR is costly to dispose, the shredder faces losses unless the disassembler removes and disposes of some of the polymer (Isaacs and Gupta 1997). Figure 4 shows the effect on profitability when the disassembler removes and disposes of polymer materials. The marginal benefit to the shredder is far less than the marginal cost to the disassembler, and because both are profitable in this scenario, there is no incentive for the disassembler to undertake this disassembly for an AIV. When a vehicle contains substantial amounts of polymers and when ASR is classified as a hazardous material, the recycling infrastructure collapses (Isaacs and Gupta 1997). AIVs represent one of the only scenarios in which the high cost of disposing of “hazardous” ASR may be offset, and the current recycling system is able to process the vehicle and still remain profitable. Another scenario that could be troubling to the shredder occurs when the cost to process hulks increases. The operation of the shredding equipment is energy-intensive, and the shredder’s profits would be linked to the cost of energy; hence, 132


the cost to process the hulk would increase from the current level of $0.0188/kg. As processing costs increase, the shredder’s profits are somewhat affected. It is not until the cost to shred is four times greater than at present, however, that the shredder’s profits for an AIV will decline to the minimum acceptable profit level of $35/vehicle. For instance, when the shredding cost is $0.088/ kg, the shredder’s profits are $36.54/AIV. In most circumstances, the revenues generated from the aluminum scrap make the shredder less sensitive to the cost of processing the hulks, thereby better ensuring that they are fully processed. There are few instances where the disassemblers fail to achieve their profitability goals. If there are few parts that can be sold from the EOL vehicle, the disassembler will not garner sufficient revenues. If the disassembler must remove and dispose of large amounts of polymers— whether to satisfy regulatory needs or the shredder’s requirements—the disassembler’s profits will suffer. Under most circumstances investigated, the shredder is able to generate substantial profits from AIVs, and no change of action would be required of the disassembler. This result aligns with the expectations of those in the recycling industry, where the disassemblers are expected to profit substantially from the AIVs (Phillips 1996).

Conclusions Unless the use of lightweight materials results in a decline in their price as secondary materials—perhaps due to the presence of contaminants—or unless the cost to dispose of the nonhazardous and hazardous portions of the vehicle increases substantially, the existing infrastructure will be able to accommodate AIVs readily. The risks are borne primarily by the shredder, and relate to the desirability of the aluminum parts. Because aluminum body panels are not easily repaired, there will probably be a significant demand for used aluminum body panels by the collision industry, or alternatively these panels could be sold by the disassembler as secondary materials. This transfers some profits from the shredder to the disassembler. To ensure the future usefulness of the secondary aluminum, the automakers should be aware that the use of magnesium and other nonferrous


metals might affect aluminum alloy recyclability and, therefore, the value of the aluminum present in vehicles. Consequently, efforts should be taken to ensure that complementary materials are used in areas where disassembly is unlikely, and when dissimilar materials must be used, the automakers should support development of costeffective means for separation and identification. If automakers are responsible for their vehicles from cradle to grave, this knowledge could promote the initial use of easily recycled materials, thus facilitating materials reclamation at the EOL, and this would significantly impact the amount of materials to be landfilled. Furthermore, to retain the value of the wrought alloys, the technologies to separate and identify the various alloys must be further developed, and the corresponding use of wrought-aluminum scrap by the primary smelters must also be promoted.

Notes 1. 1176 kilograms ≈ 2593 pounds. 2. CAFE is a fuel economy standard established in U.S. law that requires each automaker to meet a harmonic average fuel economy for all cars sold.

References Aluminum Association. 2000. Accessed September 18, 2000. Anonymous. 1995. The war on waste: There’s no armistice when it comes to recycling. Ward’s Auto World: 61. Anonymous. 1998a. 1999 passenger car and light truck aluminum content report highlights. Bloomfield Hills, MI: Ducker Research Company. Anonymous. 1998b. Magnesium in car design: Cost vs. weight savings. Automotive Engineering International, August, 53–58. Anonymous. 1998c. Technology to cleaner, lighter, safer vehicles. Automotive News, 16 February, 22L–22M. Anonymous. 1999. MIT study analyzes environmental impact of aluminum-intensive vehicle fleet. JOM 51(8): 4–6. Ashley, S. 1994. Aluminum vehicle breaks new ground. Mechanical Engineering 116(2): 50–51. Bhakta, P. 1994. Recent technology and trends in automotive recycling. JOM 46(2): 36–39.

❙

Brown, R. D., F. Ambrose, and D. Montagna. 1985. Separation of cast and wrought aluminum alloys by thermomechanical processing. U.S. Bureau of Mines. Chen, A. 1995. A product lifecycle framework for environmental management and policy analysis: Case study of automobile recycling. Ph.D. Thesis, Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, MA. Chen, A. C. and F. R. Field. 1993. Automotive polymers and their technical and economic effects on automobile recycling. In ASM/ESD Advanced Composites Conference, Dearborn, MI, 681–697. Field, F. R., J. R. Ehrenfeld, D. Roos, and J. P. Clark. 1994. Automobile recycling policy: Findings and recommendations. Davos, Switzerland: International Motor Vehicle Program, Massachusetts Institute of Technology. Froes, F. H., D. Eliezer, and E. Aghion. 1998. The science, technology, and applications of magnesium. JOM 50(9): 30–34. Hartman, A. D., S. J. Gerdemann, and J. S. Hansen. 1998. Producing lower-cost titanium for automotive applications. JOM 50(9): 16–19. Hogan, B. 1998. Steel still rules automobile manufacturing. Manufacturing Engineering 120(6): 42–49. Isaacs, J. A. and S. M. Gupta. 1997. Economic consequences of increasing polymer content for the U.S. automobile recycling infrastructure. Journal of Industrial Ecology 1(4): 19–33. Jody, B. J., E. J. Daniels, and J. A. Pomykala. 1996. Progress in recycling of automobile shredder residue. In The Annual Meeting of the Minerals, Metals and Materials Society, Anaheim, CA, February 4–8. Keebler, W. W. 1999. Audi celebrates sports car coup: TT coupe blends style, technology. Automotive News, 3 May, 28. Kirchain, R. 1999. Modeling methods for complex manufacturing systems: Studying the effects of materials substitution on the automobile recycling infrastructure. Ph.D. Thesis, Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, MA. Kurihara, Y. 1994. The role of aluminum in automotive weight reduction—Part III. JOM 46(5): 12–13. Litalien, M., F. Tremblay, A. Rosenfeld, M. Thomas, and A. Gesing. 1997. Processes for the recycling of sheet and other wrought alloys from aluminum intensive vehicles. In Applications for Aluminum in Vehicle Design, February, 45–53. Mangin, C. G. E., J. A. Isaacs, and J. P. Clark. 1996. MMCs for automotive engine applications. JOM 48(2): 49–51.


133

❙


Mascarin, A. and J. Dieffenbach. 1992. Fender material systems: A lifecycle cost comparison. In Automotive Body Panel and Bumper System Materials and Design. Warrendale, PA: Society of Automotive Engineers, Inc. Naggar, J. A. 1992. Corporate average fuel economy: An evaluation of the key concerns and conflicts in revising the standards. Warrendale, PA: Society of Automotive Engineers, Inc. Nijhof, G. H. and P. C. Rem. 1999. Upgrading nonferrous metal scrap for recycling purposes. JOM 51(8): 20–23. Oye, H., N. Mason, R. Peterson, N. Richards, E. Rooy, F. S. McFadden, R. Zabrezmol, F. Williams, and R. Wagstaff. 1999. Aluminum: Approaching the new millennium. JOM 51(2): 29–42. Phillips, M. 1996. Automotive aluminum challenges steel. Recycling Today, January, 24–38. Rechtin, M. 1999. Toyota puts spunk back in the Celica. Automotive News, 2 August, 22–23.

134


Robinson, A. 1999a. Aluminum Alliance seeks 40 percent weight cut. Automotive News, 7 June, 4. Robinson, A. 1999b. Package deal: With aluminum, you get engineers. Automotive News, 2 August, 16i–18i. Steuer, R. 1986. Multiple criteria optimization: Theory, computation, and application. New York: John Wiley & Sons. Stoffer, H. 1999. 31 Senators back higher CAFE. Automotive News, 7 June, 8. Taylor, B. 1999a. Providing the profits. Recycling Today, August, S4. Taylor, B. 1999b. Wringing out the metal. Recycling Today, November, 28–78. Winter, D. 1996. Toyota recycles ASR. Ward’s Auto World, 1 February, 57. Yamashita, H. and M. Hirano. 1998. Research on the application of aluminum door beams for automobiles. In Aluminum in Automotive Application. Warrendale, PA: Society of Automotive Engineers, Inc.