Product remanufacturing develops rapidly in recent decades due to intensified environmental legislations and economic concerns. Through remanufacturing, products/components that would otherwise head to land-fill or incineration will instead go through a set of value and material recapturing processes, including distribution, inspection, disassembly, cleaning, testing, repair, reassembly, redistribution, and remarketing or recycling. Remanufacturing allows for reusable components and recoverable materials reenter the supply chain for future reuse or new product fabrication.
Although there are numerous definitions of remanufacturing, R. Lund, describes remanufacturing as "… an industrial process in which worn-out products are restored to like-new condition. Through a series of industrial processes in a factory environment, a discarded product is completely disassembled. Useable parts are cleaned, refurbished, and put into inventory. Then the product is reassembled from the old parts (and where necessary, new parts) to produce a unit fully equivalent and sometimes superior in performance and expected lifetime to the original new product." Note that remanufacturing differs from simple repair or recovery in that a remanufactured product should meet the same customer expectation as new products in quality, warranties, life span, and functions.
Generally, the production processes of remanufacturing are comprised of nine stages: product arrival, inspection, disassembly, cleaning, testing, repairing (reconditioning), reassembly, labeling, packaging, and shipping. The flowchart of the overall process is demonstrated in Figure 1. Material arrival process for remanufacturing is a typical compound stochastic batch arrival process with varied product types and conditions. Thus, in receiving, incoming product to be remanufactured have to be classified according to its type and condition. Received products will then be briefly inspected to collect related information, and sent to inventory.
For received products, cleaning processes are required to separate undesired dirt, coating, or other contaminants from the parts to be remanufactured. Following this, the testing process is carried out to investigate the condition of the product and assign appropriate remanufacturing processes which include refurbishing, repair, reuse and material recycling. Finished goods will be labeled, packed and shipped for resale. Components that cannot be reused will be further disassembled and classified according to their material contents, and then shipped for material recycling.
Figure 1 Production flowchart
2. Remanufacturing Processes
Although processes involved in remanufacturing of different products are similar for different industries, it should be noted that due to different manufacturing technology, material and assembly methods used, remanufacturing processes vary from one industry to another. For example, auto remanufacturing industry usually employs aggressive process for part cleaning such as liquid based cleaning process to remove oils, greases, carbon deposits, and coating that was purposely engineered to last for years, thus they are hard to be separated by regular cleaning techniques. On contrary, cleaning of electronic products can employ a simple wind blast process. As such, we intend to discuss special processes used in auto remanufacturing.
2.1. Remanufacturing processes for Auto Industry:
2.1.1 Cleaning processes:
Cleaning process of mechanical parts can be grouped in to two categories: liquid based cleaning process and mechanical based processes. In liquid based processes, parts are cleaned by solutions through mechanisms such as wetting and other chemical reactions. While in mechanical processes external force is applied to the part being cleaned to separate undesired layers from the parts. It should be noted that liquid based cleaning can also employ mechanical energy to achieve a better cleaning effect.
Cleaning mechanism of liquid based cleaning method
Most of the liquid based cleaning techniques rely on following mechanisms to achieve effective cleaning: wetting; emulsification; solubilization; saponification; deflocculation; and sequestration.
Wetting mechanism is essential to any liquid based cleaning. It delivers the cleaning chemistry to contaminants to be separated. Through wetting, substrate-soil bonds are broken, so that mechanical energy can be delivered to displace and remove the contaminants. Wetting can also reduce undesired surface and interfacial tensions, allowing cleaning agent to penetrate between the contaminant and the substrate.
Emulsification is the dispersion of oils to be removed in the solvent. The main factors of emulsification include types of oil and the surfactants selected. The pH level and temperature can also affect the level of emulsification. Mechanical energy, such as vibration, ultrasonics, and turbulence are generally employed to enhance the emulsion effect. Note that emulsification does not change the chemical characters of the contaminants; however it is essential for most cleaning process in effective separation of the contaminants from the substrate.
Solubilization is a process to enhance the solubility of the contaminants in a particular solution using surface-active agents. Solubilized contaminants are then dissolved into the solution. In a typical cleaning process, cleaning agents generally solubilize a certain amount of contamination while additional contaminant is held in suspension by emulsification.
Saponification is the reaction of any organic oil containing reactive fatty acids with free alkali to form soap. Alkaline cleaners containing saponifiers rely on this process to remove some oils, including vegetable and animal fats and their derivatives. The soaps that are generated are easily removed by subsequent rinsing with water.
Deflocculation causes the breakdown of contaminants into very small particles that are then dispersed in the liquid cleaning medium and swept away. This process is similar to emulsification except it happens on a larger scale.
Sequestration is a process where undesirable ions, such as Ca+2 or Mg+2, and heavy metals are de-activated, preventing them from reacting with material that normally would form insoluble products. The classic example is the hard water scum formed when soaps are used. The scum formed is the reaction between the Ca+2 or Mg+2 ions in hard water with soap. When the water is softened, the Ca+2 or Mg+2 ions become tied or sequestered and are unable to react.
Liquid based cleaning process:
For any cleaning processes, proper cleaning equipment is required to implement the cleaning mechanisms described previously. The cleaning equipment provides not only the site for accomplishing the cleaning process; it can also provide other desired functions such as separation and collection of removed coating or dirt. In addition, most of the cleaning equipment integrates heating or mechanical vibration to provide external agitation that enhances the cleaning effectiveness.
The goal of agitation of the cleaning solution is to apply external energy to the part surface so that the cycle time and effectiveness of the cleaning process can be enhanced greatly. Agitation can be achieved by simply stirring solution with rotary stirrers. Similar effect can be achieved by rotating parts inside the solution. Stirring agitation is gentle in general and does not significantly improve cleaning effectiveness unless the chemistry is very aggressive. Nonetheless, due to its simplicity and easy to implement, it can be applies in most processes. Ultrasonic agitation uses high-frequency sound waves to achieve mechanical agitation. Ultrasonic waves can also penetrate thin layers of metal and propagate around corners to clean work pieces inside and out. Ultrasonic cleaning is usually not appropriate for thick buildups of contaminant.
Mechanical Vibration is another way of imparting energy to the cleaning site. Parts are literally moved up and down or side to side while immersed in the cleaning liquid to create shear forces between the liquid and the part surface. The more rapid the agitation, the more effective agitation becomes. Because of hydraulic "pumping" of the cleaning liquid through internal passages, part agitation can also be an effective means of cleaning inside some parts with appropriate configurations.
Based on the solution and external energy sources used, cleaning processes can be grouped as follows:
Immersion cleaning
Immersion cleaning refers to a group of the most applied cleaning methods for mechanical parts. It generally uses cleaners with high concentration. Convection current combined with external vibration, soils are removed from metal surface conveniently. This cleaning approach is particularly good for cleaning irregular shapes, box sections, tube and cylindrical configurations that cannot be penetrated using spray systems. The operation may vary from hand dipping a single part or agitating a basket containing several parts in an earthenware crock at room temperature to a highly automated installation operating at elevated temperature and using controlled agitation.
Several approaches of immersion cleaning are summarized below:
- Barrel cleaning: this approach is generally used for cleaning large quantities of small parts. Parts are placed and agitated inside a barrel that rotates in the cleaner solution.
- Moving conveyor cleaning: in this approach, parts are placed on a moving conveyor which moves parts through solution flow.
- Mechanical contact: Cleaner is applied with brushes or squeegees.
- Mechanical agitation: in this approach, parts are flooded with solution which is circulated using pumps, mechanical mixers, or ultrasonic waves.
-High pressure agitation: in this approach, a high pressure solution flow generated by pumps is applied to the parts to clean deep and blind holes as well as tubes with a small diameter.
Ultrasonic cleaning
Ultrasonic cleaning employs high frequency ultrasonic waves (20-40 KHZ) passing through liquid solutions to assist effective cleaning. Due to the gas bubbles created by ultrasonic waves inside the cleaners, ultrasonic cleaning can provide strong cleaning effects on the parts immersed in the solution. Ultrasonic cleaning is ideal for parts with complicated shapes, surfaces, and cavities that may not be easily cleaned by traditional immersion techniques.
The basic ultrasonic cleaning process generally is composed of following components: the cleaning tank, ultrasonic transducers, and the power supply.
Another similar cleaning technique is Megasonic cleaning. It uses a much higher frequency (700-1,000 kHz) acoustic energy to generate pressure waves in a liquid. Compared with ultrasonic, megasonic technique does not suffer from cavitations which is a typical drawback for ultrasonic. Less cavitations reduce the likelihood of surface damage.
Molten salt cleaning
Molten salt cleaning that has been developed for quite a few decades is widely used in nowadays industrial section. This technique is based on thermochemical or electrochemical reaction between the melt and contaminants. Compared with other immersion cleaning technologies, molten salt cleaning has many advantages that make the cleaning process more effective and faster. The first one is there is no evaporation or loss of liquid from the bath at the operating temperature because of their utterly low vapor pressure. This feature leads to less resource consuming and also the second advantage that attracts peoples' attention. That is there are no evaporative energy losses. The third comes from the high heat transfer rate and the high heat capacities of molten salts. This characteristic can make working parts get to desired temperature quickly and protect them from over-shooting. Finally, molten salt cleaning would be great helpful for post NDT inspection discussed later.
A typical salt bath equipment generally includes a sludge receiving station, salt bath furnace, water rinse tank, acid tank, and second water rinse tank. In lots of factories, this cleaning process has been totally automated.
2.2 Mechanical cleaning
Another group of cleaning technology is based on employment of mechanical force to separate contaminants from the substrate. The mechanical force can be in the forms of air blowing or exhausting, vibration, abrasion using brushes or small hard particles blasted by air. Since no chemical reaction occurs during the cleaning process, one of the most attractive benefits of mechanical cleaning is less hazardous emissions. However, due to strong mechanical forces applied to parts to be cleaned, it is also possible to damage the substrates.
Vibration cleaning
Vibration cleaning utilizes high frequency rotary oscillation to create strong vibration that overcomes the adhesive force so that dirts are separated from the parts. The dirts separated can be exhausted to a special container and can be reused. Additional abrasive bush can be combined with the vibration movement to enhance the cleaning effectiveness and reduce cycling time.
Abrasive cleaning
Abrasive cleaning use high speed propelling blade shot small hard particles on the part surface, thus cleaning contaminants by impact force. The particles used as abrasive media vary in types and sizes to meet specific cleaning scenarios. Abrasive cleaning is most commonly used method to remove heavy scale and paint on large easy to access parts. Major components of the Centrifugal blast machines include: blast wheel, work conveyor, abrasive recycling system, and a dust collection device.
Dry-Blast Cleaning
Dry-blast cleaning is also called abrasive blasting cleaning. Dry blast cleaning is considered as the most efficient and environmentally effective method for abrasive cleaning. It generally employs a 685 kPa air supply system to propel abrasive particles to separate contaminants from the parts. Different replaceable air-blast nozzles are developed with different shape and wear resistant materials. Although all metals can be cleaned through abrasive blasting processes, one should carefully select suitable abrasive medium for soft and brittle metals such as aluminium, magnesium, copper, zinc, and beryllium, to avoid damage to the part itself.
With respect to the equipment available for dry blast cleaning, people developed several types based on different material handling approach:
Cabinet machines: A cabinet is used to contain the abrasive-propelling mechanism, holds the work in position, and confines flying abrasive materials and dust. Cabinet machines may be designed for manual, semiautomatic, or completely automated operation to provide single-piece, batch, or continuous-flow blast cleaning. A single blasting cabinet is shown in figure 2.
Figure 3 A Continuous-flow Blasting Machine
Figure 2 A Blasting Cabinet
Continuous-flow machines: compared with cabinet machine, continuous flow machine uses proper conveying devices to continuously clean parts in the cabinets. These machines are used to clean coils and wires as well as castings and forgings at a high production rate. Combined with an abrasive particle recycling system, it can reuse the blast particles. Figure 3 is a typical continuous-flow blasting machine.
CO2 dry ice blasting
CO2 dry ice blast is a special dry blasting method in that it uses frozen CO2 particles or snow as abrasive media. Some parts may be sensitive to thermal changes from the pellets and should be tested first. While particles can be clean the surface at a faster rate, it can also damage the surface being cleaned. The advantage of the CO2 dry ice blasting is that they sublimate on contact with the material to be cleaned. Figure 4 shows the basic pelletization process and the removal of coatings.
Removal of Coatings
Basic Pelletization Process
Figure 4 Dry Ice Blasting
3. Nondestructive Testing
After parts are cleaned, inspection and testing procedures are followed to check if repair is required. Since parts to be remanufactured are always in different conditions, testing is generally unavoidable. Since the purpose of remanufacturing is to reuse the parts, most of the testing methods are not intended to create any damage to the part being tested. As such called, Nondestructive testing is a commonly used technique to reveal flaws and defects in a material or device without damaging or destroying the test sample.
Since nondestructive testing (NDT) is a wide group of analysis techniques used in science and industry to evaluate the properties of a material, component or system without causing damage, we would like to summarize methods being used in auto remanufacturing industry.
3.1 Methods for Nondestructive Testing
NDT methods employ techniques such as microscope, electromagnetic radiation, sound, and combined with the inherent properties of materials to detect flaws or discontinuities in the parts to be remanufactured. Microscope method is generally used to examine external surfaces of the part being tested. To test the inside of the part, methods such as electromagnetic radiation and liquid penetrant testing are generally used to examine fatigue cracks. For liquid penetant methods, a certain liquid is applies to penetrate and reveal the cracks. For non-magnetic material, fluid with fluorescent or non-fluorescing dyes is commonly used. For magnetic material, an externally applied magnetic field or electric current through the material is used. When parts have cracks, magnetic flux will leave at the area of the flaw, resulting in leakage of magnetic field at the flaw area. This leakage can be captured and used as an indication of flaws.
NDT can be further classified in to various methods and techniques. It is important to select the right method and technique for a specific part or material to ensure the performance of NDT.
Visual Testing
With its simplicity, super-low cost, convenience and other advantages, visual testing (VT) has already been applied as the most common [1] nondestructive testing method consciously or unconsciously for several centuries. It can be used as an independent testing or an auxiliary technique to make up the inadequacies of other NDT methods, such as liquid penetrant testing. The most important prerequisite to VT is pre-cleaning of the parts or assembles received in order to provide smooth and contaminants-free surfaces. There are also plenty of tools that can support the testing; they are tapes, lights, gages, magnifiers, video cameras, mirrors and so on. With these tools, cracks, corrosions, blisters, wear or physical damage existing on the surface of the parts as well as their warpage or deformation [2] could be found. However, tiny holes and inner flaws can't be detected without other NDT methods.
Liquid Penetrant Testing
Liquid penetrant inspection (LPI), or called dye penetrant inspection (DPI), applies the capillary effect of penetrants to display cracks or porosities that are invisible to naked eyes on the surfaces of nonporous materials, like metal, plastic or ceramics [3] . A general liquid penetrant test usually takes the following test procedure (shown in figure 5):
Figure 5 Procedures of Penetrant Testing [2]
1. Pre-cleaning: cleaning methods discussed earlier are used to remove any dirt, oil, grease or any other contaminants to ensure that any defects are open to the surface, dry, and free of contamination. This step is critical to this kind of testing while improper cleaning of surface leads to invalid defection or missing flaws.
2. Application of Penetrant: After parts are cleaned, penetrant, either fluorescent or visible dye is then applied to the surface. A certain period of time (5 to 30 minutes) is required to allow the penetrant to immerse into any flaws. The length of the penetration time depends on the penetrant being used, the type of material being tested, and the size of flaws being examined. Generally, smaller flaws require a longer penetration time. Excess penetrant has to be removed from the surface of the part being tested.
3. Application of developer: A developer is a chemical that draws penetrant from defects so that defects can be identified. From the stains that show up in the developer one can identify the positions and types of defects on the surface under inspection.
4. Inspection: In inspection, visible light is applied for visible dye penetrant. In contrary, for fluorescent penetrant, ultraviolet radiation is applied to the part surface being examined.
5. Post Cleaning: Cleaning is required to remove penetrant after inspection and recording of defects are finished.
To reduce the chance of misusing LPI and invalid indication, proper training for operators may be required. And attentions should be focused on thorough pre-cleaning, careful removal of excess penetrant, enough penetration and diffusion time of penetrant and observation.
Magnetic Particle Testing
As to magnetic penetrant testing (MPT), fine iron or magnetic particles, held in suspension in a suitable liquid are used as penetrant. For better performance of the inspection, the particles are usually colored and coated with fluorescent dyes visible under ultraviolet light. To apply the penetrant, the liquid suspension is sprayed or painted on to the part which is magnetized. Due to magnetic leakage at the defect area, the magnetic particles are attracted in the area of the defect. When UV light is applied, the location and size of the defect can be easily identified. Like LPI, post cleaning of particles is required while demagnetization may be added. One of limitations of this method is its application field in which there are only ferrous or ferromagnetic materials. However, it is used much more extensively than LPI for these materials. Magnetic penetrant testing method is generally a low cost inspection method and is much faster than ultrasonic testing and radiographic testing. Figure 6 shows some results of MPI under ultra-violet light.
Figure 6 Magnetic Particle Inspection
Radiographic Testing
Radiographic Testing (RT) methods use short wavelength electromagnetic radiation to penetrate materials and reveal defects. Typical radiation source is an X-ray or gamma ray machine. Since the amount of radiation emerging from the opposite side of the material can be detected and measured, variations in the intensity of radiation are used to determine thickness or defect of material. This fundamental principle (shown in figure 7) also gives this NDT method a big disadvantage that the access to the both sides of part or testing section becomes a mandatory necessity.
X-ray Film
Figure 7 Radiographic Inspection
Different from previous introduced NDT methods, radiographic inspection has the ability of detecting internal flaws and is suitable and practical to nearly all materials with no consideration of its high cost. And compared with penetrant testing, RT does demand stricter and higher degree of training since it might be of great hazardous to human and surroundings if it is operated improperly.
Eddy Current Testing
Based on the theory of eddy current, Eddy Current Testing (ECT) is widely used in detecting discontinuities in conductive materials, especially heat exchanger tubing. The principle is depicted in figure 8. When alternative electricity goes through the coil, eddy current would be generated and changed with the intensity of the current in the coil. Simultaneously, any change in the conductive specimen affecting the eddy current can also influence the current within the coil. By detecting this change, some surface and subsurface flaws could be indicated.
Figure 8 Principle of Eddy Current Inspection
An advantage over the other NDT methods is its permeability of coatings. However, ECT cannot penetrate through the whole thickness of parts and the penetrate depth could be stated in the following equation [4] :
δ =
4. A Disassembly Tree Method for Disassembly Process Planning:
The process of remanufacturing generally starts from product disassembly, which is important to residual value recovery and reduction of environmental impact resulted in recycling processes. Disassembly analysis in this regard, addresses three issues: (1) Optimal disassembly strategy that recovers maximum residual value, (2) Disassembly sequence planning, and (3) evaluation of disassembly time, cost, and disassembly difficulty rate, with component information provided.
The disassembly relationships among the components of a product to be remanufactured include component-fastener relationships and precedence relationships. Therefore, two types of graphs are needed in order to fully represent the relationships among the components of a product, namely, component-fastener relationship graph and precedence relationship graph.
Fasteners are used to attach one component to another for the purpose of assembly. Examples of fasteners include screws, rivets, inserts, etc. In a component-fastener graph, The components are represented as the vertices , where n is the number of components. Their relationships are represented as the edges , where m is the number of edges. If two components vi and vj (i j) are jointed by fasteners, then (vi, vj) E; otherwise (vi, vj) E. The graph Gc is an undirected graph. Vertices and edges in graph Gc are modeled using object-oriented techniques. While the object vertex consists of component information including its name, weight, material type, etc., the object edge consists of fastener information including the number of fasteners, fastener type, etc. For example, Figure 2 a. shows component-fastener graph of a personal computer.
Precedence graph represents the precedence relationship among the components of a product, namely, a component cannot be disassembled before certain components. Figure 2 b. shows the precedence relationship graph.
(a)
(b)
Figure 2. a: Component-fastener graph for the assembly b: Precedence relationship graph for the assembly
Disassembly tree can then be constructed based on the component-fastener graph and precedence graph. The disassembly tree consists of vertices representing an assembly or a component and information such as its name, material type, weight/volume. A vertex is decomposed into child vertices representing its child sub-assemblies or components. An edge, linking a child vertex with its parent vertex, represents the disassembly relationship between two components and information about assembly method.
The disassembly tree is constructed through searching of cut-vertices in the component-fastener graph. A cut-vertex is a vertex whose removal disconnects the graph. If a cut-vertex is found, the graph is split into two or more sub-graphs. The same procedure is repeated until no cut-vertices can be found. In this way, a pseudo-disassembly tree is generated which is showed in Figure 3 a.
The pseudo-disassembly tree is then modified by the precedence of the disassembly according to the precedence graph, and a disassembly tree can be obtained as illustrated in Figure 3 b.
In disassembly sequence planning, a popular assumption is that EOL products should be disassembled to the fullest extent possible (Beasly and Martin 1993, Zussman et al. 1994). However, based on a recent survey in the recycling industry, such assumption is not practical in many cases due to the high cost of disassembly. It is very important to find the optimal level for disassembly where the benefit of reverse manufacturing is maximized and the cost is minimized. The disassembly sequence planning can be determined after such a termination point.
a
b
Figure 3. a: Pseudo-disassembly tree. b: Disassembly tree for the assembly
Optimal disassembly planning is determined based on the cost and profit. Three types of costs and one type of profit are addressed: (1) disassembly cost which includes labor and tooling cost, (2) material reprocessing cost, i.e. cost of recycling (3) disposal cost, which includes transportation fee and landfill cost, and (4) salvage profit, which is the profit gained by means of component reuse or recycling. The cost model for determining the termination of disassembly is illustrated in figure 3.
Figure 4. Optimal disassembly termination analysis
The total cost is calculated as the sum of disassembly cost, material reprocessing cost, disposal cost, and salvage profit. The lowest point of the curve (f) representing the total cost determines the termination of disassembly where the cost is minimized, in other words, the benefit of disassembly is optimized.
5. Remanufacturing Production Planning:
The most significant characteristic of remanufacturing production system is its unstable and uncertain incoming flow. The returned products generally have a high uncertainty in arrival pattern and high variation in product type with disparate residual value. Quantity, year of model, and quality of returned products are also subject to high uncertainty. For example, the product might come from a software company that updates its computers every three months or they might come from a family replacing its 10-year-old home computer. The consequence of high uncertainty and variation of the return flow is the difficulty associated in production planning and control of the remanufacturing, which leads to increased production cost and poor economic performance.
Another major challenge of remanufacturing comes from the distinct role of the receiving inventory. On one hand, it differs from traditional ones in that customers return their post consumer products to the inventory instead of taking the product away from the inventory. In this regard, inventory is used to meet the product return demand. A redistribution cost, which does not exist in a forward manufacturing system, is incurred when the remanufacturer finds no inventory space to handle the returns. On the other hand, receiving inventory can still act as a buffer to dampen the randomness of material arrival process, thus providing a relatively stable material flow for the reverse production. However, replenishment of stocks (post consumer products) is a stochastic process with high uncertainty, while remanufacturer has little control over it. This generally results in huge safety inventory for the remanufacturers.
As in forward manufacturing, operations and processes of remanufacturing should also be aligned and optimized to maximize the total profit. This leads to following three production planning problems that need to be addressed:
1) First of all, remanufacturing system has to handle substantial number of product types. Generally these different products share one production line. Therefore, a priority based switch rule has to be developed for production planning to determine how and when to switch between different production types. The priority mechanism is generally based on following concerns. The first concern is the depreciation rate of the products or components received. Products with the highest depreciation rate should be given first consideration. The second concern is the residual value of the product. Generally, products with a high residual value should be processed first. The third concern is the environmental impact. If the product has in-transition environmental impact, it should also be processed early. The fourth concern is the market demand. If the secondary-market demand for a certain remanufactured product or component is higher, these products should be handled first. In determining when to switch, production lot size for different products with different priorities have to be determined and optimized to reduce total holding cost, set up cost and redistribution cost.
2) The second issue for remanufacturing planning and control is determination of the optimal receiving inventory capacity and safety stock level. On one hand, receiving inventory capacity set a constraint of safety stock level and the possible production run size. It also has significant impact on both stability of production and redistribution cost. With more receiving inventory space, a higher level safety stock can be allocated to improve the stability of reverse production system. This could result in better efficiency. More receiving inventory space will also reduce the chance of redistribution and associated cost. Nonetheless, excessive inventory capacity also has shortcomings - large inventory capacity increases the space cost, while higher safety inventory results in higher inventory cost.
3) The third problem is to determine the optimal workforce level and production capacity. The unstable and uncertain incoming flow of the dedicated model requires workforce level and production capacity respond to the product return demand so that excessive capacity can be avoided. However, changing capacity of any production system will always incur costs.
Obviously, effective modeling and analysis of the production model of remanufacturing system is critical to attack the problems discussed. Approaches such as Queuing networks or other mathematical modeling techniques are possible options. However, due to the special stochastic characteristics of the arrival process and the priority based switching rules in production planning, the Quequeing model has to consider both the compound bulk arrival and the priority Queuing. Analysis of priority queue with compound bulk arrival has shown to be very hard to solve. Optimization with simulation methods proved to be an effective approach and can be used in optimization of a system that possesses the characteristics described in a remanufacturing system.
