Comprehensive Model to Evaluate the Impact of ... - Springer Link

15 downloads 126460 Views 246KB Size Report
Life Cycle Model; Dedicated tools; Process-based models. 1 INTRODUCTION .... the product is the sum of all part/tool life cycle costs and environmental impacts.
Comprehensive Model to Evaluate the Impact of Tooling Design Decisions on the Life Cycle Cost and Environmental Performance Inês Ribeiro, Paulo Peças, and Elsa Henriques IDMEC, Instituto Superior Técnico, UTL, Lisbon

Abstract The design and production of dedicated tools is usually a technically complex and demanding activity, as it highly influences the manufacturing process in which they are used, from material and energy consumption to reliability and cycle time parameters. However, despite being acknowledged by the industry, it is still lacking the quantification of these impacts due to the “one of a kind” nature of these tools. The model proposed in this paper overviews the research work under development, aiming to quantify the tool life cycle impacts. By using process-based models the costs and environmental impacts are assessed and modelled to be sensitive to part and tool design changes. Keywords: Life Cycle Model; Dedicated tools; Process-based models 1

INTRODUCTION

In an increasingly competitive environment, the traditional manufacturing business focused on physical products has been shifting to embrace an integrated product-service approach. Companies are no longer just responsible for the physical product, but also for the entire product’s life cycle. In this context Life Cycle Cost (LCC) has been widely used for the last decades in a vast variety of products, projects and equipment [1, 2, 3]. In fact, in the literature there is no standard method to perform a LCC analysis and there are a vast number of proposed approaches varying in their form and scope. However, there is a common aspect regarding in most product LCC analysis: the scope is on either on the product design and on the impacts of design changes or on design alternatives throughout its life cycle [4, 5]. Some authors are focused on the manufacturing phase mainly to evaluate equipment disregarding the product design, being the product specifications only a static input to the model that evaluates only the manufacturing equipment able to produce it [6, 7, 8]. Nonetheless, a product is made of parts and almost every part requires dedicated tools in the manufacturing phase. Tools are often disregarded or included in product LCC studies as a cost input, despite the fact that the tool design, along with the product design, highly affects the manufacturing phase [9, 10]. These designs are not independent, as the product design also constrains tooling engineering solutions albeit leaving open several tool design alternatives to produce the same product. In summary, the tool life cycle enters the production phase of the product life cycle and affects part production regarding material waste, energy consumption and productiveness, among others, much beyond the mere tool investment. This leads to the need to fill a gap in LCC models developed so far. Therefore in this study a model that integrates the tool life cycle in the product life cycle is proposed. In parallel, the tool also affects the environmental impacts by affecting energy consumption and material waste. Having in mind the goal of developing a model that captures the most relevant impacts of these design decisions, it is also necessary to assess the environmental burdens of the product and tool life cycle in an early design stage. This is increasingly important given the fact that 70% of the final cost/impact of a product is determined at its design phase [11]. The vast and strong links between the tool design and the product design in addition with their effects on the performance of

manufacturing processes bring out the fact that the integration of the two life cycles cannot be a simple sum of the tool life cycle cost and environmental impacts to the product’s LCC and LCA. Moreover, a dedicated tool is a one of a kind device, being difficult to quantify its influence in the part production phase, as there is a lack of data regarding a specific tool. It is relatively easy to estimate its production cost, but it’s increasingly harder to estimate impacts going forward in the tool and product’s life cycle. This is extendable to the environmental impacts, as many of the required information, in some extension, is identical for both analyses. A possible approach to solve this problem is to use process-based cost models, which regard cost as a function of technical factors, such as cycle time, downtime, reject rate, consumed materials, equipment and tooling requirements. Process-based cost models are usually applied to a manufacturing process and model the costs considering the net of influences between technologies, operations and economic based variables. This type of models has been applied by researchers to several processes within different scopes, always with the intent to compare alternatives – either in materials, processes or product architectures [12, 13]. The philosophy and knowledge used in process-based cost models can also be used to analyse the environmental burdens. The technical factors are linked to environmental impacts, translating this way the material and energy requirements into environmental indicators. Furthermore, an additional aspect must be taken into careful consideration when integrating the life cycle of the tool into the product life cycle. If tool design influences the performance of the part manufacturing process, then tool reliability is a crucial aspect to understand. Manufacturing processes highly depend on machine reliability [14]. Whilst machines are generally standard, a great amount of data is available and reliability data is often easy to find, the same does not happen as regards dedicated tools. The reason, already explained, is that a dedicated tool is a non-standard device materializing a unique engineering solution. However, it is possible to estimate these parameters through experts’ knowledge, using a method used for machine and tools components. Some authors have developed methods to assess tool reliability through the Weibull distribution [15, 16] that can be used in evaluating dedicated tools reliability. This approach integrates as a sub-model the global LCC model, contributing to the enrichment of the LCC methodology. Other two models are also proposed for two other important aspects of dedicated tool performance: tool maintenance scheduling and energy

19th CIRP International Conference on Life Cycle Engineering, Berkeley, 2012

I. Ribeiro, P. Peças, and E. Henriques

516 consumption at the tool use/part production. Again, the one-of-a-kind characteristic of dedicated tools is the main reason driving the integration of these two models. The tool under design might never have been used, however there is empirical knowledge available at the product/tool design phase that can be used. 2

COMPREHENSIVE LIFE CYCLE MODEL (CLC)

This model proposes to capture the impacts of products and tools design decisions, linking both life cycles. As a product can be composed by several parts, each part requiring specific materials, manufacturing processes with specific tools and enabling different End-Of-Life (EOL) scenarios, several life cycles are involved. Hence, the proposed Comprehensive Life Cycle (CLC) model is applied to each part; and the total life cycle cost and environmental impacts of the product is the sum of all part/tool life cycle costs and environmental impacts. Starting with the scope of the analysis for each part, the life cycle stages to analyse are defined, as well as the manufacturing processes included in each phase. The subsequent step is to model each process using process-based models, integrating additional models when relevant in order to further explore critical aspects. These are for example tool reliability, energy estimation and maintenance models. Finally, all costs and environmental impacts are computed according to the resources requirements driven by the design specifications. 2.1

2.2

Cost Analysis of Processes

In the proposed model, processes involved in each life cycle stage are modelled using process-based cost models in order to compute the resources requirements and, further on, the costs. This is a demanding task, as a huge amount of manufacturing processes can be involved and furthermore relations between them are also considered. For this task engineering knowledge of each process is required, as the modelling is not simply a cost accounting procedure. Process-based cost models start from the description of the intended product (part material and geometry). The process(es) required for its production is(are) then modelled regarding the cycle time, resources (equipment and labour) specifications, etc. This can be obtained with theoretical and empirical relations correlating the properties of the part and the requirements of the involved technologies. By adding inputs regarding the operating conditions of a certain plant it is possible to build up the operations description, which allows computing the needed resources regarding the number of tools, equipment, operators, etc. (or, as far as equipment and operators might not be dedicated, its time consumption). Finally, having modelled all the processes and the required resources to produce the part, and introducing the price factors to each cost driver, the economical model is completed and the part cost computed (see Figure 3).

Scope of the CLC Model

The proposed model expands the product life cycle by integrating the life cycle of the tool required to produce it (see Fig. 2). According to the product and to the tool required to produce it, the processes involved in each life cycle phase are defined. As dedicated tools are allocated to only one product, the tools life cycle is exclusively allocated to the product life cycle and the total life cycle includes both. Notice that there is a common phase, the part production/tool use phase, which depends on both design decisions (tool and product), becoming a two dimensional analysis. Additionally, there are some life cycle stages that may occur outside an industrial context, as for example the part use or some EOL scenarios such as landfill or recycling. These stages need to be evaluated individually and in some cases, for example in the use phase of some household plastic parts, may be disregarded if no costs or environmental impacts are entailed.

Tool Material Production

Part Production/ Tool Use

The proposed CLC model further explores the process-based models potential, as it broaden the scope not only to the manufacturing process required to produce the part, but to all the processes involved in the part and tool life cycle. Regarding the part production/tool use phase, the process(es)’ requirements are modelled as a function not only of the part design and material, but also of the tool design. In some cases it is even possible to correlate the part design with the tool design, increasing the model flexibility and sensitivity to part design and material changes. 2.2.1 Financial Relations Used in Process-Based Cost Models

Tool Production Part Material Production

Figure 3: Process-based cost model.

Part Use

Tool EOL Figure 2: Product/Tool Life Cycle.

Part EOL

The financial model follows the approach of annual production costs, in which an annual expected demand is assumed. The variable costs, including material, labour and energy are dependent on the production volume to meet the expected demand and the fixed costs, associated with the use of tools, equipments and space, are computed by allocating an annual cost. This annual cost is simply an annuity of the investment, and the allocation rule is based on the percentage of time consumption. For further information, these financial models have been extensively used and can be consulted elsewhere [Field et al. 2007, Fixson 2005]. However, as already explained, this methodology goes beyond production costs, seeking to estimate and correlate tool or part design features with the subsequent processes in which the tools are used. With this it is possible to estimate the impacts of particular tool design alternatives in the product’s production, use and end-of-life phases.

Comprehensive Model to Evaluate the Impact of Tooling Design Decisions on the Life Cycle Cost and Environmental Performance

2.1.1.1

Total Annual Cost of a Process

In order to build the overall life cycle cost model, each process included is decomposed in several cost drivers (Eq. 1), therefore building it with a backwards methodology . Each cost driver is decomposed first in price factors and resources requirements. These resources requirements are then analysed in terms of operating conditions and process requirements. This last phase is where technological knowhow is entailed. In order to understand the process requirements it is necessary to understand the production process and if possible, to link these requirements to product specifications through physical, statistical or empirical relations between parameters. (1) 2.1.1.2

Cost Modelling Concepts

In order to better understand some relations presented in the next sections, it is necessary to introduce some concepts to be used in the cost modelling of the process. The following tables present the main definition and variables generally used in process-based cost models (table 1) and the equations to compute the variable and fixed costs of each process (table 2). The uptime is the available time for production minus the fixed resources idle time. So only during the uptime parts or products are being produced, meaning that the fixed costs should be allocated based on the uptime consumption. For example, a machine may be available for production 8 hours per day (1 shift, no breaks), but if in average it is idle 2 hours per day, it only generates value during the uptime of 6 hours per day. The costs of use the machine should then be all allocated to the uptime – that is, when parts/products are being produced. Depending whether the equipment/tool is dedicated or not, the uptime regards the production of one or more parts/products. Additionally, the uptime during the part production/tool use phase is highly dependent on the tool maintenance level, as it influences the downtime. But the tool maintenance requirements, even if it is a decision of the plant supervision, are influenced by the tool design and part design. To accommodate that a maintenance model of the tool was developed, linking part and tool design features with tool maintenance level. Regarding the dedicated tool production, the required processes are modelled considering a unit production volume (see equation 3). However, tool spare parts may be required over time during tool use/part production. The part production is highly dependent on the dedicated tool reliability (replacement of tool components is usually required). Therefore the integration of a model to capture the tool reliability cost in the LCC is also important. This model includes not only the replacements of tool components but also the probability and cost of penalties. Finally, dedicated tools influence other aspects such as engineered material waste, reject rate and energy consumption. These aspects are also further developed in order to capture the tool Cost Driver Material Cost

Term Cmaterial

517

design impacts on them. The modelling of the impacts of part and tool design in the LCC is explored in the following sections. Name

Term

Description

Downtime

DT

Idle time

Idle

Uptime

UT

Gross Volume

Vgross

Non productive time – breaks, no shifts periods, maintenance and idle. Time in which the plant /line/workstation/machine/worker is available to produce but has no work Time in which parts or products are being produced, being the plant costs allocated only in this time Annual volume of produced parts

Net Volume

Vnet

Annual volume of accepted parts

Rejected rate

rrej

Percentage of rejected parts

Cycle time

tc

Time required to produce a part

Number of setups Setup time

nsetup

Number of machine setups required between batches to produce Vnet Time of one machine setup

tsetup

Annual required time Price factor

t pi

Price factor of item i

Weight

wi

Weight of item i

Annual discount rate Social taxes

r

Interest rate

rsocial

(2)

Nr of paid months Nr of workers

nPM

Social taxes over salaries paid by company Per year

nW

Number of workers per process step

Working days

WD

Per year

Table 1 - Definitions and variables 2.3

Environmental Analysis of Processes

In order to assess the environmental impacts of the tool and product life cycle, a comprehensive methodology such as LCA is necessary. Similarly to the cost analysis, process-based models are used in order to link product and tool design to the material and energy consumption and emissions. As most of the energy and material flows were already analysed in order to assess the life cycle costs, the environmental model can be integrated in the process-based cost models, using the data previously obtained. As in LCC model, this procedure is applied to all processes of the product/tool life cycle. Other required inputs for the environmental analysis need to be added to the model and the environmental model can be developed using LCA software, SimaPro®.

General Processes

Dedicated Tool Production

_

Eq. (3)

_

Direct Labour

Clabour

Indirect Labour

Cindirect

(5)

Energy Tooling, Equipment, Building Maintenance

Cenergy

(6)

Ctooling, equip,build Cmaintenance

(4)

1 ,

,

Cmaintenance=%Ctooling, equipment, building Table 2: Variable and fixed costs.

1

1

(7) (8)

I. Ribeiro, P. Peças, and E. Henriques

518

2.4

Linking Part and Tool Design with Process Requirements

Albeit some relations are already patent in the financial relations, one of the main operational purposes of process-based models is to avoid process inputs when changing the part specifications. That is, if a change is performed in the part design, the model is supposed to accommodate this change and to compute the new cost or environmental impact. For these the development of physical, empirical or statistical relations may be required, depending on the focus of the analysis and on the processes involved. Furthermore, as this model integrates the life cycles of two elements, the product and the tool, dependencies arise between them, being the tool also related with the product. Figure 3 illustrates general relations that exist in most part/tool situations. Starting with part design, its geometry and material obviously influences the geometry, material and some components of the tool. However, it still leaves field for some freedom in tool design, as different architectures and materials, subsystems and other components can produce the same part, with different tool production cost (driven from time to produce, purchased standard components or subsystems and subcontracts required when the technology is not available in the tool producer) and different part production/tool use performance. This brings another dependency; part production/tool use phase is dependent obviously of the part design and specifications, but also of the tool design decisions. Most of these dependency relations are not possible to generalize for all part/tool life cycles, as different parts require different production processes and therefore different tools. Despite this, there are important aspects that are common in most part/tool life cycles, as the dependence of the part production process on the tool reliability, the impact of both the part and tooldesign in energy consumption, material waste, downtime due to maintenance and failures and equipment required. Additionally, the tool maintenance cost depends on the part material and tool complexity and architecture. Regarding the part/product use, its performance is mainly dependent on the part design and production, but the modelling of this stage varies from product to product. Finally, the part and tool EOL scenarios are constrained by the materials used Tool Design • Geometry • Material • Architecture • Extra components , subsystems and technology

Tool Production • Production time • Machines • Material waste • Energy • Consumables • Subcontracts

Part Design • Geometry • Material Energy submodel Physical and/or empirical relations

Part Production/Tool Use

Reliability submodel Maintenance submodel Energy submodel Physical and/or empirical relations

Part EOL

Part Use • Performance

• EOL scenarios

• Material waste • Process Reliability • Cycle time • Production Volume • Tool Maintenance • Energy • Downtime • Machine required Tool EOL •EOL scenarios

Figure 4: Impact relations of part and tool design.

and by the product/part/tool easiness to dismantle (design for dismantling), and therefore by part and tool design. In this model engineering know-how is required to model these correlations by using already established relations or by developing models, as illustrated in Figure 4. The proposed models further explore the correlations between part and tool design, namely tool reliability, tool maintenance and energy consumption in the part production. Depending on the particular processes involved, other key aspects may arise fostering the development of other correlations or models. 2.5

Developing Sub-models

When applying the methodology to a specific process, it is often required to develop specific sub-models to allow estimating outputs of the life cycle processes. Having in mind the need to link the part or tool specifications with the process requirements, it is necessary to understand the processes involved and the tacit engineering knowledge, to identify already known relations and correlations or to develop new ones. The first step in developing a sub-model is to search in existent bibliography similar studies that can serve as a base for the particular process. The most common approach is to use physical or statistical models, defining with them physical parameters that link the part or/and tool with an output of the process. The second step consists in experimental analysis in situ with the aim of obtaining experimental results, along with the application of the theoretical model to the same process. This allows two main deductions: (1) the ability of the theoretical model to estimate the output in the specific process; (2) the identification of the technological or physical parameters, not included in the model, that are responsible for the deviation between the experimental and the theoretical results. Notice that the experimental analysis can be of different types: measurements, historical data, questionnaires or surveys, depending on the particular case. The third step is to develop empirical or statistical relations to fill the gaps or to introduce new data in the theoretical models. However, if no similar studies were found in step one, it is then required to develop from the process analysis a new model from scratch, which again can be based on statistical/historical data or on empirical relations. Finally, the last step is to validate the developed model with case studies and again with in situ measurements/data collection. The described methodology can be visualized in figure 5. Also a brief example is given with the development of a tool failure model in the plastic injection moulding process. In order to exemplify the methodology proposed in above, a brief description of the development of a model regarding the failure probability of plastic injection mould components is illustrated in Figure 6. The need to develop a failure model for these kinds of dedicated tools is mainly due to their one of a kind nature, not being possible to assess the failure probability in the design stage of a new mould based on historical data. However, despite being all different according to the part to be injected, the client requirements and the design choices, the moulds can be seen as systems with mechanical components, most of them standard. Therefore, it is possible to decompose the mould and assess the failure probability of each element. Regarding the non standard components, the core and the cavity, also these elements are usually made of parts and with critical aspects as stress points, injection of abrasive materials and very thin elements, among others. Despite the lack of historical information and quantification of reliability values, companies have a qualitative knowledge regarding these critical aspects. Through bibliographic

Comprehensive Model to Evaluate the Impact of Tooling Design Decisions on the Life Cycle Cost and Environmental Performance

No

Theoretical model (process)

Experimental analysis in situ

Experimental analysis in situ

Model proposed

•Empirical or statistical relations to fill the gaps or to introduce new data in the theoretical models

Identification of relevant technological parameters (measurements, historical data, questionnaires or surveys )

Yes

Empirical/ Statistical relations from process analysis and in situ data

Similar studies?

•Ability to estimate the output in the specific process •Identification of relevant technological parameters (measurements, historical data, questionnaires or surveys )

•Linking part/tool parameters with process output (physical, statitical, empirical)

Process Analysis

Improved/ completed model

Model validation

In situ measurements/ data collection

Figure 5: Developing sub models.

Injection Moulding Similar studies? Yes

Theoretical model (process)

Process analysis Bibliographic research Approach to gathering Failure behaviour information about mechanical components based on expert knowledge using Weibull distribution [15, 16]. Application of the methodology to gather Weibull parameters to plastic injection mould components

Experimental analysis in situ

Gathering data and experts knowledge regarding failure of mould components

Improved/ completed model

Weibull parameters of injection mould components and estimation of probability of failure and replacements

519

methodology. The information was obtain through carefully conducted questionnaires to the company engineering staff. This is very useful because the Weibull parameters are allocated to elements and not moulds, being possible to use this information to other moulds. Finally, this estimation allowed the quantification of the failure probability and the need/impact of mould components replacement, with the corresponding material and energy requirements. The validation of the model was performed through the mould performance during production. The model can be then applied to other moulds, by analysing the components involved, the part geometry (regarding the stress points), the type of material and the geometry of the cavity/core elements. 3

SUMMARY

The design and production of dedicated tools highly influences the manufacturing process in which they are used. Despite being acknowledged by the industry, there is still a lack in the quantification of these impacts due to the “one of a kind” nature of these tools. This is especially critical when moving forward through the tool life cycle. In the tool design stage several relevant issues are not in a formal way, especially regarding aspects as reliability, material wastes, energy consumption, expected downtime, among others. However, there is generally sufficient theoretical knowledge and historical data and, most of all, tacit knowledge based on the experience of senior tool designers that can be gathered, treated and structured into formal relations and rules to support “good” design decisions. Finally, another factor sensitive to the tool design is the environmental impact. Increasingly important nowadays, it is normally disregarded in tool design, as the tool itself usually represents low environmental burdens. Yet, when analysing the impacts of design changes in material and energy consumption, the environmental impacts can be very sensitive to it. The model proposed in this paper, overviews the research work under development, aiming to quantify the tool life cycle impacts by using process-based models, so that the costs and environmental impacts are assessed according to the processes modelled and are sensitive to design and manufacturing changes. From the geometry of the part, material used and other design inputs, each process is modelled taking into account the manufacturing requirements and both the costs and environmental burdens driven by them. Finally, this methodology is not aiming to translate into a model the current design decisions, but aims to go further in capturing the most relevant impacts of these design decisions, even if disregarded in the process. Moreover, it is especially useful when dealing with new technologies and tool features, as the modelling of the process allows estimating their future costs and other impacts without major investments. Hence, it can be used to support more informed decisions in the tool design phase, prior to the part production. 4

REFERENCES [1]

Woodward, D.G. (1997): Life cycle costing—Theory, information acquisition and application,.in: International Journal of Project Management, Vol. 15, No. 6, pp. 335344.

Figure 6: Development of injection mould failure sub model.

[2]

research (step one), it was possible to find some reliability studies regarding mechanical components and a methodology to gather Weibull parameters based on expert knowledge [15][16]. This was then applied to injection moulds gathering in the field information required to estimate the Weibull parameters, using the pre-existent

Dahlén, P.; Bolmsjö, G.S. (1996): Life Cycle Cost analysis of the labour factor, in: International Journal of Production Economics, Vol. 46-47, pp. 459-467.

[3]

Barringer, H.P.; Weber, D.P. (1996): Life Cycle Cost Tutorial, in: Fifth International Conference on Process Plant Reliability, Houston, Texas.

Model validation

Validation with historical data

I. Ribeiro, P. Peças, and E. Henriques

520 [4]

Park, J.-H.; Seo, K.-K.; Wallace, D.; Lee, K.-I. (2002): Approximate Product Life Cycle Costing Method for the Conceptual Product Design, in: CIRP Annals Manufacturing Technology, Vol. 51, No.1, pp. 421-424.

[5]

Asiedu, Y.; Gu, P. (1998): Product life cycle cost analysis: State of the art review, in: International Journal of Production Research, Vol. 36, No. 4, pp. 883–908.

[6]

Westkämper, E.; Osten-Sacken, D. (1998): Product Life Cycle Costing Applied to Manufacturing Systems, in: CIRP Annals - Manufacturing Technology, Vol. 47, No. 1, pp. 353-356.

[7]

Chen, S.; Keys, L. K. (2009): A cost analysis model for heavy equipment, in: Computers & Industrial Engineering, Vol. 56 No. 4, pp. 1276-1288.

[8]

Jambulingam, N.; Jardine, A. K. S. (2009): Life cycle costing considerations in reliability centered maintenance: An application to maritime equipment, in: Reliability Engineering, Vol. 15, No. 4, pp. 307-317.

[9]

Peças, P.; Ribeiro, I.; Folgado, R.; Henriques, E. (2009): A Life Cycle Engineering Model for Technology Selection: a Case Study on Plastic Injection Moulds for Low Production Volumes, in: Journal of Cleaner Production, Vol. 17, No. 9, pp. 846-856.

[10] Ribeiro, I.; Peças, P.; Henriques, E. (2011): Life Cycle Approach to Support Tooling Design Decisions, in: Proceedings of the 11th International Conference on Engineering Design. Copenhagen, Denmark.

[11] Jeswiet, J.; Hauschild, M. (2005): EcoDesign and future environmental impacts, in: Journal of Materials and Design, Vol. 26, No. 7, pp. 629-634. [12] Field, F.; Kirchain, R.; Roth, R. (2007): Process cost modeling. Strategic engineering and economic evaluation of materials technologies, in: JOM, Vol. 59, No. 19, pp.2132 [13] Johnson, M. D.; Kirchain, R. E. (2009): Quantifying the effects of parts consolidation and development costs on material selection decisions: a process-based costing approach, in: International Journal of Production Economics, Vol. 119, No. 1, pp. 174–186. [14] Lanza, G.; Niggeschmidt, S.; Werner, P. (2009): Optimization of preventive maintenance and spare part provision for machine tools based on variable operational conditions, in: CIRP Annals - Manufacturing Technology, Vol. 58, No. 1, pp. 429-432. [15] Lanza, G.; Werner, P.; Appel, D.; Behmann, B. (2011): Increased trustability if reliability prognoses for machine tools, in: 18th CIRP International Conference on Life Cycle Engineering, Germany, pp. 225-228. [16] Jagger, P.; Bertsche, B. (2004): A new approach to gathering Failure behaviour information about mechanical components based on expert knowledge, in: Proceedings of the Annual Reliability and Maintainability Symposium, pp. 90-95.