Department of Product and Production Development CHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden 2014 Concept Development of a Lightweight Driver’s Seat Structure & Adjustment System Combining Optimization & Modern Product Development Methods to achieve a Lightweight Design Master’s thesis in Product Development JAKOB STEINWALL PATRIK VIIPPOLA Concept Development of a Lightweight Driver’s Seat Structure & Adjustment System Combining Optimization & Modern Product Development Methods to achieve a Lightweight Design Jakob Steinwall Patrik Viippola Department of Product and Production Development CHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden 2014 Concept Development of a Lightweight Driver’s Seat Structure & Adjustment system Combining Optimization & Modern Product Development Methods to achieve a Lightweight Design © JAKOB STEINWALL, 2014 © PATRIK VIIPPOLA, 2014 Department of Product and Production Development Chalmers University of Technology SE-412 96 Göteborg Sweden Telephone + 46 (0)31-722 1000 Cover: Render of the final seat structure design concept. Further explanation of this concept is found in section 6. Further development of the chosen concept Reproservice Göteborg, Sweden 2014 Concept development of a Lightweight Automotive Driver’s seat Structure and Adjustment system Combining optimization & modern product development methods to achieve a lightweight design Jakob Steinwall Patrik Viippola Department of Product and Production Development Chalmers University of Technology Summary The automotive industry is currently going through changes in terms of consumer buying patterns, the best selling cars have gone from large SUV’s and luxury sedans to small compact cars that are fuel efficient. In response to this auto manufacturers try to increase the fuel efficiency of their cars. The current trend is to accomplish this by reducing the overall mass of the vehicle, but this cannot compromise the cost and performance of the current solution. The purpose of this thesis has been to develop a lightweight design concept for a generic driver’s seat structure and adjustment system, in collaboration with Semcon Sweden AB. Furthermore the developed concept was to answer the question of: How could the driver’s seat mass be reduced, without compromising the safety, cost or ergonomic performance of the current solution? The approach to developing a lightweight design was to analyze the current seat design, model this design using CAD-software and analyze it using FEM, to produce a set of target specifications. Based on the specifications new concept designs were generated and optimized using topology optimization software. The generated concept designs were evaluated in the same way as the reference and further refined before a final concept selection was made using selection matrices. The chosen concept was further developed to a detail that could be compared to the reference seat design. As a final step, an evaluation of the final concept was performed using ergonomic analysis, FE-analysis, and cost analysis. The results of the analyses have shown that all target specifications have been fulfilled, and that if compared to the reference seat, the final concept is 27 percent lighter, 1 percent cheaper in terms of unit cost, able to withstand the same impact load cases, and able to fulfill the same basic ergonomic requirements. The main conclusion of this thesis is thus that it is possible to reduce the driver’s seat mass without compromising the safety, cost or ergonomic performance of the reference design. More specifically it has been shown that it is possible to achieve this goal by focusing on redevelopment of the seat structure and adjustment system, whilst leaving the remaining seat components untouched. Keywords: Concept development, automotive, optimization, lightweight, driver’s seat Sammanfattning Just nu undergår bilindustrin förändringar gällande kundernas köpvanor då de mest säljande biltyperna har övergått från att vara stora SUV:ar, och lyxiga sedaner, till små bränslesnåla bilar. På grund av detta försöker biltillverkare numer minska bränsleförbrukningen på sina bilar. Den nuvarande trenden är att åstadkomma detta genom att reducera fordonets vikt, utan att för den skull äventyra dess tillverkningskostnad eller prestanda. Syftet med detta examensarbetet har varit att utveckla ett lättviktskoncept för en förarstols stomme och justeringssystem, i samarbete med Semcon Sweden AB. Vidare skulle det utvecklade konceptet vara ett svar på frågan: Hur kan massan på en förarstol minskas utan att äventyra säkerheten, kostnaden eller ergonomin hos den nuvarande lösningen? Metoden som användes under utvecklingen av lättviktsstolen var att först analysera den nuvarande stolen, sedan modellera upp denna med hjälp av CAD-program och analysera den nya modellen i FEM, detta för att få fram en kravspecifikation. Baserat på denna kravspecifikation framställdes nya koncept vilka optimerades med hjälp av topologioptimeringsprogram. De nya koncepten utvärderades på samma sätt som referensstolen, vilket ledde till att förfinade koncept kunde fås fram innan ett slutgiltigt koncept valdes med hjälp av elimineringsmatriser. Det slutliga konceptet utvecklades vidare till en nivå där det kunde jämföras med referensstolen. Som ett sista steg utvärderades det slutgiltiga konceptet med hänsyn till ergonomi, strukturella krav (FEM) och tillverkningskostnad. Resultatet från analyserna har visat att samtliga krav i kravspecifikationen har uppfyllts, och jämfört med referensstolen så har vikten reducerats med 27 procent, kostnaden minskat med 1 procent, konstruktionen klarat alla lastfall, samt de grundläggande ergonomiska kraven har blivit uppfyllda. Slutsatsen av detta examensarbete är att det är möjligt att reducera massan hos en förarstol utan att äventyra säkerheten, kostnaden eller ergonomin hos den nuvarande konstruktionen. Mer specifikt så har det visats att det är möjligt att uppnå detta resultat genom att enbart fokusera utvecklingen på stolsramen och dess justeringssystem, och lämna övriga komponenter orörda. Nyckelord: Konceptutveckling, fordon, optimering, lättvikt, förarstol Acknowledgements The development team in this thesis, Jakob Steinwall and Patrik Viippola, would like to acknowledge and extend our gratitude towards the following people for their contributions. First of we would like to thank our supervisor, and expert of automotive interior design, Björn Rosvall for helping us immensely through numerous stages throughout the project. These would have been difficult to overcome without his advice. Furthermore we would like to thank Ulrika Berglin who helped provide us with this thesis. At Chalmers we would like to thank Hans Johannesson, which has been our supervisor and examiner, for his help with methodology and thesis guidance throughout the project. Further acknowledgements for people at Semcon are Adam Andersson for his expert advice and genuine help during different thesis stages and Anders Sundin for lending us his experience in the field of ergonomics. We would finally like to thank seat design expert Mathias Klein for his advice and help during the concept generation phase of the thesis. Terminology CAD Computer Aided Design, in this thesis the software CATIA V5 is used for concept design FEM Finite Element Method, in this thesis the software CATIA V5 is used for structural analysis Altair Inspire Topology optimization software BIW Body in White, the car body without any loose components such as doors and wheels Recliner An adjustment function that enables the seat back to pivot around the axle between the back and base SRP Seat Reference Point, the pivoting axle point between the back and the base Seat Structure The seat frame without any padding or upholstery Adjustment system Collective name of the seat’s different adjustment functions, which includes length, height and recliner. Ergonomic performance Anthropometry, how well body positions and angles interact with a product, e.g. which knee angle is regarded as comfortable in a seating position Structural performance / Safety How much force the seat can withstand without breaking, or bending too much Sill & Tunnel The sill is the beam part of the BIW that is running along the outskirts of the vehicle. The tunnel is the center beam part of the BIW. Both of them help reinforce the BIW. BESO2D Topology optimization software which only can generate two dimensional models BOM Bill of Materials, a list of all components in a product assembly Table of Contents 1. Background & Problem Identification ................................................................................................. 1 1.1 Fuel consumption and trends in automobile consumer behavior ................................................ 1 1.2 Possibilities of weight reduction in the automotive industry ....................................................... 2 1.2.1 Current trends of weight reduction in car seats .................................................................... 3 1.3 Car Seat Components and Adjustments ........................................................................................ 4 1.4 Mass allocation .............................................................................................................................. 5 1.5 Semcon Sweden AB and the thesis ............................................................................................... 6 1.6 Thesis purpose ............................................................................................................................... 6 1.7 Unique Contribution ...................................................................................................................... 6 1.8 Delimitations ................................................................................................................................. 7 1.9 Chosen Development Process ....................................................................................................... 7 1.9.1 Planning of the development process (Chapter 3)................................................................. 7 1.9.2 Concept generation (Chapter 4) ............................................................................................. 8 1.9.3 Concept selection (Chapter 5) ................................................................................................ 8 1.9.4 Further development of the chosen concept (Chapter 6) ..................................................... 8 1.9.5 Final evaluation against target specifications (Chapter 7) ..................................................... 8 2. Theory .................................................................................................................................................. 9 2.1 Competitive Benchmarking ........................................................................................................... 9 2.2 Reverse Engineering ...................................................................................................................... 9 2.3 Target Specification ....................................................................................................................... 9 2.4 Brainstorming .............................................................................................................................. 10 2.5 Function means modelling .......................................................................................................... 10 2.6 Design Optimization .................................................................................................................... 10 2.6.1 Topology Optimization ......................................................................................................... 11 2.7 Concept Combination Table (Morphological Matrix) .................................................................. 11 2.8 Weight decision matrix ................................................................................................................ 12 2.9 Concept Screening and Scoring Matrix ....................................................................................... 12 2.10 Cost Analysis .............................................................................................................................. 12 2.11 Computer Software ................................................................................................................... 13 3. Preparing the development process ................................................................................................. 14 3.1 Customer needs........................................................................................................................... 14 3.2 Analysis of Reference Seat .......................................................................................................... 15 3.2.1 Reverse engineering a current seat ...................................................................................... 16 3.2.2 CAD-model of reference seat ............................................................................................... 17 3.2.3 Forces acting upon the seat ................................................................................................. 17 3.3 Initial Product Specifications ....................................................................................................... 19 3.3.1 Establish engineering metrics ............................................................................................... 19 3.3.2 Competitive benchmarking .................................................................................................. 21 3.3.3 Target specifications ............................................................................................................. 22 4. Concept generation ........................................................................................................................... 24 4.1 Reformulation of functional criteria ............................................................................................ 25 4.2 Functional-means model analysis ............................................................................................... 25 4.3 Exploration of the solution space ................................................................................................ 28 4.4 Initial screening and optimization of sub-solution concepts ...................................................... 30 4.4.1 Ranking and sorting of ideas ................................................................................................ 30 4.4.2 Refinement and description of promising concept ideas .................................................... 32 4.5 Optimization process ................................................................................................................... 34 4.6 Generation of concepts for Adjust driver & Provide structural support ..................................... 37 5. Concept Selection .............................................................................................................................. 42 5.1 Choosing weighted screening criteria ......................................................................................... 42 5.2 Ergonomic analysis of solution concepts .................................................................................... 43 5.2.1 Adjustment system concepts ............................................................................................... 43 5.2.2 Structural support concepts ................................................................................................. 43 5.3 Concept mass analyses ................................................................................................................ 44 5.3.1 Mass analysis of adjustment concepts ................................................................................. 44 5.3.2 Mass analysis of structural concepts .................................................................................... 44 5.4 Concept cost analysis .................................................................................................................. 46 5.5 Weighted selection matrix screening .......................................................................................... 47 5.5.1 Adjustment concept selection matrix .................................................................................. 48 5.5.2 Structural concept selection matrix ..................................................................................... 48 5.6 Description of chosen solutions for adjustment and structural support .................................... 49 6. Further development of the chosen concept.................................................................................... 51 6.1. Material selection and Optimization .......................................................................................... 51 6.2. Design considerations ................................................................................................................ 53 6.2.1. Dimensions .......................................................................................................................... 53 6.2.2. Moving the recliner mount ................................................................................................. 54 6.2.3. Height of the tunnel ............................................................................................................ 54 6.2.4. Neck support ....................................................................................................................... 54 6.2.5. Supporting driver weight ..................................................................................................... 54 6.2.6. Ergonomic constraints ......................................................................................................... 55 6.3. Optimizing the final design ......................................................................................................... 55 6.3.1. Base ..................................................................................................................................... 56 6.3.2. Back ..................................................................................................................................... 57 6.3.3 Seat-to-floor attachment structure ...................................................................................... 57 6.4 Final design and assembly procedure ......................................................................................... 58 7. Evaluation of the final seat structure design concept ....................................................................... 59 7.1. Ergonomic evaluation ................................................................................................................. 59 7.2. Structural evaluation .................................................................................................................. 62 7.2.1 Base frame structure ............................................................................................................ 63 7.2.2 Back frame structure ............................................................................................................ 63 7.2.3 Seat-to-floor attachment structure ...................................................................................... 64 7.2.4. Summary of structural performance ................................................................................... 64 7.3. Mass evaluation & Comparison.................................................................................................. 65 7.4. Cost evaluation & Comparison ................................................................................................... 67 7.5. Final comparison with target specifications ............................................................................... 69 8. Discussion .......................................................................................................................................... 70 8.1 The findings and their meaning ................................................................................................... 70 8.2 Reliability ..................................................................................................................................... 71 8.3 Validity ......................................................................................................................................... 73 8.4 Lessons learned throughout the process .................................................................................... 74 9. Conclusions ........................................................................................................................................ 75 10. Recommendations........................................................................................................................... 76 11. Alternative concept and further recommendations ....................................................................... 77 12. Reference List .................................................................................................................................. 78 Appendix 1 - Calculation of load cases for the seat structure............................................................... 80 Appendix 2 - Competitive Benchmarking .............................................................................................. 81 Appendix 3 -Reformulation of Criteria .................................................................................................. 82 Appendix 4 - Brainstorming ideas ......................................................................................................... 83 Appendix 5 - Concept generation optimization loop ............................................................................ 84 Appendix 6 - Ergonomic analysis of design solution concepts .............................................................. 85 Appendix 7 - Mass analysis of design solution concepts....................................................................... 86 Appendix 8 - Cost analysis of design solution concepts ........................................................................ 90 Appendix 9 - Weighted screening criteria matrices .............................................................................. 98 Appendix 10 - Material Selection .......................................................................................................... 99 Appendix 11 - Structural analysis of reference design ........................................................................ 101 Appendix 12 - Structural analysis of final design ................................................................................ 106 Appendix 13 - Reference seat Bill-of-materials ................................................................................... 112 Appendix 14 – Complete bill-of-materials for the final concept ......................................................... 114 Appendix 15 - Assembly procedure visualization ................................................................................ 116 file:///C:/Users/PJ/Documents/Dropbox/Dropbox/Master%20Thesis/Rapportstruktur/Final%20Report_Lightweight%20Seat%20140520.docx%23_Toc389047586 1 1. Background & Problem Identification In order for the reader to understand the motive for developing a lightweight seat structure concept the current trends of the automotive industry is presented in this introductory chapter. Especially trends relating to environmental impact and consumer behavior, but also the most relevant information of lightweight design in the transportation sector is presented. During this background description the main design problem is identified, which is further clarified in the Thesis purpose sub- section. The background chapter also contains a description of a typical current car seat design and a description of the company at which the thesis has been conducted. Finally the chosen development process is presented with a brief description of the methods used. The outline for this chapter is: 1.1 Fuel consumption and trends in the automotive sector 1.2 Possibilities of weight reduction in the automotive industry 1.3 Description of the current car seat design 1.4 Mass allocation of current seats 1.5 Semcon Sweden AB and the thesis 1.6 Thesis purpose 1.7 Unique contribution 1.8 Limitations 1.9 Chosen development process 1.1 Fuel consumption and trends in automobile consumer behavior Modern cars are faced with constantly increasing demands of lowering emissions and reducing fuel consumption while at the same time improving performance. It is also common to incorporate an increasing amount of functions in an effort to make the owners life easier (Lotus 2010). Regarding emissions the transportation sector is said to account for 23 percent of the world’s CO2-emissions, a figure that is projected to increase by 40 percent by the year 2030 (ITF 2010, p. 5). At the same time fossil fuel reserves are projected to run out shortly after 2050 (Doherty 2012). This has resulted in increased costs for the consumers of automobiles, as both the base price of fuel and the governmental tax in many countries have increased significantly over the last decade (Hatt 2012). Conversely automobile buying patterns has shifted from large SUVs and luxury sedans to smaller more fuel efficient cars that cost less to use. In a survey conducted by NADA guides in USA, the country which is considered to be one of the largest markets of cars, the predominately determining factor in consumers buying their next car will be fuel consumption, even compared to such factors as quality and safety (PR Newswire 2013). Interestingly enough a large part of the respondents were willing to downsize or give up comfort functions in order to achieve lower fuel consumption. The challenge for modern car producers will be how to increase fuel efficiency while at the same time advancing the cars performance relative to competitors, all while keeping the cost of the vehicle down. The performance of the car is associated with several criteria, such as acceleration, handling, and comfort. The mass of the car plays a large role here as reducing it will have a positive impact on both performance and fuel efficiency. The major limitation to mass reduction is cost, as lightweight materials are generally more expensive than traditional engineering materials such as steel. Thus innovative design ideas that could utilize traditional materials in more effective ways would also be a potential source of mass reduction at maintained cost levels. This leads to the question: How can vehicle mass be reduced, without sacrificing cost, safety and ergonomics? 2 Part Base mass [kg] Mass reduction [kg] Mass reduction [%] Rear 60% seat 26.48 13.55 51.2 Rear 40% seat 16.41 1.49 9.1 Front driver’s seat 26.91 4.72 17.5 Front passenger seat 22.75 3.64 16.0 Seats total 92.55 23.39 25.3 1.2 Possibilities of weight reduction in the automotive industry In 2010 Lotus engineering released a major study on the capabilities of weight reduction in modern cars. Based on a detailed reengineering of a 2009 Toyota Venza, they found that by combining the best solutions on the market together with a system-level engineering methodology it was possible to achieve significant weight reduction and lower the cost of the vehicle (Lotus 2010). This was achieved without sacrificing vehicle functions, performance or safety. In summary a total of 21 percent weight reduction was achieved in a low development scenario involving 2017’s production models, and a total of 38 percent in a future scenario that would start development by the year 2020 (Lotus 2010, p. 7). Although this study acknowledges some absence of detailed analysis in cost, mass and impact performances, it reveals a large potential area of improvement within the automotive industry. Another recent study on this subject is a working paper by the International Council of Clean Transportation (ICCT) where a detailed weight reduction forecast on light-duty vehicles within EU conducted by the FEV Company was included. The paper focuses on the change of manufacturing costs due to the implementation of weight reduction and how much emission cuts costs relative to the reduction of weight (Meszler et al. 2013). In conclusion FEV estimated that 18.3 percent of the total weight could be reduced by year 2020. This study was based on the 2010-era cars Toyota Yaris, Ford Focus, Toyota Camry, Ford Transit Connect and Ford Transit. An outtake from ICCT’s paper on just the car seat reductions can be seen in table 1 below. In order to explain the results of both the Lotus study and the FEV findings, the system level approach must be understood. Weight reduction of just one component might not necessarily improve fuel economy or performance significantly, but could increase the component cost more than it contributes to performance or fuel efficiency. However if a system level approach is considered, the reduced weight of one component, e.g. the body, could lower the constraints on another component, e.g. the drivetrain. Not only will this free up resources that could be used for weight reduction but it will also produce a cascading effect. An example of a cascading effect would be that the reduction of Body-In-White (BIW) mass enables a 1.6 liter engine to be used instead of a 2.0 liter engine, which in turn would require a smaller drivetrain. This smaller engine and drivetrain would put less strain on the chassis and suspension system which subsequently could also be downsized, the result of all this downsizing is reduced mass and structural requirements. This cascading effect opens up opportunities for car manufacturers to improve fuel efficiency and performance whilst still keeping costs down. In the 2009 Toyota Venza the seat sub-system make up 39 percent of the total interior mass which is why the seats were targeted as having the greatest potential for weight reduction of the interior (Lotus 2012, p. 77). Table 1: An outtake showing potential weight reduction (Meszler et al. 2013, p. 5) 3 1.2.1 Current trends of weight reduction in car seats Current trends in weight reduction of seats include system integration and co-modeling processes, as well as the use of lightweight frame materials such as magnesium and hybrid combinations. However, a magnesium frame using conventional frame tools would increase the unit costs by as much as 50 percent according to Faurecia, which is one of the largest car seat manufacturers (Lotus 2010, p. 81). Faurecia does however have a strategy for decreasing these costs which is making use of die casting tools in the production process along with an integration of composite materials in the frame. Another lightweight material is ultra-strong polyamide plastics with woven fabrics creating new possible ways to construct lightweight seats (BASF 2014). This material is exceptionally stiff and strong, and can in some cases even replace metal as a construction material. Another technology developed by BASF a is binder technology for powder injection molding which uses metallurgy combined with a traditional plastic process that enables metal to be formed into shapes that conventionally only have been possible for plastic parts. For foams, environmentally friendly materials which are recyclable mark a clear trend within car seats. One of these is Elastoflex® W produced by the BASF chemical company which claims to also have reduced the weight by 15 percent. Composites are a potential construction material for seats as well, making the seat back entirely in composite could evidently reduce the thickness of the seat back by 15 percent and decrease mass by 20 percent (Faurecia 2014a). Since the choice of material often set the boundary for engineering designs, these new materials open up exciting possibilities for future designs. Although they provide a promising future as construction materials, most of them imply increasing production costs which contradicts the automotive companies’ often tight budgets. A faint paradox is that to reduce weight more integration of hybrid materials and complex structures are often needed, but this also implies higher costs (Faurecia 2014b). Because of this many of them are not yet feasible for market introduction. The great challenge ahead will be to reduce material and manufacturing costs, as well as creating innovative designs that enables lighter vehicle components to be produced and be used in competitive solutions. Regarding the design of the seat, the Lotus study acknowledges some potentially weight reducing areas of improvement. The back frame shape is not considered to be ergonomically optimal at the moment since the current shape requires compensational foam and suspensions to work (Lotus 2010). Conventional seat backs structures create pinch points at the front edges of the frame and needs extra padding to satisfy the customers’ requirements on comfort. Lotus suggests a more ergonomic and thinner design, much like the one used in Mercedes SLK, which would reduce the amount of padding needed and thus save weight. Another design idea that proved useful in saving weight was the integration of seat mounts into the sill and tunnel areas of the vehicle body. Traditionally front seats are mounted on steel risers and Figure 1: Standard seatback vs. new thinner seatback (Lotus 2010, p. 80) 4 tracks, which are then bolted to the floor. The floor is reinforced to accommodate the seat loads during transport and in the event of a crash. Replacing these floor mounted seat attachments and floor reinforcements could reduce the seat sub-system mass, as the sill and tunnel are already reinforced eliminating the need for extra structural reinforcement (Lotus 2010). 1.3 Car Seat Components and Adjustments When designing a new seat system it is important to understand the current system, and a good way of understanding this is to describe the different components in the system and what functions they perform. Because this thesis focuses on the driver’s seat the rear seats are not included in this system analysis. The subsystem ‘Front Seats’ consists of two seats, the driver’s seat and the passenger seat. The driver’s seat is often the heaviest as it encompasses additional comfort features such as power assisted adjustments and lumbar support. The subsystem shall provide all customers, for which the cars are designed, with the ability to adjust the seat manually or electrically to ensure that an ergonomically correct driving position is possible. It shall together with the seat belt, and side airbags protect the occupants in the event of an accident. In this section a breakdown of this sub- system is explained. The seat component sub-system consists of two major segments, the seat base and the seat back. The interface between the front seats and the BIW is a sliding rail and track segment which is bolted to the floor and the seat base frame. The seat base is built up of the lower frame, pad and base cover. The purpose of the base is to support the driver’s weight in the z-axis and to provide the main structure of the seat as the base frame connects to the BIW via a track and rail segment, as well as connecting to the back frame of the seat Figure 3: Cartesian coordinate system of a seat Figure 2: Exploded view of a typical modern car seat, BMW 3 Series 328i (A2mac1 2014) 5 Figure 4: Weight allocation of an Audi A5 3.0 Tdi (A2mac1 2014) which regarded as a critical component in terms of performance and mass. Another function of the seat base, which was discovered after discussions with an automobile expert, is to prevent the driver to slip under the seat belt during a frontal crash, a phenomenon also called submarining. Therefore the seat base is always angled between 10° – 20° in order to absorb the kinetic energy that the driver’s mass will produce during a frontal crash. The seat back consists of the rear frame, rear cover, pad and headrest which together mainly supports the driver in x-axis. It is also common in modern car seats for the seat back to incorporate an adjustable lumbar support as well some form of whiplash protection in the headrest. The reclining mechanism is usually located within this component and functions as a link between the base and back structure. This mechanism also allows a change in back angle which in this thesis will be referred to as the recliner function. On both the seat back and seat base there are side cushions sticking out from the pads which supports the driver in y-axis. The connection between the seat base and BIW usually consists of four support legs and two rails, the latter being bolted to the BIW. The support legs are moveable which allows the seat to be adjusted in height via a lever that lifts the base frame upwards. The rails can slide within each other enabling a length adjustment of the entire seat. Together with the recliner function they make up the main adjustments of a car seat. An important aspect of the seat structure is that it should be able to transmit occupant and seat belt loads to the vehicle body in case of a collision. This implies a structural restraint in terms of mass reduction as the seat must be strong enough to meet all performance requirements set by current seat designs. 1.4 Mass allocation When trying to reduce weight it is interesting to look at how the weight of the seat sub-system is distributed across the different components. As seen in the pie-charts of a selection of cars in different segments, the mass of the frame is the largest contributor to the seat sub-system total mass. It can also be seen that the power adjustment equipment make up a large fraction of the mass in seats that have this feature. The reason for choosing the seat in particular is that the seats constitute as much as 50-60 percent of a car’s interior mass (A2mac1 2014). Consequently it is believed that there is room for potential weight reduction of the seats and that reducing weight here will have a large effect on the overall interior weight, this might also produce a cascading effect on the entire vehicle’s mass as described earlier. Together with the adjustment system the seat structure is a promising candidate for mass reducing redevelopment purposes, the remaining components such as the padding, heating system, and side airbags account for a much smaller fraction of the total seat mass. Figure 5: Seat fraction of interior mass (A2mac1 2014) 6 1.5 Semcon Sweden AB and the thesis In recent years weight reduction has become increasingly important in automobile design. In order to differentiate themselves in the marketplace it is important for car producers to come up with innovative solutions that could lower weight and thus improve fuel efficiency and performance, but at the same time not add significantly to production costs. This is where Semcon enters the picture, as a company that delivers expertise and conceptual ideas to major auto developers it is in Semcon’s interest to help develop innovative ideas and strengthen their knowledge in lightweight design. Furthermore interior car design is an area that lies at the core of Semcon’s business plan, thus research in lightweight interior design is most interesting for this company. One way of advancing in the field of lightweight interior design is to collaborate with research institutions such as Chalmers University of Technology which leads to the thesis project that is at hand. This thesis is a first step towards a lightweight seat design proposal that Semcon hopes to deliver in the future. The Lightweight Car Seat is a thesis carried out by two students at MSc level. 1.6 Thesis purpose When developing a new solution for the seat design the team should not focus solely on reduction of mass but also on how to maintain the current level of safety and ergonomic performance. There is simply no purpose of having a lightweight design if it cannot fulfill the basic functional criteria that is expected of a modern car seat. Similarly a lightweight design would not yield any benefits if no car manufacturer can afford to implement it into their current or near future line-up, which implies that a cost constraint is put on the developed concept. This thesis aims to incorporate a mixed optimization and traditional product development approach to design a concept that solves the following problem: How can the driver’s seat mass be reduced, without compromising the safety, cost or ergonomic performance of the current solution? The goal is to find this solution by exploring lightweight interior design and to carry out a computer aided design of a seat structure that fulfills the following criteria:  Reduce the mass of the driver’s seat as much as possible (Goal) Without violating the following constraints:  Maximum unit cost should not exceed the current solution cost  Maintained performance, i.e. safety and comfort, compared to the current solution 1.7 Unique Contribution Recent studies indicate that the use of topology optimization, which is further explained in the theory section 2.6, early in the development phase results in significant reductions of mass (Zhu et al. 2011; Polavarapu 2008). However these studies tend to focus on the optimization method and calculations itself rather than the use of topology optimization as an aid in generating design concepts. One study investigates the use of topology optimization in the early design phase of the car body but not in the application of car seat design (Hasselblad 2011). This thesis unique contribution is to apply topology optimization as an inspirational tool, alongside traditional design methods, throughout the development process. 7 1.8 Delimitations This thesis is conducted by two students, which means that the work has to be focused on the most essential elements of the seat design. In order to prevent the scope from becoming too wide, and thus reduce the efficiency of the development process, a number of delimitations are necessary in agreement with the supervisors at both Semcon and Chalmers. The delimitations are listed below.  The thesis only includes a conceptual study and a concept selection. The end product is a finished CAD model, mass analysis and cost analysis of that concept.  The thesis does not include a detailed design. The concept includes the main parts that are needed, but detailed parts like screws are left out.  The thesis only focuses on the driver’s seat. Although some ideas might be compatible with the passenger’s seat, the reference seat is still the driver’s seat.  The main elements of interest in the development work are the seat adjustment system and the seat structure. These are focused on throughout the thesis.  The thesis does not include any prototype constructions or testing. Tests are only performed by using CATIA V5, both ergonomically and structurally (FEM).  The thesis does not evaluate any optimization tools, but rather just apply them throughout the project. 1.9 Chosen Development Process In order to achieve the goal of this thesis, to develop a concept that has less mass than the reference design without lowering safety, cost, or ergonomic performance, an iterative strategy that borrows elements from Ulrich & Eppinger (2012) was chosen. The entire project is divided into five main phases: Planning, Concept generation, Concept selection, Further development, and Final evaluation which will be described in more detail below. 1.9.1 Planning of the development process (Chapter 3) The purpose of this phase is to analyze a typical current design of a car seat in order to create a reference design for the later phases of development. Another objective is to understand the Figure 6: Concept development process map 8 customer needs and translate these into measurable targets, i.e. a set of target specifications, for the development process. Analysis is done by dismantling a physical seat as well as modeling this seat in CATIA V5 which yields measurements, mass and performance of the different components. The load cases for the most important impact scenarios are researched here as well. Customer needs are found using personal interviews with drivers and engineers at Semcon, and translated using methodology from Ulrich & Eppinger (2012). The deliverables of the preparation phase is a reference design and a target requirements specification that constitutes the basis for all subsequent development work. 1.9.2 Concept generation (Chapter 4) The focus in this phase is on exploring alternative solutions that might fulfill the goal of mass reduction, safety, ergonomics, and cost. First the functional requirements of the seat structure are understood in more detail through the use of functional-means modeling. This is followed by idea generation for the two main functional criteria Adjust driver and Provide structural support. The long list of sub-solution ideas is initially screened down to a number of feasible concepts that are developed in more detail. The ideas are digitally realized using CATIA v5 and analyzed using Altair Inspire. The use of Inspire, which is a topology optimization software, produces new concepts which means that this is an iterative process. During this process the sub-solution concepts are developed in more detail and evaluated based on their potential to reduce seat mass, FE-analysis is thus used already at this early stage. With the aid of morphological matrices corresponding to the functional- means models the sub-solution concepts are combined into 4 main concepts for the Adjust driver criteria, and 11 concepts for the Provide structural support criteria. 1.9.3 Concept selection (Chapter 5) Utilizing the requirements specification as a starting point, one concept for each of the two main functional criteria is selected, and then combined into one seat structure design concept. In order for the selection process to be as valid as possible, each concept is analyzed in terms of mass, cost, and ergonomic performance. After determining weights, i.e. importance, of the different screening criteria two concept scoring matrices are used to choose a final concept for each of the two main functional criteria. 1.9.4 Further development of the chosen concept (Chapter 6) Having determined the design solutions and overall layout of the final concept, material and manufacturing processes are chosen. This is done with the aid of materials selection software CES and an optimization approach. The final geometry is then optimized based on the selected material and manufacturing process, resulting in a final concept CAD-design. 1.9.5 Final evaluation against target specifications (Chapter 7) To verify that the thesis purpose has been fulfilled an evaluation of the final concept design against the target specifications is conducted. The final analysis consists of an ergonomic evaluation with virtual test dummies in CATIA, FE-analysis of the structural performance when subjected to the identified impact load cases, and a cost analysis of the final concept bill-of-materials. The results from this analysis is compared to the reference seat design results and then evaluated using the target specifications set up in the preparation phase. This closes the “development loop” and the delivery of a lightweight seat structure concept is completed. 9 2. Theory In this chapter different theories and methods that were used throughout the thesis are explained and presented. This is to keep the reader up to date with how the product development theories and tools currently are, and help understand the rest of the thesis better. Further descriptions of how these methods were used are stated in each correspondent chapter. 2.1 Competitive Benchmarking To assure that a product development team will be successful, performing a competitive benchmarking is vital (Ulrich & Eppinger 2012). The reason for this is to gain important insight as to what level of performance commercially available products are at, and get an overview perspective of other product designs before a product development project can begin. One way of doing this is to set up a benchmarking chart that lists all the metrics that have been acquired through the customer needs collection, and then set the values of the different competitors’ relative products for each metric. Although this process seems straight forward, gathering all the data required can be time consuming and sometimes includes purchasing other products and dismantling them. This is because data obtained from competitors own literature might be incorrect and information should always come from a neutral and independent source. Another method that can be used for benchmarking is to grade each metric performance subjectively, which can be obtained by measuring customer satisfaction and perceptions regarding the product. This can be a preferred method if gathering numeric values is difficult. 2.2 Reverse Engineering When trying to understand a product without previous knowledge or having very little understanding about it, implementing the reverse engineering method can be useful. It involves dismantling an existing product that has been identified in the benchmarking process for a new design. This can either be used to copy a design or to improve it (Evertsson 2013). The first step is to investigate the product space and main functions in order to determine the product domain which should relate to customer needs. The second step is to reconstruct the initial product in detail in order to create a Bill of Materials (BOM). This is done by tearing down the product and investigating part interfaces and assimilability. After this phase a functional analysis can be performed in order to study the energy and force flows, and discover possible sub-functions that could be cross integrated using, for instance, a morphological matrix. A new product specification can now be created which will lead design modelling and analysis into a redesign phase which often corresponds to a normal product development strategy. This final phase can either be a parametric redesign where the new design is just an improved version of the old model, an adaptive redesign where the main model is still the same but its sub-functions may have been altered, or an original redesign where knowledge from the initial model have led to an entirely new concept. 2.3 Target Specification A target specification is the first out of normally two specifications during a product development process, the last one being the final specification which is determined after the concept screening and testing. Before a target specification can be made, a translation of the customer needs and requirements into metrics and units are necessary in order for them to be measurable (Ulrich & Eppinger 2012). This is possible by first making a list of the different needs that the product has to fulfill, including the importance of each need. The needs should be formulated in a way that they do 10 not relate to existing solutions and be expressed as wishes. The next step is to define units for each need that is possible to measure, and to identify if some metrics can fulfill multiple needs and vice versa. To be able to decide which values and grades that is to be included in the set of specifications, a collection of competitive benchmarking information is vital for any project. This reassures that the finished product will be at a competitive level and profitable. The value for each metric should be set to an ideal and a marginal value, where the marginal value is the minimum requirement and the ideal is the best case scenario. 2.4 Brainstorming In able to generate new and creative ideas, concept generations often begin with an initial search in form of brainstorming conducted within a project group or a company. It is important for the participants to remain open minded and follow a clear protocol in order not to suppress any radical or unorthodox ideas. The following version is one of these protocols (Ulrich & Eppinger 2012, p. 127). 1. Suspend judgment. Subjective judgment is constantly used in everyday life to make quick and necessary decisions in order to succeed. This however inhibits the ability to come up with creative ideas and solutions. Therefore any criticisms are not allowed in a brainstorming session and all ideas are of equal importance. 2. Generate a lot of ideas. It is believed that increasing the quantity of ideas lowers the expectations of other ideas and helps stimulate new even more creative solutions to be generated. 3. Welcome ideas that may seem infeasible. Infeasible or in some cases rather silly ideas helps to stretch the solution boundaries and contributes to the ‘outside the box’-thinking. 4. Use graphical and physical media. An idea can be difficult to comprehend in just words and trough verbal communication. Therefore it is important to use sketches or other quick physical models to help translate ideas between the participants. 2.5 Function means modelling Function means modelling is a way to divide a product’s function requirements (FR) and design solutions (DS), and see which requirements corresponds to which sub-solution (Johannesson, Persson & Pettersson 2013). This is often done in a tree structure where a FR only can have one DS, but a DS can fulfill multiple FR’s. The method is used in the early stages of product development to illustrate an overview of a product’s functions and to possibly identify unnecessary DS’s that might not fulfill a function. It can also be used to better see if DS’s can be combined or satisfy multiple FR’s. 2.6 Design Optimization Optimization methods can be used to seek the optimal solution in the design space. Even if many solutions can satisfy all the constraints and requirements, optimization tools can guide the product development to the best result. The simplest form of optimization is using trial and error to come up with the best solution (Hoffenson 2013). This method can be time and work consuming but is sometimes the only feasible way to proceed, i.e. when the problem is too complex to formulate mathematically. In other cases when the problem can be formulated, a target function f(x) is set to either be maximized or minimized depending on the problem. Design requirements can be set as constraints, e.g. g(x,t)<0 and h(x,t)=0, creating the boundaries in which the optimal solution would be located. These constraints can either be inequalities or equalities. According to Hoffenson (2013), the design optimization process can be divided into five steps. The first step is to define the system and 11 design space, which largely determines the end result and is often the hardest part of an optimization procedure. The second step is to formulate the problem by deciding on what the objective is and which constraints there are in the system. Deciding on which design variables to be working with is also important along with definition of parameters, which can be assumed fixed elements, and constants which are by definition fixed. The formulation usually looks like this: min f(x) g(x,t) ≤ 0 h(x,t ) = 0 xmin ≤ x ≤ xmax The third step is to create an analytical model and simulate using FEA and optimization software tools in order to test and validate the output data. Since optimization is not often a trivial process a variety of methods can be used when sampling the different variables in the experiment models. Step number four is exploring the problem space in order to see which constraints that are active and determine the function behavior. The last step is to find the optimal solution, often by using different mathematical or numerical methods. The concept of global optimum and local optimum should also be explained. A global optimum is the best possible result in the entire available design domain; a local optimum is the best possible result in a limited section of the entire design domain. An optimization problem could have several local optima but only one global optimum; it is this global optimum that needs to be found if a truly optimal solution is to be reached. 2.6.1 Topology Optimization One special application of the FE-method is topology optimization where an initial design space is subjected to a structural optimization procedure where at least one objective is minimized, e.g. mass, without violating design requirements formulated as constraints. This technique has the ability to change dimensions and geometry of the initial design in an automated process by creating geometry vacancies or changing the outer boundary of a structure. It is generally used in the initial stage of a development process since it is a powerful tool when the objective is to find new conceptual designs. One software program that incorporates this method is Altair Inspire which has the ability to apply topology optimization to an existing CAD-model. A useful side effect of this is that it can show the designer the location of internal stresses and directions of forces, thus enabling an optimal design in both the conceptual and the detail design phases of development. 2.7 Concept Combination Table (Morphological Matrix) A concept combination matrix, or a morphological table, can generate ideas for a complete solution by combining many different sub-solutions for each sub-problem. To perform this method a table is set up where the different sub-functions creates the column titles, and below them the different solutions (fragments) can be set up (Ulrich & Eppinger 2012). This divides the different functions and solutions into categories which enables for different combinations of them to be made. By combining and comparing these types of solutions can potentially trigger new innovating thinking and clarify the problem in a new way. However, the amount of different possible combinations can be enormous if there are too many sub-functions or solutions. Therefore it is only suitable to be used when the number of columns (sub-functions) is no more than three or four. One way of reducing the amount of combinations is to deem a fragment or a sub-solution infeasible, which can be done by using a concept evaluation tool. 12 2.8 Weight decision matrix A weight decision matrix is used to decide weights that might be included in a concept scoring matrix, without any directly subjective judgment (Johannesson, Persson & Pettersson 2013). It’s performed by listing up all the criteria that is thought to be of relative importance, and then rank them with each other. If a criterion is less important than another it receives a value of 0, equally as important a value of 0.5, and if it’s more important a value of 1. This will result in an n x n-matrix and the total ranking sum of a criterion will be the final weight score, which can be interpolated into a weight scale. 2.9 Concept Screening and Scoring Matrix A concept screening matrix, or a Pugh matrix, is often the first selection tool used in order to narrow down different concepts (Ulrich & Eppinger 2012, p. 150). It uses a reference solution, often the current solution, to compare the new ideas with. If the new concept is better than the reference it receives a ‘+’ for that category, a ‘0’ if it’s about the same or a ‘-‘if worse. These rates are then added up to a net score and then ranked to finalize the evaluation. After this a decision which concepts to continue with and/or which ones to combine is made. The rates are decided with the project members’ intuition and previous knowledge and are not always easy to judge. That is why this tool only should be used as a guideline and not as a definitive selection tool. A combination with other concept screening tools is therefore often recommended, which for instance can be a concept scoring matrix. The concept scoring matrix works similar to the screening matrix, but it uses criteria weights as well. Having a weighted matrix enables for a more detailed concept selection and a better end result. 2.10 Cost Analysis There are many different ways of assessing production costs of different parts and materials. One way is to start with the BOM and try to get an estimation of cost for each part (Ulrich & Eppinger 2012). Normally a company or a manufacturer already has knowledge and experience of how much a typical standard part costs by comparing similar parts already produced or purchased. If the number of parts produced is high, for instance more than 100 000 unit per month, the production cost becomes low and usually stagnates around a known value, which also implies material prices. Another way of collect correct estimates is to soliciting price quotes from vendors or suppliers whom often can give some useful indicators of price, even though it’s not entirely correct. After gathering advice from these experts and niched manufacturers, total cost estimation for a product that uses standard part can then be achieved. However, a new product often requires new types of parts and in some sense custom parts are needed to be implemented. This is not as easy as estimating standard part costs but instead focuses on production tools and the complexity of a part. New tools are often needed which increases the fixed expenses, which in turn requires an investment from the company. Other costs as assembly costs and overhead costs could also be estimated, but are even harder to predict. Needless to say total cost estimations are time consuming and will almost always differ from reality. For a product specification early on in a development project only a rough cost assessment is required because of the uncertainties of the product outcome. A more detailed estimate is more feasible during the production ramp-up phase. 13 2.11 Computer Software Along the course of this thesis, a number of different software programs have been used to make the product development process efficient and effective. As a Computer Aided Design tool (CAD), CATIA V5 have been used to make 3D-models of different concepts and has also been the main tool to obtaining the correct mass for each concept. For simulating forces and mechanically testing the concepts, the Finite Element Method program (FEM) in CATIA V5 was used throughout the thesis. The most unique software that was used was Altair Inspire which is an optimization tool. It uses FEM as well which is imbedded in the program. It creates a model with altered geometry in a way that lets a development team see where to use mass when the goal is to reduce weight. To decide which material is optimal CES EduPack 2013 has been used. This program uses Granta’s material database and it enables material indices and cost analysis to be set up against each other. The material index that was used in this thesis is stated below and is the index of choice when trying to create a beam with high bending stiffness and low weight (Ashby 2011). This can also be referred to the goal function in an optimization problem as in chapter 2.6. 14 3. Preparing the development process For the purpose of developing a new seat structure concept design, an extensive development process has been undertaken. Prior to the actual development work, which officially started with breaking down the functionality of the driver seat and the generation of concepts as described in the Concept generation section, a planning phase was undertaken. The planning phase served as a preparation for the actual development work, with important areas of pre-study being covered. First there was a need to find out what the actual users of driver seats would expect a new concept to achieve, without this input there would be no way of knowing if the final concept design could be implemented in the automotive market. Thus knowing that the concept solution should be light was simply not enough to ensure a successful product. After finding out what the customers expected and wanted of a future solution, the current solution had to be analyzed in order to provide a valuable point of reference. This was done by reverse engineering an actual driver seat from a leading car manufacturer and creating a digital representation of its seat structure in CATIA. This model will be referred to as the reference design throughout the rest of this report. Being a part of analyzing the reference design was also the load case analysis. The load cases for the most critical impact scenarios had to be found in order to be able to evaluate the structural performance and safety of the final concept design as well as the reference design. Finally all of the information gathered in this preparation stage had to be converted into measurable targets for the subsequent development effort; this was done by establishing engineering metrics, benchmarking, and setting target specifications. The outline of this section is as following:  Customer needs  Analysis of reference seat, which details the Reverse engineering, CAD-modelling, & Load case analysis  Initial product specifications, which details the Engineering metrics, Benchmarking, & Target specifications 3.1 Customer needs In order for the subsequent development effort to be successful it was important to determine what the actual customers of car seats wanted the concept design to achieve. It was already known that the company wanted a lightweight seat design but in order to accomplish that, a new concept essentially built from scratch, had to be developed. For that purpose the goal of low mass was not enough, first the necessary functions had to be studied as these would be the main drivers for the new concept design solutions. Understanding which functionality the customers expected of a car seat, and what they thought was missing in the current solutions, was an important first step in eliciting the functional criteria that would guide the rest of the development effort. The method chosen for finding these user needs was interviews performed with different types of car drivers, and automotive engineers. Complementary to these interviews, visits to three different retailers of personal cars were also conducted. Cars of different segments were tested and investigated and interviews were performed with salesmen in the different retailer locations. In this section the elicited customer needs, found by the mentioned methods, are summarized. The customer in this case was a mix of different stakeholders primarily consisting of car drivers and the companies that produce and sell automobiles. For the driver, the comfort and safety of the seat was of the highest importance. Since drivers come in many different shapes and sizes, the comfort not only depends on the shape of the seat, but to a large extent also its adjustability. For the driver it 15 is therefore important to be able to adjust the seat in length, height and in the angle of the seat back. Many drivers also need a supportive function in the lower back, while taller drivers appreciate extra leg support at the front of the seat. For car owners, there are some additional considerations regarding the seat. The cost of the seat should be included in the price of the car as few customers choose to pay extra for higher quality seats. The fuel consumption of the car should be kept to a minimum, and so should the environmental impact. Both these demands directly translate to a lower mass of the vehicle and thus a minimum weight of the seat was desired, as long as this does not violate the previously mentioned demands on comfort and price. For the producers of automobiles, it is vital to keep costs down since the profit margin in the current car industry is very low. This equates to keeping the costs of all components down while satisfying as many customer demands as possible. In the case of seats it is important to be able to incorporate all functions that the future car owner would want, as well as to make the seat aesthetically pleasing since this improves the overall image of the car specifically and the brand in general. As already mentioned this information was gathered from a mix of potential stakeholders for car seats, utilizing personal interviews. All answers were written down, organized and compared in order to get a quantitative result from the qualitative data collection approach. Having identified potential stakeholders, and gathered qualitative as well as quantitative data on their demands and desires regarding the driver seats of cars, ten customer needs where derived: 1) The seat should be comfortable when driving for an extended period of time 2) The seat should be comfortable when driving for a short period of time 3) The seat (car) should not be too expensive 4) The seat (car) should be as light as possible in order to reduce fuel consumption and pollution 5) The seat should protect the driver in case of collision 6) The seat should be adjustable in length 7) The seat should be adjustable in height 8) The angle of the seat back should be adjustable 9) The controls for seat adjustment should be easy to find and reach 10) The seat should be aesthetically pleasing Evaluating the elicited needs it was apparent that comfort and adjustability functions were connected and that they were also important in the eyes of the main user. Adjustment functions would thus be one of the main drivers for developing new design solutions, they were named Adjust driver. The other needs found would mainly act as constraints in the subsequent concept selection processes. The most important of these was safety performance. Since the focus of this thesis was to develop a new frame structure for the seat, user need number ten involving the aesthetics of the seat would not be used in the development process. 3.2 Analysis of Reference Seat In order to provide a starting point for the development process the elicited customer needs had to be converted into measurable targets. In order to do this more information had to be gathered, specifically regarding the seat structure, functions and impact load cases. The load cases would not only be a vital measurement of performance, but would also be used as input to the optimization process. This was to ensure that the optimization procedure would not result in designs that could not meet the customers’ expectations on safety. The analysis was based on a car seat of modern 16 design that was reverse engineered to create an understanding of the structure and functionality. The structure was then modeled in CATIA and analyzed using Altair Inspire which would serve as a reference design for the subsequent evaluation of concept designs during the development effort. Prior to analyzing the structure, the dimensioning load cases and forces had to be discovered, which was done using an existing seat requirements specification combined with dynamic calculations of energy in motion. Assumptions and calculations were later confirmed by a lead seat designer in an interview. In this section all the steps of the seat analysis are explained in detail. 3.2.1 Reverse engineering a current seat In order to create a reference model of a current car seat, a physical teardown of a driver seat from a leading car manufacturer was performed. This gave a detailed insight of the different components, shapes and materials of the entire seat. The first discovery made was that the seat back structure was divided into two segments, one lower supporting structure and an upper bar bent to shape. The reason for having separated the back into two parts was believed to be an active head restraint. This is a function that creates a support for the driver’s head during rear impact by automatically releasing a spring loaded mechanism that folds the backrest forward. The function was believed to be the main source of complexity in the reference seat and a driver of both mass and cost. The connecting parts between the seat back and the base, which also served as the reclining function, were sturdy and consisted of thick steel bars. The reason for this was believed to be that the seat experiences the highest loads near the recliner during impact, which was later confirmed in the load case analysis. The lumbar support function consisted of two steel wires and fabric, it was presumed as already light and simplified. The main structures of both the seat back and seat base were constructed using stamped sheet metal, which were 1.5 millimeters thick. The bent tube supporting the upper back was also interesting as the circular cross section might be used for the concept design as well. The sturdiness and length of the tracks that attach to the BIW were also noticed, another component which was believed to suffer large forces upon impact. All parts of the seat were measured and modeled in CATIA and then weighed; this information was used in the subsequent benchmarking table. Regarding the functionality of the seat, most of the information gathered in the customer needs process was confirmed. The seat back angle could be adjusted through the use of a rotational joint connecting the back structure to the base structure. The point at which the center of this rotational joint was located is called SRP (Seat Reference Point), as it would turn out during the load case analysis this was also the point used for measuring the resulting moment acting on the seat upon rearward impact. A high strength design resulting from the loads suffered at the center of the recliner mount drives both mass and cost. The seat base structure featured a height adjustment mechanism that works by rotating an asymmetrical bar connected to the track segment. A result of this is that the joints between the track segment and the base structure, receives an added level of complexity which includes the fact that several metal bars have to be added. This drives both mass and cost. The attachment structure had Figure 7: The reference seat halfway through the reverse engineering process 17 two main functions, transfer forces from the seat structure to the floor of the car and enable longitudinal adjustment of the seat relative the steering wheel and pedals. The first function was achieved through fastening elements connected to the floor and a locking mechanism in the rails. Longitudinal adjustment was achieved by the use of rails that slide in tracks, this means that the full length of the track and rail segment, at maximum length adjustment, is more than twice the length of the seat. Having these four long reinforced metal components in the design, drives both mass and cost of the seat, as well as the mass of the car body due to necessary added reinforcements in the floor. Both the seat base and the seat back contained a spring loaded suspension grate used to absorb driver movements. Constructed mainly of metal parts these components were judged as cost effective although not necessarily optimal in terms of mass reduction. The main conclusion of this reverse engineering process was that if the adjustment functionality could be removed or accomplished in alternative ways, there would be a significant potential for reducing mass and perhaps unit cost as well. 3.2.2 CAD-model of reference seat A CAD-model was built based on interpretations and measurements of the actual seat used for the reverse engineering process; this was to generate a reference model that would be used for the subsequent design and evaluation process. The model was weighed in CATIA and thus provided a fair estimation of mass when compared to the concept designs later on, as these were also modeled and weighed in CATIA. 3.2.3 Forces acting upon the seat In order to use finite element based software such as Altair Inspire, the user have to specify input forces and constraints on the design to be optimized. It was therefore important to understand the dominant load cases acting upon the car seat structure during regular use, as well as in a crash situation. Based on physical principles of energy in motion as described in appendix 1 (Load case analysis), and information compiled from actual car seat requirement specifications, an analysis of the load cases present in the car seat was carried out. Estimated forces and load cases were afterwards confirmed during interviews with an expert in car seat design. For clarity a distinction has been made between load cases inherent in regular use, called static load cases, and load cases that are present in different types of collisions, called dynamic load cases. Static load cases: The loads depicted here are meant to represent forces that could be exposed to the seat by the user during regular use i.e. in all other scenarios other than collisions. a) The complete seat shall withstand a minimum moment load of 1600 Nm forward around the H-point, representing the resulting moment of a force applied to the back of the seat, e.g. a passenger pressing against the seat back. Figure 8: Reference seat design modeled in CATIA 18 b) The complete seat shall withstand a minimum moment load of 2050 Nm rearward around the H-point, representing the moment caused by a driver throwing his weight at the backrest while sitting down. c) The seat shall withstand an extreme load of 1600 N anywhere on its structure, in order to withstand general mishandling of the product. Dynamic load cases: These were loads that the seat structure must withstand in order to protect the driver in a crash situation, which was an essential user need as stated earlier in this report. Using the physical law of conservation of energy and applying this theory to the case of slowing down a car from an initial velocity to a standstill in a specified distance, the force exerted on the seat could be approximated. Data on velocity and stopping distance has been gathered from the previously mentioned requirement specifications which contain actual crash test scenarios. It was revealed that a safety margin of 1.5 was accounted for in these specifications which eliminated the need for additional safety margins in the calculations for this thesis. For company secrecy purposes the requirement specification could not be published within this thesis report. After analyzing the different scenarios and consulting with lead car seat designers, the realization that the seat back was the main component affected in a crash situation, was made. More specifically the recliner joint was a critical dimensioning design element since this was the seat component affected by the largest load in any situation. Based on this, only the seat back and the load cases acting upon it are detailed in this section. The remaining load cases for the base structure and attachment structure can be found in appendix 1 (Load case analysis). Rearward impact with another car, initial velocity 54 km/h. This is the load case that exerts the largest force on the seat, and the seat back in particular. The mass of the driver has to be supported by the seat back as it decelerates from an initial velocity of 54 km/h to 0 km/h in a distance of 0.3 m which is the same stopping distance as the car. In this scenario there can be no elongation of stopping distance, as opposed to the frontal impact scenario where the stretch of the belt enables the driver to achieve a longer stopping distance, effectively reducing the force of impact. Adding to this force is the inertia caused by the mass of the seat back itself. It was believed that this load case primarily determined the necessary structural requirement of the seat back and recliner structure. d) The seat back and recliner structure should remain its structural integrity while subjected to a resulting rearward moment of 2100 Nm around SRP. This corresponds to a rearward impact of 54 km/h with another car. Frontal impact with unrestrained cargo, initial velocity 50 km/h In this scenario the seat back must be able to stop an object projecting from the backseat without failing. This creates a forward resulting moment on the seat back as the mass of the object must be supported as M_rear F_rear M_unrest F_unrest Figure 9: Load Case during rear impact Figure 10: Load Case during frontal impact 19 it decelerates from 50 km/h to 0 km/h in a distance shorter than the car’s stopping distance. Due to the higher point of impact when compared to the previous load case, this load case actually results in the largest moment around the recliner mechanism. Therefore this would be the actual load case determining the necessary structural requirements of the seat frame structure. e) The seat back and recliner structure should remain its structural integrity while subjected to a resulting forward moment of 2500 Nm around SRP. This corresponds to a frontal impact of 50 km/h against a rigid wall where an unrestrained box of 18 kg would hit the upper seat back. General attachment f) The rail & track segment attachment points should withstand a minimum force of 4 kN in any direction. This represents the structural integrity needed to withstand the force absorbed by the belt and inertia of the seat structure, during a front collision with another car. The main conclusion of this load case analysis is that the seat back will have to be designed to resist both forward and rearward dynamic bending while the seat base structure will be designed mainly to handle submarining force and static pressure. During the load case analysis it was realized that there were user needs apparent in the seat design that had not been previously elicited. These were mainly functional criteria assumed to always be found in a seat design such as the base structure supporting the driver from below and the back structure from the rear. Similarly the attachment structure is responsible for transferring loads from the seat to the car body. Thus a second set of functions that would drive new concept design solutions had been found, they were named Provide structural support. Together with the functional criteria Adjust driver the main functional criteria had now been found. 3.3 Initial Product Specifications In accordance with Ulrich & Eppinger’s (2012) theory of establishing product specifications, the following steps, involving metrics, benchmarking and analysis, have been used in the planning phase of this thesis: 1. Utilizing the customer needs as a starting point, a list of engineering metrics has been established. 2. Information about leading designs have been gathered and compared using the list of engineering metrics. 3. The benchmarking in step two has been used as a basis for setting measurable targets for the development effort. 3.3.1 Establish engineering metrics Customer needs are subjective and leave too much margin for interpretation; therefore there was a need for a translation of customer needs into a set of specifications that spell out in measurable detail what the future product had to do. This was done by establishing a set of engineering metrics that would connect the customer needs to targets for the subsequent development effort. Based on the assumption that meeting these specifications would lead to a satisfaction of the associated customer needs, the metrics were linked to specific customer needs where possible. 20 Table 2: Identified customer needs and their relevant importance The metrics were derived from analyzing the list of customer needs and the necessary functions that were discovered in the reference seat analysis. Generally thinking in the way of how an engineer would express that need and how to measure it in an objective way. It was decided to only include metrics that could be practically measured by the team during the development process; otherwise it was judged that impractical metrics would simply be overlooked. For instance since it was not in the scope of this thesis to perform any extensive testing, the crash test requirements found when scanning through several auto seat requirements had to be translated into load cases that could be used as input and constraints in the optimization, and FE- software instead. Table 3: Engineering metrics translated from the customer needs Customer needs No. Need Importance 1 The seat is comfortable when driving for an extended period of time 5 2 The seat is comfortable when driving for a short period of time 3 3 The seat is affordable for a majority of car manufacturers 4 4 The seat is as light as possible in order to reduce fuel consumption and pollution 4 5 The seat protects the driver in case of a collision 5 6 The seat supports a variety of driver heights 4 7 The seat supports a variety of driver widths 3 8 The seat supports the driver's back in different angles 2 9 The seat is easy to operate 3 Metric No. Need Nos. Metric Importance Units Adjustment system 1 6,7 Length adjustment interval of H-point (X-direction) 5 mm 2 7,6 Height adjustment interval of H-point (Z-direction) 5 mm 3 8 Back angle adjustment interval (around Y-axis) 4 degrees 4 1,2 Ankle angle interval 4 degrees 5 1,2 Knee angle interval 4 degrees 6 1,2 Elbow angle interval 4 degrees 7 1,2 Clear field of view for driver within size range 5 Subjective 8 1,2 Clear sight of instrument cluster for driver within size range 5 Subjective 9 9 Time to adjust seat 2 s Base structure 10 5 Max. allowed deflection from driver weight (Z-direction) 4 mm 11 5 Max. allowed deflection from submarining (X-direction) 5 mm 12 5 Max. allowed stress during impact 5 Mpa 13 1,2 Width 3 mm 14 1,2 Length 3 mm Back structure 15 5 Max. allowed deflection during impact (Pos. X-direction) 5 mm 16 5 Max. allowed stress during impact 5 Mpa 17 5 Max. allowed deflection from driver weight (Neg. X- direction) 4 mm Has side support 3 Binary 18 5 Has whiplash protection 4 Binary 19 1,2 Width 3 mm 20 1,2 Height 3 mm Attachment structure 21 5 Max. allowed deflection from driver weight (Z-direction) 4 mm 22 5 Max allowed stress during impact (any direction) 5 Mpa General 23 3 Unit manufacturing cost 5 SEK 24 3,4 Added manufacturing cost / Reduced mass 4 SEK/kg 25 4 Potential for reducing overall car weight 4 Subjective 26 4 Total mass / Reference mass 5 % 21 Most of the metrics followed logically from the user needs, but metric no. 25 might need further explanation. This was a derivative from the background study and the philosophy of a cascading mass reduction effect as explained in the Lotus report (Lotus 2010). This is a subjective evaluation of a concepts potential to cause mass reduction in other sub-systems other than the seat. According to lead seat designers the safety metrics, relating to need no. 5, will be dimensioning for the seat frame structure which is why these have been highlighted as important. Although theory states that a metric should only be linked to one specific customer need there were some cases where one metric was inevitably associated with multiple needs (Ulrich & Eppinger 2012). This was true in the case of metric no. 25: potential for reducing overall car weight. This metric had a direct association to customer need no. 4, the seat is lightweight, but also affected the cost of the car in a positive way and thereby need no. 3 as well. The complete list of engineering metrics can be seen in table 3. 3.3.2 Competitive benchmarking Before determining what values of the engineering metrics to aim for in the subsequent development process, reference values had to be gathered. The reference seat already analyzed could supply these values but in order to develop confidence that the resulting concept design would be feasible in the automotive market, information about competitors’ products was needed as well. Since the engineering metrics would be used to evaluate the future concept design, it was natural to use these as a basis for evaluating solutions that were already in the marketplace. The main source of information about the other products was a2mac1.com. This is a company that buys new cars and completely tears them down while documenting the entire process. Data such as component mass, dimensions and placement can easily be found on their website along with detailed images that helped this analysis. However two vital sources of information that could not be found in this way was the structural performance of a component and the component cost. Since these were integral for evaluating the success of a subsequent concept design they had to be explored in some way. The lack of performance data was replaced with the assumption that all available solutions at the very least had the same structural performance requirements as the reference seat. The reference seat is older than any of the other seats analyzed, which motivated this assumption. For the cost, a cost model had to be developed. This was based on the actual cost of components in a reference seat, with penalty functions adding cost for more material used, and added complexity such as more parts, difficult shapes or other manufacturing techniques. This would then adjust the cost of the other designs accordingly. The car J seat frame for instance is made up of few components which consist mostly of stamped sheet metal, a majority of its components are not painted, and it uses less material than the other designs. As a result it is approximately 35 percent cheaper than the reference design. At this stage the cost analysis was only based on the structure of the seat as this would be the main component of interest for this thesis. When the benchmarking chart had been constructed several conclusions could be drawn and the most important ones were:  Across the solutions there was a small difference in frame mass; one reason for this could be that the overall design was similar as well.  The undercarriage containing the tracks for length adjustment varied from 2.2 kg to 5.9 kg hinting that there is room for a potential weight reduction in this section of the seat.  The lightest solutions all made use of tubular construction elements in the frame; the tubular cross section could be an important feature in terms of mass reduction. 22 Metric Importance Units Reference Car B Car E Car F Car G Car J Adjustment system Length adjustment interval of H-point (X-direction) 5 mm (+125, -125) (+125, -125) (+125, -125) (+125, -125) (+125, -125) (+125, -125) Height adjustment interval of H-point (Z-direction) 5 mm 305(+30,-30) 345(+30,-30) 320(+30,-30) 300(+30,-30) 300(+30,-30) 300(+0,-0) Back angle adjustment interval (around Y-axis) 4 degrees angle -25,(+15,-30) angle -25,(+15,-30) angle -25,(+15,-40) angle -25,(+15,-45) angle -25,(+15,-45) angle -25,(+15,-50) Time to adjust seat 2 s 3 7 7 5 7 6 Base structure Max. allowed deflection from driver weight (Z-direction) 4 mm 1 1 1 1 1 1 Max. allowed deflection from submarining (X-direction) 5 mm 4 4 4 4 4 4 Max. allowed stress during impact 5 Mpa 700 700 700 700 700 700 Width 3 mm 495 500 490 480 490 525 Back structure Max. allowed deflection during impact (Pos. X-direction) 5 mm 11.2 11.2 11.2 11.2 11.2 11.2 Max. allowed stress during impact 5 Mpa 700 700 700 700 700 700 Max. allowed deflection from driver weight (Neg. X- direction) 4 mm 1.2 1.2 1.2 1.2 1.2 1.2 Has side support 3 Binary Yes Yes Yes Only on back Yes Only on back Has whiplash protection 4 Binary Yes Yes Yes Yes Yes Yes Width 3 mm 515 530 485 520 515 535 Attachment structure Max. allowed deflection from driver weight (Z-direction) 4 mm 0.5 0.5 0.5 0.5 0.5 0.5 Max allowed stress during impact (any direction) 5 Mpa 700 700 700 700 700 700 General Unit manufacturing cost (Frame + Padding) 5 SEK 860 958 834 692 617 593 Potential for reducing overall car weight 4 Subjective 1 1 1 1 1 1 Frame mass 4 kg 7,85 11,267 7,882 7,983 7,469 7,676 Adjusters mass 4 kg 7,8 5,882 5,372 3,905 4,01 2,248 Seat mass 5 kg 25,376 23,422 18,472 18,13 18,011 14,57 Total mass / Reference mass 5 % 100 92,3 72,8 71,4 71 57,4  Although the base frame did not vary considerably in mass across the solutions, the total mass revealed large differences depending on the number of adjustability functions incorporated in the seat. If it would be possible to reduce the number of functions with maintained comfort, this could possibly result in mass savings.  The dimensions of the seat frames and components were similar across vehicle segments, which could be due to similar size of the people they were designed for.  The cost varied mainly with the complexity of the seat structure, and again with the number of adjustability functions incorporated in the seat. A fraction of the Benchmarking table can be seen in figure 11 while the full table can be found in appendix 2. 3.3.3 Target specifications The target specifications were intended to both act as a guide throughout the development effort and to enable an evaluation of the final concept design against the customer needs and purpose of this thesis. These specifications would thus act as an agreement of objectives between the team and the main stakeholders as a way to measure the success of the end result. The target values were based on the benchmarking analysis as well as the reference seat analysis, and the focus was on mass reduction. This meant setting the mass reduction targets quite aggressively, which would serve as a goal function for the optimization procedure. The other specifications such as comfort, adjustability and cost, which would be used as constraints in the optimization procedure, were set close to their current level by aiming for the mean value of the designs analyzed. The reason for choosing two different targets for cost was to make sure a reasonable basic price would be Figure 11: An outtake from the benchmarking table (for complete version see appendix 2) 23 maintained, the unit manufacturing cost target, and to establish a relationship between amount of mass reduced and the resulting tolerated cost increase, the cost/reduced mass target. The critical targets were the specifications associated with need number five, safety. The other specifications are a mix of measurable dimensions and subjective evaluations that ensures a satisfaction of customer needs. The aim had now been set for the subsequent development process. The target specifications are compiled in a spreadsheet and can be seen in table 4. Table 4: Complete target specification which the thesis is based on Spec. No. Need Nos. Statement Imp. Units Marginal Value Ideal Value Adjustment system 1 6,7 Adjust length of H-point in X-direction. Min. interval = 210 mm. 5 mm 210 220 2 7,6 Adjust height of H-point in Z-direction. Min interval = 50 mm. 4 mm 50 60 3 8 Adjust back angle around Y-axis. Min interval = 14 degrees. 3 degrees 14 >14 4 1,2 Maintain ankle angle in comfortable position. Min. interval = 90-110 4 degrees 90-100 N/a 5 1,2 Maintain knee angle in comfortable position. Min. interval = 110-130 4 degrees 110-130 N/a 6 1,2 Maintain elbow angle in comfortable position. Min. interval = 80-165 4 degrees 80-165 N/a 7 1,2 Enable clear field of view for driver. Size range = 5-95 & 5 Subjective 5-95 Percentile 1-99 Percentile 8 1,2 Enable clear sight of instrument cluster for driver. Size range = 5-95 % 5 Subjective 5-95 Percentile 1-99 Percentile 9 9 Driver is able to adjust seat easily. Max. time to adjust seat = 10 s 2 s 10 6 Base structure 10 5 Support driver when subjected to a static load of 1600 N in Z-direction. Max. allowed deflection = 1 mm. 5 mm 1 <1 11 5 Support driver when subjected to a rear impact load of 2100 Nm in X- direction. Max allowed stress = Yield limit of material. 5 Mpa Yield limit of material N/a 12 5 Support driver when subjected to an unrestrained cargo load of 2500 Nm in (-X)-direction. Max allowed stress = Yield limit of material. 5 Mpa Yield limit of material N/a 13 5 Support driver when subjected to a submarining load of 4000 N in X- direction. Max allowed deflection = 5 mm. 5 mm 5 2 14 1,2 Width of seat base 3 mm 500 N/a 15 1,2 Length of seat base 3 mm 600 N/a Back structure 16 5 Support driver when subjected to a static load of 1600 N in X-direction. Max. allowed deflection = 2 mm. 5 mm 2 <2 17 5 Support driver when subjected to a rear impact load of 2100 Nm in X- direction. Max. allowed stress = Yield limit of material. 5 Mpa Yield limit of material N/a 18 5 Support driver when subjected to a rear impact load of 2100 Nm in X- direction. Max. allowed deflection = 14 mm. 5 mm 14 N/a 19 5 Support driver when subjected to an unrestrained cargo load of 2500 Nm in (-X)-d