CHAPTER 7 VALIDATION BY PROCESS WINDOW APPROACH BASED ON THE THEORY OF INVENTIVE PROBLEM SOLVING

Transcription

1 97 CHAPTER 7 VALIDATION BY PROCESS WINDOW APPROACH BASED ON THE THEORY OF INVENTIVE PROBLEM SOLVING 7.1 INTRODUCTION TO TRIZ INNOVATIVE DESIGN METHOD Taguchi methods are experimental statistical methods to optimize a given process technology with respect to an objective function defined as: Objective function = 10 Log 10 {Ideality} = 10 Log 10 {Benefits / (Costs + Harms)} Variance is in fact reduced in presence of noise (variations in the control parameters of the process) and thus the product/process becomes "robust" and "low cost". It is primarily an optimization technique and suggests "optimum" parameter settings for best results. However, should there be trade-off situations, the ANOVA plots point to the situations requiring "trade-off". This occurs when two or more process parameters have conflicting effect on two distinct desired characteristics (technical contradiction) or when low and high levels of one single parameter result in improving one desired characteristics while the middle level gives worsening characteristics (physical contradiction). The Taguchi method thus points out clearly the technical and physical contradictions and thus helps TRIZ in the sense of identification of the problem becomes easy. TRIZ tools can then be applied to resolve the contradictions. Exactly in the opposite way, the innovative solution concepts of TRIZ can be verified, evaluated, implemented

2 98 by planning an experiment where parameter settings can be optimized and best process can be selected. The TRIZ method is developed in the former Soviet Union by G. Altshuller who had analyzed over 400,000 patents and he classified them in 5 levels, which he called "Levels of Inventions. Altshuller proposed to exclude the two extreme levels viz. level 1 and level 5 from his inventive problem solving tools. As one can see, the tools become progressively more powerful as the personnel move from level 2 to level 3 and to Level 4. The levels 2 and 3 are termed as "innovative" and level 4 as "inventive". Each level has its own defined problems and its own problem solving tools. The aim is to move towards ideality. In this sense the level 4 is not better than level 3 if level 3 solution brings it closer to ideality. Each higher level also requires more detailed analysis and resources. One of the more interesting laws of TRIZ is the Law of increasing ideality is given in Equation (7.1). It states that technical systems evolve toward increasing degrees of ideality, where ideality is defined as the quotient of the sum of the system's useful effects (U i ), divided by the sum of its harmful effects (Altshuller et al 1999 and Li et al 2007). Ideality U H i j (7.1) Useful effects include all the positive results of the system's functioning. Harmful effects (H j ) include all the negative effects such as cost, undesired increase in size, energy consumed, pollution, danger, etc. The ideal state is one in which there are only benefits and no harmful effects. It is to this state that all product systems strive to evolve. Product developers aim both to design product with greater benefits and to reduce the cost of labor, materials, energy, and harmful side effects. Normally, when improving a benefit would result in increased harmful effects, a trade-off is made, but the law of ideality drives designs to eliminate or solve any trade-offs or design contradictions. The ultimate ideal final result would theoretically be a product where the beneficial function exists but the machine itself does not. In solving

3 99 engineering problems, the engineers usually encounter the system s contradiction, i.e., when a system improves one certain engineering feature, another engineering feature worsens. As soon as the engineer encounters the contradiction, he usually adopts a compromise for handling and does not really solve the problem. Altshuller thought that problems encountered in the invention could be overcome by the scientific method. Inventions inevitably encounter problems and contradictions, and then leap from compromising dilemma. As a result, Atshuller studied worldwide famous patents to eliminate the contradiction method TRIZ through the integration of interdisciplinary technology. During the problem solving process, TRIZ applies some tools, including separation principle, substance-field analysis, 76 standard solutions. Among which, the most famous and practical one is the contradictive matrix table that compiles frequently encountered technical contradictions. Table Engineering Parameters 1. Weight of moving object 11. Tension, pressure 21. Power 31. Harmful side effects 2. Weight of nonmoving object 12. Shape 22. Waste of energy 32. Manufacturability 3.Length of moving object 13. Stability of object 23. Waste of substance 33. Convenience of use 4. Length of 14. Strength 24.. Loss of 34. Reparability nonmoving object information 5. Area of moving 15. Durability of 25. Waste of time 35. Adaptability object moving object 6. Area of nonmoving object 16. Nonmoving object Durability 26. Amount of substance 36. Complexity of device 7. Volume of moving object 17. Temperature 27. Reliability 37. Complexity of control 8. Volume of nonmoving object 18. Brightness 28. Accuracy of measurement 38. Level of automation 9. Speed 19. Energy spent by moving object 29. Accuracy of manufacturing 39. Productivity 10. Force 20. Energy spent by nonmoving object 30. Harmful factors acting on object

4 engineering parameters are given as Table 7.1 and 40 inventive principles as given in Table 7.2 to establish a contradictive matrix and provide solutions for technical contradiction (Marsh et al 2004). Table 7.2 TRIZ 40 inventive principles 1.Segmentation 11. Beforehand cushioning 21. Skipping 31. Porous materials 2.Taking out 12. Equipotentiality 22. Blessing in disguise 32. Color change 3.Local quality 13. The other way round 23. Feedback 33. Homogeneity 4.Asymmetry 14. Curvature 24. Intermediary 34. Discarding and recovering 5.Merging 15. Dynamization 25. Self-service 35. Parameter changes 6.Universality 16. Partial or Excessive action 26. Copying 36. Phase transitions 7.Nested doll 17. Another dimension 27. Cheap shortliving objects 8.Anti-weight 9.Preliminary anti action 10.Preliminary action 18. Mechanical Vibration 28. Mechanics substitution 19. Periodic action 29. Pneumatics and hydraulics 20. Continuity of useful action 30. Flexible shells and thin films 37. Thermal expansion 38. Strong oxidants 39. Inert atmosphere 40. Composite materials First of all, technical personnel propose the encountered technical problem or conflict. The designer then introduces TRIZ and transfers it into a TRIZ problem. TRIZ tools are used to obtain the general solution of that TRIZ problem and finally transfer that general solution to a solution applicable in engineering technology. The TRIZ resolving process is shown in Figure 7.1. The TRIZ problem solving process can be divided into the following steps:

5 101 Figure 7.1 TRIZ resolving flowchart Step 1: Step 2: Step 3: Confirm and clarify the encountered technical problem and difficulty that needs to be solved. The technical problem goes through TRIZ s 39 engineering parameters and becomes a TRIZ problem. Confirm whether technical contradiction or physics contradiction exists in the TRIZ engineering problem. Physics contradiction improving and worsening engineering parameter is the same parameter, i.e., self-conflicts, go to step 7; Technical conflict different conflict exist or a single improving/worsening engineering parameter. Step 4: Different parameter technical contradictions exist, then directly enter TRIZ contradictive matrix to locate the corresponding

6 102 principle. If a single improving or worsening parameter can be located only, or no conflict existed among parameters, go to step 6. Step 5: Through a TRIZ contradictive matrix (improving parameter is set as Y-axis, while worsening parameter is set as X-axis), conflict problem can obtain inventive suggested principle, go to step 8. If a corresponding position is an empty matrix, then go to step 6. Step 6: Step 7: Step 8: Step 9: Set parameter is introduced into Single Engineering Inventive Principle (SEIP) method (Chen and Liu 2001) to obtain inventive suggested principles, go to step 8. TRIZ separation principle is used to obtain suggested principle through separation of time, space, and substance. Obtained TRIZ suggested principle conducts proper selection to obtain TRIZ solution. TRIZ solution transfers to inventive design reference for solving an actual engineering problem to produce a final solution. Step 10: When obtained suggested principle cannot satisfy final demand, engineering parameter can be redefined to go through steps 1 to 9 once again and obtain new suggested inventive principle, go back to step 3. Regarding the modification motive, the contradiction matrix engineer parameters are not convenient for use due to its rather big size. In addition, not all the engineering parameters are involved in the problems. The engineering parameter closest to the problem is selected according to different directions of demands. Taking into consideration of these two points, the present study made slight modification on the selection of engineering parameters and contradiction matrix. After confirming the problems for

7 103 modification, this study first analyzed the problem, then confirmed the issues, and finally related the problems for modification with 39 engineering parameters. Several engineering parameters closest to the modification problems are selected among 39 engineering parameters. Next, these parameters are put into the matrix, which size and range depend on the number of parameters selected INTRODUCTION TO PWA BASED ON TRIZ Optimization is a mathematical discipline that concerns the finding of minima and maxima of functions, subject to so-called constraints. Today, optimization comprises a wide variety of techniques which are used to improve business processes in practically all industries. Another recent trend is the combination of optimization techniques for problems that do not lend themselves easily to one technique alone. Today, these techniques prove to deliver robust engines that provide very high quality solutions for even very large problem instances. Sand casting process, in general, involves a large number of parameters affecting the various casting quality features of the product. In foundries, the commonly used process optimization methods, including trial and error, linear and non-linear methods, numerical simulation method, Taguchi design, grey scale and RSM method (Chien et al 2011). However, all of these methods provide limited improvement of quality and economy for castings because they are only able to deal with sequential procedures, waste of substances and using of specified expensive software. In fact, the optimization of casting process involves not only the adjustment of parameters, but also the improvement of optimization procedures and measures. Therefore, the optimization process procedures and measures must be taken into account so as to obtain adoptability in optimizing the sand casting parameters without any changes using PWA based on theory of inventive problem solving (TRIZ) which actually relies on an engineer s

8 104 educational background or his own limited experience. The two optimization techniques are used to determine the optimum operating conditions for the system or to determine a region of the factor space in which operating requirements are satisfied. Most of the process optimization will be done by the specialized software with knowledge workers and it consuming more time for processing, complexity in the operational procedures, waste of all kind of resources and multiple choices of operational methods raises confusion to adopt to solve the process optimization. To overcome these difficulties, the process window approach can be used to quickly determine the impact of process changes and compare different process parameters. It is statistical that quantifies the robustness of a manufacturing process which conform the process improvement with optimized parameters of manufacturing process. In this case, a highly efficient and more practical method, which is used for sand casting process optimization is expected. A new method is developed in this work to validate the optimized process parameters by PWA and by taking advantage of innovative heuristic power of TRIZ this PWA could also treated as adoptable process optimization approach. The idea of applying TRIZ to improving the optimization process for casting a flywheel component is realized by using this method. The success of the improved optimization process PWA for the sand casting process shows that the proposed method is of great potential for foundry. 7.3 RESEARCH METHODOLOGY In the previous chapters, analysis to optimize the various critical process parameters and the interaction among them are carried out with the help of Taguchi s method of experimental design and to optimize the results obtained and to make the analysis more precise and cost effective, the most significant critical parameters are also analyzed with response surface methodology and the results are presented.

9 105 This chapter focuses on the implementation of the proposed PWA base on TRIZ concept in order to optimize the variables of the sand casting process to minimize the casting defects developed in flywheel component. Eventually, the optimized parameters obtained using Taguchi method and RSM are compared with PWA and validated in a foundry Process Window Design Analysis The quality of a green sand casting is the result of a great number of parameters. Taguchi method and RSM cannot judge and determine the effect of individual parameter on the entire process. The significant factors and/or their interactions are identified, for various trial conditions and the parameters which significantly influence the casting defects. However, the both process requires systematic procedures, software support and knowledge, more sequential steps and time. To overcome these difficulties PWA is used for validating the optimal parameters set by Taguchi and RSM approach and itself offering a simplest way of selecting the optimal parameter levels. Process Window Approach is a simple statistical tool that quantifies the robustness of a manufacturing process which conform the process improvement with optimized parameters of manufacturing process, as reported by Jim Hall and Phil Zarrow (2002). The data are available from the set of runs; the lowest probability can be confirmed by calculating the Cpk is given in Equation (7.2): C ( USL X ) 3 or ( X LSL ) 3 (Whichever is less) (7.2) pk If the mean value is centered between the LSL and the USL, then both (USL Mean) and (Mean LSL) are equal to [(USL+LSL) / 2], and both equations yield the same value of Cpk. If the mean value is not cantered, then one of these differences will be less than [(USL+LSL) / 2], resulting in a lower Cpk.

10 106 For an analogue-style specification, the lowest defects could exist at some value or sweet spot within the selected process window of Cpk but not at its center. However, this relationship would most certainly be very productspecific, so no general guidelines can be given. If such data available for a specific product, then the process setup should be adjusted to move the mean closer to this sweet spot to produce less defects. This approach is consistent with the weakest link philosophy, where an out-of-specification value at any location constitutes a defect, regardless of all others. These ranges are the process windows within which zero are minimum defects will be achieved. For some parameters, the minimum or maximum are true, binary statistical specification limits where any value inside the range is good and any value outside the range is a defect. The overall Cpk value provides an excellent metric for qualitatively evaluating processing parameter adjustments. If relationship between Cpk and quality (defects) is clearly understood, an acceptable maximum Cpk value can be established for PWA. If a modeling algorithm is available for the process, then changes can be evaluated without actual physical experiments and the ultimate method for optimizing a process setup using search engine. The search engine would change all of the setup variables, using the model s algorithms to calculate the PWA for each combination. Obviously, two fundamental requirements are needed for such a system to be successful. First, the model must be accurate for the full usable range of all setup parameters. Second it should be able to self-correct based on confirmation test data versus results predicted by its algorithms. Once the model predictions can be relied upon, the search engine must efficiently find the optimum process setup. It should evaluate the entire range of all parameters and all combinations in a reasonable time. For example, overnight computing is acceptable to optimize the casting process. If

11 107 certain parameters are more critical to the overall process or more likely to cause defects, their priorities can be increased by tightening the limit values for the PWA calculations inside the actual specification limits. Mathematically, this approach causes the value of Cpk to fall closer to the each one or not included in the process window are negligible. Figure 7.2 Flowchart for the PWA validation for optimization process

12 108 It requires confirmation run of the final set up, and it is expensive when the process yield more defects if it is directly used as optimization model. Hence, the effective use of this model to gain good results by validating the optimized process parameters with help of Taguchi and RSM analysis. This work proposes an innovative method PWA based on the TRIZ for quickly evaluating the robustness of a process set up by validating the casting process optimization, and realizes an idea applying TRIZ so as to remove the casting defects. The successful validation of sand casting process parameters optimization demonstrates the feasibility of the PWA based TRIZ method. This work hopes to provide foundry personnel with innovative design ideas under research and for practical application. Based on the above discussions, the functional flowchart is prepared up to the level of satisfy the PWA practical procedures implementation effectively and it is given in Figure Process Window Approach (PWA) PWA measures how well a sand casting process fits into a higher capability level of the process limit known as the specification limit. The specification limit is the tolerance allowed for the process obtained in the casting production in terms of Cpk and is statistically determined. In the sand casting foundry, these specification limits are known as the process window, and values that a plotted inside this window is known as the process window approach. Using PWA values, casting process performance can be accurately measured, analyzed, compared. For an analogue-style of specification (Cpk), the lowest defects could exist at some value or sweet spot within the range but not at its centre. If such data available for a specific product, then the process parameter set up should be adjusted to move the mean closer to this sweet spot to produce less defects. Once the analysis prediction can be relied upon,

13 109 the PWA must efficiently find the optimal parameter set up. It should validate the entire range of all parameters and all possible combinations. In this work, an innovative approach to validate the sand casting process optimization analysis. Optimal solution arrived from the Taguchi analysis and RSM, the set of parameters are again analyzed to its significance on the defects. Based on percent contribution, it is identified and the four significant parameters are analyzed in the response surface methodology. The PWA is a powerful tool for quickly validating the performance measurement thereby robustness of the process parameter setup of RSM. In this approach the most significant parameters moisture content, permeability, volatile content and mold pressure is selected and other parameters are kept as it is. These parameters are more critical to the overall process or more likely to cause defects. PWA considering all interaction parameters separately and gives the optimal result for the individual parameter by selecting sweet spot through process window. Table 7.3 shows the analysis of TRIZ contradictive matrix. Through the matrix, the improving parameter is set as adoptability to validate the Taguchi, RSM optimization process and this PWA is a innovative approach to find the optimal parameter level which supplement the other optimization process (improving parameter is set as Y-axis), as explained by Hsieh and Chen (2010). The undesirable secondary effect, i.e. worsening parameter is selected initially as convenient of use, area of non moving object, durability of non moving object and waste of substances. However the waste of substance is selected as worsening parameter (X-axis), among all three parameters having significant with each other and selection procedures.

14 110 Table 7.3 Contractive matrix- analysis of engineering parameters Improving factor ( Y) Worsening factor (X) Improving factor ( Y) Worsening factor (X) Improving factor ( Y) Worsening factor (X) Since TRIZ theory originates from invention in different technical fields and integrates physics, chemical, and engineering sciences, the theory is beyond field limitation and can be fully applied in other industries. In the preliminary stage and application of new technology, TRIZ can fully bring its effect into fully play, especially when there is a lack of reference data and insufficient experience. The advantages of TRIZ are that it can transfer system contradiction and conflict into a useful factor and be compiled into an effective method to solve the problem to facilitate research and for design personnel to rapidly solve the problem (Liu Feng et al 2007). The matrix modification employed in the study is illustrated by selecting the engineering parameters for improvement and worsening feature. Engineering parameter to avoid worsening feature is waste of substances and improve the engineering characteristics is adaptability. The selected parameters are inputted into the contradiction matrix of modification for

15 111 crossover to obtain the recommended number of the 40 innovation methods: 15- Dynamics, 10- Preliminary action, 02- Taking out, and 13- The other way round. Use of TRIZ contradictive matrix is shown in Table 7.4: when #35- Adaptability is selected as an improving feature and #23-waste of substance as a worsening feature, respectively. The corresponding matrix obtains suggested inventive principles as 15,10,2, and13. Table.7.4 The usage of TRIZ contradictive matrix Worsening feature # 23 Improving feature waste of substance #35 Adaptability 15- Dynamics 10- Preliminary action 02- Taking out 13- The other way round As per the direction of Atshuller study, Level 2 is selected for sand casting optimization process with minor improvements by removing some contradictions (45% of all the patents) and uses 40 principles to separate and solve technical contradictions. It requires knowledge from different areas within the same field. From the selected combination inventive principles with respect to X and Y axis and most frequent improve parameters (35- adaptability) are taken into account for selecting the recommended inventive principles for the casting process optimization problems as given in Table 7.5.

16 112 Table 7.5 Selection of inventive principles from the selected X and Y axis parameters X axis Y axis Principles Complexity of control Adaptability 15,1 Convenience of use Adaptability 15,16,34,1 Area of non moving object Adaptability 15,16 Durability of non moving object Adaptability 2,16 Reparability Adaptability 7,1,4,16 Energy spent by moving object Adaptability - Length of moving object Adaptability 35,1,2,29 Adaptability Length of moving object 14,15,16,1 Waste of substances Adaptability 15,10,2,13 Adaptability Waste of substances 15,10,2 Manufacturability Adaptability 2,13,1,32 Adaptability Manufacturability 13,15,2 Based on the inventive principles for the corresponding X and Y axis parameters, the number of repeated inventive principles are identified and from the analysis the X axis parameter is confirmed and its related inventive principles are listed according to the rank given in the Table 7.6. The selected most combination inventive principles for the respective parameters for validating and optimizing sand casting parameters are: Convenience of use (X-axis) Area of non moving object Durability of non moving object Adaptability (Y-axis) 15, 16,1,34 (principles) Adaptability 15, 16 Adaptability 2, 16

17 113 Waste of substance Adaptability 15, 10, 2, 13- selected Adaptability Waste of substance 15,10,2 Table 7.6 Confirmation of inventive principles for the problem Inventive Principle Number No of times occurrences 1 6 Segmentation 2 6 Taking out 7 1 Nested doll Name of the principles 10 2 Preliminary action 13 3 The other way round 14 1 Curvature 15 7 Dynamics 16 5 Partial or excessive actions 29 1 Pneumatics and hydraulics 32 1 Color changes 34 1 Discarding and recovering 35 1 Parameter changes The definition for the inventive principles of segmentation and partial and excessive actions will not satisfy the specific problem solution PWA which includes mathematical and analytical research for continuity. Hence, the preliminary action and the other way round principles are taken in addition with the taking out and dynamics principles. Principle 15- Dynamics: Allow a system or object to change to achieve optimal operation under different conditions. Split an object or system into parts capable of moving relative to each other. If an object or system is rigid or inflexible, make it movable or adaptable. Increase the amount of free motion.

18 114 PWA is the statistical innovative tool to validate and optimize the process parameter set up having a simple logical algorithms rather than using 0, 1 coding in DOE and RSM methods. PWA uses a simple calculation of, Cpl, Cpu and Cpk instead of calculating S/N ratio, mean and sum of squares. PWA determined by the Cpk value and it can be play coin for the PWA matrix throughout the process. It reduces and eliminates more number of operations used in various optimization techniques and dynamism of these techniques can be followed and attained without changing its effectiveness. It can be used for all models of optimization to validate the measurement effectiveness with shorter cycle time and efficiently. Principle 10- Preliminary action: Perform, before it is needed, the required change of an object (either fully or partially). Pre-arrange objects such that they can come into action from the most convenient place and without losing time for their delivery. PWA is a tool to identify the place called sweet spot where the Cpk value is greater and it denotes that the selected optimized parameters meeting the objectives. Then, a process window will form as per the algorithm to find next optimal values. The way in which the personnel can be found the area of optimal values and adjusting the process parameters according to it just like in RSM sum of squares. It can be an alternate model to the Taguchi and RSM to optimize the process parameters. Optimization analysis with Taguchi and RSM, PWA is a tool to find the optimal area of result quickly without use of any statistical software. Principle 2- Taking out: Where a system provides several functions of which one or more is not required (and may be harmful) at certain conditions, design the system so they are or can be taken out. Software based orthogonal design, experimental analysis for parameter optimization is not required for TRIZ based PWA validation

19 115 method. It requires simple Cpk and sigma values and simple logical knowledge is essential to do the process without changing its technicality. Principle 13- The other way round: Use an opposite action(s) used to solve the problem; make moveable objects fixed, and the fixed objects movable; Turn the object, system or process upside-down. PWA is a totally different concept in the form of mathematical matrix as in Taguchi orthogonal array and RSM design coding. It is also a reversible action to validate and optimize the process where the personnel approaching the tool with respective of use. Taguchi and RSM values are selected based on Smaller is better concepts, relates to casting defect. But in PWA the greater Cpk value determines the better parameter level at casting defects are less. 7.4 SUMMARY The process window approach is used to quickly determine the impact of process changes and compare different process parameters rather than using of specialized software with knowledge workers. This chapter focuses on the implementation of the proposed PWA base on TRIZ concept in order to optimize the variables of the sand casting process to minimize the casting defects developed in flywheel component. From the selected combination inventive principles with respect to X and Y axis and most frequent improve parameters (35- adaptability) are taken into account for selecting the recommended inventive principles for the casting process optimization problems is given. The detailed result analysis given in the chapter 10.