Robust konstruktion som arbetssätt inom tillförlitlighetstekniken Detta webbinarium är ett komplement till Magnus Arnérs seminarium i våras 1
Failure rate - the Bath-tub curve z(t) Failures due to Manufacturing Variation & Different use environment Failures due to Extreme stress Failures due to wear-out, fatigue, etc t 2
Failure rate - the Bath-tub curve z(t) Failures due to Manufacturing Variation & Different use environment Failures due to Extreme stress Failures due to wear-out, fatigue, etc t 3
Sources of variation: Noise On the demand side Between customer/usage Environmental System interactions On the capacity side Manufacturing Wear, fatigue,... 4
Instead of working with failure rates Make the product insensitive to variation! But when and how? Where in the complete value creating process? Systems engineering Product development Manfacturing... 5
System engineering process a very simplified picture: 3 phases Requirement Concept (generation and selection) Improvement 6
The Systems Engineering process Clausing et al. 7
The Systems Engineering Process Clausing et al. Requirement Concept Improvement Requirement Concept Improvement Requirement Concept Requirement Improvement Concept 8 Improvement
The Problem A world of variation noise factors Key Product Characteristics (Critical to Quality Characteristics), i.e. product attributes, such that excess variation is not acceptable to the customer. Find a design solution such that it is insensitive to noise factors it deliver its function irrespective of noises 9
The Parametric Design solution Belongs to the Improvement phase When The Failure Mode factory has been operating now, try to find a way out by changing the design parameters Too late? Too few degrees of freedom? We cannot always wait for the Parametric Design solution 10
An illustration: Gamlestaden factory The self aligning bearing A Creative Reliability Improvement 1907 Sven Wingquist 11
What to do in the Requirement phase? 12
What is important- where are the noise factors? QFD important characteristics Critical to Quality? Noises? Important that already in the Requirement phase we understand important sources of variation noise factors Demand from Customer Usage Usage Environment Component interaction Demand eventually creating capacity reduction But we also have to think about manufacturability and noise from manufacturing Iterative processes! 13
P-Diagram Störningar Källor till variation N Signaler* M Process Output Resultat y *Vi vill skapa olika värden I processen Styrparametrar x 14
Initiate a Variation Mode and Effects Analysis VMEA Instead of an FMEA where failure modes are investigated a VMEA goes deeper down to the sources of variation that may create this failure mode 15
What to do in the concept generation phase 16
How to think about variation in the creative concept generation phase No one systematic method like parameter design Be inspired by other creative solutions 17
Creative Robust Design 18
Creative solutions: some illustrations The self aligning bearing A Creative Reliability Improvement 1995 1907 Sven Wingquist 19
Inspiration Creative yesterday commonplace today Replacing the chain with a wire DFA - solutions 20
Inspiration Creative yesterday commonplace today DFA type solutions m 1 m 2 m 1 m 2 m m 21
Mistake proofing Inspiration for creative robustness Poka Yoke
Poka Yoke Mistake-proofing (fool proofing) avoiding inadvertent errors Shigeo Shingo (Toyota Prodution System) Simple idea Make it impossible or very hard to make it wrong If something is wrong make it easy to find out. If something goes wrong reduce the effects Simple ideas are not simple to get
Informative inspection (find a mistake before it makes any harm)
Mistake prevention Make it hard or impossible to make a mistake
Some illustrations
Prevent the influence of mistakes
Why are manhole Covers always round?
Poka Yoka Some references Shingo S. (1986). Zero quality control: source inspection and the pokayoke system, transl. by Dillion, AP. Productivity Press, New York. Grout J. (2007). Mistake-proofing the design of health care processes. (Prepared under an IPA with Berry College). AHRQ Publication No. 07-0020. Rockville, MD: Agency for Healthcare Research and Quality; May 2007. http://pokayoke.wikispaces.com
Concept generation and selection Idea generation 31
Pugh s Controlled Convergence Controlled convergence (CC) Initial number of concepts Concept Generation (CG) Initial number reduced New concepts added Further Reduction (FR) CC CG CC Further Addition (FA) FR SelectionCriteria: Including sensitivity to variation Concept selected 32
What to do in the Improvement phase? Parameter Design already discussed by Magnus Arnér You could try to use VMEA to further understand the critical sources of variation those affecting Critical to Quality characteristics 33
What to do in the Improvement phase? Parameter Design already discussed by Magnus Arnér You could try to use VMEA to further understand the critical sources of variation those affecting Critical to Quality characteristics If parameter design does not help we have to try Tolerance design but that is a different webinar! 34
Department of Quality P diagram Within the framework of Robust Design it is common to represent a product or a system with the P-diagram: The P - diagram is a useful tool, but it is 1) basically for illustration purposes 2) it is sometimes difficult to find areas or systems for which to build a P - diagram Per Johansson & Alexander Chakhunashvili 2004-04-02 35
Department of Quality P - diagram and VMEA VMEA is an attempt to address: the problem of finding the system response the problem of identification and prioritization of noise factors the problem of locating areas (subsystems) for which the identified noise factors may be harmful Before introducing the VMEA method, let us recall a few things that are important for its successful application. Per Johansson & Alexander Chakhunashvili 2004-04-02 36
Department of Quality Variation and Product Characteristics Customer Wants and Needs QFD Product Characteristics (PCs) Standard Product Characteristics (SPC) Key Product Characteristics (KPC) KPC is a product characteristic for which reasonably anticipated variation could significantly affect the product's safety, or compliance with governmental standards or regulations, or is likely to significantly affect customer satisfaction with a product; see General Motors Key Characteristic Designation System (1998). Per Johansson & Alexander Chakhunashvili 2004-04-02 37
Department of Quality Product Characteristics and Noise Factors Noise factors are sources of variation that cause deviations in important product characteristics. They can be unknown; or unobservable, although known (their observation requires unavailable or too expensive equipment); or uncontrollable, although known and observable (it is not possible to set the factors at any specific intended value). Sub-KPCs are characteristics of a system or sub-system which affect the KPC. They are generally known and controllable by engineers. Per Johansson & Alexander Chakhunashvili 2004-04-02 38
Department of Quality VMEA Procedure The four basic steps 1. KPC Causal Breakdown: Sub-KPCs and NFs Once KPC is selected, it can be usually decomposed into a number of elements called Sub-KPCs and Noise Factors. This is usually done with an Ishikawa diagram. 2. Sensitivity Assessment In the second step of the procedure engineers assess the sensitivity of the KPC to the action of each Sub-KPC and the sensitivity of each Sub-KPC to the action of Noise Factors according to the given criteria. 3. Variation Size Assessment In the third step of the procedure engineers examine Noise Factors and assess their variation size according to the given criteria. 4. Variation Risk Assessment and Prioritization Based on the assessment given in the previous two steps, a Variation Risk Priority Number (VRPN) is calculated for each Noise Factor and then for each Sub-KPC. Per Johansson & Alexander Chakhunashvili 2004-04-02 39
Department of Quality Assessment Criteria, Sensitivity Per Johansson & Alexander Chakhunashvili 2004-04-02 40
Department of Quality Assessment Criteria, Size of Variation Per Johansson & Alexander Chakhunashvili 2004-04-02 41
Department of Quality Calculation of VRPN Variation Risk Priority Number (VRPN) where S 1 indicates the sensitivity of the KPC to the action of the Sub- KPC influenced by the Noise Factor, and S 2 indicates the sensitivity of the Sub-KPC to the action of the Noise Factor and finally V indicates the Noise Factor s variation size. If a Sub-KPC is influenced by several Noise Factors, it is possible to calculate a VRPN related to the Sub-KPC as a sum of the VRPNs of the Noise Factors acting on it: Per Johansson & Alexander Chakhunashvili 2004-04-02 42
Summary Reliability management and improvement may be better off if we work with insensitivity to variation rather than with failure rates Closer to the sources of problems Easier to work with in early stages of system/product design Calls for creativity Focus on sources of variation one tool may be VMEA 43
Read more Pär Johannesson, Bo Bergman,Thomas Svensson,Martin Arvidsson,Åke Lönnqvist,Stefano Barone, Jacques de Maré (2012) A Robustness Approach to Reliability, in Quality and reliability engineering international Bergman B, de Maré J, Lorén S, Svensson T (eds). Robust Design Methodology for Reliability - Exploring the Effects of Variation and Uncertainty. Wiley: Chichester, 2009. Bergman B, Klefsjö B. Quality from Customer Needs to Customer Satisfaction. Studentlitteratur: Lund, 2010. 44