Core Tools: The Alph APQP, PPAP, FMEA, S! Jd Marhevko — Accuride Corporation, Shainin Medalist, ASQ Fellow, CSSBB, CMQ/OE, CQE, ASQ World Conference — Session 112 — MayThe FIVE Core Tools 1. APQP: Advance Product Quality Planning: Guidelines for a product quality plan to develop a product or service that satisfies the customer 2. FMEA: Failure Modes and Effect Analysis: Methodology used to ensure potential problems have been considered and addressed throughout the product and process development process (Ex. APQP). Traditionally includes the Control Plan (CP) 3. PPAP: Production Part Approval Process: Ensures product consistently meets customer engineering specification requirements during production run at the quoted production rate 4. MSA: Measurement Systems Analysis: Guidelines for assessing the quality of a measurement system where readings are replicated 5. SPC: Statistical Process Control: Basic graphing statistical tools that enable process control and capability for continual @) improvement ASOOther Sample Manuals eT ea TT SOOO Ee @) ASOCore Tool inferences in ISO/IATF 16949:2016 Core 1S0 9001:2015 IATF 16949:2016 Tool (Core Tools NOT Specified) (Core Tool Inferred/Referenced) APQP 8.1 Operational Planning 8.1.1 Operational Planning and Control and Control 8.2 Requirements for Products and Services 8.2 Requirements for 8.3 Design and Development of Products and Services Products and Services 8.4 Control of Externally Provided Processes, Products 8.3 Design and Development and Services of Products and Services 8.4 Control of Externally Provided Processes, Products and Services FMEA 6.1 Actions toAddress Risks 4.4.1.2 Product Safety and Opportunities 8.3.5 Design and Development Output 9.1, Monitoring, Measurement, Analysis and Evaluation General 6.1 Actions to Address Risks and Opportunities 8.3 Design and Develop of Products and Services (8.3.3.3, 8.3.5.1, 8.3.5.2] 8,5 Production and Service Provision [8.5. 8.7 Control of Non-Conforming Outputs [8.7.1.4, 8.7.1.5] 9.1 Monitoring, Measurement, Analysis and Evaluation ‘General 9.2.3 Manufacturing Process Audit 10.2 Non-Conformity and Corrective Action [10.2.3, 10.2.4] 40.3.4 Continual ImprovementCore Tool inferences in ISO/IATF 16949:2016 Core 180 9001:2015 Tool (Core Tools NOT Specified) CP* 8.3.5 Design and Development Outputs: 8.5.1 Control of Production and Service Provision 8.6 Release of Products and Services 8.7 Control of Non-Cenforming Outputs: IATF 16949:2016 ‘Core Tool Inferred/Referenced) 8.3.5.2 Manufacturing Process Design Output 8.5 Production and Service Provision (8.5.1.1, 8.5.1.3, 8.5.6.1.1] 8.6 Release of Products and Services 8.7 Control of Non-Conforming Outputs 9.1.1.2 Identification of Statistical Tools 9.2.2.3 Manufacturing Process Audit 10.2.3 Problem Solving Annex A. Contro! Plan PPAP 8.3.4 Design and Development Control 8.3.4.3 Prototype Program 8.3.4.4 Product Approval Process “The Control Plan is not considered a “stand alone” Core Tool. Usually paired with the P-FMEA @]Core Tool inferences in ISO/IATF 16949:2016 Core 180 9001:2015 IATF 16949:2016 Tool (Core Tools NOT Specified) (Core Tool Inferred/Referenced) SPC 9.1 Monitoring, Measurement, 8.3.5.2 Manufacturing Process Design Output Analysis and Evaluation 8.6.4 Verification & Acceptance of Conformity... 9.1 Monitoring, Measurement, Analysis and Evaluation MSA 7.1.5 Monitoring and Measurement 7.1.5 Monitoring and Measuring Resources Resources 7.1.5.1.1 MSA 7.1.5.2.1 Calibration/Verification Records 7.1.8.3 Laboratory Requirements 8.6.3 Appearance Items (inference)APQP Advanced Product Quality Planning CONCEPT INITIATION/ APPROVAL PROGRAM APPROVAL > PROT Type PILOT LAUNCH PRODUCT DESIGN AND DEV, ROCESS DESIGN AND OI et VELOPMENT PRODUCT PROCESS VALIDATION é PRODUCTION 4 st : é 2 FEEDBACK ASSE: Zs o DMAIC Lp @) DFSS ASOAPQP What is it: The management of Product Development Why do we need it: To understand what our customer wants and to fulfill those wants How is it done: Across a prescriptive “Five-Stage”, “Gated” or “Phased” approach. Other iterations exist and are also used so long as the foundational five are in place. The process is required to be cross-functional in its development and execution @] ASOThe Typical APQP Stages/Phases CONCEPT INMATION/ PROGRAM APPROVAL ‘APPROVAL PROTOTYPE Puor LAUNCH Shawn | | PRODUCT DESIGN AND DEV. PROCESS DESIGN AND DEVELOPMENT PRODUCT & PROCESS VALIDATION FEEDBACK. |ENT AND CORRECTIVE ION Planning Product Design & Process Design& Product & Process Feedback, INPUTS. Development Development Validation ‘Assessment & INPUTS NUTS inpurs ‘CAPA INPUTS Va a ay Planning Product Design& Process Design & Product & Process soot: QUTPUTS Development Development Validation Assessment & OUTPUTS OUTPUTS OUTPUTS CAPA OUTPUTSAPQP Plan & Define Phase VOC Data Design goals Marketing Strategy Reliability/Quality Goals Product/Process Preliminary Critical Assumptions Characteristics Customer Inputs Preliminary Process Flow Compliance Criteria Preliminary BOM Etc. Ete. @) ASOAPQP Product Design & Development Phase Program Approval Design Outputs APQP Outputs _DFMEA New Equipment/Tooling [DesignforMfg/Asm New FacilityNeeds Design Verification Gage/Test Requirements | Prototype Built Final Critical Characteristics _Eng Drawings/Specs Etc. Etc. ASOAPQP Product Design & Development Phase Prototype Outputs Pkg Standards/Specs MSA/AAA _Product/Process Review Management Support Process Flow Chart Cp/Cpk Plan Floor Plan Work Instructions PFMEA/DCP Ete.APQP Product & Process Validation Phase ite) em y-Tanl)(-Meleiaeleicy Significant Production Run = Packaging/Preservation MSA/AAA Production Control! Cp/Cpk Studies Quality Sign-Offs PPAP Completion Management Support Product Validation Testing Etc.APQP Feedback, Assessment & CAPA Phase Launch Outputs Reduced Variation _ Improved Customer Satisfaction _ Improved Delivery/Service _Lessons Learned Standard Work Updates Etc.2a Design FMEA Design Failure Mode Effects Analysis wie astonALL Products & Processes Fail Failure is ALWAYS a Design Requirement/Criteria Determining HOW the design will fail, WHEN it will fail, and WHY it will fail will allow a designer to incorporate failure as an acceptable design constraint Failure as an ACCEPTABLE design constraint = Customer Satisfaction = Design Quality @] ASOFMEA: Design (D) & Process (P) What is it: A risk analysis of a part or process Why do we need it: To identify the functions of a process and the associated potential failure modes, effects and potential causes. The vision is to prevent problems from occurring so that defects are not incurred and no one gets hurt. It is used to evaluate if the current planned actions are sufficient and effective How is it done: Via the utilization of a cross-functional team approach. Multiple iterations exist across industry. Within IATF, the process is required to be cross-functional in its development and execution. It is considered a “Risk-Based Thinking” (RBT) tool. It often incorporates results from other methods such as SPC, MSA, Fault Tree Analysis, etc. @] ASOFMEAS for Products & Processes There are three (3) basic cases in which an FMEA is applied: 1. New designs, new technology or new process 2. New application of existing design or process 3. Changes to an existing design or process ¢ Design FMEA: A technique which analyzes system functions within a defined boundary to address possible design weakness and potential risks of failure. DFMEA data is used in the creation of the PFMEA * Process FMEA: A technique which analyzes processes that can impact quality. These processes may be: Receiving, Handling, Manufacturing, Assembly, Storage, @) Transportation, Maintenance, Repair and Communication ASOSix (6) Steps of an FMEA (D or P) pains ‘System Function Failure Risk Optimiza Analysis Analysis Analysis Analysis -tion 1. Define Scope. Identify what is to included in the evaluation. (System, Sub-system, Component). Include relevant Lessons Learned (LL) and reference materials. Manage the five (5) T’s: 1.Team: Who will constitute the core team 2.Timing: When is it due. Gantt, lay-out timing plan 3.inTent: Why is the team there; Ensure skills/training 4. Tool: What reporting methodology will be used? Excel, Software, etc 5. Task: What work needs to be done across the six steps. ®) Consider inclusion of effective documentation for neal auditing and customer reviewoe Function Failure Optimiza ae Analysis Analysis sera -tion 2. Conduct System Analysis: Define the customer(s) wrt End Users, Assembly, Manufacturing, etc. 1. Identify and break down the design into system, sub-system, component and parts for functional risk analysis. Note: A component FMEA is a subset of a system FMEA. Ex. A brake pad is a component of a brake assembly which is a sub-system of the chassis 2. Visualize the system via block (boundary) and/or structure tree diagrams Ii il Chassis pa ‘ie sembly ‘rae asseriy Brake Pad ‘Axle Asm Brake Pat eae Pat Wheel Asm ——— Structure @ ‘te sey fain asseriy @) sock = = ASODefine ‘System > Function Failure Risk > Optimiza Scope Analysis Analysis Analysis Analysis -tion 3. Conduct Function Analysis: Insures that the specified and required functions are appropriately allocated to the system elements. A function describes WHAT the itern/ system element is intended to do. 1.Associates functions with the pertinent system elements 2. Overviews the functionality of the product 3.May describe functions in detail. May need to consider interfaces and clearances wrt physical connections, material exchange, energy transfer and data exchange 4.Allocates requirements/characteristics to individual functions e 5. Cascades internal/external customer functions with @) associated requirements for intended use ASOstn Function pales Risk Optimiza re Analysis Analysis ction 4. Conduct Failure Analysis: Identify failure causes, modes, and effects, and show their relationships to enable risk assessment. ~a Failure effects are the consequence of a failure mode = 1. Identification of potential failures assigned to functions in structural elements 2. Visualize failure relationships (FMEA spreadsheet) 3. Collaborate between the customer and suppler on effects Consider “Failure Chain’ ese" "=". g™™y war How way approach. AKA the t Failure rif Failure * Failure = Y Golden Circle Ng Sect CFEL AS Moc (FM) A couse FC) @) Marker dried out Cap Fell Off Barrel ID too Small ASODefine System Function Failure Risk Optimiza Scope Analysis Analysis Analysis Analysis ction 5. Conduct Risk Analysis. Prioritize the risks by evaluating Severity (how bad), Occurrence (how often) and Detection (how well can we find it). Aka SOD. Each is on a scale of 1-10. The multiplication of S x O x D is the RPN 1. ARisk Priority Number (RPN) is determined 2. Based on the RPN, assign preventive controls which provide information/guidance as an input to the design 3. Assign detective controls to verify and validate procedures previously demonstrated to detect the failure 4. Completed SOD assessment 5. Collaboration between customer and supplier on SeverityRPN, Criticality or Prioritization Each method of evaluation has pros and cons. There is a change in process towards an “Action Prioritization” (AP) matrix which may incorporate Criticality (S*O), RPN will be eliminated as a method of risk evaluation (AIAG, 2018) AIAG currently references the SOD tables found in the FMEA “Blue Book”. Many organizations have evolved to their own form of prioritization tables 3, 5 6 05 based on their own logic - —4th Ed SOD Summary for Design FMEA NOTE: OEs & Other businesses often use their own SOD tables. This is a MODEL Severity Criteria Occurrence Criteria ‘Opportunity for Detection Taare to ect sae anor repli sequement ote tare made et te vais opeation naar eatvesnoe-cmplana with goverment tegulnion wet airing Waryiigh ew techni new dig wih a tony = er 30 Tio seein appaianay Nocatee COI Cat Bee ortinat anahze. Detection almost imposstie 9 | Faniare rarer safey anor regasnaryrequvemens Potent fare mode aflet sate wee operation aor ives non-compliance wih goverment ‘egulnion wth warning Tigh Fare aineviable wah raw deaen pew aren hanes ey rape Tia te deta Hay Hage. Och anaiyin/ oration cana have aoa detetion capably, Vital ana net conrelned tnerpected actual operating conditions. Datecion wey remate 3 | tess ordegradation of primary function tossot mary fon Tigh Tatars ay van ew oa Ao spplesan or changin ty eyeefopertg tones Jin 0 Test design rete and prio telaunth Pode verteshn/vsin ser Sgn Weer ae pas aunch with Saunt testing. Detection a rerote 7 | fost oc dearadtion of iran Hien, Bepradiion ‘aereary anes Wi Fare ueceran wih new Gea nee aplaton ox change in ay eyefoper ing ‘encom 10 199 ot design reie and pir to launch Pod sertcaliza/leaton ter deugn free ae poo burch ith testo fale testing Detection veya, 6 | toss or deqraasion of sacontary aneon Lasso secondary uncon Moderate. Frequent ives asadaied wih sini dean or indesign simulation and tecing S08 ost design rese and prior to launch, Prodt vsctiveadation afer design feere ad ror lurch with Sepradstion tering Detection tow 5 | toss ordegraaation f econdar function eqradaian of secondary function 7 | sreayanee.Anserance ee aie nase, vie ‘pert, tem des pot conlormy And raid by most inarah (75%) ‘deraie: Geena lanes scodaied oF sine ceiga or ig smultion and ey ‘deste ap ted afar Sane ae imiar esp rn design simulation aa testing 10,00 rior te deg Fete. Prado vercsuon/aiaaton ar free and prt anon wth pal asin, Detector rr to Seng ree Prodan veV aR vandaton He SEO freee and porte nunc with test fare esting. Otecton rmoderaey rigs 3 | Ansoyance Appearance or aude nie wrhice ‘peri, tem does nt conform an noticed by mg isamen (50%), Ta Ory ited fares coated with Se idenuca evn on design smutabon testing 3 400000 Tir to den rere rodod verfictn/valaation ater Gave freee and porte auncn with degradation tsing.Dstecion Neh 2 | srsoyanee. Appevance ae aie rane, vee ‘operat, tem des nt confor ad roid by ‘Sharinnatng cstomen (25%) Te Wa ved es aS WT aT igentea evan oF desgn smelaion testing 3 30000000 Vaya Faas conaaned gh Ren] ‘Vatu anya correc Deen araly/eecion Contos have ‘org deteon eapany, viru aay gh corn ‘wth actual or epected operating conditions porto design Freee Deteon ral oppseabi kar AAAIGA Taare cans aR made cannot excur because uly prevented ough dees {ouilans Detection soit etait PEMEA 4" Edition. 2008. Chrysler LLC, Fard Mator Campany, General Motors CorperationDefine System Function Failure ortinle Scope Analysis Analysis Analysis re 6, Evaluate for Optimization, The planning and execution of actions to mitigate risk and assess the effectiveness of those actions 1. Identify necessary actions 2. Assign responsibilities and timing 3. Confirmation of effectiveness of the actions taken 4. Continuous improvement of the design Multiple other types of FMEA applications: System, Concept, Environmental/Safety, Machinery, Software, etc. @] ASODFMEA Sample Format DFMEA formats vary widely based on OE criteria and independent company expectations...Even though the AIAG will add ~8-10 more columns to the current standard, the general approach and intent will be the same; mitigate risk through failure analysis -<~ = es O wuar HOw way Faiture (Y Failure ne | Effect (FE] Mode (FM] Ca FC) @) eg ag te ye Marker dried out Cap Fell Off Sarvel ID too Small ASOOther DFMEA Sources... + http://quality-one.com/fmea/design-fmea/ * http:/www.isixsigma.com/dictionary/dfmea/ + http://www.qmii.com/LT- 133%201SO%209001_2015%20Risk%20Based%20Thinking.pdf + http://www.iso.org/iso/home/standards/iso31000.htm (ISO Risk Management) * 86 Minute Video...very detailed http://www.isixsigma.com/tools-templates/design-of-experiments- doe/mark-kiemele-interview/ + AIAG APOP for DFMEA Checklist (2nd ed)2b Process FMEA & CP PFMEA + Control Plan = Dynamic Control waa stonWhat is a DCP @ + ADCP is a blended format of a PFMEA and CP. It leverages the common columns in both tools and enables “linear” thinking across the analysis of an individual process step. It saves time and increases the security of the system « APFMEA defines, identifies, prioritizes, and eliminates known and/or potential process failures from reaching the customer. The goal is to eliminate Failure Modes and reduce their risks + ACP follows the PFMEA steps and provides details on how the “potential issues" are checked for in the process + ADCP is a living document which helps to prevent problems + It saves time and increases process security @] ASOADCP ADCP lists a sequence of tasks used to produce a product or provide a service by combining the PFMEA and CP. It: Me 2. 3. Identifies process related Failure Modes before they occur Determines the Effect & Severity of these failure modes Identifies the Causes and probability of Occurrence of the failure modes 4. Identifies the Controls and their Effectiveness 5. Quantifies the Risks associated with the failure modes 6. t 8 Develops and documents Action Plans to reduce the risks . Identifies the Type & Effectiveness of the Gaging system Determines the necessary Inspection Frequency ASOFMEA & CP in One Format The format is completed cP “Side” Pp - AA linearly from A—AA. This f= come rtooto ensures inclusion ofa gaging [*__ ‘== 205 Pr eae ee system review and eliminates ove [coed sou [ms [ae vm] ceo the need to manage 2 forms sou) occ er | asm [cs] ow | se **Many sites modify ue a ®) the format to fit their own needs ASOA Practice DCP The fit of a marker cap... | } | | 1. Look at the cap and barrel of a writing marker 2. Review the step of assembling the cap onto the barrel 3. Complete relevant lines of the DCP wrt assembly 4. There can be two general failure modes: a. The cap fits with an audible “click” and stays firmly in place. It does NOT easily pull off b. The cap does not stay secure and falls off 5. Each failure mode will have its own “DCP Stream” of information 6. Follow across the format and complete the information 7. Work in teams across the format LS) ASO4 Ed SOD Summary for Process FMEA NOTE: OEs & Other businesses often use their own SOD tables. This is a MODEL # | Severity Criteria (Customer effect) | Occurrence ‘Opportunity for Detection ] 12 | famretoree ftranor enue reqarenons rte | Y= Na a en opp Hoe LES CaM CT TL NTT tare roar oes eee operation nacre | 1 810 Deen wainon mgstaie tare wth pve epian hot wary 2 | Fate tore afta pve equrenens rte | TR Tet aa owe Hy Ha Te Cea, BET tlre moae teste ete eprao ander eestor. | 31820 Daten wveryremae senpanee with gvrnoen eaten with wenn ct dared opomary heen er neallcinte | OR Tati Brecon SOU erOCITETTE Be GENSION po TNE WaT | Sa ere oe 1830 rou ia act, orate meas Detection freee 7 | asec ccqaasoon olpimey ncion Begaionct ney] FR otter rion assur Taare made detection Hatin opener Bou Atneson, Weise speslevvedand eel pers seo ‘ut ate ozo manor ast eroceing teow sate gage Detecon ivveriow © | tose cepasitan fencondiy uncon Vode oparstia ba | NOSE rete aan po waco ire bd eto pn EAB {onemninclcomiont hintons operatic nso ‘Spear ou we arabe pap ern sabon by spear ag we Sirois pang. tat ow 5 | tossecaegaaiton ofaecondir uncon veide opeabe bat | NaS ae ste Secon tare Far vdeo Goo tos Fo yap Comersreteanfor tnciom weeawiesivenofperttmmee | 3192000 ‘up i of ara ggin 0 by tomate corona tal Steet Sou snd ntl operator Gag peromed onset and pe ac. Cee & modest 7 | anos, Appcwence ovale ae ie apart Rem | Mode ‘rote seed pom aceang Fase ade CO DO POSIT Soest eaten ara novsadoy oneoseaman NF Nie1m009 Staaten tat nl ee! roar part nok ao ror Pet ecere Green rosea Me 7 | ajnce heparan or uaa nas, coke apna em | UO Tate atta acre Fer ade deren Faby nana Seats coformrocnd oy any aor PoP) inion | turn sect crear gad wat ck atin ant pve te ocering Bete high 7 | feoonnce aspeorone or ousive noe, vai operabe nem | Cow Ee Tor eto nor problam proenbon Ev cng Raton hoon Seton actrenuntsonedtrasamicingccermsterss) | Hop00%00 | auanate enaros thao cael sr ad pret ropa pombe de T | tect aoa Waviow Fanrei | brtelon rt apieabe evar pevenbon Precis Bevenon HST T ansted ough | fefmacnelnr oeogn Oster pana canmatbemase aoe toer Dea econ _| Detection saint eta ‘PFMEA 4" Edition, 2008. Chrysler LLC, Ford Mator Company, General Motors CorporationFor Want of A Horse Cause and Effect Perspective stachsanh's era. yc PLD epee USAIN th permission 10/6/14DCP or Fire-Fight? Planning vs Fire-Fighting ‘When Planning is Secondary to Fire-Fighting Fighting is Second: | ‘Continuous Many Surprises Fire-Fighting Significantly Fewer » a Surprises: £ g Planning & iB] Sveugh OP: ‘Smoother é 2 Production Minimat a S, Planning Launch Production Planning Launch Production Project Timing Project Timing Total time is area under the curve...Estimated monies are @ 7:1 with OT, Freight, Material/Equipment changes, T&E, etc. Leverage the DCP to minimize fire-fighting after AS@ release. Partner with functional teamsCase Study: Before/After DCP Initial release and after DCP implementation of 3 products. Was planning secondary to firefighting? What kinds of losses were likely incurred? Was it worth it? 100% 90% > June: Before DCP gow, > Sept: After DCP > December: 10% Current Performance 60% +- 50% 40% @] ASO First Pass Yield —ProductA -=-ProductB -—-Product C =3 PPAP Production Part Approval Process eit saePPAP What is it: Requirements for approval of production parts Why do we need it: To make sure that we understand all of the customer requirements, and that we can meet them under actual production conditions How is it done: Based on customer direction, there are 5 levels of PPAP to secure product approval. An application “cover sheet” is called a Product Sample Warrant (PSW) which lists 18-20 different types of evidence that may be required for submission. These can be customer and/or product/process dependent. It is typical for a customer to witness a launch and review PPAP records when on-site @] ASOPPAP Levels per AIAG 4" ed. 4. 2: 3. 4, 5. Warrant with product samples and complete supporting Warrant only for appearance items Warrant with product samples and limited supporting data Warrant with product samples and complete supporting data Warrant with other requirements specified by the customer data reviewing at the supplier's manufacturing location PPAP level details are typically arranged in advance with the supplier and customer and will often depend on whether the product is a new design or another revision of a tried and true process @) ASOPPAP Components Te Design records 2. Authorized Engineering So Ps Po Change documents. Customer engineering approval Design FMEA Process flow diagrams Process FMEA Control Plan MSA Studies Dimensional results results ASO ee 11. Initial process study 12. Qualified lab documentation 13. Appearance approval report 14, Sample production parts 15. Master samples 16. Checking aids 17.Customer specific requirements (CSR) records 18.PSW 19. Bulk material requirements checklist 0. Material/performance test 20.Special process audit resultsPPAP Prep...All Hands on Deck 1. TAKES TIME and attention to DETAIL 2. Requires a cross-functional team 3. Insure a good understanding of the Customer Specific Requirements (CSRs) in advance 4, Do WELL on the Appearance Approval Reports (AARs). While the easiest “up front’, these are often the most expensive later on. Take the time to develop boundary samples and conduct Attribute Agreement Analysis (AAAs) to ensure skill . Attend to the full Measurement System Analysis (MSA) on variables metrics. Include calibration, resolution and GRR 6. Enable sufficient lead time for the DFMEA, FMEA and CP 7. Insure statistical control of significant characteristics 8. Etc. a ASOHow to Organize 1. Many customers will dictate submission formats 2. Some companies establish binders/books 3. Some use formal organizing software It is critical that: 1. More than 1 person has access/passwords 2. Proper security is enabled across those individuals 3. Proper revisions are sustained/maintained @] ASOCpk Cp/Cpk/Pp/Ppk Process Capability PrimerProcess Capability 101 * Cp/Cpk: Also called “short term’ capability which is used to reliably determine if a process is yielding good initial results by taking a representative sample size. »Cp is based on the whole breadth of the process >Cpk is based on “half” of the process * Pp/Ppk: Also known as “long term” process capability. The key difference is that there is much more data on hand for Pp/Ppk. AIAG aso Notes “90 shifts, 90 days” aDissecting the Bell Lower Spec Limit Upper Spec Limit 60 (+/-3.0n each side of the average) When there 34.2% 34.2% is ROOM for 6 sigma’s on EACH side eae of the have “60 average quality!” before the closest ee target is sf-30_99:30% hit... sea 99.90% 4h-50 oases +169 oaga07%Calculating cane ‘Cp (Pp). Measures the ability of the WHOLE bell to fit within the target limits If the whole bell i sigmas) fit within the target limits a total of 1 time, then the Cp = 1. Ideally, 2 is preferred. Cp = (USL-LSL)/ (6x) USL =6, LSL=0,c=1 ° 3 6 0123 4 5 678 pi Fen Measures the ability of FAL of a bell (3 sigmas) to fit within es average and the closest target jimit Cpky = (USL — Average) / (3 x c) Cpk, = (Average —LSL) / (3 x o) USL=6,LSL=0,0=1 Cp =(6-0)/(6xa)=1 Cpky = (6 — 5) / (3 xc) = 1/3 (0.33) Cpk, = (5 — 0) /(3 xc) = 1 2/3 (1.67)Cpk Worksheet Determine the Cp and Cpk for each situation...Remember, if the process is NOT shaped like a bell, then sigma cannot be used (without special consideration) and the Cp/Cpk cannot be properly determined In each case either the %Non-Conf average or sigma may tg na or may not orange. : 1] 5.0 | 2.50 only the specifications remain the same 2| 5.0 | 1.67 3 | 7.5 | 0.83 [4 | 5.0 [ossShift Happens Cpk of 2 is desired for initial capability Before: Cpk Long term capability is Ppk. This is the capability y | after the process experiences “life” via multiple f \ material lot changes, set up and operator \ variation, seasonality, etc. Ppk is usually als l calculated after “90 days” (or with a significant Cpk = 2.0 quantity) of process data. It is the type of product results that the /ong term process will represent It is estimated that a process will “shift” by +/- Ppk 1.5c in response to those changes. As such, if a process started ideally with a Cpk of 2.00, then it is estimated that the resultant Ppk would be 1.33 to accommodate these types of affects @] Ppk ~ 1.33 ASO4 MSA (GRR & AAA) Measurement Systems Analysis Accurate & ®) PreciseMeasurement System Analysis When we measure or make an assessment of the goodness of an item, we need to be sure that our result is correct. If it is not correct, we take two risks: > Alpha o Risk: We may inadvertently discard or rework a good item (Aw, darn) > Beta B Risk: We may inadvertently pass on a bad item (Boy, that was Bad) @] ASOWhy Do We Need to Know? We need to know how much error there is in our measurement processes for sevetal | reasons: + Prevent a and B errors + Reduce scrap/rework + Understand what process Cp/Cpk we need our processes to have + It is our JOB to ensure that our people are enabled to make the right pass/fail 5 decision EVERY time « And of course...it is an inherent part of PPAP * NOTE: EVERY item called out for measure or inspection on a control plan is REQUIRED to have an MSA analysis @) conducted. ASOMSA Types: Variable & Attribute ™ Humans usually believe what they see and do not question a value shown on an instrument. There are two typical types of variables MSA used to determine the percentage of results error: «Crossed Gage R&R (Repeatability & Reproducibility): One instrument, multiple operators and multiple part samples «Nested GR&R. Used for gage error in destructive testing There is generally one type of Attribute MSA to determine HOW right or wrong we are in our results: + Attributes Agreement Analysis (AAA) is used for items we assess visually or by go/no go or needs . to be categorized Is this window broken? It still @) opens. The wooden frame is ASO in placeHow Data Varies ©OO©© Accurate & Inaccurate Accurate but Inaccurate & Precise but Precise Imprecise Imprecise Accuracy: Generally managed by calibration includes bias (how far off), linearity (across the breadth of the measured range) and stability (holding a measure over time) Precision: Generally managed by Repeatability (gage) and Reproducibility (human) aka GR&R ASO aGeneral MSA Notes For a variables Measurement System to work, three features are equally needed: > Resolution: Ability to read the gage. (Discrimination). Resolution needs to be at least 10% of the tolerance (If not at 10% or better, additional actions are needed) >» Calibration: A check of bias, linearity and stability (performed on a regular basis) » GR&R: Amount of error in human and gage performance. Typical GR&R <= 10% error on safety features. Included in PPAP, it insures that the gage system will work as intended BEFORE the process is launched. After that, it is conducted @) on an as needed basis (verification of process, gage < system change, qualification of personnel)Resolution and Cpk What does Resolution do for you? With a “10% resolution gage", we would accept a unit that reads 10. But...it could be a 9 or an 11. We are at risk 1/3 of the time for a B error...IF the Cp/Cpk is 1 -10 0 +10 We would also reject an 11, (it XXX could be a 10 or 12). We could -10 0 +10 have an a. error 1/3 of the OD time...Again, IF the Cp/Cpk is 1 -10 0 +10 This is one of several reasons why a Cp/Cpk of 1 isn’t good enough for ASO safety featuresResolution With Cpk >1.33 Resolution with better process capability With a more capable process, if we still have a “10% gage”, the process is not likely to generate any units measuring a “10". As such, if we read an 8, it could still be a7 or 9. However, there is -10 0 +10 now minimal risk for either an o or XX _ B error, In this case, the Cp/Cpk #10 is 1.33 This is one of several reasons of why a minimum Cp/Cpk of 1.33 is @) required for safety featuresAttribute Agreement Analysis AAA Checks for the chances of 100% agreement on three features: > Within “myself”; Did | repeatedly call it good or bad in a consistent manner (even if | was wrong) » Between both me and “my peer”; Did both my peer and | repeatedly call it good or bad in a consistent manner (even if we were both wrong) > Compared to “Standard”; Did I/we get it rightAAA Quick Notes An AAA needs many Pass/Fail “Samples”; Preferably 50 or more (pass/fail/borderline). NOTE: One unit might have several samples on it An AAA is a check for accuracy in human perform- ance. The target for “Statistical Agreement” is >= 85%. Another form of Agreement is called Kappa (K). AIAG calls out for K >= 75%. AAA is done as a part of PPAP to ensure that the review process will work as intended; before the process is launched. It should be treated as a “maintenance” action with regular review to keep human assessors “calibrated”. Usually ASO quarterlyAAA: What It Looks Like AAA Gives a series of graphs to show how the operators perform in general. While 100% agreement is not feasible, (like 0% GRR Error), industry norm is 85% for Statistical Agreement Screen % Effective Score vs Attribute’ ‘Total Inspected #in Agreement 96% UCL Calculated Score |}95% Let Not an effective Statistical Agreement at < 85% This team will be in statistical agreement about 68% of the time. However, 95% of the time, they will likely range from 47% in agreement to 85% ASO i wee UCL ‘% Score vs Appraiser pa mp LCL t + + | . ‘Gemeiency 110.0% 100.0% 90.0% 80.0% 70.0% 60.0% 50.0% 40.0% 20.0% 20.0% 10.0% 0.0% + { Operator#1 Operator? Operator #3,5 SPC Statistical Process Control X-bar Chart SampleWhat's Normal? There are 6 main causes of Normal Variation for almost any type of process... This is NORMAL. Hence the “normal” or Gaussian distribution. Manpower i= Ma chine Ma terial Me thod Syke Me asurement af ay Environment Me @] ASOSPC; High Level Guidelines ao Z 5. SPC applies to both variables and attributes. It is a graph- based statistical method to analyze and control a process First step is to insure MSA effectiveness; whether for variables (GRR) or attributes (AAA) . For variables, must insure that the process is capable FIRST, prior to establishing a control chart (Cpk >= 1.33) . Determine any key patterns (common sense control) that are meaningful to your process and train to those conditions. These typically include: Shifts, Trends, Points outside of the limits After that, it’s a go/no go chart. The graphs help you to know when the processes change (whether desired or not) @) ASOAfter GRR & Cpk; Now We Can Chart Moving X and Range chart plots data across time along with its corresponding ranges. Patterns are reviewed for prevention purposes. Most Common Signals: « 5 or more points above or below the average line is considered a shift (bell has moved) +5 ormore points continuously _ increasing or decreasing is considered a trend + Any point outside of the control limits. These are considered non-normal Resene and the process spread has. ikely increased NOTE: Different references call out varying control criteria ASO 2-bar Chart Sample 12 ee Se Teo Ie I X-bar and R charts are PREVENTIVE and PREDICTIVE forms of process management. They give an advanced warning enabling proactive actionsPre-Control: No “Limits” SPC is powerful and effective. Pre-Control is a step before that. It “forces” a 1.33 Cpk by requiring the process to “pre-act” when data signals are in-spec but outside of the +/- 3 sigma range. While no control limits need to be calculated, careful communication of WHY a person needs to react and adjust the process for an in-spec part USL Red zone. Stop/Adjust process. Take next set of readings. Recover process back to Green/Yellow zones, Green zone. Run as long as results are within +/- 2 sigma of the nominal Red zone. Stop/Adjust process. Take next set of readings. Recover process back to Green/Yellow zonesAttribute Charts; With a Good AAA p-chart. A trend-based percentage chart. Must be paired with a Pareto or checksheet to execute fixes. A p-chart Spice ly follows a Weibull distribution because either 0 or 100 is optimal and a “half bell” is developed with bias towards one end or the other. c-chart. This “counts” defects per unit. Ex. A application may have 3 typos, 2 smudges and 2 areas not filled out for 8 defects on 1 item. The next one may be perfect. The c would equal 4 defects per unit. This is a highly effective method that captures detailed data. It is powerful when paired with a Pareto. Again, checksheets are often used. There is usually a high cost to capture this data. c-charts are usually “turned on/off’ to capture a timeframe of data and then rechecked later to verify the effectiveness of the fixes Trends: « 5 or more points above or below the average line is considered a shift + 5 or more points continuously increasing or decreasing is considered a trend + Any paint outside of the control limits. Spread has likely increased p and c Charts describe what happens AFTER the process has occurred. (identifying either Scrap/rework). Losses are incurred. The intent of these charts is to see if the corrective actions are workingCommon Types of SPC Charts Chart Type Primary Usage What is Charted Typical Sample Size X-Bar & R_— Routine monitoring of high _—Plots the average of ~3to6 volume manufacturing the data set and its processes range Individual & Used when only sample is Plots the value and One Moving possible. Common for the moving range of Range transactional (monthly) the current and (IMR) processes preceding values. p-Chart Routine monitoring of high Plots the percent Variable volume processes where non-conforming scrap/rework trends are critical c-Chart Used for deeply analyzing Plots the average Variable non-conformities ina number of non- product conformities in a single unitWhere The Alphabets Fit... APQP 5 Stages (or more)The FIVE Core Tools 1. APQP: Advance Product Quality Planning: Guidelines for a product quality plan to develop a product or service that satisfies the customer 2. FMEA: Failure Modes and Effect Analysis: Methodology used to ensure potential problems have been considered and addressed throughout the product and process development process (Ex. APQP). Traditionally includes the Control Plan (CP) 3. PPAP: Production Part Approval Process: Ensures product consistently meets customer engineering specification requirements during production run at the quoted production rate 4. MSA: Measurement Systems Analysis: Guidelines for assessing the quality of a measurement system where readings are replicated 5. SPC: Statistical Process Control: Basic graphing statistical tools that enable process control and capability for continual @) improvement ASOASQ The Global Voice of Quality’ Questions? Jd Marhevko Phone: (419) 704-5603 Email: JdMarhevko@AccurideCorp.com