1 AUTOMOTIVE CORE TOOLS; SPC, MSA, FMEA, APQP/CONTROL PLAN CQI Wessex - 11 December 2012John Skinnerrdaconsultancy.com
2 AUTOMOTIVE PROGRESS SINCE 1990 Best selling car in the UK in 1990 –Ford Fiesta – sold in EU1.4, 1.6 CVH engines were very harsh and had early failures (valve guide wear, dropping valve seats damaging the engine). Rust was still a problem, though better than the 1970’s & 80’s.Radio-cassette, some speaker adjustments, maybe electric windows. No a/c and definitely no climate control. No airbags, little in the way of crumple zones. Central locking on higher end models. Door mirrors had ‘remote adjustment’ on top models. Security was very poor – cars like the XR2 were being stolen by joyriders on a regular basis (epidemic levels in some areas). Car insurance costs on ‘hot’ models went through the roof. No ABS, stability control etc.Cost – around £8500 – roughly equivalent to £15000 in today’s prices
3 WHAT CHANGED?Improvements demanded by legislation (emissions, safety), safety (NCAP), competition (i.e. VW/Audi), consumer demand (phone compatibility, convenience features etc.)A thorough focus on improved product quality achieved through; computer modelling/simulation, enhanced testing techniques (bench and road), consumer feedback (JD Power), supplier capability, effective corrective action (8D)Increased use of electronics (particularly powertrain)Need to improve product reliability (reducing warranty/ product recall costs)Application of ISO/TS16949 and mandated use of Automotive Core Tools ?
4 ISO/TS16949Initiated as QS9000 in 1995, based on the requirements of ISO9001, but adds many automotive industry specific requirements (though only one additional documented procedure from ISO9001)Produced and controlled by the International Automotive Task Force (IATF)Members include; Ford, GM, VW/Audi, PSA, BMW, Chrysler, Renault, Daimler and major trade organisations across the globeRecognised and required as a prerequisite for becoming an ‘approved supplier’ to the respective automotive customers
5 AUTOMOTIVE REQUIREMENTS APQP; Advanced Product Quality PlanningFMEA; Failure Mode & Effect AnalysisControl PlanSPC; Statistical Process ControlMSA; Measurement System AnalysisThese require a team approach (cross functional teams with management commitment)Automotive Industry Action Group (AIAG) manuals define basic requirements for applicationThere are other requirements, based on the customer
6 Advanced Product Quality Planning (APQP) ISO9001 requires Planning of Product Realisation;The Automotive industry goes further;‘Some customers refer to project management or APQP as a means to achieve product realization. APQP embodies the concepts of error prevention and continual improvement as contrasted with error detection and is based on a multidisciplinary approach’.(extract from ISO/TS)
7 Advanced Product Quality Planning Why plan?What gets in the way of planning?
9 APQP – Project Scope2011 model year Range Rover – minor changes
10 APQP – Project Scope 2012/13 model year Range Rover - major project. New design, technologies, material – aluminium, using self piercing rivets and aerospace sourced epoxy adhesive
11 Advanced Product Quality Planning (APQP) Advantages;Thorough planning and improved decision makingShorter development timescalesSimultaneous engineering (design & manufacturing)Early procurement of long lead time tooling/facilitiesDefined objectives, measured as project stages are achievedDefined project gateways, with key deliverablesImproved use of resourcesSignificant cost savings (known impact on company finances)Effective feedback & corrective action (enhanced with use of computer systems)
12 Failure Mode & Effect Analysis (FMEA) Typically used at design (DFMEA) and manufacturing process planning (PFMEA) stagesFMEA - a systematic set of activities intended to:Recognise and evaluate the potential failure of a product/process and the consequential effects of failure (risk management)Identify actions that could eliminate or reduce the chance of the potential failure occurring (improvement)Document the entire processNeeds a ‘team approach’ to be successful
13 DFMEA DFMEA System FMEA Number: 001 FMEA Team: xxxxxxxx Subsystem SystemFMEA Number:001FMEA Team:xxxxxxxxSubsystemFMEA Date: (Original)xxxxxxxxxxTeam Leader:XComponentxxxxxxx(Revised)Action ResultsItem and Function/ RequirementsPotential Failure ModePotential Effect(s) of FailureSeverityClassPotential Cause(s)/ Mechanism(s) of FailureOccurrenceCurrent Controls - PreventionCurrent Controls -DetectionDetectionRPNRecommended ActionResponsibility and Target Completion DateAction Taken95225381
14 DFMEA DFMEA System FMEA Team: xxxxxxxx Subsystem Team Leader: X SystemFMEA Team:xxxxxxxxSubsystemTeam Leader:XComponentWheel hubItem and Function/ RequirementsPotential Failure ModePotential Effect(s) of FailureSeverityClassPotential Cause(s)/ Mechanism(s) of FailureOccurrenceCurrent Controls - PreventionCurrent Controls -DetectionDetectionRPNWheel bolts; Support wheel & tyre Transmit loads to stub axle/vehicle suspensionBolts fail after wheel impact with kerbLoss of wheel,loss of control of vehicle, possible fatalities, resulting legal action10CCIncorrect material grade (new hub nut design)7Design calculationsPre design freeze test to failure vaidation5350
15 Responsibility and Target Completion Date DFMEAFMEA Number:001FMEA Date: (Original)xxxxxxxxxx(Revised)Action ResultsRecommended ActionResponsibility and Target Completion DateAction TakenSeverityOccurrenceDetectionRPNConduct a full virtual analysis using design FEA & simulation software (proven performance)PTO engineering; Q1 2013Virtual anlysis completed; results acceptable to spec. xxxx1072140
16 DFMEAIgnores manufacturing issues; i.e. manufacturing producing/using parts that are to specificationCan direct design effort to critical/significant characteristics and improve design validation/ verification testing results, avoiding late design changesIdentifies special characteristics that need to be controlled in manufacturing to assure product qualityProvides a documented record of the analysis which can be used into the future (many vehicle recalls could have been prevented by effective DFMEA)Needs to be maintained as a live document; continual improvement
17 PFMEAA PFMEA will follow the stages defined for the manufacturing route from material receipt, through the manufacturing stages to despatchTypically the manufacturing route will be defined on a process flow diagramme, including locations, machines, operation sequences etc.
18 PFMEA PFMEA System FMEA Team: xxxxxxxx Subsystem Team Leader: X SystemFMEA Team:xxxxxxxxSubsystemTeam Leader:XComponentWheel hub/new designItem and Function/ RequirementsPotential Failure ModePotential Effect(s) of FailureSeverityClassPotential Cause(s)/ Mechanism(s) of FailureOccurrenceCurrent Controls - PreventionCurrent Controls -DetectionDetectionRPNOP 70; Finish machine bearing bores (CNC)/ dimensions to specificationBearing Ø oversizeBearing wear (with warning), loss of wheel, loss of control of vehicle, possible fatalities, resulting legal action. 100% of product may be scrapped9CCIncorrect tooling set up5Pre-set tooling usedPost operation gauging - 10%6270
19 Responsibility and Target Completion Date PFMEAFMEA Number:001FMEA Date: (Original)xxxxxxxxxx(Revised)Action ResultsRecommended ActionResponsibility and Target Completion DateAction TakenSeverityOccurrenceDetectionRPNInstall in-station gauging as part of machine upgradePTO engineering; Q1 2013in-station gauging integrated into machine; auto lockout on detection953135
20 PFMEAFocuses on potential for non-conforming product (in use and impact on manufacturing process including employee safety) and mistake proofing techniquesIdentification/prioritisation of potential failure modes and implementation of preventive/corrective actionFocus on special characteristicsContinual improvementAssumes product as designed will meet intentShould be extended to other areas; receiving, storage, transport, despatch etc. (complete process)Provide feedback to design (mistake proofing features etc.)
21 Control PlanA documented description of the systems and processes required for controlling productThis is a key output once the DFMEA and PFMEA analysis has been completedApplies to distinct stages; Prototype, Pre-launch and Production.Each part must have a control plan, but family control plans can be used where justified (e.g. a foundry producing many different castings)
22 Control Plan Control Plan N/A Production Control Plan Part #:N/APrepared By:xxxxPart Name/Description:Wheel Hub / new designCore Team:xxx, xxx, aaa, bbb, cccNotes:Latest Change Level:#REF!Supplier/Plant App./Date:Production Control PlanVendor ID:Other Approval/Date:Supplier Plant/Code:Oper. #Process DescriptionMachine/Device/Jig/ToolsCharacteristicsCCItem#ProductProcessSCProcess/Pdt. Spec./ToleranceG.ID70OP 70; Finish machine bearing bores (CNC)XXX CNC machining centreBearing Inner Bore diameterFinish machine; 25mm/min, CBN tip (XYZ)mm mm run out maxIn cycle gauging.
23 Control Plan xxxxx In cycle gauging. 100% Machine control datalogger Date (Original):xxxxxCustomer Engg. Appr./Date (Opt):Cust. Quality Appr./Date (Opt):Other Approval/Date (Optional):Control Plan No.:CP001MethodsEval. Mst. TechniqueSample SizeSample FrequencyControl MethodReaction PlanIn cycle gauging.100%Machine control dataloggerLock out; setter informed, machine checked and re-set. Suspect material quarantined; (Corrective Action Report - CAR)
24 Control PlanIdentifies the controls required to ensure product quality with a focus on special characteristicsDefines the reaction plans required to be implemented where non-conformance is identified (containment of product, 100% inspection to ensure process becomes stable and capable)Is an output from the FMEA process
25 Statistical Process Control Traditional inspection techniques (patrol inspection, batch sampling etc.) rely on detection, which is wasteful as time and resources are put into producing parts that are not always useablePrevention verses detection; i.e. not producing the non-conforming parts in the first place is an obvious preferred situationIf we can predict the process output, then we may be able to ensure conforming partsStatistical techniques can be used for process control
26 Process Control System Model with Feedback Statistical ProcessControl (SPC)PeopleEquipmentMaterialsMethodsEnvironmentManufacturingProcessProductCustomersIdentifyingchangingexpectations;direct feedbackon product qualityInputsFeedback from Customers
27 Goals of SPC To achieve a state of statistical control (stability) To maintain a state of statistical control (stability over time; prevention verses detection)To improve process capability
28 Statistical Process Control Taking action on the result of the output of a process (traditional inspection techniques) permits waste – rejected/reworked product, wasted resources, potential for rejects to ‘escape’ the processUnderstanding the variation of the process and applying statistical techniques allows for predictable process output (capability)To do this, we need to understand the types of variation present in the process
29 Statistical Process Control Machining process; part dimensions vary from each otherSizeIf the output is stable (only common cause variation), the results form a pattern that can be described as distributionSize
30 Statistical Process Control Common cause variation; inherent in the process – backlash/clearances, coolant feed, machine tolerancesSpecial cause variation (or assignable cause) –machine set up, material change, environment
31 Special Cause Variation Target linePredictable ?TimeSizeIf special causes of variation are present, the process output is not stable over time and is not generally predictable; we cannot always use sampling and SPC to control the process.
32 Common Cause Variation SizeTarget lineTimeIf only common causes of variation are present, the output of a process forms a distribution that is stable over time and is predictable; we can then monitor the process using sampling and SPC charts
33 Statistical Process Control _X – Range Control chart (AIAG SPC manual)
34 General Rules for Interpretation A point outside a control limit7 points in a row on one side of the centreline6 points in a row steadily increasing or decreasing2 out of 3 points more than 2 standard deviations from the centreline (same side)4 out of 5 points less than one standard deviation from the centreline (same side)
35 Process CapabilityCapability indices are able to summarise process performance as a number to reflect how well the process will meet customer requirements (specification).They will indicate;- How variable the process is (i.e. spread of results)- Where the process output is in relation tothe specification limits.
36 Process variation or spread Process CapabilityLower specificationUpper specificationProcess variation or spreadTolerance
37 Capability Indices Capability Index Cp = USL - LSL Process capabilities are an index produced by comparing the observed process variation or spread against the required tolerance. Examples include;Capability Index Cp = USL - LSL6 δ (process variation)_ _Capability Index Cpk = USL – x OR x - LSL(largest of) δ δi.e. Worst case resultCpk will indicate the position of the process relative to the specification (i.e. centering)
38 Capability Indices Cpk Approximate proportion out of spec If the process data has a normal distribution, the following can be used to interpret Cpk:Cpk Approximate proportion out of spec/1,000,000/1,000,000/1,000,000(Where the process is centred between the specification limits)
39 Attribute ChartingAttribute data is the result of inspection or testing that produce a fixed result and cannot be measured using measurement equipment e.g. number of paint defects on a door panel, number of rejected units from a functional test batch, number of weld failures on a floor-pan assemblyAttributes can also be monitored using control charts with control limits to determine long term stability
40 What are the Benefits of SPC? Properly used, control charts and SPC can:Distinguish special from the common causes of variation, as an aid to improvement in capabilityEnable the process perform consistently and predictablyProvide a common language for discussing the performance of the process (capability indices)Be used by employees for on-going control of a process
41 Measurement System Analysis (MSA) As measurement data is often used to make decisions with regard to manufacturing (and test) activities then the ‘quality’ of this data needs to be assuredNo measurement system is ‘perfect’ (i.e. measures exactly with reference to known standards each time); some variation will be evident in all systems, including human influencesA series of analytical techniques can be used to ensure that the inherent variation in measurement systems can be determined and the effects understood i.e. possibility for accepting ‘bad’ parts and rejecting ‘good’ partsIn essence, we need to understand the variation and limitations of the measurement systems we are using to enable confidence in those results (for equipment on the control plan)
43 Variable Gauge R&R study Gauge repeatability and reproducibility data sheet used for the numerical analysis of the study data (MSA software is also available).10 parts used to represent variability of the process; 3 appraisers.Typically 3 trials per appraiserComplete the study using usual equipment and individuals (random presentation of parts to avoid influence on the results)
45 Variable Gauge R&R study The study estimates the variation and percentage of process variation for the measurement system (Gauge R&R) and its components- Repeatability- Reproducibility- Part to part variation (how representative the parts are; this will influence the results).
46 Gauge R&R Acceptable Gauge R&R% results; Up to 10% - generally acceptable10 to 30% - may be acceptable based on importance of measurement feature, cost of better equipment, cost of refurbishment/repair of equipment, OR the skill level of the appraisersOver 30% - would be considered un-acceptable and improvement is requiredWhere results are not acceptable, check the data, calculations etc. to determine if there are any errors
47 Use of Graphical Techniques There are many different methods for presenting data for analysis‘A picture paints a thousand words’.The AIAG MSA manual strongly recommends the use of both calculated methods and graphical techniques.Other methods are available for specific situations and where more detail is required – Analysis of Variance (ANOVA), gauge performance curves, regression analysis etc.
48 Attribute StudiesGo/No-go gauging systems (acceptance gauging, gap gauges etc.)A series of comparative techniques are available to evaluate the effectiveness of attribute gauging systems, again using several appraisers and trials
49 Product Part Approval Process (PPAP) The intent of PPAP is to validate that products made from production tools/processes meet engineering requirements (specifications), that the processes are capable (Cpk, SPC data) and are capable of producing acceptable product consistently over time.The type and format of submission will depend on the customer requirements, but the AIAG PPAP manual has defined formats
50 PPAPTo support this, appropriate data has to be completed and be made available for review by the customer;Part Submission Warrant (PSW), DFMEA, PFMEA, Control Plan, MSA data, initial process capability results, material certification, marked up drawing/ dimensional results, engineering change documents, material/performance test results, qualified laboratory documentation, appearance approval report (if applicable), sample productApproval may require full or partial submission of the information (invariably depending on supplier approval status) reviewed at the customer location or supplier site
51 PPAP Acceptance of the PPAP detail is key for several possible events; Authority to ship production partsMeeting customer timing requirements for the projectSupport the supplier approval rating (i.e. PSW submission on time)Trigger tooling payments to the supplier (tooling will typically be owned by the customer)
52 AUTOMOTIVE PROGRESS SINCE 1990 Best selling car in the UK 2012 (to date) - Ford Fiestaex. Zetec £12495 OTR price – 3 year warranty, plus breakdown cover – sold in EU & USAAirbags x2, side airbags, knee airbag, anti-whiplash head restraints, optional curtain airbagsABS, brake force distribution, stability control, traction assist, auto brake assist, auto hazard warning lightsHigh strength steels, active seat belts, anti-submarine seats, Halogen projector head lightsImmobiliser/alarm, deadlock doorsHeated, electric remote door mirrors, auto wipers with rain sensor, alloy wheelsRadio/CD,, AUX connection for phones, IPODs, A/C etc.4 way adjustable front seatsElectronic power assist steering, advanced engines up to 85 mpg possible with some models
53 SummaryIn truth, many factors have influenced the improvement in product quality over the past 20 odd years, but the use of APQP and core tools have provided a focus on improved process capability and product quality/ reliability throughout the automotive supply chain.The aerospace industry are now utilizing many of these tools to improve supplier performance