User`s Manual
Contents
1. i Start Cpk medium speed or 100 values Analysis Analysis Reset delete A EN Jo Data show Stop simulation Data Grid Start Control oan pog Memorize Chart Analysis Analysis Current Data A typical usage scenario would be similar to the following Open a Simulator otart the simulator to generate some data Export the data to Excel for analysis Analyze the data using a statistical program such as DOE Pro or SPC XL Use the results to change the inputs if applicable to optimize the outputs So INO Iure Open a Simulator Primary Method oimWare ships with over 25 quality simulators built in To open a built in simulator select Simulator from the main menu There are five groups of simulators arranged by what concepts the simulators are designed to teach Basic Statistics Simulations designed to teach basic concepts such as t Test F Test Summary Stats etc Measurement System Analysis MSA Simulations that allow you to analyze a measurement system Design of Experiments DOE with one output Simulators that allow you to perform and analyze a DOE with a single output Design of Experiments DOE with multiple outputs Simulators that allow you to perform and analyze a DOE with multiple outputs Design for Six Sigma DFSS Simulators designed to demonstrate DFSS concepts such as parameter design and tolerance allocation WebService You can also download simulators from the SimWare Pro WebServ2. 1100 300 Reagent Conc 160 pH 50 Reagent Conc 435 Reagent Conc pH 75 Incubation Time The standard deviation decreases with Substrate Type 2 Broadcast News Audience Size 60 20 Edit Time 10 Num Cameras The standard deviation decreases with the Standard Plus edit method i e Edit Method Standard Plus Vial Filling Volume 3542 4 Active Press 1 2 Active Press 5 4 Inert Press 2 4 Inert Press 3 2 Flow Ratio 1 3 Nozzle Diameter 5 2 Nozzle Diameter 0 8 Active Press Inert Press 4 5 Active Press Nozzle Diameter The standard deviation decreases when Inert Pressure decreases Note Inert Pressure also shifts the mean DOE Multiple Output Simulations Restaurant Minutes till Order 6 1 4 Num Cash Registers 8 Tables Per Waiter 4 Num Cash Registers Tables Per Waiter Minutes till Drinks Minutes till Order 6 1 2 Tables Per Waiter 8 Music Type 7 Tables Per Waiter Music Type Minutes till Entree Minutes _till_Drinks 7 2 9 NumChefs The standard deviation is not affected by the model Space Shuttle ME Left Engine 300 58 Core Temp 68 Nozzle Temp 35 Chamber Press 75 Left Temp 44 NozzleTemp 24 CoreTemp 31 CoreTemp NozzleTemp 26 Chamber Press Left Temp The Left Engine standard deviation decreases with an increase in Neutrinos Right Engine 300 45 CoreTemp 58 NozzleTemp 55 Chamber Press 65 Right Temp 34 Nozzle Temp 34 Core Temp 41 Core Temp Nozzle Temp 36 Chamber Press Left Temp The Right Engine standard deviatio
3. 31 Control Guar SHOWING NOW X2 id 32 Summary Stats SHOW me NO Wisin 33 Box Plot Show mebowl tette teet tete 34 Measurement System Analysis ssa 36 Measurement System Analysis Simulators sennssenenennnsnnnnrrnesnrrsrrrerrresrrnnn 37 Measurement ENOL IVA pta eran on eut 37 COMPressor EE 37 Ville ge ln ET EE 38 Mn eee ee eee ee 38 Visual Clarity Advanced MS Aleida 38 SHOW Me HOWIMSA liinda 39 Uo AMSHOW TRO NOW o 40 DOE SINGS e Ee asus 42 DOE Single Output SIMULANS E 43 Aine Simulatom DOE auae toate ue diti etin oe P EE RE REOS INE decocti Os 43 Commercial Profit DOB EEN 43 Space SMP ThrBSE DOE EE 43 Temperature Control DOE EE 44 Hydraulic arcus e a 44 BOOG Anal SIS DOE sarral E 44 Broadcast News DOE 44 Viak FIDON el kourin 44 DOE Multiple Dutra 45 DOE Multiple Outputs Simulators i cssiiiiiiriniiiiirininininininininininiinininiin naian 45 Restaurant WOE Ien 45 Space Shuttle Main Engine DOE scsi 45 VENICIGIMB ACT alesana ideados 45 DFSS Design for Six Sigma sss 45 DESS SIMA clio S OTT 46 RESSO DE SO iiie de ii o d ii i im DE MM 46 S UD DE S etd ducet m pua uti eae meu mus p teu ee 47 Nuclear Reservoir DE SS nestle iii 47 Spark PIUG DESIGN DESO assente etdett etes a eee 48 Washing Machine dB gere ce m 49 Washing Machine Advanced DEG 50 MEMORIZE DATA consista 51 STATISTICAL ANA E SIS iiri ierit ain eit enti aii 51 Using SimWare SimWar
4. are resistors which are located in parallel The equation for impedance for this circuit is R4 R2 R4 Re For proper operation you need the impedance to be between 31 and 35 ohms Problem 1 Predict the Cpk and dpm for this circuit using Monte Carlo simulation software such as DFSS Master Problem 2 Use SimWare to generate 100 circuit boards and compare the results to the predicted results from the Monte Carlo Simulation Leave the inputs at their default values which is R1 Mean z 100 R2 Mean z 50 R1 SD 2 R2 SD 2 Problem 3 Go back to your Monte Carlo simulation software and use it to determine to which input the process is most sensitive Based upon these results which of the following would cause a greater increase in the Cpk 1 Reducing R1 s standard deviation from 2 to 1 2 Reducing R2 s standard deviation from 2 to 1 Problem 4 In SimWare reset the current data by selecting Control Reset Current from the main menu Reduce H1 s standard deviation to 1 Open the Cpk analysis window by selecting Stats Cpk from the main menu Generate 25 data points Memorize the current curve by selecting Control Memorize Current and then reset the current data set by selecting Control Reset Current Return R1 s standard deviation to 2 and reduce R2 s standard deviation to 1 Generate 25 more data points 46 Does this result match the anticipated results in Problem 3 Problem 5 What actions would you t
5. your results by entering the optimal values and running the Cpk Analysis on the simulator Note A 3 level design will provide optimal results Vehicle Impact Safety Exercise 1 Analyze the process by performing a DOE to determine which factors shift the mean and or standard deviation Exercise 2 Use the results of the DOE to optimize the output Confirm your results by entering the optimal values and running the Cpk Analysis on the simulator DFSS Design for Six Sigma 45 DFSS Simulators SimWare provides six simulators which are designed to demonstrate DFSS concepts Resistor DFSS Basic Monte Carlo simulation from established transfer function no DOE required and Tolerance Allocation Stun Gun DFSS Two inputs one output simulation that requires a two level DOE and Monte Carlo simulation to optimize Nuclear Reservoir DFSS Four inputs one output simulation that requires a two level DOE and Monte Carlo simulation to optimize Spark Plug Design DFSS Four inputs one output simulation that requires a three level DOE and Monte Carlo simulation to optimize Washing Machine DFSS Four inputs one output simulation that requires a three level DOE and Monte Carlo simulation to optimize Washing Machine Advanced DFSS Identical to the Washing Machine simulation except the user has control of the standard deviations of the individual inputs Resistor DFSS Two components on a printed circuit board PCB
6. 1 64 1 55 1 Select SPC XL Analysis Diagrams Box Plot from Excel s menu If you do not see the SPC XL menu item you need to start SPC XL Ensure the data range is correct and click Next gt gt Box Plot Data Box Plot Data Please enter the range for the Box Plot source data SA 1 D 26 Help Our data sets are in columns so select the radio button Data sets in Next gt gt Columns and press Next 35 Data Orientation Would you like this range of data to be analyzed as one group of data or multiple data sets in rows or columns Data Orientation C One group of data Data Sets in Rows Example of data in Columns Example of data in Rows Question A Question B Question A 47 34 4 4 2 QuestonB 4 4 5 4 3 4 e SPC XL creates the Box Plot Box Plot Titles force a force b force c force d Measurement System Analysis 36 Measurement System Analysis Simulators Measurement Error MSA A simulator that provides a True Value and Measured Value output Demonstrates the impact of a poor measurement system in terms of process capability Compressor MSA Two operators with a single replicate Micrometer MSA Three operators with two reps per operator Visual Clarity MSA Two operators with two reps per operator Visual Clarity Advanced MSA Same as the Visual Clarity simulation but with inputs that allow you to change the variance components of th
7. 10 05 5 5 10 97 10 89 8 98 9 9 9 14 9 49 9 1 9 78 10 75 11 35 9 58 943 11 33 10 57 9 94 10 07 10 14 10 21 9 4 9 15 11 81 1154 10 8 1 11 31 12 02 10 19 9 7 9 11 9 89 10 33 9 82 e Export the data to Excel by selecting Export Excel from the main menu e Now that the data is in Excel you must create a MSA template in SPC XL for analysis Select SPC XL MSA gage capability Create MSA Template from Excel s menu If you do not see the SPC XL menu item then you need to start SPC XL e Indicate that you have 3 operators 2 replicates and 10 parts Enter 13 for the Upper Spec Limit and 7 for the Lower Spec Limit 40 MSA Parameters MSA Template MSA Parameters ee Number of Operators Appriasers Naber ofRepicates ak eue E number ren UAR Specification Limits nutu Upper Spec Limit Optional UU das acteur pina E SPC XL will create an empty template to enter your data For Attribute data enter A for Accept and R for Reject Description scissa Operator 1 Operator 2 Operator 3 Reference Rep 1 Rep 2 Rep 1 Rep 2 Rep 1 Rep 2 A ESA ES AP MESA ESO IES SS HERES eee Ness Ed ES oo AS A E A E NA AS EH TI I A Zu w d 5 J o EM eo 3 eee 9 01 S UM lo obl O eee Go back to the worksheet that holds the data you exported from SimWare Select just the output data WITHOUT th
8. 51 Index B Basie E TICE 17 C Clear WISI OFZ k cisco ee hectic cca en e EEN 9 Clear Mem oT E ta er eee eee 9 Control e CH DEE MULA E a ELA E EL ag 10 E E EE 10 D DESS MI E EE 13 DESS SIMIO ES ca tl rana Las scr il lap cos wen eee ee Teme ee ro locas 47 O A II A 10 DOE Multiple Outputs Smmulators sse nennen eene eene nn nn nn nn rr 46 DOE Single Output Simulators sse eene rn rr 44 DOF Analys nd T 10 DORA a e il e Tae oe 12 13 E Export AG E 8 Export DOE Dita E ee 12 H EF ee 10 M Measurement System Analysis Simulators 000000nnnnnnneonssssseseseeteeerensssssssssssserrrereesssssssssssssrererreesssssss 38 Feu o y KEE 9 Memorize Data 52 O Opena omul d EE 7 DO ek echte ESOT OE RC OY EVENT EER A ebe 5 R Rocer lI ERNEUT 8 S Sunu lator BXeECISES osos A eee lso 14 A eet 14 e e DE A NI O MN ME E MAD AA 13 AS ele 8 T d Kr EEN 9 92 53
9. AR e Select SPC XL Analysis Tools Correlation Matrix from Excel s menu bar If you don t see the SPC XL menu you need to start SPC XL e The first step is to ensure that you have selected the correct data range Click Ok to continue Correlation Coefficient zi Correlation Coeffi cient eae a aaa n be arranged such that each sample is in a column Yu root must select two or more columns E eer xs Since the data is in columns Click on the Data in Columns button Data Orientation Please select the orientation of your data Example of data in Columns Question A Question B 4 4 Column header must be non numeric 2 3 lt lt Data in Columns 4 Example of data in Rows Row headers must be non numeric QuestionA 42 34 Data in Rows gt gt QuestionB 445 4 e SPC XL calculates the correlation and gives us the results Correlation Matrix Operators Cust Service Reps Lines Operators 0 255587 Cust Service Reps 1 0 0 169296 0 250157 Lines 1 0 0 402023 time 1 0 The correlation coefficient for Operators vs time is 997 This indicates a strong negative correlation Anything below 7 and above 7 is colored red by SPC XL to indicate a strong correlation This indicates that as Operators is increased time will decrease Further ltems to Explore 28 Is it possible that two of the inputs would end up being highly correlated Why is the diagonal of the correlation
10. If you can redesign the spark plug so that it will fire lower than 20 kV then you can enter this new and highly lucrative market space The factors that control the performance are the angle of the spark tip to the plug body Groove Angle the power of the laser during welding Welding Power the angle of the laser during welding Laser Angle and the rotation speed of the plug during welding Rotation Speed The current design has the following parameters e Groove Angle N 45 2 e Welding Power N 2 01 e Laser Angle N 15 15 e Rotation Speed N 5 5 Problem 1 Perform a DOE on the simulator to get the transfer function The recommended design is a Three Level 4 Factor Central Composite Design This will require the use of another software package that performs DOE analysis such as DOE Pro Problem 2 Using the transfer function use software that performs Monte Carlo simulations like DFSS Master and predict the current number of 48 defects from the current design Problem 3 Using the DOE software find the new input settings that will optimize the design reduce the number of defects Verify the new settings in SimWare Problem 4 Using your Monte Carlo simulation software find the optimal input settings that will optimize the design Verify the new settings in SimWare Washing Machine DFSS Two key performance metrics for a washing machine are the cleaning effectiveness and the energy consumptio
11. from the main menu Toolbar The toolbar at the top of SimWare is designed to allow rapid access to the majority of SimWare s features Begin simulation Generate 1 25 Start Cpk Start Histogram medium speed or 100 values Analysis Analysis Reset delete yp Data tH a B 6i JA B s 3 LN Ke 1 G P A C pot d s show Stop simulation Data Grid Start Control oan pog Memorize Chart Analysis Analysis Current Data Statistical Analysis Cpk Analysis To start the Cpk Analysis select Stats Cpk from the main menu Cpk Analysis will display a graphical bell curve and associated statistics The area out of specification will be red while the area in spec will be blue Note you must have variation in the data in order to get a Cpk analysis When you start the Cpk Analysis you will get a new window for each output The statistics calculated are e Mean e Standard Deviation e Cp e Cpk e DPM defects per million Sigma Capability Control Chart To create a control chart select Stats Control Chart from the main menu When you first create the control chart it will plot the data until it has enough data to calculate the control limits the default is 25 plotted points When the contro limits have been calculated they will not be reset until you reset the current data You can change the subgroup size the way out of control checking is performed for both the Xbar and R chart and the number of p
12. matrix all 1 0 If this were your process where would you set the inputs Scatter Plot Show me how e Start the simulation by selecting Start Speed Medium from the main menu e While the simulation is running slowly increase the inputs until you reach the maximum of 1200 degrees Kiln Temp Export the data by selecting Export Excel from the main menu n Excel select the data including the column headings Kiln Temp thickness f03 2 54 703 2 64 703 2 51 T03 247 703 2 55 703 2 53 804 3 03 804 3 12 804 3 14 a04 3 04 804 3 11 804 29 Select SPC XL Analysis Diagrams Scatter Plot from Excel s menu 29 If you do not see the SPC XL menu item you need to open SPC XL e Ensure that the selected data range is correct and press the Next gt gt button Scatter Chart Scatter Chart Data Please enter the range for the scatter chart A 1 8 57 Es Next gt gt e Indicate that the data is in columns by clicking on the Data in Columns button Data Orientation Please select the orientation of your data Example of data in Columns Column header must be Question A Question B EE 4 2 3 lt lt Data in Columns Example of data in Rows Row headers must be non numeric Question A quA Data in Rows gt gt Question B 4454 e SPC XL creates the scatter plot 30 Scatter Chart Kiln Temp vs thickness SE y 0 005x 0
13. the menu bar and SPC XL will perform 1 EM MS S 5 To 38 ZEN pass 99 o EE l1 E _ AM o d 1 T capability the Measurement System Analysis DOE Single Output 42 DOE Single Output Simulators Airline Simulation DOE Commercial Profit DOE Space Ship Thrust DOE Temperature Control DOE Hydraulic Hammer DOE Blood Analysis DOE Broadcast News DOE Vial Filling DOE Airline Simulation DOE A new commuter airline has established on time departures as a primary business goal You have been tasked to identify the factors that significantly affect the time of departure recommend a method of improvement and confirm the results After completion of an Ishikawa diagram the factors selected as most likely to effect the time of departure are spent on training of Jets of Employees and Overbooked The output is Minutes Late with negative values indicating that the flight left early Exercise 1 Perform a DOE on the four inputs and export the data for analysis in your DOE software Which factors shift the mean and or standard deviation Exercise 2 Based upon the results in your DOE find optimal values for the inputs Enter these values into SimWare and confirm your results For more information about DOE Pro visit http www sigmazone com doepro htm Commercial Profit DOE Exercise 1 Analyze the process by performing a DOE to determine which factors shift the m
14. 1 55 Water Height 45 Angle 48 Water Height Agitator Angle 20 Water Height 60 Agitator Angle 55 Agitator Width 39 Drum Accel Energy Consumption 250 88 Accel 44 Agitator Width 22 Agitator Angle 19 Accel Agitator Angle Basic Stats Basic Stats Simulators SimWare includes 7 simulators that are designed to teach basic statistical tools t Test Oil Refinery F Test Data Entry Correlation Call Center Scatter Plot Kiln Control Chart Car Compactor Summary Stats Delivery Time Box Plot Phone Cover t Test Oil Refinery You are in the employ of a company that owns two oil refineries Both refineries operate using electric dehydration and desalinization to refine the low sulfur crude that is provided by your customers One of your key metrics is the number of barrels of crude that can be refined per day The plant manager of the Northern Plant has proposed that his new method of handling waste byproducts has enabled the Northern Plant to produce more barrels of oil than the Western Plant Since the two plants are identical in every other way you decide to do at Test to determine if there is a statistical difference between the two plants Problem 1 Perform a t Test on the data from each plant with the Crude Price 20 per barrel to determine if there is a difference between the Northern and Western Plants Since the number of barrels that can be produced per day is dependent on the orice of crude you decide to withh
15. 3 level design will provide optimal results Broadcast News DOE Exercise 1 Analyze the process by performing a DOE to determine which factors shift the mean and or standard deviation Exercise 2 Use the results of the DOE to optimize the output Confirm your results by entering the optimal values and running the Cpk Analysis on the simulator Vial Filling DOE Exercise 1 Analyze the process by performing a DOE to determine which factors shift the mean and or standard deviation Exercise 2 Use the results of the DOE to optimize the output Confirm your results by entering the optimal values and running the Cpk Analysis on the simulator 44 Note A 3 level design will provide optimal results DOE Multiple Output DOE Multiple Outputs Simulators Restaurant DOE Space Shuttle Main Engine DOE Vehicle Impact Safety Restaurant DOE As a small restaurant owner you are responsible for timely delivery of service Using the available inputs reduce the time required for each of the services provided Exercise 1 Analyze the process by performing a DOE to determine which factors shift the mean and or standard deviation Use this information to optimize the three responses outputs Space Shuttle Main Engine DOE Exercise 1 Analyze the process by performing a DOE to determine which factors shift the mean and or standard deviation Exercise 2 Use the results of the DOE to optimize the output Confirm
16. 40 20 7936 8075 20 6195 91401 From the Excel menu select SPC XL Analysis Tools t Test Matrix Mean If you do not see the SPC XL menu item it means you need to start SPC XL 21 e The first screen SPC XL will present you with will confirm the range selected It Test Matrix t Test Mean Please enter the range for the test The data should be arranged such that each sample is in a column You must select two or more columns SBS1 SCS26 e The next will ask if your groups are in rows or columns Data Orientation Please select the orientation of your data Example of data in Columns Question A Question B Column header must be non numeric 2 3 d lt lt Data in Columns 4 Example of data in Rows Row headers must be non numeric Data in Rows gt gt i e The result will be created on a new worksheet t Test Analysis Mean P value 0 556 The simplest way to interpret a pvalue is that we have 1 pvalue 100 confidence that the means are not equal In this case we are only 44 confident that the means are not equal Had the p value been 1 then we would be 90 confident that the means are not equal The second part of the problem asks what happens if you change the crude price to 30 per barrel 22 e Change the crude price to 30 e Reset the current data by selecting Control Reset Current from the main menu This will remove the previous data from the dat
17. 9463 R 0 9899 45 thickness 25 on aso 900 950 1000 1050 1100 1150 1200 1250 Kiln Temp The high R value indicates that the model is a good fit As the Kiln temperature increases so does the thickness Further Items to Explore What is the difference between R and a p value What are the minimum and maximum values for R Control Chart Show me how e Start a control chart analysis by selecting Stats Control Chart from the main menu Generate 25 data points by selecting Start Generate 25 from the main menu e The control chart should add data until it has sufficient data to establish the control limits 31 280 15 1 mios LICL a 275 2183 MON sium e Expand the tree on the right of the control chart to see the value of the control limits In this case the UCL is 275 and the LCL is 222 with the center at 248 Further Items to Explore Export the data and repeat the analysis in SPC XL Will the control limits always be the same If not why would they change Control Chart Show me how 2 e Select the drop down box and change the value to 4 32 Summary Stats Show me how Generate 25 data points by selecting Start Generate 25 from the main menu e Export the data to Excel by selecting Export Excel e n Excel select the data using the mouse e Select SPC XL Analysis Diagrams Summary Stats Dot Plot from Excel s menu If you don t s
18. EY Digital Computations SimWare Pro User s Manual Copyright 2003 Digital Computations Inc All Rights Reserved Unauthorized Duplication Prohibited By Law Table of Contents USING SIMWABE ege Eege 5 IOWA RE TEE 5 Eur A TO 5 EEN 5 SAP E A 7 A e o CO OO MM ae eae MM 8 EXPO Dallas 8 EE EE eee eee nee ne oO o 8 WIGIMIOMNZ 6 e sxnsewsdsusnuvagesne soreness ousibenvousnuessuansaee sudeteea duspeusndeanseenduasscanduasets 9 CledrMOmollzetla ost EDIDI ID ADI D Im 9 ole le 9 Statistical ANS Siro iE uc tut 10 Se VAY aida 10 CODO GS E oce Um detur Ib bb Np uM rarer er reer cre ere erect 10 SNE 9 WE 10 Biel un E 10 DE Analysis ocu E OA E 11 Reie re oM ee o o lodo li 11 DOE VV IZA SIED PUE 11 DOE Wizard SUS DO irene tee ie tee e O as ties 11 DOE WZI SIED artritis alante 11 EXPO DUE Dd T 12 op 13 DOE PRO A 13 DIS 5 SV EE 13 SIMULATOR EXERCISES paa ees set eee 13 oimulalor EXerciS85 usus tenti dd osa aiens 14 cae MH tintin 14 BASIC Stats 17 BASIS Ee el Sidi 17 OR RB ellliel V E 17 EC EE 18 Correlation Gall CON ran AAA 18 SCE PRIOK in T Tm 19 Control Gbalt Lar GOmpactol scraps Re 19 Summary Stats Delivery TIME errian 20 BOX Plot PRRORS GOVE REO T 20 SHOW Me TOW Basilio Slats snoot 21 ETEStCSMOW qe NOW sacada 21 ETES SMOW me NOW e eo be Ende 23 Correlation Show me how 26 Scatter Plot SHOW Me NOW iaa 29 Control Chart Show me how
19. In most cases you must have statistical software to perform the analysis such as SPC XL DOE Pro and DFSS Master Basic Stats Simulators Exercises designed around basic statistical concepts such as t Test F Test Correlation Scatter Diagrams Box Plots and Summary Stats Measurement System Analysis Simulators Exercises designed to practice analyzing and evaluating measurement systems using MSA DOE Single Output Simulators Exercises that allow you to perform a DOE on a simulated process to include optimization and confirmation DOE Multiple Outputs Simulators Exercises designed to practice DOE on processes that have more than one output response DFSS Simulators Exercises which enable you to practice more advanced techniques such as Robust Design and Tolerance Allocation Most of these simulators require a good foundation in DOE for completion Solutions This section contains the underlying models for some of the simulators so you can compare your results with those in the Simulator Note that your solution will not match this solution exactly Measurement System Simulations DOE Single Output Simulations DOE Multiple Output Simulations Design for Six Sigma Simulations Measurement System Simulations Compressor Operator 1 is reading the display to the tenths digit Operator 2 is reading the display to the hundredths digit This artificially deflates the standard deviation of Operator 1 which can be easily seen b
20. a grid Now just follow the instructions from the start of this section start by generating 25 data points and you should get a result similar to the following t Test Analysis Mean The p value now shows 0 0 which means that we are over 99 9 confident that the means are different When the Crude Price is 20 per barrel there is no difference in the output from the northern and western plants However when the crude price is 30 there is a difference between the two plants Further inquiries of interest would be to determine if there is a difference when the crude price is 21 per barrel If not what if you generate 100 data points instead of 25 Why would the number of data points affect the analysis F Test Show me how e Ensure that the Project Status is set to Pre Project Pre Project e Project Complete Project Status e Generate 25 data points by selecting Start Generate 25 from the 23 menu bar Memorize this data by selecting Control Memorize from the menu bar Reset the current data by selecting Control Reset Current from the menu bar Change the Project Status to Project Complete e Pre Project 24 Project Status Generate 25 data points by selecting Start Generate 25 from the menu bar Export the data to Excel by selecting Export Excel from the menu bar The memorized and current data will be exported to Excel on separate worksheets k H Cu
21. ake in this circuit design Stun Gun DFSS You are designing the next generation of personal protection stun guns For proper operation the Stun Gun should operate between 25 000 and 30 000 pulse watts The major factors that contribute to the number of pulse watts are the power and frequency of the stun gun Each gun manufactured will have variation in both power and frequency The current design has the following parameters e Power N 2 1 e Freq N 10 3 Note Standard deviation estimates are from the previous model of stun gun Problem 1 Perform a DOE on the simulator to get the transfer function The recommended design is a Two Level Four Run Full Factorial This will require the use of another software package that performs DOE analysis such as DOE Pro Problem 2 Using the transfer function use software that performs Monte Carlo simulations like DFSS Master and predict the current number of defects from the current design Problem 3 Using the DOE software find the new input settings that will optimize the design reduce the number of defects Verify the new settings in SimWare Problem 4 Using your Monte Carlo simulation software find the optimal input settings that will optimize the design Verify the new settings in SimWare Nuclear Reservoir DFSS The level of the coolant water in a nuclear power plant is critical to the cooling and therefore safety of the plant Proper operation of the plant requi
22. ator supports For example if the current simulator has three inputs then you will not be given an option of a 2 full factorial After selecting the desired design select Next and proceed to step 2 DOE Wizard Step 2 After selecting the desired design you may be prompted to select the factors you wish to include in the DOE Note If the design you chose requires all of the factors in the current simulator you will not see step 2 select the factors using the check box next to the name of each factor and select Next to continue to Step 3 DOE Wizard Step 3 Now that you have selected the factors for the DOE you need to determine what the range for each factor will be The default range is low and high for the extremes in each factor however you may change this value to any value that is possible given the design you have chosen It is possible to choose values that are not possible and if you do you will get an advisory message For example if you have a three level design and you have a factor which is marked as Integer Only then a low of 1 and a high of 4 would not be possible because the center point would be 2 5 which is not an integer However a low of 1 and a high of 5 would be possible because the center point would be 3 which is allowed You are now ready for Step 4 DOE Wizard Step 4 The final step is to select the number of replicates You may have between 1 and 20 replicates for the DOE When done select Fi
23. e Welcome to SimWare by Digital Computations Inc SimWare is a statistical simulator which allows you to practice analyzing data and taking corrective action as a result of the analysis For information about how to get started go to the Overview page To learn about the toolbar go to the toolbar page For practice exercises go to the exercise page Getting Started Overview SimWare is a statistical simulator which allows you to practice analyzing data and taking corrective action as a result of the analysis SimWare will allow you to perform a variety of analysis such as t Test F Test Measurement System Analysis and primarily Design of Experiments DOE When you first start SimWare you will need to open a simulator from the Simulator menu item At this point the screen is divided into three main areas The top area is the toolbar menu bars the center area is were the animation and analysis is performed and the bottom area is where you can control the inputs to the simulation Main menu and toolbar Process outputs on d animated graphic gt nisaled Graphic j ado al Saz Como ExprLDa ice Fo Ee H EP L E ta ty mao RR EI A Z A HMaxial Acceleration 0 Rib Compression 0 Design Cost 3 000 E Anno Cms Sugpert Center 7 In 45 Wetica Support D insuster C In 4 Boe Dinar 2 En An Enam Luxe 35 to 601 Process inputs user adjustable Begin simulation Generate 4 25 start Histogram ey
24. e MSA Measurement Error MSA The purpose of this simulator is to demonstrate the impact of a poor measurement system on the capability of the process The simulator has two outputs the True Value and the Measured Value The true value represents the actual measurement before any measurement error has been introduced The measured value represents the measurement after the measurement error has been introduced In most applications you can not really know the true value however you can estimate the variation that is introduced by the measurement system Exercise 1 Start a Cpk analysis for this simulator and generate 25 data points with the default inputs What is the Cpk of the True Value vs the Measured Value Exercise 2 Reset the data and increase the Measurement Standard Deviation to a value of 6 Again compare the True Value vs the Measured Value Exercise 3 Reset the data and change the Measurement Bias to a value of 10 Compare the Cpk of the true vs measured values Exercise 4 Can a poor measurement system decrease the Cpk of a process Can a part that is in reality good be classified as bad Can a bad part be classified as good If your process is operating with a poor Cpk how can you tell if the problem is primarily the measurement system or the process Compressor MSA Two operators are concurrently reading the pressure of a compressor Exercise 1 Perform a measurement system analysis on t
25. e column headings and without the input column 41 Process Non repi opl rep2 jop2 repi jop2 rep2 op3 repi opi rep2 e Copy this data by selecting Edit contains the MSA template Select the cell that corresponds to Operator 1 Rep 1 Part 1 Should be cell C12 and select Edit 10 10 93 10 67 10 56 11 75 9 43 8 98 10 75 9 94 11 81 10 13 923 10 18 11 21 10 65 98 9 9 11 35 10 07 11 54 oF 9 09 9 83 10 94 11 09 10 05 9 14 9 58 10 14 10 6 9 11 9 53 3 00 9 02 10 07 10 64 11 36 10 79 11 43 10 94 9 2 11 58 11 01 3 5 10 97 10 69 9 49 9 1 39 16 9 33 11 33 10 57 10 21 94 9 15 12 11 31 12 02 9 89 10 35 9 82 Copy Go back to the sheet that Paste When you are done the MSA template will look like the following Do not be concerned if the lines are overwritten when you paste in the data Sao ojo ajojoje jej MSA Data Template 10 67 10 07 11 38 10 56 11 21 10 79 11 43 1 Operator 1 Reference Rep 1 Hep 2 9 j 1014 10 21 9 15 Il IR 10 1 10 11 75 10 65 11 09 11 58 11 01 10 05 Ka 9 y 11 35 11 33 10 57 U a 5 y NIS B E epee LAS E ees SEV For Attribute data enter A for Accept and R for Reject Description Operator 2 Operator 3 Rep 1 Rep 2 Rep 1 Rep 2 10 19 97 You are now ready for the analysis Select SPC XL MSA gage Anova Analysis from
26. e the desired glaze thickness Problem 3 Create a Cpk diagram in SimWare set the kiln to the temperature you chose in problem 2 and confirm your analysis Control Chart Car Compactor As a car compacting fiend you are obsessed with the quality of your car compaction As such you have the ability to adjust the mean and standard deviation of the resulting thickness at will However you would like to investigate if your process is in statistical control Problem 1 Leave the process mean and standard deviation the inputs at their default settings and start a control chart until the control limits have been established leave the sub group size set at the default of one What are the center UCL and LCL values for the X Xbar chart Based upon the process mean and standard deviation do those numbers make sense Show me how Problem 2 Run the simulation for 100 more data points without adjusting the mean or standard deviation Do you see any out of control conditions red points If you did why would you see out of control conditions if the mean and standard deviation didn t change Problem 3 Adjust the mean 1 5 standard deviations to the right 265 and generate 100 more data points Did the number of out of control conditions increase Problem 4 Change the subgroup size in upper right corner to 4 This will cause the control limits to be reset Generate 100 more data points to generate the control limits What are the c
27. ean and or standard deviation Exercise 2 Use the results of the DOE to optimize the output Confirm your results by entering the optimal values and running the Cpk Analysis on the simulator Space Ship Thrust DOE Exercise 1 Analyze the process by performing a DOE to determine which factors shift the mean and or standard deviation Exercise 2 Use the results of the DOE to optimize the output Confirm 43 your results by entering the optimal values and running the Cpk Analysis on the simulator Temperature Control DOE Exercise 1 Analyze the process by performing a DOE to determine which factors shift the mean and or standard deviation Exercise 2 Use the results of the DOE to optimize the output Confirm your results by entering the optimal values and running the Cpk Analysis on the simulator Hydraulic Hammer DOE Exercise 1 Analyze the process by performing a DOE to determine which factors shift the mean and or standard deviation Exercise 2 Use the results of the DOE to optimize the output Confirm your results by entering the optimal values and running the Cpk Analysis on the simulator Blood Analysis DOE Exercise 1 Analyze the process by performing a DOE to determine which factors shift the mean and or standard deviation Exercise 2 Use the results of the DOE to optimize the output Confirm your results by entering the optimal values and running the Cpk Analysis on the simulator Note A
28. ee the SPC XL menu item then you need to start SPC XL e Confirm the range is correct by clicking on Next gt gt 33 Summary Stats Dot Plot X Summary Stats Dot Plot Data Please enter the range for the Summary Stats source data e SPC XL analyzes the results and creates the summary statistics Count 25 Mean 60 46 Median 61 Mode 53 Max f4 Min 53 Range 21 5td Dev Pop 4 964836352 Std Dev Sample 5 06 7214883 Variance Pop 24 6496 Variance Sample 29 67666667 Skewness 0 632081153 Kurtosis 0 810346852 95 Conf Interval for Mean Upper Limit 62 5 164349 Lower Limit 56 56055651 99 Conf Interval for Mean Upper Limit 63 31453063 Lower Limit DI 64546131 Box Plot Show me how e Generate 25 data points by selecting Start Generate 25 from the main menu e Export the data to Excel by selecting Export Excel n Excel select the data using the mouse 34 T e force a 1 88 2 06 3 19 2 09 2 53 2 02 2 12 1 96 1 63 2 08 2 51 2 67 2 76 2 24 2 13 2 91 2 25 2 53 2 61 2 98 2 56 2 03 2 13 3 05 2 67 force b 22 2 05 2 11 1 31 2 16 1 67 2 18 201 1 85 2 04 1 8 21i 1 66 1 62 2 06 2 20 2 05 1 95 1 99 2 06 1 87 2 23 2 03 1 56 force c 1 03 0 64 1 47 1 52 1 08 1 66 1 25 1 3 1 51 1 21 1 07 1 29 1 16 1 07 1 28 1 11 1 14 147 1 19 1 44 1 05 1 36 1 12 1 23 force_d 1 71 1 86 1 82 1 66 1 7 7 1 76 1 75 1 66 1 66
29. enter UCL and LCL values How do these values compare with the control limits when the sub group size was 1 Why is there a difference Show me how Summary Stats Delivery Time You are responsible for the shipping of products from the small parts facility in your company You suspect a problem in shipping and as such you would like to collect some data and get a baseline for shipping time performance Problem 1 Run the simulation and export to Excel for analysis Analyze the shipping time and calculate the summary statistics mean standard deviation confidence intervals for the mean etc Show me how Problem 2 What is the average shipping time What is the standard deviation Problem 3 Calculate the following e Mean shipping time plus two standard deviations Mean shipping time minus two standard deviations What percent of shipments should fall between these two numbers Problem 4 What is the 95 confidence interval for the mean upper and lower limit Compare these numbers to the mean 2 standard deviations Should the numbers be the same and if not what is the difference Box Plot Phone Cover You are responsible for designing the sliding phone cover for your company s next mobile phone One of the key metrics for the phone is the force required to actuate slide the cover from the closed to the open position After a couple of months of development there are four competing designs Problem 1 Genera
30. he two operators What appears to be the problem 3 Micrometer MSA In this simulation you have three operators who all measure the same part twice The USL is 13 and the LSL is 7 Exercise 1 Perform a MSA analysis What percent of the variation is being caused by the following components e Part to Part e Repeatability e Reproducibility If this was your process what would you do first in order to increase the capability Show me how Exercise 2 Change the process mean the only input to a value of 11 and repeat the analysis Visual Clarity MSA Two technicians perform a visual clarity analysis of a sample without the use of a Secchi Disk rapid testing You have set up a MSA so that the analysis is performed twice by both operators on the same sample The sample is then measured four times by highly qualified operators using a more precise method a Secchi Disk The average of the four measurements using the Secchi Disk are the reference values Exercise 1 Perform a MSA analysis What percent of the variation is being caused by the following components e Part to Part e Repeatability e Reproducibility Is there any Bias in this measurement system If this was your process what would you do first to increase the capability Exercise 2 Change the process mean the only input to a value of 11 and repeat the analysis Visual Clarity Advanced MSA This simulator is identical to the Visual Clarity simulator
31. ice by selecting File WebService from the menu bar The WebService is intended to allow users to create their own simulators and share them with others Note that simulators that are shared in this manner are not by Digital Computations and are therefore not supported Open SimWare supports opening a simulator from a file with a simw extension This allows you to create and share simulators with others Note that simulators that are shared in this manner are not by Digital Computations and are therefore not supported Start Simulator SimWare supports two methods for running a simulation Continuous Simulation To start a continuous simulation select Start Speed and then the desired speed from the menu bar The simulation will run until you stop the simulation by selecting Start Speed Stop from the menu bar Generating a Specific Number simWare allows you to generate 1 25 100 or 1000 iterations and then stops automatically To generate a specific number select Start Generate followed by the desired number from the menu bar Export Data The data that is generated by your simulator can be exported to Excel the clipboard a comma separated file or a tab delimited file Exporting to Excel requires that MS Excel version 2000 or later be installed To export data select Export Data followed by the desired format from the menu bar What gets exported If you are exporting to Excel then both the curre
32. ill be able to access many tools for statistical analysis including Control Charting Cpk analysis TBI Histograms Paretos MSA and much more The data analysis in this help system was performed using SPC XL For more information visit Nttp www sigmazone com spcxl htm a D DOE PRO XL is the ultimate in Design of Experiments packages Multiple response capability coupled with advanced analysis and optimization features enables you to take control of your experiments all from within Microsoft Excel The data analysis in this help system was performed using DOE Pro DOE PRO XL is an Excel add in which gives users powerful yet easy to use DOE Design of Experiments capabilities From within Excel you will be able to create designs analyze designs using multiple regression plot results optimize and predict For more information visit http www sigmazone com doepro htm DFSS Master is an Excel add in which facilitates Design for Six Sigma initiatives Features include Monte Carlo simulation Design Scorecards Tolerance Allocation tables Parameter Design for factor optimization and much more For more information visit http www SigmaZone com dfssmaster htm Simulator Exercises Simulator Exercises In order to facilitate the learning of statistical concepts problems or scenarios around each simulator are available The exercises allow you to practice using SimWare and your statistical software
33. ize the design reduce the number of defects Verify the new settings 49 in SimWare Problem 4 Using your Monte Carlo simulation software find the optimal input settings that will optimize the design Verify the new settings in SimWare Washing Machine Advanced DFSS The advanced version of the simulator allows you to control the standard deviation of the inputs Two key performance metrics for a washing machine are the cleaning effectiveness and the energy consumption Competitive market pressures demand better effectiveness with lower energy consumption The specification for the model under development is a cleaning effectiveness of at least 290 with no more than 200 Watt Hours of energy consumption The factors that influence the performance metrics are Agitator Angle Agitator Width the acceleration of the drum Drum Accel and the height of the water as set by the consumer Water Height While all the factors have variation the consumer can set the Water Height to any value between 50 and 100 using the control on the unit The current design has the following parameters e Agitator Angle N 45 2 e Agitator Width N 6 01 e Drum Accel N 4 01 e Water Height N 60 1 Problem 1 Perform a DOE on the simulator to get the transfer function The recommended design is a Three Level 4 Factor Central Composite Design This will require the use of another software package that performs DOE analysis such as DOE P
34. lotted points required to establish the control limits by selecting these options on the right side of the window Histogram To create a histogram select Stats Histogram from the main menu The histogram window will appear and create a histogram of any further data generated after initial creation The mean standard deviation Cp Cpk and DPM are also calculated concurrently DOE Analysis To start a DOE select Stats DOE from the main menu The DOE functionality in SimWare enables you to quickly create designs and generate the associated dependent data output The analysis of the DOE must be performed in a DOE software package such as DOE Pro When you start the DOE Analysis the DOE Wizard will start and lead you through the process of determining what design you would like to run selecting the factors for the design selecting the lows and highs and determining the number of replicates When you have completed the wizard a new window will appear with the 10 independent data inputs and dependent data outputs This can be exported to Excel for analysis DOE Analysis DOE Wizard Step 1 The first step in the DOE Wizard is to select the type design that you would like to create You may select from a variety of 2 or 3 level designs that are built into SimWare To change from a 2 to 3 level design use the radio button on the left side of the window SimWare will only show you designs which the current simul
35. n Competitive market pressures demand better effectiveness with lower energy consumption The specification for the model under development is a cleaning effectiveness of at least 290 with no more than 200 Watt Hours of energy consumption The factors that influence the performance metrics are Agitator Angle Agitator Width the acceleration of the drum Drum Accel and the height of the water as set by the consumer Water Height While all the factors have variation the consumer can set the Water Height to any value between 50 and 100 using the control on the unit The current design has the following parameters e Agitator Angle N 45 2 e Agitator Width N 6 01 e Drum Accel N 4 01 e Water Height N 60 1 Problem 1 Perform a DOE on the simulator to get the transfer function The recommended design is a Three Level 4 Factor Central Composite Design This will require the use of another software package that performs DOE analysis such as DOE Pro Problem 2 Using the transfer function use software that performs Monte Carlo simulations like DFSS Master and predict the current number of defects from the current design While the water height has a standard deviation of 1 the user can control the water height and set it at any value between 50 and 100 Consider modeling this in your DFSS software as a uniform distribution from 50 to 100 Problem 3 Using the DOE software find the new input settings that will optim
36. n decreases with an increase in Neutrinos Vehicle Impact Safety Triaxial Acceleration 18 8 Cross Support Dia 7 Box Dia 2 3 Beam Angle 8 Box Dia Beam Angle 8 Cross Support Dia Box Dia The Triaxial Acceleration standard deviation decreases with an increase in Box Diameter Rib Compression 14 7 Cross Support Dia 2 1 Vert Support Dia 1 2 Box Dia 1 1 Beam Angle 5 Box Dia Beam Angle 84 Cross Support Dia Vert Support Dia The Rib Compression standard deviation decreases with an increase in Box Diameter Design Cost 140 27 Cross Support Dia 56 Vert Support Dia 110 Box Dia The Design Cost standard deviation is not affected by the model Design for Six Sigma Simulations Resistor The equation for impedance is on the animated graphic and is R4 R2 R1 R2 Stun Gun Pulse Watts 27500 2400 Power 125 Freq 2600 Power Freq Nuclear Reservoir Water Level 700 50 Plug Press 122 Bellow Press 66 Ball Press 100 Water Temp 96 Water Temp Plug Press 33 Bellow Press Ball Press An interaction causes water temp variation to induce significant variation in water level if the PlugPress is high 50 Spark Plug Design KV to Spark 25 2 8 Groove Angle 2 Welding Pwr 1 Laser Angle 1 7 Rotational Speed 1 6 Rotational Speed 2 1 Rotational Speed Groove Angle The standard deviation is lowest with Rotational Speed in central setting 5 due to the quadratic Very difficult to pick up 16 Washing Machine Design Cleaning Effectiveness 33
37. nish and SimWare will create the DOE Design for you You will also be asked if you would like SimWare to generate the replicates for you If you indicate Yes then the dependent data will be filled according the design At this point you should export the data to Excel for analysis in a software package such as DOE Pro Export DOE Data If you have created a DOE in SimWare then while the DOE output is visible a new menu item will be available called DOE Control This menu item allows you special control over the DOE options in SimWare To export the data select DOE Control Export then the desired format You have the option of exporting the DOE data in either DOE Pro or stacked formats Excel DOE Pro Format The entire DOE including the dependent and independent data the formatting and the coding will be exported into a DOE Pro design sheet which is ready for analysis You must have Excel 2000 or later installed to export to Excel Excel Stacked The dependent and independent data will be exported into Excel in the format that is usually required by DOE software other than DOE Pro You must have Excel 2000 or later installed to export to Excel How do I export the contents of the Data Grid See Exporting Data 12 E A DOE PRO DFSS Master aos SPC XL is a Microsoft Excel add in which gives users powerful yet easy to use SPC Statistical Process Control capabilities From within Excel you w
38. no longer than 70 seconds You have a variety of factors in your control including the number of operators the number of customer service representatives and the number of lines trunks Problem 1 Start the simulation and randomly change the of operators of customer service reps and of lines If you start a continuous simulation on medium speed you can change these inputs while the simulation is running Export the data and perform a correlation analysis in a statistical software package such as SPC XL Which inputs are correlated with time on hold Show me how Problem 2 What does a positive vs a negative correlation coefficient mean Problem 3 Where should you set the inputs in order to reduce the time on hold 18 Scatter Plot Kiln As the manager in a small arts and crafts store you are responsible for the results of the glazing process In order to attract customers you need to ensure that your glazing process provides a glaze thickness of at least 4 2 mm but no more than 4 8mm You suspect that the temperature of the kiln is important in the thickness of the glaze Problem 1 Start the simulation in continuous mode medium speed is best and slowly increase the Kiln Temp from the minimum to the maximum Export the data and create a scatter diagram Does changing the Kiln Temp change the Glaze Thickness Show me how Problem 2 Based upon your scatter diagram what is the optimal kiln temperature to produc
39. nt data and memorized data if any exist will be exported to a workbook in Excel If you are exporting to the clipboard a comma separated file or a tab delimited file then you must export the current data and memorized data separately How do I export the results of a DOE See Export DOE Data Reset Current As you generate data the data is stored in the data grid and the analysis is performed by any statistical tools that you have started If you would like to reset the analysis and delete the data from the grid select Control Reset Current from the main menu bar Reset Current does not affect any memorized data To delete the memorized data you need to Clear Memorized Memorize In many cases it is desirable to compare previous data with the current data To assist in this SimWare allows you to memorize the current data by selecting Control Memorize from the main menu e Any data in the Data Grid will be placed on a new tab entitled Memorized Data e The current Cpk curve will be represented by a single line and the Statistics for the memorized data will be retained e The current Histogram will be reflected in different colors and the statistics for the memorized data will be retained e There is no visual indication of memorized data on the control chart however the statistics will be retained Clear Memorized After you have memorized data you can clear the data by selecting Control Clear Memorized
40. old your judgement and wait until the price of crude increases to 30 per barrel Problem 2 Change the price to 30 per barrel and redo the analysis Show me how F Test Data Entry Your company takes insurance claims via telephone in centralized call centers for multiple insurance brokers Each broker has slightly different requirements in terms of the fields data for each call to be collected Over time this has lead to multiple screen formats to which the operator must adapt Your project is to make these forms as similar as possible to make the data entry easier on the operator Problem 1 Perform an F Test to determine if there has been a significant change in the standard deviation in the time required to complete a record Show me how Problem 2 Perform a t Test to determine if there has been a significant change in the mean Problem 3 The goal is for all operators to finish a call in less than 90 seconds Based upon the information you have gathered in Problems 1 and 2 predict if the process capability Cpk will be improved by implementing the project Problem 4 Perform a Cpk analysis and compare the results to your prediction Correlation Call Center You are the manager of a small call center which handles inbound inquiries from a variety of vertical industries One of your key metrics is the time customers spend on hold waiting for a customer service representative Most of your contracts require that this be
41. res that the fluid level be between 700 and 900 thousand gallons in normal operation The factors that influence the fluid level are Plug Pressure Bellow Pressure Ball Valve Pressure and Water Temperature While you can control the Water Temperature it is very expensive and you would prefer to allow it to vary anywhere between 70 and 100 degrees The current design has the following parameters e Plug Pressure N 50 1 47 e Bellow Pressure N 12 05 e Ball Valve Pressure N 145 1 e Water Temp N 91 2 5 Problem 1 Perform a DOE on the simulator to get the transfer function The recommended design is a Two Level 16 Run Full Factorial This will require the use of another software package that performs DOE analysis such as DOE Pro Problem 2 Using the transfer function use software that performs Monte Carlo simulations like DFSS Master and predict the current number of defects from the current design Problem 3 Using the DOE software find the new input settings that will optimize the design reduce the number of defects Verify the new settings in SimWare Problem 4 Using your Monte Carlo simulation software find the optimal input settings that will optimize the design Verify the new settings in SimWare Spark Plug Design DFSS Your company is trying to enter a new market for high performance spark plugs The key performance measure for this market is the kV required by the plug to generate a spark
42. ro Problem 2 Using the transfer function use software that performs Monte Carlo simulations like DFSS Master and predict the current number of defects from the current design While the water height has a standard deviation of 1 the user can control the water height and set it at any value between 50 and 100 Consider modeling this in your DFSS software as a uniform distribution from 50 to 100 Problem 3 Using the DOE software find the new input settings that will optimize the design reduce the number of defects Verify the new settings in SimWare 90 Problem 4 Using your Monte Carlo simulation software find the optimal input settings that will optimize the design Verify the new settings in SimWare Problem 5 Change the Agitator Angle SD to a value of 2 and redo the analysis How does the answer change Memorize Data Select Control Memorize from the main menu Memorize will store the current data on the Data Grid Cpk Analysis and Histogram for comparison to future generated data lt is similar to taking a snap shot of the current data To clear the memorized data select Control Clear Memorized from the menu bar Statistical Analysis SimWare supports Cpk Analysis Control Charts Histogram and Design of Experiments To start any statistical analysis select Stats then the desired tool from the main menu When you begin a statistical analysis the current data will automatically be reset
43. rrent Data Memorized Data Fi Copy the current data to a new worksheet and give it the column name Project Complete Copy the memorized data to the new worksheet and give it the column name Pre Project Select the two columns including the column headings Project Complete Pre Project 58 45 60 50 T6 64 60 12 ZA 57 T8 99 59 24 AT 50 46 30 69 50 116 44 81 57 50 63 69 61 56 63 IK bf 2 43 60 52 2b 45 59 30 56 52 54 60 41 55 641 Le e Select SPC XL Analysis Tools F Test Matrix Std Dev from Excel s menu If you don t see the SPC XL menu item you need to start SPC XL e The first window will ensure you have the correct data selected Ensure that the range of cells is correct and press Ok F Test Matrix F Test Standard Deviation Please enter the range for the test The data should be arranged such that each sample is in a column You must select two or more columns 3A51 58528 i e Next you need to tell SPC XL that the data is in Columns by clicking on the button Data in Columns 25 Data Orientation Please select the orientation of your data Example of data in Columns Column header must be non numeric J lt lt Data in Columns 4 Question A Question B 4 4 Example of data in Rows Row headers must be non numeric Question A 4 2 34 Data in Rows gt gt Question B 445 4 e SPC XL calcula
44. te at least 25 data points and export the data to Excel for analysis Create a Box Plot of the resulting data Which design requires the most force for actuation on average Which design requires the least force of actuation on average e Which design has the most variation Which design has the least variation Show me how Problem 2 Customer surveys indicate that customers prefer the feel of a phone if the force required is between 1 and 2 newtons Based upon the box plot which phone appears to meet this customer specification the best 20 Problem 3 Create a Cpk analysis and confirm which phone design has the best process capability Show Me How Basic Stats t Test Show me how This show me how requires SPC XL which is a separate product from SimWare Pro For more information click here e After opening SimWare generate 25 data points by selecting Start Generate 25 from the main menu e Export the data to Excel by selecting Export Data Excel from the main menu e Select the columns entitled northern plant and western plant A Crude Pricinorthern_plant western plant 20 dagr 5063 20 8032 T921 20 7636 T939 20 8053 T983 20 8043 5066 20 6100 13919 20 8061 T9850 20 6202 1921 20 7939 1399 20 7658 6216 20 fore 8033 20 7928 T928 20 8043 7944 20 1982 Faro 20 7939 F964 20 8015 5060 20 8043 T994 20 fei 13923 20 60505 cre 20 766 8057 20 7666 852 20 fora 6053 20 6033 80
45. tes the results and give you the p value F Test Analysis Std Dev You have 1 pvalue 100 confidence that the variances are not equal In this case since the p value is 000931 you are over 99 9 confident that the variances are not equal Further items to explore Would the answer change if you only generated 10 data points Why would fewer data points change the results Did the project improve the process Correlation Show me how e Start the simulation by selecting Start Speed Medium from the main menu e While the simulation is running change the inputs to random values Make sure you go to the extreme minimum and maximum values for the inputs while you are changing the inputs Make the changes random don t change all of the inputs to the high values and then return all of them to the low Move some up and others down while the simulation is running 26 d T u T l l l l l l aE Limas T 3 b o f T mp l l d E Cul Sarea Mapa I L e After you have generated about 100 data points stop the simulation by selecting Start Speed Stop from the main menu e Export the results to Excel by selecting Export Excel from the main menu e Select all the data including the column titles Cust Senice Reps S time H gt zl ip e un ors Line 2 T 12 6 amp 4 2 T 12 65 2 T 12 64 2 T 12 66 2 T 12 66 2 T 12 65 2 T 12 65 2 T 12 64 2 ri 12 66 3 10 13 49 a J ar 13
46. with the addition of inputs which allow you to experiment with various factors The following inputs are avallable that can be changed e Process Mean e Process Standard Deviation e Operator 1 Accuracy e Operator 1 Repeatability 38 e Operator 2 Accuracy e Operator 2 Repeatability The process mean and standard deviation control the distribution of the Reference Measurement The accuracy of the measurement represents the difference an operator will have on average from the reference measurement For example if you set the accuracy to O the operator will be on average the same as the reference value An accuracy of 1 5 would make the operator be on average 1 5 units greater than the reference value The operator repeatability directly controls the repeatability standard deviation for each operator Exercises Using the provided controls change the values of the accuracy and repeatability and perform a MSA Show Me How MSA 39 MSA Show me how The following steps are for the Micrometer simulation but the same steps can be followed for the Visual Clarity or the Compressor simulations Generate 10 parts by pressing Start Generate 1 ten times or by starting the simulator and stopping it after 10 parts have been generated n opi repi opl repa opi rei 9 3 9 23 Le 9 09 H 9 88 9 02 10 67 10 19 9 83 10 07 10 84 11 38 10 56 1121 10 94 10 79 11 43 10 94 11 75 10 65 11 09 g2 11 58 11 01 9 43 9 8
47. y creating a Cpk diagram and running 100 simulations Operator 1 doesn t have sufficient discrimination Micrometer The primary contributor to variation is repeatability with reproducibility playing a small roll also Visual Clarity Reproducibility accounts for more than 90 of the total variation Operator 2 consistently measures less than Operator 1 DOE Single Output Simulations Airline Simulation 14 Minutes Late 30 28 Num Employees The standard deviation decreases when NumuJets increases Commercial Profit Profit 1000 250 Advert 100 Training The standard deviation decreases when Percent Stocked increases Space Ship Thrust Thrust 500 266 Muon Flux 134 Plasma The standard deviation decreases with Test Pilot 2 Temperature Control Temp 500 132 lonization Rate 67 NaCl2 22 lonization Rate Oxygen Ratio 34 NaCl2 MgBer The standard deviation decreases when the Nitrogen Level decreases Hydraulic Hammer Error 100 36 Fluid Type 41 Hyd Pressure 21 Hammer Friction Surface Temp The standard deviation decreases when Rivet Density decreases The Hammer Friction Surface Temp interaction is difficult to pick out in a screening design which is common if an interaction is significant but neither of the main effects in the interaction is significant If you set Rivet Density at the low and then run a 2 with the remaining factors the interaction will be significant in the DOE Blood Analysis Optical Density
Download Pdf Manuals
Related Search