Documents

Quickref-QT4

Description
Page 1 Quick Reference Manual (Text only) For Qualitek-4, Automatic Design and Analysis of Taguchi Experiments Nutek, Inc. Bloomfield Hills, Michigan, USA. www.nutek-us.com (Last updated Jan. 2, 2007) [This brief TEXT ONLY document is intended to be used for quick reference on the use of Qualitek-4 capabilities. For detail information, use the reference manual in the form of a hard copy or the electronic format.] Content A. B. C. D. E. F. G. H. I. J. K. How To . . . . Tips Installation of Qua
Categories
Published
of 17
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
    Page 1 Quick Reference Manual (Text only) For Qualitek-4, Automatic Design and Analysis of Taguci !x eriments #utek, $nc . Bloomfield Hills, Michigan, USA. www.nutek-us.com  (Last udated !an. , ##$% & This brief TEXT ONLY document is intended to be used for quick reference on the use of Qualitek-4 capabilities. or detail information! use the reference manual in the form of a hard cop or the electronic format.#  'ontentA.How o . . . . isB.)nstallation of *ualitek-+ fo indows'.How to Stat and /it Pogam 0.n-line Alication Suot ./eiment Planning 2./eiment 0esign3.diting 2actos, Aa4s and 5esults H.Peaation of 5esults).Anal4sis of 5esults!.Pactice Session6.ou7leshooting A% &' T' $%  enu items efeed ae all in  e&periment confi'uration sceen  ( Design !x eriment * Select desi'n  menu otion to stat all 4ou e/eiment design tasks. Select automatic desi'n  when 4ou want *ualitek-+ to decide the aa4 and assign factos to the aoiate columns.'eate  'uter Array  8 Outer arra  is included in the design 4ou ha9e alead4 designed the inne aa4 (using manual desi'n %. Select outer arra  fom desi'n menu  and oceed as omted.0esign fo Dynamic +aracteristic  8 nce the inne aa4 is designed: 4ou can oceed to include the signal and noise facto in the oute aa4 74 selecting manual design 8 d4namic chaacteistic fom the design menu item.nte Results for '!+  8 o ceate ' 4ou must ente the oiginall4 collected o7se9ations fo eachciteion of e9aluation in a secial sceen. 2om the result menu item,  select multiple criteria  ('% otion. Befoe 4ou oceed with 4ou own data ent4 fo ', 4ou must make 4ouself familia with the inciles of ' fomulation (e9iew www.ko4.com;w-oec.html  and www.ko4.com;w-tec.html  %. o e9iew an e/amle PU<0.*+ included with the ogam. cale Results * o scale o tansfom esults (u, down, mega tansfomation, tuncation, etc%, select edit results  fom edit   menu item. 'lick on the transform  7utton and follow sceen omts.  $m ort Results  8 )moting esults e=uie 4ou to eae a te/t file containing the esults. 2ollow thesimle stes out lined in the tips for importin'   otion in anal sis  menu item. 1$  <utek, )nc. *ualitek-+ fo indows *uick 5efeence Manual (e/t nl4% 0ate of last udate> 11;#$;1+    Page % $nstalling Qualitek-4 in indo. /01 o install *ualitek-+ ogam, 4ou will need the ogam '0-5M and 4ou esonal fifteen digit 5egistation num7e. )f 4ou do not ha9e the coect egistation num7e, the ogam will install as a ?0emo@ 9esion allowing 4ou to use it fo designing e/eiments with L- aa4 onl4.ou will install the ogam the same wa4 as all othe indows ogam - Cust inset the '0-5M into the di9e un D>ESUP (o SUP*+% (D F '0 o flo4 di9e as alica7le%. +%&o. to tart and !xit 0rogram  ou can use the Stat 7utton of indow G to stat *ualitek-+ ogam. o stat, click the Stat  7utton, and then oint to *ualitek-+ ogam. )f 4ou ae using indow I./, then dou7le click the *ualitek-+ icon fom the indow Pogam Manage sceen. D% 'n-line A lication u ort he 7asic stes in the alication ocess ae as outlined 7elow. Bief descition and alication guidelines ae included on each of the alication stes in the *ualitek-+ on-line hel. elect 0ro2ect and lan !x eriment Look fo simle oCect to al4 fist. Aange fo the lanning;7ainstoming session. 0etemine>J 9aluation citeia and define a method to com7ine themJ 'ontol factos and thei le9els.J )nteaction (if an4%J <oise factos (if an4%J <um7e of samles to 7e tested.J /eiment logistics. !x eriment Design 0esign e/eiment K 0esci7e tial conditions.J 0etemine the ode of unning the e/eimentJ 0esci7e noise conditions fo testing samles if the design includes an oute aa4 +onduct !x eriment 'a4 out e/eimentsJ <ote eadings, calculate and ecod a9eages if multile eadings of the same citeia ae taken.J 'alculate ' using the fomula defined in the lanning session.  Analy3e Results J 0etemine facto influence (Main ffect%J )dentif4 significant factos (A<A%J 0etemine otimum condition and estimate efomanceJ 'alculate confidence inte9al of otimum efomance 1$  <utek, )nc. *ualitek-+ fo indows *uick 5efeence Manual (e/t nl4% 0ate of last udate> 11;#$;1+    Page I J AdCust design toleances 7ased on A<A +onfirm 0redicted $m roement 5un confimation tests with multile samles at the otimum condition. 'omae the a9eage efomancewith the confidence inte9al.Although *+ is designed mainl4 to hel 4ou accomlish 4ou design and anal4sis tasks on command and not to teach 4ou the method, thee is a lage 9olume of infomation a9aila7le a7out the method and its alication stategies in the reference manual  . Most of the theo4 elated items and geneal toics ha9e 7een laced unde HLP menu in the main sceen. 0iscussion toics and seach items ae desci7ed in h4ete/t so that when a click on an item it takes 4ou diectl4 into the su7Cect matte. o get some tis on what the content of the sceen mean o how to inteet the data, click on the HLP  7utton fo hel on the secific sceen. !%!x eriment 0lanning /eiment lanning is a necessa4 and imotant fist ste in the alication ocess. f couse  lanning;7ainstoming fo the oCect is done outside the ogam and 7efoe 4ou come to use the  ogam. But should 4ou need some hel with how to ca4 out the lanning session o what to discuss in the lanning session, a considea7le amount of guidelines ha9e 7een included in the main sceen HLP in *+. 5emem7e it is the fomal lanning session, which sulies answes to the  etinent =uestions 7efoe an e/eiment is designed. he main sceen HLP in *+ contains guidelines as follows>PU5PSS 2 B5A)<S5M)<3 SSS)< - Puoses of a 7ainstoming session ae man4> ã )dentif4 factos, le9els and dei9e othe etinent infomation a7out the e/eiment, collecti9el4with all in9ol9ed in the e/eiment. ã 0e9elo team effot and achie9e the ma/imum aticiation fom the team mem7es. ã 0etemine all e/eiment-elated items 74 consensus decisions.H SHUL0 '<0U' he session should 7e facilitated 74 a eson who has a good woking knowledge of the aguchi methodologies. nginees o statisticians dedicated to heling othes al4 this tool will make 7ette facilitatos.H SHUL0 HS H SSS)<he team;oCect leade should host the 7ainstoming session.H SHUL0 A<0he oCect team mem7es consisting of all those who ha9e fist-hand knowledge and;o in9ol9ement in the su7Cect unde stud4 should 7e included. 2o an engineeing design o a manufactuing ocess,  7oth the design and the manufactuing esonnel should attend. )f cost o sulie knowledge is likel4factos, then esons fom these discilines should 7e encouaged to attend (gou siNe emitting%. 1$  <utek, )nc. *ualitek-+ fo indows *uick 5efeence Manual (e/t nl4% 0ate of last udate> 11;#$;1+    Page + H MA< SHUL0 A<0he moe the 7ette. he ue limit should 7e aound 1. )t can 7e as low as . HA )S H A3<0A 25 H SSS)<he meeting agenda need not 7e shaed with all aticiants. he facilitato should t4 to comlete all discussions within the meeting time, 4et allowing enough time fo item discussions. P)'S 2 0)S'USS)<She following toics should 7e included in the agenda fo the 7ainstoming session.1. B!') 2 H SU0 (hat ae 4ou afte% ã hat is the chaacteistic of =ualit4 How do we e9aluate the o7Cecti9e ã How do we measue the =ualit4 chaacteistic hat ae the units of measuement ã hat ae the citeia (atti7utes% of e9aluation fo the =ualit4 chaacteistic ã hen thee ae moe than one citeion o thee ae se9eal atti7utes of the =ualit4 chaacteistic, ã How do we com7ine them into an 9eall 9aluation 'iteion ('% ã How is the diffeent =ualit4 citeia weighted ã hat is the sense of the =ualit4 chaacteistic ã Lowe is 7ette, nominal is the 7est, etc. . 0S)3< 2A'5S A<0 H)5 LLS ã hat ae all the ossi7le factos ã hich ones ae moe imotant than othes ae (Paeto diagam% ã How man4 factos should 7e included in the stud4 ã How to select le9els fo the factos How man4 le9els ã hat is the tade off 7etween le9els and factosI. O<)SO A5)ABLS (How to make a o7ust design% ã hat factos ae likel4 to influence the o7Cecti9e function, 7ut can not 7e ã 'ontolled in eal life. ã How can the oduct unde stud4 7e made insensiti9e to the noise factos ã How ae these factos included in the stud4+. )<5A')< SU0)S (hich factos ae likel4 to inteact% ã hich ae the factos most likel4 to inteact ã How man4 inteactions can 7e included ã Should we include an inteaction o an additional facto ã 0o we need to stud4 the inteaction at all. AS6 ASS)3<M< A<0 0S'5)P)< (ho will do what, how and when%  hat stes ae to 7e followed in com7ining all the =ualit4 citeia into an ' 1$  <utek, )nc. *ualitek-+ fo indows *uick 5efeence Manual (e/t nl4% 0ate of last udate> 11;#$;1+

ijret_110209004

Jul 23, 2017

Hello

Jul 23, 2017
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks