123450.0000.0000.0000.0000.00011σ#DIV/0!22σ#DIV/0!33σ#DIV/0!44σ#DIV/0!55σ#DIV/0!66σ#DIV/0!==結論﹕#DIV/0!#DIV/0!#DIV/0!#DIV/0!#DIV/0!線性良好﹐偏倚小0.000.000.000.000.00#DIV/0!#DIV/0!#DIV/0!#DIV/0!#DIV/0!線性可以﹐有偏倚0.000.000.000.000.00基准值X=0.000.000.000.000.00線性不良﹐偏倚大偏倚值Y=#DIV/0!#DIV/0!#DIV/0!#DIV/0!#DIV/0!對策﹕a=#DIV/0!b=#DIV/0!Y=b+ax→分別計算出『每一個基準值〈X〉』所對應的『組間的偏倚〈Y〉』標示於線性圖中Y1=######+######=######Y2=######+######=######Y3=######+######=######Y4=######+######=######Y5=######+######=######X00.0000.0000.0000.0000.000Y#DIV/0!#DIV/0!#DIV/0!#DIV/0!#DIV/0!#DIV/0!a:slopeb:interceptPARTAVG.零件均值VALUEBIAS偏倚RANGE極差9擬合優度R210%線性11123456781過程變差值斜率2公差上偏差料號下偏差VALUE基准值數據分析表線性和偏倚分析表LinearityExamplePART零件標號客戶尺寸品名0.30000.40000.50000.60000.70000.80000.90001.0000制表﹕審核﹕(藍色中有公式勿動﹐在米黃色中輸入數據即可)0.00000.10000.20000.3000000001111#DIV/0!#DIV/0!#DIV/0!#DIV/0!#DIV/0!#DIV/0!#DIV/0!#DIV/0!#DIV/0!#DIV/0!#DIV/0!#DIV/0!#DIV/0!線性良好﹐偏倚小線性可以﹐有偏倚線性不良﹐偏倚大→分別計算出『每一個基準值〈X〉』所對應的『組間的偏倚〈Y〉』標示於線性圖中線性值%線性數據分析表線性和偏倚分析表LinearityExample11日期:品名名稱料號編號規格分辯力上下限公差測量員測量編號次數1234567891019.2339.2409.2339.2369.2329.2409.2429.2409.2459.23629.2349.2409.2349.2369.2329.2369.2439.2439.2469.23739.2359.2379.2339.2309.2359.2379.2469.2409.2479.236平均值9.2349.2399.2339.2349.2339.2389.2449.2419.2469.236X1=9.2378全距0.0020.0030.0010.0060.0030.0040.0040.0030.0020.001R1=0.002919.2349.2479.2419.2349.2409.2389.2469.2469.2449.24129.2369.2439.2389.2359.2359.2379.2439.2419.2479.23739.2359.2459.2409.2379.2419.2429.2439.2419.2459.241平均值9.2359.2459.2409.2359.2399.2399.2449.2439.2459.240X2=9.2404全距0.0020.0040.0030.0030.0060.0050.0030.0050.0030.004R2=0.003819.2359.2429.2349.2349.2329.2409.2429.2439.2479.23929.2379.2449.2389.2379.2369.2369.2449.2449.2499.24139.2389.2419.2389.2369.2359.2359.2469.2439.2459.239平均值9.2379.2429.2379.2369.2349.2379.2449.2439.2479.240X3=9.2397全距0.0030.0030.0040.0030.0040.0050.0040.0010.0040.002R3=0.00339.2359.2429.2379.2359.2359.2389.2449.2429.2469.239X=9.2393Rp=零件平均值之最大值-最小值=9.246─9.235=0.0111XDIFF=操作員平均值之最大值-最小值=9.240─9.238=0.0026R=﹝R1+R2+R3)/3=X=﹝X1+X2+X3)/3=中心線CLR=R=管制上限UCLR=R*D4=0.003*2.575=管制下限LCLR=R*D3=0.003*0=中心線CLX=X=管制上限UCLX=X+A2*R=9.239+1.023*0.003=9.243管制下限LCLX=X-A2*R=9.239─1.023*0.003=9.236以上數據在X-r管制范圍內﹐可以進行GR&R分析測量次數以上數據不在X-r管制范圍內﹐說明制程不穩定﹐產品變異大tA2D3D4不可以進行GR&R分析﹐應解決制程問題后再重新收集數據21.88003.267以上數據不在X-r管制范圍內但可以排除﹐應先須完31.02302.575成偏倚和線性分析或進行穩定性(AV)分析40.72902.282以上數據研究零件變異﹐如不大可不計TV計算GR&R50.57702.115以上數據過于集中在中心線有不真實數據﹐應重新收集數據以上數據說明分辯率不足﹐應用更精密分辯力高的量測工具黃色的和藍色的框不要亂修訂﹐因里面有公式。零件編號平均值GageR&R數據資料表(B)零件全距0.00860.0000GageR&R數據資料表(A)//受測零件量具測量員3.單位/姓名:1.單位/姓名:2.單位/姓名:零件平均值AB操作員全距全距總平均0.0033數據總平均9.2393C管制圖常數平均值管制圖9.239以上數據結論﹕全距管制圖0.0033X值分析9.2209.2259.2309.2359.2409.2459.2509.25512345678910X值分析0.0000.0010.0020.0030.0040.0050.0060.00712345678910R值分析日期://由數據資料表得:R=0.0033XDIFF=0.0026Rp=0.0111允差=0.14mm操作員人數(m)=3零件數(n)=10量測次數(t)=3重複性(EV)量測次數K1%EV=100(EV/TV)EV=R*K124.56=100(0.0100.022)=0.003*3.0433.04=####%=0.01042.50%EV=100(EV/Tolerance)52.21=100(0.0100.14)=####%再現性(AV)操作員人數K2%AV=100(AV/TV)AV=√〔(XDIFFxK2)2-(EV2/nt)〕23.65=100(0.0070.022)=√〔(0.0026*2.70)2-(######2/10*3)﹞32.70=####%=0.00742.30%AV=100(AV/Tolerance)52.08=100(0.0070.14)=4.90%重複性與再現性(R&R)R&R=%R&R=100(R&R/TV)=√(0.0102+######2)=100(0.0120.02)=0.012=####%%R&R=100(R&R/Tolerance)=100(0.0120.14)=####%零件變異(PV)零件數K3PV=Rp*K323.65%PV=100(PV/TV)=0.011*1.6232.70=100(0.0180.022)=0.01842.30=####%52.0861.93全變異(TV)71.82TV=√(R&R)2+PV281.74=√(0.0122+0.0182)91.67=0.022101.62數據分級﹕不計TV值﹐計算GR&R值﹐用公差值代替1.41*PV/GR&R只能通過EV和AV情況﹐了解測量系統2.074≒2重新收集數據﹐做線性偏倚分析5個分級此數據不能用于分析過程狀況良好(%R&R10%)5個分級此數據可用于分析過程,計算TV值尚好,待改進(10%=%R&R30%)結論:%R&R=%必須改進(%R&R30%)核准:審核:填表:量測系統分析建議GaugeR&R報告%全變異(TV)量測單元分析√(EV2+AV2)