Arab Academy for Science, Technology & Maritime Transport Lecture Four Control Charts for Variables (continue ) جميع حقوق الطبع والنشر محفوظة لألكاديمية العربية للعلوم والتكنولوجيا والنقل البحري What if the Process is Out of Control If points plot out of control, then the control limits must be revised. Before revising, identify out of control points and look for assignable causes. If assignable causes can be found, then discard the point(s) and recalculate the control limits. If no assignable causes can be found then * 1) either discard the point(s) as if an assignable cause had been found or * 2) retain the point(s) considering the trial control limits as appropriate for current control. Introduction to SQC Productivity and Quality Institute S-2 Revised Control Limits for X and R Revised center line for x and R control charts xnew x xd m md Rnew R Rd m md where xd discarded subgroup means md number of discarded subgroups Rd discarded subgroup ranges Introduction to SQC Productivity and Quality Institute S-3 Revised Control Limits (Cont.) xo xnew Ro R new R σo o d2 Where d2 is a factor from the table. Introduction to SQC Productivity and Quality Institute S-4 Revised Control Limits (Cont.) Revised x chart UCLx xo A o Revised R Chart UCLR D2 o LCLx xo A o LCLR D1 o Where A, D1 and D2 are factors from the table corresponding to subgroup size n. Arab Academy for Science and Technology 5 Example 1 The following data represent the depth of keyway taken in millimeters. We wish to establish statistical control of the process using x and R charts. Twenty five samples, each of size 4, have been taken when we think the process is in control. The depth of keyway measurements data from these samples are shown in the next table. Introduction to SQC Productivity and Quality Institute S-6 Subgroup Data for Depth Sample Number Date Time 1 23/12 Measurements Average Range x1 x2 x3 x4 8:50 6.35 6.40 6.32 6.37 6.36 0.08 2 11:30 6.46 6.37 6.36 6.41 6.40 0.01 3 1:45 6.34 6.40 6.34 6.36 6.36 0.06 4 3:45 6.69 6.64 6.68 6.59 6.65 0.10 5 4:20 6.38 6.34 6.44 6.40 6.39 0.10 8:35 6.42 6.41 6.43 6.34 6.40 0.09 7 9:00 6.44 6.41 6.41 6.46 6.43 0.05 8 9:40 6.33 6.41 6.38 6.36 6.37 0.08 9 1:30 6.48 6.44 6.47 6.45 6.46 0.04 24 2:00 6.43 6.43 6.35 6.38 6.38 0.08 25 4:25 6.39 6.39 6.43 6.44 6.41 0.06 Sum 160.25 2.19 6 27/12 Ri Comment New, temporary operator … Arab Academy for Science and Technology 7 Trial Control Limits x •Also, the chart is out of control, and need revision. •R chart has to be revised first, and then chart x Xbar-R Chart for depth of keyway 1 Sample M ean 6.6 1 6.5 U C L=6.4737 __ X=6.4099 6.4 LC L=6.3461 1 6.3 1 3 5 7 9 11 13 Sample 15 17 19 21 23 25 1 0.3 Sample Range •Since the R chart indicates that the process variability is out of control, we may revise the chart. 0.2 U C L=0.1998 0.1 _ R=0.0876 0.0 LC L=0 1 3 5 7 9 11 13 Sample 15 Arab Academy for Science and Technology 17 19 21 23 25 8 Identifying Out of Control Points Test Results for Xbar Chart of X1, ..., X4 TEST 1. One point more than 3.00 standard deviations from center line. Test Failed at points: 4, 16, 20 Sample 4 : Assignable cause; New temporary operator Sample 16: Random Cause Sample 20: Assignable cause; Bad material Test Results for R Chart of X1, ..., X4 TEST 1. One point more than 3.00 standard deviations from center line. Test Failed at points: 18 Sample 18: Assignable cause; Damage oil line Introduction to SQC Productivity and Quality Institute S-9 Revised R Chart The revised center line is Rnew RR 2.19 0.30 0.0788 d m md 25 1 R0 Rnew 0.0788 R0 0.0788 0 0.03827 d2 2.059 The value of d2, for sample of size n = 4 and using the table equal d2 =2.059 Arab Academy for Science and Technology 10 Revised Control Limits for R Chart Therefore, the Revised control limits for the R chart are UCLR D2 0 (4.698)(0.03827) 0.1797 LCLR D1 0 (0)(0.038) 0 Introduction to SQC Productivity and Quality Institute S-11 Revised Control Chart for R R Chart for depth of keyway •We can revise the x chart, to control the process center. 0.25 Sample Range •Now, the R chart is in control at the stated level. i.e the variability of the process is within control. 1 0.30 0.20 UCL=0.1797 0.15 0.10 _ R=0.0788 0.05 0.00 LCL=0 1 Introduction to SQC 3 5 7 9 11 13 15 Sample Productivity and Quality Institute 17 19 21 23 25 S-12 Revised Xbar Chart The revised center line is xnew x x 160.25 (6.65 6.51) 6.3951 d m md 25 2 x0 xnew 6.3951 The value of d2, for sample of size n = 4 and using the table equal d2 =2.059 Arab Academy for Science and Technology 13 Revised Control Limits for Chart X Therefore, the Revised control limits for the x chart are UCLx x0 A 0 6.395 (1.5 0.03827) 6.4525 LCLx x0 A 0 6.395 (1.5 0.03827) 6.3377 Introduction to SQC Productivity and Quality Institute S-14 Revised Control Chart Xfor •Note that subgroup 9 is out of control. The cause responsible for this signal should be treated as random cause. Introduction to SQC Xbar Chart for depth of keyway 1 6.65 6.60 6.55 Sample Mean •The x chart is now in control at the stated level; center line and control limits. 1 6.50 1 6.45 UCL=6.4525 6.40 _ _ X=6.3951 6.35 LCL=6.3377 6.30 1 3 5 7 9 11 13 15 Sample Productivity and Quality Institute 17 19 21 23 25 S-15 Revised Control Charts Since both the revised x and R charts exhibit control, we would conclude that the process is in control at the stated level. Adopt the Revised control limits for use in on-line statistical process control. The capability analysis can now be conducted if the specifications limits on the depth of keyway is known. Introduction to SQC Productivity and Quality Institute S-16 Example 2 Estimate the mean depth of the keyway data? Estimate the process standard deviation? Introduction to SQC Productivity and Quality Institute S-17 Solution We may estimate the mean depth of the keyway data as: ˆ x 6.3951 mm The process standard deviation may be estimated using the following equation: ˆ 0 R0 0.03827 d2 Where the value of d2 for samples of size 4 is found from the table. Introduction to SQC Productivity and Quality Institute S-18 Example 3 The depth of keyway is normally distributed with mean 6.3951 mm and standard deviation 0.03827. The specification limits on the depth are 6.40 ± 0.06 mm. Estimate the fraction of nonconforming keyway. Also, calculate the capability of the process. Introduction to SQC Productivity and Quality Institute S-19 Solution p P( x LSL) P( x USL) P( x 6.34) P( x 6.46) 6.34 6.3951 x 6.46 6.3951 ) P( ) 0.03827 0.03827 P( z 1.44) P( z 1.70) P( x 0.0749 0.0446 0.1195 11.95% That is, about 11.95% [119500 parts per million (ppm)] of the piston rings produced will be outside the specifications. Introduction to SQC Productivity and Quality Institute S-20 Solution (Cont.) Note that 6 spread of the process is the basic definition of process capability. Since is usually unknown, we must replace it with an estimate. . We frequently use σ̂ as an estimate of , resulting in an estimate C p of Cp USL LSL Cp 6 6.46 6.34 0.12 0.52 6(0.03827) 0.22962 Introduction to SQC Productivity and Quality Institute S-21 Capability Using Minitab Process Capability of X1, ..., X4 LSL Target USL Within Ov erall P rocess Data LS L 6.34 Target 6.4 USL 6.46 S ample M ean 6.3951 S ample N 100 S tDev (Within) 0.03827 S tDev (O v erall) 0.0733946 P otential (Within) C apability Cp 0.52 C P L 0.48 C P U 0.57 C pk 0.48 O v erall C apability Pp PPL PPU P pk C pm 6.24 O bserv ed P erformance P P M < LS L 60000.00 P P M > U S L 130000.00 P P M Total 190000.00 Introduction to SQC 6.32 6.40 E xp. Within P erformance P P M < LS L 74966.23 PPM > USL 44957.59 P P M Total 119923.82 6.48 6.56 0.27 0.25 0.29 0.25 0.28 6.64 E xp. O v erall P erformance P P M < LS L 226405.56 P P M > U S L 188277.51 P P M Total 414683.07 Productivity and Quality Institute S-22 Solution (Cont.) This implies that the “natural” tolerance limits in the process (three-sigma above and below the mean) are well outside the lower and upper specification limits. Consequently, a very large number of nonconforming keyways will be produced. Introduction to SQC Productivity and Quality Institute S-23 Guidelines for the Design of X and R Specify sample size, control limit width, and frequency of sampling. if the main purpose of the x chart is to detect moderate to large process shifts, then small sample sizes are sufficient (n = 4, 5, or 6). if the main purpose of the x chart is to detect small process shifts, larger sample sizes are needed (as much as 15 to 25)…which is often impractical…alternative types of control charts are available for this situation… Introduction to SQC Productivity and Quality Institute S-24 Guidelines for the Design of X and R (Cont. ) If increasing the sample size is not an option, then sensitizing procedures (such as warning limits) can be used to detect small shifts…but this can result in increased false alarms. R chart is insensitive to shifts in process standard deviation. (the range method becomes less effective as the sample size increases) may want to use S chart. Introduction to SQC Productivity and Quality Institute S-25 Allocating Sample Effort Choose a larger sample size and sample less frequently? or, Choose a smaller sample size and sample more frequently? The method to use will depend on the situation. In general, small frequent samples are more desirable. Introduction to SQC Productivity and Quality Institute S-26 Charts Based on Standard Values If the process mean and variance are known or can be specified, then control limits can be developed using these values: x chart R chart UCL μ Aσ UCL D2σ CL μ CL d 2σ LCL μ Aσ LCL D1σ The constant A, d2, D1, and D2 are given in the table for various values of n Introduction to SQC Productivity and Quality Institute S-27 Control Charts for X and S First, S2 is an “unbiased” estimator of 2 Second, S is NOT an unbiased estimator of S is an unbiased estimator of c4 where c4 is a constant. The standard deviation of S is Introduction to SQC Productivity and Quality Institute σ 1 c42 S-28 Construction and Operation of X and S Control Chart If a standard is given the control limits for the x and S charts are: UCLx x A UCLS B6 CL x CL c4 LCL x A LCLS B5 B5, B6, A, and c4 are found in the table for various values of n. Introduction to SQC Productivity and Quality Institute S-29 Construction and Operation of X and S Control Chart If no standard is given (i.e is unknown), we can use an average sample standard deviation,1 m S Si m i 1 The upper and lower control limits of the S chart are UCLs B4 S CL S LCLs B3 S Introduction to SQC Productivity and Quality Institute S-30 Construction and Operation of X and S Control Chart The upper and lower control limits for the are given as chart x UCLx x A3 S CL x LCLx x A3 S where A3, B4, and B3 are found in the table. Introduction to SQC Productivity and Quality Institute S-31 Estimating Process Standard Deviation The process standard deviation, can be estimated by S σˆ c4 Where c4 is a constant found in the table for various values of n Introduction to SQC Productivity and Quality Institute S-32 Example : Piston Ring Data Arab Academy for Science and Technology 33 Solution The grand average and the average standard deviations are 25 x i 1850.028 x 74.001 mm 25 25 and i 1 25 S S i 1 Introduction to SQC 25 i 0.2350 0.0094 25 Productivity and Quality Institute S-34 Solution (Cont.) The trial control limits for the x chart is UCLx x A3 S 74.001 (1.427)(0.0094) 74.014 CL x 74.001 LCLx x A3 S 74.001 (1.427)(0.0094) 73.988 And for the S chart is UCLS B4 S (2.089)(0.0094) 0.0196 CL S 0.0094 LCLS B3 S (0)(0.0094) 0 Introduction to SQC Productivity and Quality Institute S-35 Plot the Trial Control Limits Xbar-S Chart for Piston Ring Data U C L=74.01459 Sample M ean 74.01 _ _ X=74.00118 74.00 73.99 LC L=73.98776 1 3 5 7 9 11 13 Sample 15 17 19 21 23 25 Sample StDev 0.020 U C L=0.01964 0.015 _ S =0.00940 0.010 0.005 0.000 LC L=0 1 3 Introduction to SQC 5 7 9 11 13 Sample 15 17 19 21 Productivity and Quality Institute 23 25 S-36