Sunteți pe pagina 1din 33

Establishing Process Capability

The statistical way

Process Evaluation
How good is the process Is the process capable of meeting the requirements How to evaluate any process

Statistical Method

Variation
is the greatest enemy of any process More than two third of the productive time is spent to counter it

Statistical Method
Extent of Variation in any process

Exhibits the Process Performance or

Process Capability
Lesser the variation better is the capability

Statistical Method
Understanding & Measuring Variation

6 Total Process Variation = 6


where : Standard Deviation

Statistical Method

Process Capability = (Inherent variability)

Measuring Process Capability Histogram Method Normal Probability Plotting Method Range Method

Measuring Process Capability


Histogram Method
1. Ensure that the process is in a state of control (free from assignable causes) 2. Collect a sample of 100 observations from the process on any CTQ. 3. Construct a histogram 4. Check whether the shape appears close to a bell shape (normal) 5. If yes, then compute the standard deviation () 6. If no, institute process control and recollect the sample and draw a histogram 7. Compute process capability as 6 x

Histogram
CTQ Selectivity
15 14 13 12 11 10 9 Frequency 8 7 6 5 4 3 2 1 0 < = .3 2 ( .3 4 ,.3 6 ] (.3 8 ,.4 ] ( .4 2 ,.4 4 ] ( .4 6 ,.4 8 ] (.5 ,.5 2 ] (.5 4 ,.5 6 ] > .5 8 ( .3 2 ,.3 4 ] (.3 6 ,.3 8 ] ( .4 ,.4 2 ] (.4 4 ,.4 6 ] (.4 8 ,.5 ] (.5 2 ,.5 4 ] (.5 6 ,.5 8 ] C la s s in t e r v a l S e le c t iv it y

Mean = 0.434 SD()= 0.047

Process Capability 6 * = 6 * 0.047 = 0.282

Measuring Process Capability

Normal Prob. Plotting Method


1. Ensure that the process is in a state of control (free from assignable causes) 2. Collect a sample (N) of 25 30 observations from the process on any CTQ (Xi) 3. Put them (Xis) in ascending order (j) and compute the corresponding prob. Plotting points as Pj = (I0.5)/N * 100 4. Plot the Xj and Pj values on a normal probability paper Continued

Measuring Process Capability

Normal Prob. Plotting Method


4. Check whether all the plotted points excepting one or two in the beginning and end fall in to an approximate straight line 6. If yes, then draw a line and compute as (X90 X10)/2.56 7. If no, institute process control and repeat the steps. 8. Compute process capability as 6 x

Normal Probability Plot


CTQ : Density

99
Mean: 40.0870 2.20500

95 90 80 70 60 50 40 30 20 10 5

StDev:

Percent

PC = 6 * 2.205 = 13.23

1 35 40 45

Density

Measuring Process Capability

Range Method
1. Ensure that the process is in a state of control (free from assignable causes) 2. Collect 20 30 sets (sub groups) each with a sample (n) of 1 to 5 observations from the process on any CTQ at equal intervals 3. Construct a suitable control chart. (X MR or X-bar & R) 4. Check whether the Range chart (MR or R-chart) is in a state of control.
Continued

Measuring Process Capability

Range Method
5. If yes, then compute as MR-bar / d2 or R-bar / d2 where MR-bar or R-bar = average range and d2 is a tabulated value for a given sample size (n) 6. If no, institute process control and repeat the steps. 7. Compute process capability as 6 x

Process Capability Analysis


Range Method (X-bar & R Chart) CTQ - Temperature
405 3.0SL=401.5

Sam ple M ean

395 385 X=380.2 375 365 355 -3.0SL=358.8 0 5 10 15

Subgroup

= R/d2 = 20.87/1.693 = 12.327


60 50

PC = 6 * 12.327 = 73.96
3.0SL=53.72

Sam ple R ange

40 30 20 10 0 -3.0SL=0.00E+00 R=20.87

How to use PC
Process Capability (Inherent process variation) VS Tolerance (design specified)

Process Capability VS Tolerance Process Capability Indices


Short Term Cp = Potential Capability Cpk = Achieved Capability (CPU & CPL) Long Term Pp = Potential Capability Ppk = Achieved Capability (PPU & PPL)

Common Measures for Establishing Process Capability of Performance Variables

How to use PC

Variation

Tolerance

How to use PC

6 < T
Variation (6) Tolerance (T)

How to use PC

6 = T
Variation (6) Tolerance (T)

How to use PC

6 > T

Variation (6) Tolerance (T)

Process capability Indices

(USL - LSL) Cp = 6 or T 6
Where USL Upper Specification Limit LSL Lower Specification Limit T - Tolerance - Standard Deviation(Estimated)

Potential Capability (Cp)

Process capability Indices


Potential Capability (Cp) Achieved (Cp)

Cp = (USL - LSL) 6 or T 6

(USL - X) CPU = 3 (X - LSL) CPL = 3 C = Min CPU, CPL pk


PC VS T
Process Capability = Tolerance or

6
Tolerance(T) LSL USL

PC = T or Cp = 1 Cpk = 1

PC VS T

Process Capability < Tolerance


1 1

6
Tolerance(T) LSL USL

6 < T

or

T=8

Cp = 1.33 Cpk = 1.33

PC VS T
Process Capability < Tolerance 6 < T or T=8

6
Tolerance(T) LSL USL

Cp = 1.33 Process off centered Cpk < 1.33

PC VS T

Total Process Variation < Tolerance


2

6
Tolerance(T) LSL

6 < T

or

T=10

Cp = 1.67
USL

Cpk = 1.67

PC VS T

Total Process Variation < Tolerance

6
Tolerance(T) LSL USL

6 < T

or

T=10

Cp = 1.67 Process off centered

PC VS T

Total Process Variation < Tolerance


3

6
Tolerance(T) LSL

6 < T

or

T=12

Cp = 2.00
USL

Cpk = 2.00

PC VS T

Total Process Variation < Tolerance 6 < T or T=12

6
Tolerance(T) LSL USL

Cp = 2.00 Process off centered Cpk < 1.00

Histogram
CTQ Selectivity
C T Q :S e le c tiv ity M e a n : 0 .4 3 4 S ig m a :0 .0 4 7 S p e c ific a tio n s : L S L = 0 .3 5 N o m in a l= 0 .5 0 U S L = 0 . 6 5 P r o c e s s C a p a b ility In d ic e s C p = 1 . 0 7 C p k = 0 .6 4 LSL 40 35 30 25 20 15 10 5 0 0 .2 5 N O M IN A L USL

Frequency

0 .3 0

0 .3 5

0 .4 0

0 .4 5

0 .5 0

0 .5 5

0 .6 0

0 .6 5

Process Capability Analysis


CTQ - Density Specification - 40.5 7.695 (19%)
Individual and MR Chart
3.0SL=47.46

Capability Histogram

Individual Value

45 40 35 -3.0SL=32.71 30 Obser. 0 9 10 20 3.0SL=9.059


35 40 45

X=40.09

Normal Prob Plot

M ov.R ange

6 3 0 R=2.773 -3.0SL=0.00E+00 35 40 45

Last 23 Observations
43.5 41.0 38.5 36.0 0 10 20

Values

Potential (ST) StDev: 2.45809 Cp: 1.04 Cpk: 0.99 Overall (LT) StDev: 2.28031 Pp: 1.12 Ppk: 1.06

Capability Plot
Process Tolerance
I I I I I I I I I ST LT

32.805

48.195

Specifications

Observation Number

Process Capability Analysis


CTQ - Temperature
425 400 375 350 Obser. 0 45 30 15 0 R=13.55 -3.0SL=0.00E+00 350 375 400 10 20 30 40 50 3.0SL=44.28 X=379.7

Individual and MR Chart


3.0SL=415.8

Capability Histogram

Indiv idual Value

-3.0SL=343.7 350 375 400

Normal Prob Plot

Mov .R ange

Last 25 Observations
395 380 365 350 30 40 50

Values

Potential (ST) StDev : 12.0152 Cp: 1.00 Cpk: 0.45 Ov erall (LT) StDev : 15.0578 Pp: 0.80 Ppk: 0.36
I

Capability Plot
Process Tolerance
I I I I I I I I ST LT

324

396

Specif ications

Observ ation Number

S-ar putea să vă placă și