• Nem Talált Eredményt

PROCESS CAPABILITY PROCESS CAPABILITY

N/A
N/A
Protected

Academic year: 2022

Ossza meg "PROCESS CAPABILITY PROCESS CAPABILITY"

Copied!
7
0
0

Teljes szövegt

(1)

Six Process Capability 1

PROCESS CAPABILITY PROCESS CAPABILITY

Role of Quality Engineering Role of Quality Engineering

IN CONTROL?

yes no

CAPABLE?

yes

no

upper natural tolerance limit

lower natural tolerance

upper specification limit

lower specification limit

(2)

Six Process Capability 3

Analysis of process capability Analysis of process capability

Process capability index (Potential capability)

C U SL LSL

P

= −

6 σ

− =

= LSLσ µ zlower

− =

= σ

µ

zupper USL P(x>USL)=

=

< ) (x LSL P

Example 1

In a manufacturing process the expected value of a quality characteristic is 250.727 unit, the standard deviation is 1.286 unit. The specification is 250.5 unit.

How much is the proportion of defectives in this process?

Calculate the CPcapability index!

(3)

Six Process Capability 5 LSL USL

LNTL UNTL

CP=

µ T

µ+3σ µ-3σ

0.00135 0.00135

σ 6

LSL CP =USL

LSL USL

UNTL LNTL

CP

µ T

µ+3σ µ-3σ

σ 6

LSL CP =USL

(4)

Six Process Capability 7 LSL USL

UNTL LNTL

CP

µ T µ-3σ µ+3σ

σ 6

LSL CP =USL

Corrected indices (demonstrated capability) (

PU PL

)

PK PL

PU LSL C C C

USL C

C ; min ,

; 3

3 − = − =

= σ

µ σ

µ

LNTL UNTL

CP= CPU, CPL

CPK

µ

µ+3σ µ-3σ

σ 6

LSL CP =USL

(5)

Six Process Capability 9

(

PU PL

)

PK PL

PU LSL C C C

USL C

C ; min ,

; 3

3 − = − =

= σ

µ σ

µ

LSL USL LNTL UNTL

CP=

CPU, CPL CPK

µ

T

µ+3σ µ-3σ

σ 6

LSL CP =USL

Modified process capability index Modified process capability index

capability index modified capability index

σ 6

LSL CP =USL

( )

[

2

]

=τ2

=E x T

MSE τ2 =σ2+

(

µT

)

2

( )

2

6 2

6 T

LSL USL LSL

CPm USL

− +

= −

= −

µ τ σ

related to Taguchi’s quadratic loss function

(6)

Six Process Capability 11

Example 2

Compare two processes, the specification for both is 100±1.

I. s =0.2, m=99.5, that is the center of fluctuation deviates from the nominal value

II. s =0.4, m =100, that is the center of fluctuation is the nominal value, but the fluctuation is larger

Example 3

The specification is 100±1, s =0.2. Calculate the capability indices and the proportion beyond specs (above USL or below LSL), if m is 100, 99.5 and 100.5!

Process capability and process performance (short term and long term)

Process capability and process performance (short term and long term)

σ 6

LSL CP =USL

which s?

s s

(7)

Six Process Capability 13

Estimating variance from the within-samples (short term) changes refers the internal, random fluctuation CP (potential capability)

Combining both within-samples and between-samples changes the long term fluctuation is considered PP (process performance)

P

P C

P

The process capability study is to be interpreted for in-control processes only.

Two parts of the task:

1. Stabilize the process for an acceptable time span, eliminating potential sources of fluctuation (e.g. operator, lot of raw material) 2. Compare the long term process performance with that expected How to check stability?

Hivatkozások

KAPCSOLÓDÓ DOKUMENTUMOK

A special case of stratified samples is considered where each stratum has the same number of units and from each stratum, one unit is selected in the sample with simple ran-

The low friction of treated samples under “run-out” lubrication conditions can be attributed to enhance the oil retention capability in the sliding interface due to the

The expected value and variance of the function of random variables8. The data in the samples are the results

The crystallinity changes and recrystallization process of amorphous samples were investigated by hot-humidity stage X-ray powder diffractometry (HH-XRPD) with fresh

Regarding the problem of estimating driver’s driving capability based on the measured vital parameters and the developed sensor network, we will analyze drivers’ stressors.

First we wanted to compare the samples aged with the xenon test accelerated aging process (PLA_xenontest) and the samples aged with photodegradation by a

Kruskal-Wallis test shows a significant difference between the samples (P=0.000605231; without recent samples P=0.00323383), so the collection may have targeted different

The sampling regime study included soil (8 samples) and surface water from Illatos ditch (2 samples), as well as biological samples, tree bark (Robinia