File : OCTEP / SVCF / AQMS /

AUTOMOTIVE QUALITY MANAGEMENT

SRI VIGNESHWARA COLD FORGE QSP
PLOT PB 6 /2, SIPCOT INDUSTRIAL PARK, SYSTEM
IRUNGATTUKOTTAI, KANCHIPURAM DT-602 STATISTICAL PROCESS CONTROL
105
File : iatf\ svcf \ aqms \ qsp

1. PURPOSE
To establish a procedure for improving statistical process controls & verifies the capability.

2. SCOPE
For the special characteristics identified (variable)

3. REFERENCES
QAD/ DEV

4. DEFINITIONS

5. PROCEDURE

5.1 FOR PRELIMINARY PROCESS CONTROL (PROCESS CAPABILITY)

5.1.1 This study is carried out as required by the APQP in the product & process validation stage.
Ensure that R & R study is carried out for the measurement system & it is acceptable before
conducting preliminary SPC study.

5.1.2 Histogram chart can be used to study the preliminary process performance, QAD shall identify the
other appropriate tools depending upon the process & get customer approval, if required.

5.1.3 The study should be based on 50 or more subgroup of data containing at least a total of
100 individual readings (short term).

5.1.4 Interpret the control chart for stability (control), if stability is not achieved contact customer and
determine appropriate action (Refer II of interpretation).

5.1.5 Interpret for capability (process performance): The 6_ range of a process total variation,
where _ is usually estimated by S, the standard deviation calculated using all of the individual
values (using calculator).

USL - LSL
Pp = ---------------------
6s

USL - X double bar

Ppk = Min ( --------------------------------- )
3 sigma

Or

X double bar - LSL

( ---------------------------------- )
3 sigma

NOTE: It should be used only to compare or with Cp and Cpk and to measure and prioritized
improvement overtime. If it is not specified by the customer, a Ppk value >1.67 should be achieved.
Calculate the Ppk index and take the following actions

DOCUMENT NO REVISION NO REVISION DATE PAGE OUT OF COPY NO

L2.1/AQMS-SPC 01 03/03/2022 Page 1 of 5 1
 File : OCTEP / SVCF / AQMS /
AUTOMOTIVE QUALITY MANAGEMENT
SRI VIGNESHWARA COLD FORGE QSP
PLOT PB 6 /2, SIPCOT INDUSTRIAL PARK, SYSTEM
IRUNGATTUKOTTAI, KANCHIPURAM DT-602 STATISTICAL PROCESS CONTROL
105
File : iatf\ svcf \ aqms \ qsp

RESULTS INTERPRETAION
1. Processes that The process probably meets customer requirements. After
appear stable Pp & approval, begin production and follow the approved
Ppk > 1.67 control plan

The process may not meet customer requirements. After

2. 1.33< Ppk <1.67 part approval begins production with additional attention to
the characteristic until ongoing Cpk > 1.33 is achieved.

Unstable Depending on the nature of the instability, the

process may not meet customer requirements. Special
causes.
should be identified, evaluated & whenever possible,
eliminated. Use 100% inspection & increase SPC
II. Process that
sampling until ongoing stability with a Cpk of 1.33 is
appear
demonstrated or until the customer is satisfied. Process
improvement must be given a high priority and
documented in a corrective action plan. A revised control
action plan for these interim actions must be reviewed with
and approved by the customer.

If process capability cannot be obtained by the part submission promise date, a corrective
action plan & an interim revised control plan must be developed. If required contractually,
these shall be sent to customer for review & approval. Typical corrective actions include
process improvement, tooling changes & changes to customers engineering
requirements.
For all the above cases, the process FMEA for that particular process shall be reviewed
again to re evaluate the possibility of implementing mistake-proofing activities.

5.2 ON GOING PROCESS CAPABILITY STUDY

5.2.1 Ensure the measurement is appropriate i.e. (least count of instrument is at least 1/5 th of
the expected process variation (or) 1/10 th of tolerance whichever is less).

5.2.2 Before collecting data, ensure that there are no obvious deficiencies like faulty gauge,
loose clamp, improper tool etc.,

5.2.3 Ensure R & R is carried out for the measurement system and it is acceptable.

5.3 DATA COLLECTION

5.3.1 Plan the data collection like how many samples, what frequencies, who will measure, the format
etc. also decide which process parameters needs to be monitored during study.
Generally up to 25 or more subgroups containing about 100 or more individual readings give a
good test of stability (for control chart method).
The frequency of subgroup should be based on the prior knowledge on repetitiveness of
the process & the rate of production. It gives a good estimation.

5.3.2 While collecting data, note down all the process events with respect to time.

5.3.3 Select suitable scale for the control chart. (Generally for X bar chart, the scale should be such that
the UCL / LCL should occupy approx.50% the middle of the chart. For R chart, the scale should be
double that of X bar chart.)

DOCUMENT NO REVISION NO REVISION DATE PAGE OUT OF COPY NO

L2.1/AQMS-SPC 01 03/03/2022 Page 2 of 5 1
 File : OCTEP / SVCF / AQMS /
AUTOMOTIVE QUALITY MANAGEMENT
SRI VIGNESHWARA COLD FORGE QSP
PLOT PB 6 /2, SIPCOT INDUSTRIAL PARK, SYSTEM
IRUNGATTUKOTTAI, KANCHIPURAM DT-602 STATISTICAL PROCESS CONTROL
105
File : iatf\ svcf \ aqms \ qsp

5.3.4 Calculate the avg. (x bar and range R of each sub-group.)

5.3.5 Plot the average and ranges on the control chart, connect the points with lines.

5.3.6 Ensure that the calculation and plots are correct.

5.3.7 For the charts UCL & LCL value not calculated shall be mentioned as ‘INITIAL STUDY’ on the
Date column.

5.4 CALCULATE CONTROL LIMITS:

R-chart

R1 + R2 + …. + Rn
R-bar = ------------------------------------ , N = no. of subgroups
N

Xbar.1 + Xbar.2 + … + Xbar.n

X-bar = ----------------------------------------------
N

Then,
UCLr = D4R bar
LCLR = D3R bar
D4, D3 are constants varying by sample size

5.4.1 Draw lines for the avgs. and the control limits on the charts.

5.5 INTERPRETATIONS FOR CONTROL:

5.5.1 Analyze the data for control in prediction check the existence of Special causes as per conditions
given under action on Special causes on Control chart. If the process is not under control (i.e.
approx. more than 20% of points fall outside limits) study the process using cause and effect
diagram prepared for that process.

5.5.2 Establish control by taking action on the process and repeat the study

5.6 HOMOGENISATION:

5.6.1 Recalculate the control limits with the excluded readings of out of control point affected by special
causes in the R chart. Calculate new avg. range (R bar) and confirm that all range points show
control when compared to new limits.

5.7 CALCULATE PROCESS CAPABILITY:

5.7.1 Once control is established, evaluate process capability Cp & Cpk.

DOCUMENT NO REVISION NO REVISION DATE PAGE OUT OF COPY NO

L2.1/AQMS-SPC 01 03/03/2022 Page 3 of 5 1
 File : OCTEP / SVCF / AQMS /
AUTOMOTIVE QUALITY MANAGEMENT
SRI VIGNESHWARA COLD FORGE QSP
PLOT PB 6 /2, SIPCOT INDUSTRIAL PARK, SYSTEM
IRUNGATTUKOTTAI, KANCHIPURAM DT-602 STATISTICAL PROCESS CONTROL
105
File : iatf\ svcf \ aqms \ qsp

R bar
SIGMA (σ) = ---------
d2

USL - LSL Tolerance

Cp = ------------------ = ------------------
6σ 6σ

If Cp is > 1.33 carry out the following steps, else follow step II.15.8

X-bar chart
Calculate UCLxbar = X-double bar + A2 R-bar
LCL x-bar = X-double bar - A2 Rbar

5.7.2 Follow the steps described II.12 to II.14 for X bar charts

USL -X-double bar X double bar - LSL

5.7.3 Calculate Cpk = Min (------------------------------) or (---------------------------------)
3σ 3σ

5.7.4 X-bar chart signals the variation in the process averages w.r.t. time, the recorded process events
like dressing, adjusting, tool change etc. should be correlated to the variation in the control chart.
X-bar charts generally describe the stability of the process, i.e. maintaining the process average
w.r.t. time.

5.7.5 If all points are within control limits establish UCL, LCL for both X-bar & R chart & use it for
ongoing control, mention the date on which control limits calculated in the ongoing control chart.

The MOST RECENT

ACTIONS ON THE PROCESS OUTPUT
POINT indicates that
Based on the Ongoing process capability (Cpk)
the process
Less than 1.33 ** 1.33 - 1.67 Greater than 1.67
Is in Control
100% inspect Accept product Continue to reduce product
variation.
Has gone out of control IDENTIFY & CORRECT SPECIAL CAUSE
in an adverse direction.
All individuals in the Accept product
sample are within Inspect 100% since
100% inspect Continue to reduce
specification. the last in-control part
product variation.

Has gone out of control IDENTIFY & CORRECT SPECIAL CAUSE

in an and one more
individuals in the
sample are outside 100% inspect product produced since the last in-
100% Inspect
specification control sample..

DOCUMENT NO REVISION NO REVISION DATE PAGE OUT OF COPY NO

L2.1/AQMS-SPC 01 03/03/2022 Page 4 of 5 1
 File : OCTEP / SVCF / AQMS /
AUTOMOTIVE QUALITY MANAGEMENT
SRI VIGNESHWARA COLD FORGE QSP
PLOT PB 6 /2, SIPCOT INDUSTRIAL PARK, SYSTEM
IRUNGATTUKOTTAI, KANCHIPURAM DT-602 STATISTICAL PROCESS CONTROL
105
File : iatf\ svcf \ aqms \ qsp

5.7.6 The process shall be focused at the mean to achieve Cpk to the value of Cp, depends upon the
process.

5.7.7 Continuous improvement plan shall be initiated for all special characteristics identified once if the
stability & required capability is achieved (Cpk>1.33) to reduce the variation).

5.7.8 If Cp & Cpk is less than 1.33 QAD shall initiate a corrective action with resp. & target to achieve
the desired process capability index. Till that time, 100% inspection shall be carried out at that
station / at final inspection as appropriate.

5.7.9 On going process capability shall be calculated after three month and monitored in Cpk study
plan .

6.0 DOCUMENT REFERENCE: SPC Manual

7.0 RECORD REFERENCE: 1. Control Charts.

2. Process Capability Reports.

DOCUMENT NO REVISION NO REVISION DATE PAGE OUT OF COPY NO

L2.1/AQMS-SPC 01 03/03/2022 Page 5 of 5 1