Measurement system analysis Combined
with Shewhart’s Approach
Magdalena DIERING
1, a
, Adam HAMROL
,b
and Agnieszka KUJAWIŃSKA
3,c*
1
Poznan University of Technology, Piotrowo 3, 60-965 Poznań, Poland
2
Poznan University of Technology, Piotrowo 3, 60-965 Poznań, Poland
3
Poznan University of Technology, Piotrowo 3, 60-965 Poznań, Poland
a
magdalena.diering@put.poznan.pl,
b
adam.hamrol@put.poznan.pl,
c
agnieszka.kujawinska@put.poznan.pl
Keywords: measurement system analysis, Shewhart’s control charts, statistical process control,
repeatability and reproducibility
Abstract. The paper presents new procedure of methodology for statistical assessment of
measurement systems variation (methodology known in the literature as Measurement Systems
Analysis, MSA). This procedure allows for calculation and monitoring in real time (that is on-line)
of measurement system (MS) characteristics which determine its usability for manufacturing
process control. The presented solution pointed out the gap in process control, which consists in
lack of methods for monitoring measurement processes in the on-line way. Their key point consists
of taking samples that are also needed for the process control chart for the needs of the MSA
method. This means that the samples are taken directly from the production line and during the
production process. The method is combined with the standard procedure of statistical process
control (SPC) with the use of process control charts. It is based on two control charts. The first one
is called AD-chart (Average Difference chart) and it allows to estimate the variation between the
operators and stability of the monitored measurement system. The second control chart illustrates
the %R&R index (Repeatability and Reproducibility) and allows to monitor the MS capability.
The paper also presents authors’ proposal of guidelines about the reference value for the %R&R
index calculation and assessment. Recommendations and guidelines for choosing the reference
value are based on two criteria: information about sample and manufacturing process variation and
the purpose of using MS (product or process control).
Introduction
Measurement system is one of the most important components of production systems – if it is
not capable enough, may have a significant influence on the quality of managers decisions made
about the production process. It should be stable, predictable, capable and lean. It means that MS
capability should be monitored and managed. Also, the measurement uncertainty of the obtained
results should be known.
In the Shewhart’s approach there is used a theory of stochastic processes, in which the process
model is a function of the variability of its measurement and distribution of its location [1]. Basic
tool of this approach, that is of statistical process control (SPC), to monitor process are control
charts. The idea of the charts is to assess the stability of manufacturing process – stable process is a
process, which monitored statistics are kept within control limits. The essence of observation of
image on process chart lies in its skilful interpretation. However, it should be borne in mind that the
plot on the chart shows the total variation in both the manufacturing process and the measurement
process [2]. Thus, variation of the measurement system (MS) should be known and much smaller
than the natural variation of the manufacturing process, to ensure that decisions about the process
(decisions such as: the process is stable/unstable in the statistical sense) and the product (decisions:
good item/bad item) is accurate. Thus, the aim of MS assessment is to gain information whether it is
adequate for controlling the production process and the manufactured product.
Key Engineering Materials Vol. 637 (2015) pp 7-11
© (2015) Trans Tech Publications, Switzerland
doi:10.4028/www.scientific.net/KEM.637.7
All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of TTP,
www.ttp.net. (ID: 46.175.46.69-27/01/15,12:03:49)