Research of quality impact to the product design properties and characteristics/Kokybes itakos konstruojamo gaminio vertei ir charakteristikoms tyrimas.
Cikotiene, D. ; Bargelis, A.
1. Introduction
Quality impact research becomes an integral part of new product and
process development. Customers require permanent improve of product
properties and increase variety of types. Designers seeking the higher
product value and better properties have to apply more complicated
geometrical form and qualitative parameters. These challenges raise the
probability of errors in both the product and process development, and
manufacturing. Manufacturers avoiding errors have to increase the
efforts to quality warrant, which raises the quality costs. The tendency
in many organizations is to think of quality costs in terms of what is
visible: scrap, rework, returns, warranty service, claims and the like
[1].
The reasons of quality cost, however, often lie in the early stage
of a new product and process development that are not visible, or are
sorely to foresee it. This is very urgent in high variety and low volume
manufacturing [2]. Developed new properties of a product can create
quality problems during manufacturing, if designer had not solved them
at the early product design stage. Impacts of a product modular design
on agile manufacturing [3] emphasizing creation easiest of production
and quality product alternative. The well-known Worst case and Root sum
square models [4] and Tolerance allocation method [5] in low volume
manufacturing is inadequate.
The research of this paper exams the quality affects to the new
product design properties and characteristics in high variety and low
volume manufacturing. It accented the factors causing errors and defects
in product design and manufacturing. The new methodology verifying and
estimating product quality in its life cycle phases is proposed and
discussed in this research.
2. Quality impact to product development
The developed methodology to verify and estimate product quality in
its life cycle phases is able to act and appraise the errors and their
probability in product and process development, production stages and
influence of the human factor on originate errors. Starting to appraise
errors in product development and design procedure, organization can
earlier make the necessary corrections and foresee the appropriate
decisions in later activity steps. The virtual prototyping and rapid
prototyping processes extended to the virtual reality technologies must
be widely used in this stage. New product developer particular attention
must pay to the user requirements, creating and applying specialized
techniques for this aim [6]. On the other hand the designer has to check
process cost [7] and quality cost [8] for the product developed at the
early design stage.
Product production stage realizes requirements of finished product
design and has a small influence on manufacturing and quality cost. The
latter can increase and fluctuate over tolerances, when insufficiently
operates quality management system in organization [1] or product and
process design is imperfectly made. The probability for an error to
occur in both the product and process design and manufacturing depends
on many factors as product complexity, applied inter-discipline modules,
human skills and experience, striving novelties and challenges in new
product development and so on.
Acquirement of statistical data and errors probability in product
and process development and manufacturing is very useful technique to
quality impact research. A basic assignment of errors and defects
probability and reasons can be conveniently characterized by the
following functions that map the statistical distribution in different
phases of product design and manufacturing
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (1)
where E1 is the product's design error probability, Dt is the
set of design factors influencing the errors and imperfections
greatness.
Quite often, E1 specifies a percentage measuring level and is
defined empirically or applying expert's knowledge. The knowledge
base (KB) creates many best cases of design error probability for
various products and sets of design factors. The Eq. (1) helps the
designer to find a minimal number of errors and defects seeking higher
product's properties and characteristics applying DFX methodology.
The created analogous expressions of other phases are able to estimate
their errors and defects.
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (2)
where E2 is the product's process design error probability,
[P.sub.j] is the set of process design factors influencing to the errors
and imperfections.
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (3)
where E3 is the manufacturing factor of error probability,
[M.sub.k] is set of manufacturing factors acting to the product errors
and defects.
The Eq. (3) helps managers of production divisions seeking higher
productivity and products quality in managing divisions' activity.
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (4)
where E4 is the human factor of error probability, [H.sub.1], is
the set of human factors influencing the errors and defects in all
phases of product development and manufacturing.
Eqs. (1-4) can separately estimate and forecast the errors
probability of each phase of product development and manufacturing. The
objective of mentioned equations is to specify the greatness of
available errors and defects. They are valuable seeking minimum of each
phase errors. In reality, developing and producing products, the
interfacing of phases is fixed and errors from one phase mutates to the
other increasing initial error. Considerations of sheet metal products
development and manufacturing show that the value of E1-E4 fluctuates in
the margins of 0-0.06 and depends on the product and process complexity
and production volume. The evaluation of an aggregate error mutation
among all product development phases is available as follows
E5 =[there exists](E1)U/(E2)U/(E3)U/(E4) (5)
where E5 is aggregate factor of error probability in product
development and manufacturing.
In an ideal case, E5 = 0, when E1 = E2 = E3 = = E4 = 0 or is equal
to wane small values; conversely, E5 fluctuates from 1 to 100%, i.e.
when the product has defects or errors, or at least is totally
unqualitative when errors and defects in separate product development
and manufacturing stages are not eliminated. The E5 value probability in
this case is calculated by the Evidence theory [9].
[FIGURE 1 OMITTED]
The developed methodology is suitable to consider the two types of
products: 1) composed of some assembling units, original and standard
parts and purchased components, and 2) original parts composed of
various design features with different qualitative-quantitative
parameters (Fig. 1).
The presented research has been made in both laboratory and
industry. The theoretical aspects of quality impact to the first type
product's design properties and characteristics have been created
applying industry statistical data and mathematical modeling methods.
The verification and validation tests for the developed research have
been conducted in laboratory. Some companies of Lithuania took part
conducting the acceptance tests of that research. Research results
chapter illustrates how the developed methodology suits to the real
industry problems.
3. Research results
A sequence of developed research experimental testing is shown. A
typical mechanical product--a recuperator with various original and
standard parts and purchased components has been taken. Lithuanian
company X produces these products of different capacity and different
purposes.
The main task of the recuperator is to clean, heat, to supply fresh
air, and to exhaust used air. For heating of supplied air, the power of
exhausted air is used. Air heaters of the considered air-handling units
are electrical; the units have electronic communication fans with
backward-curved impeller. More than 50 recuperators of different airflow
capacity were investigated. The investigation of different recuperators
(with maximum capacity requirements 400; 1500; 1900; 2000 [m.sup.3]/h
and other) was carried out. Analysis of characteristics--differential
pressure [DELTA]p, output and input airflow productivity V was made.
Dependences between characteristics and construction parameters of the
product were determined.
Multiple nozzles in a chamber (Fig. 2) determine flow rate. The
calculation of flow rate is made according to the methodology given in
standard EN ISO 5167-1 [10]. The purpose of these calculations was to
set up dependence of flow rate on differential pressure.
[FIGURE 2 OMITTED]
Flow rate of the fluid is calculated as follows
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (6)
where [q.sub.m] is the mass flow rate, which passes the vent during
the time unit; [p.sub.sg1] is stagnation pressure; [R.sub.w] is constant
of humid air; [[THETA].sub.sg1] is air flow, passing the fan.
[q.sub.m] = [alpha][epsilon][pi][d.sup.2]/4[square root
(2[rho][DELTA]P] (7)
where [alpha] is flow rate coefficient; s is expansibility
coefficient; d is diameter of the tap; [rho] is air density; [DELTA]p is
differential pressure.
[p.sub.sg1] = [P.sub.eSg3] + [P.sub.a] (8)
where [P.sub.a] is atmosphere pressure during the experiment.
The differential pressure [DELTA]p was changing from the maximum to
minimum [10] meaning for the recuperator, which required airflow rate
capacity V is 400 [m.sup.3]/h. According to the calculation methodic of
dependence between differential pressure and flow rate of the fluid,
given in the standard, supply and exhaust flow rate V, [m.sup.3]/h was
estimated. The results are given in the Table 1.
Analogical experiments were fulfilled with different recuperators
with different requirements for the airflow rate capacity.
Experiments of different parameters, characteristics and assembly
units influence on the supply and exhaust airflow rate capacity where
fulfilled.
Characteristic of some recuperators with different airflow's
rate capacity requirements were investigated. The dependence of airflow
rate capacity on differential pressure was set. Influence of
construction parameter's inaccuracy on the main recuperator
characteristic (airflow rate capacity) was investigated.
Fig. 3 shows the range of airflow productivity when differential
pressure is changing from maximum meaning to 0. Maximum airflow
productivity is lesser the mark (400 [m.sup.3]/h).
[FIGURE 3 OMITTED]
Fifty measurements of airflow productivity were performed. They
were grouped into clusters, completed of 5 individual measurements
(Table 2). Fig. 4 shows the diagram of average airflow dimension before
process correction. This diagram shows that the average airflow
productivity is lesser the mark.
[FIGURE 4 OMITTED]
Average meaning of the airflow productivity must be 400 [+ or -] 20
[m.sup.3]/h. Diagram of the average output airflow productivity shows,
that the average meaning is 360 m3/h. That means that these products are
not working according to the requirements and are not suitable to be
delivered for the customers.
Fig. 5 presents dimension 400 [+ or -] 20 m3/h range values before
process correction. Testing results obtained are not suitable (point 5)
because it is over warning line. Standard parts of the recuperators were
changed and experiments were repeated. Meanings after displacement of
standard parts show better result, but it is not proper.
[FIGURE 5 OMITTED]
Fig. 6 presents the diagram of average airflow dimension after
replacement of standard parts. It shows that the average meaning is
good, but the process is not steady, meaning in point 10 reaches upper
action line. After such correction, results are better, but not proper.
[FIGURE 6 OMITTED]
Second step was precision investigation of recuperator construction
parameters. Design features of recuperator's housings were
examined. Holes of the recuperators' housing were measured.
Diameter of the holes must be 7 [+ or -] 0.1 mm.
Fig. 7 illustrates the diagram of average diameter of the holes
before process correction. It shows, that meanings are not
satisfactory--meaning in point 4 is below the lower action line,
meanings in points 5 and 8--are over the warning lines. The process is
not statistically controlled, so we suppose, that variance of
constructional parameters has influence on airflow productivity of
recuperators, because loses of airflow start up because of nonquality
housing plates.
The means to improve airflow productivity were sought--other ones
replaced standard parts, precision of construction parameters was
investigated. The diagram of average airflow dimensions after process
correction is shown in Fig. 8. The process after correction became
statistically controlled--meanings are arranged in even distribution,
average meaning is in the range of tolerance, values in different points
do not exceed warning lines (Fig. 8).
The dependence between supply-exhaust airflow productivity and
differential pressure of recuperators with required airflow productivity
of 400 m3/h after process correction is presented in Fig. 9.
[FIGURE 7 OMITTED]
[FIGURE 8 OMITTED]
[FIGURE 9 OMITTED]
Airflow productivity, obtained after improving parameters of design
features (holes for cramp of recuperator housing), meets the
requirements.
4. Discussions and conclusions
The research in this paper presents quality impact to the product
design properties and characteristics. New product development is a
permanent solution at contradictions arising among product properties
and manufacturing cost. Latter parameters destine the success of a
developed product. Product development and manufacturing cost have a
significant influence on the total product value. The developed
methodology to verify and estimate product quality in its life cycle
phases can help to improve product quality in early development stages.
The method that has been described in this paper accomplishes the
objective of this research. However, this is not the only method
currently available. It has its advantages and disadvantages. The
advantages are several: the originated methodology simultaneously
verifies errors and appraises their probability in product and process
development and manufacturing stages. The mathematical formalization
helps fixing errors probability in all product life cycles. The
principal shortcoming of the developed methodology is intangible
acquirement of product and process design factors influencing the errors
and imperfections. Briefly, we can conclude as follows:
1. the created methodology to verify and estimate product quality
in its life cycle phases helps to appraise errors at the early product
development stage;
2. it is able to define errors and defects probability in product
and process development and manufacturing phases that helps to avoid
them;
3. the mathematical formalization of the developed methodology
increases chance to implementation.
Acknowledgment
EC Project "Inter-Countries Research for Manufacturing
Advancement (IRMA)" 2008-2009 partially supports this research.
Received July 10, 2009
Accepted September 21, 2009
References
[1.] Ross, J. E. Total Quality Management: Text, cases and
readings.-New York: St. Lucie Press, 1999.-550p.
[2.] Setijono, D., Dahlgaard, J. J. The value of quality
improvements.-Int. J. of Quality & Reliability Management, 2008, 25,
No3, p.292-312.
[3.] Kassi, T., Leisti, S., Puheloinen, T. Impact of product
modular design on agile manufacturing. -Mechanika. -Kaunas:
Technologija, 2008, Nr.6(74), p.56-62.
[4.] Chase, K. W., Parkinson, A., R. A survey of research in the
application of tolerance analysis to the design of mechanical
assemblies. -ADCATS Report No 91-1, April 5, 1991.-42p.
[5.] Ji, S., Li, X., Ma, Y., Cai, H. Optimal tolerance allocation
based on Fuzzy comprehensive evaluation and generic algorithm. -Int. J.
Manufacturing Technology, 2000, 11, No16, p.461-468.
[6.] Robinson, G., Ritchie, J. M., Day, P. N., Dewar, R. G. System
design and user evaluation of Co-Star: An immersive stereoscopic system
for cable harness design.-Computer Aided Design, 2007, 39, April, p.245
257.
[7.] Bargelis, A., Hoehne, G., Cesnulevicius, A. Intelligent
functional model for cost minimization in hybrid manufacturing systems.
-Informatica, 2004, 15, No 1, p.3-22.
[8.] Bargelis, A., Stasiskis A. IDEF0 modelling technique to
estimate and increase the process capability at the early product design
stage. -Mechanika. -Kaunas: Technologija, 2008, Nr.3(71), p.45-50.
[9.] Ayyub, B. M. Elicitation of Expert Opinions for Uncertainty
and Risks. -New York: CRC Press LLC, 2001. -302p.
[10.] EN ISO 5167-1:2003. Measurement of fluid flow by means of
pressure differential devices inserted in circular cross-section
conduits running full - Part 1: General principles and requirements,
2003.
D. Cikotiene *, A. Bargelis **
* Siauliai University, Vilniaus 141, 76353 Siauliai, Lithuania,
E-mail: dalia@tf.su.lt
** Kaunas University of Technology, Kqstu?io 27, 44312 Kaunas,
Lithuania, E-mail: algirdas.bargelis@ktu.lt
Table 1
Flow rate capacity of the supply and exhaust air
Supply
[DELTA], V, Exh aust V,
Pa [m.sup.3]/h [DELTA]p, Pa [m.sup.3]/h
518 0 418 0
458 65 375 65
390 105 347 101
278 194 235 171
238 215 211 198
210 228 186 213
170 255 152 236
120 290 111 274
75 318 72 301
29 348 27 328
0 362 0 361
Table 2
Maximum airflow rate capacity before process correction
Nr. 1 sample 2 sample 3 sample
400 [m.sup.3]/h 1 420.0 400.0 455.0
[+ or -] 20 [m.sup.3]/h 2 390.0 310.0 400.0
T=40[m.sup.3]/h 3 320.0 390.0 320.0
4 340.0 330.0 330.0
5 420.0 410.0 350.0
[mu] 378.0 368.0 371.0
R 100.0 80.0 135.0
[delta] 41.183 40.199 50.239
4 sample 5 sample 6 sample 7 sample
400 [m.sup.3]/h 315.0 397.0 310.0 355.0
[+ or -] 20 [m.sup.3]/h 410.0 398.0 355.0 345.0
T=40[m.sup.3]/h 380.0 330.0 320.0 310.0
360.0 400.0 400.0 380.0
310.0 320.0 300.0 390.0
[mu] 355.0 369.0 337.0 356.0
R 100.0 80.0 100.0 80.0
[delta] 38.210 36.078 36.551 28.178
8 sample 9 sample 10 sample
400 [m.sup.3]/h 300.0 310.0 300.0
[+ or -] 20 [m.sup.3]/h 325.0 325.0 315.0
T=40[m.sup.3]/h 360.0 350.0 320.0
400.0 450.0 410.0
390.0 360.0 350.0
[mu] 355.0 359.0 339.0
R 100.0 140.0 95.0
[delta] 37.947 48.826 39.038
