Machine Design Vol.11 (2019) No.3

UNIVERSITY OF NOVI SAD
FACULTY OF TECHNICAL SCIENCES

ADEKO – ASSOCIATION FOR DESIGN, ELEMENTS AND CONSTRUCTIONS
machine design
Vol.11(2019) No.3
ISSN 1821-1259, E-ISSN 2406-0666
editors:
siniša kuzmanović
MILAN RACKOV
novi sad, 2019

Publication
“Machine Design”
editors
Siniša KUZMANOVIĆ, Ph.D. Eng., University of Novi Sad, Faculty of Technical Sciences
Milan RACKOV, Ph.D. Eng., University of Novi Sad, Faculty of Technical Sciences
Publisher
University of Novi Sad, Faculty of Technical Sciences, Trg Dositeja Obradovića 6, 21000 Novi Sad, Serbia
Supported by
ADEKO, Association for Design, Elements and Constructions
CEEPUS III RS0304; CEEPUS III PL0033; CEEPUS III BG0703
Printed by
Faculty of Technical Sciences, Graphic Center – GRID, Novi Sad, Serbia
technical preparation and cover design

Ivan KNEŽEVIĆ, MSc., University of Novi Sad, Faculty of Technical Sciences
Electronic version of journal available on

www.mdesign.ftn.uns.ac.rs
journal e-mail
m_design@uns.ac.rs
Frequency
Four issues per year
CIP – Каталогизација у публикацији

Библиотека Матице српске, Нови Сад
62-11:658.512.2
MACHINE Design / editors Siniša Kuzmanović, Milan Rackov. - 2009- .-

Novi Sad : University of Novi Sad, Faculty of Technical Sciences, 2009- .- 30
cm
Тромесечно.
ISSN 1821-1259
e-ISSN 2406-0666
COBISS.SR-ID 239401991
SCIENTIFIC editorial board
Prof. Carmen ALIC, Ph.D. Prof. Sava IANICI, Ph.D. Prof. Slobodan NAVALUŠIĆ, Ph.D.
University Politehnica Timisoara, Eftemie Murgu University of Resita, University of Novi Sad,
Faculty of Engineering – Hunedoara, Hunedoara, Faculty of Engineering, Faculty of Technical Sciences,
Romania Resita, Romania Novi Sad, Serbia
Prof. Zoran ANIŠIĆ, Ph.D. Prof. Milan IKONIĆ, Ph.D. Prof. Peter NENOV, Ph.D.
University of Novi Sad, University of Rijeka, "Angel Kanchev" University of Rousse,
Faculty of Technical Sciences, Faculty of Engineering, Faculty of Transport Engineering
Novi Sad, Serbia Rijeka, Croatia Rousse, Bulgaria
Prof. Ranko ANTUNOVIĆ, Ph.D. Prof. Lozica IVANOVIĆ, Ph.D. Prof. Milosav OGNJANOVIĆ, Ph.D.
University of East Sarajevo, University of Kragujevac, University of Belgrade,
Faculty of Mechanical Engineering, Faculty of Engineering, Faculty of Mechanical Engineering,
East Sarajevo, Bosnia and Herzegovina Kragujevac, Serbia Belgrade, Serbia
Prof. Kyrill ARNAUDOW, Ph.D. Prof. Juliana JAVOROVA, Ph.D. Prof. Zoran PANDILOV, Ph.D.
Bulgarian Academy of Sciences, University of Chemical Technology and Ss. Cyril and Methodius University,
Sofia, Bulgaria Metallurgy, Deptartment of Applied Mechanics, Faculty of Mechanical Engineering,
Sofia, Bulgaria Skopje, Macedonia
Prof. Livia Dana BEJU, Ph.D. Prof. Miomir JOVANOVIĆ, Ph.D. Prof. Jose I. PEDRERO, Ph.D.
"Lucian Blaga" University of Sibiu, University of Niš, UNED,
Engineering Faculty, Faculty of Mechanical Engineering, Departamento de Mecanica,
Sibiu, Romania Niš, Serbia Madrid, Spain
Prof. Mirko BLAGOJEVIĆ, Ph.D. Prof. Imre KISS, Ph.D. Prof. József SÁROSI, Ph.D.
University of Kragujevac, University Politehnica Timisoara, University of Szeged,
Faculty of Engineering, Faculty of Engineering – Hunedoara, Faculty of Engineering,
Kragujevac, Serbia Hunedoara, Romania Szeged, Hungary
Prof. Ilare BORDEAŞU, Ph.D. Prof. Peter KOSTAL, Ph.D. Prof. Victor E. STARZHINSKY, Ph.D.
Politehnica University of Timisoara, Slovak University of Technology in Bratislava, V.A. Belyi Metal-Polymer Research Institute of
Faculty of Mechanical Engineering, Faculty of Materials Science and Technology National Academy of Sciences of Belarus, Gomel,
Timisoara, Romania STU, Trnava, Slovakia Belarus
Prof. Marian BORZAN, Ph.D. Prof. Dražan KOZAK, Ph.D. Prof. Radoslav TOMOVIĆ, Ph.D.
Universitary Centre of Baia Mare, J.J.Strossmayer University in Osijek, University of Montenegro,
Technical University of Cluj, Mechanical Engineering Faculty, Faculty of Mechanical Engineering,
Baia Mare, Romania Slavonski Brod, Croatia Podgorica, Montenegro
Prof. Radoš BULATOVIĆ, Ph.D. Prof. Kosta KRSMANOVIĆ, Ph.D. Prof. Andrei TUDOR, Ph.D.
University of Montenegro, University of Arts in Belgrade, University POLITEHNICA of Bucharest,
Faculty of Mechanical Engineering, Faculty of Applied Arts, Faculty of Mechanical Engineering and
Podgorica, Montenegro Belgrade, Serbia Mechatronic, Bucharest, Romania
Prof. Ilija ĆOSIĆ, Ph.D. Prof. Sergey A. LAGUTIN, Ph.D. Prof. Lucian TUDOSE, Ph.D.
University of Novi Sad, Chief Expert on Gears, Design and Technology, Technical University of Cluj-Napoca,
Faculty of Technical Sciences, JS Co „EZTM“, Electrostal, Moscow, Russia Faculty of Machine Building,
Novi Sad, Serbia Cluj-Napoca, Romania
Prof. Eleonora DESNICA, Ph.D. Prof. Tihomir LATINOVIĆ, Ph.D. Prof. Krasimir TUJAROV, Ph.D.
University of Novi Sad, University of Banja Luka, Angel Kunchev University of Rousse, Faculty of
Technical Faculty "M.Pupin“, Faculty of Mechanical Engineering, Agricultural Mechanisation, Department of
Zrenjanin, Serbia Banja Luka, Bosnia and Herzegovina Thermotehnics, Hydro- and Pneumotechnics,
Rousse, Bulgaria
Prof. Lubomir DIMITROV, Ph.D. Prof. Stanislaw LEGUTKO, Ph.D. Prof. Karol VELISEK, Ph.D.
Technical University of Sofia, Poznan University of Technology, Slovak University of Technology in Bratislava,
Faculty of Mechanical Engineering, Institute of Mechanical Technology, Faculty of Materials Science and Technology
Sofia, Bulgaria Poznan, Poland STU, Trnava, Slovakia
Prof. Mircea-Viorel DRAGOI, Ph.D. Prof. Nenad MARJANOVIĆ, Ph.D. Prof. Miroslav VEREŠ, Ph.D.
"Transilvania" University of Brasov, University of Kragujevac, Slovak University of Technology,
Faculty of Technological Engineering and Faculty of Engineering, Faculty of Mechanical Engineering,
Industrial Management, Brasov, Romania Kragujevac, Serbia Bratislava, Slovakia
Prof. Dezso GERGELY, Ph.D. Prof. Biljana MARKOVIĆ, Ph.D. Prof. Simon VILMOS, Ph.D.
University College of Nyíregyháza University of East Sarajevo, Budapest University of Technology and
Faculty of Engineering and Agriculture Faculty of Mechanical Engineering, Economics, Department of Machine and Product
Nyíregyháza, Hungary East Sarajevo, Bosnia and Herzegovina Design, Budapest, Hungary
Prof. Veniamin GOLDFARB, Ph.D. Prof. Štefan MEDVECKY, Ph.D. Prof. Jovan VLADIĆ, Ph.D.
Izhevsk State Technical University, University of Žilina University of Novi Sad,
Institute of Mechanics, Faculty of Mechanical Engineering, Faculty of Technical Sciences,
Izhevsk, Russia Žilina, Slovakia Novi Sad, Serbia
Prof. Ladislav GULAN, Ph.D. Prof. Athanassios MIHAILIDIS, Ph.D. Prof. Adisa VUČINA, Ph.D.
Slovak University of Technology, Aristotle University of Thessaloniki, Faculty of University of Mostar,
Faculty of Mechanical Engineering, Engineering, Lab. of Machine Elements & Faculty of Mechanical Eng. and Computing,
Bratislava, Slovakia Machine Design, Thessaloniki, Greece Mostar, Bosnia and Herzegovina
Prof. Csaba GYENGE, Ph.D. Prof. Vojislav MILTENOVIĆ, Ph.D. Prof. Rushan ZIATDINOV, Ph.D.
Technical University of Cluj-Napoca, University of Niš, Keimyung University, Department of Industrial
Faculty of Machine Building, Faculty of Mechanical Engineering, & Management Engineering,
Cluj-Napoca, Romania Niš, Serbia Daegu, South Korea
Prof. Fuad HADŽIKADUNIĆ, Ph.D. Prof. Radivoje MITROVIĆ, Ph.D. Prof. Miodrag ZLOKOLICA, Ph.D.
University of Zenica, University of Belgrade, University of Novi Sad,
Faculty of Mechanical Engineering, Faculty of Mechanical Engineering, Faculty of Technical Sciences,
Zenica, Bosnia and Herzegovina Belgrade, Serbia Novi Sad, Serbia
technical editorial board
Vasile ALEXA, Ph.D. Jana GULANOVA, Ph.D. Žarko MIŠKOVIĆ, MSc.
University Politehnica Timisoara, Slovak University of Technology, University of Belgrade,
Faculty of Engineering – Hunedoara, Faculty of Mechanical Engineering, Faculty of Mechanical Engineering,
Hunedoara, Romania Bratislava, Slovakia Belgrade, Serbia
Milan BANIĆ, Ph.D. Ivan KNEŽEVIĆ, MSc. Sorin RAŢIU, Ph.D.
University of Niš, University of Novi Sad, University Politehnica Timisoara,
Faculty of Mechanical Engineering, Faculty of Technical Sciences, Faculty of Engineering – Hunedoara,
Niš, Serbia Novi Sad, Serbia Hunedoara, Romania
Jozef BUCHA, Ph.D. Zoran MILOJEVIĆ, Ph.D. Roman RUZAROVSKY, Ph.D.
Slovak University of Technology, University of Novi Sad, Slovak University of Technology in Bratislava,
Faculty of Mechanical Engineering, Faculty of Technical Sciences, Faculty of Materials Science and Technology
Bratislava, Slovakia Novi Sad, Serbia STU, Trnava, Slovakia
Maja ČAVIĆ, Ph.D. Aleksandar MILTENOVIĆ, Ph.D. Milan TICA, Ph.D.
University of Novi Sad, University of Niš, University of Banja Luka,
Faculty of Technical Sciences, Faculty of Mechanical Engineering, Faculty of Mechanical Engineering,
Novi Sad, Serbia Niš, Serbia Banja Luka, Bosnia and Herzegovina
COPYRIGHT
Authors retain copyright of the published papers and have the right to use the article in the ways permitted to
third parties under the Creative Commons Attribution license 3.0 Serbia
(http://creativecommons.org/licenses/by/3.0/rs/).Authors grant to the publisher the right to publish the article, to
be cited as its original publisher in case of reuse, and to distribute it in all forms and media.
DISCLAIMER
The views expressed in the published works do not express the views of the Editors and Editorial Staff. The
authors take legal and moral responsibility for the ideas expressed in the articles. Publisher shall have no
liability in the event of issuance of any claims for damages. The Publisher will not be held legally responsible
should there be any claims for compensation.
from the editor
Machine Design publishes fundamental research about mechanical
engineering and design including machine elements, design fundamentals,
computer aided design, product forms, shapes and performances,
manufacturing processes and technologies, theory of materials, its structures
and capabilities, product design management, technology management,
communication and cognitive science.
The journal is a good opportunity to show and present the results of our
recent work and researching. Also, it is a chance for leader researchers and
scientists in the field of machine design from abroad to represent their
researching results. In such way, we would like to obtain insight in the
present situation of mechanical engineering in the region, to know and learn
about researching in other institutions, to compare results and find out new
solutions, as well as to make new contacts and find out mutual interests for
international cooperation and researching on a project or some topic.
The editorial board will try further to develop this publication in order to
achieve and maintain a high quality of publications, so we can receive an
Impact factor. Our goals are to be referred in international publication
databases, to provide an international medium for scientific contribution and
participation to mechanical engineers and to create a platform for the
communication between science and industry in the field of technical
sciences.
Also, we would like to promote and to encourage international cooperation,
mutual researching, projects and publishing papers between foreign partners’
institutions. Thus, we want to help better understanding and knowing about
work and researching of colleagues from all over the world.
I hope You will recognize the interest to publish Your paper in the journal
Machine Design; so, with a great pleasure, I call You to send further Your
papers for this journal. At the end of the journal we gave the instructions for
formatting and preparing the paper.
For additional information, please visit our website:
www.mdesign.ftn.uns.ac.rs
Editors,
Siniša Kuzmanović & Milan Rackov

CONTENTS:
Preliminary notes
1. Dimensional and Geometrical Tolerances in Mechanical Engineering – a Historical Review

Michael S. J. WALTER ........................................................................................................................................... 67
2. A Comparative Study of Analytical and Numerical Analyses of the Structure of the Double-Beam
Bridge Crane With the Rail in the Middle of the Main Girder
Goran PAVLOVIĆ, Mile SAVKOVIĆ, Nebojša ZDRAVKOVIĆ .......................................................................... 75
3. Models in Safety Management

László POKORÁDI .................................................................................................................................................. 85

Research papers
4. FEA Application Method for Stress Test of the Wheel for Motor and Trailer Vehicles
Eleonora DESNICA, Dejan KLJAJIĆ, Mića ĐURĐEV .......................................................................................... 95
5. Design and Optimization of Vehicle Muffler Using the Ffowcs Williams and Hawkings Model
Barhm MOHAMAD ............................................................................................................................................... 101
6. Repair of the Damaged Screwed Spindle of a Type FSP-160 Friction Screw Press by Repair
Welding
Rade VASILJEVIĆ, Dragan PANTELIĆ .............................................................................................................. 107
MANUSCRIPT FORMAT
machine design, Vol.11(2019) No.3, ISSN 1821-1259 pp. 67-74
DOI: 10.24867/MD.11.2019.3.67-74
Preliminary note
DIMENSIONAL AND GEOMETRICAL TOLERANCES IN MECHANICAL ENGINEERING –

A HISTORICAL REVIEW
Michael S. J. WALTER *
University of Applied Sciences Ansbach, Faculty of Engineering, Germany
Received (03.04.2019); Revised (10.07.2019); Accepted (12.07.2019)
Abstract: The specification of cost-optimal tolerances is ever since a major challenge for design engineers to ensure
functionality as well as economic efficiency of mechanical assemblies. However, the underlying conflict "as tight as
required, as wide as possible" is as old as the design, manufacturing of mechanical parts and their assembly itself –
going back to times, when first plain tools were produced by hand using stones and wood. Since then, a long and
challenging path has been taken by researchers and industrial experts to develop today’s effective methods and tools on
tolerance engineering. In this paper a historical review on tolerance engineering of dimensional and geometrical
tolerances in mechanical engineering is presented. Starting with the early beginnings during the age of individualized
manufacture, four ages of tolerance engineering are analyzed and its major achievements are presented and discussed.
Finally, we are looking ahead – focusing on current and upcoming trends in tolerance engineering.
Key words: Tolerancing, tolerance analysis, GD&T, quality engineering, review
1. INTRODUCTION since first tolerance-related problems arose centuries

years ago (assembly of simple wood/stone hand tools), a
Imperfect manufacturing processes result in parts that long and challenging path has been taken by researchers
differ in dimension and form its nominal geometry [1]. and industrial experts to develop effective methods and
The assembly of these non-ideal parts leads to products tools on tolerance engineering. This paper uses the term
whose aesthetic and functional qualities may not meet the tolerance engineering, which is quite common in Europe
given requirements [2]. to summarize all activities around tolerance specification,
Hence, during product development the design engineer analysis and synthesis, while the term geometrical and
has to specify appropriate tolerances to limit the dimensional tolerancing (short: GD&T) used in ASME
appearing deviations [3]. However, the specification of standard Y14.5 [7] is very popular in the USA.
optimal tolerances is a highly complex and challenging Because this long path has hardly been discussed in a
task [2, 4]. Therefore, appropriate methods and tools on scientific work, this paper details the origins and the
statistical tolerance analysis and tolerance synthesis are development of geometrical and dimensional tolerancing
commonly used during engineering design [2, 5]. by looking deep into work and literature of the four ages
These established methods allow statistical tolerance of tolerance engineering and GD&T history (Fig. 1).
simulations of mechanical assemblies and are used to Furthermore, current aspects and upcoming trends in
identify a cost-optimal tolerance design [6]. However, tolerancing are discussed.
previous:
The age of individual manufacture #1 #2 #3 #4
The four ages of tolerance engineering

Fig.1. The four ages of tolerance engineering
*Correspondence Author’s Address: University of Applied Sciences Ansbach, Residenzstr. 8, 91522 Ansbach, Germany
michael.walter@hs-ansbach.de
Michael S. J. Walter: Dimensional and Geometrical Tolerances in Mechanical Engineering – a Historical Review;
Machine Design, Vol.11(2019) No.3, ISSN 1821-1259, E-ISSN 2406-0666; pp. 67-74
2. A LONG TIME AHEAD - THE AGE OF 4. THE SPECIFICATION OF TOLERANCES

INDIVIDUAL MANUFACTURE TO LIMIT DEVIATIONS (1840 – 1920)
The history and development of tolerance engineering is While the principle of interchangeability tolerated small
closely linked to progress in the manufacture and deviations of the components, at the beginning of the 19th
assembly of mechanical products and in particular tools century efforts were made to maintain the specified ideal
and weapons [3]. First plain tools were made by hand dimensions as precisely as possible [9]. From today's
about one million years ago (simple pointed stones as point of view, this unrealistic and excessive requirement
cutting and grinding tools and for hunting and cooking on a manufactured part’s quality and precision could only
purposes) [8]. With increasing manual skills and be achieved through time- and cost-intensive production
experience, more complex parts could also be machined and extensive reworking or by accepting drastically high
and assembled around 8 000 BC (e.g. pointed stone with a scrap rates. The interchangeability in manufacture,
wooden handle) [9]. Until well into the 20th century, however, illustrated for the first time that the quality and
however, the manufacture of products was dominated by precision of products can be guaranteed even if certain
one-off manual production. The effects of appearing deviations are accepted / tolerated. The associated cost
deviations in dimension and geometry of the individual savings in manufacture led to the introduction of the so-
parts on assemblability and functionality of the product called "go"-tolerances around 1840, which were extended
were usually compensated by experienced craftsmen to the "go / no-go" tolerances about 30 years later [9, 15].
during production or at least significantly reduced in By defining the lower and upper specification limits (LSL
adequate rework. Beyond that, the mostly existing and USL) of a deviating dimension, an interval - the
individual production (quantity = 1) did not place any tolerance - was defined within which the dimension was
further requirements on the interchangeability of considered sufficient in quality and precision. However, if
individual parts. Only with the introduction of assembly the actual dimension lies outside the "go" interval, the
line production in extremely high quantities, e.g. by H. component did not meet the required quality requirements
FORD in January 1914 [10], the control of deviations by and was either reworked or had to be sorted out as reject
limiting them by means of tolerances rapidly gained in (see Fig. 2).
importance.
3. THE STIRVING FOR THE „no-go“ upper specification limit USL

INTERCHANGEABILITY OF PARTS
The idea of complete interchangeability in the production nominal value N

of arrows and bows was already born in ancient Egypt „go“
around 3 000 BC [9]. The aim was to reduce the
production effort and the arrows' individual rework to fit lower specification limit LSL
to the respective bows, which was usually necessary and
highly time-consuming. However, it was only in the 18th „no-go“
century in France that J.-B. VAQUETTE DE QRIBEAUVAL
and H. BLANC, among others, promoted the
interchangeability of weapon parts [11]. Fig.2. „go / no-go- tolerances”
The first successful implementation of interchangeability
in manufacture is wrongly attributed to the US-American At the beginning of the 20th century, the standardization
E. WHITNEY. WHITNEY appeared at the US State of components was pushed forward massively worldwide.
Congress on Jan 8th 1801, completely disassembled ten With the beginning in Great Britain (1901), another 24
muskets to their individual parts and finally assembled a organizations for standardization were founded within
new musket of randomly selected parts [12]. The US State following 31 years [9]. One of the first results published
Congress was extremely impressed and felt confirmed in in 1922 by the German Standards Committee (short: DIN)
having commissioned WHITNEY to produce 10 000 was the DIN system on dimensional fits [16], which still
muskets three years earlier. WHITNEY's presentation of the exists today in a modified form. As standardization
successful interchangeability in his company’s production progressed, the need for a uniform specification of
was repeatedly questioned in the course of the following tolerances in technical drawings was first regulated during
years. In 1979, finally, historians proved that WHITNEY the first half of the 20th century [17].
deliberately manipulated the parts of the muskets prior to The paradigm shift from the philosophy of an ideal
the demonstration and thus deceived the US State manufacture in the second half of the 19th century led to
Congress [13, 14]. Nevertheless, WHITNEY‘s massive increases in scrap rates in industrial manufacture.
demonstration should be seen as one of the most In consequence, drastically increasing efforts had been
important events for quality engineering and tolerance taken to determine expected scrap rates by adequate
engineering and as a stimulus for a multitude of mathematical methods and even further, to keep these as
subsequent research work on various aspects of tolerance low as possible. The most important contribution to this
engineering. Especially, since shortly after WHITNEY's was made by the statistical consideration of the tolerance
death his armaments company was able to realize of components and the effect of deviations on the quality
interchangeability in their production. of the end product.
68
5. THE INTRODUCTION OF STATSTICS TO Carlo-based tolerance analysis, is essential to ensure

ENGINEERING AND THE RISE OF THE reliable results [28]. The method of statistical tolerance
AUTOMOTIVE INDUSTRY (1900 – 1960) synthesis (using root sum squares; short: RSS) was first
discussed in detail by ETTINGER et al. in 1936. Further, by
At the beginning of the 20th century, statistics and the use of adequate linearity coefficients (≠ ±1), they were
probability theory found their way into almost every already enabled the observation of non-linear dimensional
science, including mechanical engineering [9]. The chains [29].
rapidly growing automotive industry in particular
demanded new methods to improve design and assembly
TOLERANCE
Tolerance
Analysis
production and thus the quality and reliability of vehicles. level
In addition to H. FORD, also M. LELAND should be
mentioned above all. The founder of the car brands
CADILLAC and LINCOLN introduced interchangeability TOL E RA NCE component
level
(through the use of standardized parts) as early as 1908,
thus enabling manufacturing contracts to be awarded to
external suppliers for the first time. As a result, every
CADILLAC and LINCOLN car could now be serviced by TOLERANCE assembly
Tolerance
Synthesis
level
independent workshops. The relevance of this innovation
for automotive industry was underlined in 1908 by the
award of the renowned DEWAR TROHPY FOR PRECISION component
AND EXCELLENCE IN MANUFACTURING by the UK ROYAL TOL E RA NCE level
AUTOMOBILE CLUB [18].
As early as in the 1920s, first statistical methods were
applied to the analysis and control of quality problems Fig.3. Tolerance Analysis and Tolerance Synthesis
and for manufacturing process control in mechanical (according to [3])
engineering [19]. So, in 1919 DUDDING began using
statistical methods to study the manufacture processes of During war, research activities were reduced enormously
light bulbs made of glass and metal [20]. Meanwhile in or focused on the production of military equipment [30].
the USA, the statistical analysis of industrial production However, in the following years the rising automotive
and assembly processes was essentially driven forward by production (especially in the USA) and the rapidly
SHEWHART, who introduced statistical process control increasing economic and thus purchasing performance in
(short: SPC) in 1924 [21]. In Germany, the industrial Europe spurred the spread of statistical tolerance
application of statistical methods was initiated in 1927 calculations in research and industrial mass production
with the publication of the book "Applications of [31]. In 1946, EPSTEIN showed the application limitations
mathematical statistics to problems of mass production" of RSS-based tolerance calculations [32]. GIBSON
[22] by Becker et al. [19]. This book and SHEWHARTs quantified the advantage of a statistical tolerance design
pioneering book from 1931 “Economic Control of Quality over a tolerance design based on worst-case tolerance
of Manufactured Product“ [23] are largely credited with analyses [33], while STÖCKER uses this potential to widen
the upheaval in the quality thinking and understanding of individual tolerances in order to reduce the resulting
German engineers in the pre-war years [19, 24]. manufacturing costs [34]. In addition, new methods (such
SHEWHART explained in [25] as early as 1925 that the as design of experiments according to [35]) and findings
quality of each product during operation is a function of on the application of statistical methods (e.g.
the qualities of the single parts used [26]. In [26], he determination of adequate sample sizes for tolerance
justified the necessity to analyze the effects of each issues by [36]) facilitated the application of statistical
individual part’s deviations on the resulting quality of the tolerance analysis to solve tolerance problems in industry
final product, taking into account the associated [21].
probability distributions and corresponding scrap rates. In addition, an increasingly holistic understanding of the
With this statement, SHEWHART justified the method of effects of dimensional and geometrical deviations on
statistical tolerance analysis in its currently known form - mechanical parts and assemblies has developed. DUDDING
even if he did not explicitly use this term in his et al. demanded from the design engineer both statistical
publications. knowledge and sufficient experience in manufacture and
In remaining pre-war years, statistical tolerance assembly of products [21]. RICE saw the main conflict in
calculation increasingly became subject of scientific tolerance engineering "as tight as required, as wide as
publications. RÜDENBERG quantified the influence of each possible" solvable only by an interdisciplinary team,
geometrical deviation on the variation of the air gap consisting of designer engineers, production experts and
between rotor and stator of asynchronous motors using statisticians [37]. BATES supported RICE’s conviction with
the so-called contribution of each deviation [27]. This the example of a successful cooperation of various
contribution corresponds to the essential result of a departments, such as engineering design, production,
statistical tolerance analysis, nowadays known as the quality assurance, controlling and management [30] and
statistical contributor. MOLINA examined the influence of evaluated in [15] the caused financial effort of this
sample size on the significance of the findings from this interdisciplinary cooperation as comparably small to the
sample and makes it clear that the definition of an cost savings resulting from the achieved improved
adequate sample size, as required, for example, in Monte allocation of tolerances.
69
6. THE RISE OF COMPUTER-AIDED optimization using the pseudo-bool programming of [58]

TOLERANCE ENGINEERING and [53] and thus drastically reduced the numerical effort
(SINCE 1960) [59], while CHASE et al. confirmed the higher
performance of numerical tolerance optimization with
With the increasing availability of mainframe computers Lagrange Multipliers compared to traditional approaches
in the early 1960s, research and development in tolerance of tolerance allocation [3]. With the introduction of the
engineering probably experienced the greatest impetus to computer, the development of software tools for tolerance
date. Existing and established methods and approaches analysis (among others [60, 61, 62]) gained in importance
[38, 39, 40] could be implemented on-the-fly and then in the following years. As a result, a large number of
applied quickly and easily to various tolerancing issues. It software tools [63, 64] were developed, which have since
is therefore not surprising that within a few years established themselves in companies worldwide. Due to
tolerance-related researchers relied massively on the variety of methods and tools developed, there was an
mainframe support. MCGHEE et al. used a mainframe to increasing need to revise existing standards and
analyze the effects of manufacturing tolerances on cost specification specifications. The revision of the
and functionality of mechanical assemblies [41]. A short Geometric Product Specification (short: GPS, [65]) by
time later HOPKINGS et al. and KNAPPE presented expert committees at both national [66] and international
computer-based assistance systems for the design level [67] deserves special mention. In addition, research
engineer, which among other things enabled the activities on tolerance engineering have increasingly
computer-aided tolerance analysis [42, 43]. In addition, grown together over the last three decades. This can be
GROSSMANN explained the advantages of the Monte Carlo seen, among others, in the establishment of various
method in the context of statistical tolerance analysis working groups, conferences and expert panels such as
using a computer [44]. the ADCATS Group in the USA, the French Groupe de
In addition to tolerance analysis, however, tolerance Recherche en Tolérancement, the CIRP Conference on
synthesis in particular benefited from the almost Computer Aided Tolerancing [68] and the Dimensional
unlimited possibilities of numerical computer support. In Management Forum of German Automotive OEMs.
1964, MOY used a numerical optimization method to
determine the optimal tolerance allocation of a 7. NEW ASPECTS AND TRENDS IN
mechanical system for the first time via numerical TOLERANCE ENGINEERING
computer power [45]. He thus founded a multitude of
future research work in the field of statistical tolerance While research in the field of virtual product development
optimization and numerical tolerance allocation (e.g. [46 - has been shifting its focus away from the traditional focus
54]) as well as on the required tolerance-cost models of "method and tool development" towards cross-domain
(Fig. 2). The contributions of SPECKHART, SPOTTS and aspects of product development for several years [69, 70],
SUTHERLAND et al. in the 1970s deserve special mention this development has only recently entered tolerance
[55, 56, 57]. The 1970s can therefore be regarded as the engineering.
decade of tolerance-cost models. Above all, environmental protection and human interests
are already subject of initial, much-noticed research work.
25 The consideration of the effects of tolerance allocation on
manufacturing process #1
the environment was first discussed in 2013 in [71] with a
view to sustainable tolerance. The 13th CIRP Conference
Manufacturing costs in €
20
on Computer Aided Tolerancing (CAT 2014) focused on
15
manufacturing process #2 "Tolerancing for disassembly and recycling" and
underlines the relevance of holistic view on tolerance
engineering. The human is no longer seen merely as a
10
user of the methods and tools developed [70]. Rather,
manufacturing process #3
peoples’ concerns and requirements as i) users of
5 tolerance analysis software and as ii) users of products
produced and thus tolerated (in the form of requirements
0 for these products’ functional, aesthetic and perceived
0 0.2 0.4 0.6 0.8 1 quality [72, 73] as well as for new methods and tools
Tolerance T in mm
[65]) are already taken into account during design and
Fig.4. Tolerance-Cost-Relations development (such as tolerance visualization in Virtual
Reality or Augmented Reality environments) [74, 75].
About 30 years after the introduction of mainframe With regard to the further development of methods and
computers into tolerance engineering, the use of compact computer-aided tools, the handling of large amounts of
personal computers has rapidly gained in importance. data should be mentioned in particular [76]. Such big data
From then on, statistical tolerance simulations could be arise, for example, in the context of integrative tolerance
carried out comparatively easily and with considerably analysis and the NASA philosophy on the “digital twin”
lower financial expense directly at the engineer's [77] in the field of industry 4.0 [78]. The results of a large
workplace. The relevance of this development has been number of numerical simulations are required for the
omnipresent in almost all areas of tolerance engineering integration of operational deviations into the tolerance
since then and has spurred progress in research and calculation. In order to keep the numerical effort as low as
development. Thus, KIM et al. simplified tolerance possible, the use of machine learning and artificial
70
intelligence has become established [79, 80, 81]. In [10] Sponsel, H.: Ford - Vom Blechesel für jeden zum
addition, the scope of performance of current CAE-tools Traumwagen für alle, Sigbert Mohn Verlag, 1.
enables the consideration of further operation-dependent edition, Gütersloh.
deviations such as thermally caused geometrical [11] Alder, K. (1997). Innovation and Amnesia:
deviations or abrasive wear [82, 83]. In addition, Engineering Rationality and the Fate of
tolerance management is increasingly understood as an Interchangeable Parts Manufacturing in France.
essential component of robust design [84, 85] and is thus Technology and Culture, 38, 2, pp. 273–311.
attracting increasing attention both in robust design [86] [12] Shewhart, W. A. (1986). Statistical method from the
and in associated work [87]. Finally, the currently biggest viewpoint of quality control, In: Deming, W. E. (Ed.),
hype in tolerancing is seen in the additive manufacturing. Dover, 1986, New York. (1. Re-print of the first press
The associated new manufacturing technologies open a from 1939)
large variety of challenges for upcoming research on
[13] Smith, M. R. (1990). Eli Whitney and the American
tolerancing for additive manufacturing [88, 89].
System, In: Technology in America: A History of
Individuals and Ideas, Pursell, C. W. (Ed.), pp. 49–
DEDICATION 65, MIT Press, Cambridge.
This paper is dedicated to Kenneth W. Chase, founder of [14] Lindsay, D. (2005). Madness in the Making: The
the ADCATS group, coryphée on tolerance engineering Triumphant Rise and Untimely Fall of America's
and author of the publications that got me into tolerance Show Inventors, iUniverse, 2. edition, Lincoln.
engineering and its history. Thank you for your work. [15] Bates, E. E. (1949). Specifying Design Tolerances –
by statistical methods. Machine Design, 21, 3; pp.
120–125.
[16] Schuh, R. (1947). Von DIN- zu ISA-Passungen,
REFERENCES Wälzlagerpassungen. Ein Lehr- und Nachschlageheft
für Werkstatt und Schule, C. F. Müller Verlag, 2.
[1] Srinivasan, V. (2005). Elements of computational Edition, Karlsruhe.
metrology, In: Computer aided design and
[17] Voelcker, H. B. (1998). The current state of affairs in
manufacturing, DIMACS Book Series #67, Janardan,
dimensional tolerancing. Integrated Manufacturing
R.; Smid, M. & Dutta, D. (Ed.), pp. 79–116,
Systems, 9, 4, pp. 205–217.
American Mathematical Society, Providence.
[2] Wartzack, S.; Meerkamm, H.; Stockinger, A.; Stoll, [18] Royal Automobile Club: Past Winners. Available
T.; Stuppy, J.; Voß, R.; Walter, M. & Wittmann, S. from: http://www.royalautomobileclub.co.uk/Motori
(2011) Lebenszyklusorientierte Toleranzsimulation ng/dewar-trophy. Accessed: 2018-08-31.
zur funktionalen und ästhetischen [19] Tobies, R. (2007). Zur Position von Mathematik und
Produktabsicherung. Konstruktion, 63, 6, pp. 63–67. Mathematiker/innen in der Industrieforschung vor
[3] Chase, K. W. & Greenwood, H. W. (1988). Design 1945 - am Beispiel früher Anwendung von
Issues in Mechanical Tolerance Analysis. mathematischer Statistik in der Osram GmbH. NTM
Manufacturing Review, 1, 1; pp. 50–59. International Journal of History and Ethics of
[4] Chase, K. W.; Greenwood, W. H.; Loosli, B. G. & Natural Sciences, Technology and Medicine, 15, 4,
Hauglund, L. F. (1990). Least Cost Tolerance pp. 241–270.
Allocation for Mechanical Assemblies with [20] Dudding, B. P. (1952). The Introduction of Statistical
Automated Process Selection. Manufacturing Methods to Industry. Journal of the Royal Statistical
Review, 3, 1, pp. 49–59. Society. Series C (Applied Statistics), 1, 1, pp. 3–20.
[5] Chase, K. W. & Parkinson, A. R. (1991). A Survey of [21] Dudding, B. P. & Jennett, W. J. (1940). Statistics and
research in the Application of Tolerance Analysis to Engineering Practice. Journal of the Institution of
the Design of Mechanical Assemblies. Research in Electrical Engineers, 87, 523, pp. 1–15.
Engineering Design, 3, 1, pp. 23–37. [22] Becker, R; Plaut, H. & Runge, I. (1927).
[6] Walter, M. S. J.; Sprügel, T. C. & Wartzack, S. Anwendungen der mathematischen Statistik auf
(2015). Least Cost Tolerance Allocation for Systems Probleme der Massenfabrikation, Springer, 1.
with Time-variant Deviations. Procedia CIRP, 27, edition, Berlin.
pp. 1–9. [23] Shewhart, W. A. (1931). Economic Control of
[7] Meadows, J. D. (2009). Geometric Dimensioning and Quality of Manufactured Produc,. D. Van Nostrand,
Tolerancing-Applications, Analysis & Measurement 1. edition, New York.
[per ASME Y14.5-2009], James D. Meadows & [24] Gasca, A. M. (2004). Organization and Mathematics:
Associates, 1. Edition. A Look into the Prehistory of Industrial Engineering.
[8] Fütterer, O. (2005). Properties of Multivariate In: Technological Concepts and Mathematical
Statistics from the Viewpoint of Industrial Models in the Evolution of Modern Engineering
Application. Proceedings of the Joint International Systems, Lucertini, M.; Gasca, A, M. & Nicolò, F.
IMEKO TC1 + TC7 Symposium, 21.09.-24.09., (Ed.), Springer, Basel.
Ilmenau. [25] Shewhart, W. A. (1925). The Application of Statistics
[9] Shewhart, W. A. (1938). Applications of Statistical as an Aid in Maintaining Quality of a manufactured
Methods to Manufacturing Problems. Journal of the Product. Journal of the American Statistical
Franklin Institute, 226, 2, pp. 163–186. Association, 152, 20, pp. 546–548.
71
[26] Shewhart, W. A. (1928). Economic aspects of [43] Knappe, L. F. (1965). A Computer Oriented
engineering applications of statistical methods. Mechanical Design System. Proceedings of the
Journal of the Franklin Institute, 205, 3, pp. 395– SHARE design automation project (DAC ’65).
405. 23.06.-25.06., Atlantic City.
[27] Rüdenberg, R. (1929). Die Beurteilung elektrischer [44] Grossman, D. (1976). Monte Carlo Simulation of
Maschinen und Apparate durch Toleranzen auf Tolerancing in Discrete Parts Manufacturing and
Grund statistischer Methoden. Zeitschrift für Assembly. Memo AIM-280, Report No. STAN-B-
Angewandte Mathematik und Mechanik, 9, 4, pp. 76-555, Stanford Artificial Intelligence Laboratory.
318–334. [45] Moy, W. A. (1964). Assignment of tolerances by
[28] Molina, E. C. (1935). Probability in Engineering. dynamic programming. Product Engineering, 21,
Transactions of the American Institute of Electrical May, pp. 215–218.
Engineers, 54, 4, pp. 423–427. [46] Edel, D. H. & Auer, T. B. (1964). Determine the least
[29] Ettinger, W. J. & Bartky, W. (1936). Basis for cost combination for tolerance accumulations in a
Determining Manufacturing Tolerances. Product drive shaft seal assembly. General Motors
Engineering, August, pp. 297–300. Engineering Journal, 11, 4, pp. 37–38.
[30] Bates, E. E. (1947). How to Increase Tolerances and [47] Bennett, G. & Gupta, L. C. (1970). Least-Cost
Obtain Closer Fits. The Iron Age, July, pp. 58–61. Tolerances I. International Journal of Production, 8,
[31] Mansoor, E. M. (1963). The Application of 1, pp. 65–74.
Probability to Tolerances Used in Engineering [48] Bennett, G. & Gupta, L. C. (1970). Least-Cost
Designs. Proceedings of the Institution of Mechanical Tolerances II. International Journal of Production, 8,
Engineers, 178, 1, pp. 29–39. 2, pp. 169–182.
[32] Epstein, B. (1946). Tolerances on Assemblies. [49] Peters, J. (1970). Tolerancing the Components of an
American Machinist, 90, 1, pp. 119–121. Assembly for Minimum Cost. Journal of
[33] Gibson, J. (1951). A New Approach to Engineering Manufacturing Science and Engineering, 92, 3,
Tolerances, The Machinery Publishing, 1. edition, pp. 677–682.
London. [50] Grey, D. S. (1970). Tolerance Sensitivity and
[34] Stöcker, F. (1960). Über Toleranzen bei Optimization. Applied Optics, 9, 3, pp. 523–526.
zusammengesetzten Massenfabrikaten. In: Beiträge [51] Bandler, J. W. (1974). Optimization of design
zu Gegenwartsproblemen der Angewandten Statistik, tolerances using nonlinear programming. Journal of
Statistische Vereinigung Köln e. V. (Ed.), pp. 127– Optimization Theory and Applications, 14, 1, pp. 99–
134, Westdeutscher Verlag, Köln. 114.
[35] Fisher, R. A. (1935). The Design of Experiments, [52] Wilde, D. & Prentice, E. (1975). Minimum
Hafner Publishing, 2. edition, New York. Exponential Cost Allocation of Sure-Fit Tolerances.
[36] Wilks, S. S. (1941). Determination of Sample Sizes Journal of Manufacturing Science and Engineering,
for Settings Tolerance Limits. The Annals of 97, 4, pp. 1395–1398.
Mathematical Statistics, 12, 1, pp. 91–96. [53] Ostwald, P. F. & Huang, J. (1977). A Method for
[37] Rice, W. B. (1944). Setting Tolerances Scientifically. Optimal Tolerance Selection. Journal of
ASME Mechanical Engineering, December, pp. 801– Manufacturing Science and Engineering, 99, 3,
803. pp. 558–565.
[38] Evans, D. H. (1974). Statistical Tolerancing: The [54] Michael, W. & Siddall, J. N. (1982). The Optimal
State of the Art – Part I. Background. Journal of Tolerance Assignment With Less Than Full
Quality Technology, 6, 4, pp. 188–195. Acceptance. Journal of Mechanical Design, 104, 4,
pp. 855–860.
[39] Evans, D. H. (1975). Statistical Tolerancing: The
State of the Art – Part II. Methods for Estimating [55] Speckhart, F. H. (1972). Calculation of Tolerance
Moments. Journal of Quality Technology, 7, 1, pp. 1– Based on a Minimum Cost Approach. Journal of
12. Manufacturing Science and Engineering, 94, 2,
pp. 447–453.
[40] Evans, D. H. (1975). Statistical Tolerancing: The
State of the Art – Part III. Shifts and Drifts. Journal [56] Spotts, M. F. (1973). Allocation of Tolerances to
of Quality Technology, 7, 2, pp. 72–76. Minimize Cost of Assembly. Journal of
Manufacturing Science and Engineering, 95, 3,
[41] McGhee, R. B. & Levine, A. (1959). Determination
pp. 762–764.
of optimum production tolerances by analog
simulation. Proceedings of the Eastern Joint [57] Sutherland, G. H. & Roth, B. (1975). Mechanism
Computer Conference, 01.12.-03.12., Boston, The design: Accounting for Manufacturing Tolerances
National Joint Computer Committee (Ed.), pp. 249– and Costs in Function Generation Problems. Journal
254. of Engineering in Industry, 97, 1, pp. 283–286.
[42] Hopkins, R. E. & Spencer, G. (1962). Creative [58] Wilde; D. (1975). Simplifying Discrete Tolerance
Thinking and Computing Machines in Optical Assignment. ASME Paper 75-DET-106.
Design. Journal of the Optical Society of America, [59] Kim, S. H. & Knott, K. (1998). A pseudo-boolean
52, 2, pp. 172–176. approach to determining least cost tolerances.
72
International Journal of Production Research, 26, 1, States. Proceedings of the 19th International
pp. 157–167. Conference on Engineering Design, Vol. 7: Human
[60] Sayed, S. E. Y. & Kheir, N. A. (1985). An efficient Behaviour in Design, 19.08.-22.08., Seoul,
technique for minimum-cost tolerance assignment. Lindemann, U.; Venkataraman, S.; Kim, Y. S.; Ion,
SIMULATION, 44, 4, pp. 189–195. W. & Malqvist, Y. (Ed.), pp. 467–476.
[61] Roy, U.; Liu, C. R. & Woo, T. C. (1991). Review of [73] Konstantinos, S.; Wickman, C. & Söderberg, R.
dimensioning and tolerancing: representation and (2015). Defining Perceived Quality in the
processing. Computer-Aided Design, 23, 7, pp. 466– Automotive Industry: An Engineering Approach.
483. Procedia CIRP 36, pp. 165-170.
[62] Craig, M. (1996). Dimensional management versus [74] Wickman, C.; Söderberg, R. & Lindkvist, L. (2001).
tolerance assignment. Assembly Automation, 16, 2, Toward non-nominal virtual geometric verification
pp. 12–16. by combining VR and CAT technologies, In:
[63] Salomons, O.; van Houten, F. & Kals, H. (1998). Geometric Product Specification and Verification:
Current Status of CAT Systems. In: Geometric Integration of Functionality - Selected Conference
Design Tolerancing, El Maraghy, H.A. (Ed.), pp. Papers of the 7th CIRP International Seminar on
438–452, Springer, Dordrecht. Computer-Aided Tolerancing, held at the École
Normale Supérieure de Cachan, France, 24–25 April
[64] Prisco, U. & Giorleo, G. (2002). Overview of current
2001, Bourdet, P.; Mathieu, L. (Ed.), pp. 301-310,
CAT systems. Integrated Computer-Aided
Springer, Berlin.
Engineering, 9, 4; pp. 373–387.
[75] Zhenyu, L. & Jianrong, T. (2005). Virtual assembly
[65] International Organization for Standardization
and tolerance analysis for collaborative design.
(2017). Geometrical product specifications (GPS) -
Proceedings of the Ninth International Conference
Geometrical tolerancing - Tolerances of form,
on Computer Supported Cooperative Work in Design,
orientation, location and run-out (ISO 1101:2017).
24.06.-26.06., Coventry, School of Mathematical and
[66] Deutsches Institut für Normung (2014). DIN-Tagung Information Sciences, Coventry University.
GPS 2014 - ISO-GPS-Normen zur Spezifikation und
[76] Montgomery, D. (2013). The Quality, Reliability and
Verifikation von Funktionsanforderungen;
Statistical Engineering Professional in the 21st
„Technische Produktdokumente“. Available from:
Century. Quality and Reliability Engineering
http://www.beuth.de/sixcms_upload/media/2247/Tag
International, 20, 1; pp. 1.
ungsunterlage-GPS- 2014.pdf. Accessed: 2014-07-
30. [77] Söderberg, R.; Wärmefjord, K.; Carlson, J. S. &
Lindkvist, L. (2017). Toward a Digital Twin for real-
[67] Nielsen, H. S. (2013). Recent developments in
time geometry assurance in individualized
International Organization for Standardization
production. CIRP Annals, 66, 1, pp. 137-140.
geometrical product specification standards and
strategic plans for future work. Proceedings of the [78] Wartzack, S., Schleich, B., Aschenbrenner, A. &
Institution of Mechanical Engineers, Part B: Journal Heling, B. (2017). Toleranzmanagement im Kontext
of Engineering Manufacture, 207, 5, pp. 643–649. von Industrie 4.0. ZWF Zeitschrift fuer
Wirtschaftlichen Fabrikbetrieb, 112, 3, pp. 170-172.
[68] Garaizar, O. R.; Anwer, N.; Mathieu, L. & Qiao, L.
(2015): Exploring the proceedings of Computer [79] Andolfatto, L.; Thiebaut, F.; Douilly, M. & Lartigue,
Aided Tolerancing CIRP Seminars and Conferences: C. (2013). On Neural Networks’ Ability to
A scientometric analysis. Proceedings of the 13th Approximate Geometrical Variation Propagation in
CIRP Conference on Computer Aided Tolerancing, Assembly. Procedia CIRP, 10, pp. 224–232.
11.-14.05, Hangzhou, Keynote-Paper. [80] Walter, M. S. J. (2016). Toleranzanalyse und
[69] Krogstie, L. (2015). Closed Loop Tolerance Toleranzsynthese abweichungsbehafteter
Engineering, Dissertation, Norwegian University of Mechanismen. Dissertation, University Erlangen-
Science and Technology NTNU Trondheim, NTNU Nuremberg. VDI Fortschritts-Berichte, 1, 437, VDI,
2015:1. Düsseldorf.
[70] Krogstie, L.; Walter, M. S. J.; Wartzack, S. & [81] Closed Loop Tolerance Engineering, Dissertation,
Martinsen, K. (2015). Towards a more Norwegian University of Science and Technology
comprehensive Understanding of Tolerance NTNU Trondheim, NTNU 2015:1.
Engineering Research Importance. Procedia CIRP, [82] Zuowei, Z.; Anwer, N.; Qing, H. & Mathieu, L.
27, pp.29–34. (2018). Machine learning in tolerancing for additive
[71] Hoffenson, S.; Dagman, A. & Söderberg, R. (2013): manufacturing. CIRP Annals - Manufacturing
Tolerance Specification Optimization for Economic Technology, 67, pp. 157–160.
and Ecological Sustainability. Smart Product [83] Pierre, L.; Teissandier, D. & Nadeau, J. P. (2009).
Engineering - Proceedings of the 23rd CIRP Design Integration of thermomechanical strains into
Conference, 11.03.-13.03., Bochum, Stark, R. & tolerancing analysis. International Journal on
Abramovici, M. (Ed.), pp. 865–874, Springer, Berlin. Interactive Design and Manufacturing, 3, 4, pp. 247–
[72] Hoffenson, S.; Dagman, A. & Söderberg, R. (2013) 263.
Tolerance Optimization of a Mobile Phone [84] Pierre, L.; Teissandier, D. & Nadeau, J. P. (2009).
Considering Consumer Demand for Quality and Variational tolerancing analysis taking
Sustainability in China, Sweden, and the United thermomechanical strains into account: Application
73
to a high pressure turbine. Mechanism and Machine [88] Ameta, G.; Witherell, P.; Moylan, S. & Lipman, R.
Theory, 74, April, pp. 82–101. (2015). Tolerance Specification and Related Issues
[85] Taguchi, G., Chowdhury, S. & Wu, Y. (2007). for Additively Manufactured Products. ASME 2015
Taguchi’s Quality Engineering Handbook. John International Design Engineering Technical
Wiley & Sons, New Jersey. Conferences and Computers and Information in
[86] Eifler, T.; Ebro, M. & Howard, T. J. (2013). A Engineering Conference - Volume 1A: 35th
Classification of the Industrial Relevance of Robust Computers and Information in Engineering
Design. Proceedings of the 19th International Conference, 02.08.-05.08. Paper No. DETC2015-
Conference on Engineering Design, Vol. 9: Design 47531, Boston, ASME.
Methods and Tools, 19.08.-22.08., Seoul, Lindemann, [89] Lienke, T.; Denzer, V.; Adam, Guido. A. O. &
U.; Venkataraman, S.; Kim, Y. S.; Ion, W. & Zimmer, D. (2016). Dimensional Tolerances for
Malqvist, Y. (Ed.), pp. 427–436. Additive Manufacturing: Experimental Investigation
[87] Howard, T. J. & Eifler, T. (Ed.) (2014). Proceedings for Fused Deposition Modeling. Procedia CIRP, 43,
of the First International Symposium on Robust pp. 286–291.
Design ISoRD 2014, 14.08.-15.08., Copenhagen.
© 2019 Authors. Published by the University of Novi Sad, Faculty of Technical Sciences.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution
license 3.0 Serbia (http://creativecommons.org/licenses/by/3.0/rs/).
74
DOI: 10.24867/MD.11.2019.3.75-84
Preliminary note
A COMPARATIVE STUDY OF ANALYTICAL AND NUMERICAL ANALYSES OF THE

STRUCTURE OF THE DOUBLE-BEAM BRIDGE CRANE WITH THE RAIL IN THE MIDDLE
OF THE MAIN GIRDER
Goran PAVLOVIĆ1, * - Mile SAVKOVIĆ2 - Nebojša ZDRAVKOVIĆ2
1 R&D Center Alfatec Ltd, Niš, Serbia
2 University of Kragujevac, Faculty of Mechanical and Structural Engineering, Kraljevo, Serbia
Abstract: This research deals with the analysis and comparison of the results obtained analytically and by using the
finite element method on the box girder of a double-beam bridge crane with the rail in the middle of the girder, where
an existing bridge crane, which is in use, was observed. Stress states in the girder were observed in this analysis, and
the emphasis was placed on the local stress states which occur in the middle of the girder due to the wheel-rail
pressure. The expressions used in the domestic literature and applied in the design of these types of structures were
used for obtaining analytical values of the results. The software package SAP2000 and the software module Autodesk
Nastran In-CAD of the software package Autodesk Inventor were used for the FME analysis, where the girder structure
was modelled by shell finite elements in both cases. In the application of the software package SAP2000, the rail was
modelled by beam finite elements, and the loads were applied on the nodes of the finite element mesh which connect
beam and plate finite elements, while in the application of the software module Autodesk Nastran In-CAD the rail was
modelled by 3D finite elements, and the loads were applied on the external rail surfaces which are in contact with the
wheel.
The aim of this research is to present the comparison between analytical and numerical results in order to justify the
application of the proposed models of FEM analysis, their advantages and disadvantages, as well as the credibility of
the obtained results in relation to the existing analytical model of calculation. In this way, it can be shown whether the
FEM results are within those obtained analytically, so that in the design of these types of structures the proposed
modelling methods could be reliably used.
Key words: Box girder, Bridge crane, Local stresses, SAP2000, Autodesk Nastran In-CAD
1. INTRODUCTION Nowadays, judging by the practice of the leading

manufacturers of crane equipment, double-beam bridge
Double-beam bridge cranes are most widely spread and cranes with the rail above the web are increasingly used
have the largest application in relation to the other types but, regardless of this fact, these structures with the rail in
of cranes, which are used in industrial halls and plants. the middle of the main girder are still present (Figure 1).
Also, there are numerous publications which deal with the
problem of analysis and optimization of double-beam
bridge cranes with the rail in the middle of the main
girder, as well as with their comparison with the
mentioned type of solution. In addition to stresses and
strains, analyses increasingly include the other important
criteria.
The analysis and optimization of the girder in a double-
beam bridge crane with the rail in the middle of the main
girder is most often performed by applying FEM, i.e. by
using 3D finite elements, [1-10, 12, 13], indicating the
importance and justification of application of the given
software packages.
The paper [1] presents the design of main girders in a
double-beam bridge crane, with the application of the
CAE platform based on ANSYS software. It shows a
Fig.1. Double-beam bridge crane with the rail in the
considerable saving in time by using this type of design,
middle of the main girder
where maximum stresses, strains and fatigue are taken
They are most often applied in load manipulation in into account. ANSYS is largely applied in these types of
production as well as for maintenance in various types of structures, which can also be seen in [2], which presents
plants. the analysis of an existing bridge crane designed
*Correspondence Author’s Address: R&D Center Alfatec Ltd, Bulevar Nikole Tesle 63/5, 18000 Niš, Serbia,
goran.pavlovic@alfatec.rs
Goran Pavlović, Mile Savković, Nebojša Zdravković: A Comparative Study of Analytical and Numerical Analyses of the Structure of the Double-
Beam Bridge Crane With the Rail in the Middle of the Main ...; Machine Design, Vol.11(2019) No.3, ISSN 1821-1259, E-ISSN 2406-0666; pp. 75-84
according to the Indian standard, according to stresses and the main girder. The research described in paper [15]
deflections, where the crane model was created in Creo dealt with the problem of optimization of main girders
software. The bridge crane designed according to the with the rail above the web, with the emphasis placed on
Indian standard is also analyzed in [3], where the maximum stresses, i.e. the occurrence of biaxial stress
comparison between analytical and numerical results was state, and the application of the given model was justified
carried out, and the model of the carrying structure of the in two examples of cranes, by using HS and SA
bridge crane was obtained in Pro/E software. Similarly to optimization algorithms. Application of the well-known
the previous research, the comparison between the analytical optimization process, the method of Lagrange
analytical method and the FEM results was performed, multipliers, was shown in [16], where an example of a
with the stresses reduced by shaping sharp edges, [4]. bridge crane was used to show how the optimum
Also, ANSYS software was applied for the optimization geometrical parameters of a box girder change depending
of the mass of the main girders in a double-beam bridge on the selection of materials and conditions in which the
crane, [5], where the saving of 19.4 % was achieved, crane opertaes, based on the stress states occuring in the
taking into account stresses and deflections of the upper flange of the girder.
structure. Similarly, the paper [6] presents the The presented publications point to the importance and
optimization of the mass of the main girders of an existing justification of considerations of this type of engineering
bridge crane and thus the realized saving was 29 %. problems as well as to the application of most diverse
The paper [7] deals with the analysis and optimization of software packages and analysis procedures.
the whole carrying structure of a double-beam bridge The aim of this research is to use an example of a solution
crane, observing the connection between the end girders of double-beam bridge crane with the rail in the middle of
and the main girders and its influence on the decrease of the girder to present the comparison between analytical
stresses in the structure, while the paper [8] provides the and numerical results and, consequently, draw certain
analysis of a double-beam bridge crane aiming at conclusions and guidelines for the design of this type of
presenting the critical zones of stresses and deflections, structures.
by using Radioss 12.0 and HyperWorks software
packages. Also by applying HyperWorks software, the 2. ANALYTICAL CALCULATION OF THE
research presented in [9] dealt with the multicriteria MAIN GIRDER
optimization of the main girders of a bridge crane, with
the optimizations done by shape, size and topology as The calculation of the main girder will be performed on
well as with the oprimization of the cross section, which the basis of strength, i.e. stress states in the middle of the
resulted in a significant saving in the mass. Varying of girder, below the rail, at the point of maximum stress,
plate thickness and positions of vertical stiffeners was according to [17].
carried out in [10], where optimization of the existing
main girders of the bridge crane, with the mass reduced
by as much as 38%, was done according to the criteria of
stresses, deflections and natural frequencies. SolidWorks
software was used here for the model of the bridge crane
structure.
In contrast to the previous research, where 3D finite
elements were used, shell finite elements were used in the
paper [11], in which the model of a double-beam bridge
crane was prepared in FEMAP software, while the static
and dynamic analyses were performed by using PAK
software.
In the research presented in [12], the model of a bridge
crane with the rail in the middle of the main girder was
compared with two modern variants of the solution with
the rail above the web, based on maximum stresses, by
applying 3D finite elements, where a slight saving
accomplished by new variants of solution of main girders
was shown. Similarly to the previous research, in the Fig.2. Cross section of the box girder with the
master's thesis [13], the author presented the analysis of presentation of stress states in the flange plate below the
the main girder of a double-beam bridge crane with the rail
rail in the middle of the girder, both analytically and
through FEM, and compared it with the variants of the A solution of double-beam bridge crane, with the carrying
girder with the rail above the web, where the advantages capacity of 10 t and the span of 7 m, will be used as an
of these solutions over the existing ones was shown. example. The diaphragms within the girders are
Unlike FEM, analysis and optimization can be performed distributed at a distance of 1 m. The thickness of all girder
in a pure analytical way, which is shown in papers [14- plates and diaphragms is 8 mm.
16]. The paper [14] compares two types of bridge cranes Figure 2 presents the cross section of the main girder of a
analytically, and a concrete example is used to show that double-beam bridge crane with the rail in the middle of
the main girders have a smaller mass if the rail is above the girder, with necessary dimensions, for the observed
the web if compared to those with the rail in the middle of example.
76
2.1. Stress calculation b2  L   z2  bk  (8)

In the analysis of stress states in the box girder, the
influence of local stress which occurs due to the pressure L
of the winch wheel on the rail, when additional stress z2  (9)
components in the z and x directions occur, will be 2
primarily observed. That is the reason why the stress state
will be observed in the middle of the girder (below the qn  1.1    g  A (10)
rail), where two load positions will be observed when
wheel 1 (F1>F2) is in the middle between two diaphragms where:
(since maximum values of local stresses occur at that R - the resulting force in the main girder,
point): position 1 is when the force F1 is in the middle of F1 , F2 - the forces of pressure on the girder, from wheel
the third zone (Figure 3), and position 2 is when F1 is in 1 and wheel 2, respectively,
the middle of the fourth zone, i.e. in the middle of the  ,  - the coefficients, according to [17],
girder span (Figure 4). M 1 , M 2 - the maximum bending moments at the points
z1 and z2, respectively,
qn - the specific weight of the girder (increased by 10 %
due to the weight of the diaphragms and the welds).
Figure 5 shows the position of wheel 1 which is located
between two diaphragms, when maximum local stresses
occur, as well as the influential zone of action of pressure
on the girder upper plate, size z0.
Fig.3. Static model of the girder for load position 1 z0  2  hš  5, cm (11)
Fig.4. Static model of the girder for load position 2
The continuation of the paper will present all necessary

relations which are used for the static calculation of the
girder.
  Q  mk Fig.5. Position of the diaphragms in the interior of the
R   g (1) girder
2
The calculation of values necessary for determination of
b e local stresses is presented in the continuation of the text.
F1  R  k 1 (2)
bk F1
N , a1  b1 (12)
96  b  I Š 1  K1 1
2
F2  R  F1 (3) 1 1

a  t13
1
3
co
F1  z1   L  z1  F2  b1  z1
M1    F1
L L N , a1  b1 (13)
(4) 96  I Š 1  K1 1
qn   L  z1  z12  1 
  a1  t13 co
2
 z ,u   z   zM   d (14)
b1  L   z1  bk  (5)
Re
a d  (15)
z1  2  a1  1 (6) 1
2
F1  L F2  b2  z2 q  L2 M
M2     n (7)
z  (16)
4 L 8 Wx
77
6  K3  N  the rail dimensions: 50x50,

 zM  (17)  the rail material: S355 (Re = 35.5 kN/cm2).
t12
The other data were taken according to the classification
6  K2  N class, according to [17].
 xM   d (18)
t12 2.2. Analytical results
Based on the input parameters, the calculation results in
 u   z ,u 2   xM 2   z ,u   xM  1.1  d (19) obtaining the value of stress in the rail σš = 23.68 kN/cm2,
where: while the permissible stress of the rail material is σš,d =
N - part of the wheel pressure force which, due to the 27.31 kN/cm2.
rail stiffness, is transferred to the plate, Table 1 shows the values of stress components in the
a1 - the distance between the diaphragms, middle of the girder upper plate as well as the equivalent
stress at the point of pressure of wheel 1 on the rail.
b1 - the inner width between the webs of the box girder
(300 mm, Figure 2), Table 1. Analytical values of stresses in the box girder
t1 - the upper flange plate thickness, σz σzM σz,u σxM σu
Case
K1 , K 2 , K 3 , co - the coefficients, according to [17], (kN/cm2) (kN/cm2) (kN/cm2) (kN/cm2) (kN/cm2)
I Š 1 - the moment of inertia of the rail for its axis, 1. 3.78 26.99 32.10
 zM ,  xM - the stress components in the longitudinal and 23.21 35.49
2. 3.89 27.09 32.13
cross directions, due to local bending, respectively,
 z ,  z ,u - the normal stress due to bending and the total
stress in the longitudinal direction, respectively, As it can be seen from Table 1, the normal bending stress
σz has a relatively small value in relation to the local
 d - the permissible stress of the girder material, stresses σzM and σxM, and therefore the emphasis is placed
Re - the yield stress for the basic material of the main on the middle of the girder, on the stress below the rail.
During the design phase and calculation of these girders,
girder,
the stress states of local stresses are not taken into
 1  1.5 - the permissible stress of the girder material, account, so that based on Table 1 it can be seen that the
 u - the equivalent (total) stress in the middle of the exceeding of stress occurs in these components (σd =
15.67 kN/cm2) and also that the total stress is exceeded
girder upper plate. (1.1·σd = 17.23 kN/cm2).
The maximum stress in the rail is determined based on the If a short diaphragm were inserted between the existing
following relation: diaphragms, in each zone, these stresses would
N1  a1 considerably decrease, so that their values would be σzM =
š    š ,d (20) 4.74 kN/cm2, σxM = 6.82 kN/cm2. In this case, the stress
6  Wx , š states would not be exceeded.
N1  F1  N (21)
3. NUMERICAL ANALYSIS OF THE MAIN
Re, š GIRDER
 š ,d  (22)
1.3 3.1. Results from SAP2000
where: Figure 6 presents a hybrid model for the FEM analysis in
N1 - part of the wheel pressure force which is connected SAP2000 software, which consists of a main girder
with the rail, modelled by shell finite elements and a rail modelled by
Wx , š - the resistance moment of inertia of the rail, beam finite elements.
 š ,d - the permissible stress of the girder material,

Re, š - the yield stress for the rail material.
The input data (characteristics of the existing crane)
necessary for the calculation of the girder are as follows:
 the carrying capacity of the crane: Q = 10 t,
 the span of the crane: L = 7 m,
 the weight of the trolley with the winch: mk = 4.49 t,
 the span of the trolley wheels: bk = 1866 mm,
 the position of the resultant force of load in relation to
the wheel 1 - trolley: e1 = 783 mm,
 the material of the girder plates S235 (Re = 23.5
kN/cm2), Fig.6. Mesh of finite elements of the girder with the rail in
 the classification class of the crane: 2, SAP2000
78
The following figures presents the diagrams of

distribution of bending moments and shear force in the
rail at the point of action of wheel 1 (point with the largest
load) for both observed load positions (Figure 7 and
Figure 8).
Fig.10. Distribution of stress in the x-direction in the

main girder for load position 1 in SAP2000
Fig.7. Static diagrams at the point of maximum load of

the rail for load position 1
Fig.11. Distribution of the total stress in the main girder

for load position 1 in SAP2000
Fig.8. Static diagrams at the point of maximum load of

the rail for load position 2
The following figures present the distributions of stress

components in the main girder for both observed load
positions (Figure 9 – Figure 14).
Fig.12. Distribution of stress in the z-direction in the
main girder for load position 2 in SAP2000
Fig.9. Distribution of stress in the z-direction in the main

girder for load position 1 in SAP2000 main girder for load position 2 in SAP2000
79
Fig.16. Maximum stress in the rail for load position 1 in

Fig.14. Distribution of the total stress in the main girder Autodesk Nastran In-CAD
for load position 2 in SAP2000
Table 2 presents the maximum values of stress

components in the main girder for both load positions,
equivalent (total) stress, as well as the calculated value of
stresses in the rail based on the maximum bending
moment and shear force (Mmax,1 = 446.81 kNcm, Fs,1 =
21.52 kN – for load position 1 and Mmax,2 = 447.36 kNcm,
Fs,2 = 21.54 kN – for load position 2).
Table 2. Values of stress in the box girder and the rail in

SAP2000
Case σz,u σxM σu σš

(kN/cm2) (kN/cm2) (kN/cm2) (kN/cm2) Fig.17. Detail of the presentation of the maximum stress
in the rail for load position 1 in Autodesk Nastran In-
1. 17.62 31.65 27.47 21.50
CAD
2. 17.78 31.65 27.48 21.52
3.2. Results from Autodesk Nastran In-CAD

Figure 15 presents a hybrid model for the FEM analysis
in the software module Autodesk Nastran In-CAD, which
consists of a main girder modelled by shell finite elements
and a rail modelled by 3D finite elements.
Fig.18. Maximum stress in the rail for load position 2 in

Autodesk Nastran In-CAD
Fig.15. Mesh of finite elements of the girder with the rail

in Autodesk Nastran In-CAD
The following figures present the distributions of stress Fig.19. Detail of the presentation of the maximum stress
components in the main girder for both observed load in the rail for load position 2 in Autodesk Nastran In-
positions (Figure 16 – Figure 31). CAD
80

Fig.23. Detail of the distribution of stress in the x-
main girder for load position 1 in Autodesk Nastran In-
direction in the main girder for load position 1, at the
CAD
point of action of wheel 1 in Autodesk Nastran In-CAD
Fig.21. Detail of the distribution of stress in the z- Fig.24. Distribution of the total stress in the main girder
direction in the main girder for load position 1, at the for load position 1 in Autodesk Nastran In-CAD
Fig.22. Distribution of stress in the x-direction in the Fig.25. Detail of the distribution of the total stress in the
main girder for load position 1 in Autodesk Nastran In- main girder for load position 1, at the point of action of
CAD wheel 1 in Autodesk Nastran In-CAD
81
Fig.29. Detail of the distribution of stress in the x-

CAD
Fig.30. Distribution of the total stress in the main girder

for load position 2 in Autodesk Nastran In-CAD
Fig.27. Detail of the distribution of stress in the z-

Fig.31. Detail of the distribution of the total stress in the

main girder for load position 2, at the point of action of
wheel 1 in Autodesk Nastran In-CAD
Table 3 presents the maximum values of stress

components in the main girder for both load positions,
equivalent (total) stress and the value of stress in the rail.
Table 3. Values of stresses in the box girder and the rail

in Autodesk Nastran In-CAD
Case σz,u σxM σu σš

(kN/cm2) (kN/cm2) (kN/cm2) (kN/cm2)
1. 13.42 20.10 17.86 30.94
2. 15.67 21.96 20.39 30.45
CAD
82
4. CONCLUSION ACKNOWLEDGMENT
This research led to the comparison between analytical A part of this research is a contribution to the projects of
results obtained on the basis of [17], for the calculation of the Ministry of Education, Science and Technological
stress states in the middle of the main girder with box Development of the Republic of Serbia, Project III 44006
cross-section, as well the results obtained by FEM, in two and Project TR 35038.
different software packages, according to different
models.
Based on Table 1 and Table 2, it is seen that the model in REFERENCES
SAP2000 software shows quite good results regarding the
stress state of the box girder in the x-direction (cross [1] Zhang, Q.; Pei, C. & Guo, J. (2012). Development of
direction), while the value of stresses in the z-direction the Traveling Crane Main Girder CAE Platform
(longitudinal direction) is considerably smaller than the Based on ANSYS, Proceedings of 2012 International
analytical value. Also, the value of stresses in the rail Conference on Measurement, Information and
obtained in this way is approximate to the analytical Control (MIC 2012), May 18-20, Harbin, China,
value. The application of the method of modelling in this ISBN: 9781457716010, pp. 1067-1071, Institute of
software package is thus justified and the analytical model Electrical and Electronics Engineers (IEEE), New
is confirmed regarding the components of wheel pressure York City
force and its distribution on the rail and the box girder [2] Paresh, P. & Kamdar, N. (2014). Design and
upper plate. Analysis of Major Component of 120 Tones Capacity
Based on Table 1 and Table 3, it is seen that the stress of EOT Crane. IJEDR, Vol. 2, No. 2, pp. 2150-2154,
components in the box girder in both directions are ISSN: 2321-9939
considerably lower than the analytical values, and based
[3] Khalidurfeasif, P. & Bharti, D. (2015). Finite
on Table 2 and Table 3, it is seen that the stresses in the
Element Analysis Of 450T EOT Crane Box Girder.
upper plate are lower than those obtained according to the
IJAERD, Vol. 2, No. 11, pp. 171-178, ISSN: 2348-
previous model. The stresses in the rail are considerably
6406
higher than those obtained analytically and the stresses
obtained in the previous model. The cause can be found, [4] Sushma, K. & Malviya, A. (2017). Study & Design
before all, in the method of modelling the rail itself and its Analysis on Girder of EOT Crane. IJARIIE, Vol. 3,
connection with the box girder, as well as in the No. 5, pp. 1441-1445, ISSN: 2395-4396
introduction of the load itself, along the upper surface of [5] Ning, Z. (2012). Structural Optimization Research on
rail. Girder of 200t Bridge Crane Based on ANSYS.
In this case, SAP2000 provides more precise results, Advanced Materials Research, Vols. 430-432, pp.
primarily because the modelling method corresponds 1708-1711, ISSN: 1662-8985
more to the analytical model of calculation. The software [6] Suratkar, A.; Shukla, V. & Zakiuddin, K. (2013).
module Autodesk Nastran In-CAD is simpler and more Design Optimization of Overhead EOT Crane Box
practical in engineering work regarding design and Girder using Finite Element Analysis. IJERT, Vol. 2,
analysis of these types of structures. Possible No. 7, pp. 720-724, ISSN: 2278-0181
modifications of the model are more easily performed in [7] Sankar, V.; Pravin, A. & Suresh, P. (2018). Analysis
Autodesk Inventor software, which also holds for a new and Design Optimisation in the Structural Member of
analysis in this module, which is not the case with the a Double Girder Overhead Crane. IJPAM, Vol. 118,
application of SAP2000, where modifications are much No. 11, pp. 211-217, ISSN: 1311-8080
more complicated. [8] Linear Static Analysis of a Double Girder EOT Crane
It can be noticed that the stresses in the plate obtained by Assembly http://www.gowra.com/resources.html -
FEM are within the limits in both software packages if Gowra Engineering Technologies Pvt. Ltd., Case
compared to the results obtained analytically, which means Study
that these procedures can be successfully used as reliable
[9] Yifei, T.; Zhaohui, T.; Wei, Y. & Zhen, Y. (2013).
ones in the design of these types of carrying structures.
Research on Energy-saving Optimization Design of
Also, it can be noticed that proper placement of additional
Bridge Crane. Eksploatacja i Niezawodnosc –
short stiffeners between diaphragms can considerably
Maintenance and Reliability, Vol. 15, No. 4, pp.
decrease the value of local stresses in the girder upper
449–457, ISSN: 1507-2711
plate.
Future research can deal with the analysis of local [10] Bećirović, A.; Vukojević, D. & Hadžikadunić, F.
stability of box girder plates, both the upper flange plate (2011). Optimization of Double Box Girder
and the webs, as well as with the influence of placing Overhead Crane in Function of Cross Section
horizontal stiffeners on the webs. The inclusion of other Parameter of Main Girders, Proceedings of XV
criteria and technological restrictions is desirable for the International Research/Expert Conference Trends in
purpose of optimization of all geometrical parameters of the Development of Machinery and Associated
the box girder. The other types of models in FEM Technology (TMT 2011), September 12-18, Prague,
analysis, the influence of selection of the type of finite Czech Republic, ISBN: 1840-4944, pp. 641-644,
elements and the method of load introduction can also be Faculty of Mechanical Engineering of University of
used in order to obtain as realistic model as possible in the Zenica, Bosnia and Herzegovina
analysis and design of such types of structures.
83
[11] Miloradović, N.; Vujanac, R. & Slavković, R. (2006). Girder Bridge Crane with the Rail Placed in the
Modeling of the Steel Structure of Overhead Middle of the Top Flange, Proceedings of IV
Traveling Crane using Numerical Methods, Conference on Mechanical Engineering
Proceedings of IV Symposium with International Technologies and Applications (COMETa 2018),
Participation Konstruisanje Oblikovanje Dizajn November 27-30, East Sarajevo – Jahorina, B&H,
(KOD 2006), May 30-31, Palić, Serbia & ISBN: 978-99976-719-4-3, pp. 343-350, Faculty of
Montenegro, ISBN: 86-85211-92-1, pp. 57-60, Mechanical Engineering East Sarajevo, Republic of
Faculty of Technical Sciences of University of Novi Srpska
Sad, Serbia [16] Pavlović, G.; Gašić, M.; Savković, M.; Bulatović, R.
[12] Sankar, V.; Vijayan, D. & Ashraf, I. (2015). & Zdravković, N. (2014). Optimization of the Box
Reducing the Structural Mass of a Real-World Section of the Main Girders of the Bridge Crane for
Double Girder Overhead Crane. IJAET, Vol. 8, No. the Case of Placing the Rail in the Middle of the Top
2, pp. 150-162, ISSN: 22311963 Flange, Proceedings of VIII International Conference
[13] Lerga, J. (2011). Modeling and Simulation Research Heavy Machinery (HM 2014), June 25-28, Zlatibor,
on the Metal Structure of Bridge Cranes, Technical Serbia, ISBN: 978-86-82631-74-3, pp. 105-112,
University - Sofia, Master thesis Faculty of Mechanical and Civil Engineering
[14] Kumar, A. (2012). Economic Design of EOT Cranes, Kraljevo, Serbia
Alard College of Engineering - Pune University, [17] Ostrić, D. & Tošić, S. (2005). Cranes. Institute for
Seminar Report Mechanization of the Faculty of Mechanical
[15] Pavlović, G.; Savković, M.; Marković, G. & Engineering of the University in Belgrade, ISBN: 86-
Zdravković, N. (2018). Mass-Optimized Design of 7083-520-7, Belgrade, Serbia
the Main Girder with Box-Section of the Double-
84
DOI: 10.24867/MD.11.2019.3.85-94
Preliminary note
MODELS IN SAFETY MANAGEMENT

László POKORÁDI *
Óbuda University, Institute of Mechatronics and Vehicle Engineering, Budapest, Hungary
Abstract: In our age safety, reliability and security are essential issues in modern engineering. Modern technical
systems should meet technical, safety and environmental protection requirements. For decision making the experts need
the opinions of other sciences, sometimes they have to consider moral questions. The paper gives a short overview of
mathematical models and tools for safety management that researched by the Author and his followers.
Key words: engineering; safety; reliability; risk; mathematical modelling;
1. INTRODUCTION But, what are the safety, reliability and risk? Let us see
their definitions:
For the modern technical management it is indispensable Safety is the freedom from hazards that can cause death,
to apply the mathematical modelling, as quantity models injury, or illness in humans; adversely affect the
in decision-making. Nowadays, the investigation of the environment, or cause damage to or loss of equipment or
uncertainty has become researchers’ area of interest in property [1].
technology, especially in that area, where some measured In technical terms, reliability is defined as the probability
data are needed to use. that a product performs its intended function without
The maintenance management can be resolved into its failure under specified conditions for a specified period of
element e.g. implement management, logistic, reliability time [25].
theory, maintenance theory; however sometimes the last Risk is the potential for shortfalls, which may be realized
one includes all of the others. in the future, with respect to achieving explicitly stated
Horváth summed in his paper the maintenance performance requirements. The risk should be defined as
management schools, according to schools keep focus on the combination of the probability of an event and its
their investigating, which are the next: organizational-, consequences [1].
condition-, quality-, reliability-, process and mathematical
A. Safety
schools [3]. The mathematical school provides help with
mathematical descriptions to decision-making. The Safety management, from the systematic point of view,
Author is a Hungarian representative of this school. deals with artificial systems that exhibit the following
The aim of this paper is to show theoretical background features:
safety, reliability and risk management; mathematical
 they are integrated;
modelling and to summarize the most important issues
(models and methods) of the Author’s and his followers’  sliding
scientific researches.  the technical systems usually are integrated, large and
The remainder of this article is structured as follows: complex;
Section 2 shows the main definitions and ideology of  mutual interdependence between system components
safety, reliability and risk management. Section 3 presents is significant;
theoretical background of mathematical shortly. Section 4  system inputs are subject to stochastic stimulation;
shows a new modular approach method to determine  the technical systems are partly automated.
sensitivity of reliability of systems with complex Safety engineering deals with preparation of decisions
interconnections. In the Section 5 a Monte-Carlo dedicated to technical and/or organizational matters that
Simulation-based method to investigate maintenance include [23]:
process is demonstrated. Section 6 describes a modified
fuzzy rule-based risk assessment method. The closing part  analysis of the problem with consideration of all the
of the paper is a summary in Section 7. essential internal and external interrelationships;
 definition of desired targets for the problem solution
2. SAFETY, RELIABILITY AND RISK with respect to the required system safety level;
 disclosing of all the possible options for the possible
All human activity involves any risk. In our age safety, final solution;
reliability and security are essential issues in the  comparison between specific solutions and rating
engineering. Modern technical systems should meet
them in terms of detrimental effect against the
technical, safety, security and environmental protection
required expenditures;
requirements.
*Correspondence Author’s Address: Óbuda University, Donát Bánki Faculty of Mechanical and Safety Engineering, Népszínház str.
8, 1081 Budapest, Hungary, pokoradi.laszlo@bgk.uni-obuda.hu
László Pokorádi: Models in Safety Management;
 selection of the most advantageous (or less The homologue model is similar to the real system
disadvantageous option). geometrically. For example, wind-tunnel model of aircraft
The risk is the combination of the likelihood and the is a homologue model. The analog model does not need to
consequence of a specified hazard being realized [12]. It be similar to the real system, but its principle has to be
is important to mention that only in case of quantitative analog with the process occurring on the real modeled
(so called hard) risk assessment method, the risk is system. As a rule, an analog computer (for example
product of probabilities and calculated “crisp” severity of “classical” robot pilot) means an analog model.
investigated hazard. This method can be used when these In the safety management practice the most widespread
factors can be determined unequivocally. For example models are the mathematical ones. The mathematical
insurance companies use it. model is the mathematical equation or system of
equations which describes the internal principles of the
B. Risk process occurring on the system from the point of view of
The risk management is the systematic application of the given investigation [13].
policies, practices, and resources to the assessment and On the basis of the characters of mathematical equations
control of risk affecting human health and safety or the which describe the behavior of the system or their
environment [12]. The experts have to decide on moral determination, the following mathematical models can be
base, during decision-making. It is very important to use distinguished in pairs [19]:
risk management methods to investigate and to minimize A. Static – Dynamic
risks which concomitance of human activity. A critical
role of the safety regulator is to identify activities The mathematical model will be a static (stationary) one
involving significant risk and to establish an acceptable if the state of the system can be described by algebraic
level of the risk. In most cases, the near zero risk should equations or differential equations which do not contain
be not achievable can be very costly. derivatives with respect to the time (see Section 5).
A dynamic (non-stationary) mathematical model
C. Reliability describes the changes of the system parameters depending
Reliability has a broad meaning in our daily life. In our on time. They can be vulgar or partial differential
daily life, reliability has a broad meaning and often means equations. It is possible that the changes of the system
dependability. parameters can be described by not only on a time
First we define terms that relate to the definition of interval, but on a transformed domain chosen practically.
reliability: B. Linear – Non-linear
 binary state: the function of a product is either success A linear model consists of only variables or their
or failure (see canonic systems); derivatives multiplied coefficients. A linear mathematical
 multistate: the function of a product can be complete model can be a linear or linearized equation or a system
success, partial success, or failure (see fuzzy dule- of equations (see section 4.2).
based systems an methods); The non-linear mathematical models are free from the
 hard failure: this is a catastrophic failure that causes requirement mentioned above. At least one of the
complete cease of a function; equations which describe process occurs in the system is
 soft failure: this is partial loss of a function that non-linear.
occurs in a multistate (degradation) product. C. Continuous time – Discrete time
The focus of risk management is the identification and In case of continuous parameter (continuous time)
treatment of risks. It is a process that includes risk- mathematical models variables take values at every
informed decision making and continuous risk instant of time.
management in an integrated framework. The mathematical model is a discrete parameter (discrete
In the safety, reliability and risk management the application time) one if its variables are defined only at particular
of decision-preparing methods based on mathematical instants of time
modelling is indispensable for decision- makers.
D. Continuous state-space – Discrete state-space
3. THE MATHEMATICAL MODELS If the variables can have all value continuously, the
mathematical model is called a continuous state-space one
A system is a collection of devices that operate on input (see Section 5).
signal x(t) (or excitation) to produce an output signal y(t) A mathematical model is considered to be of a discrete
(or response). A system may also be regarded as a state-space one if the values of the dependent variables
mathematical model of a physical process that relates the constitute a count finite or non-finite set.
input signal to the output signal [22].
By the NASA handbook [9] model is a description or E. Deterministic – Stochastic
representation of a system, entity, phenomenon, or In case of a deterministic mathematical model, the
process. In other words, it is a simplified copy of the real dependencies of output parameters on input parameters
system which emphasizes its important characteristics and can be described unambiguously in some time internal or
neglects secondary characters are not determinant ones domain of the independent variables.
from the point of view of the investigation. The abstracted Stochastic (random) mathematical model includes random
model can be analog, homologue and mathematical one. functions which can describe interdependencies between
86
dependent and independent variables of the examined f ( x1 ; x2 ; xn ) xi y xi

system (see Sections 4 and 5). K xi  K i   (1)
xi f ( x1 ; x2 ; xn ) xi y
Obviously the list mentioned above of the kinds of the
mathematical models is not complete, of course.
Mathematical model realized in the engineering practice This derivative-based approach is very efficient from a
means the synthesis of the above mentioned ones. computational standpoint; however, it does have one
serious flaw. Derivative-based approaches are only valid
F Model Uncertainty at the point that they are computed. This is acceptable for
During mathematical modelling we can meet any type and linear systems but would be of little value for systems
rate model uncertainty. exhibiting nonlinear behavior.
By the NASA handbook [9] the uncertainty is an From safety management’s point of view the sensitivity
imperfect state of knowledge or a physical variability analysis of mathematical model can show what elements
resulting from a variety of factors including, but not should be correct (by engineering decisions) to get the
limited to, lack of knowledge, applicability of maximal improvement of system or process reliability and
information, physical variation, randomness or stochastic safety or the maximal decreasing of cumulative risk (see
behavior, indeterminacy, judgment, and approximation. Section 4.).
On the one hand the real systems and processes are
precise, but complex. Additionally the large-scale and 4. RELIABILITY OF CANONIC STRUCTURES
complex systems consist of large number of inter-related WITH COMPLEX INTERCONECTIONS
subsystems. On the other hand, the mathematical model
should be simple therefore may be imprecise. The study reliability of systems has become important in
Mathematical modelling and simulation of complex risk management. The system having no simple
technical systems must include the non-deterministic interconnections is called System with Complex
features of the modeled system and its environment or Interconnection (SwCI) or complex system. These
human interaction with the system. systems cannot be simplified to single equivalent
These non-deterministic features mean that the response of the elements or blocks by a combination of series and parallel
system cannot be predicted precisely because of the existence reductions. Therefore the “traditional” (Fault Tree
of uncertainty in the system or the environment [11]. Analysis – FTA and Reliability Block Diagram – RBD)
One of the most widely recognized distinctions in methods cannot be used to determine their reliability. One
uncertainty types is between aleatory and epistemic ones. approach to computing the reliability of the SwCI is the
Uncertainties are characterized as epistemic, if the summation of the probabilities of all the operating states
modeler sees a possibility to reduce them by gathering of the system. A table listing the probability of each
more data or by refining models. possible state and its consequences for a system is
The epistemic uncertainty derives from some level of frequently referred to as the truth table.
ignorance of the physical process, the system or the Myers [7] focused on reliability modelling of complex
environment. Experts use the term epistemic uncertainty multichannel systems, such as the digital fly-by-wire
to describe any lack of knowledge or information in any aircraft control system. According to Iordache [4], a
phase of the modelling and model application. This type complex system is composed of subsystems characterized
of uncertainty may be comprised of substantial amounts by an emergent behavior from nonlinear interactions
of both objectivity and subjectivity. among the subsystems resulting in multiple levels of
The aleatory uncertainty is an inherent variation organization.
associated with the investigated system or its Pokorádi showed the adaptation of linear mathematical
environment. It is also called as variability, irreducible, diagnostic modelling methodology for setting-up of
random uncertainty. Aleatory uncertainty is primarily Linear Fault Tree Sensitivity Model (LFTSM) [15];
associated with objectivity. For example, this type Linear Sensitivity Model of System Reliability
uncertainty can be investigated by Monte-Carlo (LSMoSR) and Linear Sensitivity Model of System
Simulation (see Section 5). Unreliability (LSMoSU) [16]. The sensitivity analysis
It is important to mention that usage of linguistic shows how sensitive the output parameter is while
variables – for example during risk assessment – is a changing in any elements of the input parameters. The
source of aleatory uncertainties (see Section 6). LFTSM is a modular approach tool that uses matrix-
G Model Sensitivity algebraic method based upon the mathematical diagnostic
methodology of aircraft systems and gas turbine engines
The sensitivity analysis shows how sensitive the system’s shown by Pokorádi [14].
output parameter is while changing in any elements of the
input parameters. In other words the sensitivity coefficient 4.1. Reliability of the SwCIs
characterizes measurement of the investigated output
system parameter’s dependence upon given input one. To demonstrate the proposed sensitivity determination
Using the mathematical model of investigated system or method, firstly let us determine the reliability parameters
process relative changing of output parameters can be of the system shown by Figure 1.
determined in case of changing of input one (or ones). By The system shown in Figure 1 has five elements, A; B; C;
literatures [15] and [16], the sensitivity coefficient by D and E. Their reliability can be characterized by
input parameter xi of general function y = f(x1, …, xn) reliability ri i  L and probability of failure: pi i  L ,
f :  n   can be determined analytically by where L is set of Latin characters A; B; C; D; E.
87
In case of operating state an item is performing its equation (1), ui  pi  1  ri , then sensitivity coefficient
required function. If an item is in non-operating state, it is is:
not performing its required function. Sum of probabilities
ri n
of operating and non-operating should be unity: K ji    uk . (6)
Q j k 1
pi  ri  1 (2) k i
Table 1. Possible states of the system in Figure 1

j Element
System Qj
state
A B C D E
1 0 0 0 0 0 0
2 1 0 0 0 0 0
Fig.1. Investigated system (Source: [7]) 3 0 1 0 0 0 0

4 1 1 0 0 0 0
One of the approaches to correctly compute the
unreliability of SwCI is the summation of the 5 0 0 1 0 0 0
probabilities of all non-operating system states of the 6 1 0 1 0 0 0
investigated system. A table listing the probability of each
possible state for a system is frequently referred to a truth 7 0 1 1 0 0 0
table. The elements and the complete system can be 8 1 1 1 0 0 0
operating (designated as the 1 state) or non-operating ones
(designated as the 0 state). 9 0 0 0 1 0 0
The possible system states are summarized in the form of 10 1 0 0 1 0 1 rA p B pC rD p E
a truth table, shown in Table 1, with each element being p A rB pC rD p E
11 0 1 0 1 0 1
assigned either an operating or a non-operating state. The
Q j j   column lists the probabilities of each of the 12 1 1 0 1 0 1 rA rB pC rD p E
system states. Since the table covers all of the possible 13 0 0 1 1 0 0
combinations, the sum of all of the state probabilities
14 1 0 1 1 0 1 rA p B rC rD p E
should be 1.
15 0 1 1 1 0 1 p A rB rC rD p E
The state probabilities resulting in an operating system are
included in the rows 10; 11; 12; 14; 15; 16; 19; 20; 21; 16 1 1 1 1 0 1 rA rB rC rD p E
22; 23; 24; 26; 27; 28; 29; 30; 31 and 32. The
17 0 0 0 0 1 0
Rsys  Q10  Q11  Q12  Q14  Q15  Q16  Q19  Q20  Q21  Q22  Q23 
 Q24  Q26  Q27  Q28  Q29  Q30  Q31  Q32
(3) 18 1 0 0 0 1 0
sum of the operating system state probabilities included in 19 0 1 0 0 1 1 p A rB pC p D rE

this column is the reliability of the system. 20 1 1 0 0 1 1 rA rB pC p D rE
4.2. Creating Sensitivity Model 21 0 0 1 0 1 1 p A p B rC p D rE
The theoretical method of setting up LSMoSU can be 22 1 0 1 0 1 1 rA p B rC p D rE
read in references [16] and [17] in detail. The
probabilities of possible system states (see right column 23 0 1 1 0 1 1 p A rB rC p D rE
Table 1) can be described by
24 1 1 1 0 1 1 rA rB rC p D rE
n
Q j  f j ( p1 ; p2 ; pk )   ui ( pi ) (4) 25 0 0 0 1 1 0
i 1
26 1 0 0 1 1 1 rA p B pC rD rE
general equation, where the inner function ui ( pi ) can
27 0 1 0 1 1 1 p A rB pC rD rE
have one of the following two forms. The state of element
and system state can be the same or its inverse. 28 1 1 0 1 1 1 rA rB pC rD rE
If state of element is the same as the investigated system
29 0 0 1 1 1 1 p A p B rC rD rE
states, the intrinsic function ui  pi , then the sensitivity
coefficient is: 30 1 0 1 1 1 1 rA p B rC rD rE
K ji  1 (5) 31 0 1 1 1 1 1 p A rB rC rD rE
32 1 1 1 1 1 1 rA rB rC rD rE
For example, the 1st row of Table 1, system state and
states of all elements are non-operating – the system
states are the same. In case of functions that determine directly the
If the state of the element is supplemental of the probabilistic system parameters – see equation (3) –, the
investigated system states, the intrinsic function, seeing sensitivity coefficients can be determined by
88
Qj The vector of the independent parameters:

Kj  (7)
Psys xT  p A p B pC p D p E  (13)
Following the general determinations mentioned above and the vector of the dependent parameters by equation
we can now set up the linear sensitivity models of system (2):
unreliability.
According to references mentioned above the connection Psys 
Q 
between the independent and the dependent parameters  10 
can be described by Q11 
 
Q12 
Ay  Bx (8) Q14 
 
Q15 
Q 
equation, where A   mm and B   mn coefficient  16 
Q19 
matrices are the independent and the dependent 
Q20 

Q21 
parameters, and x   n1 , y   m1 are vectors relative y    (14)
Q22 
changing of independent and dependent parameters, n is Q 
 23 
the number of dependent parameters, m m   is the Q24 
Q 
number of the dependent parameters. Using the  26 
Q27 
 
Q28 
D  A  1B D   m n (9) Q29 
 
Q30 
relative sensitivity coefficient matrix of the investigated Q 
system, the equation  31

Q32 
y  Dx (10)
Oftentimes, during the system reliability investigation
can be used for relative sensitivity investigations [13]. only the probabilistic system parameter is of interesting.
The coefficient matrix of independent parameters: But, the other dependent probabilistic parameters can be
 0 0 0 0 0  important too, from technical point of view. If just the
 1 K10 E 
 K10 B K10C 1 probabilistic system parameter and its sensitivities are
 K11 A 1 K11C 1 K11E 
  being investigated only the first row of matrix D should
 1 1 K12C 1 K12 E 
 1 K14 B 1 1 K14 E  be used relative sensitivity vector d   m1 . In this case:
 
 K15 A 1 1 1 K15 E 
 1 K16 E 

1 1 1
 d 0,9  0,017100 0,026100 0,017100 0,107100 0,107100  (15)
 K19 A 1 K19C K19 D 1 
 
 1 1 K 20C K 20 D 1 
The Fig. 2 shows the relative sensitivities of system
K K 21B 1 K 21D 1 
B   21A  (11) failure probability in the investigated cases.
 1 K 22 B 1 K 22 D 1 
K 1 1 K 23 D 1 
 23 A 
 1 1 1 K 24 D 1 
 1 K 26 B K 26C 1 1 
 
 K 27 A 1 K 27C 1 1 
 
 1 1 K 28C 1 1 
 K 29 A K 29 B 1 1 1 
 
 1 K 30 B 1 1 1 
K 1 1 1 1 
 31 A 
 1 1 1 1 1 
The coefficient matrix of the dependent parameters:

 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
 K 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
 10
  K11 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
 
  K12 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Fig.2. Relative sensitivities of system failure probability
  K14 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
 
  K15 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0
 K
 16
0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0

The relative sensitivity coefficient vectors of system
  K19

0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0

reliability in cases of different rinom nominal values of
  K12 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0
  K 21 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 element reliabilities are shown by Fig. 2.
T
A 
 K 22 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0
 (12)
 K
23 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 From the results of modelling and analysis the following
 
 K 24 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 conclusions can be deduced:
 K 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0
 26 
 K 27 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 A. Conclusions about the Investigated SwCI
 
 K 28 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0
 K 29 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0
A.1 Increasing the reliability of elements, namely a
 
  K 30 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 reduction their failure probabilities, reduce the
 K 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0
 31
 sensitivity of the system
  K 32 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1
A2. The sensitivity of the system is “symmetrical”.
89
In each modeled case the system had the same degree of (RNSP) and Numbers of Failures (NoF) depending on
sensitivity to reliability parameters of the A and C as well required estimating uncertainty.
as D and E components. This conclusion can be deducted
from the symmetric physical structure of system. (In all
modeled situation the reliability parameters of elements
were equivalent. In case of different element parameters
this conclusion can be false.)
A3. The system has more sensitivity to reliability
parameters of “second line” D and E elements.
A4. In the case of the “first line” elements the system
has more sensitivity to reliability parameters of B.
B. Conclusions about the Proposed Method

B1. The proposed method can be used to analyze Fig.3. The graph model of investigated maintenance
sensitivity of the reliability of the SwCI. process
B2. Elements of coefficient matrices can be easily
determined. During the investigated operation process of the equipment
B3. Sensitivity parameters of the possible system states used in a large number four different (A, B, C, D) types
can be used in cases with separate professional component-related failures have been distinguished. The
meaning (for example road traffic or logistical feature of the repairs of equipment (except C type failure)
management). is a long – approximately 45 day (1080 hours) – period
because of logistical matters. This investigation is done
In this case, the Conclusions B-1 – B-4 has to be taken
from point of view of end-user; therefore the repairs are
into account from professional points of view.
characterized by Mean Repair Turnaround Times (MRTT).
B4. The disadvantage of the proposed method is that Additionally it can be established that the time of
the number of possible states of the system grows replacement of faulty equipment is negligible. So these
by a exponential function depending on the times are not taken into account during simulation
number of elements. modelling.
The main data of the failures and their repairs are
5. STOCHASTIC MODELLS AND THEIR included in Table 1. The Fig. 3 shows weighted directed
APPLICATION graph of the process. In the graph, the weights of the
edges show probability densities (failure or turnaround
rates) of changes of operational states.
The maintenance is one of the most important territories of
practical safety engineering. From the mathematical point The failure rate λi is equal to the probability of the ith
of view, the operation of technical systems and equipment failure in a unit time interval given that no failure has
is a discrete state space stochastic process without after- occurred before it [18]. The turnaround rate μj can be
interpreted analogically. The system of differential
effects, so it can be approximated with a Markov-chain. So,
equations of this process that describes the changes in
during this investigation a static, discrete state-space,
time of the probability of staying in different states can be
stochastic mathematical model was used.
Monte-Carlo Simulation (MCS) is a classical simulation determined as
technique. MC is only one particular application of another
dPW
general method, which is applicable in both deterministic   A   B  C   D PW   A PA   B PB   C PC   D PD
d
and probabilistic settings. dPA
  A PW   A PA
There is vast literature on MCS with a high number of d (15)
books and articles. During the last years, many studies have dPB
  B PW   B PB
shown MCS and its applications. For example: book of d
dPC
Newman and Barkema [10], and Kalos and Whitlock [6];  C P1   C PC
d
papers of Gál et al. [2] and Morariu and Zaharia [8]. dPD
 C P1   D PD
At the heart of MC there is a computational procedure in d
which a performance measure is estimated using samples Because of the investigated process is stationary, the
drawn randomly from a population with appropriate differential coefficients of eq. (15) are:
statistical properties. The selection of samples, in turn,
requires an appropriate random number generator. Ideally, dPW dP dP dP dP (16)
 A  B  C  D 0
the generated “random” sequence is a completely faithful d d d d d
software counterpart of the non-determinism underlying the
A further condition of the solution is the
actual process.
The aims of our investigation were applying the Markovian D
model of the investigated maintenance process and to MCS  Pi ( )  1 (17)
i W
determine availability and Required Number for Spare Part
90
probability of event space, where L set of Latin letters W where N is the number of equipment in the system (in the
A B C D) This equation expresses that the object of present case: N = 5000).
operation has to stay only in one of six states (in the Applying expected (nominal) value of the availability (PW
present case, the state space consists of them). The = 0.98072) shown by Table 2, in case of 5000 equipment
methodology of setting up and solving of the model the RNSP is 96. But, this result is “only” the expected
mentioned above can be known profoundly by publication value of RNSP. The probability that more equipment
of Pokorádi [18]. failure occurs at the same time is 0.5 (50%), which is
The model simulation is done from point of view of end- unacceptable for end-users.
user fundamentally. Thus, the most important question is On the basis of the standard normal distribution – in case
the Required Number of Spare Part (RNSP). Knowing of 10% estimating uncertainty (the probability of not
probability of the availability Pw, the RNSP can be having a spare part in case the equipment failures is 0.1):
determined by the following equation: PRNS  mW  1.29 sW (19)
 1  
RNSP    1 N  (18)
 PW  
Table 2. Essential (nominal) data of statistical analysis

Failures A B C D
Mean Time Between Failures MTBF [hour] 183627 162059 152800 179789
Failure Rate λ [hour-1] 5.446 10-6 6.171 10-6 6.545 10-6 5.562 10-6
Mean Repair Turnaround Time MRTT [hour] 1080.8 1081.1 167.13 1079.8
Turnaround Rate μ [hour-1] 9.252 10-4 9.250 0-4 5.983 10-3 9.261 10-4
Fig.7. Histogram of probabilities of the staying in state B-

Fig.5. Histogram of the availabilities type failure
In cases of numbers of failures, – on the basis of the

standard normal distribution – if the estimating
uncertainty is 10%:
PiE  mi  1.29s i (21)
The determined NoFs are shown in Table 3.
Fig.6. Histogram of probabilities of the staying in state A-

type failure
The Number of i-th type Failure (NoFi) can be determined

by the following equation:
 T  Pi 
NoFi   N (20)
 MRTTi 
where T is the length of investigational time (in the Fig.8. Histogram of probabilities of the staying in state
present case: T = 1 year = 365·24 = 8760 hours). C-type failure
91
In the engineering literature, there are many books and

papers dealing with fuzzy rule-based methods supporting
maintenance decision-making and risk assessment. The
present article follows in the spirit of publications of
Zadeh [26]; Tóth-Laufer et al. [24]; Rikalovic and Cosic
[21]; Johanyák and Kovács [5].
The general aim of Pokorádi and Portik was to work out
new fuzzy rule based risk assessment methods which
consists the a Summarized DeFuzzification (SDF) where
the opinion of experts goes through the process till the
composition part during the decision-making without
averaging the experts’ opinions at the beginning of the
Fig.9. Histogram of probabilities of the staying in state
process [20].
D-type failure
For estimators, the use of Fuzzy Rule-based Risk
Table 3. Number of failures Assessment (FRbRA) provides chance the programming
of computer applications in order to help and prepare
i PiE NoFi decisions.
A 5.959 10-3 241 To demonstrate possibility of use of FRbRA method a
B 6.748 10-3 273 case study will be shown shortly.
C 1.265 10-3 329 During risk management analysis of hydraulic systems’
D 6.086 10-3 246 risks of its hazards was assessed by fuzzy rule-based
method. Pending preparation of risk assessment, the
6. FUZZY METHOS IN SAFETY MANAGEMENT experts defined categories of severity, probability
(frequency) and risk. Then definitions, and membership
The experts often apply linguistic variables in risk functions those categories were determined, which are
assessment, gaining other significant sources of shown by figures 1 – 3. Lastly they depicted the Risk
uncertainty. Therefore the application of fuzzy set theory Assessment Matrix (see Table 4)
allows the description of these variables in decision-
making in risk management.
Table 4. Risk assessment matrix
Frequent Likely Occasional Seldom Unlikely
Catastrophic M
EH H
Critical M
Moderate H M
L
Negligible M
EH – Extra High; H – High; M – Medium; L – Low
A. Traditional FRbRA
The traditional fuzzy rule-based risk assessment method
has the next subprocesses: fuzzyfication; inference,
composition and defuzzyfication (fig. 13).
Fig.10. Membership Functions of severity sets
Fig.13. Traditional FRbRA flowchart
In the FUZZIFICATION subprocess, the membership

Fig.11. Membership functions of probability sets functions defined on the severities and probabilities of
investigated hazards are applied to their actual values, to
determine the degree of truth for each rule premise.
In the INFERENCE subprocess, the truth-value for the
premise of each rule is computed, and applied to the
conclusion part of each rule. This results in one fuzzy subset
to be assigned to each output variable for each rule. In the
COMPOSITION subprocess, all of the fuzzy subsets
Fig.12. Membership functions of the risk sets assigned to each output variable are combined together to
form a single fuzzy subset for each output variable.
92
The DEFUZZIFICATION subprocess creates a crisp not able to handle the extremeness which containing
value of criticality level from fuzzy subconclusion set. In important information about an investigated area. So the
this study the Center of Gravity DeFuzzification authors emphasizes a new fuzzy rule-based method with
(CoGDF) method is used: summarizes defuzzification to solve the problem
n B. FRbRA with summarised deffuzyfication
 zi Ai The FRbRA using Summarised Deffuzyfication (SDF)
i 1
YCOG  n
(22) method is the next: one takes the traditional risk
 Ai assessment process till the composition subprocess. In
other words one let the input opinions, where some of
i 1
where: expert teams can have extreme opinion, through
fuzzyfication, inference and composition subprocesses.
n – number of subconclusions n   ;
z  ;
zi – “centre of gravity” of the ith subconclusion i
Ai – area of the ith subconclusion Ai   .
In the assessment period of risk management experts
identified all potential hazards of system operation. Next
step was to determine probability of identified system
operation hazards statistically. Then hazards’ severities
were determined collectively – as average values of
expert team’s opinions. For risk assessment calculation
the PLutorisK V.1.1D software was used which has been
developed by Author.
The traditional FRbRA method has a disadvantage. It is
Fig.14. FRbRA with SDF flowchart (source: by [20])
Table 5. Results of fuzzy risk assessment
Traditional FRbRA with SCoGDF
Input data
FRbRA A team B team

Severity 8.5 8 9
Probability 0,005
Catastrophic 0.50 0.00 1.00
fuzification
Severity
Critical 0.50 1.00 0.00

Moderate 0.00 0.00 0.00
Negligible 0.00 0.00 0.00
Frequent 0.00
fuzzification
Probability
Likely 0.00
Occasional 0.70
Seldom 0.30
Unlikely 0.00
Extra High 0.00 0.00 0.00
Composition
Inference &
High 0.50 0.70 0.70

Medium 0.30 0.30 0.00
Low 0.00 0.00 0.00
Defuzification 5.685 6.182
One collects all this results of. This means one has n – number of sub-conclusions n   ;
converted all extreme opinions to fuzzy risk assessment. zji – “centre of gravity” of the ith sub-conclusion in
Then one should defuzzyfy for all fuzzy risk opinions, if case of the jth opinion z ji   ;
crisp value of risk is needed, where the summarized is Aji – area of the ith sub-conclusion in case of the jth
shown on Figure 5. The definition of Summarized Center opinion A ji   .
of Gravity DeFuzzification (SCoGDF) is in
Table 5 shows the “Pump failure” hazard and
m n intermediate results of its risk Traditional and SDF
  z ji A ji assessments.
j 1i 1
YSCOG  m n
(23) After the complete analysis, we showed its complete
  A ji results to all participators, risk estimators. Their general
j 1i 1 opinions were, the results of FRbRA using SDF are more
equation, where: carefully quantified of joint expertise according to the rate
m – number of opinions m   ; of experts’ group. But they cannot explain this general
opinion exactly, “only” on the basis of their experiences.
93
7. CONCLUSION [9] NASA System Safety Handbook Volume 1, System

Safety Framework and Concepts for Implementation,
The paper shown theoretical background safety, reliability Washington, D.C.: NASA, 2011.
and risk management; mathematical modelling and gives [10] M. E. J. Newman, G. T. Barkema, Monte-Carlo
an overview of the most important models and methods of Methods in Statistical Physics. New York: Oxford
the Author’s and his students’ and followers’ scientific University Press Inc. 1999.
researches in the Virtual Lab Virtual Lab of Process & [11] W. Oberkampf, et al., “Error and Uncertainty in
System Modelling and Workshop of Technical Safety Modelling and Simulation”, Reliability Engineering
Science – μβσ. & System Safety, vol. 75. pp. 333-357, 2002.
The Author’s proposed prospective scientific research [12] L. Pokorádi, “Fuzzy Logic-Based Risk Assessment”,
related to this field of applied mathematics and technical AARMS, Academic and Applied Research in
management decision making includes the study of Military Science, Vol. 1 (1), pp. 63–73. 2002.
methodologies of complex system uncertainty using other [13] L. Pokorádi, “Introduction to Mathematical
mathematical tools. Diagnostics”, Debreceni Műszaki Közlemények, vol.
2007 (1), pp 65-80. 2007.
ACKNOWLEDGMENT [14] L. Pokorádi, Systems and Processes Modelling,
Debrecen: Campus Kiadó, 2008. (in Hungarian)
The research presented in this paper was carried out as [15] L. Pokorádi, “Sensitivity Investigation of Fault Tree
part of the EFOP-3.6.2-16-2017-00016 project in the Analysis with Matrix-Algebraic Method” Theory and
framework of the New Széchenyi Plan. The completion of Applications of Mathematics & Computer Science,
this project is funded by the European Union and co- vol.1 (1). pp. 34-44, 2011.
financed by the European Social Fund. [16] L. Pokorádi, “Sensitivity analysis of reliability of
The scientific work of the Author has been helped by Systems with Complex Interconnections”, Journal of
Miss Boglárka Molnár and members of Virtual Lab of Loss Prevention in the Process Industries, vol. 32. pp.
Process & System Modeling (Dr. József Bera, Dr. Vahid 436-442, 2014.
Farrokhi, Mr. Csaba Fenyvesi) and Workshop of [17] L. Pokorádi, “Failure Probability Analysis of Bridge
Technical Safety Science – μβσ (Miss Krisztina Tráj, Mr. Structure Systems”, 10th Jubilee IEEE International
Bálint Bera, Mr. Levente Farkas, Mr. Gábor Ágoston, Symposium on Applied Computational Intelligence
Miss Fruzsina Horváth) which is gratefully and Informatics, pp. 319-322. 2015.
acknowledged. [18] L. Pokorádi, Monte-Carlo Simulation-based
Investigation of Maintenance Process,
REFERENCES Repüléstudományi Szemelvények Szolnok, pp. 37-
56. (in Hungarian)
[1] “Doc 9859, Safety Management Manual (SMM)”, [19] L. Pokorádi, R. Szabolcsi, Mathematical Models
Montréal: ICAO, 2013. Applied to Investigate Aircraft Systems. Budapest:
[2] T. Gál, L. Nagy, L. Dávid, L. Vasa, P. Balogh, Műegyetemi Kiadó, 1999.
“Technology Planning System as a Decision Support [20] T. Portik, L. Pokorádi, “The Summarized Weighted
Tool for Dairy Farms in Hungary” Acta Polytechnica Mean of Maxima Defuzzification and Its Application
Hungarica, vol. 10 (8). pp. 231-244, 2013. at the End of the Risk Assessment Process” Acta
[3] Cs. Horváth, “Thoughts about scientifically Polytechnica Hungarica, vol. 11 (3). pp. 167-180,
projections of maintenance, in maintenance keeping 2014.
focus on value – cost – competition” International [21] A. Rikalovic, I. Cosic, “A Fuzzy Expert System for
Conference Proceeding, Veszprém, 2008, pp. 51-58. Industrial Location Factor Analysis” Acta
(in Hungarian) Polytechnica Hungarica, vol. 12 (2). pp. 34-51, 2015.
[4] O. Iordache, Modelling Multi-Level Systems. Berlin: [22] M.N.O. Sadiku, Signals and Systems: A Primer with
Springer- Verlag, 2011. Matlab, New York: Taylor & Francis Group, 2016.
[5] Zs. Cs. Johanyák, Sz. Kovács, “A brief survey and [23] Z. Smalko, “Relations between Safety and Security in
comparison on various interpolation based fuzzy Technical systems”, International Journal of
reasoning methods”, Acta Polytechnica Hungarica, KONBIN. vol. 1(3), pp. 63–73, 2007.
vol. 3 (1) pp. 91-105, 2006. [24] E. Tóth-Laufer, M. Takács, I. Rudas, “Fuzzy Logic-
[6] M. H. Kalos, P. A. Whitlock Monte-Carlo Methods. based Risk Assessment Framework to Evaluate
Second Edition Weinheim: WILEY-VCH Verlag Physiological Parameters” Acta Polytechnica
GmbH & Co. KGaA, 2008. Hungarica, vol. 12(2). pp. 159-178. 2015.
[7] A. Myers, Complex System Reliability. London: [25] G. Yang, Life Cycle Reliability Engineering,
Springer-Verlag, 2010. Honoken: John Wiley &Sons, 2007.
[8] C.O. Morariu, S.M. Zaharia, “A New Method for [26] L. Zadeh, “Fuzzy Sets”, Information and Control,
Determining the Reliability Testing Period Using Vol. 8 (3), pp. 338-353, 1965.
Weibull Distribution” Acta Polytechnica Hungarica,
vol. 10 (7) pp. 171-186. 2013.
94
DOI: 10.24867/MD.11.2019.3.95-100
Research paper
FEA APPLICATION METHOD FOR STRESS TEST OF THE WHEEL FOR MOTOR
AND TRAILER VEHICLES
Eleonora DESNICA1, * - Dejan KLJAJIĆ1 - Mića ĐURĐEV1
1 University of Novi Sad, Technical Faculty “Mihajlo Pupin”, Zrenjanin, Serbia
Abstract: The wheel is historically considered one of the most important achievements in the development of human
civilization thereby bringing radical changes in a person's life, primarily in the field of transport. The wheel has to be
technically functional in terms of safety requirements for motor vehicles (it is necessary to ensure radi-axial throw
within the tolerance limits and provide required wheel load capacity on one side and reduce the total mass of the wheel
on the other side in order to reduce the total mass of inelastic resting mass of a vehicle and achieve the smallest
possible polar moment of inertia). This problem can be solved by modelling the wheel in a 3D environment by
simulating wheel load in one of the popular CAD programs. The paper presents the importance of the Finite Element
Method (FEM) with the focus on its practical application known as the FEA analysis (Finite Element Analysis) which is
used in the design of the wheel of motor and trailer vehicles in the SolidWorks software environment. The simulation of
the load pressure test and analysis of the obtained results are shown. Accordingly, the simulation of the real problem
can be performed which allows time-saving, rapid and easy correction of the identified deficiencies prior to the start of
manufacturing.
Key words: wheel, motor and trailer vehicles, FEA, simulation of stress
1. INTRODUCTION Modern development of motor vehicles and trailers, their

assemblies and components, is based on the application of
The wheels of the motor or trailer vehicles are one of the modern methods of design, modeling, calculations, tests,
most important components that directly affect the safety manufacturing and exploitation. In this paper, the three-
of vehicles in traffic, i.e. performance of the braking dimensional model of the wheel for motor vehicles and
system of a vehicle. On the other hand, they have an trailers modeled in the SolidWorks CAD software is
indirect impact on the performance of the vehicle as a presented [4,5]. After a brief introduction to the types of
whole. The importance of the wheel reflects in the fact wheels, as well as their standard dimensions with the
that the contact with the ground is achieved only through focus on the steel wheels, three calculation methods used
the tire and the wheel itself [1]. for calculating certain wheel elements but not the wheel
The wheel consists of two main parts: a rim with the disk as a whole are presented. Among them, the finite element
method is selected and then formulated in more detail.
(wheels) and an adequate tire. Assuming that this paper
Then, it is later applied for the simulation of steel wheel
will be focused only on the metal part of the wheel, we
load which is previously modeled in the SolidWorks
will address it as "the wheel". Basic function of the
environment. The emphasis is put on the main steps of the
wheels is to absorb the entire weight of the vehicle and FEA analysis such as the selection of materials for wheel
transfer it to the ground as well as to facilitate its manufacturing, definition of forces and loads, simulation
movement and maneuvering. Besides, their role is also to of stresses occurring in the wheel, calculation of the
damp impacts which originate from surface irregularities assembly’s safety factor and finally, generation of
and transfer the load on other parts of the car, thereby detailed reports with analysis of the obtained results.
enabling the uniformity of movement and stability of the After that, guidelines for future developments are given
vehicle on the road. For this reason, harmonization of with the focus on the further generation of process plans
elastic-damping properties of the wheel with the used for manufacturing the designed and analyzed steel
appropriate characteristics of the suspension and braking wheel.
system is very important [2].
By increasing vehicle performances and considering their 2. MAIN TYPES AND CHARACTERISTICS
growing application in different areas of the economy and OF WHEELS
in highly severe conditions, high requirements related to
performance, quality and reliability of wheels are also The wheels of motor vehicles can be classified into steel
increased. A larger capacity is required on one side, and and aluminum wheels. On one hand, aluminum wheels
smaller vehicle mass and polar moment of inertia on the (Figure 1a) apart from their esthetics, have smaller mass
other side. This affect a decrease of inertial brake load and higher strength compared to steel wheels, which also
and improvement of efficiency of the braking system [3]. highlights their performances. The stability of vehicles is
*Correspondence Author’s Address: University of Novi Sad, Technical faculty “Mihajlo Pupin”, Djure Djakovica bb,
23000 Zrenjanin, Serbia, desnica@tfzr.uns.ac.rs
Eleonora Desnica, Dejan Kljajić, Mića Đurđev: FEA Application Method for Stress Test of the Wheel for Motor and Trailer Vehicles;
provided with a lower percentage of twist and a small

torque regardless of starting or braking. Because of the
many advantages, aluminum wheels are now becoming
increasingly popular. They are usually made of aluminum
and magnesium alloys by casting as a single piece.
Considering materials and technologies for their
manufacturing, aluminum wheels can be classified into
several groups: cast wheels, forged wheels, chrome Fig.2. Symmetrical (a) and asymmetrical (b) rim [2]
wheels and Mg-Al wheels [6].
Steel wheels (Figure 1b) represent the standard variant of The disc can have a different axial position, i.e. the depth
wheels that is used for all vehicles. These wheels are of indentation, and therefore, we can distinguish three
sufficiently reliable and cost effective, can withstand high variants of positions which are presented in Figure 3 [3].
stresses and are easy to maintain. Since the steel is very
plastic material with expressed formability,
manufacturing process for steel wheels is easier and
cheaper in comparison to aluminum wheels. Apart from
the large mass, their main disadvantage is twisting during
driving (operation load), which cannot be seen with bare
eyes. Therefore, a vehicle driving at higher speeds and
around curves have less stability and is less safe. In this
paper, emphasis will be placed on the steel wheels [7].
Fig.3. Position of the disc in the rim [3]
Shapes and dimensions of wheels are subject to national

and international standards. Serbian standard that
provides the shape and dimensions of wheels is labeled as
SRPS M.N1.068 and manufacturers are obliged to comply
a) b) with this standard (Figure 4). Wheels are labeled similarly
as tires with two numbers expressed in inches. The first
Fig.1. a) Aluminum wheel [6]; b) Steel wheel [7] number indicates the width and the second indicates
diameter corresponding to the diameter of the tire. To
Steel wheel consists of two elements: the rim and the disc label special wheels (S), of a certain profile type (W-
which can be јоined by a separable joint (screws and shaped) in front of and behind the first number,
eyelets) or inseparable joint (welding e.g.). Inseparable alphabetical signs are placed for the appropriate wheel,
joint is applied in passenger cars, light trucks and buses. e.g. W5,00Sx20 "[8].
Separable joint is applied in heavy cargo and some special
vehicles, as well as in agricultural machinery where there
is a need for changes in the width of tire tracks. Rims
represent trough-profiled wheel rims whose main task is
to hold a tire. They can be: one-piece, separable and split
wheels. There are two basic types of one-piece rims,
symmetrical and asymmetrical (Figure 2). Symmetrical
rims are easier to manufacture, but asymmetric rims are
more suitable when considering complex designs [2].
Steel wheels are usually produced by making steel strip
profiles or by hot-rolling or cold-rolling of steel sheet
with 1.8 mm to 10 mm thickness. Discs are used to
transfer load from tires and wheel bolts down to the hub.
There are different types of discs which depend on the
purpose, sheet thickness (from 3 to 15 mm) or
requirements in terms of the depth of indentation, or by
axial displacement inside the rim. Due to its tendency to
corrode, surface of the wheel is protected with different Fig.4. Rim according to the standard SRPS M.N1.068 [8]
types of coatings that are made by painting, plasticizing,
chrome plating, etc. 3. CALCULATION OF THE LOAD OF MOTOR
In many cases, a steel wheel has through holes for
AND TRAILER VEHICLE WHEELS
ventilation and better cooling which also reduces wheel’s
mass. Modern steel wheels made of quality steel have
Although at the first place the wheel is a relatively simple
much larger ventilation holes and their appearance is
part of a motor vehicle, its calculation is a very complex
similar to aluminum wheels.
96
theoretical task. The level and distribution of stresses in Finite element method (FEM) belongs to the group of
the wheel is not only determined by the structural modern numerical methods. It can be applied for very
characteristics of the wheel and the properties of its complex forms of mechanical, structural and other parts
material, it also depends on a whole range of randomly and assemblies with high static uncertainty which are
changing parameters, such as: the level and distribution of exposed to complex loads. Applying analytical methods
internal stresses determined by a specific manufacturing for these calculations would be ineffective. The FEM is
technology for manufacturing and assembly of the wheel, based on solving the problem of a large number of
highly changing working conditions of a vehicle algebraic linear equations in a matrix form. The basic idea
regarding the conditions and characteristics of a contact of the FEM is decomposition of the continuous mass of
ground, dynamical and design characteristics of a vehicle, 2D or 3D models to discrete (finite) elements [10,11,12].
tire characteristics and the ways of stress transfer [1,9]. Considering that most of the industrial components are
For these reasons, an appropriate method for theoretical made of metal, most of the FEA calculations generally
calculation of the wheel as a whole has not yet been refer to metal components. Their analysis of the load can
developed. So far, there are design methods used for be linear or nonlinear. Linear stress analysis can be used
calculating only certain elements of the wheel as well as in the case of elastic deformation (the body returns to its
experimental - theoretical methods which include: original position after the force is removed). In this case,
 Calculation method using the theory of structures the safety factor is the main objective. In the case of the
 Calculation methods based on random variable loads analysis of the plastic deformation effects on the body,
 Calculation of wheels using finite element method non-linear analysis of loads is used (nonlinear stress
analysis). In this case, the focus is placed on tensile
3.1. Calculation of wheels using finite element strength and permanent deformation of the body. Analysis
method of the non-metallic components made of plastic or rubber
can also be achieved with the non-linear analysis.
The basic principle of calculation using the finite element
method is the following: The continual structure of the 3.2. FEA simulation method and the analysis of
wheel is decomposed into a finite number of elements of the results [13]
plate or beam shape (Figure 5), which have finite
SolidWorks simulation provides the ability to convert a
dimensions (and not of infinitely small) and are connected
to each other in a finite number of points (nodes). The modeled component into a tetrahedral and triangular
elements of mesh, beam or girder. The grid may consist of
arrangement and dimensions of the elements are such that
one or more types of elements. To improve accuracy and
they form the geometry of the structure. The behavior of
each element is described with mathematical dependency performance in certain segments of a component, user can
define and control the local mesh control for points, edges
of the node displacement and the distribution of
and surfaces of that component.
dilatations and stresses within these elements. Such
dependency is derived from the equations of the theory of In this paper, after modeling the wheel in the SolidWorks
environment, the simulation is performed, i.e. calculation
elasticity. Assuming the above mentioned, continual
of the wheel load for motor vehicles and trailers using the
structure is usually reduced to a system with a finite
FEA method. By analyzing the consisting elements of the
number of unknown node point displacements.
wheel, it is possible to determine the critical load forces
For the given load conditions and taking into account the as well as the points at which the deformation of a wheel
characteristics of particular elements and how they are material occurs. This simulation aims to determine
joined into a whole, it is possible to establish a system of whether the modeled shape of the wheel before its
linear algebraic equations by unknown displacements of manufacturing meets the required operating conditions in
all nodes. With this information, the deformed shape of terms of workload and compression. In this way, the
the whole structure is obtained, i.e. its general rigidity. subsequent changes and problems during manufacturing
are eliminated, time from the idea to the finished wheel is
shortened, and cost of the production cycle is reduced.
The result is a more reliable product that can be used with
safety and security [14,15].
Analysis of the load on wheels includes many parameters
that affect the final result - the fulfillment of working
demands for wheels in real operating conditions and the
possibility of further generation of process plans. This
includes the selection of materials for wheel manufacturing,
then, the thickness of the rim and the disk, position of the
Fig.5. Finite elements [1]
disk in the rim, position and size of the opening in the disc,
Using known node displacements and formulations of total weight of the wheel, load across the tire transferred to
individual elements, the distribution of dilatations and the wheel and pressure of the compressed air in the
stresses can be calculated for each of these elements tire. Furthermore, the detailed description of the basic steps
obtaining the complete image at the same time. of the FEA analysis for wheels of motor vehicles and trailers
Development of computer technology ensured these is given: selection of materials for manufacturing, definition
calculations to be easily solved with the help of software of forces and loads, simulation of stresses occurring in the
tools for the simulation of loads within the SolidWorks wheel, calculation of safety factors of the assembly and
environment. generation of detailed reports.
97
 Wheel material - since the steel from which the the as well as complexity. If the components are well-
rim is made Č.0261 and the steel used for producing formed and joined without overlap, the computer will
the disk Č.0361 are similar in mechanical and do the calculation of the mesh that will appear across
chemical properties, that means the wheel is made of the entire wheel assembly (Figure 7).
the same material in the program labeled as
 Simulation of wheel stresses - after completing the
"S235JRG2". By selecting this material from the list,
calculation according to the given parameters, the
all the essential characteristics of this steel: modulus
of elasticity, tensile strength, bulk density and other display will show the image of the model with loads
properties are shown in the table (see Figure 6). occurring due to the pressure and gravitational force
(Figure 8). The colors indicate the level of the load,
wherein the smallest load is dark blue, while the
largest load is red. In the case of steel used for
producing the wheel, the limiting value (yield
strength) is 235 [N/mm2], and is marked in Figure 8
on the right side.
Fig.6. Selection of steels

Fig.8. Simulation results
 Calculation of safety factors of the assembly - Factor

of safety (FOS) is calculated from the submenu by
selecting the icon "Results". The selected upper limit
of the FOS has a value of 2. In this case, all segments
of the assembly that are considered to be safe will be
colored in dark blue, while the segments that do not
meet the safety criteria will be assigned red color. In
this case, the whole wheel model is dark blue which
means that it meets the safety criteria, Figure 9.
Fig.7. Calculation of the mesh
 Definition of forces and loads - it is first necessary to

define a fixed surface that will hold the entire
assembly. In this case, this is a surface of the calotte
through which screws are mounted on the wheel hub
using tensile force. After selecting the fixed surface it
is necessary to mark the outer surface of the rim which
is exposed to the air pressure inside the tire.
Afterwards, the surfaces that will be exposed to
pressure are selected within the program and the Fig.9. Results of the calculation of the factor of safety
pressure value is entered. Since the operating pressure
in the inflated tire is 4.6 [bar], the test pressure which  Generation of reports and analysis of the results -
is 30% higher than the operating pressure is entered in After all calculations are completed, it is necessary to
create a report with all the relevant information related
the table: pi = 1.3 · pr = 1.3 · 4.6 = 6 [bar], i.e. 0.6 to the assembly, forces, and simulation results.
[MPa]. The next step is to form a mesh for calculating
loads of the wheel assembly. Setting the quality of the Figure 10 shows the characteristics of the material from
mesh depends on the type of elements, assembly size which the wheel is made.
98
4. CONCLUSION AND FUTURE RESEARCH
In this paper the possibilities of the CAD software in the

process of modelling the wheel for motor vehicles and
trailers are shown. Based on the requirements for the best
possible performances (smallest possible mass and polar
moment of inertia) as defined by the standards in terms of
dimensions, profile shape, capacity, wheel material and its
application, the 3D model of the steel wheel is formed as
Fig.10. Wheel material the basis for the further calculation. The FEA analysis of
the loads within the SolidWorks environment by
Figure 11 shows the reaction forces that occur in the simulating the real operating conditions in terms of tire
wheel. pressure that is carried on the wheel was performed. As a
result of this analysis, it was found that the modeled
wheel meets the designed level of safety as well as the
required operating requirements and thus its
manufacturing can be performed. Due to high level of
computer development and advancement of available
programming software, this method has become
indispensable in design stages of all branches of industry.
After modeling the wheel, and then obtaining the results
of the wheel load using FEA analysis, the next step in the
research would be directed towards generating process
plans for wheel production. The process planning should
be performed taking into account a minimum number of
operations with a possibility for selecting appropriate
resources, such as machines and cutting tools. Also, the
emphasis should be put on a single production cycle with
the possibility of considering time optimization of the
production cycle.
Fig.11. Reaction forces in the wheel
Figure 12 provides information about the minimum and REFERENCES

maximum load in [N/mm2] as well as the critical points in
which these loads occur. [1] Janićijević, N. & Janković, D. (1986). Developing of
methodologies for calculation of wheel for motor and
trailer vehicles. Faculty of Mechanical Engineering,
Belgrade
[2] Šekeljić, M. (1985). Analysis of existing methods
and test tables for testing the wheels, B.Sc. thesis,
Faculty of Mechanical Engineering, Belgrade
[3] Vilotić, D. (1990). Analysis of the technological and
manufacturing process of gutter like wheels and the
proposed improvements, Faculty of Technical
Sciences, Novi Sad
[4] Lombard, M. (2013). Introduction to SolidWORKS
2013, Computer library, Belgrade
[5] Bethune, J. (2013). Package Design SolidWorks
premium 2013, Computer library, Belgrade
[6] https://www.ronal-wheels.com/int/r41-s, Accessed
on: 2017
[7] https://www.adenstyres.co.nz/products/roh-blak-trak-
steel-wheels-black-price-is-each, Accessed on: 2017
[8] Serbian standard for all shapes and sizes of wheels
SRPS M.N1.068, Accessed on: 2017
[9] Heleta, M. & Cvetković, D. (2009). Fundamentals of
engineering and modern methods in engineering,
Fig.12. Results of the compressive forces Belgrade
99
[10] Letić, D. & Desnica, E. (2011). Engineering [13] Kljajić, D. (2018). Application of the CAD system in
Graphics I and II, Technical faculty “Mihajlo Pupin”, the design of a wheel for motor and trailer vehicles as
Zrenjanin a tool to optimize the production cycle, Master thesis,
[11] Geramitchioski,T. & Trajcevski, Lj. (2015). Technical faculty „Mihajlo Pupin“, Zrenjanin
Dynamic analysis of the sloping bridge using FEM, [14] Stanisavljev, S. (2017). Development of stochastic
Machine design, Vol.7, no. 1, pp. 5-10, ISSN 1821- optimization model cycle time of production in small
1259 and medium-sized enterprises, PhD thesis, Faculty of
[12] Ašonja, A.; Desnica, E. & Palinkaš, I. (2016). Technical Sciences, Novi Sad
Analysis of the static behavior of the shaft based on [15] Antonić, D. (2007). Evaluation of making products
the finite elements method under effect of different using regressive methods, Faculty of Mechanical
variants of load, Applied Engineering Letters - Engineering and Naval Architecture, Zagreb
Journal of Engineering and Applied Sciences, Vol.1,
No.1, pp. 8-15, ISSN 2466-4677
100
DOI: 10.24867/MD.11.2019.3.101-106
Research paper
DESIGN AND OPTIMIZATION OF VEHICLE MUFFLER USING THE FFOWCS WILLIAMS

AND HAWKINGS MODEL
Barhm MOHAMAD *
Faculty of Mechanical Engineering and Informatics, University of Miskolc-Hungary
Abstract: Exhaust and muffler noise are a challenging problem in the transport industry. While the main purpose of the
system is to reduce the intensity of the acoustic pulses originating from the engine exhaust valves, the back pressure
induced by these systems must be kept to a minimum to guarantee maximum performance of the engine. Modern
mufflers typically have a complex system of chambers and flow paths. There are a variety of sound dampening
mechanisms working to quiet the sound flowing through a muffler and piping system. This study presents an efficient
process to optimize the transmission loss of Audi A6 C6 2.0 TDi reactive muffler by using Ffowcs Williams and
Hawkings analytical method. The muffler has a complex partition which is perforate plate located in middle of muffler
chamber. CAD file of the muffler was established for developing FEA model in ANSYS by another commercial
advanced design software (CATIA V5). Transmission loss [dB] was calculated in range of frequency 0-2000 Hz,
sensitivity studies of design parameters were performed to optimize the transmission loss (TL) of the muffler. Software
analysis results are included in the paper. Recommendations are made for obtaining smoother transmission loss curves
for various measurement methods.
Key words: Exhaust system, Reactive muffler, Transmission loss, Ffowcs Williams and Hawkings model, Finite element
method, Noise.
1. INTRODUCTION investigation. Mohamad and Amroune 2019 used

computational fluid dynamic (CFD) tools to describe the
The design of a muffler plays an important role in the flow effects on engine exhaust chamber acoustic level, it
overall NVH performance of a vehicle. Generally, the shows the transmission loss of muffler at different
insertion loss and the transmission loss of the muffler are frequency using 1D boost solver. Mohamad et al 2019
used to assess its performance. There are numerous used transfer matrix method (TMM) to perform the
publications available in the acoustic design and the transmission loss of a muffler and the algorithm can be
measurement methods of exhaust mufflers. There are used for other part of exhaust system, the result of their
three types of mufflers available in market; Reflective study of an existing muffler has been compared with
(reactive) type, resistive (dissipative) type and hybrid vehicle level test observation data. The transmission loss
muffler. In first type wave cancelation method is used to has been optimized for new muffler design; other
reduce the noise. This is used for low frequencies. In literatures were played significant rule for validate their
second type absorbing materials are used. The energy is results. Mohamad 2019 studied several new techniques
converted into heat and then dissipated. This type is used through that literature review, the latest development done
for high frequencies. Third type is combination of on exhaust systems with regard to acoustic performance.
reflective and absorptive. We can use hybrid muffler for The objective of this research work to clear basic theory
all range of frequencies. The objective of this paper is to behind flow noise generated as a broadband noise from 0
conduct a sensitivity analysis of commercial muffler on to 2000Hz frequency range. Our analysis is found to be
three key design parameters, i.e. partition location, effective in finding the noise source location and strength
chamber volume variation, and effectiveness of inserting as well in finding the noise generation mechanism. Some
absorbent material in a reactive muffler. Mohamad et al. acoustic software tool has been applied to a variety of
2017 Presented in their technical paper review of muffler exhaust systems.
used in industry, and this review depicts flow and
temperature distribution along the muffler ducts. The 2. METHODOLOGY
techniques for different methods used in the design,
calculation and construction of muffler both 2.1. CAD model
experimentally, practically and transmission loss The geometry was implemented based on Audi A6 C6 2.0
characteristics were described. 1D calculations are much
TDi reactive muffler prototype using CATIA V5
faster, and still give a good overview of the system under
advanced design software, including inlet, outlet and
*Correspondence Author’s Address: University of Miskolc, Faculty of Mechanical Engineering and Informatics, 3515 Miskolc -
Egyetemváros, Hungary, pywand@gmail.com
Barhm Mohamad: Design and Optimization of Vehicle Muffler Using the Ffowcs Williams and Hawkings Model;
chamber. The perforate plate was placed in the middle of as ambient condition and assigned at the outlet as relative.
the rectangular shaped chamber of the muffler. The cross The details in schematic below.
section and the dimensions of muffler is explained in
Figure 1.
Fig. 2. Modeling process scheme
The standard atmosphere air state, i.e. pressure 101325 Pa

and temperature 298.5 K, was set as the free-stream
condition, which was specified at outlet. The inlet
turbulent degree is set as 0.01. The non-slip, adiabatic
condition was set at the wall
 Compute the aerodynamic field with CFD analysis
(Ffowcs Williams and Hawkings),
 Coupling a quantity using Ansys optimizer,
 Run the acoustic computation.
Fig.1. The cross section and the dimensions of muffler
2.2. Software setup

Several procedures were made to implement the modeling Fig.3. Total volume elements was 9.45 million, surface
using computational fluid dynamics tools and optimum mesh quality equal to 60 degrees and skewness Fluent
design for case study have been created using Ansys solver mesh quality < 0.9
topology optimization. The boundary conditions
considered for the flow acoustic analysis are 2.3. Specifying acoustic receivers
representative of the exhaust system advancing at
In the Acoustics Model select Define Receivers and then
constant engine speed. At the inlet into the geometry the
set Number of Receivers to 2. Enter the coordinates as in
velocity (u) and the turbulence quantities k and ϵ were
the below figure. Receiver-1 and Receiver-2 are located at
assigned. Inlet velocity 55.5 m/s, Inlet pressure 102300
the inlet and outlet boundary, respectively.
pa, mass flowrate 0.05 kg.s-1 and the pressure p consider
Fig.4. Boundary conditions and configuration of software

102
2.4. Postprocess the sound pressure signals ready to be analyzed using ANSYS Fluent’s FFT tools.
To plot Sound Pressure Level (dB) at each receiver, you
Using the FFT Capabilities; Once the sound pressure
can use from below figure 5.
signals are computed and saved in files, the sound data is
Fig.5 Postprocess the sound pressure signals
3. NOISE PROPAGATION COMPUTATION Introducing acoustic density perturbations =
3.1. Ffowcs Williams and Hawkings (FWH)

formulations (3)
Ffowcs Williams and Hawkings (FWH) formulations:

Equation (3) indicate inhomogeneous wave equation in
The approach of splitting the flow and sound fields from
turbulent source field.
each other and solving for them separately can be
The Lighthill’s equation is formally exact: it is derived
simplified further if the receiver has a straight,
from the Navier-Stokes equations without any
unobstructed view of each individual point that is a source
approximation called Lighthill’s tensor.
of noise. Sound transmission from a point source to a
receiver can be computed by a simple analytical (4)
formulation. The Lighthill acoustic analogy provides the
mathematical foundation for such an integral approach. Where Tij is the Lighthill stress tensor, t is flow time, Xn
The Ffowcs-Williams and Hawkings (FW-H) method is length component, ρ is the density, p is the pressure, un
extends the analogy to cases where sound sources and is is the fluid velocity components, ij is the Kronecker
the most complete formulation of the acoustic analogy to symbol, and τij is the viscous stress tensor (Gardner et al
date. Both methods are implemented in FLUENT (Cherng 2015) and (Mohamad 2019).
et al 2015). The FW-H function is as follows:
The acoustics wave equation can be derived from
classical Navier-Stokes conservation of mass and Navier-
Stokes equations:
Mass (1)
(5)
Momentum (2) In equation (5), is the Dirac function, and H(f) is

the Heaviside function.
103
The value of time step size determines the highest The procedure for computing acoustic power using the
frequency that the acoustic analysis reproduces. The FW-H acoustics model in ANSYS Fluent consists largely
highest frequency is calculated by the following equation of two steps. In the first step, a time-accurate flow
(6): solution is generated from which time histories of the
relevant variables (for example, pressure, velocity, and
fmax=1/ (2Δt) (6) density) on the selected source surfaces are obtained. In
Where Δt is time step size. By Δt=0.00025s the highest the second step, sound pressure signals at the user-
frequency (fmax) is equal to 2000Hz. Change max iteration specified receiver locations are computed using the source
per time step to 35 and run the calculation again. The data collected during the first step, details in figure 10-11.
calculation is continued to be converged.
Where Δt is time step size. By Δt=0.00025s the highest
frequency (fmax) is equal to 2000Hz. Change max iteration
per time step to 35 and run the calculation again. The
calculation is continued to be converged.
Sound Pressure Level (dB)= 20 log (PMeasured/PReference) (7)
The transmission loss of a muffle is given by the
following equation:
TL=SPLin - SPLout (8)
4. RESULTS
The CFD simulation is performed to obtain the averaged

flow field, with a conventional Reynolds-averaged
Navier–Stokes (RANS) method. Followed that, the noise
source fluctuations are generated by the synthetic
reconstruction of turbulence. Then, the acoustic Fig.7. Velocity streamline
perturbation equations applied on model.
The mean flow is computed by RANS. The mean flow
fields of velocity and pressure are showed in figure 6 The variation of velocity of exhaust gas flow from the
below. inlet pipe to the outlet pipe. The velocity of exhaust gas is
The static pressure distribution is indicated by various increased due to the high pressure in the chamber and
coloured contours and in Fig. 6 the static pressure pushes the exhaust gas out to the atmosphere.
distribution along the muffler.
Fig.8. Surface acoustic power level
Fig.6. Pressure contour
The pressure rise in the expansion chamber is due to stack

up of gas flow as the cross-sectional area between the
chamber and the outlet pipe is changed. However, at the
Fig.9. Surface acoustic power level of ducts
outlet.
104
Surface acoustic power level (dB) on different surface is 80
shown in figures 8 and 9. The maximum generated noise

is about 65 dB and is on exit pipe. 70
Transmission loss (dB)

60
100
50
90 Inlet
40
80
Sound Pressure Level (dB)
30
70
0 500 1 00 0 1500 2000
F req u ency (H z)
60
Fig.12. Transmission loss of the muffler at t=1.57925s

50
40
Based on topology optimization analysis system, the
noise can be further attenuated by reducing inner structure
30
0 500 1000 1500 2000 (diameter of tail pipe) of the muffler. The structural
Frequency (Hz)
performance of the muffler refers to the structural
characteristics of the muffler like installation dimension.
Fig.10. Sound pressure level at the inlet duct at Details in table 1.
t=1.57925s
Table 1. Comparison of parameters before and after
optimization
40 Parameters Original Optimized
value value
30 Outlet Inlet diameter Din (mm) 60 56
Outlet diameter Dout (mm) 65 59
20 Hole diameter Dh(mm) 2 1.7
Sound Pressure Level (dB)
Mean SPL (dB) 90 85

10
0
5. CONCLUSION
-10
The nonlinear acoustic solver is implemented to model
noise generation and transmission from an initial
-20
statistically steady turbulent flow, which provided by
RANS (Reynolds-averaged Navier-Stokes) simulation,
-30
0 500 1000 1500 2000 and to simulate the noise in near field. The radiated far-
Frequency (Hz)
field noise of the muffler was predicted by Ffowcs
Williams-Hawking (FW-H) acoustic analogy. Ffowcs
Fig.11. sound pressure level at the outlet duct at Williams and Hawkings (FWH) acoustic analogy relies
t=1.57925s on the acoustic analogy and takes into account moving
fluid as sound sources. The mechanisms of the vortex and
sound generation were revealed, and results indicate that
the perforated tube muffler has much lower flow induced
Transmission loss is simulated by resolving the pressure noise level. Also, the results show that decreased with the
waves traveling through the fluid, as presented from inlet pipe diameter, the acoustic energy and the sound
figure 12, the transmission loss result of the exhaust pressure level of the flow noise at the end of the inlet pipe
muffler, the sound elimination concerning the exhaust gradually increased. the CAE models should be
orifice noise reached 60 dB at frequency band of (1000- experimentally verified by accurate testing methods.
1500) Hz, which met the sound elimination performance
of the exhaust system.
105
REFERENCES [4] Mohamad, B. (2019). A review of flow acoustic

effects on a commercial automotive exhaust system,
[1] Mohamad, B., Szepesi, G., Bolló, B. (2017). Review Mobility and Vehicle Mechanics, 45, (2), pp.1-4.
Article: Modelling and analysis of a gasoline engine [5] Cherng, J., Wu, W., Ding, P., Hebbes, M., Zhang, H.
exhaust gas systems, 5th International Scientific (2015). Design optimization of vehicle muffler
Conference on Advances in Mechanical Engineering, transmission loss using hybrid method, SAE technical
University of Debrecen-Hungary, pp. 345-357. paper 2015-01-2306, DOI:10.4271/2015-01-2306.
[2] Mohamad, B., Amroune, S. (2019). The analysis and
[6] Gardner, B., Mejdi, A., Musser, C., Chaigne, S.
effects of flow acoustic in a commercial automotive
Macarios, T. (2015). Coupled CFD and vibro-
exhaust system, Advances and Trends in Engineering
acoustic modeling of complex-shaped mufflers
Sciences and Technologies III, Proceedings of the
accounting for non-uniform mean flow effects, SAE
3rd International Conference on Engineering
technical paper 2015-01-2313, DOI:10.4271/2015-
Sciences and Technologies (ESaT 2018), September
01-2313.
12-14, 2018, High Tatras Mountains, Tatranské
Matliare, Slovak Republic, pp. 197-202. [7] Mohamad, B. (2019). A review of flow acoustic
[3] Mohamad, B., Karoly, J., Kermani, M. (2019). effects on a commercial automotive exhaust system -
Exhaust system muffler volume optimization of light methods and materials. Journal of Mechanical and
commercial passenger car using transfer matrix Energy Engineering, 3(2), pp. 149-156. doi:
method, International Journal of Engineering and 10.30464/jmee.2019.3.2.149.
Management Sciences (IJEMS), 4, 132-139.
106
DOI: 10.24867/MD.11.2019.3.107-112
Research paper
REPAIR OF THE DAMAGED SCREWED SPINDLE OF A TYPE FSP-160 FRICTION SCREW

PRESS BY REPAIR WELDING
Rade VASILJEVIĆ1, * - Dragan PANTELIĆ2
1 Faculty of Maritime Academic Studies, Serbia
2 Auto-Valve, Užice, Serbia
Abstract: The paper presents the approach to the repair of the damaged screwed spindle of a type FSP-160 friction
screw press. In the previous paper (Maintenance Forum 2018), a brief description of the problem was presented. In this
paper, a description of the problem of repair the screwed spindle in the friction screw press is improved and expanded.
The described repair of the screwed spindle is realized in practice. After the conducted control tests, the screwed
spindle was mounted on the press and the pressure force of the press was tested. Finally the press was put back into
exploitation.
Key words: press, screwed spindle, exploitation, fracture, repair
1. INTRODUCTION initial cracks is possible. These cracks allow that the

working stresses at the local places of the cracks
In this paper describes the repair of the screwed spindle of overcome critical stresses, which leads to the fracture of
the FSP-160 friction screw press of the nominal the spindle during exploitation. The screwed spindle it is
workforce of 1600 kN, which represents the basic made with triple square thread.
structural element of the press for achieving the working
stroke of the deformation and the return stroke, and to
receive all the stresses that occur during operation. The
shorter version of this paper was presented at Conference
"Maintenance 2018" – Reference [1]. Here, an illustration
of a defined problem is supplemented and improved.
Friction screw presses are used for hot and cold shaping
of materials. Figure 1 shows the construction of a friction
screw press [2]. More details about this press are in the
References [2-4].
The problem of material fatigue of mechanical elements is
discussed in the References [4-10]. A screwed spindle is a
typical example of a mechanical element that is prone to
damage due to material fatigue and fracture. Apart from
the fatigue of the material, other factors (e.g. bearings and
their lubrication) affect the integrity and life of the
screwed spindle [11]). Damage to the bearings can lead to
damage to the screwed spindle. Combining the impact of
initial cracks with the impact of damaged bearings can
lead to the failure of the screwed spindle.
The screwed spindle of the friction screw press is an
important structural element that has a function of
achieving an operating stroke for deformation and a return
stroke and to receive all the strain that appear while
working. In other words, the screwed spindle is exposed
to variable load during the exploitation of the friction Fig.1. Friction screw press [2]
screw press. Figure 2 shows the 3D model of the screwed
spindle in functional state. During the exploitation of the press, the vibrations of the
During this operation of the screwed spindle, its screwed spindle appear which are undesirable. The
vibrations appear which are undesirable. In accordance analysis of the load of the screwed spindle and the source
with the load of the screwed spindle and the source of of vibrations indicate that the lower slick part of the
vibration, the lower unthreaded part of the screwed screwed spindle is a critical local site on which the initial
spindle is a critical local spot where the appearance of cracks can occur.
*Correspondence Author’s Address: Faculty of Maritime Academic Studies, Bulevar vojvode Putnika 7, 11000 Belgrade, Serbia,
r.r.vasiljevic@gmail.com
Rade Vasiljević, Dragan Pantelić: Repair of the Damaged Screwed Spindle of a Type FSP-160 Friction Screw Press by Repair Welding;
screwed spindle. Repair is a precisely defined sequence of

operations that should be consistently implemented in
order to restore lost operational ability to worn and/or
broken machine parts. Concretely, it is necessary to
recover the working capability of the broken spindle of
the subject press. In accordance with the defined problem
and examination of the state of the damaged screwed
spindle, a thermal method of repair was chosen.
Concretely, the welding method was chosen.
The repair of the damaged screwed spindle included the
Fig.2. 3D model of the screwed spindle - Inventor following main operations:
 Calculation of the conceptual solution of the welded
These cracks allow working stresses in local cracks to connection;
overcome critical strains, which further leads to spindle  Making of the technical documentation (working
fracture during exploitation. The problem of this nature drawings for the completion of the main-healthy part
has arisen in the company Auto-Valve J-SC from Uzice. of the screwed spindle (1), the extension of the
During the exploitation of the friction screw FSP-160, the screwed spindle (2) and the repaired screwed spindle),
fuse spindle failed, so the pressure gauge (piston) with the on which basis the repair of the press FSP-160 will be
slick part of the screwed spindle remained in the lower performed [12];
end position, while the second part of the screwed spindle  Machining of the main (healthy) part of the screwed
with the flywheel made a return stroke in the upper end spindle at the fracture point in order to make the
position. In other words, the fracture of the screwed connection to the attachment;
spindle of the subject press occurred just above the  Machining a screwed spindle extension;
bearing after 15 years of use. In accordance with the  Heat treatment a screwed spindle extension [12];
previously defined problem, the following subchapters  Forming a "V" groove with defined dimensions prior
provide an engineering approach to the realized repair of to reparative welding (Figure 4); The connection of
the screwed spindle of the friction screw press type FSP- the healthy part of the screwed spindle (1) with the
160 of the nominal workforce of 1600 kN. screwed spindle extension (2) was achieved with a
small overlap (ø50H7/n6);
2. VISUAL EXAMINATIONS OF THE  Connecting by repair welding the healthy part of the
SCREWED SPINDLE AFTER FAILURE IN screwed spindle with the screwed spindle nut (Figure
EXPLOITATION 5). The welded joint is realized by welding of middle
quality [5];
After the failure in the exploitation of the screwed spindle  Fine machining of the welded screwed spindle
of friction screw press FSP-160, an analysis of its state (reduction of the extension of the thread to the final
was carried out. References [6-10] were used as the basis. dimensions) [12] (Figure 6); and
As the first step, in order to determine the state of the  Inspection of the performed repair.
damaged screwed spindle of FSP-160 friction screw Machining of the defined operations is carried out on a
press, a preliminary visual examination was carried out. type HN-500 lathe with a 2000 mm spikes spacing. The
As the second step, the condition of the screwed spindle welding surfaces (groove cones) on the healthy part of the
was tested. The preliminary visual examinations of the screwed spindle (1) and its extension (2) are formed by a
state of the damaged spindle of the press pointed to the mechanical process. Machining of the cone of the groove
presence of serious faults (discontinuity) of the type of is performed by turning and milling. Machining of the
cracks, fractures and deformations on the surface of the hole ø50×50 mm at the end of the main part of the
screwed spindle. The concentration of these faults was at screwed spindle (1) is carried out by boring and turning.
the lower end of the screwed spindle which is associated The machining of the cylindrical surface ø50×50 mm of
with the pressure gauge. Some of the mentioned faults the extension of the screwed spindle (2) is carried out by
(cracks, deformations and breaks) are shown in Figure 3. turning.

Fig.3. Faults on the screwed spindle



3. CONCEPTUAL SOLUTION FOR REPAIR



OF THE DAMAGED SCREWED SPINDLE
Taking into account how serious the damage to the

screwed spindle end (which is attached to press piston) is,
Fig.4. Forming a "V" groove: 1 – main (healthy) part
the conceptual solution for the repair of the damaged
screwed spindle, 2 – extension of screwed spindle
screwed spindle is to do reparative welding of the
108
minimum. Welding technology defines all the basic

elements of welding and inspection of welded joints.
From the aspect of ensuring the required quality is
obtained during the execution of welding works it is
necessary to strictly follow the provisions of the welding
technology, which are made according to ISO
(International), EN (European) and SRPS (Serbian)
standards in the field of welding.
4.1. Basic material
The screwed spindle of the friction screw press is made of
high quality steel with a finely processed self-propelled
Fig.5. 3D model of repaired screwed spindle after
screw. Basic material (old material) i.e. screwed spindle
connection of healthy part (1) and extension (2)
H162×3×40 made of steel for improving 50CrMo4 has
the following chemical composition:
After reparative welding, the machining (turning and
abrading) of the screwed spindle extension to the final
Table 1. Chemical composition of basic material [13]
dimensions is carried out.
Element C Mn Cr Mo Si P S
Percent-
0.5 0.65 1.05 0.22 0.25 0.03 0.04
age [%]
The steel 50CrMo4 belongs to a group of improved steels.

Since this steel has a carbon content of C=0.5% (Table 1),
which is greater than 0.3-0.4% C, this means that it has
poor weldability [14].
The mechanical properties of the basic material are:
Table 2. Mechanical properties of basic material [12]

Fig.6. 3D model of repaired screwed spindle after
machining the extension on the final dimensions Stretching Tensile
Dilatation
Mech. range strenght Hardness
Amin
Prior to the realization of repair of the screwed spindle of properties Re Rm [HB]
[%]
press, the following preparatory activities are carried out: [N/mm2] [N/mm2]
 Disassembling the friction screw press in order to Value 600 850÷890 13 248÷266
remove the damaged screwed spindle [2];
 Cleaning of all surfaces of the screwed spindle with
For the extension of the screwed spindle (2) (Figures 4
degreasing agent in cold conditions FAMIN ECO, as
and 5), the same material was chosen as the one used for
well as cleaning from asphalt and resinous particles
the original screwed spindle [14] i.e. for healthy part (1).
that are collected in the threads of the spindle and the
corners of the screw; Type of weld
 Providing the equipment for the processing of a
According to the given conceptual solution for the repair
healthy part of the screwed spindle and making the
of the screwed spindle, the connection of a healthy part of
extension;
the screwed spindle and attachment is achieved through a
 Providing the equipment for joining by welding a
butt welded joint. The healthy part of the screwed spindle
healthy part of the screwed spindle and extension;
and the extension lie in the same plane, which means they
 Providing the equipment for gas preheating and
are faced to each other. Since the thicknesses of the parts
heating of welded joints and contact measuring of the
which are being connected are larger (s=(ø115-ø50)/2
temperature;
mm) the V- groove is applied.
 Providing fire resistant material to cover the welded
joint after welding;
 Forming the team for the doing the repair of the 4.2. Welding process
screwed spindle. The following factors were analysed in the selection of
the welding process of a healthy part of the screwed
4. WELDING TECHNOLOGY spindle to the extension:
 Quality and dimensions of the basic material,
In addition to the described possible repair of the screwed  Quantity and quality of the screwed spindle,
spindle of the friction crimped press, the possible welding  Available equipment, and
technology is described. Inevitable companions of repair  Geometric complexity of the screwed spindle.
are additional strains and deformations [9]. Optimal In accordance with these factors, the REL procedure
application of all of the procedures is sought while (manual electric arc welding) was applied. Welding was
repairing (for example, welding). This reduces the carried out with the appliance Varstroj VARUS 400
influence of additional strains and deformations to a manufactured by Gorenje. Welding devices have been
109
previously checked from the aspect of working of the suitability for exploitation in three phases, as
environment safety and from the aspect of maintaining the follows:
required welding parameters. Electrical characteristics A. Inspection before welding,
must be reliably maintained and, if necessary, changed B. Inspection during welding, and
with the possibility of their reading and control. C. Inspection after welding.
Inspection before welding covers:
4.3. Additional material  Insight into the quality assurance of basic, additional
While selecting the additional material for the manual and auxiliary materials,
electric arc welding process of a healthy part of the  Checking the geometrical dimensions of parts of a
screwed spindle and the extension, the following factors healthy screwed spindle and extension,
were analysed:  Checking the geometric shape of the groove and joint,
 Quality and dimensions of the basic material, and  Checking surface cleaning,
 Complexity of construction and welding position.  Checking devices and equipment for welding,
EUWELD 303, by manufacturer "EUWELD" [15] has  Checking the professional ability of the welders,
been selected for the manual electric arc welding process.  Certificate of the qualification of the welders (SRPS
EUWELD 303 is a special high alloyed austenitic-ferrite EN 287-1),
electrode for welding of hard-to-weld, easy-to-weld  Checking the measures taken to ensure safety at work,
steels, high-strength steel, armor steel, springs steel,  Verification of the prescribed welded joint welding
spindles and similar. It is especially applicable for technology and material quality (SRPS EN ISO
welding of various types of steels (e.g. unalloyed or 15614-1),
alloyed structural steels with high alloy chromium and  Insight into the documentation on the contractor's
chromium nickel steels, for direct welding or overlapping authorization to perform welding work (SRPS EN ISO
welding of manganese steel as well as welding with other 3834).
steels). That is an electrode which successfully weld Inspection during welding covers:
steels of an unknown composition and origin. She solves  Checking the application of the prescribed elements of
many welding problems. Welding with EUWELD 303 technology (welding parameters, welding mode).
does not require a special technique, it is sufficient to Inspection after welding covers:
apply the same technique as with stainless steel electrodes  Dimensional inspection of the screwed spindle as a
to determine that manipulation is easier than with other whole and control of the geometric shape of the
types. The welded joint and clean metal of the seam made welded joint - 100%,
by the EUWELD 303 electrode, regardless of the high  Visual control of welded joints - 100%,
content of chromium, prevents carbon diffusion and  Magnetic or penetrant inspection of welded joints in
hardening in the heat-extinguishing zone, which prevents the range of 100% for quality level B and 10% for
the formation of cold and hot cracks or porosity while quality level D (SRPS ISO 5871),
welding cold and oiled parts during repair. The metal  Radiographic or ultrasonic inspection of welded joints
weld seam done by EUWELD 303 electrode does not in the range of minimum 75% for quality level B
change the characteristics to temperature of 1180 °C. (SRPS ISO 6520).
Mechanical properties of pure metal seams are (Table 3): After welding, no heat treatment is required to eliminate
residual stresses.
Table 3. Mechanical properties of additional material
(filler metal) [15] 5. TECHNICAL AND ECONOMIC ANALYSIS
Stretching Tensile It was necessary to decide whether to repair the damaged
Dilatation
Mech. range strenght Hardness screwed spindle or to obtain a new screwed spindle (new
Amin
properties Re Rm [HB] spare part). For this aim, a technical and economic
[%]
[N/mm2] [N/mm2] analysis of the justification of the repair of the damaged
Value 680 900 40 220 screwed spindle was carried out.
5.1. Technical analysis
The used parameters of manual electric arc welding The technical indicator of the justification of the repair of
procedure are: the screwed spindle is analysed, i.e., the efficiency of
 Type of welding current, AC ("+" or "-") or DC, safety against dynamic fracture due to fatigue while push
 Welding current strength, I=40…150 A, the welded joint SD.
 Voltage of the arc, U=20…32 V. In order to determine the technical justification of repair
The metal seam was created using the EUWELD 303 in the form of reparative welding, it is necessary to
electrode which does not react to aggressive calculate the welded joint of the conceptual solution of
environments. The exceptions are hydrofluoric acid, the reparative welding of the screwed spindle. The
hydrochloric acid and zinc chloride. The select electrode connection between the main – healthy part (1) and the
is not necessary to be dried before use. extension (2) of the screwed spindle was achieved by
4.4. Inspection welding using a butt V - bevel groove (Figure 7), which is under
the action of pressing force loaded on the pressure. At this
During the repair work, the welding of the screwed location, the screwed spindle is not action by turning and
spindle of the press is carried out with a final assessment bending moment.
110
SD 
    DM  2.68  1.25...1.5 (9)
 
The efficiency of safety (7) is greater than the required
minimum Smin=1.25…1.5 [5,16], so that the cross section
of the bevel groove is well dimensioned and there is no
danger of fracture occurrence due to fatigue in the cross-
section of the bevel groove. In other words, by reviewing
the technical indicator there is justification for reparative
welding of the damaged screwed spindle.
5.2. Economic analysis
The economic indicators of the justification of the repair
Fig.7. Welded joint of the healthy part (1) and extension of the screwed spindle such as the price of the new
(2) of the screwed spindle of the friction screw press screwed spindle Cn,NV and as the price of the repaired
screwed spindle Cr,NV were analysed. Purchase price of
The starting data for the welded joint calculation are: the new screwed spindle from MIN FAM J-SC Nis [15]
 Maximum pressing force, Fmax=1600 kN, on 05.12.2016. was 3.715,00 EUR. The price of the
 Material main - healthy part (1) and extension (2) of repaired screwed spindle was 477,00 EUR. The ratio of
the screwed spindle, steel 50CrMo4, the price of the new screwed spindle and the price of the
 Diameter of the screwed spindle at the place of the repaired screwed spindle RCNR is equal:
welded join, d=115 mm,
 Thickness bevel groove, a=27.5 mm, Cn ,NV 3.715, 00
 Quality of bevel groove, middle - C, RCNR    7.8 (10)
Cr ,NV 477, 00
The number of changes cycle in the stress in the working
life welded joint, nΣ>ND≈3·106. The working stress on the Taking into account the ratio between the purchase price
pressure of the welded V - bevel groove is: of the new screwed spindle and the price of the repaired
screwed spindle (10), there is justification that the
F 1600 103  N  damaged screwed spindle is repaired by repair. By this
   189.947  mm 2  (1) way, savings are significant. To support the accepted
Ag 8419.125
decision, there is a fact that repair of the screwed spindle
can be carried out in its own facility within 4 working
d 2
Ag 
4
1  2   8419.25  mm 2  (2) days, while the deadline for the production of a new
screwed spindle is 30 working days given by the company
MIN FAM. Also, there is a need for the friction screw
d  2a
  0.434783 (3) press to be fixed as soon as possible in order to avoid
d significant halt and delay in the manufacturing process
The critical stress or dynamic endurance on the pressure and the final delivery of valves. The justification of the
of the welded V - bevel groove is: repair of the damaged spindle from an economic point of
view is even greater because Auto-Valve J-SC posses
    DM (4) great number of friction screw presses. The basis for the
examination of the economic aspects of repair was also
 D ( 1) M  N 
 DM   509.091  (5) found in Reference [18].
1  cot   tan  M  mm 
2
Finally, based on the technical and economic analysis of
the justification of the repair of the damaged screwed
 N  spindle, it is concluded that there is a technical and
 D ( 1) M   D ( 1)  1   2  280  (6) economic justification for the repair.
 mm 
2
 6. TESTING THE PRESS AFTER REPAIR OF

cot   sr  0.5 (7) THE SCREWED SPINDLE
g
In accordance with Sections 3 and 4, the repair of the
   damaged screwed spindle of the FSP-160 friction screw
tan  M  1  1  2 D ( 1)   1   2  0.9 (8)
  D (0)  press is carried out the using method of reparative
 welding.
Dynamic endurance for stretching of the standard tube of After completing of all inspections of the repaired
50CrMo4 amounts to: σD(-1)=400 N/mm2, σD(0)=700 screwed spindle, the friction screw press was installed,
N/mm2 [16]. The correction factors taken under the which included the installation of the repaired screwed
influence of the stress concentration caused by the welded spindle. Afterwards, trial tests were carried out on the
joint and bevel groove failure are: ξ1=1, ξ2=0.55 [5,16]. operational force of the press, which confirmed that the
The efficiency of safety against dynamic fracture due to repair of the screwed spindle was successfully performed.
fatigue while pressing the welded connection is: After this press was put back into exploitation (Figure 8).
111
Proceedings of Maintenance Forum 2018, May 24.-

26, Belgrade, ISBN 978-86-84231-43-9, Vasić, B.
(Ed.), pp. 1-6, Institute for research and design in
commerce & Industry, Belgrade
[2] Mašinska industrija Niš (***). Frikcione zavojne
prese FZP-80 i FZP-160 - Uputstvo za rukovanje i
održavanje, Niš
[3] Musafija, B. (1988). Obrada metala plastičnom
deformacijom, Svjetlost, ISBN 86-0100520-9,
Sarajevo
[4] Miltenović, V. (1999). Mašinski elementi, Mašinski
Fig.8. Repair of screwed spindle in exploitation fakultet, ISBN 86-80587-12-5, Niš
[5] Ognjanović, M. (2000). Konstruisanje mašina,
7. CONCLUSION Mašinski fakultet, ISBN 86-7083-389-1, Beograd
[6] Ristivojević, M. & Ristivojević, M. (2013).
This paper presents the realized repair of the screwed Ispitivanje mašinskih konstrukcija, Zavod za
spindle of the friction screw press type FSP-160 of the udžbenike i nastavna sredstva, ISBN 978-86-17-
nominal workforce of 1600 kN. In order to reduce the 18268-5, Beograd
influence of additional strain and deformation to the [7] Ergović, I. (1989). Sanacija pukotina zubnog vijenca
minimum during the repair of the screwed spindle, the mlina cementa reparaturnim zavarivanjem.
optimal application of the procedures was sought. Održavanje mašina i opreme, Vol.18, No. 5, (299-
By applying the described repair, the screwed spindle of 303), ISSN 0350-1647
the subject press was successfully repaired. The screwed [8] Vukanović, D; Sabolić, V; Antić, M. (1989).
spindle was repaired to working condition i.e. the press Popravka navarivanjem i zavarivanjem oštećenih
was re-enabled to further exploitation. It was first pogonskih vratila valjačkog stana. Održavanje
confirmed by trial tests of the work force of the press. mašina i opreme, Vol.18, No. 5, (304-3307), ISSN
Finally, the confirmation that the repair of the screwed 0350-1647
spindle has been successfully realized is the fact that the [9] Savković, M.; Gašić, M; Arsić, M. & Petrović, R.
press with the reconditioned spindle is in operation for (2011). Analysis of the axle fracture of the bucket
two years without any problems. wheel excavator. Engineering Failure Analysis,
A technical-economic analysis has shown that there is Vol.18, (433-441), ISSN 1350-6307
justification for the repair of the damaged screwed [10] Domazet, Ž.; Krstulović-Opara, L; & Stupalo, M.
spindle. The paper emphasa the economic significance of (2005). Fatigue cracks and failures in cement
the repair of the damaged screwed spindle. Application of industry, shipbuilding and power plant facilities.
repair welding of the screwed spindle of the press has Engineering Failure Analysis, Vol.12, (819–833),
resulted in significant direct and indirect savings in ISSN 1350-6307
comparison to the purchasing of a new screwed spindle. [11] Todorović, M. (2015). Osnovi održavanja, Fakultet
The process of screwed spindle repair was carried out in a inženjerskih nauka, ISBN 978-86-6335-034-2,
much shorter period compared to the delivery period of Kragujevac
the new screwed spindle. [12] Mašinska industrija Niš (1979). Tehnička
In order to reduce the likelihood afresh failure of the dokumentacija frikcione zavojne prese FZP 160 -
screwed spindle in subject press, more adequate Vreteno, Niš
maintenance is recommended. In other words, more [13] Savić, Z. i grupa autora (1992). IMP1 – Inženjersko
adequate lubrication of the bearings is recommended [9]. mašinski priručnik, Zavod za udžbenike i nastavna
Also, routine tests and controls of the press is sredstva, ISBN 86-17-01767-8, Beograd
recommended. By this be reputed to be other presses from [14] Manojlović, B. (1982). Mašinski materijali I,
the family of friction screw presses in the production plan Mašinski materijali, Beograd
of company Auto-Valve.
[15] EUWELD D.O.O (ex YUWELD),
The aim of this paper was to illustrate the process of
http://www.euweld.com, Dodatni materijali za
repair of the screwed spindle. The continuation of this
zavarivanje, Accessed on: 2015
paper will show the following research results:
[16] Savić, Z. i grupa autora (1992). IMP2 – Inženjersko
 Metallographic examination of the surfaces of the
mašinski priručnik, Zavod za udžbenike i nastavna
fracture of the screwed spindle,
sredstva, ISBN 86-17-01767-X, Beograd
 Analysis of the stress state of the repaired screwed
spindle. [17] MIN FAM A.D. (ex MIN), http://www.minfam.com
[18] Vaserman, R. (2012). Kako se štede milioni tehnikom
REFERENCES reparaturnog zavarivanja i tehnikom zavarivanju
mašinskom održavanju, Beoknjiga, ISBN 978-86-
[1] Pantelić, D. & Vasiljević, R. (2018). Repair of 7694-375-3, Beograd
screwed spindle on friction screw press FZP-160,
112
machine design, Vol.11(2019) No.3, ISSN 1821-1259 pp. 81-xx
DOI: 10.24867/MD.11.2019.3.73-80
Research paper
INSTRUCTIONS FOR THE PREPARATION OF THE FULL PAPER FOR THE JOURNAL
”MACHINE DESIGN”
Petar PETROVIC1, * - Szabo MIKLOS2 - Anna KRENCZYK3
1 University of Novi sad, Faculty of Technical Sciences, Novi Sad, Serbia
2 University of Miskolc, Faculty of Mechanical Engineering, Miskolc, Hungary
3 Silesian University of Technology, Dept. of Engineering Processes Automation and Integrated Manuf. Systems, Gliwice, Poland
Received (dd.mm.yyyy); Revised (dd.mm.yyyy); Accepted (dd.mm.yyyy)
Abstract: Instruction for the preparation of the camera ready papers for the journal MACHINE DESIGN is presented.
These pages provide You with an example of the layout and style, which we wish You to adopt during the preparation of
Your paper. Only papers prepared according to these instructions will be published.
Please, make an abstract as short as possible summarizing the paper in max 300 words, but including sufficient detail
to permit a reasonable evaluation.
Key words: up to five words
1. INTRODUCTION bold, italic and in the same line write max five key words
- italic.
This is an instruction for the preparation of the camera
ready paper for the journal MACHINE DESIGN. The Table 1. Editorial instructions
publication is published on English. The paper can be sent
during whole year as a Word document in electronic form
on e-mail: m_design@uns.ac.rs. For more information No Item Explanation
please visit: www.mdesign.ftn.uns.ac.rs.
Use A4 format (210 x 297). Only
2. FORMAT 1. Paper size
this paper size will be accepted.
The text should be written on page A4 format (210 x 297)
The length of paper is not limited,
using fonts Times New Roman, 10 pt, single spacing, two
2. Length but the paper have to be on even
columns (5mm between columns). Margins: top - 20 mm,
number of pages.
bottom - 15 mm, inside - 25 mm, outside - 15 mm, using
mirror margins. Use mirror margins, and leave 20
The length of paper is not limited, but the paper has to be mm as margins at the top, 15 mm
on even number of pages. 3. Margins
at the bottom, 25 mm on inside and
15 mm as outside margins.
2.1. Title
The title of the paper should be written with capital The main body of the text should
letters, 14 point, bold, font Arial Narrow. Leave one blank be written in two columns, each
line (10 point) and write authors’ first name (Arial Page
4. left and right justified. The width
Narrow, 10 point, bold) and last name (Arial Narrow, 10 Layout
of column is 82.5 mm and spacing
point, bold, capital letters) with number in superscript. between columns is 5 mm.
The first author should also be superscripted with “*” and
his correspondence address should be given in footer on
the same way as it is presented in example. The paper The 10 point Times Roman font in
5. Fonts
should be prepared according to the explanations 1.0 line spacing is recommended.
provided in Table 1.
2.2. Abstract 2.3. Style
After authors’ names write the authors’ name of the After one line (section break continuous) write section
institution determined by superscript (Arial Narrow, 9 headings (11 points, bold, capital letters), subsection
point). After two blank lines write “Abstract:” – Times headings and sub-subsection headings (11 points, bold),
New Roman, bold, italic and in the same line continue left justified. Before and after section headings use 10
with text of abstract (italic), not to exceed 10 lines. Skip point space and before and after subsection and sub-
one line and write “Key words:” – Times New Roman subsection headings use 5 point space.
*Correspondence Author’s Address: University of Novi Sad, Faculty of Technical Sciences, Trg Dositeja Obradovica 6, 21000 Novi
Sad, Serbia, ppetrovic@tu.ac.rs
Petar Petrovic, Szabo Miklos, Anna Krenczyk: Instructions for the Preparation of the Full Paper for the Journal ”Machine Design”;
Machine Design, Vol.X(201X) No.X, ISSN 1821-1259, E-ISSN 2406-0666; pp. XX-XX
Inserted picture and diagrams should have a good quality M3

M1
and resolution at least 300 dpi.
At the end, leave two blank lines and write heading P1
“References” (11 points, bold, capital letters). Please
follow the instructions for each type of references, given
on the next page. Please, give a large number of
references. You can cite some already published papers
from References of MACHINE DESIGN and thus You
M2
could increase the rating of our journal.
…
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxxxxx
…xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx Fig.2. Bad quality, distorted figure with too small letters
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx …
…… 4. EQUATIONS
Equations should be placed flush- left with the text

margin and should be preceded and followed by one
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx blank line (5 point).
f
3. TABLES AND FIGURES S (t )    d 3  s
i 1
i (1)
3.1. Tables If they are numbered make sure that they are numbered
consequently. Put the numbers in parentheses flush with
Tables should be presented in the form shown in Table 1 the right- hand margin level with the last line of the
of this instruction. The title of the table should be equation.
numbered and written above the table (10 point, italic).
Before table title use 10 point space and after table title
5. CONCLUSION
use 5 point space. If a table is too wide to fit into one
column (82.5 mm), You can put it in one pagewidth
Please, remember the ‘fine- tuning’: do not end the page
column (170 mm). If a table is too long to fit into one
with a heading; do not start a page with an incomplete line;
page, the table number and headings should be repeated
do not underline any part of the text. Please try to align
on the next page before the table is continued.
columns by horizontal at top and bottom of each page.
Alternatively, the table can be spread over two
The full text papers should be sent during whole year as a
consecutive pages turned 90° without repeating the
Word document in electronic form on journal e-mail
headings.
address: m_design@uns.ac.rs.
3.2. Figures
The title of the figure should be numbered and written
REFERENCES
below the figure (10 point, centered, italic). Before and
after figure title use 10 point space. All figures should be
[1] Spotts, J., Blitzer, W. & Gingrich, T. (2007). Name of
clearly displayed by leaving single line of spacing above
Book in Italics, Publisher, ISBN, Place of Publication
and below them. All notations and lettering should be no
[2] Kazakcci, A., Tsoukias, A. & Park, H. (2010). Name
less than 2 mm high. The use of heavy black, bold
of Paper, Available from: http://www.exact-address-
lettering should be avoided as they will look unpleasantly
of-site Accessed: YYYY-MM-DD
dark when printed. Pictures should be inserted in the
[3] Cohen, B., Milbank, A. & Romney, J. (2009). Title of
paper as JPEG or TIFF format (resolution at least 300
conference paper, Proceedings of xxx xxx, conference
dpi).
data, location, ISSN, ISBN, Randolph, B. (Ed.), pp.
11-19, Publisher, City
M1 M3 [4] Bachmann, A. (2010). Name of paper. Name of
P1 Journal in Italics, Vol., No., (month and year of the
edition) page numbers (first-last), ISSN
[5] Harbaugh, K. & McClain, A. (2008). Name of paper,
In: Name of Book in Italics, Name(s) of Editor(s),
(Ed.), page numbers (first-last), Publisher, ISBN,
Place of publication
M2 [6] *** (2009) http://www.nastit.hu - The National
Statistics Institute, Statistics Year Book, Accessed
on: 2010-05-18
[7] *** (2005) TFOD machine & transmission trends
ISEP Redding Publishing Company Assembly, pg.68,
Fig.1. Good quality figure with clear lettering ISSN: 1250-3176, Vol 38, No.4, April 2010
2

Machine Design Vol.11 (2019) No.3

Încărcat de

Informații document

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Machine Design Vol.11 (2019) No.3

Încărcat de

Drepturi de autor:

Formate disponibile

UNIVERSITY OF NOVI SAD

FACULTY OF TECHNICAL SCIENCES

novi sad, 2019

technical preparation and cover design

Electronic version of journal available on

CIP – Каталогизација у публикацији

MACHINE Design / editors Siniša Kuzmanović, Milan Rackov. - 2009- .-

Siniša Kuzmanović & Milan Rackov

1. Dimensional and Geometrical Tolerances in Mechanical Engineering – a Historical Review

3. Models in Safety Management

Barhm MOHAMAD ............................................................................................................................................... 101

DIMENSIONAL AND GEOMETRICAL TOLERANCES IN MECHANICAL ENGINEERING –

Received (03.04.2019); Revised (10.07.2019); Accepted (12.07.2019)

Key words: Tolerancing, tolerance analysis, GD&T, quality engineering, review

1. INTRODUCTION since first tolerance-related problems arose centuries

The four ages of tolerance engineering

2. A LONG TIME AHEAD - THE AGE OF 4. THE SPECIFICATION OF TOLERANCES

3. THE STIRVING FOR THE „no-go“ upper specification limit USL

The idea of complete interchangeability in the production nominal value N

5. THE INTRODUCTION OF STATSTICS TO Carlo-based tolerance analysis, is essential to ensure

6. THE RISE OF COMPUTER-AIDED optimization using the pseudo-bool programming of [58]

A COMPARATIVE STUDY OF ANALYTICAL AND NUMERICAL ANALYSES OF THE

Received (02.05.2019); Revised (10.09.2019); Accepted (12.09.2019)

1. INTRODUCTION Nowadays, judging by the practice of the leading

2.1. Stress calculation b2  L   z2  bk  (8)

Fig.4. Static model of the girder for load position 2

The continuation of the paper will present all necessary

6  K3  N  the rail dimensions: 50x50,

 š ,d - the permissible stress of the girder material,

The following figures presents the diagrams of

Fig.10. Distribution of stress in the x-direction in the

Fig.7. Static diagrams at the point of maximum load of

Fig.11. Distribution of the total stress in the main girder

Fig.8. Static diagrams at the point of maximum load of

The following figures present the distributions of stress

Fig.9. Distribution of stress in the z-direction in the main

Fig.16. Maximum stress in the rail for load position 1 in

Table 2 presents the maximum values of stress

Table 2. Values of stress in the box girder and the rail in

Case σz,u σxM σu σš

3.2. Results from Autodesk Nastran In-CAD

Fig.18. Maximum stress in the rail for load position 2 in

Fig.15. Mesh of finite elements of the girder with the rail

Fig.20. Distribution of stress in the z-direction in the

point of action of wheel 1 in Autodesk Nastran In-CAD

Fig.29. Detail of the distribution of stress in the x-

Fig.30. Distribution of the total stress in the main girder

Fig.27. Detail of the distribution of stress in the z-

Fig.31. Detail of the distribution of the total stress in the

Table 3 presents the maximum values of stress

Table 3. Values of stresses in the box girder and the rail

Case σz,u σxM σu σš

MODELS IN SAFETY MANAGEMENT

Received (02.03.2019); Revised (24.06.2019); Accepted (26.06.2019)

Key words: engineering; safety; reliability; risk; mathematical modelling;

dependent and independent variables of the examined f ( x1 ; x2 ; xn ) xi y xi

Table 1. Possible states of the system in Figure 1

Fig.1. Investigated system (Source: [7]) 3 0 1 0 0 0 0

sum of the operating system state probabilities included in 19 0 1 0 0 1 1 p A rB pC p D rE

Qj The vector of the independent parameters:

The coefficient matrix of the dependent parameters:

B. Conclusions about the Proposed Method

Table 2. Essential (nominal) data of statistical analysis

Fig.7. Histogram of probabilities of the staying in state B-