Transfer inovácií 31/2015 2015
SIMULATION OF MAINTENANCE AND REPAIR OF MACHINES
Ing. Alena Pešková software Snoopy and Pipe2. It was found out that
doc. Ing. Jozef Svetlík, PhD. their manageability depends on considerable
Technical University of Košice knowledge of simulation technology. Also this led
Faculty of Mechanical Engineering us to choose a simulation tool Insight Maker [4].
Department of Production Systems Condition for work with the program is to create a
Němcovej 32, 042 00 Košice, Slovakia user account. Created simulations can be saved or
e-mail: [email protected] published. Published simulation can be consulted
with other users of the program. The current area of
simulation program is wide. It is possible to find
Abstract published simulations of the economy, banking,
healthcare, supply theory, solutions to production
We present the simulation tool Insight problems, etc. It is also possible to create groups by
Maker in this paper applied to problems of topic of interest where new users can participate.
maintenance and repair of machines .It is also The simulation tool also offers tutorials for quick
suitable for users who do not have programming familiarization with the program. The big
experience. This open software allows advantage is that if we do not have much
communication with the spreadsheets, where experience with this program, we can refer to
simulation results can be processed. We demostrate experienced users (forum, e-mail) when necessary.
the possibilities of Insight Maker in the illustrative Insight Maker is a useful tool especially for those
example of dynamic elimination of machine who work with Excel or other spreadsheets. [7]
failures. We propose to plan a maintenance in the The simulation outputs can be exported in the form
situations when elimination of the failures has of *.csv and then can be worked on in this format.
precedence over planned preventive maintenance.
The results of simulations are graphs that describe
THE SIMULATION MODEL MAINTENANCE
the creation of failures and repairs of machines
AND REPAIR OF MACHINES
during their preventive maintenance.
We demostrate the simulation tool on
Key words: Insight Maker, dynamic simulation,
illustrated example. We limited for just basic
maintenance and repair
elements in Insight Maker, which are necessary for
the simulation of dynamic systems.
INTRODUCTION
For those users of simulation tools who The formulation of the problem
do not have programming experience it is advised Let us consider a situation where it is
to look for the simplest program to eliminate this necessary to dynamically simulate activities during
disadvantage. For example it is possible to use the maintenance and repair of machines in continuous
graphical program Vensim [1]. More articles and work in given monitored period:
work, in which the authors found the It is given the number of repairmen that
implementation of this program to solve the
is necessary for carrying out regular maintenance of
problem of maintenance were abroad created. Let machine and to eliminate the failures. One person
us mention dissertation [2] "Modelling Industrial per day is allocated to maintenance of machines and
Maintenance Systems and the Effects of Automatic
he will carry it out at regular scheduled intervals at
Condition Monitoring", where maintenance is known time. If the worker does not have scheduled
solved from a system point of view. The main maintenance, is inactive unless a failure occurs.
results confirm that the subjective character of the Workers are mutually substitutable. No rest break
failure, available information of machine and for worker is considered. The failures are
repeatability maintenance are the most important
eliminated as soon as possible after the occurrence,
factors in the management of a maintenance and it is known intensity of their occurrence and the
system. Author effectively use Vensim in average time for elimination (random occurrence).
modelling workers allocation in maintenance,
It is considered such strategy between cooperation
elimination of component failures, inspection of activities of maintenance and repair that elimination
preventive maintenance and supply chain of failures takes precedence over preventive
modeling. The disadvantage of the program is that
maintenance, which may be excluded. The
it is commercial, freely available is only for a simulation results should be time development
limited time. indicators:
In paper [3] we investigated possibilities
number of free repairmen,
of simulation renewal and maintenance using Petri
nets. We tried to experiment with open source
94
Transfer inovácií 31/2015 2015
number of implemented repairs of machines, we have to multiply the intensity by this
machines, number.
number of preventive maintenances of
machines.
In the simulation we have chosen
monitored period - 10 days i.e. 240 hours, the
number of repairmen - 5, number of machines - 10,
scheduled interval of preventative maintenance - 4
days (96 hrs.) taking 3 hours, the failure occurence
is generated via Poisson flow with an average gap
between failures - 5 days (120 hrs.), the average
time to elimination of failure is 3 hrs. and it is
modelled by exponential distribution. We see that it
is good to measure time in hours.
1. MODEL OF MAINTENANCE
Fig. 2 Simulation diagram of repair model
Now we can unify these two independent models
into one.
Fig. 1 Simulation diagram of maintenance model
3. MAINTENANCE AND REPAIR MODEL
At first we will create a simple
In this model we will unify maintenance
deterministic model of maintenance with one
with repair and we will additionally assume that
maintenance man. We need for it two elements
also time repair of machine is stochastic and it is
called Primitive in Insight Maker, that is Stock and
modelled by exponential distribution.
Variable . On Fig. 1, we have a diagram that
contains Maintenance such as warehouse where the
failures are cummulated and Interval indicating the
time between two immediately following
requirements for maintenance. In the variable
Interval we have stored constant the size of 96 hrs.
And in variable Scheduled time we have stored
constant 3 hrs. Time calculation of maintenance is a
little bit more difficult. We need to actually achieve
a state, where after each multiple of interval of
maintenance, maintenance of the machine was
carried out. After the simulation we obtain
development of maintenance of one machine,
maintenance is repeated after 96 hours at a time
0,96,192 in range of 3 hours. As we can see, in that
deterministic model the difficulty of its creation is
focused on proposal of diagram graph of primitives
with flows and information links (arrows - oriented
edges) between them.
2. MODEL OF REPAIR Fig. 3 Simulation diagram of maintenance and
We have stochastic flow of machine repair model
failures in case of machine repair, time on their
elimination is constant for a start. Input flow of Maintenance is here affected by number
failures on Fig.2 would be for one machine of machines on which is necessary to carry out
determined by Poisson distribution with intensity preventive maintenance and by the number of free
equal to the inverted value of average length of repairmen. The most difficult place in the diagram
gaps between failures, that we have stored in on Fig.3 is probably condition in flow, where
variable Average gap. But since we have 10 maintenance started. Performing maintenance is
95
Transfer inovácií 31/2015 2015
conditionally depends on the existence of at least Technology, Information and Computer
one free repairman at the scheduled time of Systems in Automation, Espoo, 2004, 122p.
preventive maintenance. Modelling exponential
time of service is analogical to modelling input
flow of failures. After the simulation we get graph
of the development required indicators of activities,
shown on Fig.4.
Fig. 4 Development of monitored indicators of [3] Pešková, A.: Modelling Maintenance and
maintenance and repair model Renewal in Petri Nets, In: Applied Mechanics
and Materials: ROBTEP 2012, 11th
Mentioned graph of the development International Conference on Idndustrial,
indicators can be imported to a file in table form Service and Humanoid Robotics, Štrbské
and then statistically processed. Pleso, Slovakia, 14-16 November 2012, Vol.
282 (2013), pp. 282-286, ISBN 978-
CONCLUSION 303785603-1, ISSN 1660-9336.
In this paper we deal with only with [4] Roe, S. F.: Insight Maker: A General-purpose
dynamic modelling of maintenance and repair of Tool for Web-based Modeling Simulation,
machines at priority repair of machine against Simulation Modelling Practice and Theory 47,
maintenance. In future work, we plan to focus on 2014, pp. 28–45.
analysis of the various strategies of coordination of [5] Kreheľ, R., Straka, Ľ., Krenický, T.:
these activities in terms of economic efficiency. In Diagnostics of production systems operation
case of incorporation of reliability indicators of based on thermal processes evaluation
machines it will be needed to familiarize also with 2013. In: Applied Mechanics and Materials.
simulation tools based on agents, which puts even Vol. 308, p. 121-126. - ISSN 1660-9336.
greater requirements on users. First experience with
the simulation tool Insight Maker persuaded us that [6] Krenický, T., Salokyová, Š., Dobránsky, J. :
it is very good choice for users with minimal Experimentálna analýza v diagnostike
programming knowledge. We appreciate the prevádzky technických zariadení, Košice, TU ,
amount of available examples on the internet from 2015, 204 s. ISBN 978-80-553-2004-5.
which inspiration may be drawn, as well as friendly [7] http://insightmaker.com (cit. 27.6.2015)
forum of Insight Maker users. The fact that it is a
web application independent on the web browser
This work was supported by Project
will please not only beginners.
VEGA n. 1/0124/15 Research and development
of advanced methods for virtual prototyping of
References production machines.
[1] http://Vensim.com (cit. 27.6.2015)
[2] Honkanen, T.: Modelling Industrial
Maintenance Systems and the Effects of
Automatic Condition Monitoring},
dissertation, Helsinki University of
96
