=Paper=
{{Paper
|id=Vol-3838/paper2
|storemode=property
|title=A Discrete event approach to the simulation of selective laser sintering 3D printing large scale production
|pdfUrl=https://ceur-ws.org/Vol-3838/paper2.pdf
|volume=Vol-3838
|authors=Zahra Isania,Maria Pia Fanti,Giuseppe Casalino
|dblpUrl=https://dblp.org/rec/conf/viperc/IsaniaFC24
}}
==A Discrete event approach to the simulation of selective laser sintering 3D printing large scale production==
https://ceur-ws.org/Vol-3838/paper2.pdf
A discrete event approach to the simulation of selective
laser sintering 3D printing large scale production
Zahra Isania1,†, Maria Pia Fanti1,*,† and Giuseppe Casalino1,†
1Polytechnic University of Bari, Via Orabona, 5- 70125 Bari, Italy
Abstract
Various industries, especially construction and building renovation, need custom components in
quantities that traditional 3D printing methods are unable to supply. This article addresses the
challenge of expanding the Additive Manufacturing (AM) production scale, emphasizing careful
consideration in scaling strategies. In particular, it provides insights for optimizing AM factory
productivity through discrete event simulation. The article models the setup and processing
times for a selective laser sintering case study, examining the variability in machine and operator
numbers for cost reduction and increased daily production, in two scenarios.
The first scenario is the reference for comparison with the one proposed in this paper. The
working hours and days, and the type of machines in the production line are the same in both
scenarios. In the proposed scenario, the effect of the number of operators and machines was
studied by FlexSim software and compared with the referenced scenario in terms of the cost,
daily production count, and average waiting time. It offers a practical roadmap for successful AM
production scaling.
Keywords
Additive Manufacturing, 3D Printing, Selective Laser Sintering Machine, Discrete Event
Simulation, Layout Guidelines, Cost Reduction 1
1. Introduction
Additive Manufacturing (AM) and digital fabrication are frequently used in the construction
industry and other manufacturing domains to achieve a variety of objectives, including
waste reduction, cost and time savings, and architectural freedom [1]. However, because
the industry is distinct and conservative, digital technologies have been gradually adopted
as auxiliary tools for conventional, well-established processes. Through improvements in
mechanical, physical, and chemical qualities as well as dimensional precision, AM
technologies are transforming the manufacture of components [2].
VIPERC2024: 3rd International Conference on Visual Pattern Extraction and Recognition for Cultural Heritage
Understanding, 1 September 2024
*Corresponding author.
† These authors contributed equally.
zahra.isania@poliba.it (Z. Isania); mariapia.fanti@poliba.it (M. P. Fanti); giuseppe.casalino@poliba.it (G.
Casalino)
0000-0002-0832-7425 (Z. Isania); 0000-0002-8612-1852 (M. P. Fanti); 0000-0003-1774-4631 (G.
Casalino)
© 2024 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR
ceur-ws.org
Workshop ISSN 1613-0073
Proceedings
Even though AM has historically been good at creating complex geometries[3], a typical
drawback is that it only deposits one material at a time, which results in geometry-
dependent performance outcomes. The dynamic manufacturing landscape is changing
dramatically due to the impact of sophisticated technology, making small-scale production
more affordable and resource efficient. The emergence of AM as a direct production
technique forces businesses to reevaluate existing manufacturing strategies' locations and
approaches [4].
A variety of goals have been the focus of earlier research on AM production planning,
including decreasing production costs [5], delays [6], and processing times [7], as well as
maximizing profit and resource use [8]. Technology must be included in production
planning and control procedures for AM to become industrialized. This will increase output,
reduce costs, and improve product quality [12].
Using a variety of techniques from production planning and scheduling optimization in
conventional manufacturing could help address these issues. Computer simulation and
other optimization methods have shown to be quite beneficial for conventional
manufacturing systems in terms of both design and operation. A significant breakthrough
in dynamic system modeling was the advent of Discrete Event Simulation (DES) modeling,
which made it possible to incorporate real-world unpredictability for production
scheduling [9]. With successful applications in bottleneck identification and manual
assembly [10], where it was combined with operational management to address lot size
issues and lower production costs, DES has demonstrated a high degree of effectiveness in
simulating and resolving complicated queuing problems. Owing to its effectiveness, DES
offers a viable way to analyze and improve AM process planning.
DES has several benefits, but it cannot automatically carry out iterative optimization on
its own. As such, it requires the incorporation of appropriate optimization techniques to
attain more improvements. The advantages of combining DES with optimization tools to
address complicated production difficulties have been shown in conventional
manufacturing environments [11]. Building on this achievement, FlexSim, a DES software,
has the potential to improve AM production system design and planning. It is feasible to
create AM production planning techniques that are more reliable and effective by
combining the advantages of both approaches, eventually improving the capabilities of
additive manufacturing systems [12].
With a particular focus on the Selective Laser Sintering (SLS) process, this research
presents a simulation-optimization approach to handle scheduling and bottleneck
identification challenges in additive manufacturing. Through the application of productivity
increase strategies from conventional production to additive manufacturing, this study
highlights the value of integrating FlexSim with DES. Using this method can help make well-
informed decisions about the configurations of the equipment, which will ultimately save
costs.
The paper is structured as follows: The problem statement in the context of SLS
processing is fully described in Section 2, where it will be thoroughly examined and then
improved. In Section 3, the research methodology is explained in detail along with how it
relates to the findings in Section 4. Section 4 explores the ramifications of these findings.
Section 5 concludes the analysis and makes recommendations for further research.
2. Problem Statement
The main goal of this study is to present a method for increasing production rate by
adjusting layout and parameters. To achieve this, the production system described in [13]
is examined, and the potential applications of a specific mathematical technique are
explored to optimize workstation layouts for the integration of additive manufacturing
technologies, specifically SLS and SLM (Selective Laser Melting). The project comprised
designing a prototype design based on selected 3D printing techniques and a hypothetical
production scenario.
Because 3D printing can create products with intricate patterns and greater production
runs, investor interest in the technology is predicted to increase. Furthermore, the
construction of enterprises focused mostly on 3D printing machinery. As such, it is
imperative to establish clear rules and norms for designing layouts conducive to the
seamless operation of additive manufacturing devices. According to the material in
reference[13], SLS technology is the primary focus of this endeavor.
A hypothetical production schedule was created for the computational study, with an
emphasis on the use of 3D printing equipment to manufacture a certain product type, with
a special emphasis on SLS-produced goods.
3. Methodology
3.1. Reference model
Before making any modifications, FlexSim should first simulate the reference model. The
production plan called for producing 20,000 units of the product a year, according to work
[13], Other presumptions include:
1. The SLS machines' simultaneous printing capacity is eight units.
2. There are 250 working days in a year.
3. There are two production shifts.
4. Working day utilization factor (0.9375), considering work breaks (assumed to be
from 12:00 PM to 12:30 PM).
The recommended number of machines, together with their setup and processing time,
are displayed in Table 1 along with the number of machines that were found to be required
using a particular formula. Additionally, in Fig. 1 the factory's structure is outlined.
The layout plan was meticulously designed to optimize the workstation layout's efficacy;
the plot form allows for product receipt on one side of the production hall and material
supply on the other. This configuration was purposefully designed to allow materials to flow
continuously, increasing overall operational efficiency. The goal of this design is to create a
productive and well-organized work environment that will help to increase the overall
performance and productivity of the manufacturing plant.
A method for representing business processes that understandably incorporate formal
characteristics is the Business Process Represent and Notation Diagram (BPMN). Its visual
approach helps identify issues such as infinite loops, for preserving correctness and
coherence in process descriptions [14]. The preceding layout's description of the SLS
production process is visually shown in Fig. 2, which facilitates comprehension and
improves clarity.
Table 1
Recommended Machinery from ref [13]
Machine Names Quantity Setup Time (min) Process Time (min)
Powder Mixing 2 10 15
Sifter 1 10 5
SLS Device 4 30 320
Cabin Sandblaster 1 10 5
Grinder 2 15 8
Varnishing 1 10 5
Packing Station 2 10 7
Figure 1: Exemplary Layout Plan for 3D Printing machines and devices, modified from
[13].
The production simulation task consists of three basic phases, as follows (Fig. 2).
1. Preprocessing: Precise handling and control of powders for additive manufacturing
are critical in the field of powder management. To maintain uniformity and quality,
handling, preparation, and storage are required. Several machineries are crucial to
the production process in the model setup. The input warehouse serves as the
starting point, receiving and holding raw materials needed to create parts in
accordance with orders from customers. The Powder Selection station is responsible
for carefully evaluating and choosing only flawless powders so that only materials
of the highest quality move on to the next phase. After that, the Powder Mixing
station, also known as a powder blender or mixer, is essential for blending dry
powders uniformly and preserving product quality during the manufacturing
process. The Sifter is an additional crucial station that is accountable for eliminating
foreign objects, confirming the size of the powder grains, and enabling ongoing
quality inspections. It guarantees maximum utilization for the manufacturing
process, breaks up product lumps, and separates coarse materials.
2. Processing: This section focuses on the SLS procedure. SLS Instrument: The SLS
Device is a key part of the manufacturing process. It is designed to work smoothly
with Industry 4.0 environments, making it a production platform that looks forward.
The SLS technique, developed by Desktop Manufacturing Corporation [15], uses a
laser beam to solidify powdered materials layer by layer. A 3D CAD model is loaded,
converted to STL format, and then sliced into layers [15]. Before laser sintering,
plastic powder is placed and heated, causing the powder to fuse selectively by the
cross-section of the model. The product is cleaned and may go through finishing
procedures like varnishing and sandblasting after printing [16]. The use of 3D
printing equipment, such as the EOS GmbH FORMIGA P 110 and related support
equipment, at a recently built facility is the basis of this essay. The SLS technique is
used by these gadgets to operate [17].
Figure 2: BPMN Diagram of the SLS Workflow
3. Post-processing: The equipment used at this stage is essential for the manufactured
parts' packing, protection, and refinement. Abrasive materials are propelled at high
speeds by the Cabin Sandblaster to clean and prepare surfaces. To attain desired
forms and finishes, surface grinders are essential tools for refining SLS-printed
objects. Protective coatings are applied through vanishing to prevent environmental
damage. Careful packaging is required during packing to simplify storage and
guarantee product safety throughout transit. Before being distributed, completed
goods are centralized and stored in the output warehouse. The goal of these post-
processing sub-processes is to improve the final product's quality and appearance
by smoothing out surface flaws, applying protective coatings, and improving the
surface finish on printed parts.
3.2. Modelling in FlexSim software
The reference production line's layout plan, created to satisfy FlexSim software's processing
requirements, is the main topic of this article. 21 processors, 4 temporary storage spaces, 1
pallet temporary storage zone, and 1 input and output warehouse are needed to build the
model. The production line's FlexSim simulation model is shown in Fig. 3. Building the
model and precisely adjusting its parameters are essential to the simulation's success.
Materials in FlexSim are routed to the powder selection buffer from the input warehouse.
The powder is fed into the powder mixing machines and blended after they have finished
setting up. Following the processing time, the mixture is moved to a sifter machine, which
separates the larger grains from the smaller ones. The mixture is then transported to a
buffer because, in this work, the SLS machines have 320 minutes to prepare 8 parts at once.
This is accommodated by putting a buffer before the SLS machine and using FlexSim's
"Round Robin if Available" setting in the buffer's output. Once the parts are ready from the
SLS machines, another combiner is added. To guarantee that eight parts enter the machine,
are created, and exit at the same time, the combiner is configured in "batch" mode in this
phase. Then, another combiner machine is set up, but this time, the options section has the
"pack" option chosen. To finish, distinct separator machines are placed next to each SLS
machine and configured for "unpacking." Because of software constraints, the combiner
cannot directly pick batch mode, and the separator cannot unpack, so this layout design is
required. To accomplish the packing before the separator starts the unpacking process, an
additional combiner is placed in between them.
Figure 3: Simulation Model of SLS production line in FlexSim.
Next, two buffers are added to FlexSim: one for gathering the pallets and the other to
collect the assembled components. After that, these pallets are sent to the warehouse so
they can be used again on a production line. After that, the components move to the cabin
sandblaster, where they go through a procedure using highly accelerated abrasive
compounds. This stage improves the pieces' overall finish by smoothing down uneven
surfaces and removing surface defects. The parts are moved to the grinder station once they
are finished. Here, the SLS pieces' surface polish is improved, any lingering support
structures or build platform attachments are eliminated, and any dimensional errors that
might have happened during printing are fixed. Before proceeding to the next level, this
careful procedure makes sure that the parts fulfill the requirements and quality standards.
Subsequently, the parts are queued to prevent the creation of bottlenecks. After that,
they proceed to the varnishing station to receive a layer of ornamental or protective varnish.
The items are cleaned and, if desired, prepped before coating to ensure an even application.
Before the pieces are packaged for storage or additional processing, thorough quality
inspections are carried out after drying or curing. The components then head to the packing
station. They are packaged here in compliance with standardizing procedures. They are
delivered to the output warehouse for distribution after being confirmed.
3.3. Improving the process of the reference model
As stated before, the primary goal is to determine the optimal machine count by examining
various scenarios to increase production. The purpose of these scenarios is to determine
how many SLS, sifter, cabin, varnishing, and packing machines are needed to reach output
levels that are at least as high as the reference [13].
Before describing the scenarios, it is crucial to determine the feasibility of adjusting the
number of SLS machines. To this aim, the reference layout design is used as the basis for
evaluating the SLS machines' performance. A low machine usage rate raises the potential
for a reduction in the total number of SLS machines. The ideal number of SLS machines
found from this assessment is then used to define the scenarios.
Procedures for handling bottlenecks and equipment configurations are important
aspects of manufacturing processes that have a big impact on production efficiency. To
increase production line productivity, remove bottlenecks, and maximize equipment
utilization rates, it is crucial to examine variables including equipment processing time,
blocking time, and idle time [18]. The finished system model seeks to enhance equipment
use while minimizing idle rates. The crucial step in determining the production line's
operating state involves assessing the number of SLS machines in operation.
The number of machines in the first test was configured according to the guidelines
provided in the article, focusing on output computation using four SLS machines over 250
days. In the second test, all machine quantities remained consistent with the article's
specifications, except for the SLS machines, which were reduced to three, to determine the
output within the same 250-day period. Similarly, in the third iteration, the number of SLS
machines was reduced to two, while still adhering to the article's specified machine
quantities.
Four parameters—the sifter, cabin sandblaster, varnishing, and packing machines—are
used to evaluate the number of other machines in the FlexSim model. For these parameters,
both lower and upper limits have been established to evaluate various scenarios within the
flexible framework of FlexSim. Specifically, the lower and upper bounds for these machines
are as follows: cabin sandblaster machines range from 1 to 3, sifter machines range from 1
to 2, varnishing machines range from 1 to 3, and packing machines range from 1 to 2. By
acquiring the output of the best SLS, which was examined in Table 2, the optimal number of
these parameters was found.
Table 2
Scenarios for Machines
Scenarios # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Cabin
1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2
Sandblaster
Sifter
1 1 1 1 1 1 2 2 2 2 2 2 1 1 1 1 1 1
Machine
Varnishing
1 1 2 2 3 3 1 1 2 2 3 3 1 1 2 2 3 3
Machine
Packing
1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2
Station
Scenarios # 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
Cabin
2 2 2 2 2 2 3 3 3 3 3 3 3 3 3 3 3 3
Sandblaster
Sifter
2 2 2 2 2 2 1 1 1 1 1 1 2 2 2 2 2 2
Machine
Varnishing
1 1 2 2 3 3 1 1 2 2 3 3 1 1 2 2 3 3
Machine
Packing
1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2
Station
Furthermore, an analysis was carried out by considering the impact of utilizing a
distribution function on the deterministic processing time values found in Table 2. It was
assumed that the distribution was normal, with the mean values matching those in the
reference and a standard deviation set at 20% of the mean (std = mean * 0.2).
4. Results
The production rate that resulted from building the model using the reference model's data
was identical to that of the reference model. After that, the validation was confirmed.
4.1. System Simulation Analysis for SLS machine
Table 3 presents the outcomes of these experiments. It illustrates how, after 250 days,
employing 4 SLS machines yields a total production of 23,438 pieces. However, according
to FlexSim, the equipment utilization graphical representation in Fig. 4(a) shows that the
processing times for SLS machines 1, 2, 3, and 4 are, respectively, 58%, 57.97%, 57.96%,
and 57.96%. Furthermore, these machines' collection percentages are 33.86%, 32.68%,
33.90%, and 32.68%, respectively.
A total of 23,309 objects are produced, according to a probability distribution analysis
performed on four SLS machines. SLS machines 1, 2, 3, and 4 have collection percentages of
33.15%, 33.44%, 33.87%, and 33.09%, and processing speeds of 58.04%, 57.79%, 57.37%,
and 58.14%, respectively, as shown in Figure 4(b).
Following a trial that lasted 250 days, a comparable analysis using three SLS machines
produced a production volume of 23,446 units. The experiment's equipment utilization
rates are displayed in Figure 5(a). 13.32%, 13.34%, and 13.37% are the collection
percentages, and 77.33%, 77.31%, and 77.27% are the machine processing durations that
were measured.
Table 3
SLS Machine Configuration Output Comparison
Parts Produced in 250 Day
Experiment Number of SLS
Number Machines Without distribution With distribution
1 4 23,438 23,309
2 3 23,446 23,315
3 2 18,094 18,003
Figure 4: Production Analysis for 4 SLS Machines in 250 Days: (a) Fixed Processing Time,
(b) Distribution-Based Processing Time
23,315 units were produced in a follow-up test with distribution-based processing. The
experiment's equipment utilization rates are shown in Figure 5(b). The evaluations of the
machine processing times are 76.76%, 76.78%, and 76.77%, and the collection percentages
are 13.72%, 13.71%, and 13.71%.
When the research was expanded to include a scenario with only two SLS devices, 18,094
units were produced over 250 days. A similar normal distribution study, conducted with
the same parameters, resulted in 18,003 units.
Three SLS machines are the best option, according to a thorough investigation.
Regardless of distribution-based or fixed processing durations, the results roughly agree,
indicating that three SLS machines make up the ideal configuration. This setup guarantees
sufficient production after the 250-day term in addition to increasing daily operating
efficiency. It is also more cost-effective to use three machines than four. Because it optimizes
production and reduces the need for additional machinery, using a three-machine
arrangement is economical.
Figure 5: Production Analysis for 3 SLS Machines in 250 Days: (a) Fixed Processing
Time, (b) Distribution-Based Processing Time
4.2. Comparing different scenarios
The results across several scenarios are presented in Table 4. A review of the data reveals
three distinct production rate levels.
Table 4
Production Rate Comparison
Production Production
Scenario # Scenario #
* (Mean, std) ** * (Mean, std) **
1 20688 (20562.93, 9.96) 19 20691 (20567.03, 9.98)
2 23446 (23312.16, 9.96) 20 23450 (23452.21, 9.10)
3 20688 (20686.57, 10.27) 21 20691 (20682.39, 10.19)
4 23446 (23313.64, 9.90) 22 27135 (27126.44, 99.00)
5 20688 (20688.14, 10.36) 23 20691 (20684.56, 10.31)
6 23446 (23313.70, 9.98) 24 27135 (27126.34, 99.06)
7 20691 (20566.93, 9.99) 25 20688 (20563.07, 10.00)
8 23450 (23440.57, 8.80) 26 23446 (23313.53, 10.27)
9 20691 (20691.17, 10.30) 27 20688 (20664.27, 10.05)
10 23450 (23451.51, 10.50) 28 23463 (23321.06, 10.96)
11 20691 (20692.76, 10.37) 29 20688 (20667.24, 10.31)
12 23450 (23451.47, 10.50) 30 23463 (23320.96, 10.96)
13 20688 (20563.03, 10.02) 31 20691 (20567.03, 9.98)
14 23446 (23313.46, 10.32) 32 23450 (23452.26, 9.07)
15 20688 (20664.77, 10.48) 33 20691 (20682.36, 10.18)
16 23463 (23320.99, 10.89) 34 27135 (27126.29, 98.96)
17 20688 (20667.29, 10.03) 35 20691 (20684.67, 10.16)
18 23463 (23320.97, 10.93) 36 27135 (27126.37, 98.91)
* With deterministic processing time
** Normal distribution for the processing time
The target output rate of 20,000 units within the allotted 250 days is shown in the first
row. The desired results are achieved in Scenarios 1, 3, 7, and 13. These scenarios are
distinguished by having the highest production rates while using the fewest machines
compared to the others. The reason scenario 1 is the best of them all is that it uses the fewest
equipment to create this much. As an alternative, scenario 2 appears to be the best option
and the second tier indicates a production potential of 23,000 units. But, at a production
rate approaching 28,000 units, the third-tier deviates considerably from our goal. Besides,
reaching this rate would require an increased number of machines, which would be against
our objectives. The same data was examined under different circumstances in which a
normal distribution was used to define the processing time. The output rates of both
techniques are rather close to one another, as Table 4 illustrates.
The ideal situation is shown by the results for both scenarios as follows. The
investigation shows that the 20,000 unit production benchmark specified in the article may
be met with three SLS machines and four machines in the sections that were evaluated. But
as scenario 2 illustrates, to reach a greater output, a production rate of 23,000 units can be
reached thanks to the cooperation of three SLS units and five machines located in the
researched stations.
Two important considerations must be made before making the final choice. First, if
selling 20,000 units at the lowest possible cost is our top goal, then scenario 1 is the greatest
option. However, since using five machines yields the highest production, scenario 2 is the
greatest choice if our objective is to ensure steady costs for all analyzed stations (packing
machines, varnishing, sifter, and cabin sandblaster). Determining which parameters are
most essential to us is what thus makes a difference.
5. Conclusions
This study has addressed the complexities of scaling up AM production, emphasizing the
necessity of meticulous planning and strategy in expansion efforts. The benefits and
challenges of applying AM technology can be very large and include the construction
industry, particularly the field of cultural heritage rehabilitation. By leveraging discrete
event simulation through FlexSim software, the intricacies of optimizing AM factory
productivity, focusing on a selective laser sintering case study have been explored. Key
findings from the study include the impact of operator and machine variability, a clear
roadmap for reducing costs and increasing output, enhanced efficiency through strategic
workforce and machine adjustments, the reduction in waiting times, and streamlining
production. Finally, the paper has proved the importance of using advanced simulation
tools to predict and mitigate potential bottlenecks and inefficiencies, ensuring smooth
transitions during AM production scale-up.
References
[1] Shah, J, B Snider, T Clarke, S Kozutsky, M Lacki, A Hosseini." Large-scale 3D printers for
additive manufacturing: design considerations and challenges" The International Journal of
Advanced Manufacturing Technology (2019): 3679-3693.
[2] Gao, Wei, Yunbo Zhang, Devarajan Ramanujan, Karthik Ramani, Yong Chen, Christopher
B Williams, Charlie CL Wang, Yung C Shin, Song Zhang, Pablo D Zavattieri." The status,
challenges, and future of additive manufacturing in engineering" Computer-Aided Design
(2015): 65-89.
[3] Loh, Giselle Hsiang, Eujin Pei, David Harrison, Mario D Monzón." An overview of
functionally graded additive manufacturing" Additive Manufacturing (2018): 34-44.
[4] Zhou, Feng, Yangjian Ji, Roger Jianxin Jiao." Affective and cognitive design for mass
personalization: status and prospect" Journal of Intelligent Manufacturing (2013): 1047-
1069.
[5] Li, Qiang, Ibrahim Kucukkoc, David Z Zhang." Production planning in additive
manufacturing and 3D printing" Computers & Operations Research (2017): 157-172.
[6] Kucukkoc, Ibrahim, Qiang Li, Naihui He, David Zhang, Scheduling of multiple additive
manufacturing and 3D printing machines to minimise maximum lateness, in: Twentieth
International Working Seminar on Production Economics, Innsbruck, Austria, 2018, pp.
237-247.
[7] Zhang, Yicha, Alain Bernard, Ramy Harik, KP Karunakaran." Build orientation
optimization for multi-part production in additive manufacturing" Journal of Intelligent
Manufacturing (2017): 1393-1407.
[8] Kucukkoc, Ibrahim, Qiang Li, David Z Zhang, Increasing the utilisation of additive
manufacturing and 3D printing machines considering order delivery times, in: 19th
International working seminar on production economics, 2016, pp. 195-201.
[9] Hodoň, R, M Kovalský, M Gregor, P Grznár, New approaches in production scheduling
using dynamic simulation, in: IOP Conference Series: Materials Science and Engineering,
IOP Publishing, 2018, pp. 012023.
[10] Budde, Lukas, Shuangqing Liao, Roman Haenggi, Thomas Friedli." Use of DES to
develop a decision support system for lot size decision-making in manufacturing
companies" Production & Manufacturing Research (2022): 494-518.
[11] Jahangirian, Mohsen, Tillal Eldabi, Aisha Naseer, Lampros K Stergioulas, Terry Young."
Simulation in manufacturing and business: A review" European journal of operational
research (2010): 1-13.
[12] Avventuroso, G, Ruben Foresti, Marco Silvestri, E Morosini Frazzon, Production
paradigms for additive manufacturing systems: A simulation-based analysis, in: 2017
International Conference on Engineering, Technology and Innovation (ICE/ITMC), IEEE,
2017, pp. 973-981.
[13] Kowalski, Arkadiusz, Robert Waszkowski." Layout guidelines for 3D printing devices"
Applied Sciences (2020): 6333.
[14] Chinosi, Michele, Alberto Trombetta." BPMN: An introduction to the standard"
Computer Standards & Interfaces (2012): 124-134.
[15] Deckard, Carl R." Method and apparatus for producing parts by selective sintering"
(1991).
[16] Stansbury, Jeffrey W, Mike J Idacavage." 3D printing with polymers: Challenges among
expanding options and opportunities" Dental materials (2016): 54-64.
[17] Installation Conditions FORMIGA P 110. Laser-Sintering System for Plastics, Krailling,
Germany" (2013).
[18] Cheng, Qiang, Hongchao Shen, Hongyan Chu, Zhifeng Liu, Caixia Zhang, Jiaxiang Ren,
Research on logistics simulation and optimization of die forging production line based on
flexsim, in: Journal of Physics: Conference Series, IOP Publishing, 2020, pp. 022063.