Performance evaluation of the 3d printing system through fault tree analysis method (FTAM)

This study focuses on the performance analysis of the 3D printing process using cementitious material on equipment installed at the Structures Laboratory (LabEst-UnB) of the University of Brasilia. A research direction was defined for the 3D printing with the aim of identifying which steps need to be further developed and which improvements should be made for this specific process to evolve. Following the execution of the proposed experimental program, evaluations of cementitious material pieces printed with the InovaHouse3D’s Alya 130 printer were conducted, documenting the encountered issues and classifying them into failure groups. The causes, consequences, and affected components were defined and used in the application of the Fault Tree Analysis (FTA) method in the search for critical system events. It was observed that the most critical system failures were related to the material composition and printing parameters, focusing on events that trigger variations in the material's consistency. Thus, through this work, it was possible to identify improvement opportunities and suggest scientific research topics to enhance the 3D printing process according to the established priority levels.


Introduction
The revolution in the way cementitious material is used can be referred to, according to Roussel [20], as digital construction, automated construction, or simply 3D printing, a term that encompasses additive manufacturing present in this and other industries.Additive manufacturing is defined by the American Society for Testing and Materials-ASTM as a process of joining materials to form an object based on information from a 3D model, layer by layer.This technology has been increasingly improved, becoming economically viable for industrial, commercial, and even domestic applications.
The technologies that allow automation and digital fabrication in the construction industry are still in the development phase.Digital construction is part of the so-called Industry 4.0 (or Fourth Industrial Revolution) and seeks to enable completely digitized manufacturing, as well as an optimized and logical process using well-developed tools, as stated by [13].The 3D printing of cementitious material is a digital additive manufacturing construction process with great potential in the construction industry and will be the focus of this research.Factors that influence the printing performance, as well as the properties of the produced elements, still require studies to enable the best results, as will be addressed throughout this work.
The large number of factors that influence the printing and its complexity give rise to various failures that can lead to severe consequences, ranging from the production of poor-quality elements and the disposal of printed specimens to the damage of the equipment used.The mentioned problems, if not overcome, can lead to the waste of raw materials, the production of structures with much lower quality than those executed by traditional construction and the excessive complication of the construction process, diametrically opposed to the proposed technology.
Due to these challenges to be faced for the implementation of 3D printing, the study of various parameters, both of the material and of the printing, and of the essential steps, proves to be a demand for the improvement of the 68 Page 2 of 18 presented technology, seeking to identify and analyze the factors that lead to failure and their causes and consequences in order to avoid them.In this sense, this work proposes to study the printing process using the machine Alya 130 developed by InovaHouse3D, in an analysis considering the production steps, parameters and materials used, identifying, in an exploratory way, improvement opportunities through reliability analysis methodology.For this purpose, the Fault Tree Analysis Method (FTA) was selected.According to [3], in this method, the causes of an event are deduced and related to obtain a visual model of how failures contribute to the main accident, and it is possible to identify which ones are more important within the system.

3D Printing
The term "concrete" is trivially used when discussing 3D printing.However, it is important to clarify that this technology uses lightweight cementitious materials, without the use of coarse aggregates, which are replaced by other fine materials such as clay and sand.This is justified by the limitations of both the pumping and extrusion systems, which are incompatible with coarse aggregates due to the risks of clogging and damage to the printing system.Therefore, despite being referred to as concrete printing, in order to accurately reflect the material used, this work will prefer the use of the terms "3D printing of cementitious materials", "3D printing for civil construction" or simply "3D printing".
3D printing is a complex process that requires specific steps to be followed in order to avoid failures and achieve the desired results.[10,11] mention three stages for cementitious material printing: data preparation, material preparation, and the printing process itself.
During data preparation, the component is designed in a 3D CAD model and then sliced into layers of a desired thickness.The printing path for each layer is generated to produce a G-code program file compatible with the printing machinery.Material preparation involves mixing the batch and moving it through the pumping system from the container to the extruder.This material is then extruded into continuous filaments, layer by layer, using the machinery until the entire element is printed.
According to Lim et al. [12], a material is considered printable if it can be extruded continuously and smoothly for an acceptable open time, reaching the intended height without collapse and possessing flexural and compressive strength to fulfill its intended function without issues.[22] state that coordinating the material properties with printing parameters is necessary yet challenging to obtain the benefits of 3D concrete printing.

Parameters
Several factors and parameters must be taken into account to achieve good printing performance, and inappropriate selection can result in poor print quality or even catastrophic failure of the printed element.Based on the findings of [10,11].[12], and [21], the fundamental parameters for 3D printing of cementitious materials are: -Rheological properties of the material: yield stress, viscosity, and structure-forming rate.As a relatively new technology, there are no specific standardized processes for defining, characterizing, and testing 3D printing materials.Qualitative tests are commonly used for extrudability and buildability, and there is no completely adequate standard test, although there are proposals for quantitative tests.

Pumpability
According to [12], the pumpability of cementitious material refers to the ability to move the material through its delivery system (pump, hose, among others) easily and reliably [15].state that for this property to be adequate, the system must ensure uninterrupted material flow through the pumping system.To achieve satisfactory pumping capacity [12], [15], and Rousset et al. (2018) assert that the material must: -Have appropriate fluidity; -Possess low viscosity and low yield stress, allowing it to be easily moved; -Be able to form a lubrication layer composed of water and fine particles segregated from the material on the surface of the pipes and system equipment.

Extrudability
According to Lim et al. [12] and [10,11], extrudability is the ability of a cementitious mixture to pass through the extrusion system and be deposited through the nozzle in an easy and reliable manner, without cracks or splits along the element.For Hou et al. [8], the material needs to have an acceptable degree of workability so that it can be extruded from the printer nozzle uniformly, continuously, without blockages, cracks, and segregation.Also, to ensure that the filament is stable and retains its shape after extrusion, it is important that it develops an adequate yield strength quickly to withstand the stresses generated by gravity, according to Roussel [20].According to [15], the extrudability of a cementitious material depends on the composition and printing parameters, such as extrusion rate, printing speed, and nozzle geometry.

Buildability
The buildability parameter presents the most varied definitions.Some of these definitions for buildabiliry are as follows: -It consists of the material having enough strength in the fluid state to resist deformation when under load, having the ability to retain shape and support subsequent layers [12] and [22], -It is the ability to print a certain number of layers or a certain height without significant deformation of the first filaments, which should not collapse under the weight of subsequent layers [10,11], -The resistance to deformation during printing and the ability of the deposited material to retain its shape [9].
According to [14], buildability is a process that depends not only on the composition of the material but also on the printing process, taking into account printing parameters such as layer interval, nozzle speed, filament section, and the geometry of the printed element.If all the parameters that affect buildability are not in harmony, collapse of the structure will occur.To meet the requirements for good buildability, Han et al. (2021) state that workability and consistency must be restricted to a lower level, presenting a stiffer material, which has adverse effects on the mechanical properties and extrudability of the material.

Workability
In general, workability is expressed by rheological properties, mainly yield stress and plastic viscosity, and refers to the performance of the material during the printing process and the ease with which it can be used.The desired fluidity for good workability is one that allows the material to be smoothly transported from the container to the nozzle.The way to measure workability varies, and this parameter has a direct relationship with pumping capacity and extrudability.[10,11] concluded that workability is greatly influenced by the dosage of setting modifiers such as accelerators, retarders, and superplasticizers, with the latter being the most crucial for maintaining good workability and strength while using a low water-cement ratio.

Open time
Open time is the period during which the material maintains its suitable workability to be pumped through the printing system and extruded, generating a high-quality filament.According to Lim et al. [12], it is the period in which all the characteristics of the cementitious material remain within acceptable tolerances for the printing to occur.[10,11] describe it as the association of workability with time, noting that the reduction in workability over time coincides with the difficulty of printing an adequate filament.Ideally, the open time should be short enough to allow the layers to bond and acquire adequate adhesion, yet long enough for the lower layer to gain strength and support the subsequent layers.

Printing parameters
According to [10,11], [12], and [21], parameters related to the system and machinery play an essential role in 3D printing, and selecting these parameters incorrectly can negatively affect the quality.These parameters are: -Pumping system, which must have enough power to make the cementitious material exceed the yield stress to move through the printing machinery with a constant flow, according to Tay et al. [21], -Nozzle size and shape, which, according to Tay et al.
[21], dictate the shape of the printed filament section, altering the contact area between layers and potentially affecting layer adhesion., -Layer height, which is directly influenced by the yield stress, since the filament's ability to retain its shape under stress depends on it, according to Duda [5].
Additionally, layer height is affected by the nozzle lifting height between layers and the pressure generated by material extrusiont, -Extrusion rate, which depends on the characteristics of the pumping and extrusion systems used and, according to Joh et al. [9], -Nozzle speed during printing, which is controlled by the machinery's code and limited by the system responsible for movement in the Cartesian axes and the pumping system's characteristics; -Printing window, which is the time interval between the printing of two layers and is one of the most important printing parameters.

Mechanical properties
According to Tay et al. (2019), the mechanical properties of 3D-printed structures are greatly affected by the printing process, which generates elements with anisotropic characteristics, that is, they exhibit different strengths when positioned in different orientations.The difference from conventionally molded elements is that the latter are isotropic and exhibit properties distributed in all directions.It is expected that 3D-printed elements will have the highest strengths for stresses perpendicular to the printing plane.Tay et al. ( 2019) state that the mechanical characteristics of 3D-printed elements are directly linked to parameters such as viscosity, waiting time between layer printing, and contact area between layers.The most commonly studied mechanical properties for 3D-printed elements are the compression strength of a cubic test specimen, the flexural tensile strength of prismatic test specimens, and the adhesion between layers.

Geometry
Roussel [20] asserts that buckling failure will occur in vertical slender elements composed of a single filament above a certain height, regardless of the properties of the cementitious material.Therefore, when designing largescale structures, [22] state that caution must be taken as stability and mechanical capacities can be negatively affected when the designed internal void is too large.To address this issue, internal filling geometries should be studied and utilized in designs, enabling hollow structures that are lighter, more stable, and with adequate strengths that can exhibit unique mechanical and functional properties.According to [22], these hollow structures would be preferable for large-scale construction, as 56 to 90% of the strength of solid elements can be obtained using only 48 to 63% of the material.
Alternative solutions for single filament walls, which have more stability and less risk of buckling failure, exist and can favor other parameters such as constructability.Authors such as Joh et al. [9] and Daungwilailuk et al. (2021) propose composite wall printing with two parallel filaments and an internal filling as an appropriate solution.The authors evaluated the effect of using a lattice-shaped internal filling to provide support during block printing.It was determined that the use of such an artifact reduced the occurrence of buckling collapse and therefore improved constructability.Decreasing the span length, which is divided by the internal geometry, is what Joh et al. [9] identified as a factor for improving constructability.

Fault tree analysis method
The objective of this study is to investigate problems encountered during the printing process related to distinct failure modes.Therefore, there is a need to establish a method for classifying and systematizing these events.In this context, the most appropriate method is the Fault Tree Analysis (FTA) method, which is a deductive top-down analysis method where the causes of an event are deduced and related to obtain a visual model of how failures, human errors, and external factors contribute to the main accident, event, or failure.This method was created in the aerospace industry and then popularized by its use in other industries, being the most commonly used method for reliability calculations as it translates a physical system into a logical diagram.
According to [4], the Fault Tree Analysis (FTA) method can be used to determine the cause of an accident by discovering combinations of events that can lead to this problem, which may not be discovered through other forms of analysis.Another purpose is to graphically display the results, making it easy to identify weaknesses and how they lead to undesired events or, in an adequate system, to show that all causes have been considered.The steps to be followed when using a reliability analysis system are, according to [7]: -Definition of the system: which consists of determining the top event, the physical limitations of the system, as well as its components, interactions, specifications, and failure modes, both considered and ignored; -Construction of the logical model of the system: it is necessary to follow a top-down analysis, where the top event is the undesired event, which can be represented by a rupture, explosion, or other forms of adverse events.The top event will reflect the exact situation that one wants to analyze, with a relationship between smaller events through the use of logic gates to indicate the steps followed to the top event; -Qualitative and quantitative analysis: which involves determining the system's failure mode and indicating which components share a common failure cause; -Formulation of conclusions and recommendations.
Logical gates are fundamental blocks that perform operations, relating one or more input variables to only one output variable at a time.The input variable represents a cause event, and the output represents a consequence event.When combined, logical gates create logical circuits, which represent the system, and can be used to explain the interaction between different elements.The events related by logical gates are divided and represented by different symbols.These events are the primary or basic, the top, the intermediate, and the secondary or underdeveloped events.The symbols used to represent the logical gates and events are presented in Table 1.

Flow chart of the process
The steps followed in the development of this paper are presented in the Fig. 1.

Experimental procedures
In order to collect evidence of the failures presented in the process, it is necessary to print elements.As a proposal to evidence and evaluate the extrudability and buildability failures of the material, as well as the other process failures, four elements were modeled to be printed.During the

Symbology Element DefiniƟon
Logical "AND" gate Indicates that all input events need to occur for the output event to happen.
Logical "OR" gate Indicates that only one of the input events can be the cause of the output event.

InhibiƟon logical gate
Indicates that an output event will only occur if the input event happens accompanied by the restricƟve event.
Transfer "from" logical gate Indicates that the fault tree has been transferred from another place, represenƟng the input of the system.
Transfer "to" logical gate Indicates that the fault tree will be transferred to another place, represenƟng the output of the system.
Basic/primary event Indicates a failure in the design environment independent of other primary events, not being caused or causing any other primary event.
Top event Indicates the main undesired event that is the subject of analysis.

Intermediate event
Indicates the result of interacƟon between events, describing the resulƟng failure in greater detail.

Secondary event
Indicates events that do not have enough informaƟon to be developed or that do not have causes.

RestricƟve event
Indicates the event associated with the inhibiƟon gate, presenƟng a condiƟon that must be met for the output to occur.

EX2
Linear two-layer element aimed at evaluaƟng failures related to extrudability.

ED2
Block element without internal filling aimed at evaluaƟng failures related to constructability in an unstable element under the most unfavorable circumstances.

ED1
Block element of 20 cm height with internal filling aimed at evaluaƟng failures related to constructability in a stable element under the most favorable circumstances.
The EX1, EX2, and ED2 elements were printed only once, as in this single attempt it was possible to record all the failures related to them.On the other hand, the ED1 element, which aimed to print a stable element, required process adjustments in several print attempts for its completion, being printed over three attempts printing of these elements, the failures were documented for analysis.This approach aims to provide insight into the nature and causes of the failures, facilitating improvements in the concrete 3D printing process.The proposed elements are presented in Table 2.

Machinery
The machine used for the project was the Alya 130, a 3D printer developed by InovaHouse3D and used for research and development projects.It consists of a machine controlled by a Cartesian coordinate system, i.e., computer numerical control (CNC), where each axis is controlled by stepper motors.To control the machine, a code in the G-Code language was used, a language developed for CNC machines, which specifies the position of each axis and the amount of material used.The machine is controlled through the PronterFace software, provided by PrinRun, 1 .which enables the created code to be transmitted from the computer to the internal command system of the 3D printer.In this Fig. 2 Pronterface interfaceThe object to be printed was modeled in a CAD software, Autocad provided by Autodesk (https:// www.autod esk.com/ produ cts/ autoc ad/ overv iew? term=1-YEAR& tab= subsc ripti on), that allows exportation in the STL (STereo Lithography) format.This file is then used in a slicing software called PrusaSlicer, provided by Prusa3D (https:// www.prusa 3d.com/ page/ prusa slicer_ 424/), which divides the model into 2D layers and is also responsible for defining the printing path.Finally, the slicing software generates a code in the G-code, a language that can be read by the machinery and that will coordinate the cartesian movements of the machine, as well as its speed.The generated code must always be first tested with the machine empty to ensure that all parameters have been defined and exported correctly, and that communication with the machinery is being carried out adequately.Therefore, the test print in the empty state ensures that there is no material waste with a faulty print due to code or machinery errors.The inputs in this software are the layer height and the 3D model.The interface of the software is shown in Fig.
3 As a method of material storage, pumping, and extrusion, the B10 mortar pump from Betomaq was used.The pump used has its own material storage container, with a capacity of 30L, and the hose from the pumping system was connected to the extruder system of the 3D printing machine.The extruder system is composed of a metal cylinder that leads to a conical nozzle, with a 2 cm outlet diameter.The extrusion rate was controlled by the pump itself, being kept as close as possible to the horizontal movement speed of 100 mm/min, to ensure the continuous and uniform deposition of the material.The height at which the nozzle was raised when starting the deposition of a new layer was maintained at 2 cm, equivalent to the extrusion nozzle diameter.Initially, the printing window was kept at zero, with the printing of each new layer immediately after the previous one was completed, only with occasional pauses when it was necessary to replenish the materials in the storage tank by performing a new batching.The parameters used are summarized in Table 3.

Mix proportions and mixing
Dosage is the factor with the greatest influence on the properties of the material, both in the fresh and hardened state, defining the rheology of the mixture used.In addition, the material must meet the requirements of the essential parameters mentioned previously.Therefore, a study of the mixtures used in various projects is necessary to understand the problems and solutions encountered.The cementitious material used for 3D printing, as mentioned by Padilha [16], is generally composed of a combination of paste, formed by binders and water, fine aggregate, which can be sands, clays, and fibers, and chemical additives, such as setting modifiers, viscosity modifiers, and superplasticizers.
According to Khan (2020), the material should not have coarse aggregates, and the fine aggregates will have a diameter limited by the capacity of the pumping and extrusion system.Additionally, it is desirable that the 3D printing material have a continuous particle size distribution to improve parameters, and the use of silica fume and fly ash to replace a portion of the cement in the mixture is recommended.The use of these substitution materials benefits the material because they have rounded particles that help fill voids and improve workability.
For the determination of the mixture to be used, no laboratory tests were performed at this stage.The mixture used was defined based on a literature review, especially considering Padilha's [16] studies, due to similarities between the processes used by the author and those presented in this work.Additionally, this mixture had already been used in conjunction with the Alya machine, and acceptable results were obtained.Considering the exploratory nature of this project, a simple mixture without many additions that would increase the product cost was sought.
Thus, the mixture of the material used in this project is composed of Portland cement CP-II, fine washed sand sieved through a 10-mesh sieve, with a maximum particle size of 1.68 mm, using Axton plasticizer additive.The sand should be sieved twice to ensure that the maximum diameter of 2 mm allowed by the pumping system is respected.The mixture can be seen in Table 4.
The material was mixed in a CS 120 L concrete mixer before being deposited in the pump container used.The steps for the mixing procedure were based on the NBR 16,541 standard (ABNT, 2016), with adaptations, following these procedures: -Moisturize the concrete mixer to prevent water loss to the walls; -Add all dry material (cement and sand) to the concrete mixer; -Turn on the concrete mixer until the dry materials are homogenized; -Add 75% of the water and the superplasticizer; -Activate the concrete mixer for 90 s; -Let the mixture rest for 30 s; -Add the remaining water; -Turn on the concrete mixer for another 90 s or until the mixture is homogeneous.After the mixture is ready, it should be poured into the container attached to the pump and the printing should be initiated, always keeping the material in motion and performing a new mixture to refill the container when necessary.

Printing process
After the material was prepared and stored in the pump container and the code was dry-tested, the printing process began following the steps below: -The printer must be connected to a power outlet and, through a USB cable, connected to the computer that will control it; -In the PronterFace software, the machine must be connected and online; -The machine must be positioned at the starting point, which will be defined in the software as the "zero" point; -The code will then be opened in the software; -When starting the code printing process, the pump must be turned on; -The printing process is expected to continue until completion, evaluating the stability conditions of the printed element and refilling the container with a new mixture whenever necessary.
The recording of failures was performed through the registration of images and videos during the printing process.When necessary, screenshots from the videos were taken, which may consequently have lower quality.These photos and captures were then edited so that the failures were clearly indicated.

Printing results
The printed elements EX1 and EX2 can be seen in Fig. 4. No difficulties were encountered during printing.
Element ED1 was printed seeking the most unfavorable situation, and its collapsed form can be observed in Fig. 5.To achieve this, the layer interval was kept at zero so that the material did not develop internal structure.Additionally, a situation in which the material needed to be less  consistent was simulated, aiming to further reduce the stability of the element.
Element ED2 was printed three times to evaluate the failures of a stable and complete element.The first attempt, presented Fig. 6, was interrupted on the third layer as the element quickly deformed and collapsed.Adjustments were made to enable the complete printing of the element, including an increase in the layer interval, consistency, and a reduction in printing speed.Thus, two additional complete prints were made, which can be viewed in Figs. 7, 8.
The main failure events were documented based on the printed elements with images and videos recorded during the printing.These events were then grouped, with their possible causes, consequences, and solutions identified and analyzed in depth in the process of building the Fault Tree Analysis.The identified groups were: -Discontinuities in the filament which can be seen on Figs. 9, 10; -Variations in the section, which can be in the dimensions of the filament between layers of an element, as seen on Fig. 11, or in one layer, as seen on Fig. 12; -Material buildup that can be visualized on Fig. 13; -System clogging; -Buildability failures that can be caused by excessive deformation of lower layers (Fig. 14), geometric nonconformity (Fig. 15), or deformation due to height (Fig. 16); -Insufficient adhesion between layers.
It should be noted that during the printing process, there were moments of machine-related failures or errors, such as sudden stop of printing due to code and software errors or poor contact in the machine or pump circuits.In these cases, it was necessary to simply restart the system or check the connection points of the printing machine or control panel.As these were only machinery or software initialization problems, these situations were included in the failure tree as secondary events, not requiring a more in-depth approach.For the mentioned cases, photographic records cannot be made, and a simple check of the systems, which must be performed before each print, is shown to be the solution to the problem.Such situations will not be further investigated and will be included in the failure tree as secondary events since they do not originate from complex causes requiring evaluation and study and do not generate significant consequences.

Fault tree analysis
The first step to build the fault tree consists of determining the top event, physical limitations of the system, components, interactions, specifications, and failure modes.The system in this paper is defined as the printing process using a 3D printing machine, which can be divided into five subsystems: the material system, the control system, the machine system itself, the pumping system, and the extruder system.A schematic representation of the printing process and subsystems can be seen in Fig. 17.
The considered subsystems can then be divided into their components or, in some cases, stages.In addition, knowing the components allows decomposing each subsystem into a schematic representation.The complete decomposed system can be assembled, as presented in Fig. 18.
It is then necessary to categorize failures according to the components affected by them and determine to which class such failure belongs.Regarding the classes, failures were divided into top event, basic, secondary, and intermediate events.The top event is chosen as the occurrence that covers the largest number of possible failures, which is the disposal of the specimen.The intermediate events are the result of interactions guided by logical gates between other events.The basic events are the lowest level in the fault tree, not caused by any other event and independent of other basic events.Finally, secondary events are those that cannot be developed either due to lack of information or not having a specific cause.
Regarding the affected components, failures were associated with software, machine, pump, extruder, code, material, and printed element.Furthermore, in case the failure is caused by interference from elements not related to the subsystems, it was associated with an external component.The "material" component was subdivided into general component, mixture, ratio, and storage.The "code" component was subdivided into general, geometry, parameters, and origin.Finally, the "machine" component was subdivided into parts and cables.The other components did not present subdivisions.
The complete fault tree has 224 elements, which is too large to be presented and can be found in Felfili (2023), therefore Table 5, which was used to build the fault tree, will serve as an visual representation of the events, possible consequences, their connections, associated components and type of event considered Table 6.
For qualitative analysis, basic events were analyzed, which alone or accompanied by only one other basic event can directly lead to the occurrence of the top event, and the number of times these appear in the tree.It is possible to observe that the events that appeared most frequently in the minimum cut analysis were mainly those with a causal relation to the variation of material consistency, largely due to the selection of the mixture and printing parameters.Therefore, this failure and the events related to it should be a focus when developing the study plan.
In addition to the analysis of minimum cuts, it was considered necessary to perform a component-based analysis of the components that presented failures and the total sum of occurrences of events related to these components.This analysis was carried out by combining data from**, resulting in Table 7.
This data can then be combined into a column chart, presented in Fig. 19, for a better understanding of the results.It is noted that most of the failures are concentrated in the Material, Code, and Printed Element components.The Printed Element component cannot be decomposed, but the Code and Material components can, according to the component classifications proposed.
Following the subdivision presented previously, the Material component is decomposed into general failure, mixture failure, trace failure, and storage failure.The number of occurrences of each is presented graphically in Fig. 20, indicating the percentages according to the total quantity of material failures.
It is evident, then, that the vast majority of failures associated with the material are due to the trace, making it   proposals, their priority, and their sequencing is presented in Fig. 23.

Conclusions
To study the performance of the 3D printing system, parts were printed according to an experimental program.The failures that occurred during the process were documented and later analyzed using the Fault Tree Analysis Method (FTA).Two linear parts were printed to evaluate failures related to extrudability, and two blocks were printed to evaluate failures related to constructability.Thus, it was possible to evaluate all components of the system, process steps, parameters, and problems related to them.
The general failure groups identified during the execution of the proposed experimental program were discontinuities in the filament, variations in the filament and layer section, accumulation of material at filament points, system blockage, constructability failures, and insufficient layer adhesion.
The probable causes, consequences, and affected components were defined for each failure group.With this information, the fault tree was assembled, following the steps required by the method, to analyze the events and relationships existing between them.
After evaluating the results obtained, critical failures and severely affected components were identified.It was observed that the events that appeared with the highest priority in the minimum cut analysis were those with a causal relationship with material consistency variation.In component analysis, most of the failures were concentrated in the Material, Code, and Printed Element components.Generally, failures associated with the material were due to the trace, while those associated with the code were concentrated in the parameters.
Based on these conclusions, future study topics were proposed to expand knowledge, improve the 3D printing process, and enhance the quality of the produced elements.Thus, this work identified several opportunities for improvement in the 3D printing process, involving the enhancement of each component of the system.Optimizing these aspects is fundamental, contributing to the evolution of technology and increasingly enabling its potential for application in the construction industry.

68 Page 6 of 18 Fig. 1
Fig. 1 Flowchart of the process

Fig. 18
Fig. 18 Schematic representation of the complete system

Fig. 19
Fig. 19 Chart number of appearances per component failure The proposed topics were: -Comparative study of mixes; -Comparative study of printing parameters; -Comparative study of tests; -Comparative study of formulas; -Definition of the material characterization test program; -Definition of the test program for specific 3D printing properties; -Definition of the test program for strength properties; -Trace definition; -Evaluation of formula accuracy; -Procedures for block and wall tests; -Determination of the most appropriate geometries; -Proposal of modules.

Table 1
System failures

Table 2
Proposed Elements EX1Linear one-layer element aimed at evaluaƟng failures related to extrudability.

Table 3
Printing parameters

Table 6
No. of appearances per failure event

Table 7
No of appearances per