Fire performance in buildings: academic insights and perspective analysis

: Accelerated growth of urbanization and the consequent rise in the number of buildings results on an increase in rapid constructions without minimum criteria that enable the safe building performance and how it behaves during episodes of fire. In this scenario, this paper aims to analyze research trends and the status quo of building fire performance in the last five years, evaluating perspectives for research, and proposals for future directions through a Systematic Literature Review (SLR) approach associated with bibliometric analysis. The analysis was carried out based on the Web of Science database under parameters such as authors, countries and regions, journals, research areas, and keywords. For each rule, among all the results retrieved, those that stood out the most in the stipulated period were evaluated. It was found that research on fire performance in projects is increasing, with a total of 402 published works. In the analysis by Systematic Literature Review, there was a trend of research concerning materials and structural systems, evaluating the performance, behavior, resistance, and safety of buildings under fire conditions. Ultimately, the results pointed out a possible evolution of the trend in research on methodologies and intelligent control systems applied to the management of fire emergencies.


INTRODUCTION
The construction industry is one of the main contributors to the global economy (KIFOKERIS; XENIDIS, 2017). Given the large participation in this sector, the management of a construction project involves the use of various resources to achieve project objectives related to attributes such as quality, duration, cost, function, and durability (ZHANG et al., 2019).
In building construction, the parties involved work in a flexible and dynamic environment that supports interactive processes based on knowledge and responsibility. However, those responsible for the fire safety design of buildings are insufficiently involved in this interactive process (MA; WU, 2020;MALUK, 2017). In the buildings' management phase, fire has always been a significant threat to their safe operation (MA; WU, 2020). Every year, building fires cause numerous deaths, along with a serious economic and social impact worldwide (LUCHERINI; MALUK, 2019). As an example, 38% of the causes of fires in the United States are residential (NATIONAL FIRE DATA CENTER, 2019). In the year 2018 alone, there were a total of 379,600 residential fires, which resulted in 2790 deaths, 11,525 injuries, and a material loss of $8.2 billion.
Thus, one must consider appropriate fire safety strategies, fire prevention regulations in particular. Each country has proper regulations, with parameters and requirements, for the analysis of structural systems under fire and smoke conditions, such as the International Building Code (ICC, 2018) adopted in the USA, the EN 13501-1 adopted in the European Union (EUROPEAN COMMITTEE FOR STANDARDIZATION [CEN], 2018), and the Performance Standard (ASSOCIAÇÃO BRASILEIRA DE NORMAS TÉCNICAS [ABNT], 2013) adopted in Brazil.
Since the development of building codes, fire safety design and regulation in the building structural systems design have been based on the concept of compliance. In this way, the design looks at individual building elements, evaluating what is needed to meet the acceptability criteria presented in building codes to ensure that buildings provide an assumed, though typically unquantified, level of fire safety (LAW; BEEVER, 1995;MALUK, 2017). More specifically, current fire codes are prescriptive for passive building construction and active fire protection systems in many places (CHOW, 2015). As such, a building, when ensuring fire performance, must: (i) enable the safe exit of occupants; (ii) ensure conditions for the employment of public rescue with timely (iii) prevent or minimize damage to the building itself, to adjacent buildings, to public infrastructure and the environment (ABNT, 2013).
Fire performance research was first addressed at the end of the twentieth and the beginning of the twenty-first century, and since then, it has been presenting a growing number of proposed works in the area. Even so, fire performance in buildings is a recent concern and a gap in research. Moreover, according to (MALUK; WOODROW; TORERO, 2017), there are still adversities in allying Fire performance in buildings: academic insights and perspective analysis the fire safety community -both in research and practice -with other areas of building design. In addition, the authors highlight the greatness and potential benefits of this relationship.
About the trend allied to technologies, the literature has been recurrent in pointing out that performance assurance leads to a consideration of normative requirements or other performance parameters demanded by users since the feasibility analysis and initial studies (COTTA; ANDERY, 2018). This requires the integrated and simultaneous development of architectural design and engineering (KAMARA; ANUMBA; CUTTING-DECELLE, 2007). In addition, (BRÍGITTE; RUSCHEL, 2016) indicate the need for a systemic view between the various variables that make up the performance requirements.
Too often, fire design professionals participate only in the non-essential part of a building's design process, sometimes to obtain regulatory approval-restricting the design to align with prescribed fire safety measures that are supposed to provide adequate fire safety. The result is a suboptimal relationship between the overall design and the fire safety design in many buildings (MA; WU, 2020; MALUK, 2017). Thus, according to (MALUK; WOODROW; TORERO, 2017), the fire safety community (both in research and practice) recognizes the need for integration to the other design fronts and has reacted to the continuous evolution of buildings. In recent decades, global efforts have been made to develop and implement performance-based approaches to fire safety design.
Therefore, it is important to evaluate the current conditions regarding research on building fire performance, as well as the applications that are contributing to the promotion and development of the proposed analysis.
This paper aims to analyze the trends in research and the status quo of building fire performance in the last five years, evaluating perspectives for research in the area, in the construction sector, by Systematic Mapping of the Literature associated with bibliometric analysis.

METHOD
The research trends and the status quo of fire performance in buildings will be evaluated in this research from aspects such as global contributions, leading countries and regions, most productive institutions, journals, authors, leading research areas, collaboration patterns between countries/regions and institutions, most cited articles and historical maps of keywords of authors, international standards, and important topics. To understand the current status quo, an analysis of papers published in the last 5 years will be considered.
Systematic Literature Review (SLR) is used to identify the issues and explore new research approaches regarding building fire performance. In addition, Bibliometric Analysis is adopted for data collection and analysis.

Method of analysis and data collection
The flowchart of the data collection method used is shown in figure 1. The data collection and analysis were carried out in the following parts, with their respective steps: determination of the research question adopted in the research, data collection, data analysis, and data visualization.
According to (KITCHENHAM, B.;CHARTERS, 2007), the systematic mapping study is complementary to the systematic review, which in turn is characterized as a review of a broad character and results in the primary studies in a specific area, thus seeking to identify possible evidence available in that area (CHEN et al., 2019). Also, according to the authors, systematic mapping is a method whose goal is to build a classification scheme and structure in a field of interest. Systematic mapping studies are used to structure a research area, typically providing visual summaries (outcome maps), while systematic reviews are focused on collecting and synthesizing evidence (PETERSEN; VAKKALANKA; KUZNIARZ, 2015).

Research protocol
The execution of the systematic mapping or RSL is presented in figure 2 (KITCHENHAM, B.;CHARTERS, 2007), as the definition of research protocol with its respective guidelines, prior to the execution of the literature review. It will be followed in these guidelines, which include activities grouped into three main phases: (i) planning, (ii) conducting, and (iii) reporting.

Research Process
The goal of a systematic review is to find as many primary studies as possible related to the research question, using an unbiased search strategy.
The rigor of the search process is a factor that differentiates systematic reviews from the usual ones (KITCHENHAM, B.;CHARTERS, 2007). In this study, the strategy for identifying published research was the use of WoS.
WoS is widely regarded as a standard tool for generating citation data for scientific research and other evaluation purposes. Its flagship collection includes more than 12,000 authoritative, high-impact academic journals worldwide, including the natural sciences, engineering, biomedicine, social sciences, arts, and humanities (LI et al., 2020).
In the present study, the main concepts, i.e., title (TS), abstract, and keywords, were addressed in topics as well. To elaborate the search string, the terms building, fire, performance, and standard were highlighted. In addition, synonyms and relevant variants were added to the search, resulting in TS = ((building OR house* OR residential) AND (fire) AND (performance) AND (code OR regulation OR standard* OR certification)).
Exclusion and inclusion criteria were adopted to obtain consistent results: (i) inclusion: papers that are journal articles in English, Spanish and Portuguese languages; (ii) exclusion: duplicate studies; studies older than five years. It is worth mentioning that the search string adopted may not cover all the existing synonyms for the term fire performance standards in buildings and, thus, may be insufficient to reach all the studies in the area. In addition, it should be noted that the WoS database may not publish all studies, and consequently, not be shown in this research.
To manage the data extracted from the documents, we used the VOSviewer software -a data mining and visualization platform that creates maps based on the network data and can be visualized, explored and indicate useful information that lies behind the data.

RESEARCH QUESTIONS
The research questions specified at the beginning of the systematic mapping (KITCHENHAM, B.;CHARTERS, 2007) are as follows: • Q (i): What is the frequency of publications in Web of Science (WoS) journals regarding building fire performance in the last five years? • Q (ii): What is the category of the publications? • Q (iii): Which are the most cited journals related to the specific theme? • Q (iv): What are the most common keywords for the proposed theme? • Q (v): Which countries stand out in the research of fire performance in buildings? • Q (vi): Which authors have worked with the proposed theme?

RESULTS
When analyzing Q (i), we obtained 402 papers published from January 2015 to April 2020. Of these, 402 documents, 381 articles, and 21 reviews. Regarding the analysis of the results obtained, 64 countries contributed to the research field of fire performance in buildings. Rejane Martins VIEGAS; João Paulo LIMA; Michele Tereza CARVALHO; Caio Frederico e SILVA It can be seen that the trend in the growth of published research over time, for the specific research field, is an average of 81 published papers per year (in the last 5 years), which can be seen in figure 3. To answer question Q (ii), 76 different WoS classification categories resulted from the 402 papers on building fire performance, where the same paper can be classified in more than one category. Thus, the WoS categories labeled in the same papers are connected. Table 1 shows the top ten WoS categories by a record count for the topic covered.
The Civil Engineering area represents the vast majority of the category list, with a total of 40.30 % of the papers published in the last five years, followed by Building Construction Technology, Multidisciplinary Materials Sciences, and Multidisciplinary Engineering. The remaining categories contributed to less than 10 percent of papers published on the specified topic. The categories Energy and Fuels, Thermodynamics and Sustainable Technologies lead the average number of citations per paper, with 7.05; 5.30 and 4.95 citations per publication, respectively. It is also possible to evaluate the h-index of each category, highlighting Civil Engineering with an h-index of 15 (15 publications with at least 15 citations each), followed by Building Construction Technology with h-index = 12 and Multidisciplinary Materials Sciences with h-index = 11. The other categories have h-indexes less than 10. When analyzing all papers, there is a total of 1542 citations, an average of 3.7 citations per paper, in an h-index = 18.
The results in Table 1 also point to a concern for research within Engineering on the term Fire performance in buildings: academic insights and perspective analysis fire performance in buildings. Many papers (62.40% of the total) fall within the fields of Civil Engineering, Building Construction Technology, and Multidisciplinary Materials Science. Thus, one can associate the recurrence of research in the last five years to the search for fire performance in buildings through the minimum strength of the materials used in structural systems, as required by national and international standards. In question Q (iii), academics and those interested in building fire performance must know in which journals the search scare presents recent publications relevant to the topic. In the proposed search field, 233 journals contributed to the scientific literature. Table 2 presents the list of the 10 most recurrent journals in the analyzed topic.  The application of studies on civil construction materials also can be highlighted, as proposed by (ABBAS et al., 2019) who suggested predictive relationships based on Artificial Neural Networks for the fire performance of high-strength concrete and compared their results with codes, standards, and other research. Also, one can notice the presence of works that evaluate fire safety codes in buildings such as (GISSI; RONCHI; PURSER, 2017) and (GRIMWOOD; SANDERSON, 2015).
Finally, attention is brought to works focused on Fire Safety Design and Management, such as (ARIYANAYAGAM; MAHENDRAN, 2015) which evaluated the potential of integrating fire safety into modern building design; and (ABNT, 2013) which presented fire design rules for cold-formed steel frame walls exposed to realistic design fire curves. The keywords defined by the author reflect the main focus and trend of the research on Fire Performance in buildings. To discuss question Q (iv), 1956 author keywords were analyzed from the retrieved results. Keywords with the same meanings were unified into a single word. Publications that have no author keywords might not have been included in this analysis. A total of 1615 (82.56%) keywords were used Fire performance in buildings: academic insights and perspective analysis only once, which demonstrates a wide range of research interests in building fire performance. Table 3 shows the most frequently occurring keywords. Figure 4 presents a bubble graph used to evaluate the keyword development trend.

Figure 4 -Keyword Trends by Year
Source: Self-Elaboration At the top of the graph are the years of publication. The number in each bubble is the annual number of publications for each keyword listed on the abscissa axis. The larger the bubble is, the more publications for each topic were found. Performing a vertical comparison of the sizes of the bubbles identifies the trending keywords for each year. The size of the bubbles horizontally shows the possibility of determining the growth trend of each keyword over time. It can be seen that performance is the most used keyword with 84 recurrences, which had a significant growth between 2016 and 2020. It should also be noted that, even joining similar words, performance is still widely associated in different compound keywords, such as performance-based design (12th in the recurrence ranking, defined in 16 papers) and fire performance (18ª in the recurrence ranking, which was used in 13 papers), as well as examples not shown in Table 3 like "performances analysis, structural performance, and building performance".  Finally, it is noticed that, throughout the positioning, terms related to materials such as steel, concrete and wood appear. In addition, the recurrence of structural elements such as beams and columns is also noted. For most of these keywords, there is a tendency to stabilize in recurrences over the years, highlighting the possible associations of terms that result in new keywords, as previously presented.
In Q(v), verifying the results obtained by countries or regions that stood out in relation to the subject, the co-participation among them is of paramount importance to researchers in order to establish a research network. Table 4 presents the ranking with the main countries that contributed to the fire performance theme in buildings in the past five years. The Republic of China is the most productive region, with a total of 72 publications (21.82%) in the last five years, followed by the United States, with 71 (21.52%) publications. In the third position is Australia, with 52 (15.76%) publications.
Brazil is in the 32nd position in the ranking, with 3 publications (0.91%). Analyzing the average of citations, Portugal stands out, with an average of 11.38 citations per article (91 citations), followed by Switzerland, with an average of 8.56 citations per article (77 citations), and by Italy with an average of 6, 90 citations per article (207 citations). Australia, China, and Italy stand out for total citations, with respectively 286, 256, and 237 citations in the past five years. Finally, comparing the h-index by country, Australia and Italy stand out, with an h-index = 9.
By analyzing Table 4, it can be concluded that Australia is the most active country that cooperated with other countries in 40 studies, with Scotland, China, England, and the United States, in particular. England, which comes in second place (34 works in partnership), also maintains close relations with Scotland, the Republic of China, Sweden, and Australia. Fire performance in buildings: academic insights and perspective analysis Brazil 3 0,91% Finally, on Q (vi), the main authors who contribute to a certain area have a high reputation, and their works can inspire scholars to identify research directions. Thus, they are listed as the main most productive authors based on the number of publications on building fire performance (Table 5). Table 5 also presents the main scientific contributions of the authors with the highest number of papers. Among the main authors presented, Mahen Mahendran, Thomas Gernay, and Wojciech Wegrzynski, are from universities in Australia, the United States, and Poland: countries that rank 3rd, 2nd, and 16th in the production of publications in the world. Among the correlations between the main authors, Mahen Mahendran and Anthony Ariyanayagam have a total of 5 papers in common. In addition, Takafum Noguchi and Hideki Yoshioka also stand out with 5 papers in common, in the past 5 years.

DISCUSSIONS
By analyzing the results, it can be seen that more than 60% of the publications on fire performance in buildings were published in the last 5 years. Moreover, the increasing trend rates presented in most of the results in Figures 3 and 4 and Tables 1 to 5, show a high interest in research related to the theme. The emergence of collaborations between different research areas can be attributed to the development of methods for evaluating building fire performance and the diversity of practical problems to be applied. This growing relationship between the different areas is directly associated with the challenges of meeting the requirements of multiobjective optimization techniques in current building designs (HIDALGO; WELCH; TORERO, 2015). Similarly, this is the scope of the fire performance standards given as an example in Section 1 (ICC, 2018; CEN, 2018; MA; WU, 2020) that guarantees the different performance criteria and, in a multi-objective and conflicting manner, must be met, in isolation (MA; WU, 2020).
The characteristics and trends of research in fire performance in buildings vary slightly over the years, as can be easily seen in Table 3 and Fig. 4. However, associating the results, it can be seen that the concern with materials and structural systems will be the basis for studies that analyze the performance, behavior, resistance, and safety of buildings under fire conditions. As an example, we highlight one of the hot papers obtained through the retrieved data: A review of the fire behaviour of pultruded GFRP structural profiles for civil engineering applications (CORREIA; BAI; KELLER, 2015).
Furthermore, in a distributed trend, each country that researches fire performance addresses different techniques when analyzing materials or systems, or even a combination of them. This is associated with the fact that each region has specific climate conditions, availability of materials, and construction techniques. Thus, it shows the possibility that researchers from different countries or institutions can work together to further promote research on fire performance in buildings.
In general, there is a perceived lack of correlation between fire performance research and current methodologies applied to fire emergency management. Such methodologies also allow intelligent monitoring and continuous and accurate observation of fire conditions. Despite not being recurrent in the search parameters associated with data collection, one can highlight the papers BIM-based building fire emergency management: Combining building users' behavioral decisions (MA; WU, 2020), The Evaluation of Building Fire Emergency Response Capability Based on the CMM (MA; TAN; SHANG, 2019) and one of the retrieved hot papers, BIM integrated smart monitoring technique for building fire prevention and disaster relief (CHENG et al., 2017).
There is a lack of detailed guidance for conducting performance-based fire engineering analysis for the built environment, in particular, how to identify and specify performance criteria, fire scenarios, and fire design. Rejane Martins VIEGAS; João Paulo LIMA; Michele Tereza CARVALHO; Caio Frederico e SILVA

CONCLUSION
This research on fire performance in buildings was conducted for scientific literature published between the years 2015 to 2019. It was based on the Web of Science database. Applying the RSL study made it possible to survey of the research characteristics on the subject, identifying topics, journals, countries, and trending authors among 407 articles. The application of RSL associated with bibliometric analysis permitted the interpretation of the results and trends found.
The products of this research pointed out a tendency to analyze the elements, systems, and materials used in buildings by numerical and experimental tests proposed in standards, which guarantees a performance-based design for buildings.
It was identified that the analysis of fire performance in projects is growing, but when compared to the numerous publications that investigate the quality of projects based on performance, it presents a significantly low number. This is justified by the fact that many design professionals still do not recognize the topic of fire safety as an explicit design variable.
Finally, it is concluded that performance-based design should be seen and practiced as a multidisciplinary area to be developed and that for this, it is of utmost importance to evaluate the status and future applications of building fire performance analysis in different research areas, trends, and collaborations (technologies that enable multi-objective optimization analysis).