Application of Experimental Heat-Detector Database to Fire Dynamics Simulator Input Parameters for Performance-Based Design

Ho-Sik Han; Ga-Yeong Yoon; Hyo-Yeon Jang; Cheol-Hong Hwang; Bryan Klein

doi:10.7731/KIFSE.671b52ad

Int J Fire Sci Eng > Volume 40(1); 2026 > Article

Han, Yoon, Jang, Hwang, and Klein: Application of Experimental Heat-Detector Database to Fire Dynamics Simulator Input Parameters for Performance-Based Design

Research Paper

International Journal of Fire Science and Engineering 2026;40(1):24-33.

Published online: March 31, 2026

DOI: https://doi.org/10.7731/KIFSE.671b52ad

Application of Experimental Heat-Detector Database to Fire Dynamics Simulator Input Parameters for Performance-Based Design

Ho-Sik Han¹, Ga-Yeong Yoon², Hyo-Yeon Jang³, Cheol-Hong Hwang^4,^†, Bryan Klein⁵

¹Post-Doctoral Researcher, Department of Fire Safety Research, Korea Institute of Civil Engineering and Building Technology, Hwaseong 18544, Republic of Korea

²Assistant Researcher, Research and Development Strategy Team Quality Certification Part, Fire Insurance Laboratories of Korea, Yeoju 12661, Republic of Korea

³Ph. D. Candidate, Department of Disaster Prevention, Graduate School, Daejeon University, Daejeon 34520, Republic of Korea

⁴Professor, Department of Fire and Disaster Prevention, Daejeon University, Daejeon 34520, Republic of Korea

⁵Senior Support Engineer, Thunderhead Engineering & Apex Erudition, Anacortes, WA, 98221, USA

^†Corresponding Author, TEL: +82-42-280-2592, FAX: +82-42-280-2596, E-Mail: chehwang@dju.ac.kr

Received March 4, 2026 Accepted March 19, 2026

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Accurate prediction of fire detector activation time is essential for ensuring the reliability of Performance-Based Design (PBD) for fire and life safety analysis. In fire modeling tool such as the Fire Dynamics Simulator (FDS), the thermal response characteristics of heat detectors are represented by key input parameters, including activation temperature and response time index (RTI). However, these parameters are often specified using default values without sufficient validation against the actual performance of individual detectors. This study reviews and analyzes an experimental database of the thermal response characteristics of fixed-temperature heat detectors to establish a rational approach for selecting input parameters in fire modeling. The database comprises activation temperatures and RTI values measured under controlled conditions using the fire detector evaluator in previous studies. The results indicate that both activation temperature and RTI vary considerably depending on the detector type and manufacturer. Previous validation studies showed that using these measured values in FDS simulations improved agreement with observed detector activation behavior compared with simulations using default parameters. These findings highlight the importance of using experimentally validated detector input data to enhance the predictive capability of fire models and support more reliable PBD applications.

Keywords: Performance-based design (PBD); Fire dynamics simulator (FDS); Fixed-temperature heat detector; Response time index (RTI); Activation temperature

1. Introduction

The application of performance-based design (PBD) has expanded significantly with the increasing complexity and scale of modern buildings. In PBD, fire safety is evaluated by comparing the available safe egress time (ASET) and the required safe egress time (RSET), which are typically estimated through fire and evacuation simulations [1-3]. Among the RSET components, detection time is critical as it determines the initiation of occupant response and evacuation [3-5]. Therefore, predicting the fire-detection time accurately is vital for reliable PBD fire-safety assessments [2,5,6].

The fire-detection time is governed by the performance of fire-detection systems, among which heat detectors are widely used in various building types [7,8]. The activation characteristics of heat detectors are primarily defined by activation temperature and response time, which are governed by the thermal properties of the sensing element and the surrounding flow conditions [8,9]. Experimental studies, including full-scale fire tests and wind tunnel experiments, have demonstrated that detector response varies significantly depending on the fire growth rate, distance from the fire source, and ventilation conditions [10-12]. Heat detectors tend to exhibit delayed responses in the early fire stages compared with other detection technologies, highlighting limitations in early fire-detection performance [11,13]. These findings indicate that detector activation is governed by complex thermal-fluid interactions instead of a simple temperature threshold.

The activation of heat detectors is governed critically by the fire plume and ceiling jet flow, which determine the temperature and velocity fields at the detector location [14,15]. The thermal response of the sensing element is typically described by a heat-balance equation, and the response time index (RTI) is widely used as a representative parameter to characterize detector sensitivity [16,17]. The RTI, which reflects the combined effect of thermal inertia and convective heat transfer, has been extensively used in both sprinkler and detector-response modeling [16,18]. However, several studies demonstrated that the RTI may vary under different temperature and velocity conditions, particularly in low-velocity or low-temperature environments [19]. This variability introduces additional uncertainty into the prediction of the detector activation time.

In fire simulations, particularly those utilizing fire dynamics simulator (FDS), RTI and activation temperature are used as key input parameters to predict detector activation time [20,21]. However, in practice, reliable detector-specific input data are typically unavailable, thus resulting in the widespread use of default values [8,22]. Previous studies demonstrated that RTI and activation temperature can vary significantly depending on the detector type and manufacturer [23], where such variability can result in substantial differences in the predicted activation time [8,24]. Furthermore, simulation results using experimentally measured detector properties have indicated improved agreement with full-scale fire test results [23], emphasizing the importance of accurate input parameters. Although recent efforts have focused on measuring detector properties using wind tunnel and fire-detector evaluator (FDE) systems along with constructing corresponding databases [23,25], practical limitations remain in applying such data directly to design owing to cost, time, and data availability constraints. Therefore, a systematic review of heat-detector response characteristics and input-parameter variability is necessary to establish a rational approach for applying these parameters to fire simulations and improving the reliability of predicting detection time in PBDs.

2. Heat-Detector Response Theory and Modeling

2.1. Thermal response of heat detectors

Heat detectors are activated when the temperature of a sensing element exposed to fire-induced hot gases reaches a predefined threshold. This response is governed primarily by convective heat transfer and depends on both the thermal properties of the detector and the surrounding flow conditions [8,16]. The thermal behavior of the sensing element is commonly described by the lumped heat-balance equation given in Eq. (1):

(1)

mcdTedt=hA(Tg-Te)

where T_e and T_g denote the temperatures of the sensing element and surrounding gas, respectively; h is the heat-transfer coefficient; m is the mass; c is the specific heat; and A is the surface area [16]. This formulation assumes convective heat transfer dominance and disregards conductive losses, which may affect accuracy under certain fire conditions [16,26].

To simplify detector-response characterization, the RTI is widely used as a lumped parameter model to represent the combined effect of thermal inertia and convection [16,18]. The RTI enables the prediction of activation time based on gas temperature and velocity and is commonly adopted in both sprinkler and heat detector modeling [17,18]. However, experimental studies demonstrated that the RTI may vary with velocity, temperature, and detector configuration [19,24]. Additionally, conductive heat losses can become significant under low-velocity conditions, thus causing deviations from the classical RTI model [16,19]. These limitations indicate that detector response cannot be fully represented by a single constant parameter.

The RTI is experimentally determined from the measured activation time, flow velocity, and temperature rise, as expressed in Eq. (2):

(2)

RTI=-tau-12ln(1-(△Ta△Td)

where t_a is the activation time, u¯ the average flow velocity, △T_a the temperature increase of the sensing element at activation, and △T_d the temperature increase of the surrounding gas. This formulation shows that the RTI is inversely related to detector sensitivity a lower RTI corresponds to a faster response.

Importantly, experimentally measured RTI values have been shown to vary over a wide range depending on the detector type and manufacturer, thus indicating that using a single representative value may not adequately capture detector behavior. Therefore, the RTI affects the prediction of detector activation when applied to fire-modeling environments, where local flow conditions must be considered.

2.2. Modeling of heat-detector activation in fire simulations

When applied to fire simulations, the thermal-response characteristics described above are critically governed by fire-induced flow fields, particularly ceiling jet behavior. Hot gases elevate due to buoyancy, impinge on the ceiling, and spread radially as a ceiling jet, thereby forming temperature and velocity fields that dictate detector response [14,15]. Empirical correlations developed by Alpert and Heskestad are commonly used to estimate these conditions and provide a basis for detector-activation analysis [14,15].

In fire simulation tool such as FDS, detector activation is modeled using an RTI-based formulation that integrates the local gas temperature and velocity with the thermal-response characteristics of the detector [20,21]. Activation is predicted when the calculated sensing element temperature reaches a specified activation temperature. Although this approach is practical and widely applied, its accuracy depends significantly on the reliability of the input parameters, particularly the RTI and activation temperature [8,23].

Previous studies demonstrated that variations in these parameters can significantly alter the predicted activation time [23,24]. Sensitivity analyses indicate that detector response is influenced by fire characteristics, such as the heat release rate, as well as by modeling conditions, including the local flow field and mesh resolution [3]. Although experimental methods, including wind tunnel- and FDE-based measurements, provide more reliable detector properties [25], their practical application remains limited owing to cost and data availability.

Overall, heat-detector activation modeling in fire simulations is governed by the interaction between thermal-response theory and fire-induced flow dynamics. Although RTI-based models offer a practical framework, their predictive accuracy is inherently affected by simplified assumptions and uncertainties in input parameters.

3. Limitations in Current Practice

Despite the theoretical framework and modeling approaches described in the previous section, significant limitations remain in the practical application of heat-detector models in PBD. In particular, the prediction of detector activation time is significantly affected by uncertainties in the input parameters and simplifications inherent in current modeling practices.

One of the primary limitations arises from the use of default input parameters in fire simulations. In many practical design cases, detector-specific properties such as the RTI and activation temperature are not available, thus resulting in the widespread use of generic or default values provided in simulation tool [8,22]. However, experimental studies demonstrated that these parameters vary significantly depending on the detector type, manufacturer, and sensing mechanism [23]. Such variability can result in substantial discrepancies in the predicted activation time, including under identical fire conditions [8,24]. Consequently, the use of default values introduces systematic uncertainty into fire simulation results and reduces the reliability of RSET estimation in PBD.

Another important limitation is the inherent variability of the RTI. Although the RTI is commonly regarded as an intrinsic detector property, several studies demonstrated that it varies with flow velocity, temperature, and installation conditions [19,24]. Deviations from the classical RTI assumption become more pronounced in low-velocity or non-uniform flow environments, where conductive heat losses and complex heat-transfer mechanisms are critical [16,19]. This indicates that the conventional RTI-based modeling approach may not fully capture the dynamic response of heat detectors under realistic fire conditions.

In addition to parameter uncertainty, limitations arise from the interaction between detector response and fire environment conditions. Detector activation is governed not only by the RTI and activation temperature but also by the fire growth rate, heat-release rate, and ventilation conditions [3,12]. Variations in these factors can significantly alter ceiling jet characteristics, thus resulting in varying thermal exposures at the detector location [14,15]. Consequently, even under identical detector properties, the activation time can vary considerably depending on the fire scenario. This coupling between fire dynamics and detector response further complicates accurate predictions in simulation-based designs.

Current modeling practices for heat-detector activation generally rely on simplified RTI-based formulations. These formulations facilitate the practical estimation of detector response times; however, they commonly assume idealized convective heating conditions and do not explicitly account for conductive heat losses or detailed detector structural effects [16,18]. Such simplifications may introduce uncertainty when detector models are applied to complex geometries or realistic fire scenarios. Researchers have attempted to mitigate this uncertainty by experimentally measuring detector properties using wind-tunnel or FDE systems [23,25]. The findings revealed that using measured RTIs and the activation temperature can improve the prediction of detector activation time [23]. However, the direct measurement of detector-specific properties is generally impractical in routine PBD applications as it requires a dedicated apparatus, experimental expertise, and additional time and cost. Therefore, an experimental detector-property database must be developed and applied to provide reliable input parameters without requiring additional testing at the design stage.

Current heat-detector modeling is characterized by uncertainties in the input parameters, variability in the detector response, and simplifications in modeling assumptions. These limitations reduce the reliability of fire simulation results and highlight the necessity for a systematic approach to selecting and applying detector input parameters. Addressing these issues is essential for improving the accuracy of detection-time prediction and enhancing the reliability of PBD for fire and life-safety analysis.

4. Experimental Data and Database of Heat-Detector Properties

4.1. Selection of heat detectors

The heat detectors considered in this study were selected based on their prevalence and practical use in fire protection systems used in South Korea. Previous investigations on fire-detection systems in South Korea identified commonly used fixed-temperature heat detectors across various building types [23]. Based on these findings, a representative set of detectors was selected to ensure that the experimental data reflected realistic design conditions.

Fixed-temperature heat detectors, which are widely installed in buildings and frequently considered in PBD analyses, were selected. Detectors from multiple manufacturers and with different structural characteristics were included to capture variability in response behavior. This approach allows the dataset to represent a broad range of detector properties instead of a single product-specific characteristic.

By incorporating multiple detector types, the resulting database provides a practical basis for general fire-simulation applications. This is particularly important in design scenarios where the exact detector specifications are not clearly defined, thus requiring the use of representative or averaged input parameters.

4.2. Source and measurement method of experimental data

The experimental data used in this study were obtained from previous studies where heat-detector properties were measured using an FDE [23,25]. The FDE system was designed to generate controlled flow and temperature conditions within a duct, thus allowing for a reproducible evaluation of detector-response characteristics.

As shown in Figure 1, the FDE comprises a flow-generation section, a heating unit, and a test section where detectors are installed. The system maintains a uniform velocity and temperature distribution, which is essential for the reliable measurement of the RTI and activation temperature. During the experiments, the detectors were exposed to controlled thermal environments, and their activation behavior was recorded to determine the RTI and activation temperature.

In this study, previously measured data were used; detailed descriptions of the experimental apparatus, calibration procedures, and measurement methodology are available in the literature [23]. Therefore, application of the resulting data is emphasized herein.

4.3. Measured detector properties and database

The experimental dataset included the RTI and activation-temperature values for multiple fixed-temperature heat detectors. Table 1 summarizes the measured properties of representative detectors, including the detector type, RTI, and activation temperature. The results indicate that both parameters varied significantly across different detectors. The classification terms used in this study follow Korean standards and are included to reflect practical detector usage; however, the measured activation-temperature ranges are consistent with low-temperature heat detectors commonly used in international practice.

Figure 2 presents the distribution of RTI values measured experimentally. The RTI values exhibited a wide range, from approximately 9.8 to 59.3 (m⋅s)^1/2 for thermistor-type detectors and up to 477.9 (m⋅s)^1/2 for bi-metal detectors, thereby indicating substantial variability in detector thermal sensitivity. Notably, the measured values did not center around the commonly assumed default value in FDS (approximately 100 (m⋅s)^1/2), thus highlighting a significant discrepancy between default assumptions and actual detector characteristics. The significantly higher RTI observed for the bi-metal detector reflects its inherently slower thermal response due to its larger thermal mass and structural characteristics.

Figure 3 shows the distribution of activation temperatures for the tested heat detectors. The measured activation temperatures ranged from 60.4 ℃ to 69.4 ℃, with an average value of approximately 65.3 ℃. Although this range generally aligns with nominal activation temperatures provided by manufacturers, deviations by several degrees were observed, thus suggesting that even detectors with similar nominal ratings may exhibit different activation behaviors.

These variations are consistent with previous findings, i.e., detector properties are not constant but depend on manufacturer specifications and structural characteristics [23]. The variability in the RTI and activation temperature directly affects the predicted activation time in fire simulations, thus highlighting the limitations of using default input values.

To support practical applications, statistical summaries of the measured data were derived, as listed in Table 2. The mean values can be used as representative inputs when detector-specific information is unavailable, whereas the range of values provides insight into the uncertainty associated with the detector response. The total statistical values were governed significantly by the inclusion of bi-metal detectors, which exhibited significantly higher RTI values than the thermistor-type detectors.

The constructed database effectively captures the variability of the heat-detector properties and provides a reliable foundation for improving the accuracy of detector activation modeling in fire simulations. This study demonstrates that using experimentally derived data is essential for reducing uncertainty in detection-time prediction and enhancing the reliability of performance-based fire-safety design.

5. Application of Experimental Database to FDS Input Parameters

The use of default input parameters for heat detectors in fire simulations is a major source of uncertainty in predicting detector activation time. As presented in Section 4, both the RTI and activation temperature exhibit significant variability depending on the detector type and manufacturer. Because the RTI and activation temperature directly control the thermal response of the detector, higher values generally result in delayed detector activation. In PBD, such a delay increases the detection-time component of the RSET and reduces the calculated safety margin between the ASET and RSET. Therefore, the conventional practice of applying fixed default values may not only cause discrepancies in the predicted activation time but also uncertainties in evacuation-safety evaluation.

Hence, the experimental database developed in this study provides a rational basis for selecting heat-detector input parameters in FDS. When detector-specific information is available during the design stage, the direct use of experimentally measured RTIs and activation temperatures is recommended. This approach enables an accurate representation of the thermal-response characteristics of the detector and minimizes the uncertainty associated with generalized assumptions. Additionally, the use of measured properties has been shown to improve the agreement between predicted and observed activation behaviors, thus further supporting the validity of this approach.

By contrast, in many practical design scenarios, the exact type and specifications of the heat detector may not be clearly defined at the analysis time. In such cases, using representative values derived from experimental databases provides a practical alternative. Statistical parameters, such as the mean RTI and activation temperature, can be used as generalized input values that reflect the typical behavior of commonly used detectors. Compared with arbitrary default values, this approach offers a more physically grounded and consistent basis for simulations.

Furthermore, the variability observed in the experimental data can be incorporated into the analysis to evaluate the uncertainty of the detector response. By considering the range of RTI and activation-temperature values, designers can perform sensitivity analyses to assess the effect of input-parameter variations on the predicted activation time. This enables the evaluation of both realistic and conservative scenarios, which is particularly important in PBD requiring explicit consideration of safety margins.

Based on these considerations, a structured framework for selecting heat-detector input parameters is proposed, as illustrated in Figure 4. The framework integrates detector-specific information and statistical data derived from experimental measurements, thus enabling flexible applications depending on the level of available information. When detailed detector specifications are provided, measured properties are directly applied; otherwise, representative mean values and variability ranges for the relevant detector type or category are used for sensitivity analysis to ensure consistency and robustness.

The proposed approach reduces reliance on arbitrary assumptions and enhances the consistency of input-parameter selection in fire simulations. Incorporating experimentally derived data into modeling enables a more reliable prediction of detector activation time and improves the accuracy of fire-safety assessment. Finally, this study provides a practical, scientifically grounded methodology that bridges the gap between theoretical detector modeling and real-world applications in PBD.

6. Conclusion

This study investigated the applicability of experimentally derived heat-detector properties for improving input-parameter selection in fire simulations. The analysis of experimental data demonstrated that both the RTI and activation temperature varied considerably depending on the detector type and manufacturer. This variability indicates that the common practice of applying fixed default values in FDS can introduce significant uncertainty in the prediction of detector activation time.

Hence, a structured approach for selecting input parameters was proposed based on the availability of detector-specific information. When detailed detector specifications are available, the direct use of experimentally measured properties enables a more accurate representation of thermal-response behavior. In cases where such information is not available, representative values derived from an experimental database provide a rational alternative to arbitrary default assumptions.

The results revealed that using experimental input data improved the reliability of detector activation prediction and enhanced the accuracy of RSET estimation in performance-based fire-safety design. Additionally, incorporating the variability of detector properties into sensitivity analysis allowed for a more comprehensive assessment of uncertainty in the fire simulation results.

This study demonstrated that the reliability of fire simulations depends significantly on the selected detector input parameters. Replacing fixed default values with experimentally derived or statistically representative data can significantly reduce uncertainty and improve the robustness of fire-safety assessments. The proposed framework provides a practical, scientifically grounded basis that bridges the gap between simplified detector modeling and realistic fire scenarios.

Notes

Author Contributions

Conceptualization, C.H.H. and H.S.H.; methodology, C.H.H. and H.Y.J.; software, H.S.H. and G.Y.Y.; validation, G.Y.Y. and H.Y.J.; formal analysis, H.S.H.; investigation, G.Y.Y.; resources, G.Y.Y.; data curation, G.Y.Y. and H.Y.J.; writing—original draft preparation, H.S.H. and G.Y.Y.; writing—review and editing, H.S.H. and B.K.; visualization, H.S.H. and G.Y.Y.; supervision, C.H.H.; project administration, C.H.H.; funding acquisition, C.H.H. All authors have read and agreed to the published version of the manuscript.

Conflicts of Interest

The authors declare no conflict of interest.

Acknowledgments

This study was partly supported by the National Fire Agency of Korea through field-based firefighting activities supporting the technology development project, MPSS-Fire Safety-2015-66.

Figure 1.

Schematic diagram of fire-detector evaluator system used for measuring heat-detector properties under controlled flow and temperature conditions.

Figure 2.

Distribution of response time index (RTI) for representative fixed-temperature heat detectors.

Figure 3.

Distribution of activation temperatures for representative fixed-temperature heat detectors.

Figure 4.

Proposed framework for selecting heat-detector input parameters in FDS based on experimental database and detector information availability.

Table 1.

Measured Properties of Representative Fixed-Temperature Heat Detectors

Sample	Heat Sensing Mechanism	Korean Classification^*	Activation Temperature (℃)	RTI (m⋅s)^1/2
A	Thermistor	Special Class	64.0 ± 1.03	13.5 ± 1.64
B			62.6 ± 0.94	17.3 ± 2.79
C			67.0 ± 0.41	11.7 ± 3.01
D			64.1 ± 0.08	9.8 ± 2.88
E		Class 1	68.4 ± 0.63	25.5 ± 1.89
F			60.4 ± 0.20	33.4 ± 7.81
G			67.7 ± 1.08	38.3 ± 5.85
H			62.6 ± 0.28	30.9 ± 8.58
I			66.5 ± 0.20	59.3 ± 6.57
J	Bi-Metal		69.4 ± 0.23	477.9 ± 20.3

^* Note: “Special Class” and “Class 1” are Classification Categories Defined in Korean Approval Standards for Fixed-Temperature Heat Detectors. These Classifications are Based on Detector Sensitivity and Do Not Directly Correspond to the ISO 7240-5 Classification System [7], Which Categorizes Detectors Based on Static-Response Temperature Ranges. However, the Activation Temperature Ranges Observed in This Study are Generally Consistent With Those of Low-Temperature Heat Detectors (e.g., A1-A2) in ISO 7240-5.

Table 2.

Statistical Summary of Measured Heat-Detector Properties

Parameter	Detector Type	Mean	Std. Dev.	Max.	Min.
RTI (m⋅s)^1/2	Thermistor / Special Class	13.1	3.20	17.3	9.8
	Thermistor / 1 Class	37.5	13.0	59.3	25.5
	Bi-Metal / 1 Class	477.9	20.3	-	-
	Total	71.8	143.5	477.9	9.8
Activation Temperature (℃)	Thermistor / Special Class	64.4	1.85	67.0	62.0
	Thermistor / 1 Class	65.1	3.46	68.4	60.4
	Bi-Metal / 1 Class	69.4	0.23	-	-
	Total	65.3	2.95	69.4	60.4

References

1. International Organization for Standardization (ISO), "Fire Safety Engineering-Technical Information on Methods for Evaluating Behaviour and Movement of People (ISO 16738)", ISO, Geneva, Switzerland (2009).

2. Society of Fire Protection Engineers (SFPE), "SFPE Handbook of Fire Protection Engineering", 5th ed., Springer, New York, USA (2016).

3. H. Y. Jang and C. H. Hwang, “Methodology for DB Construction of Input Parameters in FDS-Based Prediction Models of Smoke Detector”, Journal of Mechanical Science and Technology, Vol. 34, No. 12, pp. 5327-5337 (2020), https://doi.org/10.1007/s12206-020-1133-0.

4. National Fire Protection Association (NFPA), "NFPA 72: National Fire Alarm and Signaling Code", NFPA, Quincy, MA, USA (2019).

5. D. A. Purser, “ASET and RSET: Addressing Some Issues in Relation to Occupant Behavior and Tenability”, Fire Safety Science, Vol. 7, pp. 91-102 (2003), https://doi.org/10.3801/IAFSS.FSS.7-91.

6. International Organization for Standardization (ISO), "ISO 23932-1:2018, Fire Safety Engineering—General Principles—Part 1: General", ISO, Geneva, Switzerland (2018).

7. International Organization for Standardization (ISO), "ISO 7240-5:2018, Fire Detection and Fire Alarm Systems—Part 5: Point Type Heat Detectors", ISO, Geneva, Switzerland (2018).

8. H. W. Park, J. H. Cho, S. Y. Mun, C. H. Park, C. H. Hwang, S. C. Kim and D. G. Nam, “Measurement of the Device Properties of Fixed Temperature Heat Detectors for the Fire Modeling”, Fire Science and Engineering, Vol. 28, No. 1, pp. 37-43 (2014), https://doi.org/10.7731/KIFSE.2014.28.1.037.

9. T. Cleary, “Results From a Full-Scale Smoke Alarm Sensitivity Study”, Fire Technology, Vol. 50, pp. 775-790 (2014), https://doi.org/10.1007/s10694-010-0152-2.

10. H. C. Ryu and D. H. Kim, “Study on the Operation Characteristics of Heat Detectors Through Fire and Wind Tunnel Experiment”, Journal of Korea Society of Disaster Information, Vol. 11, No. 2, pp. 203-209 (2015), https://doi.org/10.15683/kosdi.2015.11.2.203.

11. S. H. Hong, “An Experimental Study on the Response Characteristics of Fire Detector for Early Stage Fire Detection in Warehouse”, Fire Science and Engineering, Vol. 30, No. 3, pp. 41-47 (2016), https://doi.org/10.7731/KIFSE.2016.30.3.041.

12. S. H. Sakong, S. K. Kim, C. H. Lee and J. J. Jung, “A Study on the Response Characteristics of Fire Detector by Full-Scale Experiment of Fire Phenomena in the Row House”, Fire Science and Engineering, Vol. 23, No. 3, pp. 67-72 (2009).

13. D. T. Gottuk, M. J. Peatross, R. J. Roby and C. L. Beyler, “Advanced Fire Detection Using Multi-Signature Alarm Algorithms”, Fire Safety Journal, Vol. 37, No. 4, pp. 381-394 (2002), https://doi.org/10.1016/S0379-7112(01)00057-1.

14. R. L. Alpert, “Calculation of Response Time of Ceiling-Mounted Fire Detectors”, Fire Technology, Vol. 8, pp. 181-195 (1972), https://doi.org/10.1007/BF02590543.

15. G. Heskestad, "Fire Plumes, Flame Height, and Air Entrainment", SFPE Handbook of Fire Protection Engineering, Springer, New York, USA (2016). https://doi.org/10.1007/978-1-4939-2565-0_13.

16. G. Heskestad and R. G. Bill, “Quantification of Thermal Responsiveness of Automatic Sprinklers Including Conduction Effects”, Fire Safety Journal, Vol. 14, No. 1-2, pp. 113-125 (1988), https://doi.org/10.1016/0379-7112(88)90049-5.

17. S. Nam, “Predicting Response Times of Fixed-Temperature, Rate-of-Rise, and Rate-Compensated Heat Detectors by Utilizing Thermal Response Time Index”, Fire Safety Journal, Vol. 41, No. 8, pp. 616-627 (2006), https://doi.org/10.1016/j.firesaf.2006.06.004.

18. FM Global, "Property Loss Prevention Data Sheets 5-48: Automatic Fire Detection", FM Global, Norwood, MA, USA (2005).

19. A. T. Pomeroy, "Analysis of the Effects of Temperature and Velocity on the Response Time Index of Heat Detectors", Master's Thesis, University of Maryland, College Park, USA (2010).

20. K. McGrattan, R. McDermott, M. Vanella, E. Mueller, S. Hostikka, J. Floyd and C. Paul, "Fire Dynamics Simulator Technical Reference Guide, Volume 1: Mathematical Model", NIST Special Publication 1018-1, National Institute of Standards and Technology, Gaithersburg, MD, USA (2025).

21. K. McGrattan, R. McDermott, M. Vanella, E. Mueller, S. Hostikka, J. Floyd and C. Paul, "Fire Dynamics Simulator User’s Guide", NIST Special Publication 1019, National Institute of Standards and Technology, Gaithersburg, MD, USA (2025).

22. T. G. Cleary and D. D. Evans, "Fire Detector Performance Predictions in a Simulated Multi-Room Configuration", NIST Special Publication 965, National Institute of Standards and Technology, Gaithersburg, MD, USA (2002). https://doi.org/10.6028/NIST.SP.965.

23. G. Y. Yoon, H. S. Han, S. Y. Mun, C. H. Park and C. H. Hwang, “DB Construction of Activation Temperature and Response Time Index for Domestic Fixed-Temperature Heat Detectors in Ceiling Jet Flow”, Fire Science and Engineering, Vol. 34, No. 3, pp. 35-42 (2020), https://doi.org/10.7731/KIFSE.103eea8f.

24. A. Tsui and M. J. Spearpoint, “Variability of Sprinkler Response Time Index and Conduction Factor Using the Plunge Test”, Building Services Engineering Research and Technology, Vol. 31, No. 2, pp. 163-176 (2010), https://doi.org/10.1177/0143624410363064.

25. K. S. Chin, "Development of Bench-Scale Testing of Sprinkler and Smoke Detector Activation/Response Time", Fire Engineering Research Report, University of Canterbury, New Zealand (2002).

26. D. D. Evans, "Ceiling Jet Flows", In SFPE Handbook of Fire Protection Engineering, 2nd ed., Section 2, Chapter 4, National Fire Protection Association, Quincy, MA, USA (2005).