Mineral-insulated, metal-sheathed, Type-K thermocouples are used to measure the temperature of various items in high-temperature environments, often exceeding 1000degC (1273 K). The thermocouple wires (chromel and alumel) are protected from the harsh environments by an Inconel sheath and magnesium oxide (MgO) insulation. The sheath and insulation are required for reliable measurements. Due to the sheath and MgO insulation, the temperature registered by the thermocouple is not the temperature of the surface of interest. In some cases, the error incurred is large enough to be of concern because these data are used for model validation, and thus the uncertainties of the data need to be well documented. This report documents the error using 0.062" and 0.040" diameter Inconel sheathed, Type-K thermocouples mounted on cylindrical surfaces (inside of a shroud, outside and inside of a mock test unit). After an initial transient, the thermocouple bias errors typically range only about +-1-2% of the reading in K. After all of the uncertainty sources have been included, the total uncertainty to 95% confidence, for shroud or test unit TCs in abnormal thermal environments, is about +-2% of the reading in K, lower than the +-3% typically used for flat shrouds. Recommendations are provided in Section 6 to facilitate interpretation and use of the results. .
Polymer foam encapsulants provide mechanical, electrical, and thermal isolation in engineered systems. It can be advantageous to surround objects of interest, such as electronics, with foams in a hermetically sealed container in order to protect them from hostile environments or from accidents such as fire. In fire environments, gas pressure from thermal decomposition of foams can cause mechanical failure of sealed systems. In this work, a detailed uncertainty quantification study of polymeric methylene diisocyanate (PMDI)-polyether-polyol based polyurethane foam is presented and compared to experimental results to assess the validity of a 3-D finite element model of the heat transfer and degradation processes. In this series of experiments, 320 kg/m3 PMDI foam in a 0.2 L sealed steel container is heated to 1,073 K at a rate of 150 K/min. The experiment ends when the can breaches due to the buildup of pressure. The temperature at key location is monitored as well as the internal pressure of the can. Both experimental uncertainty and computational uncertainty are examined and compared. The mean value method (MV) and Latin hypercube sampling (LHS) approach are used to propagate the uncertainty through the model. The results of the both the MV method and the LHS approach show that while the model generally can predict the temperature at given locations in the system, it is less successful at predicting the pressure response. Also, these two approaches for propagating uncertainty agree with each other, the importance of each input parameter on the simulation results is also investigated, showing that for the temperature response the conductivity of the steel container and the effective conductivity of the foam, are the most important parameters. For the pressure response, the activation energy, effective conductivity, and specific heat are most important. The comparison to experiments and the identification of the drivers of uncertainty allow for targeted development of the computational model and for definition of the experiments necessary to improve accuracy.
Polymer foam encapsulants provide mechanical, electrical, and thermal isolation in engineered systems. In fire environments, gas pressure from thermal decomposition of polymers can cause mechanical failure of sealed systems. In this work, a detailed uncertainty quantification study of PMDI-based polyurethane foam is presented to assess the validity of the computational model. Both experimental measurement uncertainty and model prediction uncertainty are examined and compared. Both the mean value method and Latin hypercube sampling approach are used to propagate the uncertainty through the model. In addition to comparing computational and experimental results, the importance of each input parameter on the simulation result is also investigated. These results show that further development in the physics model of the foam and appropriate associated material testing are necessary to improve model accuracy.
Expert panels comprised of subject matter experts identified at the U.S. National Laboratories (SNL, ANL, INL, ORNL, LBL, and BNL), universities (University of Wisconsin and Ohio State University), international agencies (IRSN, CEA, JAEA, KAERI, and JRC-IE) and private consultation companies (Radiation Effects Consulting) were assembled to perform a gap analysis for sodium fast reactor licensing. Expert-opinion elicitation was performed to qualitatively assess the current state of sodium fast reactor technologies. Five independent gap analyses were performed resulting in the following topical reports: (1) Accident Initiators and Sequences (i.e., Initiators/Sequences Technology Gap Analysis), (2) Sodium Technology Phenomena (i.e., Advanced Burner Reactor Sodium Technology Gap Analysis), (3) Fuels and Materials (i.e., Sodium Fast Reactor Fuels and Materials: Research Needs), (4) Source Term Characterization (i.e., Advanced Sodium Fast Reactor Accident Source Terms: Research Needs), and (5) Computer Codes and Models (i.e., Sodium Fast Reactor Gaps Analysis of Computer Codes and Models for Accident Analysis and Reactor Safety). Volume II of the Sodium Research Plan consolidates the five gap analysis reports produced by each expert panel, wherein the importance of the identified phenomena and necessities of further experimental research and code development were addressed. The findings from these five reports comprised the basis for the analysis in Sodium Fast Reactor Research Plan Volume I.
As the capabilities of numerical simulations increase, decision makers are increasingly relying upon simulations rather than experiments to assess risks across a wide variety of accident scenarios including fires. There are still, however, many aspects of fires that are either not well understood or are difficult to treat from first principles due to the computational expense. For a simulation to be truly predictive and to provide decision makers with information which can be reliably used for risk assessment the remaining physical processes must be studied and suitable models developed for the effects of the physics. The model for the fuel evaporation rate in a liquid fuel pool fire is significant because in well-ventilated fires the evaporation rate largely controls the total heat release rate from the fire. This report describes a set of fuel regression rates experiments to provide data for the development and validation of models. The experiments were performed with fires in the fully turbulent scale range (> 1 m diameter) and with a number of hydrocarbon fuels ranging from lightly sooting to heavily sooting. The importance of spectral absorption in the liquid fuels and the vapor dome above the pool was investigated and the total heat flux to the pool surface was measured. The importance of convection within the liquid fuel was assessed by restricting large scale liquid motion in some tests. These data sets provide a sound, experimentally proven basis for assessing how much of the liquid fuel needs to be modeled to enable a predictive simulation of a fuel fire given the couplings between evaporation of fuel from the pool and the heat release from the fire which drives the evaporation.
The dimensionless extinction coefficient, Ke, was measured for soot produced in 2m JP-8 pool fires. Light extinction and gravimetric sampling measurements were performed simultaneously at 635 and 1310nm wavelengths at three heights in the flame zone and in the overfire region. Measured average Ke values of 8.41.2 at 635nm and 8.71.1 at 1310nm in the overfire region agree well with values from 8-10 recently reported for different fuels and flame conditions. The overfire Ke values are also relatively independent of wavelength, in agreement with recent findings for JP-8 soot in smaller flames. Ke was nearly constant at 635nm for all sampling locations in the large fires. However, at 1310nm, the overfire Ke was higher than in the flame zone. Chemical analysis of physically sampled soot shows variations in carbon-to-hydrogen (C/H) ratio and polycyclic aromatic hydrocarbon (PAH) concentration that may account for the smaller Ke values measured in the flame zone. Rayleigh-Debye-Gans theory of scattering for polydisperse fractal aggregate (RDG-PFA) was applied to measured aggregate fractal dimensions and found to under-predict the extinction coefficient by 17-30% at 635nm using commonly accepted refractive indices of soot, and agreed well with the experiments using the more recently published refractive index of 1.99-0.89i. This study represents the first measurements of soot chemistry, morphology, and optical properties in the flame zone of large, fully-turbulent pool fires, and emphasizes the importance of accurate measurements of optical properties both in the flame zone and overfire regions for models of radiative transport and interpretation of laser-based diagnostics of soot volume fraction and temperature.
The dimensionless extinction coefficient, K{sub e}, was measured for soot produced in 2 m JP-8 pool fires. Light extinction and gravimetric sampling measurements were performed simultaneously at 635 and 1310 nm wavelengths at three heights in the flame zone and in the overfire region. Measured average K{sub e} values of 8.4 {+-} 1.2 at 635 nm and 8.7 {+-} 1.1 at 1310 nm in the overfire region agree well with values from 8-10 recently reported for different fuels and flame conditions. The overfire K{sub e} values are also relatively independent of wavelength, in agreement with recent findings for JP-8 soot in smaller flames. K{sub e} was nearly constant at 635 nm for all sampling locations in the large fires. However, at 1310 nm, the overfire K{sub e} was higher than in the flame zone. Chemical analysis of physically sampled soot shows variations in carbon-to-hydrogen (C/H) ratio and polycyclic aromatic hydrocarbon (PAH) concentration that may account for the smaller K{sub e} values measured in the flame zone. Rayleigh-Debye-Gans theory of scattering for polydisperse fractal aggregate (RDG-PFA) was applied to measured aggregate fractal dimensions and found to under-predict the extinction coefficient by 17-30% at 635 nm using commonly accepted refractive indices of soot, and agreed well with the experiments using the more recently published refractive index of 1.99-0.89i. This study represents the first measurements of soot chemistry, morphology, and optical properties in the flame zone of large, fully-turbulent pool fires, and emphasizes the importance of accurate measurements of optical properties both in the flame zone and overfire regions for models of radiative transport and interpretation of laser-based diagnostics of soot volume fraction and temperature.
The thermal hazard posed by large hydrocarbon fires is dominated by the radiative emission from high temperature soot. Since the optical properties of soot, especially in the infrared region of the electromagnetic spectrum, as well as its morphological properties, are not well known, efforts are underway to characterize these properties. Measurements of these soot properties in large fires are important for heat transfer calculations, for interpretation of laser-based diagnostics, and for developing soot property models for fire field models. This research uses extractive measurement diagnostics to characterize soot optical properties, morphology, and composition in 2 m pool fires. For measurement of the extinction coefficient, soot extracted from the flame zone is transported to a transmission cell where measurements are made using both visible and infrared lasers. Soot morphological properties are obtained by analysis via transmission electron microscopy of soot samples obtained thermophoretically within the flame zone, in the overfire region, and in the transmission cell. Soot composition, including carbon-to-hydrogen ratio and polycyclic aromatic hydrocarbon concentration, is obtained by analysis of soot collected on filters. Average dimensionless extinction coefficients of 8.4 {+-} 1.2 at 635 nm and 8.7 {+-} 1.1 at 1310 nm agree well with recent measurements in the overfire region of JP-8 and other fuels in lab-scale burners and fires. Average soot primary particle diameters, radius of gyration, and fractal dimensions agree with these recent studies. Rayleigh-Debye-Gans theory of scattering applied to the measured fractal parameters shows qualitative agreement with the trends in measured dimensionless extinction coefficients. Results of the density and chemistry are detailed in the report.
The dimensionless extinction coefficient (K{sub e}) of soot must be known to quantify laser extinction measurements of soot concentration and to predict optical attenuation through smoke clouds. Previous investigations have measured K{sub e} for post-flame soot emitted from laminar and turbulent diffusion flames and smoking laminar premixed flames. This paper presents the first measurements of soot K{sub e} from within laminar diffusion flames, using a small extractive probe to withdraw the soot from the flame. To measure K{sub e}, two laser sources (635 nm and 1310 nm) were coupled to a transmission cell, followed by gravimetric sampling. Coannular diffusion flames of methane, ethylene and nitrogen-diluted kerosene burning in air were studied, together with slot flames of methane and ethylene. K{sub e} was measured at the radial location of maximum soot volume fraction at several heights for each flame. Results for K{sub e} at both 635 nm and 1310 nm for ethylene and kerosene coannular flames were in the range of 9-10, consistent with the results from previous studies of post-flame soot. The ethylene slot flame and the methane flames have lower K{sub e} values, in some cases as low as 2.0. These lower values of K{sub e} are found to result from the contributions of (a) the condensation of PAH species during the sampling of soot, (b) the wavelength-dependent absorptivity of soot precursor particles, and, in the case of methane, (c) the negligible contribution of soot scattering to the extinction coefficient. RDG calculations of soot scattering, in combination with the measured K{sub e} values, imply that the soot refractive index is in the vicinity of 1.75-1.03i at 635 nm.
Fires pose the dominant risk to the safety and security of nuclear weapons, nuclear transport containers, and DOE and DoD facilities. The thermal hazard from these fires primarily results from radiant emission from high-temperature flame soot. Therefore, it is necessary to understand the local transport and chemical phenomena that determine the distributions of soot concentration, optical properties, and temperature in order to develop and validate constitutive models for large-scale, high-fidelity fire simulations. This report summarizes the findings of a Laboratory Directed Research and Development (LDRD) project devoted to obtaining the critical experimental information needed to develop such constitutive models. A combination of laser diagnostics and extractive measurement techniques have been employed in both steady and pulsed laminar diffusion flames of methane, ethylene, and JP-8 surrogate burning in air. For methane and ethylene, both slot and coannular flame geometries were investigated, as well as normal and inverse diffusion flame geometries. For the JP-8 surrogate, coannular normal diffusion flames were investigated. Soot concentrations, polycyclic aromatic hydrocarbon (PAH) laser-induced fluorescence (LIF) signals, hydroxyl radical (OH) LIF, acetylene and water vapor concentrations, soot zone temperatures, and the velocity field were all successfully measured in both steady and unsteady versions of these various flames. In addition, measurements were made of the soot microstructure, soot dimensionless extinction coefficient (&), and the local radiant heat flux. Taken together, these measurements comprise a unique, extensive database for future development and validation of models of soot formation, transport, and radiation.
A series of experiments was performed to better characterize the boundary conditions from an inconel heat source ('shroud') painted with Pyromark black paint. Quantifying uncertainties in this type of experimental setup is crucial to providing information for comparisons with code predictions. The characterization of this boundary condition has applications in many scenarios related to fire simulation experiments performed at Sandia National Laboratories Radiant Heat Facility (RHF). Four phases of experiments were performed. Phase 1 results showed that a nominal 1000 C shroud temperature is repeatable to about 2 C. Repeatability of temperatures at individual points on the shroud show that temperatures do not vary more than 10 C from experiment to experiment. This variation results in a 6% difference in heat flux to a target 4 inches away. IR camera images showed the shroud was not at a uniform temperature, although the control temperature was constant to about {+-}2 C during a test. These images showed that a circular shaped, flat shroud with its edges supported by an insulated plate has a temperature distribution with higher temperatures at the edges and lower temperatures in the center. Differences between the center and edge temperatures were up to 75 C. Phase 3 results showed that thermocouple (TC) bias errors are affected by coupling with the surrounding environment. The magnitude of TC error depends on the environment facing the TC. Phase 4 results were used to estimate correction factors for specific applications (40 and 63-mil diameter, ungrounded junction, mineral insulated, metal-sheathed TCs facing a cold surface). Correction factors of about 3.0-4.5% are recommended for 40 mil diameter TCs and 5.5-7.0% for 63 mil diameter TCs. When mounted on the cold side of the shroud, TCs read lower than the 'true' shroud temperature, and the TC reads high when on the hot side. An alternate method uses the average of a cold side and hot side TC of the same size to estimate the true shroud temperature. Phase 2 results compared IR camera measurements with TC measurements and measured values of Pyromark emissivity. Agreement was within measured uncertainties of the Pyromark paint emissivity and IR camera temperatures.
The primary objective of the Safety and Survivability of Aircraft Initiative is to improve the safety and survivability of systems by using validated computational models to predict the hazard posed by a fire. To meet this need, computational model predictions and experimental data have been obtained to provide insight into the thermal environment inside an aircraft dry bay. The calculations were performed using the Vulcan fire code, and the experiments were completed using a specially designed full-scale fixture. The focus of this report is to present comparisons of the Vulcan results with experimental data for a selected test scenario and to assess the capability of the Vulcan fire field model to accurately predict dry bay fire scenarios. Also included is an assessment of the sensitivity of the fire model predictions to boundary condition distribution and grid resolution. To facilitate the comparison with experimental results, a brief description of the dry bay fire test fixture and a detailed specification of the geometry and boundary conditions are included. Overall, the Vulcan fire field model has shown the capability to predict the thermal hazard posed by a sustained pool fire within a dry bay compartment of an aircraft; although, more extensive experimental data and rigorous comparison are required for model validation.
Fires in aircraft engine nacelles must be rapidly suppressed to avoid loss of life and property. The design of new and retrofit suppression systems has become significantly more challenging due to the ban on production of Halon 1301 for environmental concerns. Since fire dynamics and the transport of suppressants within the nacelle are both largely determined by the available air flow, efforts to define systems using less effective suppressants greatly benefit from characterization of nacelle air flow fields. A combined experimental and computational study of nacelle air flow therefore has been initiated. Calculations have been performed using both CFD-ACE (a Computational Fluid Dynamics (CFD) model with a body-fitted coordinate grid) and WLCAN (a CFD-based fire field model with a Cartesian ''brick'' shaped grid). The flow conditions examined in this study correspond to the same Reynolds number as test data from the full-scale nacelle simulator at the 46 Test Wing. Pre-test simulations of a quarter-scale test fixture were performed using CFD-ACE and WLCAN prior to fabrication. Based on these pre-test simulations, a quarter-scale test fixture was designed and fabricated for the purpose of obtaining spatially-resolved measurements of velocity and turbulence intensity in a smooth nacelle. Post-test calculations have been performed for the conditions of the experiment and compared with experimental results obtained from the quarter-scale test fixture. In addition, several different simulations were performed to assess the sensitivity of the predictions to the grid size, to the turbulence models, and to the use of wall functions. In general, the velocity predictions show very good agreement with the data in the center of the channel but deviate near the walls. The turbulence intensity results tend to amplify the differences in velocity, although most of the trends are in agreement. In addition, there were some differences between WLCAN and CFD-ACE results in the angled wall regions due to the Cartesian grid structure used by the WLCAN code. Also, the experimental data tended t o show poorer resolution near the walls of the transition ducts. The increased uncertainty in the data highlights some of the challenges in getting data near the walls due to the low signal to noise ratio. Overall, this effort provided a benchmark case for both the WLCAN and CFD-ACE codes for the application of interest.
As part of the full scale fuel fire experimental program, a series of JP-8 pool fire experiments with a large cylindrical calorimeter (3.66 m diameter), representing a C-141 aircraft fuselage, at the lee end of the fuel pool were performed at Naval Air Warfare Center, Weapons Division (NAWCWPNS). The series was designed to support Weapon System Safety Assessment (WSSA) needs by addressing the case of a transport aircraft subjected to a large fuel fire. The data collected from this mock series will allow for characterization of the fire environment via a survivable test fixture. This characterization will provide important background information for a future test series utilizing the same fuel pool with an actual C-141 aircraft in place of the cylindrical calorimeter.