Establish Indoor Fire Load Model Database And Random Fire Scenes

In modern buildings,a large number of reinforced concrete structures and steel structures are used.When a fire occurs,changes in the overall stress and deformation properties of structures are prone to cause collapse,resulting in huge economic losses and casualties.In practice,the occurrence of fire is affected by many factors,such as fire load density,opening factor,etc.Therefore,the occurrence of fire is random.Existing fire load models such as the ISO 834 standard fire curve cannot accurately describe the random fire scenes.They can only represent fires in a specific environment,but do not represent most of the potential fires in modern buildings.Therefore,the construction of random fire scenes can be used to more comprehensively predict the fire and prevent collapse of the structure due to fire.This paper summarizes the fire load model,and conducts random analysis of the key parameters in the indoor parametric fire curve.The probability distribution of the random parameters is given.Based on computer simulation,Latin hypercube sampling is used to analyze the uncertainty of the random parameters in the parametric fire curve to realize the construction of random fire scenes.The MySQL database and Python programming language are used to build a database and write application software to implement various functional modules in the database.The fire load model mainly includes the standard fire curve,real fire curve and indoor parametric fire curve.The fire load,opening factor and wall material characteristics are selected as the key factors for indoor fires,and the small-compartment parametric fire curves used in Europe and the large-space parametric fire curve used in China are taken as examples.Parametric studies are conducted on indoor parametric fire curves to determine the random and deterministic parameters in small-compartment and large-space fires,and to give the probability distribution function of random parameters.Among them,the random parameter of fire load density in the small-compartment parametric fire curve follows a normal distribution.The opening factor follows a log-normal distribution.The wall material characteristics are uniformly distributed.In the large-space parametric fire curve,the random parameter of fire load density is normally distributed,and the material flammability coefficient is uniformly distributed.The fire area and fire growth coefficient are log-normally distributed.The random sampling and Cholesky decomposition in Latin hypercube sampling are used to analyze the randomness of parametric fire curves.The KS test and sampling error analysis are used to verify the sampling algorithm.By taking the prediction model of concrete freezing service life as an example,the three variables of water-binder ratio,concrete air content and fly ash content are sampled,and the sampling results are used to verify the accuracy of the sampling algorithm.According to the sampling results of random parameters in the indoor parametric fire curve by Latin hypercube sampling,the SPSS and Origin software are used to fit the probability distribution of the key variables in the sampling results,and a fire curve with probability characteristics is given to realize the uncertainty analysis of the parametric fire curve for small compartments and large spaces.The Python language and MySQL database are used to realize the query and drawing of standard fire curve,real fire curve,parametric fire curves,maintenance of real fire curve database,and construction of Latin hypercube sampling and indoor-fire random scenes.The database is designed according to the acquisition system requirements,conceptual model design,logical structure design,physical structure design and database implementation.The SQL script of the ER model is exported through the MySQL Workbench software,which is input into the MySQL software to realize the design and construction of the database for standard fire curve,real fire curve and parametric fire curve.According to the designed database,various functional modules of application software are designed.Based on the development platform built by Py Charm and Anaconda and the Python programming language,these functional modules are realized and the written programs are written,debugged and run.