Urban Public Places, Crowds Of Risk Theory And Applied Research

With the sustainable and stable development of China's economy, urban public venues have burden more and more activities as the carrier of economy and culture, such as commerce, amusement, culture, transportation, sports and religious activities etc. Each activity is coupled with numerous of crowd gathering. Crowd incidents occurred frequently during the past decades, crowd safety has attracted high attention of the people. The main expression of the crowd risk for the large urban venues is crowds crush and trample disaster. As the typical disaster of crowd, crush and trample disaster occurred independency or as the ramification of other disasters, such as conflagration, explosion, leakage of poisonous gases etc. The evolvement of crowd disaster has the characteristic of dual mechanism, the occurrence and development is both have the aspect of certain and the aspect of uncertain. In this thesis, based on the research for the occurrence mechanism, evolvement mechanism consequence of crowd risk in large urban venues, we will try to find the inherent rules and propose effective measures to prevent or relieve the risk.Based on the analysis of the data for crowd risk during the past one hundred years, PMCDT (psychology, movement, communication, density and trigger factor) model which can describe the occurrence mechanism of crowd crush and trample disaster is proposed. The relationship between the threshold of lethal force and the duration time of the force has been expressed by a fitting function. Aimed at special disasters scenarios, the function of crowd density distribution is also founded. According to the actual circumstance of the venues, force models of crush disaster on plan land and on stairs are established separately. The relationship between duration time, casualty range and consequence is researched based on certain scenarios; finally, the quantitative model of the casualties is proposed. It is proved that the main casualties of the disaster is concerned with the spatial structure of the venues and the density of the crowd, the duration time of the disaster is not the main factor.The common phenomena concerns with crowd risk are researched. Firstly, the jam phenomenon of the evacuation route and the recovery process is investigated. Function of crowd density-crowd flow rate is proposed based on the characteristic of crowd dynamics, during the research process, crowd flow rate is taken as variable of crowd density to research on the recovery mechanism of the jam. It is proved the recovery process of the route is concerned with the initial crowd density, the duration time of the jam. The recovery time is approximate to the jam duration time when the initial density is lower than 3 p/m2, it is hard to recovery once the density reached 3-4p/m2, different functions of crowd density-crowd flow rate are used to compare the research results.It is find the bi-directional flow can affect the maximum crowd flow rate and the walking speed under maximum crowd flow rate, but hardly have effect on the free walking speed. The effect of the bi-directional flow is extremely strong when the difference of the crowd flow ratio between the both sides is large. The effect becomes decrease with the decrease of the gap between the crowd flow ratios of both sides. The maximum crowd flow rate is the minimum when the crowd flow ratio is 0.5. Further more, jam is easy to occur when the crowd flow ratio is between 0.7 and 0.8 under the maximum crowd flow rate.DTEM(discrete time evacuation model) is proposed to define the evacuation capacity of single exit building and its density during the evacuation process. The influence of the crowd density on the evacuation capacity of the egress is concern in the model, the evacuation processs are defined as C type and L type according to the trajectory of the evacuation. It is proved the evacuation time of C type is slightly lower than the L type with the same width of the egress. The difference between the two types will decrease with the width decrease of the egress. The results are appromixately to the results of simulation. Further more, it is also find the relationship between evacuation capacity and the effective width of the building egress is not linar.The model can be used to define the crowd density at the egress of different time during the evacuation, and confirm the optimum width of the egress.As the most common disaster type in urban public venues, fire is selected as the disaster scenario to research the crowd risk during the evacuation in a large supermarket. Event trees technology is used to define the probability of the different fire scenarios, CFAST is used to define the ASET of each scenario, methodology of simulation and crowd dynamics experiment are both used to define the RSET of different scenarios, and then the definition of the casualties. It is proved the crowd flow rate that derived from the crowd dynamics experiment is much larger than the result derived from the Simulex software when the crowd density is among the scope of 2.5 and 3.5p/m2; but the latter will lower than the former when the crowd density research 4.5 p/m2. The difference of the ERL (expectation risk of life) which derived from the two circumstances can reach two scalars. Prolong the pre-evacuation time suitable can prevent the overcrowding during the evacuation and thus can prevent the durst out of the crush and trample disasters. The difference of the evacuation time between different evacuees decrease with the increase of the pre-evacuation time; the set of obstacle in the evacuation route can decrease the evacuation time in high density evacuation, on the contrary, it will prolong the evacuation in low density.SSM(Soft System Mythology)is used to analysis the psychology of different roles who involves in the Miyun disaster, which happened in Feb, 2003. As the largest crush and trample disaster, the management insufficiency of the activity and the design deficiency of the bridge are researched separately, the main factors that resulted in the disaster are defined. Computer simulation and the revised crowd theory are all used to research the gathering process of the crowd; the results coincide with each other. The results show the process of crowd gathering is swift and the process of the disaster is also extraordinary swift. The whole process of the crowd gathering is only 23min and the whole process of the disaster is only 8min. So it is vital to strength the management of the key points in the venues. This can decrease the frequency of the disaster effectively.