浅成岩体对热液成矿空间制约的复杂性的计算模拟

Focusing on the intrusions of Tongguanshan, Tianebaodan, Qingshanjiao, Xinwuli, and Yaojialing and their related orebodies in the Tongling district, this program is designed to apply the computational modeling methods to probe the complex spatial constraints from these shallow intrusions on the hydrothermal mineralization and their mechanism. Based on comprehensive analyzing the related documented geological information and data from our special geological investigation, geochemical assay and geophysical survey, we plan to simulate the fine 3D shape of these intrusions and their related orebodies, ore-hosting stratum units and ore-controlling faults on the MICROMINE-GOCAD platform through the knowledge-driven modeling method of integrating multi-sourced data under multi-scaled constraints. According to the results from the geological research and the above 3D geometric modeling, we plan to construct the computable dynamic 3D models of emplacement processes and cooling processes for these intrusions. These 3D dynamic models of intrusion-emplacement and of intrusion-cooling processes will be under computation by the distinct element algorithm in PFC3D and explicit finite difference algorithm in FLAC3D respectively. The computational results of these dynamic models are used to test and to modify the geological hypotheses about intrusions’ emplacement and intrusions’ cooling. By using of the methods of WofE and Fractal and Multi-Fractal statistical analysis, the data resulted from the above 3D geometric modeling and dynamic modeling experiments will be processed in the 3D GIS to analyze their corresponding association and statistical regularities, special to quantificationally reveal the spatial association between the topographic variety of intrusions and the mineralizing rooms, and to probe into that how the environmental and dynamic factors act as constraints on the location and scale of mineralizing rooms and their complexity. We will finally construct a series of theory models of intrusion’s controlling on the location and scale of mineralizing rooms in different tectonic setting.

以铜陵地区铜官山、天鹅抱蛋、青山脚、新屋里和姚家岭等岩体以及与之相关的矿体为主要研究对象，以计算模拟为主要手段探究岩体对成矿空间的复杂制约关系及其原因。在综合已有资料和针对性地质、地球物理和地球化学研究的基础上，以知识驱动多级约束下多源数据融合的方法，在MICROMINE-GOCAD平台上模拟各岩体及相关矿体、赋矿地层、成矿断层等的三维形态。进而分别建立各岩体侵位和岩体冷却演化的动力学模型，以离散元法和有限差分法分别进行岩体侵位和岩体冷却的动力学计算模拟，以模拟结果验证和完善有关岩体侵位和冷却演化的推断。在三维GIS平台上利用证据权重法、分形－多重分形统计法对三维形态模拟结果和动力学计算模拟结果进行空间对应分析及统计分析，定量分析岩体形态变化与成矿空间的关系，探究岩体侵位－冷却演化过程中环境和动力因素对成矿空间定位和规模的制约及其复杂性，最终建立岩体侵位和形态制约成矿定位和规模的理论模式。

浅成岩体所成热液矿床在金属矿床中占有非常重要的地位，但矿体的空间定位规律复杂，岩体在矿体定位中所起的作用也十分复杂。在传统地质调查、磁组构测量和地球化学测试的基础上，本项目以计算模拟为主要手段研究了铜陵地区的铜官山、狮子山和凤凰山以及北安第斯成矿带特里姆贝拉等浅成岩体成矿系统。磁组构特征显示成矿岩体都发生过多期次的侧向的非均匀增生，岩浆的各向异性侵位和岩体的非均匀增生制约着非均匀的成矿作用。考虑了岩浆侵位模式约束的岩体成矿系统三维形态模拟结果可视化地表征了岩体成矿系统的三维结构，显示岩体接触带的曲率变化越大、产状倾向岩体、走向与岩体冷却时的最大主应力方向一致的部位越有利于成矿。岩体所成矿床的规模与岩体的规模并没有直接关系，但与岩体的形态及围岩的构造密切相关。岩体形态越复杂、围岩呈平缓的背斜构造更有利于形成大规模矿体。岩体冷却过程的力-热-流耦合动力学计算模拟结果揭示了容矿空间形成的动力学机制，发现力-热-流耦合形成的扩容空间是最有利于矿体定位的容矿空间，岩体冷却时所处的构造应力环境、岩体接触带的形貌和产状、围岩的岩性组合的渗透率组合是决定扩容空间的规模和空间分布的重要因素。地球物理测量获得的电阻率数据能为推断岩体接触带及矿体的位置提供一定的信息，但并没有一一对应关系。岩体成矿系统的所有地质特征数据、地球物理探测数据、动力学计算模拟获得的虚拟数据都有助于了解岩体成矿的复杂性，而机器学习是挖掘这些数据与成矿关系的最佳方法。为了实现机器学习同时挖掘动力学计算模拟结果与实地探测结果数据，创立了坐标追踪-网格再生的三维数据融合方法。应用这一方法，以机器学习中的人工神经网络、支持向量机和随机森林的算法挖掘了铜官山、狮子山、凤凰山和安庆矿田等岩体成矿系统的特征及探测信息与成矿之间的关系，建立了定量三维预测模型，发现随机森林预测模型的预测效果最好。可推广应用到类似的浅岩体成矿系统。

内容获取失败，请点击重试