基于多球充填的玉米籽粒群体建模方法研究

发布时间：2019-03-13 11:11

【摘要】：玉米籽粒形状复杂,尺寸也不尽相同,如何建立其离散元法分析模型成为当今研究的一个热点。特别是,如何建立一个合理的、近似于实际玉米籽粒的籽粒群体模型仍然需要深入研究。本文以两个品种玉米籽粒为研究对象,测试了其物理力学性质,提出了基于多球充填的单个籽粒和玉米籽粒群体的建模方法。通过玉米籽粒的堆积、跌落、筛分等过程的仿真分析和试验结果的对比,验证本文所提出的玉米籽粒建模方法的可行性和有效性。本文的主要工作和结论如下。(1)以先玉335和白单678玉米籽粒为例,对其物理性质和力学性质(包括籽粒的几何形状、特征尺寸、密度、含水率、静摩擦系数和碰撞恢复系数等)进行测试与分析,并分析了含水率对玉米籽粒尺寸、密度的影响。研究表明,玉米籽粒的形状可以分为宽楔形、窄楔形、椭球锥和不规则形状4种类型。其中,宽楔形和窄楔形玉米籽粒占90%以上。因此,在对同一品种玉米籽粒群体建模时,可选择楔形玉米籽粒建立其几何模型;定义楔形籽粒的上底尺寸为主尺寸,其他尺寸根据与主尺寸间的关系式计算得到。(2)通过三维扫描仪对玉米籽粒进行扫描,获得玉米籽粒的三维点云数据,通过手动充填组成球的方法建立籽粒的模型。此外,本文提出一种基于多球充填的单个籽粒建模方法,通过定义并控制球心距的不同,对两个品种的宽楔形、窄楔形和楔形玉米籽粒分别建立了4种不同充填球数目的籽粒模型。(3)通过高速摄像技术记录玉米籽粒的下落堆积过程,并通过离散元法分析软件EDEM对4种籽粒模型的下落堆积过程进行仿真,通过仿真结果与试验结果的对比,验证本文提出的玉米籽粒建模方法的可行性和有效性。此外,通过对比容积密度的仿真与试验结果,并考察4种籽粒模型仿真计算的CPU时间,分析出较为合理的充填球数目的籽粒模型。研究表明,当充填球数目达到14-16时,所建立的玉米籽粒模型效果较好。此外,还通过对籽粒模型的跌落过程、筛分过程的研究和参数灵敏度的分析,进一步验证本文提出的籽粒建模方法的可行性。
[Abstract]:The shape and size of maize kernels are complex and different. How to establish the discrete element analysis model of maize seeds has become a hot topic in today's research. In particular, how to establish a reasonable, similar to the actual corn grain population model still needs to be studied in depth. In this paper, the physical and mechanical properties of two maize varieties were tested, and the modeling method of single grain and maize kernel population based on multi-ball filling was proposed. The feasibility and effectiveness of the corn kernel modeling method proposed in this paper are verified by the simulation analysis of the process of accumulation, dropping and screening of corn seeds and the comparison of the test results. The main work and conclusions of this paper are as follows. (1) taking Xianyu 335 and Baidan 678 maize seeds as an example, the physical and mechanical properties (including the geometric shape, characteristic size, density, moisture content of the seeds) were studied. The static friction coefficient and collision recovery coefficient were tested and analyzed, and the effects of moisture content on grain size and density of maize were analyzed. The results showed that corn grains could be divided into four types: wide wedge, narrow wedge, ellipsoid cone and irregular shape. Among them, wide wedge and narrow wedge corn grains accounted for more than 90%. Therefore, when modeling the grain population of the same variety of maize, we can select the wedge corn kernel to establish its geometric model. The upper and bottom dimensions of wedge-shaped grains are defined, and other dimensions are calculated according to the relationship between the main dimensions. (2) three-dimensional point cloud data of corn seeds are obtained by scanning corn seeds with a three-dimensional scanner. The model of grain was established by filling the forming ball manually. In addition, a single grain modeling method based on multi-ball filling is proposed in this paper. By defining and controlling the difference of spherical center distance, the wide wedge of two varieties is defined and controlled. Four seed models with different filling number were established for narrow wedge and wedge corn seeds respectively. (3) the falling accumulation process of corn seeds was recorded by high-speed camera technology. The falling accumulation process of four grain models was simulated by discrete element analysis software EDEM. By comparing the simulation results with the experimental results, the feasibility and effectiveness of the corn kernel modeling method proposed in this paper were verified. In addition, by comparing the simulation and test results of volume density, and investigating the CPU time calculated by four kinds of grain models, a more reasonable grain model with the number of filling balls was analyzed. The results showed that the corn kernel model was effective when the number of filling balls reached 14-16:00. In addition, through the drop process of grain model, screening process and parameter sensitivity analysis, the feasibility of the grain modeling method proposed in this paper is further verified.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：S513

【参考文献】