搜索

x

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

基于稳定性SPH-SWE数值模型的真实感流体动画实时模拟

邵绪强 梅鹏 陈文新

引用本文:
Citation:

基于稳定性SPH-SWE数值模型的真实感流体动画实时模拟

邵绪强, 梅鹏, 陈文新

Real-time simulation of realistic fluid animation based on stable SPH-SWE numerical model

Shao Xu-Qiang, Mei Peng, Chen Wen-Xin
PDF
HTML
导出引用
  • 流体动画模拟的真实感与实时性一直是流体模拟研究中的热点. 针对在复杂地形场景中不稳定的流体表面运动现象, 本文提出一种基于地形差异的自适应流体速度控制力, 建立了稳定性光滑粒子流体动力学方法以求解浅水方程数值模型. 首先, 将模拟域从三维降至二维表面来降低计算量, 通过粒子的密度大小表示其水深高度值; 其次, 采用变长光滑搜索半径, 确保搜索邻域内的粒子数目稳定在固定范围内, 提高模拟精度; 最后引入一种基于地形差异的自适应流体速度控制力, 根据粒子密度大小的实时变化来确定计算速度控制力的地形研究范围, 通过插值计算粒子运动前后时间步所处地形位置的差异来修正粒子的速度和位置. 本文使用屏幕空间流体渲染方法对流体表面进行绘制, 避免了表面网格的提取与重建, 流体的运动数值计算和渲染均被加载到GPU上并行化执行, 实验结果表明在达到实时交互级别的同时, 本文方法有效地改善了在复杂地形场景中流体表面的不稳定运动现象, 同时流体模拟过程中密度与压强的分布均匀.
    The reality and real-time performance have always been the research hot-point of fluid simulation. Aiming at the unstable fluid surface motion in the scenes with complex terrain, in this paper, we propose an adaptive fluid velocity control force calculation model based on terrain difference, and a stable SPH numerical model for solving the shallow water equations is established. In this proposed numerical model, firstly we reduce the simulation domain from three-dimensional space to two-dimensional surface for reducing calculation quantity, and the water depth is represented by the density of particles at the meantime. Secondly, to ensure that the number of neighborhood particles is stable within a fixed range and to improve the accuracy of simulation, we apply a variable smoothing length to our numerical model. Then, an adaptive fluid velocity control force calculation model is introduced based on terrain difference, in which the velocity and position of particles are corrected by calculating the terrain difference caused by particle movement between each time step. The coordinates on terrain used for the calculation of terrain difference are dynamically chosen by the density of the particles. To improve the real-time performance of simulation, a screen space fluid rendering method is used to refrain the extraction and reconstruction of fluid surface. The numerical calculation and fluid surface rendering both load on the GPU for parallel execution. The simulation result shows that the proposed method can effectively improve the unstable fluid surface movement in scenes with complex terrain while reaching a real-time interaction level. The density and pressure are evenly distributed during the simulation.
      通信作者: 梅鹏, 423744730@qq.com
    • 基金项目: 河北省自然科学基金(批准号: F2020502014)、中央高校基本科研业务费 (批准号: 2021MS095)和国家自然科学基金(批准号: 61502168)资助的课题
      Corresponding author: Mei Peng, 423744730@qq.com
    • Funds: Project supported by the Natural Science Foundation of Hebei Province, China (Grant No. F2020502014), the Fundamental Research Funds for the Central Universities of Ministry of Education of China (Grant No. 2021MS095), and the National Natural Science Foundation of China (Grant No. 61502168).
    [1]

    Harada T, Koshizuka S, Kawaguchi Y 2007 Proceedings of the 23rd Spring Conference on Computer Graphics Budmerice, Slovakia, April 26−28, 2007 p191

    [2]

    Mastin G A, Watterberg P A, Mareda J F 1987 IEEE Comput. Graph. 7 16Google Scholar

    [3]

    Chentanez N, Müller M, Kim T Y 2015 IEEE T Vis. Comput. Gr. 21 1116Google Scholar

    [4]

    Monaghan J J 1994 J. Comput. Phys. 110 399Google Scholar

    [5]

    Brodtkorb A R, Sætra M L, Altinakar M 2012 Comput. Fluids 55 1Google Scholar

    [6]

    Monaghan J J 1992 Annu​. Rev. Astron. Astr. 30 543Google Scholar

    [7]

    Swegle J W, Hicks D L, Attaway S W 1995 J. Comput. Phys. 116 123Google Scholar

    [8]

    Ata R, Soulaïmani A 2005 Int. J. Numer. Meth. Fl. 47 139Google Scholar

    [9]

    Rodriguez-Paz M, Bonet J 2005 Comput. Struct. 83 1396Google Scholar

    [10]

    Chang T J, Kao H M, Chang K H, Hsu M H 2011 J. Hydrol. 408 78Google Scholar

    [11]

    Xia X L, Liang Q H, Pastor M, Zou W L, Zhuang Y F 2013 Adv. Water Resour. 59 25Google Scholar

    [12]

    Chentanez N, Müller M 2010 Symposium on Computer Animation Goslar, Germany, July 2−4, 2010 p197

    [13]

    De Leffe M, Le Touzé D, Alessandrini B 2010 J. Hydraul. Res. 48 118Google Scholar

    [14]

    Chládek M, Ďurikovič R 2015 Comput. Graph. 53 170Google Scholar

    [15]

    Solenthaler B, Bucher P, Chentanez N, Müller M 2011 VRIPHYS 11: 8th Workshop on Virtual Reality Interactions and Physical Simulations Lyon, France, December 5−6, 2011 p39

    [16]

    Lee H, Han S 2010 Visual Comput. 26 865Google Scholar

    [17]

    张海超, 郑丹晨, 边茂松, 韩敏 2016 物理学报 65 244701Google Scholar

    Zhang H C, Zheng D C, Bian M S, Han M 2016 Acta Phys. Sin. 65 244701Google Scholar

    [18]

    Capecelatro J 2018 J. Comput. Phys. 356 174Google Scholar

    [19]

    van der Laan W J, Green S, Sainz M 2009 Proceedings of the 2009 Symposium on Interactive 3D Graphics and Games Boston, Massachusetts, February 27−March 1, 2009 p91

    [20]

    Müller M, Solenthaler B, Keiser R, Gross M 2005 Proceedings of the 2005 ACM SIGGRAPH/Eurographics Symposium on Computer Animation Los Angeles, California, July 29−31, 2005 p237

    [21]

    Fujisawa M, Nakada T, Mikawa M 2017 J. Inform. Processing. 25 486Google Scholar

    [22]

    Müller M, Charypar D, Gross M H 2003 Symposium on Computer Animation Goslar, Germany, July 26−27, 2003 p154

    [23]

    Liu M B, Liu G R, Lam K Y 2002 Shock Waves 12 181Google Scholar

    [24]

    Akinci N, Ihmsen M, Akinci G, Solenthaler B, Teschner M 2012 ACM T. Graphic. 31 1

    [25]

    Müller M, Schirm S, Duthaler S 2007 Proceedings of the 2007 ACM SIGGRAPH/Eurographics Symposium on Computer Animation Goslar, Germany, August 2, 2007 p9

    [26]

    dos Santos Brito C J, Almeida M W S, Vieira-e-Silva A L B, Teixeira J M X N, Teichrieb V 2017 2017 19th Symposium on Virtual and Augmented Reality (SVR). IEEE Curitiba, Brazil, November 1−4, 2017 p309

  • 图 1  地形引起粒子堆积、无序飞溅趋势

    Fig. 1.  The tendency of particle accumulation and disordered splashing caused by terrain.

    图 2  速度控制力计算模型

    Fig. 2.  The calculation model of velocity control force.

    图 3  地形整体平坦但局部凹陷严重

    Fig. 3.  A terrain which is flat on the whole but severely depressed locally.

    图 4  基于地形差异的自适应流体速度控制力计算模型

    Fig. 4.  An adaptive fluid velocity control force calculation model based on terrain difference.

    图 5  用于表示刚体的SPH粒子

    Fig. 5.  The SPH particles used to represent a rigid body.

    图 6  相机与流体表面关系示意图

    Fig. 6.  Schematic diagram of the relationship between camera and fluid surface.

    图 7  SSFR算法执行流程图

    Fig. 7.  The flow diagram of our SSFR algorithm.

    图 8  施加控制力前后效果对比 (a) 未施加控制力; (b) 施加控制力

    Fig. 8.  The comparison before and after applying control force: (a) Without adding control force; (b) with control force.

    图 9  流体与几何体交互 (a) 流体与长方体交互; (b) 流体与圆柱体交互

    Fig. 9.  Fluid interacts with geometry: (a) Fluid interacts with the cuboid; (b) fluid interacts with the cylinder.

    图 10  爱荷华河场景的四个时刻模拟渲染图 (a) 0 s时刻; (b) 30 s时刻; (c) 90 s时刻; (d) 120 s时刻

    Fig. 10.  Fluid surface rendering of the Iowa river scene at four moments: (a) 0 s; (b) 30 s; (c) 90 s; (d) 120 s.

    图 11  赤水河谷场景的四个时刻模拟渲染图 (a) 15 s时刻; (b) 30 s时刻; (c) 45 s时刻; (d) 60 s时刻

    Fig. 11.  Fluid surface rendering of the Chishui river valley scene at four moments: (a) 15 s; (b) 30 s; (c) 45 s; (d) 60 s.

    图 12  爱荷华场景水淹情况示意图 (a) 依据历年水淹范围统计的淹没概率图; (b) 2008年6月爱荷华雨洪淹没范围图; (c) 水深25 ft, 流量31200 cfs情况下淹没范围图; (d) 本文淹没模拟结果示意图

    Fig. 12.  The flood inundation maps of iowa scene: (a) The inundation probability map based on the statistics of the flooding range over the years; (b) the flood inundation map of iowa in June, 2008; (c) map of inundation range under the condition of 25 ft water depth and 31200 cfs water flow; (d) the simulated inundation map of our method.

    表 1  实验参数取值

    Table 1.  The values of parameters in experiment.

    物理量单位
    粒子半径 $ r $0.5m
    时间步长 $ \Delta t $ 0.02s
    初始粒子间距 $ {r_0} $1m
    初始搜索半径 $ {h_0} $2.66m
    静密度 $ {\rho _0} $1000kg/${{\rm{m}}^3}$
    粒子质量 $ m $1000kg
    气体常数 $ k $10
    黏度系数 $ \mu $0.1
    下载: 导出CSV

    表 2  实际水深与模拟水深数据对比

    Table 2.  Comparison between actual and simulated water depth

    点位模拟水深/m实测水深/m相对误差/%
    A1.291.439.79
    B1.511.658.48
    C1.971.922.60
    D1.501.407.14
    下载: 导出CSV

    表 3  各实验场景的平均模拟帧率

    Table 3.  The average simulation frame rate of each scene.

    场景粒子总数/103地形
    分辨率
    渲染前帧率/(ft·s–1)渲染后帧率/(ft·s–1)
    图8121024×10247034
    图9(a)12.21024×10246428
    图9(b)12.11024×10246530
    图1012512×5126832
    图1112256×2566833
    图12(d)601024×10247535
    下载: 导出CSV
  • [1]

    Harada T, Koshizuka S, Kawaguchi Y 2007 Proceedings of the 23rd Spring Conference on Computer Graphics Budmerice, Slovakia, April 26−28, 2007 p191

    [2]

    Mastin G A, Watterberg P A, Mareda J F 1987 IEEE Comput. Graph. 7 16Google Scholar

    [3]

    Chentanez N, Müller M, Kim T Y 2015 IEEE T Vis. Comput. Gr. 21 1116Google Scholar

    [4]

    Monaghan J J 1994 J. Comput. Phys. 110 399Google Scholar

    [5]

    Brodtkorb A R, Sætra M L, Altinakar M 2012 Comput. Fluids 55 1Google Scholar

    [6]

    Monaghan J J 1992 Annu​. Rev. Astron. Astr. 30 543Google Scholar

    [7]

    Swegle J W, Hicks D L, Attaway S W 1995 J. Comput. Phys. 116 123Google Scholar

    [8]

    Ata R, Soulaïmani A 2005 Int. J. Numer. Meth. Fl. 47 139Google Scholar

    [9]

    Rodriguez-Paz M, Bonet J 2005 Comput. Struct. 83 1396Google Scholar

    [10]

    Chang T J, Kao H M, Chang K H, Hsu M H 2011 J. Hydrol. 408 78Google Scholar

    [11]

    Xia X L, Liang Q H, Pastor M, Zou W L, Zhuang Y F 2013 Adv. Water Resour. 59 25Google Scholar

    [12]

    Chentanez N, Müller M 2010 Symposium on Computer Animation Goslar, Germany, July 2−4, 2010 p197

    [13]

    De Leffe M, Le Touzé D, Alessandrini B 2010 J. Hydraul. Res. 48 118Google Scholar

    [14]

    Chládek M, Ďurikovič R 2015 Comput. Graph. 53 170Google Scholar

    [15]

    Solenthaler B, Bucher P, Chentanez N, Müller M 2011 VRIPHYS 11: 8th Workshop on Virtual Reality Interactions and Physical Simulations Lyon, France, December 5−6, 2011 p39

    [16]

    Lee H, Han S 2010 Visual Comput. 26 865Google Scholar

    [17]

    张海超, 郑丹晨, 边茂松, 韩敏 2016 物理学报 65 244701Google Scholar

    Zhang H C, Zheng D C, Bian M S, Han M 2016 Acta Phys. Sin. 65 244701Google Scholar

    [18]

    Capecelatro J 2018 J. Comput. Phys. 356 174Google Scholar

    [19]

    van der Laan W J, Green S, Sainz M 2009 Proceedings of the 2009 Symposium on Interactive 3D Graphics and Games Boston, Massachusetts, February 27−March 1, 2009 p91

    [20]

    Müller M, Solenthaler B, Keiser R, Gross M 2005 Proceedings of the 2005 ACM SIGGRAPH/Eurographics Symposium on Computer Animation Los Angeles, California, July 29−31, 2005 p237

    [21]

    Fujisawa M, Nakada T, Mikawa M 2017 J. Inform. Processing. 25 486Google Scholar

    [22]

    Müller M, Charypar D, Gross M H 2003 Symposium on Computer Animation Goslar, Germany, July 26−27, 2003 p154

    [23]

    Liu M B, Liu G R, Lam K Y 2002 Shock Waves 12 181Google Scholar

    [24]

    Akinci N, Ihmsen M, Akinci G, Solenthaler B, Teschner M 2012 ACM T. Graphic. 31 1

    [25]

    Müller M, Schirm S, Duthaler S 2007 Proceedings of the 2007 ACM SIGGRAPH/Eurographics Symposium on Computer Animation Goslar, Germany, August 2, 2007 p9

    [26]

    dos Santos Brito C J, Almeida M W S, Vieira-e-Silva A L B, Teixeira J M X N, Teichrieb V 2017 2017 19th Symposium on Virtual and Augmented Reality (SVR). IEEE Curitiba, Brazil, November 1−4, 2017 p309

  • [1] 许晓阳, 周亚丽, 余鹏. eXtended Pom-Pom黏弹性流体的改进光滑粒子动力学模拟. 物理学报, 2023, 72(3): 034701. doi: 10.7498/aps.72.20221922
    [2] 张仁强, 蒋翔宇, 俞炯弛, 曾充, 宫明, 徐顺. 格点量子色动力学蒸馏算法中关联函数的计算优化. 物理学报, 2021, 70(16): 161201. doi: 10.7498/aps.70.20210030
    [3] 唐富明, 刘凯, 杨溢, 屠倩, 王凤, 王哲, 廖青. 基于图形处理器加速数值求解三维含时薛定谔方程. 物理学报, 2020, 69(23): 234202. doi: 10.7498/aps.69.20200700
    [4] 蒋涛, 黄金晶, 陆林广, 任金莲. 非线性薛定谔方程的高阶分裂改进光滑粒子动力学算法. 物理学报, 2019, 68(9): 090203. doi: 10.7498/aps.68.20190169
    [5] 蒋涛, 陈振超, 任金莲, 李刚. 基于修正并行光滑粒子动力学方法三维变系数瞬态热传导问题的模拟. 物理学报, 2017, 66(13): 130201. doi: 10.7498/aps.66.130201
    [6] 张海超, 郑丹晨, 边茂松, 韩敏. 一种基于二维光滑粒子法的流体仿真方法. 物理学报, 2016, 65(24): 244701. doi: 10.7498/aps.65.244701
    [7] 刘虎, 强洪夫, 陈福振, 韩亚伟, 范树佳. 一种新型光滑粒子动力学固壁边界施加模型. 物理学报, 2015, 64(9): 094701. doi: 10.7498/aps.64.094701
    [8] 马理强, 苏铁熊, 刘汉涛, 孟青. 微液滴振荡过程的光滑粒子动力学方法数值模拟. 物理学报, 2015, 64(13): 134702. doi: 10.7498/aps.64.134702
    [9] 雷娟棉, 杨浩, 黄灿. 基于弱可压与不可压光滑粒子动力学方法的封闭方腔自然对流数值模拟及算法对比. 物理学报, 2014, 63(22): 224701. doi: 10.7498/aps.63.224701
    [10] 蒋涛, 任金莲, 徐磊, 陆林广. 非等温非牛顿黏性流体流动问题的修正光滑粒子动力学方法模拟. 物理学报, 2014, 63(21): 210203. doi: 10.7498/aps.63.210203
    [11] 林晨森, 陈硕, 李启良, 杨志刚. 耗散粒子动力学GPU并行计算研究. 物理学报, 2014, 63(10): 104702. doi: 10.7498/aps.63.104702
    [12] 蒋涛, 陆林广, 陆伟刚. 等直径微液滴碰撞过程的改进光滑粒子动力学模拟. 物理学报, 2013, 62(22): 224701. doi: 10.7498/aps.62.224701
    [13] 苏铁熊, 马理强, 刘谋斌, 常建忠. 基于光滑粒子动力学方法的液滴冲击固壁面问题数值模拟. 物理学报, 2013, 62(6): 064702. doi: 10.7498/aps.62.064702
    [14] 达朝究, 冯爱霞, 龚志强, 宋健. 缓变下垫面对浅水方程的动力学订正. 物理学报, 2013, 62(3): 039202. doi: 10.7498/aps.62.039202
    [15] 杨秀峰, 刘谋斌. 光滑粒子动力学SPH方法应力不稳定性的一种改进方案. 物理学报, 2012, 61(22): 224701. doi: 10.7498/aps.61.224701
    [16] 马理强, 刘谋斌, 常建忠, 苏铁熊, 刘汉涛. 液滴冲击液膜问题的光滑粒子动力学模拟. 物理学报, 2012, 61(24): 244701. doi: 10.7498/aps.61.244701
    [17] 马理强, 常建忠, 刘汉涛, 刘谋斌. 液滴溅落问题的光滑粒子动力学模拟. 物理学报, 2012, 61(5): 054701. doi: 10.7498/aps.61.054701
    [18] 蒋涛, 欧阳洁, 赵晓凯, 任金莲. 黏性液滴变形过程的核梯度修正光滑粒子动力学模拟. 物理学报, 2011, 60(5): 054701. doi: 10.7498/aps.60.054701
    [19] 刘谋斌, 常建忠. 光滑粒子动力学方法中粒子分布与数值稳定性分析. 物理学报, 2010, 59(6): 3654-3662. doi: 10.7498/aps.59.3654
    [20] 廖臣, 刘大刚, 刘盛纲. 三维电磁粒子模拟并行计算的研究. 物理学报, 2009, 58(10): 6709-6718. doi: 10.7498/aps.58.6709
计量
  • 文章访问数:  6081
  • PDF下载量:  81
  • 被引次数: 0
出版历程
  • 收稿日期:  2021-07-05
  • 修回日期:  2021-08-06
  • 上网日期:  2021-08-17
  • 刊出日期:  2021-12-05

/

返回文章
返回