A 3rd-Order WENO Scheme for Stencil Smoothness Indicators Based on Mapping
-
摘要: 加权本质无振荡(weighted essentially non-oscillatory, WENO)格式能否具有低耗散特性及高分辨率特性, 关键在于光滑探测子的构造.该文针对三阶WENO格式的光滑探测子进行修正, 通过最光滑的探测子, 构造出了一个关于子模板光滑探测子的映射函数.在该函数作用下, 减小了欠光滑模板的光滑探测子, 进而增大了欠光滑模板的非线性权重.这明显地降低了格式的数值耗散, 提高了格式的分辨率.一系列数值测试表明,基于映射的模板光滑探测子的三阶WENO格式比传统的三阶WENO-JS3和WENO-Z3格式具有更高的分辨率.Abstract: The key to whether the WENO scheme can achieve optimal convergence accuracy and maintain essential no-oscillation characteristics near discontinuities lies in the construction of smoothness indicators. The smoothness indicator of the 3rd-order WENO scheme was modified through the construction of a mapping function to correct the smoothness indicators on each candidate stencil with the smoothest indicator. Under the influence of this mapping function, the smoothness indicator of the under-smooth stencil was reduced, thereby the nonlinear weight of the under-smooth stencil was increased. Then the numerical dissipation of the scheme was significantly lowered and its resolution was improved. A series of numerical examples demonstrate that, the new 3rd-order WENO scheme for smoothness indicators based on mapping has higher resolution than the classical WENO-JS3 and WENO-Z3 schemes.
-
Key words:
- hyperbolic conservation law /
- WENO /
- mapping /
- smoothness indicator /
- nonlinear weight
-
表 1 线性对流方程(25)在初值(26a)下, 不同格式在t=2.0时的L1误差和收敛阶
Table 1. L1 errors and convergence rates with t=2.0 of different schemes for linear advection eq. (25) with initial data (26a)
N WENO-JS3 WENO-Z3 WENO-JS3-M WENO-Z3-M L1 error (order) L1 error (order) L1 error (order) L1 error (order) 25 7.60E-2(-) 5.31E-2(-) 6.88E-2(-) 3.93E-2(-) 50 2.45E-2(1.63) 1.31E-2(2.02) 1.73E-2(1.99) 9.25E-3(2.09) 100 5.82E-3(2.07) 2.90E-3(2.18) 3.26E-3(2.41) 1.97E-3(2.23) 200 1.05E-3(2.47) 5.56E-4(2.38) 3.65E-4(3.16) 3.90E-4(2.34) 400 2.24E-4(2.23) 7.82E-5(2.83) 4.67E-5(2.97) 6.32E-5(2.63) 800 3.20E-5(2.81) 1.45E-5(2.43) 5.09E-6(3.20) 6.76E-6(3.22) 表 2 线性对流方程(25)在初值(26a)下, 不同格式在t=2.0时的L∞误差和收敛阶
Table 2. L∞ errors and convergence rates with t=2.0 of different schemes for linear advection eq. (25) with initial data (26a)
N WENO-JS3 WENO-Z3 WENO-JS3-M WENO-Z3-M L∞ error (order) L∞ error (order) L∞ error (order) L∞ error (order) 25 1.62E-1(-) 1.17E-1(-) 1.39E-1(-) 9.77E-2(-) 50 6.65E-2(1.28) 4.49E-2(1.38) 5.34E-2(1.38) 3.58E-2(1.45) 100 2.45E-2(1.44) 1.59E-2(1.50) 1.66E-2(1.69) 1.22E-2(1.55) 200 7.18E-3(1.77) 5.16E-3(1.62) 3.49E-3(2.25) 4.16E-3(1.55) 400 2.54E-3(1.50) 1.31E-3(1.98) 4.71E-4(2.89) 1.22E-3(1.77) 800 4.10E-4(2.63) 4.21E-4(1.64) 1.51E-4(1.64) 2.45E-4(2.32) 表 3 线性对流方程(25)在初值(26b)下, 不同格式在t=2.0时的L1误差和收敛阶
Table 3. L1 errors and convergence rates with t=2.0 of different schemes for linear advection eq. (25) with initial data (26b)
N WENO-JS3 WENO-Z3 WENO-JS3-M WENO-Z3-M L1 error (order) L1 error (order) L1 error (order) L1 error (order) 25 8.18E-2(-) 5.47E-2(-) 2.91E-2(-) 1.82E-2(-) 50 2.88E-2(1.51) 1.50E-2(1.87) 8.37E-3(1.80) 4.65E-3(1.97) 100 7.19E-3(2.00) 3.27E-3(2.20) 1.99E-3(2.07) 1.14E-3(2.03) 200 1.63E-3(2.14) 5.32E-4(2.62) 2.47E-4(3.01) 2.16E-4(2.40) 400 2.88E-4(2.50) 1.02E-4(2.38) 4.40E-5(2.49) 2.80E-5(2.95) 800 3.08E-5(3.23) 9.35E-6(3.45) 5.47E-6(3.01) 3.87E-6(2.86) 表 4 线性对流方程(25)在初值(26b)下, 不同格式在t=2.0时的L∞误差和收敛阶
Table 4. L∞ errors and convergence rates with t=2.0 of different schemes for linear advection eq. (25) with initial data (26b)
N WENO-JS3 WENO-Z3 WENO-JS3-M WENO-Z3-M L∞ error (order) L∞ error (order) L∞ error (order) L∞ error (order) 25 1.88E-1(-) 1.31E-1(-) 1.70E-1(-) 1.09E-1(-) 50 7.83E-2(1.26) 5.15E-2(1.35) 6.19E-2(1.46) 4.08E-2(1.42) 100 3.09E-2(1.34) 1.81E-2(1.51) 2.06E-2(1.59) 1.47E-2(1.47) 200 1.12E-2(1.46) 5.11E-3(1.82) 4.64E-3(2.15) 4.81E-3(1.61) 400 3.27E-3(1.78) 1.69E-3(1.60) 1.46E-3(1.67) 1.16E-3(2.05) 800 5.87E-4(2.48) 2.79E-4(2.60) 2.08E-4(2.81) 1.51E-4(2.94) 表 5 不同WENO格式关于若干一维Riemann问题的时间成本
Table 5. The time costs of a number of 1D Riemann problems with different WENO schemes
problem WENO-JS3 WENO-Z3 WENO-R3 WENO-JS3-M WENO-Z3-M Sod 0.146 5(1.00) 0.146 9(1.00) 0.175 5(1.20) 0.223 0(1.52) 0.224 9(1.53) Lax 0.195 3(1.00) 0.193 0(0.99) 0.230 9(1.18) 0.303 9(1.55) 0.305 1(1.56) Shu-Osher 0.261 1(1.00) 0.266 6(1.02) 0.316 0(1.21) 0.413 1(1.58) 0.414 7(1.59) blast 0.577 6(1.00) 0.599 3(1.04) 0.706 5(1.22) 0.900 3(1.56) 0.912 9(1.58) LeBlanc 9.414 0(1.00) 9.648 9(1.02) 11.559 4(1.23) 14.681 1(1.56) 14.889 6(1.58) -
[1] LIU X D, OSHER S, CHAN T. Weighted essentially non-oscillatory schemes[J]. Journal of Computational Physics, 1994, 115(1): 200-212. doi: 10.1006/jcph.1994.1187 [2] HARTEN A, ENGQUIST B, OSHER S, et al. Uniformly high order accurate essentially non-oscillatory schemes, Ⅲ[J]. Journal of Computational Physics, 1987, 71(2): 231-303. doi: 10.1016/0021-9991(87)90031-3 [3] JIANG G S, SHU C W. Efficient implementation of weighted ENO schemes[J]. Journal of Computational Physics, 1996, 126(1): 202-228. doi: 10.1006/jcph.1996.0130 [4] SHU C W, OSHER S. Efficient implementation of essentially non-oscillatory shock-capturing schemes, Ⅱ[J]. Journal of Computational Physics, 1989, 83(1): 32-78. doi: 10.1016/0021-9991(89)90222-2 [5] HENRICK A K, ASLAM T D, POWERS J M. Mapped weighted essentially non-oscillatory schemes: achieving optimal order near critical points[J]. Journal of Computational Physics, 2005, 207(2): 542-567. doi: 10.1016/j.jcp.2005.01.023 [6] BORGES R, CARMONA M, COSTA B, et al. An improved weighted essentially non-oscillatory scheme for hyperbolic conservation laws[J]. Journal of Computational Physics, 2008, 227(6): 3191-3211. doi: 10.1016/j.jcp.2007.11.038 [7] HA Y, KIM C H, LEE Y J, et al. An improved weighted essentially non-oscillatory scheme with a new smoothness indicator[J]. Journal of Computational Physics, 2013, 232(1): 68-86. doi: 10.1016/j.jcp.2012.06.016 [8] FAN P, SHEN Y Q, TIAN B L, et al. A new smoothness indicator for improving the weighted essentially non-oscillatory scheme[J]. Journal of Computational Physics, 2014, 269: 329-354. [9] YAN Z G, LIU H Y, MAO M L, et al. New nonlinear weights for improving accuracy and resolution of weighted compact nonlinear scheme[J]. Computers & Fluids, 2016, 127: 226-240. [10] KIM C H, HA Y S, YOON J H. Modified non-linear weights for fifth-order weighted essentially non-oscillatory schemes[J]. Journal of Scientific Computing, 2016, 67(1): 299-323. doi: 10.1007/s10915-015-0079-3 [11] CASTRO M, COSTA B, DON W S. High order weighted essentiallynon-oscillatory WENO-Z schemes for hyperbolic conservation laws[J]. Journal of Computational Physics, 2011, 230(5): 1766-1792. doi: 10.1016/j.jcp.2010.11.028 [12] ARÀNDIGA F, BAEZA A, BELDA A M, et al. Analysis of WENO schemes for full and global accuracy[J]. SIAM Journal on Numerical Analysis, 2011, 49(2): 893-915. doi: 10.1137/100791579 [13] 王亚辉. 求解双曲守恒律方程的三阶修正模板WENO格式[J]. 应用数学和力学, 2022, 43(2): 224-236. doi: 10.21656/1000-0887.420091WANG Yahui. A 3rd-order modified stencil WENO scheme for solution of hyperbolic conservation law equations[J]. Applied Mathematics and Mechanics, 2022, 43(2): 224-236. (in Chinese) doi: 10.21656/1000-0887.420091 [14] WANG Y H, DU Y L, ZHAO K L, et al. Modified stencil approximations for fifth-order weighted essentially non-oscillatory schemes[J]. Journal of Scientific Computing, 2019, 81(2): 898-922. doi: 10.1007/s10915-019-01042-w [15] WANG Y H, DU Y L, ZHAO K L, et al. A new 6th-order WENO scheme with modified stencils[J]. Computers & Fluids, 2020, 208: 104625. [16] WANG Y H. Improved weighted essentially non-oscillatory schemes with modified stencil approximation[J]. Computational and Applied Mathematics, 2023, 42(2): 82. doi: 10.1007/s40314-022-02075-y [17] ZENG F J, SHEN Y Q, LIU S P. A perturbational weighted essentially non-oscillatory scheme[J]. Computers & Fluids, 2018, 172: 196-208. [18] WU X, ZHAO Y. A high-resolution hybrid scheme for hyperbolic conservation laws[J]. International Journal for Numerical Methods in Fluids, 2015, 78(3): 162-187. doi: 10.1002/fld.4014 [19] WU X, LIANG J, ZHAO Y. A new smoothness indicator for third-order WENO scheme[J]. International Journal for Numerical Methods in Fluids, 2016, 81(7): 451-459. doi: 10.1002/fld.4194 [20] XU W, WU W. An improved third-order WENO-Z scheme[J]. Journal of Scientific Computing, 2018, 75(3): 1808-1841. doi: 10.1007/s10915-017-0587-4 [21] WANG Y H, DU Y L, ZHAO K L, et al. A low-dissipation third-order weighted essentially nonoscillatory scheme with a new reference smoothness indicator[J]. International Journal for Numerical Methods in Fluids, 2020, 92(9): 1212-1234. doi: 10.1002/fld.4824 [22] 王亚辉. 基于新的参考光滑性指示子的改进的三阶WENO格式[J]. 应用数学和力学, 2022, 43(7): 802-815. doi: 10.21656/1000-0887.420194WANG Yahui. An improved 3rd-order WENO scheme based on a new reference smoothness indicator[J]. Applied Mathematics and Mechanics, 2022, 43(7): 802-815. (in Chinese) doi: 10.21656/1000-0887.420194 [23] 徐维铮, 吴卫国. 三阶WENO-Z格式精度分析及其改进格式[J]. 应用数学和力学, 2018, 39(8): 946-960. doi: 10.21656/1000-0887.390011XU Weizheng, WU Weiguo. Precision analysis of the 3rd-order WENO-Z scheme and its improved scheme[J]. Applied Mathematics and Mechanics, 2018, 39(8): 946-960. (in Chinese) doi: 10.21656/1000-0887.390011 [24] LAX P D. Weak solutions of nonlinear hyperbolic equations and their numerical computation[J]. Communications on Pure and Applied Mathematics, 1954, 7(1): 159-193. doi: 10.1002/cpa.3160070112 [25] SOD G A. A survey of several finite difference methods for systems of nonlinear hyperbolic conservation laws[J]. Journal of Computational Physics, 1978, 27(1): 1-31. doi: 10.1016/0021-9991(78)90023-2 [26] WANG Y H, GUO C. Improved third-order WENO scheme with a new reference smoothness indicator[J]. Applied Numerical Mathematics, 2023, 192: 454-472. doi: 10.1016/j.apnum.2023.07.006 [27] WOODWARD P, COLELLA P. The numerical simulation of two-dimensional fluid flow with strong shocks[J]. Journal of Computational Physics, 1984, 54(1): 115-173. doi: 10.1016/0021-9991(84)90142-6 -