TY - JOUR
T1 - Development of a Less Dissipative Interface Variable Reconstruction to Solve the Euler Equations by Q Learning Method
AU - Huang , Shing-Ian
AU - Wu , Sheng Chang
AU - Yang , Tsung-Yu
AU - Ting , Chi-Heng
AU - Wu , Yi-Jhen
AU - Niu , Yang-Yao
JO - Communications in Computational Physics
VL - 1
SP - 160
EP - 199
PY - 2024
DA - 2024/07
SN - 36
DO - http://doi.org/10.4208/cicp.OA-2022-0224
UR - https://global-sci.org/intro/article_detail/cicp/23300.html
KW - High resolution scheme, Euler equation, ATM method, discontinuity and detonation, Q-learning.
AB - <p style="text-align: justify;">In this study, we propose a blend of the average of THINC-EM and MUSCL
(ATM) methods based on the AUSMD scheme for solving detonation wave problems.
It is well known that the simulation of the detonation problems can produce incorrect
shock information or strong spurious due to the stiff source term. Accurate simulation
of detonation problems plays a crucial role in the design of detonation engines. The
proposed ATM method combines the MUSCL and THINC-EM methods with different
weighting functions, the optimized parameters of which are determined by the Q-learning method in order to accurately capture detonation waves, shock waves, and
expansion fans. To validate the proposed numerical method, one and two-dimensional
shock tube and the detonation tube and nozzles are chosen as benchmark test cases.
Our numerical results show that the proposed the ATM type AUSMD scheme has
great potential for handling more complex detonation problems and pulse detonation
engine flow problems.</p>