トップページ    計算科学研究センターとは    人材育成    計算科学eラーニング    R-CCS国際シンポジウム    第1回R-CCS国際シンポジウム    “Validation of alternative technology by direct turbulence simulation for towing tank experiment”

The 1st R-CCS International Symposium 18 Feb,2019
"Validation of alternative technology by direct turbulence simulation for towing tank experiment"

In the 1990s, many researchers of Computational Fluid Dynamics in the marine engineering field actively worked on the development of a new turbulence model in the framework of Reynolds-averaged Navier Stokes (RANS) Equation method. They mainly tuned to match the hull resistance and the flow pattern in the propeller section with the towing tank experiment. In the 2000s, some specialized tools were developed to enable ship designers to use easily in practice and began to be utilized widely in the early stages of hull designing. In the 2010s, estimation of the interference effect between hull and the propeller by self-propelled simulation became feasible as well as the resistance increase in wave and also cavitation characteristics. On the other hand, tank test has not been completely replaced by the simulation due to the accuracy and reliability problem of RANS. The most critical problem of RANS is that the turbulence model has to be tuned each time a new type of hull-form comes out by using towing tank test. And this tuning requires enormous effort and moreover, it sometimes occurs puzzling situation such as hull resistance matches, but the wake-flow distribution does not match, or vice versa. The purpose of this research is to develop an estimation technology of ship-propulsion performance with accuracy and reliability equivalent to towing tank test using Fully-resolved Large Eddy Simulation. This is a shift from the turbulence model dependent type to high-fidelity first principle approach. We have been conducting validation and verifications using results of towing tank test for several kinds of ship-types in collaboration with eight domestic major shipyards. This project started in 2011 and scheduled to complete in 2018. The cost of computational resource for a single vessel including both resistance and self-propulsion case estimated in the year 2012 was about 190 million yen, but estimation in 2018 is 3.7 million yen. Compared to towing tank tests costing from 5 million yen to 10 million yen per one vessel, now the price has been sufficient to satisfy the investment advantage of shipyards. For practical use in the real sense, we need more time for transition period to shift from tank test to computational simulation and activities related to this will also be introduced in the presentation.

Tatsuo Nishikawa
(Shipbuilding Research Centre of Japan)

資料

講師プロフィール

名前
Tatsuo Nishikawa
所属
Shipbuilding Research Centre of Japan

他の講義を探す

(2019年2月18日)