Full scale CFD simulation for marine design: a review by Milovan Peric

Computational fluid dynamics (CFD) is becoming widely used in marine design. Generally, CFD is run at model scale to compare with testing tank validation data. However, translating model-scale results to full-scale is difficult since Reynolds number scaling is not observed. This means that an optimal design at model scale often does not match full-scale requirements. Studying interactions between different scale vessel components such as hull-propeller interactions, or energy saving devices, is also not easy at model scale. 

In this white paper, Professor Milovan Peric examines some common reservations about running CFD at full scale and encourages full-scale analysis of marine designs under realistic operating conditions. The paper looks at the effects of Reynolds number scaling, as well as computational mesh requirements, and shows examples of full-scale CFD simulation on complex cases. In many cases, full-scale simulation is more accurate and reliable than alternatives and leads to a greater understanding of design performance.

Scale analysis examples

The author examines scale analysis examples such as the effects of Reynolds number scaling, as well as computational mesh requirements, and shows examples of full-scale CFD simulation on complex cases. In many cases, full-scale simulation is more accurate and reliable than alternatives and leads to a greater understanding of design performance.

Analysis and design of marine structures

Full-scale analysis and design of marine structures and complete systems require digital
twin technology. With robust software, simulation can be performed to mirror realistic operating conditions to ensure proper design is found. Many experienced users in the industry successfully apply CFD simulations with full-scale, read more from this review to learn the benefits of full-scale CFD.

Scale prediction software

Use cases of full-scale prediction software, and much more, can be found in this marine industry resource. In many cases, full-scale prediction is more reliable than scaling up model experiments.