In the field of marine engineering, biological animals move much more efficiently than traditional mechanical ships. When comparing a standard boat, fish swimming mechanics use far less energy to travel through heavy water. Because of this high efficiency, modern engineers want to build robotic fish for deep ocean exploration. However, testing physical robots in water tanks is expensive and slow. Therefore, researchers use a Fish Swimming CFD computer simulation to study the complex fluid dynamics before manufacturing the real machine. By utilizing a highly accurate Fish Swimming fluent simulation, designers can clearly measure the invisible water pressure and friction acting on the flexible body. This advanced CFD Analysis of Fish Swimming helps engineers choose the correct motor size and battery capacity for the robot. To successfully simulate a bending body inside a computer fluid environment, engineers must use moving grids. To learn the exact methods for setting up these changing computer boundaries, please explore our detailed Dynamic mesh tutorials.

Figure 1: Conceptual design of a Carangiform robotic fish inside a multiphysics computer environment, demonstrating the engineering approach to bio-inspired marine technology.

Simulation Process:  Dynamic Mesh Fluent Setup for Marine Kinematics

To accurately simulate the aquatic motion, we created a precise 3D digital model of a fish. Inside the Fish Swimming ANSYS Fluent software, we divided the surrounding water volume into a fine grid containing exactly 2,364,141 computational cells. To measure the microscopic friction correctly, we generated very thin cells against the surface of the fish skin, achieving a strict boundary layer y+ value. We then programmed the incoming water velocity to flow continuously at 0.3 m/s to match the natural forward swimming speed of the animal.

To create the physical swimming movement inside the Fish Swimming fluent solver, we applied the dynamic mesh tool. A standard solid grid cannot bend. Therefore, we wrote a specific User-Defined Function (UDF) code that commands the computer grid to bend in a natural “S-shape.” This code sets the tail beat frequency to 3 Hz, meaning the tail swings side to side three times per second. Because this continuous bending motion can easily destroy the grid, the software uses smoothing and remeshing mathematical algorithms. These rules carefully stretch and rebuild the 2.36 million cells at every small time step, ensuring the simulation calculates the fluid forces without failing.

Figure 2: Grid motion over time, displaying the successfully deforming computational cells around the bending fish tail.

Post-processing: Analysis of Hydrodynamic Forces and Swimming Efficiency

To understand the exact physical mechanics of the swimming motion, we must deeply analyze the velocity contours and the continuous force plots. By reading this data, engineers can determine how the flexible tail creates forward motion and how much physical resistance the robot will face. We begin by examining the 2D and 3D velocity magnitude contours. As the fish travels forward at an inlet speed of 0.3 m/s, the water flows very smoothly over the front head and the main body. The contours display calm cyan and green colors, indicating a steady local velocity between 0.3 and 0.5 m/s. This smooth flow means the water layer remains attached to the skin, which mathematically keeps the aerodynamic form drag very low. However, the fluid behavior changes significantly at the rear of the body. At the 1.5-second and 3.0-second time marks, the tail rapidly swings to the left and right. The simulation visually records bright red, high-velocity water jets reaching speeds of 0.6 to 1.0 m/s exactly at the tail section. In academic fluid dynamics, this specific pattern of fast water is known as a reverse von Kármán vortex street. This physical phenomenon proves that the tail successfully accelerates the slow water backward, which creates the necessary reaction force to push the fish forward.

While the velocity contours show the creation of thrust, we must review the drag force plot to measure the total hydrodynamic resistance. Throughout the simulation, the drag force line oscillates up and down exactly 15 times, which perfectly matches the 3 Hz swinging frequency of the tail. The computer calculates that the cycle-averaged drag force over time is -0.4 N. However, the exact data shows an important kinematic result: during the fastest part of the tail swing, the drag force line briefly crosses into positive numbers, reaching exactly +0.1 N. A positive force value in this specific axis proves that the tail temporarily generates a pure forward thrust that completely overcomes the water friction.

Figure 3: Drag force time history, showing the periodic oscillation between hydrodynamic resistance and pure thrust generation (+0.1 N).

Figure 4: Lift force time history, proving the excellent vertical stability of the fish model with an average force of +0.005 N.

Figure 5: Velocity contours, displaying the smooth attached flow over the body and the high-velocity red jets (1.0 m/s) shed by the tail.

Simultaneously, we must evaluate the lift force plot to confirm the vertical stability of the swimming model. If the lift forces are unbalanced, the robotic fish will continuously sink to the floor or float to the surface. The dynamic data reveals very small vertical force changes, smoothly bouncing between a maximum of +0.020 N and a minimum of -0.010 N. When we average these fluctuations over the full swimming cycle, the final lift force is only +0.005 N. This near-zero mathematical result proves that the 0.39-meter fish maintains excellent vertical stability, swimming in a straight horizontal path without wasting extra energy fighting vertical displacement.

Ultimately, marine engineers combine all of this highly accurate force and velocity data to calculate the required mechanical power for the robotic system. To continuously fight the steady -0.4 N drag force and maintain the constant forward speed of 0.3 m/s, the mathematical simulation dictates that the swimming motion requires Watts of power. This specific data point is the most important result for a robotic manufacturer.