Specification Design and Performances Using Computational Fluid Dynamics for Mini-Remotely Operated Underwater Vehicle

— Remotely Operated Underwater Vehicle (ROV) is a submersible robotic system, used to examine various underwater characteristics and controlled by operators from shore. The applications of ROV are widely diverse, especially for exploration and marine industry. This research is to investigate pressure and drag on underwater vehicles using the CFD approach. A low-cost mini ROV with a 300 mm length was used for the research and to determine the hardware and material type required for prototype production by observing pressure and drag from the CFD simulation result. Simulation is solved using CFD software with RANSE method and Shear Stress Transport based k- ɷ was used as turbulence setting. The result indicated that to pull the mini ROV from the surface to a depth of 50 m requires a thrust of 3,9 kg and the mini ROV receiving the pressure of 7,7 bar.


I. INTRODUCTION
Remotely Operated Underwater Vehicle (ROV) is a submersible robotic systems, used to examine various underwater characteristics and controlled by operators from shore [1].With complex, dangerous and limited areas urgently explored, there is an urgent need for an underwater machine that can replace humans to complete underwater detection. ROV were developed to perform resource exploration tasks in the ocean [2]. The applications of ROV are widely diverse, such as the oil and gas industry, discovery, aquaculture, marine biology, and military purpose [3] [4][5] [6].
Numerous of ROV designs are assembled around or inside cubic structured frame along with a buoyant on top body. The robotic vehicle utilized umbilical data cable as electrical cable and controlled by operator positioned on surface vessel.
The heavier tools are settled down on lowest possible position to keep the point of center buoyancy is higher than center of gravity, thus adequate stability is obtained. Sufficient stability of ROV meant ROV has capability to withstand against in disturbance for instance moment of rolling and pitch in longitudinal and lateral axis. Steady hydraulic rods suited of lifting and carrying particular equipment for specific purpose and placed in the fore along with cameras and lights.
Underwater ROVs are generally divided categories based on size, weight, capacity, or performance. Two common design of ROV which widely use are: the ROV Micro-Class with weigh not higher than 3 kg, are used especially in narrow area where a diver might not be capable to enter. And the ROV Mini-Class that weigh about 15 kg are also used as a diver alternative, Inspection-Class are typically rugged ROVs for commercial or industrial use, data acquisition and observation ROVs, Light weight duty category has generally no more than 50 hp on propulsion, While Heavy duty type has less than 220 hp with the ability to capability lift at least two handlers and work at depths up to 3500 m, and Trenching & Burial work class is the largest with more than 200 hp propulsion, with the ability to take a cable, put down sleds and working at depths until 6000 m in several cases [7][8][9] [10]. The ROV design and construction is a robust solution to encounter the various require of an ROV to be used in wide implementations, it is compact, handy to use and inexpensive, and allow for confined space exploration. The main form of the body is designed to encounter the particular utilization needs. Commonly, almost ROV attached by torpedo form and an hydrodynamic main body was used at high-speeds [11]. Another type is the torpedo-less form, worked mainly in remote control vehicles (ROVs) that are often worked for shorter period of time purpose or the assessment of other large subsea areas like huge ice floe or water-dams [12]. Toward to emphasize the performance in water resistance, Computational Fluid Dynamics (CFD), drives valuable part in Unmanned vehicle design. Nowadays, CFD approach has been widely applied in underwater vehicle design, as it can simulate the fluid flow field around the bodies and to find out with comprehend the physics of fluid flow phenomenon. CFD captured the flow field such as vortex in very small to major scales which is hard to obtain through field experiments.
CFD methods enable to predict and diagnostic technique for identifying the cause of particular problem in physic phenomena. The latest CFD approaches provide wealth data generated emphasize to ISSN 2355-3286 visualize the implication of the result requires considerable skill. Due to this CFD become as an integral part of the design process.
From the economic cost, numerical simulation can deliver result within sufficiently short time span in the design process, thus optimization design can be detailed in addition to addressing this issue during the design process. Numeric simulation is considered low cost analysis hence became early stage to understanding along side with experimental test which can be over cost and intensive time.
In this project, the CFD analysis assists in investigate and demonstrate variable model for the hull body form based on initial prototype [13]. In this research, a design of Mini-ROV is developed especially the principal of body device in each element and electric parts is embedded. Furthermore, the hydrodynamic parameter and analysis was provided based on CFD simulation.

A. Numerical Modelling
Hydrodynamics reference input was set for the mini-ROV. Therefore simulation of the flow around model is accomplished. The moment coefficient of symmetry flow in both xy plane and xz plane tend to be zero. Physical properties of this simulation conducted at Re= 8.5 x 10 5 , ρ = 1025 kg/m 3 , ν = 1 x 10 -6 m 2 /s and U = 1 m/s or Fr = 0.6. CFD simulation is used to analyze the total force and pressure difference between horizontal dive, vertical dive, and diagonal dive. The CFD simulation were run on 1:1 scale, 3-Dimension ROV model as shown in Figure 1 and main dimensions of mini-ROV model shown in Table 1.

B. Governing Equation RANS
CFD simulation using Reynolds averaged Navier Stokes (RANS) equations implemented in this project, attained by governed general laws for mass, momentum and energy. This method the flow considered as incompressible. The equation of RANS for incompressible flow shown in the following: The derivation of RANS equation from instantaneous of Navier-Stokes equation is the Reynolds decompositions, which refers to separation flow (like velocity u) into mean (time averaged) and fluctuating components (like ′ ) equal to zero.
The subscripts i and j defined the i ' and j ' components of the cartesian coordinate respectively, ̅ and ̅ defined the time averaged velocity and pressure respectively, t is the time and ρ is the density of the fluid then ̅̅̅ is represent the mean viscous stress tensor.
For the turbulent model adopted k-ɷ based Shear Stress Transport (SST). Standard k-ɷ turbulence model is basically combination of k-ɷ model [14]. The two equation model along with conservation law solved two-transport partial differential equations which acquired turbulence energy diffusion and convection energy. The variables of transport was turbulent kinetic energy (k) indicated turbulent energy and the other was specific turbulent dissipation rate (ɷ) also known as turbulence scale.
Later k-ɷ modified with Shear Stress Transport (SST) in order to avoid the k−ω problem of being sensitive to the inlet free-stream turbulence properties [15]. The modified model has capability carried out the transport of the shear stress and satisfied adverse pressure gradient boundary layers and separating flows.

C. Computational Domain
When conducting CFD numerical simulation, computational domain has significant impact for defining circumstance around the body which engaged with physical condition in fluid dynamics constraint. The space in boundary condition should broad enough to avoid non realistic physic phenomena which lead disrupted the result. The domain spaces where the simulation of numerical solution will be calculated, adequate space let fluid flow surround geometry model did not interfere with boundary domain themselves. The boundary need be discrete in numerous mesh element with structured or unstructured mesh form thus simulation will solved with fluid dynamics equation. The domain came with several shape, in this simulation hexahedron domain was applied. The computational domain for simulation of the ROV model which expressed 2.5L forward the fore of main hull, 4.5L backward the aft of main hull, 2L aside of main hull, 1.5L above upward of the support frame, and 1.5L downward under of the support frame as shown in Figure 2.

D. Boundary and Initial Condition
When attempted CFD analysis, the input of set constraint to the boundary value must be defined and connect the simulation model with its surroundings.
Inlet is defined as flow velocity or known as velocity inlet. The turbulent length and turbulent intensity were given for fully developed flow. Outlet identified as pressure outlet where the gauge pressure is none or equal to zero. The flow was considered a fully develops, there is no switch turn out in the flow direction and the outlet is designed long distant from the geometrical model disturbances. Wall defined as cube around model. The no-slip condition was determined at whole surface of hexahedron, hence the velocity on surface is zero and there is no stream flow passes through the surface of cube. The final boundary is defined as symmetry was placed at the bottom, top and both sides of the computational domain. The conditions at symmetric boundary are no flow across boundary. The normal velocity and normal gradient of all variable are zero it acted as a mirror that reflects all the flow distribution to the other side. Again any variables have same value and gradients at the same distance from the boundary.

E. Meshing
CFD is a way of analyzing fluid flows numerically on a computer by a set of algebraic equations. The algebraic equations are obtained by discrete the partial differential equations (PDEs), which may be conservation of mass, momentum, energy etc. The solution is attained at discrete points and hence the computational domains need to be discrete into discrete areas or volume which is the mesh. For this reason, the mesh resolution is of great importance.
The mesh is discrete domain in which the flow problem is to be solved. The flow properties are calculated in every mesh and the density of the cells for instance the mesh resolution is of great importance. Basically, the grid could be either structured or unstructured. In an unstructured mesh such as unregulated tetrahedral, the cells are arranged in an apparently random fashion in contrast to a structured mesh. Mesh generator was carried out to generated mesh for RANS solution.
Unregulated tetrahedral mesh first was generated using Mesh Modeler around the mini-ROV model as shown in Figure 3. Smaller element size of unregulated tetrahedral was used for the ROV model and for the region around the ROV model with refinement is designate to capture the fluid flow properties and to achieve a finer result. The mesh outside the model, defined as fluid then larger mesh built gradually to reduced time computational aspect.
The mesh number is affected the solution, therefore a convergence process was used in order to achieve mesh independent solutions. If mesh independent is obtained, further refinement of the element does not affect the solution.

F. Numerical Solver Setup
The fluid flow analysis is expressed using CFD based RANS method. Turbulence setting using Shear Stress Transport based k-ɷ with turbulence option first order is used to calculate the simulation. Steady state was used for simulation with total elements about 1 million to satisfy mesh independent criteria. This simulation was converged with Residual RMS level 1x10 -5 was used to investigate. The residual is fundamental measures of an iterative solution convergence, as it directly quantifies the error in the solution of the system of equations. The properties of numerical simulation for fluid flow analysis of ROV model as shown in Table 2.

A. Mesh Indepedence
The research was performed using the ROV model with horizontal dive simulation to obtain grid independence solution. Mesh independence used for validating simulation result. The sum of element analyze was used to determine the effect of element number on the calculated total force. To build the mesh, the sum of mesh from domain was sequentially reformed, whilst keeping grow rate parameters and the total element ROV models the same. The results of ROV model at different sum of element for Fr 0.6 from simulations are shown in Table 3. In this case, 6 difference meshes was used to validate mesh independence.  The solution converged successfully, with variation of total force as shown in Table 3. In addition, the difference in the total force between Mesh 5 and Mesh 6 are not greater than 2 % [15]. Therefore, the sum of elements on Mesh 5 was chosen and used in all cases of the simulation.

B. The Result of CFD Simulation
Numerical analysis of the ROV model was performed. Computational data was acquired using RANSE code and Turbulence model was selected. The model was run at three particular moves for instance horizontal, diagonal and vertical dive. All cases were run at Fr 0.6 which equals to 2 knots. The model tested in fully submerged simulation which equal 0 to 10 m of depth. The CFD simulation generated physic contour visualization around body mini-ROV model. The movement created phenomena that plot flow pattern in various contour lines. It has color bands that discrete colored regions while the display of a variable on a locator (such as a boundary) shows a finer range of color detail gradually. The contour of water superficial velocity and water velocity are presented in Figure 4  Rear body has transom formed, where body has not streamlined which lead the fluid particle did not fully transmitted far away behind body hence suddenly suffered small turbulence [16].
In CFD simulation, the value of total drag and total pressure caused by ROV movement was obtained directly, the result shown in table 4, 5 and 6. It shows the value at each step Froude number, with unit Newton in total drag and Pascal in total pressure respectively. This value became major lead reference in determining motor thrust. Since the mini-ROV mostly operated in horizontal dive, hence the value at service speed (Fr 0.6) was discussed. Total drag and total pressure 11.863 N and 94.704 Pa. It is meant that motor installed at ROV must be has a thrust of 4 kg, to run ROV at operational speed.  Figure 5(a) shows of value total drag in Newton unit toward rate of Froude number (Fr). From this graph total drag of three movement dive was increased at every step Froude number. The vertical and diagonal dive has similar coincide trend which sharply increased from Fr 0.3 to Fr 0.6 and then slowly growth until Fr 0.8. Contrast with horizontal dive which considered slightly increased and then steadily increase.
From the trends show the vertical dive has highest value and the horizontal dive has lowest value of total drag at the service speed or Fr 0.6. The total drag of horizontal is less than 12 N and vertical dive has almost 19 N. Both of dives motion has significant difference up to 32%. Diagonal dive has slightly lower than vertical dive. This different caused by ROV motion against fluid flow that commonly called angle of attack.
In other hand Figure 5(b) shown graph of the total pressure value in Pascal (Pa) unit toward rate of Froude number (Fr). The trends of three movement model has similar pattern. Each of motion has significant rose at each step Froude number. This graph displays diagonal dive has highest quantity and the vertical dive has lowest quantity at service speed model. The detail visualization about affected of total pressure around body ROV has shown in Figure 6(a), 6(b), and 6(c). This CFD simulation captured physic phenomena around body due to fluid flow respectively. When mini ROV travelled through fluid flow, the pressure force acting on each element of body surface summed over entire body to produce total pressure. The pressure arose from viscous of fluid flow.
The figure 6(a) illustrated total pressure of body mini-ROV which simulation capture the pressure mostly affected at side of body. The figure 6(b) demonstrated total pressure of body mini-ROV when moved diagonally. It shown that diagonal dive has affected pressure on almost surface area and has biggest total pressure value among the movement. When body moved diagonally both side and bottom affected pressure that push perpendicular integrally. The last figure 6(c) described vertical dive. Same as horizontal dive, it has one side affected major pressure area, which is bottom region of mini ROV. This pressure acted normally to the body.

A. Material Production
According to CFD analysis result regarding pressure value, the material of ROV main body was selected. Polyvinyl Chloride (PVC) with category rigid PVC known as Unplasticied PVC (PVC-U) was chosen to accommodate the requirement value of pressure. UPVC materials are used in the broad transportation of drinking water, water sewage and soil, underground drainage and industrial applications and eco-friendly which meant allow to disposable. UPVC has high hardness mechanical properties of tensile strength up to 51 MPa, thus body ROV allowed to free dive below the sea conducted with CFD result.

B. Controller Hardware
In Addition, Mini ROV was propelled by thruster with BLDC type of brushless motor fitted 4 kg of thrust power to overcome due to impact of the resistance where acted around mini ROV body up to 3.562 kg at Fr 0.6. Thus Mini ROV was allowed to travel under the sea. The velocity and axial rotation of BLDC motor controlled by 12 Volt BLDC motor driver circuit then operated by Arduino UNO R3.
V. CONCLUSION ROV has major role for environmental research and surveying underwater purpose for evidence, especially the micro-ROV with light body geometry can reach through narrow and dangerous area. It provides a costeffective alternative to fully utilize research grants and funding. The investigation of mini ROV using simulation CFD based RANSE software was successfully determined. Hydrodynamic parametric has been accomplished particularly in total force and total pressure. CFD simulation provides comprehend visualization of fluid flow around body thus performance of ROV can be examined.
When performing numerical analysis, it is important to ensure that the numerical errors are as small as possible. Except from the error associated with the fact that the solution is attained at discrete points in the domain, there are three important sources of error. First the governing equations need to be able to describe the fluid flow in a satisfactory way, this lead to the second source of error which is the algebraic equations and lastly the physical model used in the simulation (e.g. turbulence models).
The total force is used to determined amount of motor thrust with around weigh 4 kg. Then total pressure is employed the load pressure can be hold up about 8 bar when ROV operated and diving depth into underwater. The mini ROV succeed fulfill of design and performance requirement. Further research for mini ROV still developed to adjust in many research fields.