Ansys Fluent Setting For Free Surface nx5g

Numerical and Experimental Roll Response and Decay of a Surfaced Submarine

5.

CONCLUSIONS AND RECOMMENDATIONS

5.1.

SUMMARY AND CONCLUSION

5.1.1.

SUMMARY

The main object of this research project was to develop a method of application of a six degree of freedom (6 DOF) model in the CFD package ANSYS CFX Limited research using this package for a 6 DOF has been found; hence this is a relatively new system of application for ANSYS. This research is intended to be able to compliment model test experiments and give an early indication on what type of motion could be expected before model tests are applied.

•

A literature review investigated prevIOus work undertaken, specifically in regards to previous research projects completed at the AMC and CFD research into simulating a floating body's motion with 6 DOF.

•

Using the ANSYS package, a scaled CFD model was created; the parameters were derived from the physical scaled experimental model. This scaled CFD model was to simulate the roll motion of a submarine. a

A 2D model was used to investigate the capabilities of the package and familiarise the author with of the package.

a

A 3D model was used to investigate the capability of replicating the model experiments with a CFD simulation. Roll decay at a starting angle of IS' was simulated.

•

Experimental tests of a scaled experimental model were conducted in the AMC towing tank. a

Roll decay experiments with three tail configurations (x tail, + tail and bare tail) with initial starting angles (10, 15,25,30,35,40 and 50 degrees) were investigated.

a

The models response in a regular beam sea was also investigated where the casings slots were fully open and fully closed.

•

The similarities, differences and limitations between the CFD simulations and experimental results were investigated.

•

The differences between two CFD verSIOns, ANSYS CFX-Pre vll and v12 were investigated and compared.

55

Numerical and Experimental Roll Response and Decay of a Surfaced Submarine

5.1.2.

CONCLUSION

After investigating experimental roll decay and beam sea responses and creating a method to analyse roll motion in a CFD package, the following conclusions can be made:

•

The experimental roll decay results showed the following: o

The experimental study into the roll decay indicates that the x-tail configuration was generally the most effective at damping roll at various starting angles.

o

The bare tail was very ineffective at resisting roll; this was expected as it has no appendages.

•

The following was found from the physical scaled model experiments in regular beam sea results: o

The natural frequency of the scaled submarine models with casing slots fully closed was siguificantly more than when the scaled submarine models casing slots were fully open.

•

The 3D CFD model simulations indicate the following: o

0

Even though the results for the 15 roll decay case could not be completely validated, the results show a close correlation with the experimental results.

o

It is possible to get the correct roll response if the initial conditions are applied

correctly. o

The position of the VCG and VCB are critical to acquire a realistic roll response.

o

The heave motion of the model will affect the roll motion. An initial run ranging from 5 to 10 seconds could be implemented before the roll decay simulation to reduce the affects of heave motion. (Note: the beam sea response simulations doesn't need this time as it takes 5 to 10 seconds for the waves to induce a motion.)

o

Surface roughness was present it applied a larger shearing force that was important to model the damping characteristics of the submarine.

•

In regard to the differences in the two versions of the CFD package(ANSYS CFX-Pre vll

and vI2): o

It was found that the application of cylindrical motion within the ANSYS vll

package had severe limitations as this motion could only be achieved by moving a point in two linear directions around a node. Further work could develop a better method for creating cylindrical motion, but the new package ANSYS v12 makes this unfeasible. o

The ANSYS v12 package has a rigid body module built into it which allowed an easy application of rotation and 6 DOF. This allowed the model, with the correct inputs to react to the conditions applied to simulate the model response.

56

Numerical and Experimental Roll Response and Decay of a Surfaced Submarine This study should help future projects to develop similar projects using CFD packages to investigate vessels motions. It proves that CFD based programs can be used to predict 6 DOF motion. It also laid the groundwork for future projects looking at wave induced motions and roll decay into a CFD package.

5.2.

RECOMMENDATIONS FOR FUTURE WORK

This is an ongoing project by the DSTO/AMC and in regards to the CFD work the options for future work are many. The following recommendations have been made to help future projects to further investigate certain areas of interest and how to improve the current results:

•

Other similar 6 DOF CFD projects indicate that the tet mesh is not the best meshing method to use; therefore it is recommended to use a hex meshing via a program such as ICEM. This will help reduce the mesh element number without compromising accuracy and it will decrease the time it takes to solve.

•

A further investigation into optimising the mesh so it will not crash the solver easily and the ability to apply finer mesh on areas of interest, such as the waterline without effecting mesh deformation.

•

Accurately determine the model waterline, KG, trim angle and Ixx so these error factors are minimised in the CFD simulations.

•

Determine the LCG, Iyy and I" so that the pitch, yaw and surge can be accurately applied in the CFD simulations.

•

Further development of the two domain model so more complex geometry can be applied. The current two domain model has many free surface errors. This model tended to be more robust than the single domain model and warrants further investigation.

•

The response of a submarine in a range of sea states and corresponding wave frequencies within the CFD package.

•

The development of a full submarine model with bilge keels, tail fins and main fin for CFD predictions. The internal casing can be an extra factor that could be investigated. This type of investigation can use the results obtain from Lund [2005], El-Atm [2006] and Davies [2007] physical model scaled experiments. Resizing the fins to see its effectiveness.

57

Numerical and Experimental Roll Response and Decay of a Surfaced Submarine •

An investigation of the internal tanks within the actual model and see what effect it has on the

roll in a sea state in both physical experiments and using a CFD package.

•

An investigation into the fluid flow and pressures around the submarine when it is subject to a

sea state or in a roll decay simulation within CFD.

•

A simulation into a rising or sinking submarine. This may include emergency surfacing and the effects of waves on motions in both physical experiments and using a CFD package.

In respect to the physical model scaled experiment investigation:

•

During the scaled model experiments the weights within the model move at high roll angles. An adequate system for securing the weight tray to the submarine model should be created.

•

The models are stationary and are not subject to forward speed. The motion of the submarine in a sea state with forward speed should be investigated.

•

A new generic model should be created out of a new material such as aluminium or carbon fibre so that the has more control over the model distribution of weight, and radius of gyration.

58