Proceedings of PACAM XIV

14th Pan-American Congress of Applied Mechanics

March 24-28, 2014, Santiago, Chile

MODELING AND SIMULATION OF THE BOLLARD PULL TEST FOR

VESSELS USING COMPUTATIONAL FLUID DYNAMICS
Josu Hernndez Pedroza, josue61290@gmail.com
Carlos Plazaola, carlos.plazaola@utp.ac.pa
Adn Vega S.,adan.vega@utp.ac.pa
Laboratorio Especializado en Procesos de Unin y Manufactura
Facultad de Ingeniera Mecnica, Universidad Tecnolgica de Panam

Abstract. This work is focused on the development of a methodology for modeling and simulation of thrust analysis on
vessels, particularly what is known as the Bollard Pull Test, using computational fluid dynamics. The Navier-Stokes
equations are solved by using ANSYS FLUENT which is based on a Finite Volume formulation along with the Reynolds
Averaged Navier-Stokes (RANS) and the SST k-w turbulence model. Two geometrical domains were created, a fixed
one and a rotational one. A study is carried out for the standard test conditions using steady and transient methods,
also a parametric study is done under non standard conditions for situations where the test environment can not
comply with the ideal conditions for the test. Thus, the effect of dimensions of the test basin, depth and marine
currents were taken into account. A validation of the results by comparing with existing data was made. Analysis for
several propeller-nozzle diameters were obtained and compared to the theoretical results with very good agreement.
Keywords: bollard pull, advance coefficient, thrust, torque
1. INTRODUCTION
Computational Fluid Dynamics (CFD) can be used with full advantage for propeller design and related problems in
naval engineering. Solution Methods like Reynolds Averaged Navier-Stokes (RANS) along with turbulence models are
successfully applied for solving a variety of problems. Recently Galeano et al. (2012) have used RANS for solving
vessels propulsion problems. Funeno (2009) used CFD for evaluating the performance of azimutal propellers.
According to Lam (2012) the SST k-w model is one of the best for propeller analysis. Maksoud y Heinke (2002)
suggested the use of very large domains for low advance speeds. Mertes y Heinke (2008) indicated that scale models
are not always convenient.
The Bollard Pull Test is used as a base to determine the maximum thrust capacity for a vessel. The test requires field
instrumentation and the ship presence to verify that it is capable of producing the design thrust, if the requirements are
not satisfied it will be necessary to take the project to the drawing board with all its consequences.
In this work a methodology for the analysis and simulation of Bollard Pull Test is developed and with this a valuable
tool for design of the Propeller-Nozzle configurations will be available.
2. CAD MODEL
A propeller model has been selected to carry out the analysis and simulations. It is a propeller Series Wageningen
Ka 4-70 along with a Nozzle Type 19A. For several years MARIN (Maritime Research Institute Netherlands) have been
testing this Propeler-Nozzle configuration. The tip clearance is 0.4 % of the propeller diameter. This was the tip
clearance used by Oosterveld (1970) who obtained nondimensional coefficients graphs for different coefficients. The
propeller design is based on the dimensions of the Ka series according to Carlton (2007). These are given in Table 1.
Table 1. Propeller specifications
Specifications
Diameter (m)
Pitch ratio
Blade area ratio
Blade number

Value
2
1
0.7
4

A geometric model for the hull of the vessel was developed, however, for the simulations, the geometric model just
up to the waterline was necessary, this can easily modified for analysis to different depth.

Proceedings of PACAM XIV

14th Pan-American Congress of Applied Mechanics

March 24-28, 2014, Santiago, Chile

Table 2. Ship specifications

Specifications
Displacement (ton)
Volume of displacement (m3)
Draught (m)
Length (m)
Beam (m)

Value
686.231
669.494
4.016
28.6
10.034

3. COMPUTATIONAL MODEL
The analytical solution of the Navier-Stokes equations is practically impossible. Therefore a numerical approach to
the solution is taken. In this case it is based on the Reynolds Average Navier-Stokes (RANS) in which the flow
variables are divided into two components, a mean component and a fluctuating component. The computational domain
has been designed with the objective of describing the ideal environment for the Bollard Pull test. The domain has a
parallelepiped domain with dimensions of 400x200x44 meters. The forward velocity of the vessel is zero for a Bollard
Pull Test. Therefore the forward coefficient is equal to zero (J=0).

where Va is the forward velocity, n are revolutions per second and D the diameter
The thrust coefficient (KT) and the torque coefficient (KQ) are given by

where T is the propeller-nozzle thrust force and is the fluid density.

Q represents the propeller torque.

3.1. Mesh
Two domains were created using ANSYS Design Modeler. A stationary domain which includes the fluid, hull,
nozzle and the gear box, and a rotational domain that surrounds the propeller. ANSYS Meshing was used to generate
the mesh with the corresponding specifications.
A non-structured tetraehedric mesh in Patch Independent modality was used, in this modality the mesh is first
created on the faces of the domain and on the edges thereafter. In this way skewness asymmetry is reduced and at the
same time the mesh quality is improved. For the propeller zone asymmetry, 0.75 or less is recommended. In this work
asymmetry of 0.6948 was obtained. In Figure 1 an example of the tetrahedric meshing used is shown.

Figure 1. Vessel mesh

Proceedings of PACAM XIV

14th Pan-American Congress of Applied Mechanics

March 24-28, 2014, Santiago, Chile

3.2. Boundary conditions

A set of boundary conditions is required for the developed domain, these are specified with the objective of
representing the field conditions as close as possible to the real situation. The boundary conditions are key to the
solution of the equations, to satisfy the continuity equation and the convergence of the simulation. Thus, the boundary
conditions are given in Table 3.
Table 3. Boundary conditions
Zone
Hull/Propeller/Nozzle/Gear box
Water surface
Bottom
Outer boundaries

Boundary conditon
No Slip
Symmetry
No Slip
Hydrostatic pressure

3.3. Steady state simulations

The Multiple Reference Frames scheme is used for the steady state simulations, in such scheme it is posible to
establish translational and rotational speeds. In this study a speed has been established for the rotational domain. For a
reference frame with constant rotational speed it is possible to make a transformation of the equations of motion to
obtain a solution in a stationary reference frame.
At the domains interface, a local reference frames transformation is made so that the flow variables can be used in
adjacent zones. The MRF approach does not takes into account the relative motion of one zone with respect to adjacent
zones, the mesh remains fixed. Through the calculations the moving part is frozen in a given position and the propeller
flow field is obtained, Zhang et al. (2008). That is why the MRF scheme is also known as Frozen Rotor Approach. The
convergence criteria was established in a residual size of the order of 1x10-4. A convergence test was carried out using
1x10-5, the results were similar to those obtained using 1x10-4.
Table 4. Thrust force stationary state
Force
Normal (N)
Viscous (N)
Total (N)

Hull
-10109.94
-54.19
-10164.13

Propeller
186496.31
-1085.63
185410.68

Nozzle
188382.44
-788.42
187594.02

Gear box case

-22849.08
-90.69
-22939.78

Table 5. Propeller torque stationary state

Torque
Normal (N.m)
Viscous (N.m)
Total (N.m)

Propeller
61893.85
1775.58
63669.43

Figure 2. Pressure contour of suction side (left) and pressure side (right)

Full vessel
341919.73
-2018.93
339900.8

Proceedings of PACAM XIV

14th Pan-American Congress of Applied Mechanics

March 24-28, 2014, Santiago, Chile

Figure 3. Shear stress contour of suction side (left) and pressure side (right)
3.4. Transient simulations
For the simulation in unsteady state the sliding mesh scheme was used, in this case the unstable interactions of
relative motion are included, in general the sliding mesh scheme is considered with higher precision but it requires a
higher computational demands.
In the simulations the steady state solution was used as initial condition, the time step interval was taken as 0.07
seconds and 20 iterations for each interval (time step). A 30 seconds simulation time was taken, since that is the interval
used in field tests. The total simulation time was of about 85 hours.
.

Figure 4. Vessel Thrust in transient state

Figure 5. Propeller torque in transient state

Proceedings of PACAM XIV

14th Pan-American Congress of Applied Mechanics

March 24-28, 2014, Santiago, Chile

4. PARAMETRIC STUDY
A parametric study was carried out with the objective of finding the impact of conditions that differ of the ideals on
the thrust force. The following effects were considered:
1.
2.
3.
4.

Effect of the free distance on both sides of the vessel.

Effect of depth.
Effect of the fixed point-vessel distance.
Effect of marine currents.

The study reveals that at least 100 meters on both sides are required to get results close to the ideal conditions.
For 20 meters or more the depth has no effect on the thrust force. In the case of the distance between the fixed point
and stern, at least 100 meters are required. It was found that when the magnitude of the velocity of marine currents is of
the order of 0.2 m/s or higher the effect on the thrust force can be noticed.
4.1. Comparison of Results
As a way of validation of the method, simulations were carried out without including the hull and gear box
interactions. Thus, the experimental results of the Wageningen series were compared with the simulation results
showing very good agreement. The results are given in Table 6 for a 2 m diameter.
Table 6. Results for a 2 m diameter
Results
Simulation thrust (N)
Theory thrust (N)
Thrust error (%)
Simulation torque (N.m)
Theory torque (N.m)
Torque error (%)

20 rad/s
87379.28
87237.54
0.16
15592.34
14622.67
6.63

30 rad/s
196868
196284.46
0.3
34928
32901
6.16

40 rad/s
350447.77
348950.15
0.43
62090.76
58490.69
6.15

5. CONCLUSIONS
A methodology for analysis and simulation of Bollard Pull Test taking into account the interaction propeller-nozzlehull has been successfully established. The methodology was validated comparing the results with previous data. The
thrust force was determined for simulated ideal test conditions. Unsteady and steady state simulations were carried out
and a comparison of computational time and precision of the results leads to conclude that the steady state solution
shows good precision and shorter computation time.
It was found that the thrust force is affected when using reduced volumes for the test. I was also found that thrust is
reduced when there are marine currents with velocity magnitudes above 0.2 m/s.
6. ACKNOWLEDGEMENTS
My sincere thanks to Mr. Carlos Plazaola for giving me the opportunity to conduct this research, for your
thoughtful advice, support and time given. Dr. Adn Vega for the unconditional help and advice. The whole team at
LEPUM laboratory.
7. REFERENCES
Carlton, J. S., 2007. Marine Propellers and Propulsion, Second Edition. Global Head of Marine Technology and
Investigation, Lloyds Register. Butterworth-Heinemann. USA. pp. 117-118.
Funeno, I., 2009. Hydrodynamic Optimal Design of Ducted Azimuth Thrusters. First International Symposium on
Marine Propulsors smp09, Trondheim, Norway, June.
Galeano et al., 2012. Experimentos Numricos CFD en Propulsin Naval. VII Symmtechnaval 2012. IPIN Cuba.
Lam, W. H., 2012. An effective method for comparing the turbulence intensity from LDA measurements and CFD
predictions within a ship propeller jet.
Maksoud, M., Heinke, H. J., 2002. Scale Effects on Ducted Propellers. 24th Symposium on Naval Hydrodynamics.
Fukuoka, Japan.
Mertes, P., Heinke, H. J., 2008. Aspects of the Design Procedure for Propellers Providing Maximum Bollard Pull.
Suntec Convention Centre, Singapore Organised by the ABR Company Ltd.

Proceedings of PACAM XIV

14th Pan-American Congress of Applied Mechanics

March 24-28, 2014, Santiago, Chile

Oosterveld, M. W. C., 1970. Wake Adapted Ducted Propellers. NSMB Wageningen Publication No. 345. June.
Zhang et al., 2008. Progress in Analysis of Viscous Flow around Podded Propulsor. China Ship Scientific Research
8. RESPONSIBILITY NOTICE
The authors are the only responsible for the printed material included in this paper.