Measurements and Computations for Blade Spindle Torque · PDF fileMeasurements and Computations for Blade Spindle Torque of Controllable Pitch Propellers in Open Water ... propeller

Third International Symposium on Marine Propulsors smp’13, Launceston, Tasmania, Australia, May 2013

Measurements and Computations for Blade Spindle Torque of Controllable Pitch Propellers in Open Water

Isao Funeno1, Christiaan Pouw2, René Bosman2

1Kawasaki Heavy Industries, Kobe, Japan

2Maritime Research Institute Netherlands (MARIN), Wageningen, The Netherlands

ABSTRACT

For Controllable Pitch Propellers (CPP), it is essential to

predict accurately its performances such as propeller

thrust, torque and furthermore propeller blade spindle

torque in all practical operating conditions. As a first step,

a blade spindle torque sensor built in the propeller hub

was developed in order to obtain fundamental data. The

sensor was verified thoroughly to enable us to extensively

measure the blade spindle torque in open water conditions.

The sensor is devised such that it is only sensitive for

measuring the blade spindle torque and not for

interference of the thrust and torque from the propeller

blade acting on the sensor. While, by versatile CFD

simulation techniques, we analyzed the flow around the

CPPs blade in open water condition at off-design pitch

settings and for high-propeller loading conditions, which

is normally difficult for potential theory-based methods to

analyze correctly. The blade spindle torque measurements

and the flow simulation technique by means of CFD were

extensively applied to two CPP blades with different

skew distributions. Comparing the results of the

measurements and the computations, the agreement is

very good but there are still some issues to be addressed

further.

Keywords

CPP, spindle torque sensor, CFD, blade spindle torque,

propeller, off-design

1 INTRODUCTION

Recently, CPPs (Controllable Pitch Propellers) have

increasingly been installed in many vessels such as

ferries, RO/ROs, ROPAXs, shuttle tankers and others. It

is important to create databases and tools in order to

provide hydrodynamically-designed blades with high-

efficiency. Especially for the CPPs, it is essential to

predict accurately its performances such as propeller

thrust, torque and furthermore propeller blade spindle

torque in all practical operating conditions at the initial

design phase. In particular, it is necessary to establish

advanced methods to measure and compute all the above-

mentioned quantities of propeller performance in design

and off-design conditions. However, in the past, there

have been only a few publications about marine propeller

blade spindle torque in various operating conditions

(Boswell et al 1975, Pronk 1980, Ito et al 1984, Jessup et

al 2009, Koushan et al 2011 and Dang et al 2012). The

reason for so few publications is mainly due to difficulty

of the measurements and computations of blade spindle

torque.

As a first step, for the purpose of obtaining fundamental

data, after discussing fully specifications of measuring

system between Kawasaki Heavy Industries (KHI) and

Maritime Research Institute Netherlands (MARIN), a

blade spindle torque sensor built in a CP propeller hub

was newly developed at MARIN in order to measure

propeller thrust, propeller torque and blade spindle torque

on one key-blade simultaneously. The measuring sensor

was manufactured using a state of the art micro-fabricated

device and was verified thoroughly to enable us to

extensively measure the blade spindle torque in open

water conditions. The sensor is devised such that it is only

sensitive for measuring the blade spindle torque and not

for interference of the thrust and torque from the propeller

blade acting on the sensor. Also an advanced signal

processing system was taken into account. The

measurements of blade spindle torque are described in

detail in the next chapter.

On the other hand, along with the remarkable progress of

computer hardware systems, CFD (Computational Fluid

Dynamics) has been applied increasingly to propeller

blade designs. Especially the RANS-based CFD codes are

utilized to compute the complex flow around various

types of propulsors and their results are very helpful for

designers to comprehend the flow in detail (Funeno

2009). It is a logical step to also apply CFD simulation

techniques to analyze the flow around the CPP blades in

open water condition at off-design pitch settings and for

high-propeller loading conditions. These calculations are

normally difficult for potential theory-based methods to

analyze correctly. In order to assess the effectiveness of

the computation of the blade spindle torque, the

measurements and the flow computation by means of

CFD were extensively applied to two kinds of CPP blades

with different skew distributions among others, with

which fast vessels such as RO/ROs are equipped.

Comparing the results of the measurements and the

381

computations, the agreement is very good, but there are

still some issues to be addressed further.

Moreover, it is very

designers that the measurements and the computation by

CFD are complimentary to each other.

combining measurements and CFD can provide

comprehension of complex flow phenomena, which is no

longer limited to the

only. In particular,

can be used not only

also to complement the results

to compute correctly

complement the

which are not

essential for the CP propeller blade design

the pros and cons of

The computational method of blade spindle torque is

described in

chapters the results are

drawn.

2 MEASUREMENTS

2.1 Sensor and test set

To be able to conduct open water tests and measure

spindle torque

be placed inside the hub and connect

The requirements on the

from the discussions with KHI are

practical ones.

First of all, the sensor should be able to measure spindle

torques as accurate as possible. T

is preferred to measure forces which are not too small, a

larger diameter of

drag forces acting on the blade introduce bending

moments in the sensor, to avoid

moments with the measurement of the spindle torque, the

sensor should be designed such that it is resistive to the

bending moments.

From a practical point of view, the sensor should be able

to fit inside the hub of the propeller blade. Also, the

whole test set

measure the spindle torque

Based on the

the range β

1500 rpm, where

propeller blade section at 0.7

practical propeller diameter of 240 mm on model scale

the expected spindle torque and bending moments c

estimated. The resulting requirements are

spindle torque up to 4 Nm and to withstand bending

moments up to 15 Nm due to the lift and drag forces

acting on the blade.

2.2 Test set

The open water test set

to incorporate measuring an additional sensor and is

shown in Figure


issues to be addressed further.

, it is very important



measurements and CFD can provide


longer limited to the estimation of propeller performance

In particular, the measured

can be used not only to validat

complement the results

compute correctly by CFD

complement the accurate results

which are not always measured

for the CP propeller blade design

the pros and cons of both the measurements

tational method of blade spindle torque is

described in detail in the third chapter, in the

the results are

MEASUREMENTS

Sensor and test set


spindle torques, a sensor had to be developed which could

be placed inside the hub and connect

The requirements on the

the discussions with KHI are

practical ones.


as accurate as possible. T


larger diameter of the propeller is preferred. The lift and


moments in the sensor, to avoid



bending moments.



whole test set-up should be able to operate in air to

measure the spindle torque

Based on the foreseen test program (open water tests for

β=0 to 90 degree

, where β :effective advance angle of a

propeller blade section at 0.7

propeller diameter of 240 mm on model scale

xpected spindle torque and bending moments c




acting on the blade.

Test set-up sensor

The open water test set-up in use at MARIN was adapted


in Figure 1.



important for propeller blade



measurements and CFD can provide


estimation of propeller performance

measured blade spindle

validate the computed ones, but

complement the results which are

CFD. On the contr

accurate results from measurements

asured correctly

for the CP propeller blade design

the measurements


the third chapter, in the

the results are discussed and conclusion

MEASUREMENTS

Sensor and test set-up requirement


, a sensor had to be developed which could

be placed inside the hub and connected

The requirements on the sensor design

the discussions with KHI are


as accurate as possible. To accomplish this,


the propeller is preferred. The lift and


moments in the sensor, to avoid crosstalk





up should be able to operate in air to

measure the spindle torques due to centrifugal forces.

foreseen test program (open water tests for

degree with rotational speeds up to

:effective advance angle of a

propeller blade section at 0.7R) and


xpected spindle torque and bending moments c




up sensor

up in use at MARIN was adapted




for propeller blade


CFD are complimentary to each other. Generally

measurements and CFD can provide a



blade spindle torques

the computed ones, but

are not always easy

On the contrary, CFD can

from measurements

correctly. Therefore, it is

for the CP propeller blade design to recognize

the measurements and the CFD


the third chapter, in the successive

and conclusion

up requirements

To be able to conduct open water tests and measure blade


ed to a key blade.

design which followed

the discussions with KHI are mechanical and


o accomplish this,




crosstalk of these






due to centrifugal forces.


with rotational speeds up to


) and the maximum


xpected spindle torque and bending moments can

estimated. The resulting requirements are: to measure






for propeller blade


Generally,

a good



torques

the computed ones, but

not always easy

CFD can

from measurements

. Therefore, it is

recognize

and the CFD.


successive

and conclusions are

To be able to conduct open water tests and measure blade


key blade.

which followed

mechanical and


o accomplish this, as it




of these


sensor should be designed such that it is resistive to these




due to centrifugal forces.


with rotational speeds up to


maximum

propeller diameter of 240 mm on model scale,

an be

to measure





This set

sensor at the propeller side and slip rings at the strut side

to allo

acquisition system

belt by an electrical engine which is placed on top of the

strut on the carriage. The

was modified with a flange to mount the

with the spindle torque sensor. A water

connects the sensor in the hub with the cable inside the

hollow shaft and allows for easy connecting and

disconnecting of the propeller

propeller can be don

more detailed overview of the propeller and the

sensors

thrust/torque sensor and the spindle torque sensor

The spindle torque sensor itself is designed such that it

can withst

to reliably

in Figure

connect to the bottom and top face. The bottom face

connects to the key blade

Figure

torque sensor

Figure

sensors.

Figure

MARIN

This set-up consists of a hollow shaft with a thrust/torque


to allow the sensors to be connected to the data

acquisition system


strut on the carriage. The





disconnecting of the propeller

propeller can be don


sensors and Figure 3 shows



withstand large bending moments and still be capable

to reliably measure

igure 3 on the right and has


connects to the key blade

Figure 3 Thrust / torque sensor

torque sensor (right).

Figure 2 Detailed overview of

sensors.

Figure 1 Sketch of open water test setup in use at

MARIN, flow is coming from right to left.

up consists of a hollow shaft with a thrust/torque


w the sensors to be connected to the data

acquisition system. The shaft is driven through a toothed


strut on the carriage. The standard





disconnecting of the propellers. Changing the pitch of the

propeller can be done within an hour.


and Figure 3 shows



large bending moments and still be capable

measure spindle torque. The sensor can be seen

on the right and has


connects to the key blade with four bolt

hrust / torque sensor

(right).

Detailed overview of

ketch of open water test setup in use at

, flow is coming from right to left.




. The shaft is driven through a toothed


standard thrust/torque sensor

was modified with a flange to mount the propeller hub

with the spindle torque sensor. A water-tight connector



. Changing the pitch of the

within an hour. Figure


and Figure 3 shows a detail image




spindle torque. The sensor can be seen

on the right and has four small flanges which


with four bolt holes

hrust / torque sensor (left) and spindle

Detailed overview of propeller and three


, flow is coming from right to left.






thrust/torque sensor

propeller hub

tight connector




igure 2 shows a

more detailed overview of the propeller and the three

a detail image of the

thrust/torque sensor and the spindle torque sensor.




small flanges which


holes, the top

and spindle

propeller and three







thrust/torque sensor

propeller hub

tight connector




shows a

three

the




small flanges which


, the top

382

face clamps in the hub and allows for setting the correct

pitch. Two strain gauges are placed on the inner side in

between the four flanges of the sensor. To other gauges

are placed in between the opposite flanges, this will

cancel bending moments but amplify spindle torque. The

sensor is placed in a free flooded chamber in the hub with

a water tight connection for the cable. No parasite friction

from bearings or sealing is present.

After the placement of the strain gauges and soldering

lacquer threads for the blade transducer, the transducer is

coated with a special water tight coating which stays

flexible to avoid hysteresis and creep. When this process

was finished, the transducer was tested in water for three

days to check if the transducer was still watertight. After

three days, the insulation value should be more than 500

MΩ otherwise the measurements could be influenced.

2.3 Calibration / quality

The sensor was calibrated in a special set-up by applying

different loads in different directions. The (ideal) behavior

of the sensor in measuring the applied load is expected to

be linear. The linear behavior to the applied loads is

shown in Table 1. The relative error increases for small

loads and stays below 0.2544 %.

Table 1 Linear behavior to applied load with spindle torque

sensor.

Applied load

(Nm)

Measured load

(Nm)

Linearity error

(Nm * 10-3)

Relative error

(%)

-3.2627

-2.6101

-1.9576

-1.3051

-0.6525

0.6525

1.3051

1.9576

2.6101

3.2627

-3.2628

-2.6085

-1.9551

-1.3030

-0.6509

0.6521

1.3043

1.9577

2.6115

3.2657

-0.1128

1.5927

2.4567

2.0585

1.6604

-0.3982

-0.7963

0.0677

1.3525

3.0579

0.0035

0.0610

0.1255

0.1577

0.2544

0.0610

0.0610

0.0035

0.0518

0.0937

Additional checks have been performed to determine

crosstalk when bending moments due to the side forces

(thrust and torque) applied on 0.7R and centrifugal force

are acting on the sensor. Table 2 shows the crosstalk of

the bending moment (left) and the closstalk due to the

centrifugal force (right) with the spindle torque which

should be zero in this case.

Based on the above results, a maximum error in

measuring spindle torque of 0.005Nm is expected. The

maximum load capability of the transducer is 6Nm.

Additional load cases with a combination of bending

moments, side forces and centrifugal force have been

performed as well, and it was found that the crosstalk

errors were in the same order of magnitude as presented

in the tables above.

To determine the spindle torque due to the hydrodynamic

forces acting on the blade, the spindle torques due to

centrifugal forces have to be subtracted from the total

measured spindle torque. This spindle torque due to the

centrifugal force can be calculated when the weight

distribution of the blade including the sensor is known. It

is also measured for different rotational speeds in the

towing tank as shown in Figure 4.

To reduce the influence of the centrifugal force on the

spindle torque in the first place, the propeller blades are

made of aluminum. After manufacturing of the blades, its

geometry and pitch setting have been checked by means

of an optical scan as described by Dang et al (2012).

Figure 5 shows the comparison between measured and

calculated centrifugal spindle torque coefficients (based

on the properties of CAD model), KSC in air, in which a

good correlation between the two is found.

KSC is defined as follows,

(1)

Where, QSC: centrifugal spindle torque, ρ:fluid density,

n: propeller revolution, D: propeller diameter.

Applied

bending

moment (Nm)

Crosstalk

error

(Nm * 10-3)

Applied

centrifugal

force (N)

Crosstalk

error

(Nm * 10-3)

1.1628

2.3256

3.4883

4.6511

5.8139

-0.9853

-2.1524

-3.9842

-6.0854

-7.0401

29.4375 3.4106

Table 2 Crosstalk of bending moment in combination

with side force (left) and crosstalk due to centrifugal

force (right) with spindle torque sensor.

Figure 4 Model test set-up to measure spindle torque in air

(the hose is to lubricate the seals in the open water test set-

up).

Figure 5 Spindle torque coefficients in air.

-1.4E-03-1.4E-03-1.4E-03-1.4E-03

-1.2E-03-1.2E-03-1.2E-03-1.2E-03

-1.0E-03-1.0E-03-1.0E-03-1.0E-03

-8.0E-04-8.0E-04-8.0E-04-8.0E-04

-6.0E-04-6.0E-04-6.0E-04-6.0E-04

-4.0E-04-4.0E-04-4.0E-04-4.0E-04

-2.0E-04-2.0E-04-2.0E-04-2.0E-04

0.0E+000.0E+000.0E+000.0E+00

2.0E-042.0E-042.0E-042.0E-04

4.0E-044.0E-044.0E-044.0E-04

-50-50-50-50 -45-45-45-45 -40-40-40-40 -35-35-35-35 -30-30-30-30 -25-25-25-25 -20-20-20-20 -15-15-15-15 -10-10-10-10 -5-5-5-5 0000 5555 10101010

KK KKSC

SC

SCSC[[ [[ -- --]] ]]

Alpha (deg)Alpha (deg)Alpha (deg)Alpha (deg)

Ksc (measured)Ksc (measured)Ksc (measured)Ksc (measured)

Ksc (calculated)Ksc (calculated)Ksc (calculated)Ksc (calculated)

383

3 COMPUTATIONAL METHOD

3.1 General

The governing equations to solve are the RANSE

(Reynolds Averaged Navier-Stokes Equations) and the

mass continuity equation for incompressible viscous fluid.

The whole computational domain around the propeller

was divided into a vast number of minute cells. The

equations were discretized based on the finite volume

method to fulfill conservation of momentum and mass for

all the cells (Ferziger and Peric 2002). Next, the equations

were solved numerically by using the SIMPLE method.

The effect of turbulence was introduced by the

mathematical turbulence model of the k-omega/SST two

equations with wall functions for the high Reynolds

number type.

The effect of propeller rotation on the flow field was

considered by the multi-reference frame method, which

introduces centrifugal forces and Coriolis forces in a

relative coordinate system fixed at a rotating propeller as

body forces in solved RANSE. The computational

domains were divided into a static zone and a rotational

zone, in which the rotational body forces were introduced.

Only the domain around the blade, boss and boss cap

were incorporated in the rotational zone. Almost all the

computations were carried out in steady mode.

Additionally, the computations with low advance speed

and nearly neutral pitch setting condition was carried out

in unsteady mode with the second-order implicit method

in order to examine the possibility of more accurate

results. Also the effect of free surface was neglected and

cavitation was not considered.

The versatile commercial CFD software, STAR-CCM+

was adopted, which provides all the functions mentioned

above.

3.2 Grid generations and boundary conditions

The computations were conducted for various blade pitch

settings ranging from the design pitch to reverse pitch

with the same propeller rotational direction. Accordingly,

all the grid generations were accomplished through 3D-

CAD data for each blade pitch setting, which were

modeled for a propeller open water test unit including a

blade, a boss, a boss cap and a tube covering a drive shaft.

The unstructured grid method was applied for the grid

generation due to the complicated blade geometry at low

and reverse pitch setting conditions. All the

computational domains were constructed with only one

blade model, a so-called sector model due to the

computation in propeller open water conditions. The

center of propeller was located at the middle of the

computational domains, since besides the forward flow,

reverse flows were expected at low and reverse pitch

settings. As a result, both inlet and outlet boundaries were

very far from the center of propeller. However, in zero

advance speed conditions (bollard pull conditions), the

inlet boundaries were set at stagnation pressure condition

and the outlet boundaries were set at static pressure

condition. Also the outer surfaces in the radial direction

were set at slip wall condition and both the side surfaces

in a propeller rotational direction of the sector models

were set to cyclic condition. The locations of the outer

boundaries were about 8.2*D at the inlet and outlet planes

and about 4.9*D at the outer plane in radial direction from

the center of the propellers, where D is diameter of

propeller. The total number of cells was about 2 million

for the two CPPs as described in the following section.

Figure 6 shows one example of the surface mesh of the

CPP used in the computations.

3.3 Computational conditions

As mentioned before, the computations of the flow

around the CPPs were conducted in open water conditions

for the purpose of obtaining the fundamental data such as

the propeller thrust, propeller torque and blade spindle

torque. Also all the computations were conducted in

model scale in order to compare directly with the

measured data and assess the computational accuracy.

Consequently the operational conditions were as follows,

(1)Constant pitch angle fixed at the design pitch, α=0

degree from J=0.0 to KT=0.0

Name of CPP CPP1 CPP2

Diameter (mm) 240.0 208.5

Expanded area ratio 0.735 0.760

Skew angle (degree) 49.0 46.1

Number of blades 4 4

Boss ratio 0.29 0.32

Figure 7 CPP models at design pitch.

(b) CPP2 (a) CPP1

(a) Design pitch

(αααα=0 degree)

(b) Nearly neutral pitch

(αααα=-30 degree)

Figure 6 Examples of surface mesh (CPP1).

Table 3 Main particulars of CPP (Model scale)

384

(2)Various pitch angles shifted fromα=0 degree to -35

degree in J=0.0, 0.1 and 0.2

Where, J: advance coefficient, KT: propeller thrust

coefficient, α : deviation blade angle from the design

pitch. The applied CPP blades were the following two

contemporary types as shown in Table 3 and Figure 7.

The propeller revolution of both CPPs was set at 900 rpm

consistent with the model tests.

4 RESULTS and DISCUSSIONS

First, the results for the CPP with the design pitch were

verified in open water condition. Figure 8 shows a

comparison between the measured and the computed data

of CPP1 as an example. In the figure, the measured and

computed propeller thrusts, T, propeller torques, Q and

hydrodynamic spindle torques per blade, QSH are reduced

to the following dimensionless coefficient.

,

ρ,

2

Also the propeller advance coefficient, J and the propeller

efficiency, ηo are defined as follows,

, !

2"

3

Where VA: propeller advance speed. Positive QSH denotes

torque to increase blade pitch angle.

For the overall range of J, all the computed KT, KQ and

KSH agree very well with the measured ones. However,

around J=0.0 the agreement seems to become less. It is

considered that the flow around the blades in the normal

ahead operational range is nearly in shock-free entry

condition, but that at around J=0.0 it is mainly in large

angles of attack with possible stall, which normally brings

a prominent separation vortex from the leading edge.

Next, the results of the CPPs with the off-design pitch

settings from ahead to halfway astern were verified for

J=0.0 to 0.2 in open water condition. Figure 9 shows a

comparison between the measured and computed data for

CPP1 with α. As a whole, all the computed KT, KQ and

KSH agree with the measured ones quite well. However,

around the neutral pitch condition, α=-30 degree at J=0.1

and 0.2 the agreements are not so good. Normally CPPs

Figure 8 Propeller open water characteristics

(CPP1, αααα=0 degree).

-0.10-0.10-0.10-0.10

0.000.000.000.00

0.100.100.100.10

0.200.200.200.20

0.300.300.300.30

0.400.400.400.40

0.500.500.500.50

0.600.600.600.60

0.700.700.700.70

0.800.800.800.80

0.900.900.900.90

1.001.001.001.00

1.101.101.101.10

1.201.201.201.20

0.00.00.00.0 0.10.10.10.1 0.20.20.20.2 0.30.30.30.3 0.40.40.40.4 0.50.50.50.5 0.60.60.60.6 0.70.70.70.7 0.80.80.80.8 0.90.90.90.9 1.01.01.01.0 1.11.11.11.1 1.21.21.21.2

Advance coefficient : JAdvance coefficient : JAdvance coefficient : JAdvance coefficient : J

Kt (Exp)Kt (Exp)Kt (Exp)Kt (Exp)

10Kq (Exp)10Kq (Exp)10Kq (Exp)10Kq (Exp)

Eta_o (Exp)Eta_o (Exp)Eta_o (Exp)Eta_o (Exp)

10Ksh(Exp)10Ksh(Exp)10Ksh(Exp)10Ksh(Exp)

Kt (CFD)Kt (CFD)Kt (CFD)Kt (CFD)

10Kq (CFD)10Kq (CFD)10Kq (CFD)10Kq (CFD)

Eta_o (CFD)Eta_o (CFD)Eta_o (CFD)Eta_o (CFD)

10Ksh(CFD)10Ksh(CFD)10Ksh(CFD)10Ksh(CFD)

Kt,

Kt,

Kt,

Kt,10Kq, Eta_o, 10Ksh

10Kq, Eta_o, 10Ksh

10Kq, Eta_o, 10Ksh

10Kq, Eta_o, 10Ksh

(b) J=0.1

-0.20-0.20-0.20-0.20

0.000.000.000.00

0.200.200.200.20

0.400.400.400.40

0.600.600.600.60

0.800.800.800.80

1.001.001.001.00

1.201.201.201.20

-35.0-35.0-35.0-35.0 -30.0-30.0-30.0-30.0 -25.0-25.0-25.0-25.0 -20.0-20.0-20.0-20.0 -15.0-15.0-15.0-15.0 -10.0-10.0-10.0-10.0 -5.0-5.0-5.0-5.0 0.00.00.00.0

Kt, 10Kq, 10Ksh

Kt, 10Kq, 10Ksh

Kt, 10Kq, 10Ksh

Kt, 10Kq, 10Ksh


Kt(Exp)Kt(Exp)Kt(Exp)Kt(Exp)

10Kq(Exp)10Kq(Exp)10Kq(Exp)10Kq(Exp)


Kt(CFD)Kt(CFD)Kt(CFD)Kt(CFD)

10Kq(CFD)10Kq(CFD)10Kq(CFD)10Kq(CFD)


(c) J=0.2

-0.20-0.20-0.20-0.20

0.000.000.000.00

0.200.200.200.20

0.400.400.400.40

0.600.600.600.60

0.800.800.800.80

1.001.001.001.00

1.201.201.201.20

-35.0-35.0-35.0-35.0 -30.0-30.0-30.0-30.0 -25.0-25.0-25.0-25.0 -20.0-20.0-20.0-20.0 -15.0-15.0-15.0-15.0 -10.0-10.0-10.0-10.0 -5.0-5.0-5.0-5.0 0.00.00.00.0

Kt, 10Kq, 10Ksh

Kt, 10Kq, 10Ksh

Kt, 10Kq, 10Ksh

Kt, 10Kq, 10Ksh








Figure 9 Propeller open water characteristics with off-

design pitch settings (CPP1).

(a) J=0.0

-0.20-0.20-0.20-0.20

0.000.000.000.00

0.200.200.200.20

0.400.400.400.40

0.600.600.600.60

0.800.800.800.80

1.001.001.001.00

1.201.201.201.20

-35.0-35.0-35.0-35.0 -30.0-30.0-30.0-30.0 -25.0-25.0-25.0-25.0 -20.0-20.0-20.0-20.0 -15.0-15.0-15.0-15.0 -10.0-10.0-10.0-10.0 -5.0-5.0-5.0-5.0 0.00.00.00.0

Kt, 10Kq, 10Ksh

Kt, 10Kq, 10Ksh

Kt, 10Kq, 10Ksh

Kt, 10Kq, 10Ksh








385

have a peculiar pitch distribution at neutral pitch

compared to the design pitch as shown in Figure 10, in

which the negative pitch ratio is distributed from the

blade tip to about 70% of the propeller radius, R, while

the positive pitch ratio remains in the other blade root

region. This results in very complex flow around the

whole blade in the ahead operating condition from J=0.1

to 0.2. The forward flow arises around the blade root

region, while the reverse flow arises around the blade tip

region. The flow is seemed to be closely similar to that in

a crash back operation (Verma et al 2011).

Figure 11 shows the computed streamlines around the

whole blade for a pitch angle α=-30 degree at J=0.2,

which is shown in the absolute coordinate system. It is

found from the figure that the flow is very complicated

due to the leading edge separation vortex and the reverse

flow around the blade tip. Figure 12 shows the velocity

vectors on the center section parallel to the propeller

generator line in the same condition. It is found that there

is strong vortex shedding from the leading edge in the

face side and the reverse flow involved by the vortex.

Figure 13 shows the contour of the pressure coefficient,

Cpn in the same section, which is defined by the following

equation,

$% & ' &(

)

4

Where p: static pressure, p0: standard pressure. It is found

that there are the lowest pressure region in the vortex core

of the leading edge separation vortex and the negative

pressure region both on the back side of blade root region

and the face side of blade tip region. Figure 14 shows the

contour of Cpn on both the face and back blade surface. It

is found that there is a lowest pressure area around the

Figure 10 Pitch ratio distribution at design pitch and αααα

=-30 degree (CPP1).

0.20.20.20.2 0.30.30.30.3 0.40.40.40.4 0.50.50.50.5 0.60.60.60.6 0.70.70.70.7 0.80.80.80.8 0.90.90.90.9 1.01.01.01.0

Pitch ratio : P/D

Pitch ratio : P/D

Pitch ratio : P/D

Pitch ratio : P/D

r/Rr/Rr/Rr/R

Alpha = -30 deg.Alpha = -30 deg.Alpha = -30 deg.Alpha = -30 deg.

Design pitch (Alpha=0.0 deg.)Design pitch (Alpha=0.0 deg.)Design pitch (Alpha=0.0 deg.)Design pitch (Alpha=0.0 deg.)

0.00.00.00.0

(+)(+)(+)(+)

((((----))))

Figure 11 Streamlines at J=0.2 and αααα =-30 degree

(CPP1).

Figure 12 Velocity vectors in center section along

generator line at J=0.2 and αααα=-30 degree (CPP1).

Figure 13 Pressure contour in center section along

generator line at J=0.2 and αααα=-30 degree (CPP1).

(b) Face side (a) Back side

Figure 14 Pressure contour on blade surface at J=0.2

and αααα=-30 degree (CPP1).

386

leading edge on the face side. The negative pressure area

from the front part on the face side leads to the negative

hydrodynamic spindle torque decreasing the blade pitch.

However, as shown in the Figure 15 which is a

comparison between the computed hydrodynamic spindle

torque coefficients and the measured ones, which are

shown with error bars of confidence interval of 95%, the

agreement is not so good. In order to improve the

computational accuracy furthermore, some additional

computations were carried out with a finer grid system of

about 8 million cells and with different turbulence

models. However, any more accurate results were not

obtained. Actually, the computation with α=-30 degree at

J=0.2 was changed further from a steady to an unsteady

mode. As shown in the Figure 15, an inadequate result

was obtained. A next step would be to perform the

computation with the techniques of LES or DES. To

conclude, similar results of the open water characteristics

were also found for CPP2.

Considering the centrifugal spindle torques, QSC at each

blade pitch setting by using 7.6t/m3 in the blade material

density of Aluminum Bronze Castings, the total spindle

torques, QSHC were assessed using the measured

hydrodynamic spindle torques. Where, the above-

mentioned data is reduced to the following dimensionless

coefficient,

+

,5

. 6

Figure 16 shows a variation of the spindle torque

coefficients with α for CPP1 as an example at J=0.0.

From the figure, the maximum total spindle torque seems

to arise at about -20 degree from the design pitch, which

is about 10 degree away of the neutral blade angle.

Further, the maximum blade spindle torques of both the

CPP1 and the CPP2 were verified by using the computed

and the measured data. Figure 17 shows the comparative

results of both the CPPs at J=0.0. The tendency of the

computed results by CFD agrees rather well with that of

the measured ones. As a result, the CPP2 has more than

20% of the maximum blade spindle torque from the

CPP1. Many factors are expected as the cause of the

difference in the blade spindle torque. It seems that

mainly the distribution of skew causes the difference.

Figure18 shows the distribution of skew angle, θS with r/R

of both the CPP1 and the CPP2. It is considered that

compared with the CPP1, a more forward skew

distribution of the CPP2 causes a larger maximum blade

spindle torque than the CPP1.

Finally, the reason to combine the measurements and the

computations by CFD is not only to validate the

Figure 15 Hydrodynamic spindle torque coefficient

per blade at αααα=-30 degree (CPP1).

-0.005-0.005-0.005-0.005

-0.004-0.004-0.004-0.004

-0.003-0.003-0.003-0.003

-0.002-0.002-0.002-0.002

-0.001-0.001-0.001-0.001

0.0000.0000.0000.000

0.000.000.000.00 0.050.050.050.05 0.100.100.100.10 0.150.150.150.15 0.200.200.200.20 0.250.250.250.25 0.300.300.300.30

Ksh

Ksh

Ksh

Ksh

Advance coefficient : JAdvance coefficient : JAdvance coefficient : JAdvance coefficient : J

Exp.Exp.Exp.Exp.

CFD (Steady)CFD (Steady)CFD (Steady)CFD (Steady)

CFD (Unsteady)CFD (Unsteady)CFD (Unsteady)CFD (Unsteady)

Figure 16 Measured spindle torque coefficients per

blade at J=0.0 (CPP1).

-0.006-0.006-0.006-0.006

-0.004-0.004-0.004-0.004

-0.002-0.002-0.002-0.002

0.0000.0000.0000.000

0.0020.0020.0020.002

0.0040.0040.0040.004

0.0060.0060.0060.006

-35.0-35.0-35.0-35.0 -30.0-30.0-30.0-30.0 -25.0-25.0-25.0-25.0 -20.0-20.0-20.0-20.0 -15.0-15.0-15.0-15.0 -10.0-10.0-10.0-10.0 -5.0-5.0-5.0-5.0 0.00.00.00.0

Ksh, Ksc, Kshc

Ksh, Ksc, Kshc

Ksh, Ksc, Kshc

Ksh, Ksc, Kshc


KshcKshcKshcKshc

KshKshKshKsh

KscKscKscKsc

Figure 17 Comparative results of maximum spindle

torques at J=0.0.

50%50%50%50%

60%60%60%60%

70%70%70%70%

80%80%80%80%

90%90%90%90%

100%100%100%100%

110%110%110%110%

120%120%120%120%

130%130%130%130%

CFD Exp

CPP1 CPP2

Rate of m

axim

um

Rate of m

axim

um

Rate of m

axim

um

Rate of m

axim

um

spindle torque

spindle torque

spindle torque

spindle torque

Figure 18 Skew angle distributions (CPP1 and CPP2).

0.20.20.20.2

0.30.30.30.3

0.40.40.40.4

0.50.50.50.5

0.60.60.60.6

0.70.70.70.7

0.80.80.80.8

0.90.90.90.9

1.01.01.01.0

-20-20-20-20 -10-10-10-10 0000 10101010 20202020 30303030 40404040

r/R

r/R

r/R

r/R

θs (deg)θs (deg)θs (deg)θs (deg)

CPP1CPP1CPP1CPP1

CPP2CPP2CPP2CPP2

387

computed results with the measured ones, but also they

can complement each other. In particular, in Figure 9

around α=-10 degree the expected hydrodynamic spindle

torques are around zero and therefore difficult to measure

accurately due to very small signal gains, but for the CFD

this is easy to compute them, as it is not yet an extreme

off-design condition but in nearly normal pitch settings.

On the other hand, around α=-15 to -35 degree the CFD

is difficult to apply due to the extreme off-design

condition leading very complicated flow around the

blades, but here the measured results can provide more

reliable results due to large signal gains.

5 CONCLUSIONS

It is very important for a CPP design to provide blades

with not only high propulsive performance, excellent

cavitation performance and sufficient strength but also

less blade spindle torque which is very difficult to

estimate accurately by the means of common potential

theory-based methods but also experimentally for the

whole operational range, especially at low advance speed

around neutral blade pitch settings.

For these reasons, both the highly-accurate measurements

with state-of-the-art techniques, and highly-advanced

computations with updated CFD simulation techniques

for accurate estimation of the hydrodynamic blade spindle

torque in open water condition have been developed.

Applying these to two contemporary CPP blades for

RO/RO vessels, the computed results as a whole agree

with the measured ones very well. However, it is found

that there are discrepancies between the computed and

measured results for low advance speed around the

neutral pitch setting due to the very complicated flow

around the blades, which should be investigated further

by using more advanced computational techniques.

Also it is very important for the propeller blade design not

only to validate the computed results with the measured

ones but also to complement each other. Therefore it is

very essential to assess the pros and cons of the

measurements and the computations by CFD and to

utilize them effectively, especially at estimating the blade

spindle torques.

In the future, it is necessary that the presented

measurements and computations should be enhanced such

as to comprehend the complex flow around the blades as

in non-uniform inflow for behind conditions, in cavitating

conditions, interaction with a rudder and transient

operations of CPPs.

ACKNOWLEDGEMENTS

We would like to thank Mr. J. Takasu and Mr. K.

Nakagawa (Kawasaki) for their stimulating discussions

and the review of this paper. Also we gratefully

acknowledge the members of MARIN for conducting the

tank tests.

REFERENCES

Boswell, R.J., Nelka, J.J. and Kader, R.D. (1975).

‘Experimental spindle torque and open-water

performance of two skewed controllable pitch

propellers’. David W. Taylor Naval Ship Research

and Development Center Report No.4753.

Dang, J., Brouwer, J., Bosman, R. and Pouw, C. (2012).

‘Quasi-Steady Two Quadrant Open Water Tests for

the Wageningen Propeller C- and D-Series’.

Proceedings of the Twenty-Ninth Symposium on

Naval Hydrodynamics, Gothenburg, Sweden.

Ferziger, J.H. and Peric, M. (2002). Computational

Methods for Fluid Dynamics (3rd Edition). Berlin:

Springer-Verlag. ISBN 3-540-42074-6.

Funeno, I. (2009). ‘Hydrodynamic Optimal Design of

Ducted Azimuth Thrusters’. Proceedings of the First

International Symposium on Marine Propulsors,

SMP’09, Trondheim, Norway.

Ito, M., Yamasaki, S., Oku, M., Koizuka, H., Tamashima,

M. and Ogura, M. (1984). ‘An Experimental Study of

Flow Around CPP Blade (3rd Report): Measurement

of CPP Blade Spindle Torque’. Journal of the Kansai

Society of Naval Architects, No.192, pp.81-91.

Jessup, S., Donnelly, M., McClintock, I. and Carpenter, S.

(2009). ‘Measurements of Controllable Pitch Propeller

Blade loads Under Cavitating Conditions’.

Proceedings of the First International Symposium on

Marine Propulsors, SMP’09, Trondheim, Norway.

Koushan, K., Spence, S. and Savio, L. (2011). ‘Ventilated

Propeller Blade Loadings and Spindle Moment of a

Thruster in Calm Water and Waves’. Proceedings of

the Second International Symposium on Marine

Propulsors. SMP’11, Hamburg, Germany

Pronk, C. (1980). ‘Blade spindle torque and off-design

behavior of controllable pitch propellers’. Graduation

thesis TU Delft.

Verma, A., Jang, H. and Mahesh, K. (2011). ‘Large Eddy

Simulation of the Effect of Hull on Marine Propulsors

in Crashback’. Proceedings of the Second

International Symposium on Marine Propulsors.

SMP’11, Hamburg, Germany

388

Documents

Measurements and Computations for Blade Spindle Torque · PDF fileMeasurements and Computations for Blade Spindle Torque of Controllable Pitch Propellers in Open Water ... propeller