You are on page 1of 37

Department of Naval Architecture & Ocean Engineering

Universidade Federal do Rio de Janeiro

EEN-408 Hidrodinâmica Aplicada II

4. LINEAR EQUATIONS OF SHIP


MANOEUVRING MOTIONS

Prof. Claudio Rodríguez


Junho, 2019
Learning objectives

o Know the physical meanings and characteristics of the linear hydrodynamic


derivatives in the equations of ship manoeuvring motion;

o Master the method of analyzing the course-keeping ability and turning ability
based on the linear equations of ship manoeuvring motion;

o Know how to derive the Nomoto model and its application in course control
problems;

o Understand and master the relationship between the K, T indexes in the Nomoto
model and ship manoeuvrability.

Slide 2
Introduction
The equations of motions (mathematical models) can be used to
simulate any standard manoeuvres, and hence to predict ship manoeuvrability.

m(u  vr  xG r 2 )  X 0  X u (u  U )  X v v  X r r  X u u  X v v  X r r  X    ...
m(v  ur  xG r)  Y0  Yu (u  U )  Yv v  Yr r  Yu u  Yv v  Yr r  Y   ... Eq. (1)
I z r  mxG (v  ur )  N 0  N u (u  U )  N v v  N r r  N u u  N v v  N r r  N    ...

In the equations above, the hydrodynamic forces/moments acting on a ship in


manoeuvring motion have been expressed in series, with the hydrodynamic derivatives
as coefficients.

Under the assumption that the changes in velocities and accelerations 


     due to the manoeuvring motion and the
rudder angle  are small, the high-order terms in the expressions of hydrodynamic
force and moment can be neglected (righ side of eq. 1).

For consistency, the left side of the eq. (1) should be also linearized by neglecting
higher order terms, i.e.:    So:

Slide 3
Introduction
Then:

mu  X u (u  U )  X v v  X r r  X u u  X v v  X r r  X  
m(v  Ur  xG r)  Yu (u  U )  Yv v  Yr r  Yu u  Yv v  Yr r  Y  Eq. (2)

I z r  mxG (v  Ur )  N u (u  U )  N v v  N r r  N u u  N v v  N r r  N  

Note that, at the initial steady state of forward motion, the hydrodynamic
forces/moments at X0, Y0, N0 have been considered zero.

These simplified equations (linear equations) of ship manoeuvring motions may be


used to analyze the ship manoeuvrability, however, they can be further simplified
when the port-starboard symmetry of the ship’s hull is taken into account in the
hydrodynamic derivatives.

Slide 4
Analyses of linear hydrodynamic derivatives
The linear hydrodynamic derivatives represent the force or moment per unit velocity or
acceleration induced by the motion (velocity or acceleration) of the ship in a single
degree-of-freedom (i.e., the other degrees of freedom are restrained). Mathematically,
they are the gradients (or tangents) at the origin of the curve of force/moment plotted
against the component of the applied velocity or acceleration.

X Y
 X v  0 (symmetry)  Yv
v v 0
v v0

N
 Nv
v v 0

small (uncertain sign) Slide 5


Analyses of linear hydrodynamic derivatives

X Y
 Xu Y  0 (symmetry),  Yu  0
u u 0 u u 0

N
N  0 (symmetry),  Nu  0
u u 0

Slide 6
Analyses of linear hydrodynamic derivatives

X
 X r  0 (symmetry) Y
r  Yr
r 0
r r 0

small (uncertain sign)

N
 Nr
r r 0

Slide 7
Analyses of linear hydrodynamic derivatives

Y
 Y
  0

X
 X   0 (symmetry)
  0

N
 N
  0

Obs: Note the convention for the sign of the rudder angle: (+) to starboard. Slide 8
Analyses of linear hydrodynamic derivatives
According to the analyses of the flow around the ship under each of the velocities,
accelerations and rudder angle applied to the ship, we can conclude that:
o Due to port-starboard symmetry of the ship, X v , X v , X r , X r , Yu , Yu , N u , N u , X  are null.

o X u , Yv , Yr , N v , N r are added mass coefficients analogous to, respectively, A11, A22, A26, A62 , A66
in seakeeping.

o X u , Yv , Yr , N v , N r are damping coefficients analogous to, respectively, B11, B22, B26, B62 , B66 in
seakeeping.

o Y , N  denote the control force and moment derivatives induced by the rudder, designed to
be as large as possible . When the rudder angle is defined as (+) to starboard, Y  0, N   0 .

o 𝑌 ̇ > 𝑋 ̇ . Typically, 𝑋 ̇ ≈ − 0.05~0.15 . 𝑚; 𝑌 ̇ ≈ − 0.90~1.20 . 𝑚, m is the ship’s mass. Also:

“Uncertain” refers to (+) or (-) values and are associated to the resultant influence of bow/stern shapes. So,
the sign will depend on the geometry of the particular ship being analyzed.
Slide 9
Linear manoeuvring equations
After the assumption of port-starboard symmetry on the hydrodynamic derivatives,
the manoeuvring equation of motion becomes:
(m  X u )u  X u (u  U )  0
(m  Yv )v  (mxG  Yr )r  Yv v  (mU  Yr )r  Y  Eq. (3)
(mxG  N v )v  ( I z  N r )r  N v v  (mxGU  N r ) r  N  
Sometimes it is more convenient to use the nondimensional form of the equations of
motions. By dividing the first two equations of eq. (3) by  and the last
equation by  , where  is the mass density of water, L is the ship length and U
is the ship speed at the initial steady state of forward motion the nondimensional
form of the linear equations of ship manoeuvring motion is obtained:

(m ' X u )u ' X u (u ' U ')  0


(m ' Yv)v ' (m ' xG  Yr)r ' Yvv ' (m 'U ' Yr)r '  Y
(m ' xG  N v )v ' ( I z  N r )r ' N vv ' (m ' xG U ' N r )r '  N 

Slide 10
Nondimensional manoeuvring equations

where:

Slide 11
Analysis of (inherent) dynamic stability
To analyze the inherent dynamic stability of the ship (i.e. the dynamic stability without
control actions), we set  = 0 and assume that the motions are induced by a small
disturbance. The surge equation in eq. (3) is uncoupled from the sway and yaw
equations. Therefore, it can be solved separately. On the other hand, the sway and
yaw equations are coupled each other and hence have to be solved together.

surge:
(m  X u )u  X u (u  U )  0 Eq. (4)

sway & yaw:


(m  Yv )v  (mxG  Yr )r  Yv v  (mU  Yr ) r  0 Eq. (5)
(mxG  N v )v  ( I z  N r )r  N v v  (mxGU  N r )r  0

or in matrix form (sway & yaw) :


 (m  Yv ) (mxG  Yr )   v   Yv (mU  Yr )   v   0  Eq. (6)
 (mx  N ) ( I  N )   r     N     
(mxGU  N r )   r   0 
 G v z r     v
Slide 12
Analysis of (inherent) dynamic stability
Surge
Eq. (4) can be rewritten as:

d (u  U ) Xu Eq. (7)
 (u  U )  0 (ordinary differential equation of first
dt (m  X u ) order with constant coefficients)

whose solution is given by:


Eq. (8)
u  U  C3e3t

Xu
where: 3  C3 must be determined from the initial conditions
(m  X u )

 3t
Since Xu < 0, ̇ > 0 we have 3 < 0. Therefore, as t  , the term e on the
right side of Equation (6) will vanish, resulting in u  U, i.e., the longitudinal motion
of the ship will finally tend to the original steady state with a constant forward
speed.

Slide 13
Analysis of (inherent) dynamic stability
Sway & yaw
Eq. (5) can be rearranged as:
(m  Yv )v  Yv v  (mxG  Yr )r  (mU  Yr )r
Eq. (9)
(mxG  N v )v  N v v  ( I z  N r )r  (mxGU  N r )r
Solving for and :

Eq. (10)

Differentiating the first equation of Eq. (10) with respect to time and letting the result
equal to the second equation, we eliminate and from Eq. (10) and obtain:
B C
Ar  Br  Cr  0, or  r  r  r  0 Eq. (11)
A A
where:
A  (m  Yv )( I z  N r )  (mxG  N v )(mxG  Yr )
B  Yv ( I z  N r )  N v (mxG  Yr )  (m  Yv )(mxGU  N r )  (mxG  N v )( mU  Yr )
C  Yv ( mxGU  N r )  N v (mU  Yr ) Slide 14
Analysis of (inherent) dynamic stability
Eq. (11) is an ordinary differential equation of second order with constant coefficients,
whose solution is given by:
r  r e 1t  r e2t
1 2
Eq. (12)

where r1 and r2 are integral constants to be determined by the initial condition, 1 and 2
are given by:
1 B C
2
 B Eq. (13)
1,2        4 
2 A  A A
 
Similarly, if we eliminate and from Eq. (5), we obtain:
B C
Av  Bv  Cv  0, or v  v  v  0 Eq. (14)
A A
with the solution:
v  v1e 1t  v2 e2t Eq. (15)

where v1 and v2 are integral constants to be determined by the initial condition, 1 and 2
are given by Eq. (13).

Slide 15
Analysis of (inherent) dynamic stability
If both 1 and 2 are negative real numbers or imaginary numbers with negative
real parts, we have e1t , e2t  0 as t  . That means, after a small disturbance,
r  0 and v  0 as t  , the ship will finally tend to a new straight course with
constant forward speed. At the meantime, the ship will have a heading deviation
from the original heading which is given by:
t0

 (t )   r (t )dt Eq. (16)


0
where t0 is the moment when r vanishes. Obviously, a ship with better inherent dynamic
stability will tend to a new straight course faster and will have smaller heading deviation.

From the above analysis, it can be seen that a ship affected by a small disturbance
can not keep its original course without control action. It can only turn into a new
straight course with a heading deviation from the original course if it has inherent
dynamic stability. Contrarily, if the ship is unstable, the yaw rate given by Eq. (12) will
not vanish as t  , therefore the ship will turn into an unsteady turning motion.

In summary the criterion of inherent dynamic stability is: “both 1 and 2 are negative
real numbers or imaginary numbers with negative real parts”.
Slide 16
Analysis of (inherent) dynamic stability

Slide 17
Further analysis of (inherent) dynamic stability
1 B C
2
 B
1,2       4 
2 A  A A
 
To ensure both 1 and 2 are negative, two essential conditions must be satisfied:

A) C/A > 0. If C/A < 0, the square root is larger than B/A, with that, either being B/A
positive or negative, one of the values of  will be always positive.

B) B/A > 0. Starting with C/A > 0, B/A < 0 will make both 1 and 2 positive.

Mathematically, the criterion for inherent dynamic stability can be expressed as:
1  2  0 and 1   2  0
Then:
1 B C 1 B C C
2 2
B  B
1   2        4        4    0
2 A  A A 2 A  A A A
   
1 B C 1 B C
2 2
B  B B B
1   2        4         4     0  0
2 A  A A 2 A  A A A A
    Slide 18
Further analysis of (inherent) dynamic stability
Because:

A is always positive:
A  (m  Yv )( I z  N r )  (mxG  N v )(mxG  Yr )  A  0

B is also positive:
B  Yv ( I z  N r )  N v (mxG  Yr )  (m  Yv )(mxGU  N r )  (mxG  N v )(mU  Yr )  B  0

Therefore, the criterion for inherent dynamic stability is reduced to:

C  Yv ( mxGU  N r )  N v ( mU  Yr )  0 Eq. (17)

Or in nondimensional form: stability derivatives


Eq. (18)
C   Yv(mxG U   N r )  N v (mU   Yr)  0
Obs: The larger the magnitude of the positive C , the better is the inherent dynamic stability
Slide 19
Further analysis of (inherent) dynamic stability
The criterion can also be expressed in terms of moment arms. To do so eq. (17) should
be divided by Yv (mU  Yr )  0

C  Yv ( mxGU  N r )  N v (mU  Yr )  0

(mxGU  N r ) N v
 0
(mU  Yr ) Yv
lr  lv  0

Nv
lv 
Yv
(mxGU  N r )
lr 
(mU  Yr )
Neutral point
Slide 20
Nomoto model and K, T indexes
From the hydrodynamic force models, now we will derive the response models. These
models are used to study control problems in ship motions or to design control devices,
such as rudders. So, turning ability and course-changing ability will now be analyzed.
In eq. (3), sway and yaw equations can be re-arranged as follows:
( m  Yv )v  Yv v  Y   ( mxG  Yr )r  (mU  Yr )r
( mxG  N v )v  N v v  N    ( I z  N r )r  (mxGU  N r )r Eq. (19)

Solving Eq. (19) for and yields:

Eq. (20)

Differentiating the first equation of Eq. (20) with respect to time and letting the result
equal to the second equation, we obtain:

Slide 21
Nomoto model and K, T indexes
T1T2 
r  (T1  T2 )r  r  K   KT3 Eq. (21)
Yaw motion response
where: to the control action
(m  Yv )( I z  N r )  (mxG  N v )( mxG  Yr ) (2nd order model)
T1T2 
Yv ( mxGU  N r )  N v (mU  Yr )
Yv ( I z  N r )  N v (mxG  Yr )  (m  Yv )(mxGU  N r )  (mxG  N v )(mU  Yr )
T1  T2 
Yv ( mxGU  N r )  N v ( mU  Yr )
Yv N   N vY
K
Yv (mxGU  N r )  N v (mU  Yr )
Y ( mxG  N v )  N  (m  Yv )
KT3 
Yv (mxGU  N r )  N v ( mU  Yr )
KT3 Y (mxG  N v )  N  (m  Yv )
T3  
K Yv N   N vY
The coefficients on the left side of Eq. (11) can be compared to those in Eq. (21), so that
the following relationships can be established:

Slide 22
Nomoto model and K & T indexes
Assuming that the ship is approximately fore-and-aft symmetrical, we have:
xG  0, Yr  0, Yr  0, N v  0, N v  0
So that:
N I  N r m  Yv m  Yv I z  N r
K  , T1  T2   z  , T3   , T1  T2  T3  
Nr Nr Yv Yv Nr

On the other hand, from the third equation of Eq. (3) we get:
( I z  N r ) r  N r r  N  

Or, re-arranging terms:


( I z  N r ) N
 r  r    
Nr Nr
Denoting T = T1+ T2 – T3, we have: Nomoto’s model
Yaw motion response

Tr  r  K 
to the control action
Eq. (22) (1st order model)

Eq. (22) was firstly derived by japanese professor Nomoto in 1957. K and T are called
manoeuvrability indexes.
Slide 23
Stability under rudder action
Let’s assumed that the ship is initially in its straight course with a steady forward speed;
the rudder is ordered to turn to a certain angle 0 during the time t0 with a uniform
rudder turning rate and then held fixed. The rudder angle is described by:

First phase: first phase

1 K  0t
r  r  K 0   
t

0  t  t0 r t  T  1  e T 
T T t0 t0    
K 0 t2 
t

With initial condition (at t = 0): r=0    Tt  T 2
(1  e T
) 
t0 2 
K 0   
t
r  1  e 
T
t0   Slide 24
Stability under rudder action
N I  N r
Recalling again that: K  , T  z
Nr Nr

and N  0

Then: K  0, T  0
Therefore, during first phase, for a certain time t and for smaller T:
t
 decreases faster
e T

K 0  t
0  
t
larger turning rate
r  t0  T (1  e )e 
T T
t0  
K 0   
t larger yaw acceleration (1st phase)
r  1  e 
T
t0  
Smaller T, better course-changing
(initial turning) ability!
Slide 25
Stability under rudder action
Second phase:

1 K
r  r  0 t > t0
T T

with initial condition (at t=t0):


second phase
K 0  0 
t
r  t0  T  Te T 
t0  

K 0  t
0  
t
r  t0  T (1  e T )e T 
t0  
t
K 0  t02 t0
 
t
  (t0 )   rdt  t0 (t  T )   T (e  1)e 
2 T T

t0
t0  2 
K 0  tT0   Tt
r   e  1 e
t0  

Slide 26
Stability under rudder action
Recalling that: N I  N r
K  , T  z
Nr Nr

and N  0

Then: K  0, T  0
Therefore, for t  :
t

e T
0
steady yaw rate (steady turn)
r  K 0  r0
 t 
  K 0 t  (T  0 )  heading angle changing linearly with time
 2 
r  0
D0 2V0
V0  r0  R  K 0  D0 
2 K 0 Larger K, better turning ability!
Slide 27
Inherent dynamic stability Nomoto model
Yaw motion without rudder action:
Tr  r  0
whose solution is given by: t
r  r0e T
with initial condition (at t = 0): r = r0

As t   the yaw rate will vanish after the disturbance. Smaller T will make the yaw rate
vanishing faster and hence give a better inherent dynamic (straight course) stability.

In summary:
o K  turning ability;
o T  inherent dynamic stability & initial turning (course-changing) ability;
o Good manoeuvrability  large K & small T.

However, typically a large K usually accompanies a large T; or


contrarily, a small K accompanies a small T.

Recall that the above conclusions have been obtained assuming that the ship has fore-aft symmetry.

Slide 28
Analysis of K & T
N I  N r
Recalling that: K  , T  z
Nr Nr
Or, in nondimensional form: L U 
K     K , T    T
U  L
To obtain a large K (good turning ability), we wish a small Nr. However, this will result
in a large T (poor inherent dynamic stability and poor course-changing ability) at the
same time. Therefore, we cannot improve the ship manoeuvrability only by changing
Nr. On the other hand, obtaining large N (through rudder design) may result in good
turning ability without loss of other manoeuvring performances.

To obtain a small T (good dynamic stability and course-changing ability), one might try
to design the ship with small moment of inertia and small added moment of inertia.
However, the moment of inertia and added moment of inertia are mainly determined
by the principal dimensions and ship form coefficients which are chosen according to
other ship performances. Hence, there is no big space for choosing. For large ship with
full form such as VLCC, T is usually very large due to the large inertia. Therefore, such
ships are usually dynamically unstable and have poor course-changing ability.

Slide 29
Analysis of turning ability based on linear equations
Recalling in eq. (3):
(m  Yv )v  (mxG  Yr )r  Yv v  ( mU  Yr )r  Y 
(mxG  N v )v  ( I z  N r ) r  N v v  ( mxGU  N r ) r  N  

It is assumed that a ship originally travels with a steady forward speed in a straight
course, and the rudder is ordered to a certain rudder angle to starboard or port with a
constant rudder turning rate and then held fixed. Ultimately, the ship tends to a steady
turning motion.

From the perspective of ship motions, the whole process can be divided into three
phases:

 rudder-turning phase;

 transition phase;

 steady turning phase.

Slide 30
Analysis of turning ability based on linear equations

Slide 31
Analysis of turning ability based on linear equations

Slide 32
1st phase: rudder-turning phase
At this phase (0  t  t0), the rudder angle  is given =0.t/t0, where 0 is the rudder
angle applied and t0 is the time of rudder-turning. The control force and moment of the
rudder will produce sway acceleration and yaw acceleration. But the time of rudder-
turning is so small and the inertia the ship is so large that at this phase no notable sway
speed and yaw rate are induced. So:
(m  Yv )v  (mxG  Yr )r  Y 
(mxG  N v )v  ( I z  N r ) r  N  
Solving:
(-)
Y
v  
(m  Yv )
 N K
r   
( I z  N r ) T
(+)
For the simplification it was assumed that:
xG  0, Yr  0, Yr  0, N v  0, N v  0

Slide 33
2nd phase: transition phase
At this phase:

Then:
(m  Yv )v  (mxG  Yr )r  Yv v  ( mU  Yr ) r  Y 0
(mxG  N v )v  ( I z  N r ) r  N v v  ( mxGU  N r ) r  N  0

At the transition phase, the control force and moment remain constant, whereas the
kinematical quantities of the ship change with time. We can solve the equations of
motions by numerical integration method, for example, by the Runge-Kutta method,
to obtain the kinematical quantities and .

At the initial transition stage, the ship moves laterally to the outside (other side as the
rudder deflection) under rudder force action and turns to inside (the same side as the
rudder deflection) under rudder moment action. With the development of
and , the hydrodynamic force and moment acting on the ship hull increase
gradually. Ultimately, the hydrodynamic force and moment acting on the hull and the
control force and moment due to the rudder angle reach equilibrium, the ship then
enters to the next phase, i.e., the steady turning phase.
Slide 34
3rd phase: steady-turning phase
At this phase where and are the steady sway
velocity and the steady yaw rate. Then:
Yv v0  (mU  Yr ) r0  Y 0
 N v v0  (mxGU  N r )r0  N  0
Solving:
(mU  Yr ) N   (mxGU  N r )Y
v0  0
Yv (mxGU  N r )  N v ( mU  Yr )
Yv N   N vY
r0  0
Yv (mxGU  N r )  N v (mU  Yr )

The denominator in the above equations is the index C described in eq. (17). Then:

(mU  Yr ) N   ( mxGU  N r )Y


v0  0
C
Y N  N vY
r0  v  0
C
Slide 35
3rd phase: steady-turning phase
The steady-turning diameter can be obtained
from: V0  C 
D0 D  2
V0  r0 0  
2  0  Y N
v   N Y
v  

Better directional (dynamic) stability


where: V0  U  v  U
2 2
0 implies in worse turning ability, and
viceversa!

If C > 0 (ship dynamically stable), for a positive rudder


angle (rudder to starboard), we have r0 > 0 (turning to
starboard); whereas for a negative rudder angle (rudder
to port), we have r0 < 0 (turning to port). This situation is
what we expect.

On the other hand, if C < 0 (ship dynamically unstable),


for a positive rudder angle (rudder to starboard), we have
r0 < 0 (turning to port); whereas for a negative rudder
angle (rudder to port), we have r0 > 0 (turning to
starboard). That means, for a ship which is dynamically
unstable, with rudder to starboard, the ship will turn to
port; whereas with rudder to port, the ship will turn to
starboard, i.e., the ship is uncontrollable.
Slide 36
Questions?

Prof. Claudio Rodríguez


claudiorc@oceanica.ufrj.br
Federal University of Rio de Janeiro (UFRJ), Brazil

You might also like