Investigating the behavior of a function modeling the depth

of a towfish
Rationale
This investigation aims to investigate the behavior of a function
that models the depth of a towfish traveling at a constant velocity.
As shown in figure 1 towfish are missile-shaped probes towed at
specific depths behind boats to study the seafloor at depths out
of reach from active sonar on the surface (Wikmark). Towfish are
a significant resource for subsea industry, military endeavors, and
scientific investigation (Wikmark). Despite the applicability of
towfish there has been little development in their operational
methods in the last 20 years (Wikmark). Currently, towfish
manage their depth by varying the length of rope they are tethered
to, and micro-adjustments are done by use of fins and engines, Figure 1 Schematic showing a towfish vessel system
essentially depth is mechanically varied by direct application of (Townsend)
force to the towfish (Wikmark). This blunt approach may be
outdated and inefficient, and there is great room for development. Recently there has been great
developments in the aerospace industry with companies like Boeing revolutionizing airplane design with
the new X-37B spaceplane (Wall). The significance of the X37-B is that it generates lift by using the
aerodynamics of its body rather than wings, which has proven to be an efficient space shuttle design (Wall).
Perhaps the concept of using a body’s inherent aerodynamic properties to its advantage can be applied to
towfish design as well. The equivalent to the X-37B in towfish design would be designing a towfish that
varies its depth by changing its environment rather than applying a brute force directly to itself, for example
increasing the drag experienced by a towfish would cause its depth to decrease. To apply this new method
of towfish depth management, a mathematical model of how varying the environmental conditions a
towfish is under, changes its depth must be explored. This technology could potentially save companies,
and researchers millions and is thus essential to be investigated. This exploration is however only concerned
with the basics of how a towfish responds to changes in its environment, more specifically its only
concerned with mathematically modeling how the towfish behaves when at a constant velocity. But the
investigation strongly urges other researchers to build on the concepts and ideas presented in this
investigation.
In this investigation a function that models the depth of a towfish based on the behavior of the vector
forces acting on the towfish was investigated. The success of the function was then evaluated by
investigating the behavior of the function when its argument was set as the towfish’s velocity and its
geometry. The investigation produced the following function, whose behavior was analyzed and shown to
be effective when modeling the effect of varying the towfish’s velocity under ideal conditions:

2
𝑛 1 2
𝑓 = ∑ √1 − 2 𝜌𝐶𝐷 𝐴𝑣⃗
2
𝑖=1 √ 1 2 2
( (2 𝜌𝐶𝐷 𝐴𝑣⃗ ) + (𝑔⃗(∆𝜌𝑖𝑉𝑛 + 𝜌𝑉 − 𝑚)) )
𝑖

The function f models the depth of the towfish as any one of the parameters is varied (under ideal
conditions). Due to the expanse of this investigation, it was split into several components which were
explored individually these 3 sections of the exploration are:
1. Conjecturing a function that models the behavior of a towfish tethered to a straight wire (a very
simplified model)

Page | 1
 2. Conjecturing a function that models the behavior of a towfish tethered to a wire that changes shape
in response to its environment (a more realistic model)
3. Exploring the behavior of the function

Introduction to the task

When a towfish is towed by a towing
vessel, as shown in figure 2, there are a
variety of vector forces acting on the
system. If the vector sum of all forces
acting on the system is 0 at any given time,
the system will have a defined shape (the
towfish will have a constant position), but
if the vector sum is greater or less than 0,
the system will move until it reaches a state
where the vector sum of all forces are 0
(equilibrium)(Allum and Talbot). Thus it's
essential to understand the behavior of the
forces acting on the system to conjecture a Figure 2 free body diagram showing all the forces acting on the system. where vectors
function modeling the depth of the are presented in bold
towfish. The forces shown in figure 2 are
explained in greater detail below.
The vector forces acting on the towfish at any time are the force of gravity, buoyancy, and drag. These
vector forces act on the towfish as well as the rope connecting the towfish to the towing vessel (the vector
forces acting on the towfish are shown in figure 2). Gravitational force is the fundamental force of attraction
between any 2 masses in the universe (Allum and Talbot), it can be calculated on earth using Newtons
second law (Allum and Talbot):

⃗⃗⃗⃗
𝐹𝑔 = −𝑚𝑔⃗

Where ⃗⃗⃗⃗
𝐹𝑔 is the gravitational force acting on the object, m is the objects mass, and 𝑔⃗ is the acceleration due
to gravity (Allum and Talbot).
The buoyant force is an upwards force that always acts in opposition to the force of gravity (Allum and
Talbot). The buoyant force is caused by a difference in pressure between the bottom of a vessel and the
top when submerged in a fluid (Allum and Talbot). Mathematically the buoyant force can be expressed
through Archimedes principle which states that the buoyant force acting on an object when submerged is
equal to the weight of the fluid displaced, which is summarized in the below relation:

⃗⃗⃗⃗⃗
𝐹𝐵 = 𝜌𝑔⃗𝑉

Where ⃗⃗⃗⃗⃗
𝐹𝐵 is the buoyant force acting on the object, 𝜌 is the density of the fluid, 𝑔⃗ is the acceleration due
to gravity, and V is the volume of the displaced fluid (Allum and Talbot). The vectors ⃗⃗⃗⃗
𝐹𝑔 and ⃗⃗⃗⃗⃗
𝐹𝐵 are shown
in figure 2 to only act on the towfish and towing vessel, although they act on the wire as well. The ideal
wire is neutrally buoyant so the vector sum of ⃗⃗⃗⃗
𝐹𝑔 and ⃗⃗⃗⃗⃗
𝐹𝐵 is 0, thus no vectors are shown on the ideal wire.
While the real wire, is negatively buoyant which is shown by the force in the figure denoted by 𝑝⃗. 𝑝⃗ is the
distributed net force of gravity, which is the vector sum of ⃗⃗⃗⃗𝐹𝑔 and ⃗⃗⃗⃗⃗
𝐹𝐵 acting on every point in the wire
(Allum and Talbot). 𝑝⃗ is the force that causes the curved shape of the real wire which is known as a catenary
curve (Svirin).
The third force that acts on the towfish is the force of drag. Drag is a frictional force caused by collisions
between particles of the fluid and the object in motion, during each of these collisions the object transfers

Page | 2
some of its kinetic energy to the particles of the fluid, thus drag always acts opposite to the direction of
motion (Allum and Talbot). Mathematically drag can be expressed using the drag equation:
1
𝑅⃗⃗ = 𝜌𝐶𝐷 𝐴𝑣⃗ 2
2
Where 𝑅⃗⃗ is drag, 𝑣⃗ is the velocity, 𝜌 is the density of the fluid, 𝐶𝐷 is the drag coefficient of the object, and
A is the reference area, which is the cross-sectional area of the object on the plane perpendicular to the
direction of motion (Allum and Talbot).
The vector forces acting on the towing vessel are what causes the towfish to move. Since the investigation
is only concerned with the behavior of the towfish when at a constant velocity the vector sum of the force
⃗⃗⃗⃗
𝐹𝑒 (engine output) and 𝑅⃗⃗ drag can always be assumed to be 0 which causes the towfish to be at a constant
velocity.
All these forces can be expressed in terms of the force of internal tension in the ideal or the real wire. Thus
the behavior of the function for the depth of the towfish will be mathematically treated as wholly dependent
on the forces presented in the diagram in terms of internal tension.

Towfish dynamics
The simplest possible model of a towfish system
that to some extent reflects reality is a towfish
tethered to an ideal wire (Wikmark). An ideal wire
is a neutrally buoyant wire that behaves like a rigid
body, meaning that the application of a force to the
body doesn’t change its shape (Wikmark). The first
model considered in this investigation was the ideal
wire model as it’s the simplest model that reflects
reality, and can thus serve as a good starting point
for further exploration.
To conjecture a function modeling the behavior of Figure 3 Free body diagram showing vector forces acting on the idealised
system
the idealized system the following 2 axioms were
considered along with figure 3:
1. For the towfish to have a defined constant position the vector sum of all forces acting on the
system must be 0
2. In an ideal wire, the internal tension vector must be colinear with the wire.
The first axiom is valid at all times because if the vector sum of forces acting on the towfish is greater than
or less than 0 the unbalanced force will cause the system to move until it gets to a position where the vector
sum of forces is 0. The significance of this axiom is that at all times the vector sum of 𝑇⃗⃗, ⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝑅(𝑣), 𝐹 ⃗⃗⃗⃗ ⃗⃗⃗⃗⃗
𝑔 , and 𝐹𝐵
must be 0. Thus:

⃗⃗ = 𝑅⃗⃗ + ⃗⃗⃗⃗
−𝑇 𝐹𝑔 + ⃗⃗⃗⃗⃗
𝐹𝐵

This property of the tension vector is significant when considered along with the second axiom; since the
⃗⃗ determines the angle the
tension vector and the ideal wire are always colinear; the direction of the vector 𝑇
wire will take to the x-axis, and thus determines the depth of the towfish, as shown in figure 4. Since 𝑅⃗⃗ only
has a non-zero x component, and ⃗⃗⃗⃗⃗
𝐹𝐵 and ⃗⃗⃗⃗
𝐹𝑔 only have non-zero y components the tension vector 𝑇 ⃗⃗ colinear
with the wire is:

Page | 3
 𝑅⃗⃗
⃗⃗ = − (
𝑇 )
⃗⃗⃗⃗⃗
𝐹𝐵 + ⃗⃗⃗⃗ 𝐹𝑔

Since the vector 𝑇⃗⃗ determines the angle the ideal

wire takes to the x axis, by applying Pythagoras
theorem the depth of the towfish can be given as
follows by considering the wire as the hypotenuse
of a right angled triangle as shown in figure 4:

𝐷 = 𝑙 sin 𝜃
Where D is the depth of the towfish, l is the length
of the wire, and θ is the angle between the tension
vector and the x axis. To make this function
Figure 4 Sketch showing the theoretical triangle formed between the
wire and the axes
applicable to the sensory data available to towfish
vessels a function for θ was explored.
The angle between the tension vector and the x axis can be expressed in 2 ways by considering the x and y
components of 𝑇⃗⃗ in cartesian form:

⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝑅(𝑣) 𝑥 = |𝑇 ⃗⃗| cos 𝜃 = −𝑅⃗⃗
⃗⃗ = − (
𝑇 )={
⃗⃗⃗⃗⃗
𝐹𝐵 + ⃗⃗⃗⃗ 𝐹𝑔 ⃗⃗| sin 𝜃 = −(𝐹
𝑦 = |𝑇 ⃗⃗⃗⃗⃗ ⃗⃗⃗⃗
𝐵 + 𝐹𝑔 )

Thus the function for the angle between the tension vector and the x axis that uses relevant sensory data
is:

⃗⃗⃗⃗⃗
−(𝐹 ⃗⃗⃗⃗
𝐵 + 𝐹𝑔 )
𝜃 = 𝑠𝑖𝑛−1 ( )
|𝑇⃗⃗|

Or:

−𝑅⃗⃗
𝜃 = 𝑐𝑜𝑠 −1 ( )
⃗⃗|
|𝑇
Thus the function that models the depth of a towfish tethered to an ideal wire is:

⃗⃗⃗⃗⃗
−(𝐹 ⃗⃗⃗⃗
𝐵 + 𝐹𝑔 )
𝐷 = 𝑙𝑠𝑖𝑛 (𝑠𝑖𝑛−1 ( ))
|𝑇⃗⃗|

Which can be simplified to:

⃗⃗⃗⃗⃗
−(𝐹 ⃗⃗⃗⃗
𝐵 + 𝐹𝑔 )
𝐷=𝑙
|𝑇⃗⃗|

Which in expanded form is:

−𝑔⃗(𝑉𝜌 − 𝑚)
𝐷=𝑙
2
√(1 𝜌𝐶𝐷 𝐴𝑣⃗ 2 ) + (𝑔⃗(𝑉𝜌 − 𝑚))2
2
Where D is the depth of the towfish, 𝑙 is the length of the wire, 𝑔⃗ is gravitational acceleration, V is the
volume of the towfish, 𝜌 is the density of water, CD is the drag coefficient of the towfish, A is the cross
sectional area of the towfish, 𝑣⃗ is the velocity of the towfish, and m is the mass of the towfish.

Page | 4
The function D(𝑣⃗) is shown plotted in figure 5 with parameters set as realistic values (the exact values
chosen are unimportant):
Figure 5 shows the function D(v)
plotted on a cartesian co-ordinate
system. Figure 5 shows that the
function D(v) takes a bell shape that
rapidly converges to y=0. Figure 5
supports the accuracy of the function,
because the wire length was set to 100
and figure 5 shows that the maximum
depth that can be reached by the
towfish is 100 m. It also shows that
the function D(v) is very responsive to
changes in v, which makes sense, as
velocity in the drag equation is
squared.
Although the accuracy of this function
is qualitatively supported by figure 5
Figure 5 Plot of function D(v) showing the depth of the towfish plotted
against its velocity
its still only able to model the depth of
the towfish under ideal conditions
where the flow velocity of the fluid is constant, and the wire pulling the towfish is considered as an ideal
non rigid body. Thus, the function likely estimates how a function modeling the depth of a towfish would
behave, however it doesn’t actually model how the depth of a towfish changes in response to its
environment. Thus this exploration was continued to explore how a function modeling a towfish tethered
to a non rigid wire behaves.
Towfish with a real wire dynamics
The next logical question to ask is how the behavior of the function changes when the towfish is tethered
to a real wire as shown in figure 6. A real wire is a wire that behaves like a non-rigid body, which are bodies
that change shape in response to the vector forces acting on them (Wikmark). To conjecture a function
that models the depth of a towfish attached to a real wire the following axioms of the wires behavior must
be considered:
1. For a non-rigid body to have a defined
constant shape the vector sum of all
forces acting on it must be 0
2. Tension vectors are always tangential
to the point in a wire they act on
The first axiom can be summarized in terms of
internal tension; all the forces acting on a point
in the wire can be expressed in terms of one
tension vector, and the vector sum of all the
tension vectors on all the points in the wire must
equal 0 for the wire to have a defined constant Figure 6 free body diagram showing the vectors acting on the towfish
system when the towfish is tethered to a real wire
shape. The second axiom is of special
importance, it essentially says that the gradient
of a tension vector at point x is the same as the gradient of the line at point x. Thus when the sum of the
internal tension vectors in the wire is 0 the following function f(x) should model the shape of the wire:

⃗⃗⃗⃗𝑥 𝑑𝑥 + 𝑐
𝑓(𝑥) = ∫ 𝑇

Page | 5
 ⃗⃗⃗⃗𝑥 is the tension vector tangential to the wire at point x. due to axiom 1 stated above, 𝑇
Where 𝑇 ⃗⃗⃗⃗𝑥 is equal
and opposite to all the vector forces that act on any point below point x:

𝑅⃗⃗ + ⃗⃗⃗⃗⃗
𝑅𝑥
⃗⃗⃗⃗𝑥 = − (
𝑇 )
𝑝⃗ + ⃗⃗⃗⃗⃗
𝐹𝐵 + ⃗⃗⃗⃗
𝐹𝑔

Where 𝑅⃗⃗ is the drag acting on the towfish, and ⃗⃗⃗⃗⃗

𝑅𝑥 1 is the drag acting on a small segment of wire at which
point x is on (drag cannot act on a point), and 𝑝⃗ is the distributed force of gravity and buoyancy acting on
any point >x, ⃗⃗⃗⃗⃗𝐹𝐵 is the buoyant force acting on the towfish, and ⃗⃗⃗⃗
𝐹𝑔 is the force of gravity acting on the
towfish. Ideally this vector would be investigated to conjecture a function f(x) modelling the depth of a
towfish in response to various parameters, however this is impossible without advanced computing
software, because the x component of the vector contains a loop dependency. The function f(x) depends
on 𝑇 ⃗⃗⃗⃗𝑥 which depends on ⃗⃗⃗⃗⃗
𝑅𝑥 , but ⃗⃗⃗⃗⃗
𝑅𝑥 depends on the cross sectional area of wire exposed to the fluid thus
⃗⃗⃗⃗⃗
𝑅𝑥 is dependent on f(x) and vice versa, thus an insolvable loop occurs. This loop is however eliminated if
the wire is considered to be very thin, making ⃗⃗⃗⃗⃗
𝑅𝑥 is negligible so the tension vector on any point in the thin
wire is:

𝑅⃗⃗
⃗⃗⃗⃗𝑥 = − (
𝑇 )
𝑝⃗ + ⃗⃗⃗⃗⃗
𝐹𝐵 + ⃗⃗⃗⃗
𝐹𝑔

The vector 𝑇⃗⃗⃗⃗𝑥 can be used to conjecture a function f(x), but since vector calculus is out of the scope of this
exploration a workaround had to be explored, which was done by considering the conditions allowing 𝑇 ⃗⃗⃗⃗𝑥
to model the wire. As mentioned in the first axiom of the wires behavior, a non-rigid body will only have a
constant shape when the vector sum of all forces acting on it is 0. Thus, the only condition that allows the
function f(x) to model the depth of the towfish is when:

⃗⃗⃗⃗𝑥 = 𝑇
𝑇 ⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝑥+∆𝑥

For any values of x and x+∆x along the wire. By considering the vector 𝑇 ⃗⃗⃗⃗𝑥 and figure 6 it becomes clear
that the only component of 𝑇 ⃗⃗⃗⃗𝑥 that changes between a point x and x+∆x is 𝑝⃗. This is because 𝑝⃗ is the only
vector that is directly intersecting the wire as seen in figure 6. Since the only parameter that changes between
points in the wire is 𝑝⃗ and 𝑝⃗ is a parameter of the y component of 𝑇 ⃗⃗⃗⃗𝑥 the conditions that allow f(x) to model
the depth of the towfish can be specified to:

|𝑇 ⃗⃗⃗⃗⃗⃗| = |𝑇
⃗⃗⃗⃗𝑥 | sin 𝛼 + |∆𝑝 ⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝑥+∆𝑥 |𝑠𝑖𝑛𝛽

For any values of x and x+∆x along the wire. Where

α and β are the angles between the vectors and the x
axis, and ⃗⃗⃗⃗⃗⃗
∆𝑝 is the difference in 𝑝⃗ between points x
and x+∆x, this important property is displayed clearly
in figure 7.
Figure 7 Free body diagram showing how 𝑝⃗ changes between 2Since the only parameter that changes between 2
points in the wire points in the wire is 𝑝⃗, the rate of change of 𝑝⃗ thus
directly determines the rate of change of f(x) and its
behavior was thus further investigated to conjecture f(x).

𝑝⃗ is the distributed force of gravity acting on any point on the curve and is given by:

1
The vector ⃗⃗⃗⃗⃗
𝑅𝑥 is not shown on figure 6 because its neglected as explained later in the paragraph

Page | 6
 𝑝⃗ = (𝜌𝑤𝑎𝑡𝑒𝑟 − 𝜌𝑤𝑖𝑟𝑒 )(𝑉𝑔⃗)
Where 𝜌 is density, V is the volume of the wire below the given point in the wire, and 𝑔⃗ is gravitational
acceleration. The function above shows that 𝑝⃗ is a multiple of 𝑔⃗ so the x component of 𝑝⃗ will always be 0
just like 𝑔⃗, so the function for 𝑝⃗ can be considered a scalar function. When treating the function for 𝑝⃗ as
scalar function it takes the form of a straight line function (y=mx):
|𝑝⃗|(𝑉) = |𝑔⃗|(𝜌𝑤𝑖𝑟𝑒 − 𝜌𝑤𝑎𝑟𝑡𝑒𝑟 )(𝑉)

Where V is the argument of the function and |𝑔⃗|(𝜌𝑤𝑎𝑟𝑡𝑒𝑟 − 𝜌𝑤𝑖𝑟𝑒 ) is the gradient of the function. Thus,
the rate of change of V directly determines the rate of change of 𝑝⃗ which determines the rate of change of
f(x). To use the volume of wire below any specific point in the wire to conjecture f(x) the volume of wire
below point x would have to be expressed in terms of x, which is impossible since volume of wire below
point x determines f(x) but f(x) determines the volume of wire below x. This makes the function f(x) an
insolvable loop in its current state.
Thus the function f(x) cannot be expressed as:

⃗⃗⃗⃗𝑥 | 𝑑𝑥 + 𝑐
𝑓(𝑥) = ∫|𝑇

However the function f(x) can be approximated by splitting the wire into n segments of length l that are
assumed to be straight. Then by calculating the height of each straight segment using basic trigonometric
techniques, and summing up the height of every segment of the wire an approximation of f(x) is created
(an approximation of the depth of the towfish). Since each segment has defined dimensions, the volume of
wire below any segment number c can be calculated as the product of the segment number and the volume
per segment.
Thus, the problem with the exact model of f(x) is solved using the approximation of f(x). To approximate
f(x) the same method can be applied but instead of considering the integral of the vector 𝑇 ⃗⃗⃗⃗𝑥 as the
expression for f(x), the expression for f(x) can be approximated by summation of the change in y value (the
height) of each segment of the wire. This is expressed mathematically as:
𝑛

𝑓 = ∑ ∆𝑦𝑖
𝑖=1

Where ∆y is the segment’s height, and n is the total number of segments. To find ∆y the properties of the
vector ⃗⃗⃗⃗
𝑇𝑐 can be applied. The vector ⃗⃗⃗⃗ ⃗⃗⃗⃗𝑥 but it is the tension vector tangential
𝑇𝑐 is the same as the vector 𝑇
to segment number c rather than point x. Thus ⃗⃗⃗⃗
𝑇𝑐 is:

𝑅⃗⃗
⃗⃗⃗⃗
𝑇𝑐 = − ( )
𝑝⃗ + ⃗⃗⃗⃗⃗
𝐹𝐵 + ⃗⃗⃗⃗
𝐹𝑔

If each segment of wire is straight and considered to be 1 unit long, since tension vectors in a wire are
always tangential to the wire; the tension vector at segment c is colinear with the segment number c. Thus
the height of any segment c is equal to the y component of the unit vector in the direction of ⃗⃗⃗⃗
𝑇𝑐 , which is
mathematically expressed as:

⃗⃗⃗⃗
𝑇𝑐
∆𝑦𝑐 = | | sin 𝛼
⃗⃗𝑐 |
|𝑇
since:

Page | 7
 ⃗⃗⃗⃗
𝑇𝑐
| |=1
⃗⃗𝑐 |
|𝑇

∆𝑦𝑐 = sin 𝛼

Where α is the angle between the vector ⃗⃗⃗⃗

𝑇𝑐 and the x axis. Thus, for a wire with one-unit long straight
segments
𝑛

𝑓 = ∑ sin 𝛼𝑖
𝑖=1

Where angle α is equal to the angle formed by the intersection of the vector ⃗⃗⃗⃗
𝑇𝑐 and a place holder vector
⃗
parallel to the x axis 𝑋:

⃗⃗⃗⃗
𝑋 ∙ ⃗⃗⃗⃗
𝑇𝑐
𝛼 = 𝑐𝑜𝑠 −1 ( )
|𝑋⃗||𝑇⃗⃗⃗⃗𝑐 |

Since 𝑋⃗ is parallel to the x axis:

1
𝑋⃗ = ( )
0
𝑥 𝑐𝑜𝑚𝑝𝑜𝑛𝑒𝑛𝑡 𝑜𝑓 ⃗⃗⃗⃗
𝑇𝑐
𝛼 = 𝑐𝑜𝑠 −1 ( )
⃗⃗⃗⃗𝑐 |
|𝑇
Thus:
𝑛
𝑥 𝑐𝑜𝑚𝑝𝑜𝑛𝑒𝑛𝑡 𝑜𝑓 ⃗⃗⃗⃗
𝑇𝑐
𝑓 = ∑ sin (𝑐𝑜𝑠 −1 ( ))
⃗⃗⃗⃗𝑐 |
|𝑇
𝑖=1
𝑖

Which can be simplified to:

𝑛 2
𝑥 𝑐𝑜𝑚𝑝𝑜𝑛𝑒𝑛𝑡 𝑜𝑓 ⃗⃗⃗⃗
𝑇𝑐
𝑓 = ∑ √1 − ( )
⃗⃗⃗⃗𝑐 |
|𝑇
𝑖=1
𝑖

by use of the following trigonometric simplification:

𝑠𝑖𝑛2 𝑥 + 𝑐𝑜𝑠 2 𝑥 = 1
Thus:

𝑠𝑖𝑛𝑥 = √1 − 𝑐𝑜𝑠 2 𝑥
Hence:

sin( 𝑐𝑜𝑠 −1 𝑥) = √1 − 𝑐𝑜𝑠 2 (𝑐𝑜𝑠 −1 𝑥) = √1 − 𝑥 2

Page | 8
Thus, the function that approximates the depth of the towfish tethered to a real wire in its fully expanded
form is:

2
𝑛 1
𝜌 𝐶 𝐴𝑣⃗ 2
𝑓 = ∑ √1 − 2 𝑤𝑎𝑡𝑒𝑟 𝐷
2
𝑖=1 √(1 𝜌𝑤𝑎𝑡𝑒𝑟 𝐶𝐷 𝐴𝑣⃗ 2 ) + (𝑔⃗(∆𝜌𝑖𝑉𝑛 + 𝜌𝑤𝑎𝑡𝑒𝑟 𝑉 − 𝑚))2
( 2 )

The function f presented above only

estimates the depth of the towfish tethered
to a real wire, when the wire is split into n
number of segments of length 1m. The
function f is limited to estimating the depth
of the towfish when the wire is split into 1-
meter long segments because the unit vector
of ⃗⃗⃗⃗
𝑇𝑐 was used to derive it.
In figure 8 each of the parameters have been
given a realistic value (for the investigation
it’s not of importance to mention the value)
and the argument of the function has been
set as the velocity of the towfish 𝑣⃗. Figure 8
shows that the function f just like D (the Figure 8 Plot of function f(v) showing the depth of the towfish when
function for the depth of the towfish tethered to a real wire plotted against its velocity
tethered to an ideal wire) takes the shape of
a bell function rapidly converging to 0, and just like D it is supported by the fact that the maximum depth
the towfish can reach is the same as the length of the wire. It can also be seen in figure 8 that the function
D and f behave similarly as the velocity of the towfish is varied, which further supports the validity of f and
D since they both model the same thing, but with different simplifications and assumptions, this pattern is
further highlighted in figure 9.

Figure 9 shows that the function f(𝑣⃗) and D(𝑣⃗) are almost identical when the real wire is very thin (0.001
mm), which is consistent with the assumptions used to derive the functions. The function D(𝑣⃗) models
the towfish depth with a neutrally buoyant straight wire, and when the real wire is very thin, the effect of
the distributed force of gravity acting on the wire will
be very low since the distributed force of gravity is
expressed by:

𝑝⃗ = 𝑔⃗(∆𝜌𝑖𝑉𝑛 )
And when the radius of the wire is small Vn will be
small thus the effect of the distributed force of
gravity will be small. Thus when the radius of f(𝑣⃗) is
very low d(𝑣⃗) is a good approximation of f(𝑣⃗).
However as seen in figure 10 when the radius of the
wire considered in f(𝑣⃗) is increased the discrepancy
between the f(𝑣⃗) and D(𝑣⃗) is greatly increased, yet
Figure 9 plot of function f(v) with a wire of radius 0.001 m the behavior of the function is not significantly
and D(v)
changed. Since the only difference between D(𝑣⃗) and
f(𝑣⃗) is the consideration of 𝑝⃗, it can interestingly
enough be concluded that f(𝑣⃗) is simply a translation of D(𝑣⃗). another interesting observation from figure
10 is that when the radius of the wire goes beyond 0.01 m the rate at which the function converges to 0

Page | 9
decreases significantly. Which indicates that varying the velocity of the towfish for large wires is not a valid
system of depth control as very large velocity would be needed.
Behavior of the function as the geometry of the
towfish is varied
Previously in the investigation the velocity of the
towfish has been set as the argument of the function,
and its behavior has been analyzed however varying
the geometry of a towfish is another parameter
which is vital to explore, as it too has a large effect
on the depth of the towfish. To investigate the
behavior of a towfish as the geometry is varied could
Figure 10 Plot of the function D(v) in blue and f(v) with be done for any towfish shape, however this
different wire radii (Radious 0.01m: Green line, Radious exploration only concerns itself with the teardrop
0.1m: Red line, Radious 1m: Grey line)
towfish design, as shown in figure 11, as it’s the most
modern design (Wikmark). The geometry of the
towfish has a large impact on the drag acting on the towfish as well as the buoyant force acting on the
towfish. Changing the geometry of the towfish has a large impact on the drag acting on the towfish because
one of the parameters of drag is A (the cross sectional area of the towfish). Changing the geometry also
impacts the buoyant force because one of the parameters of the buoyant force is V (volume of the towfish).
Since changing the geometry of the towfish changes
2 different parameters of the function f, these 2
parameters mut be expressed in terms of 1
parameter. This one parameter can be the parameter
r (radius of the towfish). But since the towfish
considered in this investigation is teardrop shaped,
which is an irregular shape both V and A cant be
simply expressed in terms of r since there is no
standard volume equation available for a teardrop Figure 11 Image showing teardrop shape towfish (Arnero)
shape. Given that the teardrop towfish is defined by
a function y(x) where y is the radius of the towfish at any point x, the parameters A and V can be defined
using the derivative and integral of y(x). The parameter A can be defined as:

𝐴 = 𝜋𝑦𝑚𝑎𝑥 2
Where ymax is the maximum radius of any point in the towfish. And V can be defined as the volume of
rotation of the function y(x):
𝑏
𝑉 = ∫ 𝜋𝑦 2 𝑑𝑥
𝑎

ymax is the y value of the maximum point of the function y(x), which can be found using the first derivative
test:

𝑑𝑦
𝑦𝑚𝑎𝑥 = 𝑦 ( = (0))
𝑑𝑥

Which can be expressed in terms of the inverse function of the derivative of y:

𝑑𝑥
𝑦𝑚𝑎𝑥 = 𝑦 ( (0))
𝑑𝑦

Page | 10
 𝑑𝑥
Where is the inverse function of the derivative of y. using the above expression for y, the function can
𝑑𝑦
f can be expressed using the function y that describes the shape of the towfish:
2
2
1 𝑑𝑥
𝑛 𝜌 𝐶 𝜋 (𝑦 ( (0))) 𝑣⃗ 2
2 𝑤𝑎𝑡𝑒𝑟 𝐷 𝑑𝑦
𝑓 =∑ 1−
2
𝑖=1 2
√(1 𝜌𝑤𝑎𝑡𝑒𝑟 𝐶𝐷 𝜋 (𝑦 (𝑑𝑥 (0))) 𝑣⃗ 2 ) + (𝑔⃗(∆𝜌𝑖𝑉𝑛 + 𝜌𝑤𝑎𝑡𝑒𝑟 ∫𝑏 𝜋𝑦 2 𝑑𝑥 − 𝑚))2
2 𝑑𝑦 𝑎
√ ( )

This version of the function shows how the radious of the towfish at any point x is related to the depth of
the towfish. As an example plot of this function for a varrying towfish radious, the function was set as:
1 𝑥+1
𝑦(𝑥) = √1 − 𝑥 2 (1 − )
2 2
Which is a funtion showing how the radious of a teardrop shaped towfish varries across its body. To plot
figure 12 showing how varrying the radious of the towfish changes its depth, the function y(x) was solved
for A and V so the function returned to its original form:

2
𝑛 1 2
𝑓 = ∑ √1 − 2 𝜌𝑤𝑎𝑡𝑒𝑟 𝐶𝐷 𝑥𝐴𝑣⃗
2
𝑖=1 √(1 𝜌𝑤𝑎𝑡𝑒𝑟 𝐶𝐷 𝑥𝐴𝑣⃗ 2 ) + (𝑔⃗(∆𝜌𝑖𝑉𝑛 + 𝜌𝑤𝑎𝑡𝑒𝑟 𝑥𝑉 − 𝑚))2
( 2 )

Where x represents a multiple of the radius of the towfish at any point.

In figure 12 all parameters other than x were set as
constant. Interestingly figure 13 shows that
varying the geometry of the towfish has a far
smaller effect on the depth of the towfish than
varying the velocity of the towfish. Also
interestingly it shows that when varying the
geometry of the towfish the function doesn’t
converge to 0 as it does when varying the velocity
of the towfish, the limit of f(x) as x approaches
infinity is 1.33, which shows that at some point
varying the geometry of the towfish has no effect
on the depth of the towfish.

Figure 12 Plot of f(x) showing how multiplying the radius at

any point of the towfish by x effects the depth of the towfish

Conclusion
The goal of this exploration was to investigate a function that models the depth of the towfish in response
to several parameters. Throughout this investigation, trigonometry and vector techniques have been used
to conjecture a function that models the depth of a towfish in response to varying its environmental
conditions. Throughout this exploration 2 functions were found which model the depth of the towfish.

Page | 11
First a function that estimated its depth when assuming that the wire connected to the towfish is straight
and rigid was found:

−𝑔⃗(𝑉𝜌 − 𝑚)
𝐷=𝑙
2
√(1 𝜌𝐶𝐷 𝐴𝑣⃗ 2 ) + (𝑔⃗(𝑉𝜌 − 𝑚))2
2
This function largely acted at as a stepping stone for further exploration however, interestingly it was later
found that this function was a very close estimate to the function that models the depth of the towfish
tethered to a non-rigid wire, which is significant as this function required far less computing power to plot
than the function for the non rigid wire. It was also found that when increasing the wires radius past 0.01
m the discrepancy between the function modeling the rigid and non rigid wire became so vast that using it
becomes inaccurate. To deal with this a function that models the towfish when tethered to an ideal wire
was found too:

2
𝑛 1 2
𝑓 = ∑ √1 − 2 𝜌𝑤𝑎𝑡𝑒𝑟 𝐶𝐷 𝐴𝑣⃗
2
𝑖=1 √(1 𝜌𝑤𝑎𝑡𝑒𝑟 𝐶𝐷 𝐴𝑣⃗ 2 ) + (𝑔⃗(∆𝜌𝑖𝑉𝑛 + 𝜌𝑤𝑎𝑡𝑒𝑟 𝑉 − 𝑚))2
( 2 )

However this function is only an estimate of the depth of the towfish when the wire tethered to the towfish
is considered as a collection of 1 meter rigid wire segments attached together rather than a long continuous
non rigid wire. This is the largest limitation of the function, its unknown how much the error of this
function is because there is no empirical data available, however the consulting engineer (Frank Wikmark
of marex consult) confirmed that the behavior of the functions derived in this investigation is consistent
with his experience.
Although this exploration isn’t fully comprehensive it provides a base function for future investigations
into advanced towfish techniques. Future explorations into this topic are encouraged to explore how
turbulence and complex water currents impact the behavior of the functions found in this exploration.
Although this exploration wasn’t as vast as originally anticipated, it managed to explore the basic concepts
and behaviors of a function modeling the depth of a towfish.

Citations
Allum, John, and Christopher Talbot. Physics For The Ib Diploma. 2nd ed., Hodder Education, 2014.

Allum, John, and Christopher Talbot. Physics For The Ib Diploma Option C: Engineering physics. 1st ed.,
Hodder Education, 2014.

Arnero. Teardrop Shape. 2007, https://commons.wikimedia.org/wiki/File:Teardrop_shape.svg. Accessed

22 Jan 2020.

Harcet, Josip. Mathematics Higher Level. Oxford University Press, 2012.

Svirin, Alex. "Equation Of Catenary". Math24, 2020, https://www.math24.net/equation-catenary/.

Accessed 6 Dec 2019.

Townsend, Gordon. Schematic Showing A Towfish System. 2015,

https://slideplayer.com/slide/4898657/. Accessed 13 Jan 2019.

Wall, Mike. "X-37B: The Air Force's Mysterious Space Plane". Space.Com, 2020,
https://www.space.com/25275-x37b-space-plane.html. Accessed 20 Dec 2019.

Wikmark, Frank. "Interview With Engineer from marex consult". Sandefjord, 2019.

Page | 12
Page | 13