Sei sulla pagina 1di 13

Modeling the velocity of a raindrop

Ben Lynch and Gavin Lommatsch


May 6, 2011

Abstract
Mathematical modeling of a falling raindrop and the derivation of
the motion of the raindrop as it is affected by air resistance. The non
spherical shape of the drop will be taken into consideration to produce
an accurate model

Introduction

As a raindrop falls, it forms a shape different from popular belief; the resulting shape from all of the forces causes the raindrop to form almost a perfect
sphere at diameters under 1.25mm and at larger diameters the sphere flattens out into an oblate spheroid which is somewhat similar in shape to that
of a hamburger bun. When a drop reaches diameters above 8mm it starts to
break up into smaller pieces therefore giving us a range up to 8mm for the
starting diameter of a raindrop. As a rain drop falls, it attains a constant
velocity where it experiences no acceleration. This is called terminal velocity,
when the forces of gravity are opposed equally by the resistive forces on the
raindrop.

Figure 1: d = sin(/x) + 2/Rt + gz + (pi )t (pa )t

1.1

Background

In this project, simplifying assumptions were made. One of these assumptions was that the raindrop fell from a cumulonimbus cloud at low altitude
and therefore the changes in air pressure, air density and gravity are ignored.
The raindrop is also falling with other drops and therefore the surrounding
humidity is high enough that the evaporation of the raindrop as it falls will
be negligible and therefore ignored. The raindrop is also not affected by any
extraneous forces such as gusts of wind or gaining additional mass from cohering with other raindrops.
3

With these assumptions, the main force acting on the drop is air resistance.
It is caused by air molecules colliding with the water molecules, friction between the air and water molecules as the air passes around the drop, and
by the low pressure wake behind the object. The main factor in all of these
cases is the shape of the object. A raindrop, however, has a complicated
shape to determine because of the previously mentioned hamburger shape.
The article Derivation of the Shape of Raindrops derived the shape of a drop
from the Laplace-Young equation. The derivation is fairly lengthy and since
the purpose of this project is to find the velocity of the raindrop, we will not
go through the derivation. We will use much of the data generated in this
article.
In fluid mechanics, a number known as the Reynolds number is used to give
a measure of the ratio of inertial forces to viscous forces and therefore subsequently quantifies the relative importance of these two types of forces for
given flow conditions. The Reynolds number is found from the equation
v0 d
with as the density of the fluid, a is the viscosity of the same
Re =
a
fluid, and v which is velocity. As one may see, the Reynolds number is dependent on the velocity at which the object is traveling. For low Reynolds
numbers (Re << 1) associated with a sphere, the fluid flows around the
sphere without separating. Therefore, the drag coefficient, which is found
empirically, is inversely proportional to the Reynolds number. For moderate
Reynolds numbers (103 < Re < 2 105 ), a separation is formed between the
fluid and the sphere causing a helical wake. The drag coefficient continues
to decrease with the increasing Reynolds numbers until at 103 it becomes a
constant (Cd = .3) up to 2 105 . The Reynolds numbers for raindrops falls
inside of this range, so we will use Cd = .3 as the drag coefficient.

Figure 2: Air flow around a sphere at different Reynolds numbers; (a)Re <<
1, (b) 103 < Re < 2 105 , (c) Re > 2 105

Equations

We will use Newtons Second Law to develop our equation. It states that
the net forces on a particle are equal to the time rate of change of its linear
momentum.
d
Fn = (mv)
dt
Therefore,
Fn = mv
Now we can begin to sum the forces on the left side of the equation.

The force due to gravity is defined as:


Fg = mg
The other force acting on the drop is air resistance which is defined by the
following equation:
1
F d = Cd a v 2 A
2
Cd = total drag coefficient.
a =Air density (at 300K, 1 bar)=1.161kg/m3
v = velocity
A = cross sectional area
Here is where things get a bit tricky. The area that is used in the equation
is the cross-sectional area, but due to the hamburger shape, the normal
equation for the volume of a sphere cant be used. Therefore, we had to
use the area, volume, and Reynolds number calculated by the article for the
diameters 1mm through 8mm.
Putting these equations together:
Fn = Fd Fg

1
mv = Cd a v 2 A mg
2
Dividing out m gives us our differential equation.
Cd a v 2 A
v =
g
2m

2.1

Solving the equations

In order to solve the equation, we will make a few substitutions to make


things a little easier to manage. First of all, the terms, Cd , a , and A are all
constants and therefore we will let r = 12 Cd a A
The equation then becomes
dv
r
= v2 g
dt
m
6

Factoring out -

Now let 2 =

r
,
m

mg
r

mg
r
dv
= (
v2) .
dt
r
m
r
and a = m .
dv
= a( 2 v 2 )
dt

Separating the variables and integrating both sides,


Z
Z
dv
= a dt.
2 v2
The left side is a partial fraction integral,
1
A
B
=
+
( + v)( v)
+v v
1
2
1
B=
2
A=

Therefore the integral becomes,






Z
1
1
1
1
+
dv = a dt.
2 + v
2 v
Evaluating the integrals,
1
(ln | + v| ln | v|) = a t + C.
2
Using properties of logarithms,
ln |

+v

| = 2at + C.
v

Raising both sides to the e,


+v

= e2at+C .
v
7

Simplifying the right side,


+v
= Ae2at .
v
To solve for v, let B equal the right side,
+v
=B
v
+ v = B Bv
v + Bv = B
From this,
(B 1)
.
B+1
Substituting everything back in gives the final equation,
v=

q
v(t) =

2
gm
(Ae
1
C A
2 d a
r

1 + Ae

g1
2 Cd a A t
m

1)

g1
2 Cd a A t
m

The limit as t is,


v(t) =

2.2

gm
1
C A
2 d a

Explanation

We then plugged the values of our constants into the equation and graphed
the solution. To do this we created a Matlab file that graphed the solutions
at diameters from 1mm to 8mm. We ran power regressions on the area and
volume versus time to create equations dependent on diameter for area and
mass.
A = 3.3108d2.21672
m = 1000 .957251 d3.09275
Cd = .3
a = 1.161
g = 9.81
8

d(mm)
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0

Re V (mm3 )
263
.510
863
4.238
1593 14.956
2267 36.822
3012 73.805
3625 129.603
4230 204.228
4834 315.783

A(mm2 )
.791
3.339
8.022
15.295
25.716
39.165
56.116
80.270

Figure 3: Cross-sectional Area vs. Diameter

Figure 4: Volume vs. Diameter


The resulting solution from the differential equation is an exponential
decay equation which means that as time progresses on, the velocity will
eventually become a constant. This constant velocity is the terminal velocity
of the drop. This is expected to happen because of the increasing drag force
with increasing velocity. We plugged in the appropriate constants to see if
our model is realistic and used the area and volume regressions to calculate
the area and mass for each diameter. We then graphed the solutions for the
diameters from 1mm to 8mm on one graph and observed the velocities over
time. As expected, the solutions accelerated rapidly for about 2 to 4 seconds,
then hit their terminal velocities. The terminal velocities increased with
diameter, but the increases were getting smaller. This suggests that there
may be a limit in the terminal velocities for large diameters, but according
to our model, it is well outside a reasonable size for a drop.

10

Figure 5: Velocity Vs. Time with diameters ranging from 1mm-8mm

Conclusions

With this model, we can predict the velocity of a raindrop with a given
diameter. The step would be to empirically test the velocity and compare it
with the models predictions. If the results didnt match the model accurately,
the model could be revised to take into account some of the assumptions that
were ruled out. One way to reduce the error could be to take into account the
evaporation of the drop as it falls. This would introduce a change in the mass
of the drop into the equation which would make it much more complicated,
but it would be an interesting problem to solve.

11

Figure 6: Terminal Velocity vs. Diameter

%Raindrop program
t = linspace(0,10,500);
v_t =[];
di =[];

for d =.001:.001:.008
di = [di;d];
c = .3; % drag coefficient
da = 1.161; %density of air
g = 9.81; % Gravitational acceleration
m = 1000*.957251*d.^3.09275; %Mass of the the drop
A = 3.3108*d.^2.21672; %Surface area of the drop
r = .5*.3*1.161*A; % Air resistance constants coefficient
v_t = [v_t;-sqrt((m*g)/r)];
v =
(sqrt((m*g)/r).*(exp(-2.*sqrt((g*r)/(m)).*t)-1))./(1+exp(-2*sqrt((g.*r)/(m)).*t))
plot(t,v,r)
hold on

12

grid on
pause
end
xlabel(time (seconds))
ylabel(velocity (m/s))
figure
plot(di,v_t,b)
grid on
xlabel(diameter (m))
ylabel(terminal velocity (m/s))

3.1

References

Brian Lim. Derivation of the Shape of Raindrops. Cornell University: School of Applied and Engineering Physics, 19 May
2006. PDF.
E. Reyssat, F. Chevy, A.-L. Biance, L. Petitjean, and D. Quere.
Shape and Instability of Free-falling Liquid Globules. A Letters Journal Exploring The Frontiers of Physics. EPL Journal,
2007. Web.
Chean Chin Ngo and Kurt Gramoll. FLUID MECHANICS THEORY. ECourses. Web. http://www.ecourses.ou.edu/cgibin/ebook.cgi?doc=.
John Polking, Albert Boggess, and David Arnold, Differential
Equations with Boundary Value Problems, Second Ed., Prentice Hall, New Jersey, 2006.

13

Potrebbero piacerti anche