2.2. Calculations
Classical methods of theoretical calculations of fluid mechanics were used as the basic research methods. The finite element method, a numerical method for solving partial differential equations of integral equations, was also used. The problems of hydrodynamics were solved by the system of Navier–Stokes equations.
When calculating the behaviour of a fluid in a region limited by the model (this region may or may not have inlets and outlets connecting it to a fluid outside the computational domain), the so-called fictitious regions method is used, i.e., formally, a computational grid is constructed in a parallelepiped region covering the model with the fluid inside [
9,
10]. Calculations are carried out only in cells that fall into the computational domain, that is, into a space filled in accordance with the statement of the problem by the fluid medium and solid (if it calculates heat transfer). In the cells outside the calculated area, calculations are not performed. This approach allows for the calculation of flows in very complex channels without complicating the algorithm for solving the problem.
A flat grid of profiles is a set of identical airfoils or plates periodically located in the plane (
Figure 3). In [
11], a brief theory of the grid of profiles, and a detailed theory of straight grids, both single and double, are presented.
All sources for the calculation of the Lenyov hydrobelt set forth a theory of potential currents, but even at a low velocity, the flow behind the grid is vortex.
In contrast to the flow around a single plate, in an infinite distance in front of and behind the grid of plates the speeds are generally different both in magnitude and in direction. The grid plate not only changes the velocity, but also turns it.
In
Figure 3, we denote the velocity of water far from the grid as ϑ
∞, then near the grid it will be ϑ
∞ − ϑ
1, and behind the grid ϑ
∞ − ϑ
2. Then, ϑ
1 and ϑ
2 are the lost velocities, i.e., ϑ
∞ − ϑ
1 is the flow velocity at infinity in front of the grid and ϑ
∞ − ϑ
2 is the flow velocity at infinite distance behind the grid. Let us consider the current tube in the plane of the drawing. The tube is formed by any two lines of current shifted relative to each other in the direction of the axis of the grid at the distance equal to the step t. The entire flow through the grid can be divided into such current tubes, since the flow has the property of periodicity with a period equal to a step (
Figure 4).
Let us apply the Euler form change momentum theorem [
12]:
where
is the main force of plate pressure on the flow;
ρ is the water flow density;
P is pressure-the force acting on a liquid in an arbitrarily chosen volume;
is introduced differential operator. For the current tube with side t and unit depth in the case of steady flow, we have
where
is difference in pressures in front and behind the plane of the grid.
Taking the selected current tube as the control surface and its cross sections at infinite σ1 and σ2 parallel to the axis of the grid and equal in length to the step.
The vector is directed perpendicular to the axis of the grid. Consequently, is the product of the projection of speed and direction perpendicular to the axis by step, and this is the volume flow of water in 1 second. Hence or .
Assuming the flow at infinity to be irrotational, the Bernoulli theorem can be applied [
12,
13]:
Replacing the squares of the moduli of velocities by the squares of the vectors of these velocities, we obtain
Let us introduce the average vector velocity
and the velocity of flow deviation, characterizing grid deviation
Equation (2) will be rewritten in the form:
In Equation (8), the expression on the right is the decomposition of the double vector product. Consequently,
In Equation (10), according to Equation (16),
⊥
and so
is perpendicular to the flow plane and
. Then
Equation (11) determines the magnitude of the vector of pressure force on the plate, and Equation (10) shows that the vector is perpendicular to the vector .
It is possible to write the vector
in the form similar to Zhukovsky formula for a single profile [
13]:
where
is unit normal vector to the plane;
is contour velocity circulation.
To do this, the unit vectors of the grid axis and the normals to the plane of the drawing should be introduced, directing them so that , , form a trihedron aligned with the right coordinate system. Noting that is parallel to the axis of the grid, we can derive , and therefore, .
To determine the velocity circulation along the contour composed of sections
σ1 and
σ2 and two current lines shifted by the step, the integral
should be taken in the positive direction shown in
Figure 3. Given that
σ1 =
σ2 =
t∙1, we get
since the integrals over current lines shifted by a step are equal in magnitude and opposite in direction. So,
In our opinion, this conclusion is not correct. The fact is that liquids and gases transfer pressure in all directions and, therefore, according to Equation (3), the force should be perpendicular not to the direction of flow, but to the plate. The disadvantage of this solution is also that it gives a paradoxical result: when the plate is installed perpendicular to the flow, according to (14) , the force R will be maximal. In fact, the force acting on the plate along the OY axis will be zero.
Even at low velocities, the flow is vortex behind the plate. For this reason, the model of potential fluid flow around the plate is applicable only when the angles are no more than 20°. In our case, the angles are close to 45°, which is proved by the calculations below.
The task of wrapping the plate at 90° to the flow was considered by G. Kirchhoff [
14]. Later, Reilly solved this task for any angle. Reilly’s formula for the force perpendicular to the plate is [
13,
15]:
where
is liquid density;
is the angle between the velocity vector at infinity in front of the plate and the plane of the plate. In this case, the force of interest that moves the plate in the direction perpendicular to
will be equal to
Analysis of Equation (16) shows that the maximum will be at the angle of 39–40°. Although in [
13] it is stated that Equation (15) at small angles gives the force
Fy 4 times lower than the theory of N.Ye. Zhukovsky. It is used below unchanged, since the division of the denominator by 4 does not change the dependence of
Fy on the angle
.
A similar result can be obtained in another way. In [
16], the problem of the cylinder vortex resistance was considered and the drag coefficient was obtained,
lx = 1.24. It is known that by a conformal mapping it is possible to map the area outside the plate to the area outside the circle with radius R (
Figure 5).
Map has the form of
where
and
are planes;
is linear fractional function (conformal mapping of the region outside the plate to the region outside the circle of radius R), i.e.,
Z is obtained from
ξ from the linear fractional substitution (transformation);
is fractional linear transformation.
Equation (17) displays a plate that is perpendicular to the flow. If instead of l we put lsin α, then the processing of Equation (17) maps the plate to the same circle.
Then and .
In this case, the angle between the plate and the flow, giving a maximum of Fy is 45°. According to Equation (16), the coefficient at is 0.25.
The next step is obtaining the ratio between the plate length and the grid spacing. Let’s consider the flow of water, taking into account the motion of the grid. The interaction of water with the grid leads to the fact that the water velocity varies near the grid. If we denote the water velocity away from the grid by
, then near the grid it is
where
is the lost velocity (
Figure 6).
Further, the grid moves upward with the speed
U, and the flow acquires the velocity
U1, directed downwards (
Figure 6). The water velocity relative to the plate will be
U +
U1, and the modulus of the relative velocity is determined by Equation (18):
The total force acting on the grid along the y axis is:
where
μ is the angle of attack of the relative velocity. The same force acts on water, which acquires the speed
U1 along the
y axis. The force
acts on the entire grid, which is equal to Equation (19), where
b is the plate length,
l is the plate width,
t is the grid spacing,
n is the number of plates. From here
where
Cy is determined from Equation (16)
. When obtaining
Cx = 0.5 at 45°, similarly we get
Using Equations (20) and (21), for
W we get
Substituting Equation (22) into (19), we can determine the power developed by the grid:
If divide this equation by the flow velocity, the coefficient of water energy use can be found:
Since then .
At α = 45°, . Then or .
After substituting
,
and
W from the Equation (20) and taking into account that
l/t < 1 and
< 1, we get:
Complexity of the dependence of
ξ from
and
l/
t makes it difficult to analyze (25). In this light the values of
ξ were calculated at
l/
t = 0.4; 0.6; 0.8; 0.9; 1.0 and at different values of
. Results of the calculation are presented in
Figure 7.
The dependences obtained in
Figure 7 allow to draw the following conclusions:
- −
the ratio of the width of the plate to the distance between the plates (grid step) must be equal to 0.9 (l/t = 0.8) to achieve a maximum ξ;
- −
maximum ξ is reached at the ratio of the grid speed to the flow velocity equal to = 0.4;
- −
the calculated maximum value ξ = 0.186, at the velocity of = 4 m/s with a square meter area, 400 watts of energy can be obtained.
In front of the grid, the flow velocity varies. The water is decelerated and its velocity becomes . Now it is necessary to analyze in detail how the water velocity changes after passing through the first and the second grids. After passing through the first grid, the water velocity will change, it will acquire the velocity and U2.
Before the second grid, the flow is decelerated again and its velocity will be
and U
2–U
3 (the “–” sign is because the second row plate is angled to the flow at 45°). After passing through the second grid, the velocity will change again. Below, we define these velocity changes (
Figure 8).
The force acting on the grid along the OY axis is: .
The same force acts on water and creates the velocity U1. Then it is equal to . Thus .
Since the force according to the Reilly Equation (26) is equal and directed perpendicular to the plate,
,
, and at 45°
.
The fluid is inhibited by the plate and the resistance force acts on the plate
On the other hand, this force creates the velocity and .
Hence at
, since
.
Now
U2 and
should be calculated. The formation of vortices can be taken into account by the so-called added mass [
16]:
where
S2 is plate area.
The velocity of vortex moving towards the flow is
/2. Then the flow velocity behind the plate is
and the force perceived by the grid is equal to
The mass of the flow involved in the interaction with the grid is
Then the energy conservation law can be written in the form:
where the expression on the left is the energy of the flow incident on the grid, the first term on the right is the energy perceived by the grid, the second is the energy carried away by the flow, the last term corresponds to the energy consumed for the formation of vortices.
Taking into account that
, we get
Approximately, we can assume that , considering the low value of .
On the other hand, the impulse
lost by water in 1 second must be equal to the force acting on water
,
is the area through which water passes to the plate (b = 1), where
μ is the angle of attack of relative velocity. Since
μ = 45 −
β (
Figure 9), then
Considering that
, we can assume that
. Then at
Cx = 0.5
Similarly, in the projections on the axis OY we get
and
Considering that
, Equations (35) and (36), we obtain a biquadratic equation for
W, the solution of which is
Then .
Since
, and
, then
The flow past the first grid falls on the second grid. The second grid slows down the incident flow and gives it the velocity along the OY axis. When calculating the velocities
and
W2, it is necessary to take into account that in the Lenyov apparatus the second grid moves in the opposite direction. The flow falling on the second grid has absolute velocities
and U
2 (
Figure 10). Therefore, the angle of attack of both absolute and relative velocities will change.
The absolute velocity
is the sum of
and
. The angle that it forms with the plate is the angle of attack for the second plate. From
Figure 10 it can be seen that
. In the formulas of
Cx and
Cy sin
α and cos
α appears. We determined these functions using the results of calculations for the first grid.
By the formulas of trigonometry we have .
Since ; .
To calculate
and
it is necessary to get
and
, as well as
. According to
Figure 4, this velocity is equal to
The relative flow velocity for the second grid is
and
play the roles of
and
, and
and
play the roles of
and
. Therefore, reasoning similarly, we can write
Substituting
and
into Equation (40) we obtain the biquadratic equation, the solution of which is
Now it is possible to determine the coefficient of water energy use for the second grid:
To determine
, it is necessary to calculate
from
Since W2 has been calculated, determination of and is not difficult: , . Substituting them into Equation (41), we get .
It remains to determine . Substituting in Equation (43) the incoming values, we get .
Therefore, the whole Lenyov apparatus can give only 0.316, i.e., one third of the flow energy:
Thus, based on the calculated model, we can conclude that:
- −
after passing through the Lenyov unit, the water velocity slows down and is about 0.8 of the velocity before the unit. The velocity behind the unit can increase significantly, in the conditions of elevation difference, according to the formula at h = 0.5 m and = 1 m/s, the velocity = 3.28 m/s without viscous friction;
- −
from one square meter of the grid at = 1 m/s, one can get no more than 200 W. The flow power is . Net power is Nnet = 500 × 0.32 = 160 W;
- −
to obtain a net power of 16 kW, the water flow velocity must be at least = 4.5 m/s.
The calculated data indicate the need to modify the structure with regard to the derived optimal parameters of the distance between the grids, the angle of inclination, etc.