OR-16-012
Investigating the Possibility of Using a
Tesla Turbine as a Drive Unit for an
Automotive Air-Conditioning Compressor
Using CFD Modeling
Ali M. Hasan, CEng
Member ASHRAE
ABSTRACT
This paper investigates the possibility of using a Tesla
turbine as an alternative air-conditioning compressor drive
unit in vehicles as an example rather than the traditional
method of linking the compressor rotor to the vehicle engine
crankshaft belt or chain. Interest in Tesla turbines has recently
gained momentum because of the possibility of obtaining high
efficiency in energy conversion, as claimed by the inventor
Nikola Tesla. This paper investigates the possibility of driving
the air-conditioning compressor using the hydraulic energy
available in the engine lubrication oil piping provided by the
engine oil pump. The engine oil pressure can be used to drive
the Tesla turbine driving the compressor. Tesla turbines, also
known as bladeless or boundary layer turbines, were invented
in the early twentieth century but were never commercially
exploited. Tesla turbines are known for their ease of construc-
tion and low erosion rate, because the rotating parts experi-
ence no direct impact from fluid-borne abrasive matter.
For this research, a software package was used to create
the model. Computational fluid dynamics (CFD) simulations
were conducted by varying turbine inlet fluid flow rate, turbine
inter-disk spacing, and turbine disk outlet opening diameter.
The results are shown in tables and graphs presented in
this paper. Selected graphical presentation showing fluid
streamlines and the turbine/oil flow physics are included.
Results are discussed for the turbine suitability as a drive unit
for the air-conditioning compressor.
This investigation shows that Tesla turbines are not suit-
able for automotive air-conditioning compressors when
powered by relatively low-pressure engine oil flow rate
supplied by the engine oil pump. This investigation also shows
that the performance of such machines can be influenced by a
Ali M. Hasan is a senior engineer at KEO International Consulting Engineers, Doha, Qatar.
146
Published in ASHRAE Transactions, Volume 122, Part 1
number of factors, such as inter-disk spacing, fluid viscosity,
and turbine disk opening diameter. Efficiency of Tesla turbine
drives can be enhanced by harmoniously matching these
factors.
Tesla turbines are easy to manufacture and are suitable
for energy recovery where high mass flow rate is available for
using as turbine driving fluid. Tesla turbines are easy to adapt
and can cover a range of flow rate capacities. Using such a
system for energy conversion can lower capital equipment
costs and improve the turbine lifespan, contributing to sustain-
able engineering development.
INTRODUCTION
Tesla turbines, invented and patented by Nikola Tesla in
the early twentieth century, have seen little progress. However,
this invention is beginning to attract interest because of its
construction simplicity and suitability in handling various
types of fluids. Researchers are now investigating ways to
improve the turbine efficiency and establish its limits.
Research continues in establishing equations to describe the
fluid mechanics of these turbines. This paper investigates the
possibility of using such drive units in typical engine operating
conditions. A combination of different factors—inter-disk
spacing, disk openings, and varying fluid inlet flow rates—
were considered. Detailed, close-up graphical images were
produced, aimed at giving a better understanding of how Tesla
turbines (also known as boundary layer or bladeless turbines)
work. These graphical presentations clarify how energy from
the moving fluid body is transferred into rotational mechanical
energy.
A computational fluid dynamics (CFD) model was used
to investigate how the performance of such drives can be
© 2016 ASHRAE
enhanced with minimum change to current engine designs.
The turbine inlet fluid flow rates were selected to fall within
typical automotive engines’ lubrication oil output. The reason
for this approach is to minimize introducing design changes to
existing engine oil pump circulation systems. This paper
refers to similar technical papers and books on this subject,
which are based on using fluids such as air, flue gas, and water
as the turbine energizing fluid. Viscous fluid or engine oil was
used as the turbine driving fluid in this example. The turbine
in this paper was investigated for possible application as a
drive unit for an automotive air-conditioning compressor.
This paper depends largely on graphical presentation with
the aim of providing detailed description of the theories
behind such turbines, how efficiency can be improved, and the
suitability of the drive unit as an automotive air-conditioning
compressor.
Tesla turbines, known for their construction simplicity,
can provide an acceptable level of energy conversion. Tesla
turbines are relatively robust and easy to maintain, with just
one moving part within the turbine housing. Tesla turbines can
be used in areas where fluids containing abrasive or solid parti-
cles are encountered. For an example, engine solid deposits
that may exist in engine lubrication oil.
TESLA TURBINE
Historical background on Tesla turbines can be seen in an
article titled “From the Complex to the Simple,” published in
1911 by Scientific American:
A marked step was taken in the simplification of prime
movers when Watt's cumbersome beam engine, with its
ingenious but elaborate parallel motion, gave way to the
present standard reciprocating type, with only piston rod,
cross head and connecting rod interposed between piston
and crank. An even greater advance toward ideal
simplicity occurred when, after years of effort by inven-
tors to produce a practical rotary, Parsons brought out his
compact, though costly, turbine, in which the energy of
the steam is developed on a zig zag path through multi-
tudinous rows of fixed and moving blades.
And now comes Mr. Tesla with a motor which bids fair
to carry the steam engine another long step toward the
ideally simple prime mover - a motor in which the fixed
and revolving blades of the turbine give place to a set of
steel disks of simple and cheap construction. If the flow
of steam in spiral curves between the adjoining faces of
flat disks is an efficient method of developing the energy
of the steam, the prime mover would certainly appear to
have been at last reduced to its simplest terms.
The further development of the unique turbine which we
describe elsewhere will be followed with close attention
by the technical world. The results attained with this
small high-pressure unit are certainly flattering, and give
reason to believe that the addition of a low pressure
turbine and a condenser would make this type of turbine
as highly efficient as it is simple and cheap in construc-
tion and maintenance.
ASHRAE Transactions
Published in ASHRAE Transactions, Volume 122, Part 1
Tesla turbines are typically made of a number of disks
mounted on a shaft. Disks are equally spaced and have a
number of holes around the shaft. As fluid enters the turbine
casing and flows along the disk surfaces, the disks begin to
rotate. The incoming fluid stream is concentrated toward the
outer portion of the disk surfaces, away from the disk center,
providing turning movements. Fluid viscous forces act along
the disks surfaces and cause the disks to rotate. Figure 1 shows
the investigated model, a three-inlet ports model. According to
Lampart and Jedrzejewski (2010), the three inlet ports gave the
highest efficiency results. Efficiency also depends on other
factors such as disk surface finish and disk thickness. Typical
applications include energy recovery of exhaust gas and wind
turbines (Podergajs 2011).
Tesla turbines are generally simple to construct and the
amount of machining works are relatively low when compared
with other turbines. Centrifugal turbines, for example, have
more complex geometry. Hobbyists interested in Tesla
turbines are known to have constructed such a turbine using
computer CDs. Nam (2012) states: “Despite its inefficiencies,
the Tesla turbine still has several advantages. The production
costs for disks are much lower than for blades, and the overall
design is simpler and more easily manufactured.” See the work
by Rice (2003) on its inefficiencies.
The Tesla turbine uses the drag effect caused by fluids
moving along a surface are influenced by the boundary layer
conditions. Table 1 shows how varying fluid conditions can
impact the boundary layer thickness. Details of calculations
are shown in the Results and Calculations section.
COMPUTATIONAL FLUID DYNAMICS ANALYSIS
The turbine was modeled as shown in Figure 1. Fluent®
software was used with the following assumptions:
• Engine oil fluid was selected from the Fluent software
database.
• Fluid temperatures were assumed at 100°C (212°F),
near engine operating oil temperatures. See the Discus-
sions section for this temperature selection.
• Aluminum selected from the software database was cho-
sen as the disk and shaft material. Note: Material tem-
perature properties are assumed acceptable for the
purpose of simulation, and are not the subject of discus-
sion of this paper.
• From the Fluent menu, the energy option “on”; k-omega
(a common turbulence model); and the aluminum mate-
rials wall roughness constant, 0.5, were selected.
• Convergence was set for 1000 iterations.
• Shaft speed and torque results were determined using
the CFD postprocessor.
• Suitable meshing was used to build a sufficient layer of
fluid mesh to fill the inter-disk spacing and obtain a good
approximation of boundary conditions. In this example,
41,849,086 cells were used to mesh the model, shown in
Figure 1, with an inter-disk spacing of 3 mm (0.12 in.).
147
 Table 1. Calculating Boundary Layer Thickness, , for the Purpose of Optimizing Inter-Disk Spacing

v,  Laminar,  Turbulent, x, , ,
Re
m/s (fps) mm (in.) mm (in.) m (ft) kg/m3 (lb/ft3) Pa·s (centipoise)

5 2.00E+00 3.53E+00 0.075 940 1.00E–02

3.53E+04
(16.4) (7.99E–02) (1.41E–01) (0.246) (58.68) (1.00E+01)

10 1.41E+00 3.07E+00 0.075 940 1.00E–02

7.05E+04
(32.8) (5.65E–02) (1.23E–01) (0.246) (58.68) (1.00E+01)

15 1.15E+00 2.83E+00 0.075 940 1.00E–02

1.06E+05
(49.2) (4.61E–02) (1.13E–01) (0.246) (58.68) (1.00E+01)

20 9.99E–01 2.67E+00 0.075 940 1.00E–02

1.41E+05
(65.6) (3.99E–02) (1.07E–01) (0.246) (58.68) (1.00E+01)

25 8.93E–01 2.56E+00 0.075 940 1.00E–02

1.76E+05
(82) (3.57E–02) (1.02E–01) (0.246) (58.68) (1.00E+01)

30 8.15E–01 2.47E+00 0.075 940 1.00E–02

2.12E+05
(98.4) (3.26E–02) (9.87E–02) (0.246) (58.68) (1.00E+01)

40 7.06E–01 2.33E+00 0.075 940 1.00E–02

2.82E+05
(131.2) (2.82E–02) (9.31E–02) (0.246) (58.68) (1.00E+01)

50 6.32E–01 2.23E+00 0.075 940 1.00E–02

3.53E+05
(164) (2.53E–02) (8.91E–02) (0.246) (58.68) (1.00E+01)

60 5.77E–01 2.15E+00 0.075 940 1.00E–02

4.23E+05
(196.8) (2.31E–02) (8.59E–02) (0.246) (58.68) (1.00E+01)

70 5.34E–01 2.08E+00 0.075 940 1.00E–02

4.94E+05
(229.6) (2.14E–02) (8.33E–02) (0.246) (58.68) (1.00E+01)

80 4.99E–01 2.03E+00 0.075 940 1.00E–02

5.64E+05
(262.4) (2.00E–02) (8.11E–02) (0.246) (58.68) (1.00E+01)

100 4.47E–01 1.94E+00 0.075 940 1.00E–02

7.05E+05
(328) (1.79E–02) (7.75E–02) (0.246) (58.68) (1.00E+01)

• Initially, one model was simulated with 3 disks of 300 Assessing Boundary Layer Thickness—Theoretical
mm (12 in.) diameter. A mass flow rate per a port of The following equations were used to produce data shown
1.12E–4 m3/s (39.52 ft3/s) was initially assumed, as in Table 1. The boundary layer thickness for laminar flow can
shown in Table 2. Simulations then continued as shown in be calculated by the following equation developed by Blasius
Tables 2, 3, and 4, varying the mass flow rate flowing (1908):
 = --------------------
through inlets. Disk geometry was changed as indicated 5.0  x -
(1)
and explained in Figures 2 and 3. This was carried out to Re
where
probe for the optimum performance.
   v  x 
• Outlet pressure was assumed at standard atmospheric Re = ----------------------------- (2)

pressure, 1 bar at 100°C (212°F). Rearrange Equation 1 as shown in Equation 3 to obtain
the Re value. This provides an estimate of the oil flow rate
RESULTS AND CALCULATIONS value and thus provides an estimate of the velocity component,
v , given in Equation 2:
This section is divided into two parts. The first part shows  5.0  x 
Re = --------------------- (3)
how the boundary layer thickness along a flat surface was esti- 
mated using established equations. The second part shows how Similarly, for turbulent flow:
the CFD-generated results were used to calculate the turbine  0.382  x -
 = -------------------------- (4)
1/5
efficiency. Figures 4 to 7 show turbine graphical results. Re

148 ASHRAE Transactions

Published in ASHRAE Transactions, Volume 122, Part 1
Figure 1 Tesla turbine model 3D image with three inlet ports. The three turbine disks, with outlet holes evenly spaced around
the shaft, can be seen in the darker-colored shading. The disk casing and inlet ports are shown with a lighter-colored
shade.

Table 2. Three-Disk Turbine with 30 mm (1.2 in.) Diameter Disk Openings; Inter-Disk Spacing = 2 mm (0.08 in.)

m· per
P1, P2, P3, Speed, T, Shaft Power, Fluid Power, ,
inlet,
Pa (lb/in2) Pa (lb/in2) Pa (lb/in2) rad/s N·m (lb·ft) W W %
kg/s (lb/s)

0.1 9.10E+00 9.10E+00 1.06E+01 2.41E–04

3.20E+00 7.70E–04 2.85E+00 0.03
(0.22) (1.32E–03) (1.32E–03) (1.54E–03) (1.78E–04)
0.2 1.93E+01 1.90E+01 2.01E+01 4.58E–05
3.93E+00 1.80E–04 1.16E+01 0.00
(0.44) (2.80E–03) (2.76E–03) (2.92E–03) (3.38E–05)

0.28 2.65E+01 2.65E+01 2.92E+01 2.31E–04

6.00E+00 1.39E–03 2.28E+01 0.01
(0.616) (3.84E–03) (3.84E–03) (4.24E–03) (1.70E–04)

0.3 2.76E+01 2.80E+01 3.30E+01 1.27E–03

8.00E+00 1.01E–02 2.63E+01 0.04
(0.66) (4.00E–03) (4.06E–03) (4.79E–03) (9.36E–04)

0.33 3.07E+01 3.08E+01 3.54E+01 1.00E+00

7.73E+00 7.74E+00 3.17E+01 24.43
(0.726) (4.45E–03) (4.47E–03) (5.13E–03) (7.38E–01)

0.35 3.27E+01 3.26E+01 3.61E+01 1.00E+00

9.00E+00 9.01E+00 3.52E+01 25.62
(0.77) (4.74E–03) (4.73E–03) (5.24E–03) (7.38E–01)

0.4 3.80E+00 3.77E+01 4.21E+01 5.74E–04

8.87E+00 5.09E–03 3.31E+01 0.02
(0.88) (5.51E–04) (5.47E–03) (6.11E–03) (4.23E–04)

0.5 1.85E+01 1.87E+01 2.18E+01 3.15E–04

6.93E+00 2.19E–03 1.17E+01 0.02
(1.1) (2.68E–03) (2.71E–03) (3.16E–03) (2.33E–04)
Refer to Figure 2 for disks openings.
Graph of efficiency against mass flow rate per an inlet is shown in Figure 11, top graph.

ASHRAE Transactions 149

Published in ASHRAE Transactions, Volume 122, Part 1
Table 3. Three-Disk Turbine with 35 mm (1.4 in.) Diameter Disk Openings; Inter-Disk Spacing = 2 mm (0.08 in.)

m· per
P1, P2, P3, Speed, T, Shaft Power, Fluid Power, ,
inlet,
Pa (lb/in2) Pa (lb/in2) Pa (lb/in2) rad/s N·m (lb·ft) W W %
kg/s (lb/s)

0.09 1.77E+01 1.78E+01 1.92E+01 1.02E–04

3.80E+00 3.86E–04 4.88E+00 0.008
(0.198) (2.57E–03) (2.58E–03) (2.78E–03) (7.49E–05)

0.1 1.97E+01 1.98E+01 2.14E+01 9.78E–05

4.27E+00 4.17E–04 6.04E+00 0.007
(0.22) (2.86E–03) (2.87E–03) (3.10E–03) (7.22E–05)

0.105 2.07E+01 2.09E+01 2.25E+01 6.09E–05

4.33E+00 2.64E–04 6.67E+00 0.004
(0.231) (3.00E–03) (3.03E–03) (3.26E–03) (4.49E–05)

0.11 2.08E+01 2.15E+01 2.56E+01 4.29E–02

5.40E+00 2.32E–01 7.40E+00 3.128
(0.242) (3.02E–03) (3.12E–03) (3.71E–03) (3.16E–02)

0.13 2.50E+01 2.56E+01 2.98E+01 3.18E–04

6.00E+00 1.91E–03 1.04E+01 0.018
(0.286) (3.63E–03) (3.71E–03) (4.32E–03) (2.35E–04)

0.15 3.00E+01 3.01E+01 3.24E+01 5.81E–04

6.00E+00 3.49E–03 1.38E+01 0.025
(0.33) (4.35E–03) (4.37E–03) (4.70E–03) (4.29E–04)

0.17 3.36E+01 3.40E+01 3.72E+01 1.11E–04

7.33E+00 8.16E–04 1.77E+01 0.005
(0.374) (4.87E–03) (4.93E–03) (5.40E–03) (8.20E–05)

0.2 4.04E+01 4.05E+01 4.28E+01 1.14E–04

7.40E+00 8.44E–04 2.45E+01 0.003
(0.44) (5.86E–03) (5.87E–03) (6.21E–03) (8.41E–05)
Refer to Figure 2 on disks openings.
Graph of efficiency against mass flow rate per inlet is shown in Figure 11, middle graph.

Table 4. Three-Disk Turbine with 35 mm (1.4 in.) Diameter Disk Openings; Inter-Disk Spacing = 3 mm (0.12 in.)

m· per
P1, P2, P3, Speed, T, Shaft Power, Fluid Power, ,
inlet,
Pa (lb/in2) Pa (lb/in2) Pa (lb/in2) rad/s N·m (lb·ft) W W %
kg/s (lb/s)

0.01 1.00E+00 1.00E+00 1.10E+00 1.33E–05

2.93E–01 3.90E–06 3.07E–02 0.0127
(0.022) (1.45E–04) (1.45E–04) (1.60E–04) (9.80E–06)

0.02 2.02E+00 2.03E+00 2.10E+00 2.31E–05

6.00E–01 1.39E–05 1.22E–01 0.0114
(0.044) (2.93E–04) (2.94E–04) (3.05E–04) (1.71E–05)

0.03 3.10E+00 3.10E+00 3.10E+00 1.12E–06

8.00E–01 8.96E–07 2.76E–01 0.0003
(0.066) (4.50E–04) (4.50E–04) (4.50E–04) (8.26E–07)

0.04 3.70E+00 4.10E+00 6.00E+00 5.94E–04

5.67E+00 3.36E–03 5.47E–01 0.6150
(0.088) (5.37E–04) (5.95E–04) (8.70E–04) (4.38E–04)

5.10E+00 5.10E+00 5.20E+00 3.20E–07

0.05 (0.11) 1.40E+00 4.48E–07 7.63E–01 0.0001
(7.40E–04) (7.40E–04) (7.54E–04) (2.36E–07)

1.02E+01 1.02E+01 1.03E+01 8.89E–06

0.1 (0.22) 2.73E+00 2.43E–05 3.04E+00 0.0008
(1.48E–03) (1.48E–03) (1.49E–03) (6.55E–06)

2.08E+01 2.09E+01 2.25E+01 1.05E–04

0.2 (0.44) 5.20E+00 5.48E–04 1.27E+01 0.0043
(3.02E–03) (3.03E–03) (3.26E–03) (7.77E–05)

3.08E+01 3.08E+01 3.10E+01 4.81E–05

0.3 (0.66) 8.00E+00 3.85E–04 2.75E+01 0.0014
(4.47E–03) (4.47E–03) (4.50E–03) (3.55E–05)
Refer to Figure 2 on disks openings.
Graph of efficiency against mass flow rate per inlet is shown in Figure 11, bottom graph.

150 ASHRAE Transactions

Published in ASHRAE Transactions, Volume 122, Part 1
 (a) (b)

Figure 2 Example of a single port inlet model: (a) side view of turbine with dimensions as shown and (b) front view of turbine
with dimensions. Note: A single port inlet was selected for clarity of description.
Dimensions:
1 Inlet face projects 160 mm (6.4 in.) from the disk’s central y-axis.
2 Disk circular hole diameter is 30 mm (1.2 in.) with center of hold 30 mm (1.2 in.) from shaft surface.
3 The disk is 300 mm (12 in.) in diameter and 2 mm (1.8 in.) thick. The inter-disk spacing is 2.5 mm (0.1 in.), then
adjusted to 3.0 mm (0.12 in.). See Tables 2 to 4 for details.
4 The outer casing diameter is 320 mm (12.8 in.).
5 The inlet port face dimensions are 10 by 35 mm (0.4 by 1.4 in.).
6 The diameter of the outlets on both sides is 130 mm (5.2 in.) The face extends 10 mm (0.4 in.) from the wall casing.
7 The shaft diameter is 20 mm (0.8 in.).
8 The casing width varies, with the case wall to disk surface spacing equal to inter-disk spacing. A 2 mm (0.08 in.)
clear spacing from the disk surface was assumed.

Rearranging Equation 4 to obtain Re yields the following:
Re 1 / 5 = --------------------------
0.382  x -

It is clear from Equation 1 that  depends on the Re value
and that, as , x, v, and  in Equation 2 change, the Re values
change and subsequently influence the value of . The param-
eters on the right-hand side of Equation 2 (, x, v, and )
depend on fluid temperature. Table 1 shows how the lubrica-
tion oil velocity effects the boundary layer thickness.
Table 1 was used as an initial guide to assess what fluid
velocities or flow rates can achieve a boundary layer thickness
equal to the inter-disk spacing, or arrange the inter-disk spac-
ing such that the inter-disk-spacing equals the thickness of the
boundary layer.
The following equations were used to develop Tables 2 to 4.
Equations 6 to 8 were used to calculate shaft power, fluid
power applied across the turbine, and turbine efficiency.
Shaft power was calculated by multiplying turbine torque
(T) by rad/s. The results in W were tabulated in Tables 2 to 4.
(5) Shaft power = T × rad/s (6)
Fluid power transferred to turbine was calculated by
multiplying fluid density by gravity by pressure by flow rate.
Results are in W.
Fluid power =  × g × H × Q (7)
According to Massey (1983), for a turbine,  × g × H × Q
represents the energy given up by the fluid in a unit of time
passing through the machine.
Turbine efficiency was calculated by dividing power
output by power input and then multiplying the product by 100
to obtain a percentage efficiency.
 or Efficiency = [(T × rad/s) ÷ ( × g × H × Q)] × 100 (8)
Tables 2 to 4 summarize the results obtained for turbine
efficiency. Results were plotted as shown in Figure 11.

ASHRAE Transactions 151

Published in ASHRAE Transactions, Volume 122, Part 1
Figure 3 Model with various inlet ports, side views shown. Disks were set to rotate freely about the z-axis as the inlet flow rate
and temperature changes turbine performance were observed. (a) Two equally spaced inlet ports. (b) Three equally
spaced inlet ports. (c) Four equally spaced inlet ports. Note: In this paper, a three-port model was considered.

The following example shows how the first row of Table 2
was obtained:
Shaft power = 2.41E–4 × 3.2 = 7.7E–4 W
Fluid power = 899 × 9.81 × 899 ×
[10*(9.1+9.1+1.06e1)/100,000) × 1.12E–04] = 2.85 W
Efficiency = 7.7E–4 ÷ 7.7E–4 = 0.03%
Table 2 shows peak efficiency is 25.62%, at a flow rate of
0.35 kg/s (0.77 lb/s) per inlet.
Assessing the Suitability of the Turbine-Generated
Power in Operating the Air-Conditioning
Compressor
Typical direct-expansion air-conditioning systems have a
coefficient of performance (COP) value of 3.5; this can vary
depending on manufacturer and type of compressor.
COP = Cooling capacity output ÷ Power input (9)
Entering the highest achieved turbine shaft power, found
in the sixth row from the top in Table 2, the shaft power equals
9 W, and inputting the COP value of 3.5 into Equation 9 gives
the cooling capacity output of 31.5 W. This figure is too low
to provide cooling power for a typical-size passenger car.
DISCUSSION
Engine lubricant oil temperatures and pressure can vary
and depends on type of engine. Mehlhoff (2005) describes that
the oil pressure in most engines should be about 0.68 bar (10
psi) per every 1000 revolutions per minute, peaking around
374 to 442 kPa (55 to 65 psi).
According to ASHRAE Handbook—Systems and Equip-
ment (2008), high-quality lubricating oils are generally

152 ASHRAE Transactions

Published in ASHRAE Transactions, Volume 122, Part 1
Figure 4 Postprocessor graphics output for a 100°C
(212°F) inlet lubrication oil temperature with a
flow rate of 0.13 kg/s (0.29 lb/s) per inlet port. The
disk rotates counterclockwise under the action of
the inflowing lubrication oil. Top: Velocity
streamlines starting from the center as the lubri-
cation oil enters the turbine casing and then
through the disk openings and along the disk
surfaces. Lubrication oil moves away from inlets
radially toward the center outlet. The outlet
shows how the pressure is concentrated. Bottom:
Pressure volume rendering image showing rela-
tively high pressure near the inlets, and the pres-
sure dropping toward the outlet. This pressure
drop represents the transfer of energy in the form
of oil pressure and heat to mechanical rotational
energy. See Figure 6 for changes in oil tempera-
ture.
required for operation between 70°C and 93°C (158°F and
199.4°F), with longer oil life expected at lower temperatures.
Zhou (2009) indicates that, for typical engine oil pressure
requirements, the oil pump is sized to meet the low speed and
hot oil conditions. Zhou gives an example where the engine oil
pump had to deliver 1.0 bar oil pressure at 1000 rpm. The oil
pressure level had to be controlled by a PRV set to limit the
engine oil supply at approximately 2.7 bars at 7000 rpm. The
oil pump flow rate and pressure cover the engine lubrication
oil needs across a range of engine speeds and oil lubrication
temperatures.
ASHRAE Transactions
Published in ASHRAE Transactions, Volume 122, Part 1
Figure 5 Postprocessor graphics output for a 100°C
(212°F) lubrication oil inlet temperature with a
flow rate of 0.01 kg/s (0.22 lb/s) per an inlet port.
Top: Temperature rendering model with one
outlet’s streamlines selected for clarity. The disk
rotates counterclockwise under the action of the
oil flow. Bottom: A sectional view of the top image
along the y-axis. The relatively lower oil
temperature near the outlet is clearly visible. At
this point, the oil has passed some of its energy to
the turbine in the form of pressure and heat. This
transfer of energy produces rotational mechani-
cal energy.
According to Miranda (2014), typical bulk engine
temperature is approximately 100°C (212°F) for street-driven
vehicles. Bulk operating temperature can vary depending on a
number of factors; engine construction/condition, how it is
operated, and weather conditions. The construction of the
vehicle engine cooling system, air flow over the engine, and
the oil system design can affect bulk oil engine temperature.
The emphasis is on bulk oil temperature because there are
usually locations within the engine that can be cooler or hotter
than the bulk temperature.
Based on the above published information, the following
inputs for the CFD Tesla turbine simulation were assumed:
engine oil temperature at typical operating engine conditions
of 100°C (212°F) and oil pressure levels applied at the turbine
inlet kept within the abovementioned pressure levels, 20 to
100 kPa (2.94 to 14.7 psi). These inlet pressure levels can be
153
Figure 6 A close-up view of a 0.13 kg/s (0.29 lb/s) lubrication oil flow per port at 100°C (212°F) supplied through the turbine
port. The two images show how a Tesla-type turbine works, how the fluid’s viscosity physics are used, and boundary
conditions. The turbine casing texture is switched off and the disk’s image presented for clarity. The velocity stream-
lines indicate the path of the lubrication oil as it enters the inlet ports, moves along the disks’ surfaces, and then
out of the disks’ holes. The image to the left shows three disks. The image to the right shows a cut-through sectional
view of the image to the left around the XY plane showing how the fluid streamlines move along the disk’s surface
(background disk in left image) then through the disk’s circular openings and out.
• a = Lubrication oil entering the inlet
• b = The fluid moving through the disk spacing and along the disk surface
• c = Fluid moves out of the holes around the turbine shaft
• d = Fluid moves out through the outlet

seen in Tables 2 to 4. Redirecting part of the pumped oil flow
toward the Tesla turbine will no doubt lower the flow rate and
oil pressure to the engine components; however, such changes
can be accommodated by adjusting the pressure regulating
valves and/or a higher flow rate oil pump. See Figures 8, 9, and
10 for illustrations in how such a system can be constructed.
Quoting from Rice (2003):
Tesla designed, built, and tested such machines but was
not able to achieve industrial applications. In the subse-
quent years, many investigations have been carried out to
determine the performance and efficiency of this type of
turbo machinery. These have been both analytical and
experimental. Most of the investigations have had a
certain limited application as the objective, with regard
to size and speed as well as the nature of the operating
fluid. However, some of the investigations have tried to
establish the generalized performance of Tesla-type
turbo machines. In general it has been found that the effi-
ciency of the rotor can be very high, at least equal to that
achieved by conventional rotors. But it has proved very
difficult to achieve efficient nozzles in the case of
turbines. For pumps and compressors, efficient diffu-
sions after the rotor has proven difficult to achieve. As a
result, only modest machine efficiencies have been
demonstrated. Principally for these reasons the Tesla-
type turbo machinery has had little utilization. There is,
however, a widespread belief that it will find applications
in the future, at least in situations in which conventional
turbo machinery is not adequate. This includes the use
with very viscous fluids, fluids containing abrasive parti-
cles, and two phase fluids.
A description by Crocker and Cudina (2007) states that
Tesla pumps have two or more rotating parallel disks in which
a viscous drag is created. The disks must rotate very fast,
usually above 10,000 rpm, in order for the viscous drag to
generate the forces to move the liquid. Note: this text refers to
a pump which is essentially the same model. Such units are
known for working as pumps and turbines with minor adjust-
ments (Nam 2012).

154 ASHRAE Transactions

Published in ASHRAE Transactions, Volume 122, Part 1
Figure 7 Images show velocity streamlines moving out of the disk openings. Lubrication oil per inlet port is 0.13 kg/s
(0.29 lb/s) at 100°C (212°F). The image to the left shows a sectional view about the XY plane. Lubrication oil
streamlines start from the inlet port and then head toward the disk openings located around the center of the shaft.
The image to the right shows a zoomed-in image of the cut-through image shown on the left hand side, focusing
on the dash line circle. This close-up image shows the path of the velocity streamlines exiting the disk openings.

Figure 8 A typical layout for an engine oil lubrication system. The picture shows a proposal in how to tap in to the existing
oil pressure lines immediately downstream of the oil pump outlet. Figure 9 shows a line diagram describing system
(The Autocar Handbook 1935).

ASHRAE Transactions 155

Published in ASHRAE Transactions, Volume 122, Part 1
Figure 9 Schematic describing the system. The PRV-M can
be a motorized valve. The PRV-M is controlled by
a temperature thermostat located within the
passenger compartment that controls the pres-
sure supplied to the Tesla turbine. This pressure
variation controls the turbine speed and therefore
varies the compressor speed, thus avoiding the
need to manufacture variable displacement
compressors that are relatively more complex and
costly when compared with fixed displacement
compressors. Fixed displacement compressors
coupled with Tesla turbines can have their refrig-
erant outputs varied by varying the oil pressure
supplied to the Tesla turbine. Note: In belt drive
compressor units, variable displacement
compressors are designed to change the refriger-
ant gas output required to meet the cooling
demand. Therefore, they can reduce unnecessary
energy consumption and contribute to improve-
ment on vehicle fuel delivery.
The following factors given by Lampart and Jedrzejewski
(2010) influence the performance of the turbine or in this
paper, the air-conditioning compressor drive unit:
• Inter-disk spacing
• Number of disks and diameter, and disk surface finish
• Number and shape of nozzles
• Rotational shaft speed
• Engine oil inlet pressure, temperature, velocity and
angle
• Erosion of turbine elements
• Materials used, for example, composites, ceramic mate-
rials, bronzes, aluminum alloys
• Lubrication oil technical specifications
Lampart and Jedrzejewski (2010) experimented on
turbine velocities in the regions of 9000 and 18,000 rpm using
fluid gas.
Nam (2012) states:
The flow of fluid in the turbine can also be reversed into
a pump, and the disks do not suffer from cavitation issues
156
Published in ASHRAE Transactions, Volume 122, Part 1
Figure 10 An example of an automotive air-conditioning
compressor. The dashed circled clutch is a belt
driven by the engine fan belt. The clutch unit
engages the compressor once electrically ener-
gized by the vehicle’s climate control system.
Such air-conditioning compressors can be
supplied with or without this clutch unit. Figure
courtesy DENSO International.
that bladed turbines often have. These features lend
themselves well to potential uses in the generation of
power from geothermal steam and particle-laden indus-
trial gas. Other potential uses include blood pumps as
well as in wind turbines. Though the Tesla turbine has
not yet proven to be more effective than traditional
turbines, there are still potential practical uses.
A paper by Podergajs (2011) mentions some of the draw-
backs and advantages of Tesla turbines:
Its other drawbacks are shear losses and flow restric-
tions. But that can be advantages when flow rates are low.
Tesla’s design can also be used, when a small turbine is
needed. Efficiency maximized when boundary layer
thickness is approximately equal to inter disk spacing.
So at high flow rates, more disks are needed, which
means a larger turbine. Because thickness of boundary
layer depends on viscosity and pressure, various fluids
cannot be used. Disks need to be as thin as possible to
prevent turbulence at disk edges.
On the basis of the abovementioned references, the inter-
disk spacing used in this model was set to match the calculated
boundary layer thickness shown in Table 1. Measuring the
fluid velocity flowing in between turbine disks is not an easy
task; this highlights the practicality of CFD simulation. By
varying the turbine fluid flow and plotting the turbine output
results generated by the software post-processor, efficiency
ASHRAE Transactions
was pinpointed. Therefore it can be said that CFD analysis is
the most practical method that can be used in finding out the
optimum matching of inter-disk spacing and boundary layer
thickness. If the engine oil temperature is 100°C (212°F) and
the initially assumed in–between disks velocity is 30 m/s
(98.4 ft/s), then the inter-disk spacing can be set at 2.5 mm
(0.1 in.). This assumption will be reassessed once the CFD
analysis is complete and run for different lubrication oil veloc-
ity. Tables 2 to 4 summarize the results and show how the effi-
ciency peaks when the in-between disks lubrication oil
velocity produces a boundary layer thickness equal to the
inter-disk spacing.
Figures 3 to 7 show CFD-generated graphical results
highlighting possible turbine configurations and how such
turbines work. Figure 8 shows a typical automotive engine
layout and possible oil supply for the Tesla turbine. Figure 9
shows a schematic on how such a system can work. Figure 10
shows a typical belt-driven air-conditioning compressor.
Figure 11 shows graphs of efficiency versus inlet mass flow
rates for different turbine configurations.
The results in Tables 2 to 4 show that the generated shaft
power is insufficient to drive the air-conditioning compressor.
The highest possible cooling power that can be achieved is
31.5 W. The reason for this low shaft power output can be
explained by the turbine’s known relatively low torque output.
Tesla turbines are low torque/high speed machines. It is the
high turbine speed that compensates for low torque output
when producing power, or the high rads/s component of Equa-
tion 6 that ramps up the turbine shaft power output.

CONCLUSION
CFD methods are the most effective methods in optimiz-
ing the Tesla turbine efficiency. The CFD simulation was
carried out successfully with the following facts determined:

• Tesla turbines are low torque/high speed machines. The

turbine power output can be maximized with higher tur-
bine speeds. This cannot be achieved with the consid-
ered engine oil flow rates.
• Tesla turbines’ efficiency depends on the following fac-
tors:
• Fluid type, dynamic viscosity, and density
• Fluid temperature
• Number of disks Figure 11 Graphs of efficiency versus lubrication oil mass
• Disk surface finish flow inlet per a port, for three slightly different
• Disk thickness geometries. Top: Inter-disk spacing is 2.5 mm
(0.1 in.) with disk outlet hole diameters sized at 30
• Disk surface areas
mm (1.2 in.). See Figure 2 item 2 for dimensions.
• Disk exhaust openings Middle: As above, but with outlet holes’ diameter
• CFD analysis can be used to optimize inter-disk spacing increased to 35 mm (1.4 in.). Bottom: Inter-disk
for a specific fluid flow and temperature by setting the spacing is 3 mm (0.12 in.) with outlet hole diam-
inter-disk spacing to match the fluid boundary layer eters sized at 30 mm (1.2 in.) See Figure 2 item 2
thickness. for dimensions.

ASHRAE Transactions 157

Published in ASHRAE Transactions, Volume 122, Part 1
RECOMMENDATIONS Autocar Handbook, The. 1935. 13th ed. Illustration: Over-
Explore other hydraulic drive types, such as hydraulic head cam engine with forced oil lubrication. London:
motors that do not need higher engine oil flow rates. Hydraulic Iliffe & Sons.
motors should be capable of giving high shaft power at low Blasius, H. 1908. Grenzschichten in Flüssigkeiten mit
pressure under 1.0 bar, and engine lubrication oil temperature kleiner Reibung. Journal of Applied Mathematics and
at 100°C (212°F). Physics 56: 1–37.
Crocker, M. and M. Cudina. 2007. Hand Book of Noise and
ACKNOWLEDGMENTS Vibration Control. pp. 897. University of Ljublana. Pub-
The author would like to thank Dr. Lik F. Sim for setting up lished by John Wiley and Sons Inc., USA.
and providing access to the FLUENT software and Mr. Lampart, P., and L. Jedrzejewski. 2010. Investigations of
Abdelkader Benzamia of Flowpak, Doha, Qatar aerodynamics of Tesla bladeless micro-turbines. Journal
(www.flowpak.net) for the guidance and input on best practices of Theoretical and Applied Mechanics 49(2): 477–99.
in CFD mesh. Massey, B.S. 1983. Mechanics of Fluids. Fifth edition. pp.
533. UK: Van Nostrand Reinhold.
NOMENCLATURE Mehlhoff, B. 2005. Pumped up: Lubrication systems. Irvine,
m· = mass flow rate CA: Super Chevy. www.superchevy.com/how-to/148-05
P1, P2, P3 = pressure at inlets, regardless of inlet orientation 06-lubrication-systems/.
P = pressure Miranda, T. 2014. Personal communication. Senior engineer,
Ps = shaft power Castrol North America.
Q = inlet flow Nam, S. 2012. The Tesla turbine. Stanford, CA: Stanford
Re = Reynolds number University. http://large.stanford.edu/courses/2012/ph
S = speed of turbine 240/nam1/.
T = torque
Podergajs, M. 2011. The Tesla turbine. Ljublana, Slovenia:
v = velocity of fluid
University of Ljubjana. http://www-f1.ijs.si/~rudi/sola/
x = distance from start of boundary layer to center of
disk radius Tesla_turbine.pdf.
 = dynamic viscosity Rice, W. 2003. Chapter 14—Tesla turbomachinery. In Hand-
 = density of fluid book of turbomachinery. Earl Logan, Jr., ed. Tempe, AZ:
 = boundary layer thickness Arizona State University.
 = efficiency Scientific American. 1911. From the complex to the simple.
Scientific American 105: 290.
REFERENCES Zhou, Q. 2009. Engine lubrication system analysis and oil
ASHRAE. 2008. ASHRAE Handbook—Systems and Equip- pump design optimization. pp.1 Whitley, Coventry, UK:
ment. Atlanta: ASHRAE. Jaguar and Land Rover Engineering Centre.

158 ASHRAE Transactions

Published in ASHRAE Transactions, Volume 122, Part 1
Reproduced with permission of the copyright owner. Further reproduction prohibited without
permission.