Sunteți pe pagina 1din 11

Applied Thermal Engineering 27 (2007) 145155

www.elsevier.com/locate/apthermeng

Mathematical modeling of the working cycle of oil injected rotary twin


screw compressor
N. Seshaiah *, Subrata Kr. Ghosh, R.K. Sahoo, Sunil Kr. Sarangi
Cryogenics and Gas dynamics Laboratory, Department of Mechanical Engineering, National Institute of Technology,
Sector-2, NIT Campus, Rourkela 769008, Orissa, India

Received 12 July 2005; accepted 8 May 2006


Available online 7 July 2006

Abstract

Oil injected twin-screw air and gas compressors are widely used for medium pressure applications in many industries. Low cost air
compressors can be adopted for compression of helium and special gases, leading to signicant cost saving. Mathematical analysis of oil
injected twin-screw compressor is carried out on the basis of the laws of perfect gas and standard thermodynamic relations. Heat transfer
coecient required for computer simulation is experimentally obtained and used in performance prediction, when the working medium
being air or helium. A mathematical model has been developed for calculating the compressor performance and for validating the results
with experimental data. The ow coecients required for numerical simulation to calculate leakage ow rates are obtained from e-
ciency verses clearance curves. Eect of some of the compressor operating and design parameters on power and volumetric eciencies
have been analyzed and presented.
 2006 Elsevier Ltd. All rights reserved.

Keywords: Screw compressor; Oil injection; Heat transfer coecient; Leakage; Flow coecients

1. Introduction India, it is necessary that all machines operate eciently.


This can be achieved only when machine performance is
The screw compressor is a positive displacement well understood and is predictable. Unlike other compres-
machine that uses a pair of intermeshing rotors housed in sors, the mechanism of gas compression in an oil injected
a suitable casing to produce compression. Screw compres- screw compressor is extremely complex. It is dicult to
sors are capable of high-speed operation over a wide range estimate the compressor performance analytically. On the
of operating pressures. In a screw machines, oil is deliber- other hand, experimental studies are prohibitively expen-
ately injected into the compression chamber to improve sive because a new rotor needs to be fabricated using
and to provide sealing, lubrication, corrosion resistance expensive machining techniques for every change in rotor
and cooling eect. Rotary dual screw compressors are geometry.
widely used in industry for air and gas compression and
for refrigeration applications. They are particularly suit- 2. Modeling of compressor cycle
able for compression of air and helium used in small and
intermediate size cryogenic refrigerators and liqueers. A Analysis of volumetric and power eciencies is essential
computerized method for generation of rotor proles and to estimate the suitability of a compressor for a particular
analysis and performance has been suggested by Singh application. The main objective of the present performance
et al. [1,2]. Due to the high cost of energy, particularly in analysis is to develop a numerical model to ascertain the
suitability of a commercially available air compressor
*
Corresponding author. Tel.: +91 943 710 9766; fax: +91 661 246 2999. for dierent gas compression applications. Eciency of
E-mail address: seshuet@yahoo.com (N. Seshaiah). any compressor depends on the processes involved in the

1359-4311/$ - see front matter  2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.applthermaleng.2006.05.007
146 N. Seshaiah et al. / Applied Thermal Engineering 27 (2007) 145155

Nomenclature

A heat transfer area (m2) r pressure ratio (dimensionless)


Ac leakage clearance area (m2) S sealing line length along the rotor (m)
Ae experimental adiabatic power required for the T temperature (K)
compressor (W) t time (s)
Af cross sectional area of female rotor groove (m2) T1 temperature of gas at the end of suction process
Am cross sectional area of male rotor groove (m2) (K)
As input power to the compressor (W) Tg temperature of gas in the working space (K)
At theoretical adiabatic power of the compressor Tl temperature of oil in the working space (K)
(W) Toil mean temperature of leaked oil in the suction
a clearance between lobe tip and housing (m) cavity (K)
C ow coecient (dimensionless) Ts inlet gas temperature (K)
cl specic heat of lubricating oil (J/kg K) ts time required for suction process (s)
cp specic heat of gas at constant pressure (J/kg K) U internal energy (J)
cv specic heat of gas at constant volume (J/kg K) V volume of the working space (m3)
D rotor diameter (m) Vt rotor tip speed (m/s)
e interlobe clearance (m) V1 inducted gas volume at inlet condition (m3)
h heat transfer coecient between gas and oil (W/ V2 leakage gas volume at inlet pressure (m3)
m2 K) Vt1 geometrical volume of one pair of male and fe-
i specic enthalpy (kJ/kg) male cavities (working chamber) (m3)
Hg enthalpy of gas (J) W gas work (J)
k ratio of specic heats (dimensionless) wt lobe tip width (m)
Mg mass of gas in the working space (kg)
Ml mass of oil in the working space (kg)
Greek symbols
Mt1 theoretical gas mass in a pair of male and female
gv volumetric eciency (dimensionless)
cavities at the end of suction process (kg)
ga adiabatic eciency (dimensionless)
Mil interlobe leakage mass leaked into the suction
gtv theoretical volumetric eciency (dimensionless)
cavity during previous compression process (kg)
gev experimental volumetric eciency (dimension-
m leakage mass ow rate through ow path (kg/s)
less)
mt theoretical mass ow rate in to the suction
gta theoretical adiabatic eciency (dimensionless)
chamber at suction condition (kg/s)
gea experimental adiabatic eciency (dimensionless)
me actual discharged gas mass rate (kg/s)
u oil to gas mass ratio (dimensionless)
mb leakage mass ow rate through blowhole (kg/s)
h rotor rotational angle (deg)
md leakage gas mass ow rate through clearance be-
U male rotor wrap angle (deg)
tween rotor end and casing wall (kg/s)
b modied adiabatic index (dimensionless)
mil leakage gas mass ow rate through interlobe
l dynamic viscosity of oil (N s/m2)
clearance (kg/s)
q density (kg/m3)
mlr leakage mass ow rate of oil through rotor tip
C time for compression process (s)
housing clearance (kg/s)
s cycle time of compressor (s)
mgl net gas mass leakage rate from a pair of grooves
during compression process (kg/s)
mgt total gas mass leakage rate per second (kg/s) Subscripts
mdt theoretical discharged gas mass rate (kg/s) 1 upstream
N rotational speed of rotor (RPS) 2 downstream
P pressure in the working space (N/m2) b beyond ow path
Pd discharge pressure (N/m2) f female rotor
Ps suction pressure (N/m2) g gas
Q heat transferred between gas and oil (J) i gas or oil going in to the working space
q leakage volume ow rate through ow path l oil
(m3/s) m male rotor
R gas constant (J/kg K) o gas or oil going out of the working space
Rm modied gas constant of oil gas mixture (J/
kg K)
N. Seshaiah et al. / Applied Thermal Engineering 27 (2007) 145155 147

working cycle. Major processes of screw compressor are Under ideal conditions, the heat lost by the oil lm must be
suction, compression and discharge. Analysis of these equal to heat gained by the fresh gas mass. From Eqs. (1)
individual processes is essential to model the compressor and (2), the mean temperature of inducted gas at the end of
performance. suction process can be obtained as

2.1. Suction process hAT oil  T s ts


T1 Ts 3
M 1 cp
Volumetric eciency of the compressor greatly depends
on the amount of gas mass inducted into the suction cavity The amount of gas mass inducted into the geometrical vol-
during suction process. This, in turn, depends on the tem- ume at suction condition (Ps, Ts) is
perature of the cavity walls at steady state. Analysis of the P s V t1
suction process gives the average temperature of the gas in M ts 4
RT s
the suction cavity at the end of suction process. Model of
the working chamber during suction process is shown in Vt1, the geometrical volume of a pair of male and female
Fig. 1. Since the pressure and temperature uctuations dur- rotor cavities is dened [3] as below:
ing suction process are generally small, the following quan-
tities are assumed to be constant during the suction V t1 Am Af L
process: Am, and Af being the cross sectional areas of the male and
female cavities respectively and L the rotor length.
Inlet velocities of gas and oil. The gas mass in the suction cavity at condition (Ps, T1)
Inlet temperature of gas and oil. can be estimated by the expression
Pressure drop across the inlet port. Ts
Rate of heat ow from gas to oil (or from oil to gas). M t1 M ts 5
T1
The quantity of gas mass inducted into the suction cav- This gas mass is the sum of fresh charge inducted (M1) and
ities depends on the temperatures of inlet gas and the cavity the mass leaked through interlobe clearance (Mil)
walls. During the suction process the cavity wall tempera- M t1 M 1 M il 6
ture is higher than the inducted gas temperature because
of the heat ow from the compressed gas during the com- Eliminating the total mass Mt1 between Eqs. (5) and (6),
pression process. The cavity wall is covered with a lm of the fresh gas mass inducted during suction process is ob-
lubricating oil which leaks from the adjoining compression tained as below:
space when the oil is injected. The amount of heat trans- Ts
M 1 M ts  M il 7
ferred from the hot suction cavity wall (lubricating oil lm) T1
to the inducted gas during suction process can be estimated
Substituting this value of M1 value in Eq. (3), the following
by the expression
quadratic equation is obtained in terms of the gas temper-
Q M 1 cp T 1  T s 1 ature T1:
If the temperature rise of the inducted gas is small com-  cp M il T 21 T 1 cp M t1 T s cp M il T s  hAT oil  T s ts 
pared to the temperature dierence between the lubricating
oil lm and the inducted gas, the heat transfer between the  cp M t1 T 2s 0 8
leaked oil and inducted gas may be written as below:
Q hAT oil  T s ts 2 2.2. Compression and discharge process

Both the compression and the discharge processes are


unsteady processes. Thermodynamic properties of the gas
and oil vary continuously during compression process.
The gas in the working chamber is compressed to a high
pressure by the rotational movement of the rotors. To sim-
plify the analysis, it is assumed that the oil and gas are sep-
arate uids, and only heat is exchanged between them. The
discharge port is so located that the cavities connect to the
discharge port when the pressure in the working chamber
reaches the designed discharge pressure and discharge con-
tinues till the male rotor lobe completely disengage from
female rotor groove.
Fig. 1. Model of the working chamber of an oil injected twin-screw The following factors are taken into account in the
compressor cavity at the end of the suction process [4]. model:
148 N. Seshaiah et al. / Applied Thermal Engineering 27 (2007) 145155

1. volume change due to rotor rotation, Substituting (11)(14) in (10), and rearranging, the rate of
2. mass and enthalpy ows of gas entering or leaving the change of working gas temperature is obtained
working space through discharge port and leakage   
dT g k  1T g dV P b mgi
paths,   qli qlo kT gb  T g
dt Vg dt P Mg
3. mass and enthalpy ows of oil, entering or leaving the 
working space through discharge port and leakage mgo hA
 k  1T g  T g  T l 15
paths, Mg cv M g
4. heat exchange between gas and oil. The rst term on the right-hand side of the above equation
relates to the change in volume including the leakage rate
To simplify the calculations, the following assumptions of oil. The second and third terms represent the eect of
are made: gas leakage into and out of the compressor cavity respec-
tively. The last term is derived from the heat transfer from
Gas and oil temperatures are homogeneous at any gas to the oil.
instant in the working space. The rate of change of net gas volume of the working
Gas and oil never change phase. chamber [4] can be written as below:
Pressure is uniform throughout the working space at
any stage. dV g dV
 qli qlo 16
The working gas is an ideal gas. dt dt
Oil is an incompressible uid. The rate of change of gas mass due to internal leakage is
Heat exchange between gas and oil is in proportion to given by the expression
the temperature dierence between them.
dM g
A pressure uctuation across the discharge port is mgi  mgo 17
dt
negligible.
The equation of state of perfect gas may be written as
Fujiwara and Osada [4] derived the fundamental equa- follows:
tions based on standard thermodynamic laws and the laws PV g M g RT g
of perfect gas. The following equations are derived with
some modications and other details are introduced as The dierential form of the above equation can be written
needed. as follows:
 
The rst law of thermodynamics for unsteady ow of dP 1 dV g dM g dT g
gas through the working chamber can be expressed as P RT g RM g 18
dt Vg dt dt dt
dEv dQ dW Substituting Eqs. (15)(17) in Eq. (18), the rate of change
mgi igi  mgo igo  9
dt dt dt of pressure is obtained as follows:
Ev is the sum of internal, potential and kinetic energies.   
Assuming the potential and kinetic energies of gas to be dP 1 dV
kP qlo kP b  P b P qli
negligible, from the above equation the change in internal dt Vg dT

energy in time dt can be computed as T gb PV g PV g PV g hA
k mgi  k mgo  T g  T l 19
dU g dQ  dW dH g 10 T gM g Mg cv M g T g

Change in internal energy can also be expressed as a func- The rate of change of oil temperature is obtained from en-
tion of mass and temperature change and dened as below: ergy balance in terms of leakage oil temperature and heat
transferred from the gas. The increase of oil temperature
dU g cv M g dT g cv T g dM g 11 in the working chambers is due to heat gained from the
The change of enthalpy due to leakage can be expressed as leakage oil (which is at higher temperature) and the heat
gained from the gas under compression.
dH g cp T gb dM gi  cp T g dM go 12
Assuming the potential and kinetic energies to be negli-
The gas work may be expressed in terms of geometrical vol- gible, the enthalpy change of oil in the working chamber
ume change, and oil volume change due to leakage. Since can be expressed as follows:
the oil is an incompressible uid, the gas work is expressed dH l dW l dQ dJ l 20
as
where dJ1 is the heat lost by the leaked oil.
dW P dV Pqlo dt  P bi qli dt 13
The enthalpy change of oil in time dt may be expressed
Heat exchange between the gas and the oil in time dt is as
assumed to follow the Newtons law of cooling and is dH l M l cl dT l 21
expressed as
The work done by the oil is zero (i.e dW = 0), since oil is an
dQ hAT g  T l dt 14 incompressible uid.
N. Seshaiah et al. / Applied Thermal Engineering 27 (2007) 145155 149

The heat gained by the oil in time dt is given by the (2) The gas/oil mixture ratio is same in all leakage paths
expression except at the lobe tip clearance and equal to the mix-
dQ hAT g  T l dt 22 ture ratio passing through discharge port.

The oil leakage in to the working chamber is at higher tem- The apparent ratio of specic heats of oil and gas mix-
perature than the oil in the cavity. The energy lost by ture is calculated from Ref. [4]
leaked oil is given by the expression
cp /cl
dJ l mli cl T lb  T l dt 23 b 27
cv /cl
Substituting Eqs. (21)(23) in Eq. (20), the rate of change The modied gas constant of the mixture is also calculated
of oil temperature is obtained as follows: using Ref. [4]
dT l mli hA R
T lb  T l T g  T l 24 Rm 28
dt M l M l cl 1/
The rate of change of oil mass in the working chamber due The mass ratio of oil to gas in the working chambers as
to leakage as follows: well as through all leakage paths is homogeneous and equal
dM l to the mixture ratio in the working volume [5] and is ex-
mli  mlo 25 pressed as below:
dt
Ml ml
/ 29
2.3. Leakage calculation of gas and oil M g mg
The average leakage area is determined by multiplying
Leakage is a major concern in screw compressors. The sealing line length with an average gap (clearance) for each
major leakage paths identied during compression process type of leakage. Normally, the average gap/clearance is
are leakage through interlobe clearance, through blowhole, determined from the actual clearance measurements in
rotor tip-housing clearance and clearance between rotor the compressor. The discharge or ow coecients are
discharge end and casing wall [57]. Leakage mass through empirically selected to account in the presence of oil.
interlobe clearance will go directly into the suction cham- At the lobe tip, it is assumed that the clearance lls with
bers. Except at the lobe tip clearance, leakage gas and oil the oil due to the action of centrifugal force, and the oil
are uniformly mixed and the ow is isolated from any heat leakage ow is in single phase. The leakage ow rate of
exchange with their surroundings while maintaining ther- oil can be calculated using the equation of incompressible
mal equilibrium. In this study, the leakage rate is assumed viscous ow through a narrow gap [10] as follows:
to follow the well-known equation for ow through a con-  
vergent nozzle [8,9] V t a P 1  P 2 a3
mlr Sql  30
v 2 12ll wt
u  2 b1

u2b rb  r b
CAc P 1 t The leakage gas mass rate into the working chambers is
m mg ml p through leading blowhole, and through clearance between
T1 b  1Rm
 b1
b discharge end of rotors and casing wall from male and
2 female rotor leading cavities. It can be calculated using
for r > 26
b1 the below expression:

and mgi mgbi mgdmi mgdfi =1 / 31


v The leakage gas mass going out of the working chambers is
u  b1
u  b1 through trailing blowhole, interlobe clearance, and through
u 2 b1 b
CAc P 1 tb b1 2 the clearance between lagging cavity end and casing wall,
m p for r 6
T1 Rm b1 of male and female rotors at discharge end and is expressed
where r PP 21 , P1 and P2 being upstream and down stream as
pressures respectively. mgo mgbo mgilo mgdmo mgdfo =1 / 32
Due to the presence of oil, exact properties of oilgas
mixture are not known. However, the properties are calcu- Similarly, the leakage rate of oil mass into and out of the
lated based on well justiable assumptions and compared working chambers is given by the relations
with experimental data. By comparison with laboratory mli /mgi mlrmi mlrfi 33
tests, the following assumptions for dierent types of leak-
mlo /mgo mlrmo mlrfo 34
age paths have been shown to be the most appropriate [5]
Eqs. (26)(34) are adequate to calculate the rate of change
(1) The gas/oil mixture in all leakage paths is of gas and oil mass during compression and discharge
homogeneous. process.
150 N. Seshaiah et al. / Applied Thermal Engineering 27 (2007) 145155

2.4. Heat transfer coecient 1.06

Relative volumetric efficiency


Pd /Ps =8.65;Tli=308K Air
1.05
The heat transfer coecient between gas and oil is essen-
tial to estimate the rate of heat transfer between gas and 1.04
y = 0. 0026x + 0. 2416
oil. Fujiwara and Osada [4] determined the heat transfer 1.03
coecient using experimental performance data. In this 1.02
study, the same methodology is followed.
The heat transfer coecient between gas and oil is 1.01
dened as follows: 1
kP s V t1 dgev 0.99
h 35 292 296 300 304 308 312 316 320
k  1Ats dT s
Suction temperature (K)
Eq. (35) relates h to the tangent of the gevTs curve. The
experimental test setup used to determine heat transfer Fig. 3. Experimental volumetric eciency curve against suction temper-
coecients and to measure various operating parameters ature (eciencies are relative to suction temperature of 294 K).
is shown in Fig. 2. Applying experimental test data, the
change in volumetric eciency with inlet gas temperature 1.06

Relative volumetric efficiency


for air and helium are presented in Figs. 3 and 4 respec- Pd /Ps =8.65;Tli=308K Helium
1.05
tively. The slopes of these lines give the dg ev
dT s
value which
1.04
can be directly substituted in the numerical simulation to
nd the heat transfer coecient between gas and oil. The 1.03 y = 0. 0025x + 0. 2299

values of dgdT
ev
for air and helium are 0.0026 K1 and 1.02
1 s
0.0025 K respectively (as shown by the straight lines in 1.01
the gures). The experimental heat transfer coecients 1
for air and helium are found to be 2.141 kW/m2 K, and 0.99
1.4656 kW/m2 K respectively. 0.98
However, there is no exact information available con- 300 304 308 312 316 320 324
cerning the heat transfer area. Fujiwara and Osada [4] Suction temperature (K)
dened the representative heat transfer area between gas
Fig. 4. Experimental volumetric eciency shown against suction temper-
and oil as below: ature (eciencies are relative to suction temperature of 302 K).
2=3
A V t1 36
The time for suction process depends on male rotor rota- 2.5. Time for compression process
tion speed and dened as follows:
The time for compression and discharge process
1 depends on male wrap angle, male rotor rotational speed
ts 37
Nm and number of male lobes.

Fig. 2. A view of the experimental setup used to measure various operating parameters of screw compressor.
N. Seshaiah et al. / Applied Thermal Engineering 27 (2007) 145155 151

The time taken for compression and discharge process in Theoretical volumetric eciency in terms of mass ow rate
seconds is given by the expression can be dened as
  mdt
1 uU 1 gtv 46
C 38 mt
N m 360 nm
The experimental gas mass ow rate will be less than the
2.6. Total cycle time for suction and compression processes theoretical inducted mass rate due to imperfect nature of
ports, wall friction and other frictional losses apart from
The time for the entire compression process including leakage loses. Experimental volumetric eciency may be
suction and compression stages also depends on the male calculated after measuring the actual discharged mass ow
wrap angle, male rotor rotational speed and the number rate.
of male lobes. The experimental volumetric eciency is be dened as
This cycle time in unites of seconds can be expressed as me
gev 47
below: mt
 
1 uU 1
s 1 39 3.2. Adiabatic eciency
Nm 360 nm

The theoretical adiabatic eciency is dened as


3. Eciencies
At
gta 48
Many standard eciency denitions exist that qualify As
the mass ow and power performance characteristics of The theoretical adiabatic power required for the compres-
compressors. Volumetric and power eciency denitions sor is given by the expression
are taken from standard thermodynamic textbooks and " k1 #
from Ref. [11]. Pd k
At mdt cp T s 1 49
Ps
3.1. Volumetric eciency
The experimental adiabatic power Ae can be interpreted as
The theoretical gas mass inducted into any pair of male the power required to compress the gas adiabatically that
and female rotor grooves at suction condition can be produces the actual discharged mass ow rate at the given
expressed as pressure ratio.
The experimental adiabatic eciency is dened as
P s V t1
M ts 40 Ae
RT s gea 50
As
The theoretical gas mass inducted into the working cham- The experimental power required for the compressor is
bers at an average temperature of T1 becomes given by the expression
" k1 #
Ts Pd k
M t1 M ts 41 Ae me cp T s 1 51
T1 Ps
Thus the theoretical gas mass inducted per second at suc-
The actual power consumed by the compressor is inclusive
tion condition comes out as
of that consumed by the compressor cooling fan, transmis-
mt M t1 N m nm 42 sion and electric motor losses, as well as mechanical losses.

The net gas mass leakage rate from a pair of male and 4. Outline of computer program
female rotor grooves is calculated as given below:
At steady state, all the changes in the compressor work-
mgl mgi  mgo 43
ing chambers/cavities are related to rotor turning, and the
The total gas mass leakage rate depends on the number of state in a groove varies as function of the rotor turning
lobes on the male rotor and its rotational speed. The net angle/time. Therefore, if the change in a state of one pair
leakage rate per second is dened by the relation of grooves is calculated, the states in all the grooves can
be known. All the equations obtained above are used for
mgt mgl  N m  nm 44 suction and discharge process simulation. Inputs required
The theoretical discharged gas mass rate at an average suc- for simulation are listed in Tables 1 and 2. For the sake
tion condition can be estimated as below: of simplicity, the outlet pressure (discharge pressure) is
assumed to be constant. The ow chart for the computer
mdt mt  mgt 45 program employed to model the governing equations is
152 N. Seshaiah et al. / Applied Thermal Engineering 27 (2007) 145155

Table 1 shown in Fig. 5. Step-by-step calculations start from the


Rotor specications and operating conditions of prototype compressor end of the suction process.
Prole Sigma To calculate the leakage ow rate, the state of the
Combination of number of teeth 5+6 grooves beyond the leakage path must be known a prior,
Outer diameter of male rotor (mm) 72 but this information is not available at the beginning of
Outer diameter of female rotor (mm) 54 the calculation. Therefore, to start with, the state in the
Wrap angle (deg) 300
Rotor length (mm) 90
grooves is computed assuming no leakage of gas. Subse-
Theoretical volume of a pair of male and 34.2 quently, the leakage ow is calculated and the state is cor-
female rotor grooves (cm3) rected following the RungeKutta procedure. This
Built in volume ratio 5 calculation is iterated until the system converges. Finally
Suction pressure (MPa) 0.1 volumetric and adiabatic eciencies are calculated.
Discharge pressure (MPa) 0.88
Rotational speed of male rotor (rpm) 4350
Supplied oil temperature (K) 308 5. Error analysis
Supplied oil rate (l/min) 23.6
It has been observed that almost all the error arises from
the mass ow rate measurement. Gas volume ow rates are
Table 2 measured with a rotameter and temperature, and pressure
Flow coecients and clearances at dierent leakage paths are measured with platinum resistance thermometers and
Coecient at interlobe clearance 0.65 pressure transducers respectively. The gas density is deter-
At lobe tip and casing bore 0.7 mined using an equation of state as a function of pressure,
At leading blowhole 0.7 temperature and gas composition. A more realistic expres-
At lagging blowhole 0.6
Between rotor end and casing wall 0.4
sion for the overall uncertainty can be predicted by the root
Interlobe clearance (mm) 0.027 mean square error [14].
Clearance between lobe tip and casing bore (mm) 0.03
Clearance between rotor end and end plate (mm) 0.03 5.1. Uncertainty in determination of volumetric eciency
Leading blowhole area (mm2) 2
Lagging blowhole area (mm2) 2
The uncertainty in volumetric eciency determination is
calculated as below:
s
2  2
ogev ogev
Start
Dgev Dme Dmt
ome omt
Assuming that the uncertainties in the measurement of
Input reference condition and suction temperature and pressure,
and thermo-physical properties are negligible, the uncer-
tainty in experimental mass ow rate can be expressed as
Initialize
s
2  2  2
ome ome ome
Dme DV m DP s DT s
oV m oP s oT s
Calculate suction The uncertainty associated in measuring the theoretical
process mass ow rate is in the measurement of theoretical volume
of one pair of male and female rotor grooves and that in
Calculate compression and
the speed measurement of the male rotor.
discharge process using step The uncertainty in the theoretical mass ow rate mea-
by step procedure surement is calculated as follows:
s
 2  2  2  2
omt omt omt omt
No
Dmt DV t1 DN m DP s DT s
Converged oV t1 oN m oP s oT s

Yes
Calculate volumetric and 5.2. Uncertainty in determination of adiabatic eciency
adiabatic efficiency
The uncertainty in the measurement of adiabatic e-
ciency is computed as below:
Stop s
2  2
ogea ogea
Fig. 5. Flow diagram for solution of governing equations of a screw Dgea DAe DAs
compressor. oAe oAs
N. Seshaiah et al. / Applied Thermal Engineering 27 (2007) 145155 153

The uncertainty in calculation of experimental adiabatic 96 Air (exp)


power is expressed as Tli=308K Air (calculated)
Helium (exp)
s 94 Helium (calculated)
 2  2  2  2
oAe oAe oAe oAe
DAe Dme DT s DP d DP s 92
ome oT s oP d oP s

(v %)
90
and the uncertainty associated in measuring the system
power is calculated as 88
s
 2  2
oAs oAs 86
DAs Dne Dte
one ote
84
2 3 4 5 6 7 8 9
One data point from the test facility has been taken from
Pressure ratio (Pd /Ps)
the multiple data of various measurable parameters such
as suction and discharge pressure, temperature, RPM of Fig. 6. Variation of volumetric eciency with discharge pressure at a xed
male rotor, time taken for a specied number of revolu- injected oil temperature and other inlet conditions.
tions of energy meter, and ow rate measurement. The
average value of the samples at same operating condi- eciency at higher discharge pressures is negligible. This
tions is recorded as a data point. The standard deviation happens because the screw compressor does not have any
of the average value is recorded as a component of uncer- clearance volume at the end of compression process. Our
tainty for that data point. This component represents result is similar to those results obtained by Osada and
the observed random eect associated with a particular Fujiwara [4] and by Stosic et al. [12]. The dierence in the-
measurement. oretical and experimental values is attributed to be imper-
As an example, for air, the uncertainties in the measure- fect nature of ports, wall friction and other frictional losses
ment of various parameters are observed as follows: apart from leakage loss. Variation of adiabatic eciency
DVm = 0.1 L/min; DPs = 0.001 bar; DPd = 0.0015 bar; with discharge pressure is estimated using the procedure
DTs = 0.2 C; DNm = 0.4 rev/s; Dme = 0.015 kg/min; described and shown in Fig. 7. The nature of change of adi-
Dne = 0.12 rev/min; Dte = 0.02 s; Dmt = 0.017 kg/min. abatic eciency with discharge pressure is similar for
Substituting the above values of dierent measurable helium and air. The gure shows that there exists an opti-
quantities and the corresponding uncertainties in the rele- mum pressure ratio for the highest adiabatic eciency.
vant equations, the uncertainties in the measurement of The inuence of interlobe clearance on the PV diagram
volumetric and adiabatic eciency are calculated. of a 5 + 6 rotor combination using air and helium as a
The error in the measurement of volumetric and adia- working media is presented in Figs. 8 and 9 respectively.
batic eciency for air are calculated to be The nature of the PV diagram is similar to those of Refs.
[8,13] for adiabatic compression. The result shows that the
Dgev 0:054 for gev 0:89 area of PV diagram decreases with increase of interlobe
Dgea 0:039 for gea 0:5781 clearance. The decrease is more severe for helium than
for air. This is because of the higher decrease of mass in
An error analysis has been performed for all sets of the compression chamber caused by the increased interlobe
readings and the level of uncertainties is found to be of gas leakage at higher interlobe clearance. Helium being a
the same order. Conservative estimates of the uncertainties,
which are generally applicable to all the test results, are as
follows: air (exp)
62 Air (calculated) Tli =308K
Dgev Helium (exp)
Helium (calculated)
6:0% 58
gev
54
and
50
(a%)

Dgea 46
7:0%
gea
42
38
6. Discussion
34

Theoretical and experimental studies have been con- 30


2 3 4 5 6 7 8 9
ducted on a commercially available air compressor using
Pressure ratio (Pd /Ps)
both air and helium as working uids. The variation of vol-
umetric eciency with discharge pressure for air and Fig. 7. Variation of adiabatic eciency with discharge pressure at
helium is presented in Fig. 6. The variation of volumetric constant inlet condition at a xed injected oil temperature.
154 N. Seshaiah et al. / Applied Thermal Engineering 27 (2007) 145155

11 650
e=0.03mm Tg, Ml/Mg=40
Tli=308K;RPM=4350 Tl, Ml/Mg=40 Pd /Ps =8.65;RPM=4350
10 e=0.09mm Tg, Ml/Mg=50
e=0.12mm 600 Tl, Ml/Mg=50
9 Tg, Ml/Mg=60 Discharge

Temperature (K)
Tl, Ml/Mg=60
port opens
8 550 Tg, Ml/Mg=70
Tl, Ml/Mg=70
Tg, Ml/Mg=80
7 500
Tl, Ml/Mg=80 Temp. Profile
Tg, Ml/Mg=90
Pd /Ps

of Gas
6 Tl, Ml/Mg=90
Tg, Ml/Mg=100
5 450 Tl, Ml/Mg=100
Tg, Ml/Mg=110
Tl, Ml/Mg=110
4 400
Tg, Ml/Mg=120
Tl, Ml/Mg=120 Temp. Profile
3 of Oil
350
2
1 300
0 0 0.002 0.003 0.005 0.006 0.008 0.009 0.011 0.012 0.014 0.015
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Compression time (s)
Vg /V t1
Fig. 11. Variation of gas and oil temperatures during compression process
Fig. 8. Eect of interlobe clearance on PV diagram prole of air at a at dierent oil to gas mass ratio of helium at a xed pressure ratio and inlet
xed injected oil temperature and inlet conditions. conditions.

compression is transferred to the oil when the discharge


11
10 Tli=308K;RPM=4350
e=0.03mm
e=0.09mm
port opens as shown in the gures. The rise in oil temper-
9
e=0.12mm ature is relatively less due to its high specic heat capacity.
8 Also, the rise in temperature of helium gas is relatively
7 more than that of air. This eect is attributed to the higher
specic heat ratio (c = 1.67) for helium and relatively less
Pd /Ps

6
5 heat transfer coecient value (1.4656 kW/m2 K). The value
4 of oil to gas mass ratio used in the simulation model are
3 derived from the experimental data. This ratio is much
2 higher for helium compared to air. The theoretical values
1 of gas temperatures obtained during compression process
0
0 0. 1 0. 2 0. 3 0. 4 0. 5 0. 6 0. 7 0. 8 0.9 1
(for both air and helium) are higher than the expected val-
ues. This is because it has been assumed that during com-
Vg/Vt1
pression process, the heat of compression is transferred
Fig. 9. Eect of interlobe clearance on PV diagram prole of helium at a only to oil. In reality, oil and gas are mixed and a signi-
xed injected oil temperature and inlet conditions. cant fraction of the heat will be transferred to the screw
rotors and the casing. In fact, the compressor system has
low molecular weight gas can easily escape through narrow a cooling mechanism, where the heat of compression is
gap leading to higher leakage loss. continuously dissipated to the environment.
The inuence of injected oil quantity on temperature of The inuence of interlobe clearance and rotational speed
gas and oil is shown in Figs. 10 and 11 for air and helium of male rotor is presented in Fig. 12. The decrease of volu-
respectively. It is observed that enhancement of oil to gas metric eciency with an increase of interlobe clearance is
mass ratio beyond a certain limit, has little eect on tem- more severe at lower rpm for helium than air as shown in
perature of the compressed gas. Most of the heat of gas gure, because of its lower molecular weight, the gas can

600 RPM=3000(Air)
Tg, Ml/Mg=10
95 Pd /Ps =8.65;Tli =308K RPM=4350(Air)
Tl, Ml/Mg=10 Pd /Ps =8.65;RPM=4350 RPM=5000(Air)
Tg, Ml/Mg=12
RPM=3000(He)
550 Tl, Ml/Mg=12
RPM=4350(He)
Tg, Ml/Mg=14
85
Temperature (K)

Tl, Ml/Mg=14 RPM=5000(He)


Tg, Ml/Mg=16 Discharge
500 Tl, Ml/Mg=16
port opens
Tg, Ml/Mg=18
( tv %)

Tl, Ml/Mg=18
Tg, Ml/Mg=20
75
450 Tl, Ml/Mg=20 Temp. Profile
Tg, Ml/Mg=22 of Gas
Tl, Ml/Mg=22
Tg, Ml/Mg=24 65
400 Tl, Ml/Mg=24
Tg, Ml/Mg=26
Tl, Ml/Mg=26 Temp. Profile
of Oil 55
350

300 45
0 0.002 0.003 0.005 0.006 0.008 0.009 0.011 0.012 0.014 0.015 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
Compression time (s) Interlobe clearance (mm)

Fig. 10. Variation of gas and oil temperatures during compression process Fig. 12. Inuence of RPM on volumetric eciency with interlobe
at dierent oil to gas mass ratio and inlet conditions. clearance at a xed injected oil temperature and inlet condition.
N. Seshaiah et al. / Applied Thermal Engineering 27 (2007) 145155 155

100 soring this work. The technical inputs from Mr. Trilok
RPM=3000(He)
Pd /Ps =8.65;Tli =308K RPM=4350(He) Singh, Head Cryogenics Division, BARC are also grate-
RPM=5000(He)
96 RPM=3000(Air) fully acknowledged. The authors are thankful to the
RPM=4350(Air)
RPM=5000(Air) Mechanical Engineering Department, NIT, Rourkela, for
providing all testing facilities and guidance.
92
(tv %)

References
88
[1] Pawan J. Singh, A.D. Onuschak, A comprehensive, computerized
method for twin screw rotor prole generation and analysis, in:
84
Proceedings of the Purdue Compressor Technology Conference,
Purdue, USA, 1984, pp. 519527.
80 [2] Pawan J. Singh, G.C. Patel, A generalized performance computer
0 1 2 3 4 5 6 7 8 9 10 11 12 13 program for oil ooded twin-screw compressors, in: Proceedings of
Blowhole area (mm2) the Purdue Compressor Technology Conference, Purdue, USA, 1984,
pp. 544553.
Fig. 13. Eect of blowhole area on volumetric eciency at dierent RPM [3] P. Koelet, Industrial Refrigeration Principle Design and Applications,
at a xed injected oil temperature and inlet condition. McGraw Hill, 1992.
[4] M. Fujiwara, Y. Osada, Performance analysis of oil injected screw
compressor and its application, International Journal of Refrigera-
easily escape through narrow gap. The blowhole area is tion 18 (4) (1995) 220227.
[5] B. Sangfors, Computer simulation of oil injected twin-screw com-
inherent in the contour of the rotor. The eect of blowhole pressor, in: Proceedings of the Purdue Compressor Technology
area on the estimated volumetric eciency is presented in Conference, Purdue, USA, 1984, pp. 528535.
Fig. 13. Volumetric eciency variation with increase in [6] A. Pietsch, S. Nowotny, Thermodynamic calculation of a duel screw
blowhole area on air is less at higher rpm, and somewhat compressor based on experimentally measured values taking super-
higher for helium, the dierence can largely be attributed charge into account, in: International Compressor Engineering
Conference, Purdue, USA, 1990, vol-I, pp. 4450.
to helium being a light gas. [7] Huagen Wu, Ziwen Xing, Pengcheng Shu, Theoretical and experi-
mental study on Indicator diagram of twin screw refrigeration
7. Conclusions compressor, International Journal of Refrigeration 27 (4) (2004) 331
338, June.
[8] M. Fujiwara, K. Kasuya, T. Matsunaga, W. Makoto, Computer
A numerical model of the oil injected dual screw com- modeling for performance analysis of rotary screw compressor, in:
pressor has been developed covering both suction and com- Proceedings of the Purdue Compressor Technology Conference,
pression-discharge steps to study the inuence of design Purdue, USA, 1984, pp. 536543.
and operating parameters on the compressor volumetric [9] M. Fujiwara, H. Mori, T. Suwama, Prediction of the oil free screw
and adiabatic eciencies. Heat transfer coecient between compressor performance using digital computers, in: Proceedings of
the Purdue Compressor Technology Conference, Purdue, USA, 1974,
gas and oil has been determined from experimental obser- pp. 186189.
vations in which volumetric eciency decreases with [10] Robert W. Fox, Alan T. McDonald, Introduction to Fluid Mechan-
decrease of inlet temperature. Flow coecients have also ics, second ed., John Wiley & Sons, 1973.
been determined from experimentally obtained compressor [11] N. Stosic, I.K. Smith, A. Kovacevic, Retrot N Rotors for ecient
eciencies. The inuence of operating and design parame- oil ooded screw compressors, in: International Compressor Confer-
ence, Purdue, USA, 2000, pp. 917924.
ters is studied and some quantitative results have been [12] K. Ueno, K.S. Hunter, Compressor eciency denitions. Available
obtained to convert commercially available oil injected from: <www.vairex.com> May 12, 2003.
air screw compressor for helium compression. [13] Huagen Wu, Xueyuan Peng, Ziwen Xing, Pengcheng Shu, Experi-
mental study on PV Indicator diagrams of twin screw refrigeration
Acknowledgements compressor with economizer, Applied Thermal Engineering 24 (10)
(2004) 14911500.
[14] R.P. Taylor, B.K. Hodge, C.A. James, Estimating uncertainty in
The authors are grateful to the Board of Research in thermal systems analysis and design, Applied Thermal Engineering 1
Nuclear Sciences (Department Atomic Energy) for spon- (1) (1999) 317.

S-ar putea să vă placă și