POLITECNICO DI MILANO
Scuola di Ingegneria Industriale
Corso di Laurea Magistrale in Ingegneria Energetica
Experimental investigation on the impact of
materials and lubricants on the performance of a
sliding-vane rotary air compressor
Anno Accademico 2014-2015
Relatore: Ing. Gianluca VALENTI
Tutor aziendale: Ing. Stefano MURGIA
Tesi di Laurea di:
Giacomo FERRARI Matr. 787736
To my grandparents Anna, Elda, Renato and Vito,
because they have been role models for me,
and I am extremely grateful to them.
I
Acknowledgments
First and foremost, I would like to express my gratitude to Ing. Gianluca Valenti for the
experience he has given me, and especially for involving me into this admirable project
of collaboration with Ing. Enea Mattei.
My sincere thanks goes to Ing. Stefano Murgia for his continued help and support with
this thesis. I also would like to thank Filippo for providing me with CAD drawings, Paolo
and Luca for the assistance at the test bench and for the frequent changes in vanes
and lubricants.
I thank Daniele, Ida and Lorenzo for their precious friendship during this study
experience.
I thank my sister Sofia, my brother Giovanni and my grandparents Anna, Elda, Renato
and Vito for their dedicated support.
Finally, I would like to dedicate a special thanks to my parents, Elisa and Ruggero, for
patiently guiding and encouraging me along all my life path.
Desidero innanzitutto ringraziare l’Ing. Gianluca Valenti per l'esperienza che mi ha
trasmesso e, soprattutto, per avermi coinvolto in questo importante progetto di
collaborazione con la società Ing. Enea Mattei.
Ringrazio l'Ing. Stefano Murgia per la sua costante disponibilità e il suo fondamentale
aiuto nella realizzazione della tesi. Un ringraziamento a Filippo per la produzione dei
disegni CAD, così come a Paolo e Luca per l’assistenza al banco prova e per i cambi di
olio e palette.
Ringrazio inoltre Daniele, Ida e Lorenzo per aver condiviso con me questa esperienza
di studio.
Ringrazio mia sorella Sofia, mio fratello Giovanni e i nonni Anna, Elda, Renato e Vito
per il loro instancabile supporto.
Infine dedico un ringraziamento speciale ai miei genitori, Elisa e Ruggero, perché mi
hanno pazientemente accompagnato e sostenuto lungo tutto il mio cammino.
III
Index
ACKNOWLEDGMENTS ............................................................................................... I
INDEX ........................................................................................................................... III
FIGURES ....................................................................................................................... V
TABLES ...................................................................................................................... VII
SOMMARIO................................................................................................................. IX
ABSTRACT .................................................................................................................. XI
CHAPTER 1. INTRODUCTION ........................................................................... 13
1.1 Context ...................................................................................................................... 13
1.2 Sliding-Vane Rotary Compressor (SVRC) .................................................................... 15
1.3 Problem definition .................................................................................................... 18
1.4 Objectives ................................................................................................................. 18
1.5 Methodology ............................................................................................................. 18
1.6 Structure of the thesis ............................................................................................... 19
1.7 Bibliographic Review ................................................................................................. 20
CHAPTER 2. EXPERIMENTAL CAMPAIGN .................................................... 23
2.1 Compressor Equipment ............................................................................................. 23
2.2 Air - Oil Circuit ........................................................................................................... 26
2.3 Laboratory Instrumentation ...................................................................................... 27
2.4 Test Acquisition ......................................................................................................... 28
IV
2.5 Experimental Campaign ............................................................................................ 30
2.6 Vane materials .......................................................................................................... 31
2.7 Lubricants ................................................................................................................. 34
2.8 Pressures ................................................................................................................... 35
CHAPTER 3. DATA ANALYSIS ........................................................................... 37
3.1 Operational conditions (ISO 5167) ............................................................................ 38
3.2 Inlet Conditions ......................................................................................................... 49
3.3 ISO 1217 .................................................................................................................... 50
3.4 Propagation of uncertainties ..................................................................................... 51
3.5 Compatibility Check and Averaged Values ................................................................ 54
CHAPTER 4. SOFTWARE TOOLS ..................................................................... 57
4.1 MATLAB Code “XisoRS” ............................................................................................. 57
4.2 VBA Excel “XisoRS_base.xlsm” .................................................................................. 61
CHAPTER 5. RESULTS AND DISCUSSION ...................................................... 63
CHAPTER 6. CONCLUSIONS .............................................................................. 69
CHAPTER 7. FUTURE WORK ............................................................................ 71
BIBLIOGRAPHY ........................................................................................................ 73
APPENDIX A .............................................................................................................. 75
V
Figures
Figure 1.1 World atmospheric concentration of CO2 and average global temperature
change. (IEA, 2013) ........................................................................................................ 13
Figure 1.2 A taxonomy of different types of gas compressors. ..................................... 14
Figure 1.3 Cross section of a sliding-vane compressor. [3]............................................ 15
Figure 1.4 P-V diagram which represents the compression process. [3] ...................... 16
Figure 2.1 Picture of the experimental compressor outfitting. (Mattei®, 2014) ........... 24
Figure 2.2 CAD Drawing of the pumping unit. (Mattei®, 2014) ..................................... 25
Figure 2.3 Instrumentation rig. (Mattei®, 2014) ............................................................ 28
Figure 2.4 Drawing of the pumping unit. (Mattei®, 2014) ............................................. 31
Figure 2.5 Contact edges of the vane. (Mattei®, 2015) ................................................. 32
Figure 3.1 Passages to calculate performances. ............................................................ 37
Figure 3.2 ISA 1932 with d > 2/3D. (ISO 5167) ............................................................... 39
Figure 3.3 Schemes for iterative computation. (ISO 5167 - 1 - Annex A, 2003) ............ 43
Figure 3.4 Scheme of circuit main sections. .................................................................. 45
Figure 3.5 Scheme of the steps to calculate wet air mass flow rate. ............................ 48
Figure 3.6 Examples of analysis of compatibility. .......................................................... 55
Figure 4.1 Flow chart of the “XisoRS” code structure. .................................................. 58
Figure 4.2 Introductory comments of the VAPsatp MATLAB function. ......................... 60
Figure 4.3 Data collected in XisoRS_base Excel file. ...................................................... 61
Figure 4.4 Compatibility check and t-Student analysis. ................................................. 62
Figure 4.5 Performances Charts worksheet in “XisoRS_base.xlsm”. ............................. 62
Figure 5.1 Cast Iron ISO 1217 volumetric flow rate at 7.5 bar(g). ................................. 64
Figure 5.2 Cast Iron ISO 1217 shaft power at 7.5 bar(g). ............................................... 64
Figure 5.3 ISO 1217 shaft power at 7.5 bar(g): cast iron (CI) and aluminium (Al). ........ 65
Figure 5.4 ISO 1217 volumetric flow rate at 7.5 bar(g): cast iron (CI) and aluminium
(Al). ................................................................................................................................. 66
VI
Figure 5.5 ISO 1217 shaft specific energy at 7.5 bar(g): cast iron (CI) and aluminium
(Al). ................................................................................................................................. 67
VII
Tables
Table 2.1 M111H dimensions. ....................................................................................... 23
Table 2.2 Operating features of the sliding-vane rotary compressor M111H. .............. 24
Table 2.3 Instrumentation List ....................................................................................... 27
Table 2.4 Properties of chosen vane materials. (S. Murgia et al., 2015) ....................... 33
Table 2.5 Properties of chosen lubricants. (S. Murgia et al., 2015) ............................... 35
Table 3.1 Hypothesis over the wet air flow. .................................................................. 46
Table 3.2 Coefficients for Langen’s equations. .............................................................. 47
Table 3.3 Dry air molar composition. ............................................................................. 47
Table 3.4 Direct measurement uncertainties ................................................................ 54
Table 5.1 Variations comparison. .................................................................................. 65
Table 5.2 Mechanical power variation comparison changing lubricants at 7.5 bar(g). . 66
Table 5.3 Performance parameters values in every configuration. .............................. 68
IX
Sommario
I compressori volumetrici, e in particolare quelli rotativi a palette, sono ampiamente
utilizzati nel mondo dell’aria compressa. Come in molti altri settori industriali,
l’innovazione tecnologica degli ultimi anni ha permesso di intraprendere nuove linee di
sviluppo volte alla ricerca di una sempre maggiore efficienza energetica. Nello
specifico, lo scopo del presente lavoro è di studiare sperimentalmente come palette di
materiali differenti (ghisa e alluminio con superficie anodizzata) e come quattro diversi
tipi di oli lubrificanti (caratterizzati da differenti indici di viscosità e diverse
concentrazioni di additivi) influenzino le prestazioni di un compressore a palette di
capacità media. Tali prestazioni sono analizzate attraverso il calcolo di tre parametri
principali: portata volumetrica, potenza meccanica assorbita all'albero motore e lavoro
specifico meccanico. Tali parametri sono calcolati e corretti in accordo a quanto scritto
rispettivamente nelle norme ISO 5167 e ISO 1217. Il calcolo della propagazione
dell’incertezza, dai dati direttamente letti dagli strumenti fino al calcolo dei risultati
finali, e il controllo della compatibilità dei risultati sono eseguiti secondo le norme ISO
5168 e ISO IEC Guide 98 (Guide to the Expression of Uncertainty in Measurement,
GUM). L'elaborazione dei dati per le 400 prove eseguite è svolta usando un adeguato
codice di calcolo scritto tramite il software MATLAB®. Inoltre, la successiva analisi dei
valori medi ottenuti (con un livello di confidenza del 95%) è svolta tramite un foglio di
calcolo sviluppato in ambiente VBA Excel. I risultati mostrano come a livello di media
campionaria i lubrificanti considerati non influenzino apprezzabilmente la portata
volumetrica di aria. Al contrario, si osserva una variazione dello 0.5% della potenza
meccanica con le palette in ghisa e del 1% con le palette in alluminio. In aggiunta,
confrontando i due materiali si notano scostamenti nell’assorbimento di potenza
meccanica compresi tra 1.3% e 2.5%. In conclusione, la combinazione migliore è quella
composta dalle palette in alluminio e dall’olio sintetico poli-alfa-olefine.
Parole chiave: compressori volumetrici a palette, ISO 1217, materiale palette, olio
lubrificante, potenza meccanica
XI
Abstract
Positive-displacement compressors and, among them, sliding-vane rotary machines
are widely used in the compressed air sector. Pursuit of more efficient energy
utilization has become a major goal in this sector as in many other industrial fields.
The aim of the present activity is the experimental investigation into the influence of
two vane materials (cast iron and aluminium with anodized surface) and of four
commercial lubricants (characterized by different viscosity indexes and additives
concentrations) on the performance of a mid-capacity sliding-vane rotary compressor
at five different operating pressures. Performance is characterized by the study of the
volume flow rate, the absorbed mechanical power and the mechanical specific energy
evaluated according to the international standard ISO 5167 and ISO 1217. Propagation
of uncertainties is calculated according to ISO 5168 for direct measurements and to
ISO IEC Guide 98 (Guide to the Expression of Uncertainty in Measurement, GUM) for
combined quantities and for the compatibility check of results. After carrying out more
than 400 tests, averaged values are calculated and analysed through the development
of a MATLAB® program and a proper VBA Excel workbook. The results of this campaign
indicate that the considered lubricants do not affect appreciably the volumetric flow
rate. On the other hand, different lubricants determine a variation of about 0.5 % of
the mechanical power with cast iron vanes and of 1% with aluminium vanes, while
changing the specific material determines a variation of between 1.3% and 2.5%. The
best performance is achieved by aluminium vanes and a synthetic poly-α-olefin
lubricant.
Keywords: sliding-vane rotary compressor, ISO 1217, vane material, commercial
lubricant, mechanical power
13
Chapter 1. Introduction
The present activity is a breakthrough within the industrial-academic collaboration
between Politecnico di Milano and Ing. Enea Mattei S.p.A.®. The company produces
positive displacement sliding-vane rotary compressors (SVRC) for industrial
applications and it occupies a leading position in the field. Especially, the R&D
department is involved in multiple research projects, with energy efficiency
enhancement as one of the main objectives. The aim of this work is the experimental
investigation on the influence of two vanes materials and four different commercial
lubricants on the performances of a mid-capacity sliding-vane rotary compressor. This
study advances in a global context of energy efficiency improvement, but it is also
aimed to a reduction of the costs associated to the vanes production.
1.1 Context
The estimated global electricity use for compressed air accounts for 4-5% of the
industrial total consumption, in particular, in Europe, 100 TWh of electric power per
year are consumed [1]. In “Redrawing the Energy Climate Map” IEA proposes the
implementation of policy measures that can help keep the door open to the 2 °C
target through to 2020 at no net economic cost (Figure 1.1Errore. L'origine
riferimento non è stata trovata.). Within these measures, the most important effort in
favour of emissions savings, should aim at the adoption of specific energy efficiency
Figure 1.1 World atmospheric concentration of CO2 and average global temperature change. (IEA, 2013)
14
measures. [2].
The following section presents how recent technological progresses allow sliding-vane
rotary compressor (SVRC) industry to move in this direction of efficiency. After a brief
overview over the world of compressed air, SVRC technology is generally described, in
order to introduce the objectives of this experimental project.
Compressors overview
Compressed air is a beneficial form of energy in many ways. It is clean and safe, easy
to store and transport, and is very useful for diverse industrial applications: from
operating screwdrivers and similar tools to creating movements and lifting, or for
blowing surfaces clean, moving and cooling materials, food industry, aluminium
foundries and automotive applications.
A compressor is the mechanical equipment that takes in ambient air and increases its
pressure, powered by an electric prime mover. The amount of compressed air being
produced is regulated by the installation of a control system, and the presence of an
appropriate treatment apparatus grants contaminants removal (residual oil,
condensed vapour water) from the compressed air.
In particular, the compressor core unit can vary in type and size from a small one of
2.5 kW to huge systems with more than 250 MW. As shown in Figure 1.2, there are
two basic compressor types: dynamic and positive-displacement.
Dynamic compressors. They impart velocity energy to continuously flowing air
by means of impellers rotating at very high speeds. The velocity energy is
changed into pressure energy both by the impellers and the discharge
Figure 1.2 A taxonomy of different types of gas compressors.
15
diffusers. In the centrifugal-type compressors, the shape of the impeller
blades determines the relationship between air flow and the pressure
generated, and the compression ratio depends on the shaft angular speed.
Positive-displacement. Within this type of compressors, a given quantity of air
or gas is trapped in a compression chamber and the volume which it occupies
is mechanically reduced, causing a corresponding rise in pressure prior to
discharge. At constant speed, the air flow remains essentially constant with
possible variations in discharge pressure. The reciprocating compressor is an
intermittent flow machine that operates at a fixed volume in its basic
configuration through an alternating movement of a piston inside a cylinder.
Rotary compressor is lighter in weight than the reciprocating compressor and
does not exhibit the shaking forces of the reciprocating type, making the
foundation requirements less rigorous. Even though rotary compressors are
relatively simple in construction, the physical design can vary widely.
1.2 Sliding-Vane Rotary Compressor (SVRC)
Sliding-vane compressors are positive-displacement rotary machines which have an
important place in the world of compressed air considering their high reliability. SVRC
incorporates three main components: a stator, a rotor and the vanes (Figure 1.3):
Figure 1.3 Cross section of a sliding-vane compressor. [3]
16
The stator, obtained by single fusion, is the shell which contains the compressor
core and the lubrication channels and reservoir.
The rotor, which is the only rotating element, is mounted eccentrically in a
slightly larger hollow cylinder (so that the rotor is tangent to the stator in one
point), and it has a series of radial slots that hold a set of vanes.
The vanes are thin fins as long as the rotor, which are free to move radially
within the rotor slots as the rotor revolves.
Operating principles
The inner compression chamber is closed by a frontal and a rear lid and it contains the
whole process of compression. While operating, vanes maintain contact with the
stator wall by both a centrifugal force, generated as the rotor turns, and by a bottom
up boost provoked by pressurised air. The space between a pair of vanes and the rotor
and the cylinder wall form crescent-shaped cells. As the vanes cross the inlet port, gas
is trapped inside the cells at the minimum pressure. Air is then moved and
compressed circumferentially as the vane pair moves toward the discharge port,
delivering air at the maximum pressure, demanded by utilities.
The rotational shaft speed is an important parameter to be taken into account during
the machine design. In fact, the centrifugal force acting on the vanes is strictly
dependent on the shaft speed of the compressor. Consequently, the quality of the
contact between the vanes and the stator is affected by the variation of this
Figure 1.4 P-V diagram which represents the compression process. [3]
17
parameter. A low shaft speed is not able to grant enough air-tight between the top of
the vane and the stator, generating leakages between contiguous cells. At very high
rotational speed, the thin lubricant layer which wraps the vanes risks to be broken.
Therefore, wear rises over materials, causing excessive power losses due to friction.
For efficient compression to take place, the port location must be matched to the
pressure ratio dictated by the application. Figure 1.4 shows an indicator diagram of a
compression cycle. If the port has been optimized for a ratio of , the
compression line is a smooth curve from point 1 to 2. If the external pressure is higher
than the pressure for which the port was designed, so that , then when
the port opens at point 2, discharged air will return to the compressor from the line
and must again be expelled from the compressor. This energy waste is represented by
the red area to the left of the line 0-2. Conversely, if the external pressure ratio is
lower than the pressure ratio for which the port was cut, where , then
the gas will be overcompressed to point 2 and when the port opens, it will expand to
point u. The lost energy is represented by the red area to the right of the line u-2. [3]
Lubrication
As mentioned, this type of compressor must have an external source of lubrication. A
pump is not required for this injection because the pressure of the oil after separation
from the air is sufficient to be re-injected. Oil acts during the compression as:
Lubricant: oil controls the tribology of the compressor: wear, friction and
lubrication of mechanical parts. Actually, the rotor spins into two bushes
which need to be lubricated. As well, the friction among rotor, vanes and
stator affects performances and can provoke some material consumption.
Sealing agent: radial gaps between vanes and the stator, and axial clearances
between vanes and stator need to be filled in order to reduce losses in airflow.
Thermal ballast: during the real compression, heat is transferred to the lubricant
which remains almost totally in the liquid form, which has a great thermal
capacity (thanks to its high density and heat capacity) and large exchange
surface. This liquid significant quantity mitigates the gas temperature rise and,
hence, its compression work.
By absolving these three functions, the oil temperature is induced to increase, as a
consequence of heat exchange from wet air during the compression, worn and friction
of the vanes, sliding and bouncing along rotor and stator surfaces. Lastly, lubricant
executes a protective action on the metallic parts of the compressor, covering them
and, therefore, avoiding corrosion.
18
1.3 Problem definition
As already mentioned, new strategies are undertaken in order to pursue more
efficient power consumptions, rather than the well-known solidity. Therefore, there
could be two possible paths towards improving efficiency:
The first path, more thermodynamical, should operate on the thermal cycle of
compression by means of multiple intercooled stages, in order to maintain a
low air temperature. This presents plant difficulties due to the system’s
complexity. Previous works have produced changes in the size of the oil drops
injected into the compression chamber in order to increase the heat exchange
process
On the other hand, the mechanical path should make it possible to improve the
performances of the machine without varying the cycle of the gas (wet air)
along the compressor. In this dissertation two parallel roads have been taken:
the first one consists in using a different lubricant during the process of air
compression, the second one consists in using a different vane material for the
purpose of the compression.
1.4 Objectives
This investigation aims to study experimentally the impact of two vane materials (cast
iron and aluminium with anodized surface) and of four commercial lubricants
(characterized by different viscosity indexes and additives concentrations) on the
performance of a mid-capacity sliding-vane rotary compressor at five different
operating pressures. Analysis is carried out of 400 experimental tests on a SVRC.
In particular, performance is characterized by the study of the volume flow rate, the
absorbed mechanical power and the mechanical specific energy. Therefore, the main
objectives are the identification of:
possible connections between the SVRC performances and the two different
tested materials;
connections between the SVRC performances and the commercial lubricants
which have been selected for tests;
1.5 Methodology
Tests vary in lubricant, vane material and delivery pressure. They are divided into 40
cases; each of them is replied 10 times in order to grant the repeatability and to
increase the confidence of results. Performance parameters (flow rate, mechanical
19
power, and mechanical specific energy) have been calculated following a standard
procedure which is consistently applied within Mattei®.
Hypothesis is done to proceed with calculation of the air flow through the circuit and
at different conditions. Wet air is considered as an ideal mixture of gases: dry air and
vapour. Dry air is considered as an ideal gas and vapour is considered to follow pure
water behaviours described in IAPWS Formulation.
In order to calculate wet air flow through the compressor, ISO 5167 is applied at the
flow measurement device. This Standard approaches the problem of finding the mass
flow rate through a differential pressure. Afterwards, operational flow rate is
converted into Free Air Delivery (FAD, at the inlet air conditions), before being
standardize at ambient ISO 1217 conditions together with the mechanical power. This
latter conversion permits the comparison between the values of performance, as data
were collected at the same conditions of temperature, pressure and relative humidity.
Uncertainties are calculated for direct measurements according to each instrument
technical datasheet. Propagation of uncertainties is calculated according to ISO 5168
and to ISO IEC Guide 98 (Guide to the Expression of Uncertainty in Measurement,
GUM) for combined quantities and for the compatibility check of results.
To implement the whole calculation procedure, a MATLAB® program is compiled
within this study, starting from the structure of earlier versions, which were developed
within preceding thesis projects [4] [5]. Its functions have a standard input and output
format which simplifies the organization of the code. Moreover, uncertainty
calculations are added within each single function. Lastly, a VBA Excel spreadsheet
workbook is created to collect all the results, to execute the compatibility analysis and
to build graphs of averaged values of performance.
1.6 Structure of the thesis
The following sections are structured as below:
chapter 2 describes in detail the setting up of the experiment. The test bench,
the measuring instruments (their working principle and their uncertainties),
and the data collection procedure are characterized. Furthermore, the
experimental campaign is described: the mid-size compressor utilised to
conduct the tests, the two different materials (cast iron and aluminium with
anodized surface), and four commercial synthetic lubricants (called A, B, C and
D throughout this dissertation).
chapter 3 explains the hypotheses, the Standards and the formulations which
are applied to analyse data and the uncertainty calculation procedure.
20
chapter 4 describes the MATLAB code and the VBA Excel workbook, both
developed during this thesis project.
chapter 5 reports performance results from data analysis and comments on
performance trends. Also some considerations about flow rate measurement
variability with atmospheric conditions are proposed.
chapter 6 presents the conclusions of the project, evaluating whether or not the
objectives have been achieved or not.
chapter 7 proposes suggestions for possible future work and development on
the Mattei® sliding vane compressor.
1.7 Bibliographic Review
Three previous thesis projects are taken into consideration in order to proceed with
this dissertation. In particular, this work resumes the results obtained by Recalcati in
terms of methodology and in terms of vane materials
P. Calvi – Experimental investigation on the effetcs of the oil injection through
an atomised-oil injection system, 2011-2012
In these work, results of a comparison between convectional solid-flow injection and
innovative atomised-oil injection system tested with an experimental rig are shown:
atomization should generate small oil droplets to increase heat exchange and vane
cooling, leading to a decrease of compression work. Nozzles cause the compressor a
saving of energetic consumption and a specific energy increment when they’re
positioned in the closest position to the zone of maximal compression. Keeping oil
nozzles and adding a gear pump to pressurize the oil circuit an increase of air flow is
obtained and in the first part of the compression isothermal process is reached, but
energetic consumption is higher due to the pump. [6]
M. Miggiano – Experimental Campaign for the development and the study of
sliding-vane compressor with modified stator geometry, 2012-2013
In this thesis performance of a modified positive-displacement vane compressor is
evaluated optimizing the stator geometry and a useful calculation tool to process data
is developed. Five versions of the compressor were considered and their performances
compared to estimate the improvement due to the new design of the stator and the
use of a larger intake valve, a wider intake port and a larger exhaust. The volume flow
rates were calculated by using a MATLAB® ad-hoc developed code, taking into
consideration the ISO1217 and ISO5167-3 standards. This code also allows to
determine the specific energy of the compressor and to evaluate the error of the
measurement system. During the tests, the used measuring chain proved some limits
21
concerning the measurement accuracy, therefore, a new highly accurate configuration
of measurement of the flow rate was studied, with the error reduced to 1.55%. [4]
M. Recalcati – Experimental investigation of sliding-vane compressor
performances with diverse vane materials, 2013-2014
This thesis shows how vanes made of different materials (kevlar, glass fiber,
aluminium, drilled iron and iron) affect a displacement compressor performance.
Volumetric flow rate, required shaft power and shaft specific energy were first
calculated and then corrected following ISO 1217 [2] and ISO 5167 [3] guidelines. Data
spreadsheets were later fed to an appropriate programming script written to work
with both Matlab® and GNU/Octave. Results show that the required shaft specific
energy sample average is roughly the same for glass fiber, aluminium and iron, while
kevlar and drilled iron display more significant differences. This work concludes that
simply reducing a vane weight not only is insufficient to improve a compressor
performance, but it can even lead to worse performances. [5]
23
Chapter 2. Experimental Campaign
The laboratory setup, used to carry out the tests within this investigation, includes the
experimental rig with the compressor bench (main unit + electric motor), represented
within the scheme in Figure 2.1. Along this section are firstly presented the air flow
measurement circuit and an oil cool-down circuit. Then, a description of the
instrumentation used to acquire data is proposed and, more in detail, the compressor
equipment. Finally, the data acquisition procedure and the different test
configurations are presented.
2.1 Compressor Equipment
The compressor bench is composed of a compressor with an air-oil candle type
separator, an electric motor mastered by a control panel and finally a radiator
followed by a condensate separator.
Compressor
The device being tested is a Mattei® M111H compressor. Its characteristics are
summarized below:
Packaged compressor. This is a compressor which, according to ISO 1217,
integrates the power source and its transmission, the oil tank and the oil
separation and the air-condensate separator. It is fully piped and wired
internally, including auxiliary items of equipment (in this case only the cooling
radiator) [7].
ERC 22 L outfitting configuration. It is not soundproof and it has a 8 bar(g)
maximum pressure, 7.5 bar(g) operational pressure and requires 22 kWe of
power supply.
This compressor is selected for this project in virtue of its:
mid-sized dimension, which would allow an extension of results to any of
Mattei® compressors (whatever the size);
Table 2.1 M111H dimensions.
Model Length Width Height Weight
[mm] [mm] [mm] [kg] ERC 22 1580 580 970 325
24
good performances in terms of specific energy (around 6.3 kW min /m3);
good solidity and reliability, due to a long experience of test and operation.
The core unit includes the stator and the rotor of the compressor, as described in the
introduction for a standard SVRC (1.2Sliding-Vane Rotary Compressor (SVRC)). The air
suction section is located on the frontal lid, while the delivery section is located on top
of the chamber, linked to the separation unit.
The removal of the frontal lid allows the easy replacement of vanes, while lubricants
are discharged through a lateral valve and refilled from the top of the compression
chamber, in order to set the desired test condition. Furthermore, during every
lubricant replacement, the oil circuit is cleaned thoroughly to prevent contaminations.
Motor
The compressor is coupled with a 22 kWe asynchronous 3 phase motor through a
Figure 2.1 Picture of the experimental compressor outfitting. (Mattei®, 2014)
Table 2.2 Operating features of the sliding-vane rotary compressor M111H.
Rated flow rate, l/min* 3500
Rated working pressure, bar(g) 7.5 Rated power, kWe 22
Nominal rotor speed, rpm 1500 *Standard conditions as per ISO 1217 Standard acceptance
25
flexible joint which grants good alignment and low power absorption.
The whole machine supplies constant air flow at the nominal rotor speed of 1500 rpm.
Such speed is set by the 50 Hz (4 electric poles) frequency of the electric power grid.
Oil Separation
The outside of the compressor shows two stacked cylinders (Figure 2.1Figure 2.):
The lower one contains the pumping unit and acts as primary mechanical oil
separation and, at the same time, as oil tank (Figure 2.2). The air intake enters
the frontal lid, whereas the channel of cold oil flows in through the lateral
surface of the stator. When the air-oil mixture exits the chamber, the flow
meets a labyrinth path which makes it decelerate and change direction. In this
way, oil deposits and it can be easily directed toward the radiator, whereas
the air flow heads toward the upper cylinder.
The upper cylinder, placed above the main body of the machine, consists of 3
coalescent filters for the secondary oil separation. They absorb the residual oil
vapours from the air flow. Finally, the air duct leaves the upper cylinder after
the filtration step, heading towards the radiator.
Figure 2.2 CAD Drawing of the pumping unit. (Mattei®, 2014)
26
Hence, oil separation is distributed over multiple steps: the first one, which is
mechanical, is the more substantial, while the second one is reduced.
Radiator
As just mentioned, the experimental version of the machine is equipped with an
aluminium air-radiator. This unit is divided into two sections under the same fan
airflow: one for the oil cooling and the other one for the air cooling. The fan is
powered by the same electric motor as the compressor, which is connected to the
power grid. It also slightly serves as a motor cooler, as its generated air-stream is
directed toward the motor shell.
Moreover, the radiator is equipped with a condensate separator (the first along the air
circuit) to discharge water. At this point of the circuit, water forms due to the heavy
compression stage. At the same time, temperature increment during compression is
not enough to permit the water vapour to keep its gas phase: it condensates, thus
becoming liquid water. This process is what makes the hypothesis of saturated air
possible (100% of Relative Humidity).
2.2 Air - Oil Circuit
BWet air enters from the frontal lid into the compression chamber, it is compressed
and it flows out mixed with lubricant towards the mechanical separation. After this
step, air goes through the coalescent oil filters and, almost without oil, it flows
towards the radiator. At this point, there is a first condensate separator, a compressed
air tank and a second condensate separator before the flow measurement pipe.
As mentioned in section 1.2, oil acts during the compression as lubricant, sealing agent
and thermal ballast. These three functions induce the temperature of the oil to
increase, as a consequence of heat exchange, worn and friction of the vanes, which
slide and bounce along the stator surface.
Oil starts its path within the machine from the injection into the compressor chamber
through an injection case, which has 5 holes axially arranged (standard Mattei®
injectors), which make the oil enters the compression chamber radially, mixing with
air. Air and oil mixture at very different temperature condition causes a loss within the
efficiency of the compression. Thus, holes are located in the last compartments (250°
from the suction section in the rotation direction), where air is already at high
pressure and high temperature.
Leaving the compression chamber, the high temperature air-oil flow gets into a cavity
around the pumping unit, filled with maze tunnels. Here, a mechanical separation
27
takes place and for gravity action, oil sinks to the bottom whereas air occupies the top.
Oil is drained out of the interspaces and send into the radiator, where it is cooled
down by the fan.
However, some oil gets trapped in the air flow and it requires a further separation.
This happens in the upper cylinder of the machine, through three coalescing oil filters
which release air almost without oil (1-3 ppm, with 1ppm= 1.2 mg/m3 at Free Air
Delivery conditions, F.A.D.). The oil held by these filters is redirected directly to the
chamber, with no refrigeration.
2.3 Laboratory Instrumentation
In order to verify experimentally the performances of materials and lubricants, a test
rig based on a commercial 22-kW SVRC was assembled. The experimental rig, shown
in, employs the necessary instrumentation to measure air temperatures and pressures
along the compression, the delivered volume flow rate, lubricant temperatures and
pressures along the process, and, finally, shaft torque and rotational speed. The
experimental setup is design to evaluate the compressor performances while varying
the delivery pressure.
In particular, the mechanical power is calculated as product of shaft torque and
rotational speed measured using a flange torque meter installed between the
compressor and the electric motor. A Kistler 4504B1KB1N1 torque meter (Full Scale of
1000 Nm, Accuracy ±0.5% FS) is controlled by the evaluation instrument Kistler CoMo
Torque Type 4700A.
The electronic data acquisition is performed thought the National Instruments cDAQ-
9178 wired to a personal computer. LabView Signal Express 2011® is used to carry out
the measurements.
Table 2.3 Instrumentation List
Instrumentation Measured quantity Accuracy Model Manufacturer
Barometer p_barometric ±1 mbar 102 Fischer Manometer p_chamber ±0.05 bar - Spriano Hygrometer UR ±1% Supratherm Barigo Thermocouple T_in ±0.1°C T Model Tersid Thermocouple T_chamber ±0.1°C T Model Tersid Thermocouple T_device ±0.1°C T Model Tersid Water column p_static ±1mmH2O - - Water column p_differential ±1mmH2O - - Torque meter Torque ±0.1 Nm 4504B1KB1N1 Kistler Digital RPM Meter Shaft speed ±1 rpm - IDF
28
Shaft speed value has been determined through a Digital rpm meter with an accuracy
of 1 rpm.
A Fischer barometer and a Spriano analogical Manometer have been used to evaluate
respectively room pressure and chamber pressure.
Temperature values have been estimated with three thermocouples located
respectively:
at the suction port (input temperature);
at the oil separation (chamber temperature);
at the pressure differential device (device pressure).
Finally, to gauge relative humidity a Barigo Hygrometer has been used.
2.4 Test Acquisition
The collection of data follows a very standard procedure described below. It consists
of general steps, typical for every compressor analysis, and specific steps for this test
bench.
Figure 2.3 Instrumentation rig. (Mattei®, 2014)
29
Instrumentation check
1) Before turning the apparatus on, it is highly recommended to check that all the
instrumentation is accessible and ready to operate as required.
2) All the electric devices have to be plugged in: computer, temperature display,
torque/rpm viewer. They in turn have to be connected to the corresponding sensors:
thermocouples and torque meter.
3) The presence of water into the columns of the pressure gauge has to be double
checked. The of connection of this latter to the pressure differential device has to be
checked as well: the upstream intake to the static column, and the downstream intake
to the differential one. Both of them have to be closed with the respective valve in
order to avoid the spillage of water when powering the system.
Ignition
4) First of all, the recirculation valve at the suction section has to be open, to allow
the open circuit operation at the start.
5) Then it is possible to plug in the measurement system (220V switch), taking
care to set to zero the electric devices.
6) At this point also the 380V switch can be turned on, wait for the digital control
panel to be ready, and finally push the power on button.
7) Let the compressor run for 10-20 seconds and then the recirculation valve has
to be turned off. It is good practice to set the throttling valve at the maximum
opening, in order to reduce the risk to reach a very high pressure as long as the
machine is cold.
8) Next step is to regulate the throttling valve to obtain the required backpressure,
e.g. 7.5 bar-g.
9) Now it is possible to open the water column pressure gauge valves.
10) When the speed condition is stable (after about 1h) at the previously set
backpressure, the chamber temperature should be constant (or very slowly fluctuating
around a value). It is now possible to acquire test data, reporting them onto a proper
“Test Reports” paper, or directly in the input Excel file for the MATLAB code “RnD”.
Shutdown
11) Before turn off the power, it is good practice to open again the throttling valve
down to the minimum chamber pressure. Then, also the recirculation valve has to be
opened.
30
12) When the chamber pressure reaches the 1.5 bar-g is it possible to turn the
power off from the digital control panel.
Rig rearrangement
13) All the valves between the differential pressure device and the water columns
have to be closed.
14) It is possible to turn off the 380V switch.
15) After the computer shut down, also the 220V switch can be turned off.
2.5 Experimental Campaign
This investigation is based on an extensive experimental campaign. Research is carried
out involving 400 tests, performed on the Mattei® M111H Compressor. Tests are
performed combining:
2 types of vane materials: cast iron and aluminium;
4 types of lubricants: A, B, C, D;
5 values of pressure: 6.5, 7, 7.5, 8, 8.5.
Each combination of material, lubricant and pressure is repeated 10 times in order to
obtain the repeatability for all the 40 possible conditions. For each condition, tests are
carried out on at least two different days (from 9:00 to 18:00), so that one may trace
the average behaviour of the machine, independently from ambient pressure,
temperature or humidity conditions at different times of the day.
In order to delete the effect of any correlation between tests, their sequence is
randomized. The only fixed parameters are:
Chamber operating temperature of about 85°C
Nominal Rotor speed of about 1500 rpm
Modifications of parameters are introduced by operating on the compressor bench
(interchanging vane materials and lubricants) and on the pressure regulating valve
(placed immediatly after the air pressurized tank).
31
2.6 Vane materials
Besides the rotor, vanes are the only part of the compressor subject to motion during
the whole compression process. For this reason, in addition to their reduced thickness
(of about 5 mm, Figure 2.4), they are the most critical components of a SVRC and they
are among the main objects of study in this field.
Due to the force system (in Figure 2.5) that generates during the operation of the
compressor, vanes are exposed to:
Friction, of the vane which slides along the slot (two contact edges along the
rotor section) and hits the stator shell with its radius of curvature (R = 9.5
mm).
Wear, as a result of friction, causes the adaptation of the vanes to their slots
and to the stator contact during operation. In particular, the most significant
abrasion can be noticed on the back of the vane (due to the back of the vane
sliding on the rotor upper edge) and on the top of it (where the vane hits the
stator). Wear leads to an auto-adjustment of the contact resistance. This
phenomenon ensures less contact between surfaces (vanes-rotor, vanes-
stator) and after some time of running test, performances start improving: in
particular, the power consumption decreases by about 2%.
Figure 2.4 Drawing of the pumping unit. (Mattei®, 2014)
32
In this specific compressor (M111H), there are 7 double vanes, making a total of 14
“half-vanes”. As a consequence of its mid-size, double vane configuration is necessary
to avoid rotation or flexion of the vanes within each slot. On the other hand, the
double vane configuration presents one more leakage source in between the two half-
vanes, which can produce a loss in flow rate (for some more elastic material it is
possible to apply a single vane through the entire rotor).
In order to try to improve the impact of the vanes on the performance parameters of
the compressor (flow rate and power consumption), a different vane material is
tested, rather than the standard one, so that the two options are:
cast iron vanes (CIV), standard
aluminium alloy vanes (AlV), non-standard
Figure 2.5 Contact edges of the vane. (Mattei®, 2015)
33
The main features of the chosen materials are listed in (Table 2.4).
Cast iron
Cast iron is primarily composed of iron (Fe), carbon (C) and silicon (Si). Its structure is
crystalline and relatively brittle but overall it offers better mechanical properties than
aluminium. In fact, it is readily machined and, additionally, the machined surfaces are
resistant to sliding wear thanks to their hardness.
This is the most adopted material for this type of compressor because of its surface
hardness, which allows a good worn resistance. Furthermore, it has low coefficients of
expansion, so that gaps are minimized and during transient phases (especially
powering the compressor) leakages are reduced. For these reasons this material is
usually used to produce the vanes for SVRC.
Furthermore, cast iron has:
high thermal conductivity, it allows good heat transmission between contiguous
air volumes (more uniform temperature along the vane section);
low modulus of elasticity, it makes vanes brittle and fragile and therefore less
subject to deformation under stress;
ability to withstand thermal shock, this property is useful for fast ignition stages,
during which temperature increases greatly in a short time.
Aluminium
On the other hand, aluminium is remarkable because of its low density. As the weight
of the vane directly affects the friction losses of the SVRC, this dissertation investigates
experimentally the effect of the reduction of the vane weight on the compressor
efficiency.
Table 2.4 Properties of chosen vane materials. (S. Murgia et al., 2015)
Properties Cast iron Aluminium alloy
Superficial treatment None Anodization (20 μm) Density, kg/m3 7200 2700
Specific heat*, J/kg/K 460 920
Thermal conductivity at 100°C, W/m/K 48.5 180
Coefficient of expansion*, μm/m/K 11.7 24
Modulus of Elasticity, GPa 120 70
Weight of a single vane 175 65
Roughness index, Ra 0.38 0.36
*Between 20°C and 100°C
34
Other characteristics of aluminium are:
a good thermal and electrical conductivity, higher than cast iron;
a lower modulus of elasticity in comparison with cast iron, which makes this
alloy highly frigile;
a moderately high coefficient of expansion (aluminium vanes need longer axial
gap between them and the lids)
Moreover, the aluminium alloy used during the investigation is subjected to an
improving of its mechanical properties using a finishing superficial treatment called
anodization. This treatment (invented in 1923) consists in an irreversible
electrochemical process, which causes a thin oxide coating to develop on the material
surface. This oxide coating improves the mechanical characteristics of the vane surface
and avoids corrosion. Anodization also improves also the surface hardness and, thanks
to the properties of the superficial oxide coating, increases the affinity with lubricants.
The aluminium vanes used in the experimental investigation are characterized by an
oxide coating about 20 μm thick.
2.7 Lubricants
Almost all positive-displacement air compressors require oil to:
Cool, subtracting heat within every single volume of compression produced and
therefore acting as thermal ballast;
seal gaps, preventing from leaking in between two contiguous air volumes;
lubricate internal components (vanes and brushes), reducing wear and friction
between the moving parts of the compressor.
A correct lubrication ensures the reliability of the equipment, actuating regulation
systems. It prevents corrosion and wear, by protecting internal metal parts, and
contributes to a reduction in energy consumption. Finally, it filters and cleans,
removing from particles contained in the incoming air.
During the investigation, 4 commercial lubricants have been considered and their
effects on the SVRC performance have been evaluated. A sample for each type of oil
has been analysed by an Oil Quality Laboratory. The formulations and properties of
the considered lubricants are reported in Table 2.5:
A, an ISO-VG 68 diester-based synthetic lubricant, usually used on the SVRC
B, an ISO-VG 100 Poly-α-olefin synthetic lubricant
C, an ISO-VG 100 diester-based synthetic lubricant
D, another ISO-VG 100 diester-based synthetic lubricant with no additives
35
Lubricant viscosity affects the performances of air compressors in several working
conditions: for example, viscosity determines the gap clearance (between rotor and
lids, between the vanes and the lids, ...). With appropriate viscosity, equipment:
can be started in low-temperature environments (typically in winter or in cold
locations)
can be kept running in high-temperature conditions (for example in a hot
environment)
The four lubricants are selected considering:
the operating temperature, which is around 85°C inside the compression
chamber: oil viscosity is around 10 Centistokes(cst) at 100°C for the four
lubricants (correspondent to 0.1 Poise);
the speed at which vanes are moving with respect to the rotor and the stator;
the load upon the compressor components.
This is because properties of lubricants also influence the tribology of the SVRC, which
considers the interaction of surfaces in relative motion as well as the effects of
friction, wear and lubrication. Ultimately, focus is placed on the relationship between
the utilization of a specific lubricant and the compressor energy consumption. On the
contrary, the durability of lubricants and therefore their performances are not
evaluated within this investigation, but it could be a possible future project.
2.8 Pressures
For each combination of vane materials and lubricants, tests at 5 different pressures
have been executed, between 6.5 and 8.5 bar(g) by a 0.5 step, as previously reported.
The M111H compressor has 7.5 bar(g) as optimal operational pressure. Nevertheless,
an overview showing compressor performances under different pressure values is
Table 2.5 Properties of chosen lubricants. (S. Murgia et al., 2015)
Properties Test Method A B C D
Type - Synthetic Synthetic Synthetic Synthetic Base - Diester Poly-α-olefin Diester Diester
Additives concentration - High Low Medium none
Viscosity, cSt @ 40°C ASTM D7042 68 91 96 95
Viscosity, cSt @ 100°C ASTM D7042 10 14.8 10.7 9.2
Viscosity index ASTM D2270/ISO 2909 120 170 96 63
Total acid number, mg KOH/g ASTM D664 0.17 0.03 0.18 0.11
Density, kg/m3 ASTM D7042/ISO 12185 951 849 957 954
36
aimed at highlighting other possible trends of the performance parameters (such as air
flow rate, absorbed power and specific energy).
37
Chapter 3. Data Analysis
After data collection, the entire database is created using the MATLAB “XisoRS”
developed during the analysis stage. The output of the code returns performance
values for each test, within fundamental parameters:
volume flow rate of compressed air
mechanical power
mechanical specific energy (which is the” power-flow rate” ratio)
at ambient inlet standard ISO 1217 conditions:
Inlet air pressure: 1 bar(a)
Inlet air temperature: 20 °C
Relative water vapour pressure: 0 bar
In particular, air volume flow rate is firstly determined considering the standard ISO
5167, and is converted under standard conditions according to standard ISO 1217.
Finally, the uncertainties of the values of wet air flow rate, power consumption and
specific energy were calculated in accordance with ISO IEC Guide-98 (GUM 1995).
This calculation procedure makes it possible to obtain resulting values of performance
parameters which can then be compared with different configurations or different
compressors. Starting ambient conditions of temperature, pressure and humidity
(operational conditions) are not aligned between tests. For this reason ISO 1217
provides a “correction” method in order to standardise tests and make them
comparable with each other. Passages of this conversion are represented in the table
in Figure 3.1.
Figure 3.1 Passages to calculate performances.
38
This chapter describes in detail the calculations for the sections encircled in red
dashed line in Figure 3.1:
Operational conditions (ISO 5167)
Inlet Conditions (FAD)
ISO 1217 conditions
3.1 Operational conditions (ISO 5167)
Operational conditions are conditions under which each single test is performed.
Hence, they can change from one test to the next even during the same day, for
example: form midday to late evening. However, calculation of power and flow rate in
operational conditions (according to ISO 5167) is the starting point to achieve their
“corrected” values under the ambient inlet standard ISO 1217 conditions.
Power
In particular, mechanical power, is the power absorbed from the compressor shaft,
calculated as follows:
(1)
where:
C, is the motor torque, measured through the torque meter;
, is the shaft angular speed, measured through the digital RPM meter;
Flow rate
Flow rate in operational conditions is the actual quantity of compressed air passing
through the flow measurement device. It could be directly measured through
appropriate mass/volumetric flow meters (for example: vortex flow meter) or through
the procedure specified within the standard ISO 5167, as in this investigation.
ISO 5167 is the standard which regulates the calculation of the mass flow rate of a
fluid which flows along a conduit, suggesting simplified industrial methods. The latter
require the utilisation of pressure differential devices (orifice plates, nozzles and
Venturi tubes) inserted in circular cross-section conduits running full.
39
Hypotheses, under which ISO 5167 is applicable, are considered satisfied. In particular,
they are:
the stability of a subsonic flow throughout the measuring section of the
pressure differential device;
the possibility to consider the fluid as single-phase;
the fluid may be either compressible or considered as being incompressible;
the no-pulsation of the flow, in order to avoid the inconsistency of the flow
measurement;
the presence of a fully developed flow.
ISO 5167 consists of four parts, however, only part 1 and part 3 are useful in these
project conditions:
Part 1, gives general principles and requirements of measurement and
uncertainty that are to be used in conjunction with Parts 2 to 4.
Part 3, specifies ISA 1932 nozzles, which is the differential pressure device
utilised within this investigation.
Figure 3.2 ISA 1932 with d > 2/3D. (ISO 5167)
40
Figure 1 shows the cross-section of an ISA 1932 nozzle at a plane passing through the
centreline. The nozzle consists of a convergent section with rounded profile, and a
cylindrical throat.
There are specified limits of pipe size and a Reynolds number within which this type of
nozzle should strictly be used:
;
( where is the diameters ratio across the device d1/d2 and it is
equal to 0.8 for M111H test rig);
(Reynolds number for );
(dimensionless roughness for = 0.8);
.
ISO 5167 is based on the application of Bernoulli’s principle to a close pipe segment.
Hypotheses under which Bernoulli’s principle is applicable are:
One-dimensional flow;
Steady state;
Homogeneous fluid;
Horizontal conduit;
No heat exchange with surroundings (adiabatic condition);
Compressible flow.
Assuming that all the assumptions are satisfied, it is possible to consider the
conservation of energy in between the two sections:
Upstream, which has diameter d1 (it coincides with pipe diameter D);
Downstream, which has diameter d2 (it corresponds to the throat diameter).
They are respectively located before and after the differential pressure device. Hence,
the energy balance can be written as:
(2)
where:
p1 is the static absolute upstream pressure;
p2 is the static absolute downstream pressure;
w1 is the upstream fluid velocity;
w2 is the downstream fluid velocity;
ρ is the fluid density at the measurement device.
41
For the continuity equation it is:
(3)
where:
is the upstream (pipe) section;
is the downstream (throat) section.
Equations can be combined to give the volumetric flow rate, , written as:
(4)
The streamline constriction into a generic throttling device generates a decrease in
static pressure. The more volumetric flow rate circulates, the bigger the pressure
difference is. The equation represents the ideal volumetric flow rate, considering the
absence of rotational flow which could compromise the uniformity of velocity
distribution and the adiabaticity of the flow. Actually, the hypothesis of one-
dimensional flow (undisturbed) is correct only in sections which are sufficiently distant
from the upstream and the downstream sections. These two “sufficiently distant
sections” are not precisely definable as the contracted flow length changes
progressively varying the flow rate and the device diameter.
In order to take this effect into account, the Standard ISO 5167 requires lateral
pressure tappings in close proximity of the upstream and downstream sections and
provides for two adjustment coefficients:
, which is called “flow coefficient”
, which is called “expansibility factor”
The first one relates the actual flow rate to the theoretical flow rate through the
device (whether the flow is incompressible or compressible), while the second one
considers the possible fluid density variation through the device (given a compressible
flow). For that reason, the flow coefficient, α, is defined as follows:
(5)
where is the discharge coefficient. In case an ISA 1932 device is utilised, is given by
the equation (7):
42
(6)
where Reynolds number, , is the dimensionless parameter expressing the ratio
between the inertia and viscous forces in the upstream pipe, and it is defined as:
(7)
The expansibility factor, , in case of ISA1932 with p2/p1 0.75, is calculated by means
of the following equation:
(8)
where:
- , is the heat capacity ratio of the fluid at the differential pressure device;
- , is the pressure ratio
At this point the mass flow rate, qmassic, can be obtained from:
(9)
And combining equation (5) and (10) it is possible to write the mass flow rate as
follows:
(10)
Equations (7), (8) and (11) are strictly interdependent, so that the only procedure to
solve this system is to undertake an iteration process, as illustrated in Figure 3.3. After
calculating the invariant , ISO 5167 suggests that = ∞ as the “first guess” value
(turbulent flow). Starting from this value it is possible to calculate the constant and
the “first guess mass flow rate”, through which “second guess ” can be calculated,
and so on until the precision is reached (fixed at 10-5 within this investigation).
Furthermore, from the equations above, it is possible to understand the necessity to
know:
density,
43
viscosity,
heat capacity ratio,
of the fluid at the upstream pressure tapping.
As there is no possibility to directly measure the fluid density directly, it is calculated
by using the appropriate equation of state from the knowledge of the absolute static
pressure, absolute temperature and composition of the fluid at that location. Even
viscosity depends on the composition of the fluid, its temperature and pressure, while
heat capacity ratio depends on composition and temperature.
In particular, the air flow rate is considered to be wet air, namely a two-component
system formed by:
dry air, treated as an ideal gas;
water vapour, treated as perfect fluid;
Figure 3.3 Schemes for iterative computation. (ISO 5167 - 1 - Annex A, 2003)
44
As a result, their combination is treated as an ideal mixture of gases. Single
components of an ideal mixture do not interact with each other, so that mixture
extensive properties are the sum of single components extensive properties (e.g.: the
number of moles = + ). Likewise, pressure of the mixture is equal
to the sum of the component partial pressures, which are the pressures that
components would have at the same temperature, if they were occupying the whole
mixture volume. As a consequence it is possible to apply Dalton’s law to the mixture:
(11)
This grants the validity of Raoult’s law:
(12)
Moreover, in an ideal mixture, the behaviour of water during the saturation process is
not modified by the presence of dry air, so that:
(13)
Data collected within each test, makes it possible to outline the thermodynamic status
of the wet air flow at:
Input. In this section of the circuit a thermocouple detects , a barometer
measure the ambient pressure, considered equal to , and an hygrometer
gauges the environment relative humidity, .
Output. Temperature is detected inside the compression chamber ( ) in
order to keep the oil temperature monitored. Also pressure ( ) is
measured inside the last cell at highest pressure through a manometer.
Upstream. A water column is connected to the differential pressure device. One
tapping is located before the device throat and it measures .
Downstream. The other tapping of the water column is connected to the
downstream internal surface after the differential pressure device, in order to
detect .
Delivery . 5 diameters downstream the differential pressure device is, according
to ISO 5167, a last thermocouple is located in order to evaluate .
The condensate separator section is undefined in temperature and pressure. However,
it would be useful to know the thermodynamic status of wet air at the separator in
order to estimate the quantity of condensate which the separator removes from the
main wet air flow.
45
For ISO 5167 the primary device temperature shall preferably be measured at least at
the distance of 5 D, downstream the device. It is possible to extend this limit to the
separator section, neglecting temperature losses between the differential pressure
device and the separator:
(14)
Another strong hypothesis is required in order to know the pressure at the
condensate separator. Although actually there is a pressure drop in between the
compressor chamber ( ) and the separator, ideally it is possible to consider the
same pressure in the two sections:
(15)
The strongest hypothesis of this analysis is the assumption that wet air flow is
saturated at the separator. This expectation could be considered valid because the
compression is followed by a slight cooling process from to . For
this reason it is possible to assume water condensation, at least where there is contact
between the flow and the colder solid parts, as the walls of the separation unit.
Assumptions (14), (15) and (16) allow us to calculate the water vapour fraction at the
separator as follows:
(16)
Therefore, with saturated air flow, =1:
(17)
Figure 3.4 Scheme of circuit main sections.
46
Finally, for Raoult’s law, mole fraction is defined as:
(18)
At this point, since wet air is considered as an ideal gas mixture, it is possible to
assume the conservation of the mole number along the conduit between the
separator and the device upstream section (since no mechanical actions and no
chemical reactions occur after the separator), and therefore the conservation of mole
fractions:
(19)
From the knowledge of it is possible to calculate density
( ), heat capacity ratio ( ) and viscosity
( of wet air at the differential pressure device. Firstly, under the
hypothesis of ideal mixture of gases it is possible to write:
(20)
where , is the molar mass of wet air:
(21)
Table 3.1 Hypothesis over the wet air flow.
Hypothesis Consequences
Wet air is an ideal mixture of gases Validity of the Raoult’s law - Validity of the gas state equation
No temperature losses between separator and differential pressure device
=
No pressure losses between the separator and the compressor chamber
=
Saturated flow at the condensate separator =
( )
No mechanical actions and no chemical reactions =
47
composed by the water molar mass = 18.0153 kg/kmol and by the dry air
molar mass = 28.9641319 kg/kmol. Both parameter values are calculated
from atomic weights listed in IUPAC 2005 – Technical Report [8]. Dry Air molar
composition is given by constituent, taken from "U.S. Standard Atmosphere (1962)".
Minor components have been omitted, thus nitrogen molar fraction has been rounded
up to account for such loss, as in Table 3.3.
Secondly, the heat capacity ratio is given by the fraction between the pressure
capacity and the volume one:
(22)
where heat capacity at constant pressure, , and heat capacity at constant volume,
, are estimated through the equations of Langen, (23) and (24), as a function of the
absolute temperature of the gas and of the correspondent coefficients listed in Table
3.2.
(23)
(24)
Thirdly, viscosity of wet air is calculated starting from dry air viscosity and water
vapour viscosity values, according to equation (52) in "Thermophysical and transport
properties of humid air at temperature range between 0 and 100 °C" [9] applied to a
Table 3.2 Coefficients for Langen’s equations.
Gases a a’ b
Nitrogen (N2) 0.236 0.165 3.8·10-5
Oxygen (O2) 0.203 0.144 3.8·10
-5
Carbon dioxide (CO2) 0.199 0.154 8.6·10-5
Argon (Ar) 0.124 0.079 0
Water vapour (h2o) 0.372 0.262 23.80·10-5
Table 3.3 Dry air molar composition.
Gases x MM, kg/kmol
Nitrogen (N2) 0.7809 14.0067 Oxygen (O2) 0.2095 15.9994
Carbon dioxide (CO2) 0.0003 28.0101
Argon (Ar) 0.0093 39.948
48
two component system (namely dry air and water vapour). Since mass flow is constant
from condense separation unit, then both mass and molar quantities are constant too.
Under the hypothesis of ideal gas, it is thus possible to use equation (20), whereby the
mole fraction at the device is the same of that at the separator unit.
(25)
where:
, calculated according to the "Transport Phenomena 2nd ed." for ideal
gasses, tables E.1 and E.2 at pages 864-866, and eq. (1.4-14) at page 26 (see
example page 28) [10].
, calculated according to "Release on the IAPWS Formulation 2008 for the
Viscosity of Ordinary Water Substance (September 2008)" [11], passing
through the calculation of water vapour viscosity according to "Revised
Release on the IAPWS Industrial Formulation 1997 for the Thermodynamic
Properties of Water and Steam (August 2007)" [12].
Finally, it is possible to calculate the wet air mass flow rate at the pressure device
, according to the iterative method described in Figure 3.3. Passages of
the whole procedure are summarized in Figure 3.5, starting from the mole fraction of
vapour at the condensate separator.
Figure 3.5 Scheme of the steps to calculate wet air mass flow rate.
49
3.2 Inlet Conditions
At this point, in order to apply ISO 1217 coefficients, it is necessary to calculate the
entire wet air volumetric flow rate which enters the compressor chamber, condensate
included. Since the separator unit removes condensate from the wet air flow without
collecting it, next step is to estimate this loss in condensate, while dry air flow rate is
constant along the entire circuit.
Flow rate
Since wet air mass flow rate at the differential pressure device is known, it is possible
to calculate wet air number of moles. Therefore, knowing temperature, pressure and
relative humidity at the inlet section, it is possible to do a proportion between the
water mole fraction at the pressure device and at the inlet.
Firstly, it is necessary to know the wet air number of moles, at the device,
which does not include water extracted at the separator:
(26)
Secondly, conditions at the inlet section are necessary to add the quantity of water
initially contained within the air flow entering the compressor. In particular from the
knowledge of the inlet relative humidity, it is possible to calculate the inlet
molar fraction :
(27)
Finally, passing through the calculus of dry air mass flow rate, it is possible to write a
proportion referred to the dry air number of moles (1- ) respectively at the device
and at the inlet, written as follows:
(28)
where is the total volumetric quantity of wet air which enters the
compressor at the inlet conditions (Free Air Delivery, FAD). This flow rate includes the
condensate trapped within the separator unit located after the radiator, and it is now
ready to be “corrected” into standard ISO 1217 conditions.
50
3.3 ISO 1217
The International Standard ISO 1217 defines the measurement procedure to
characterize any volumetric compressor. It disconnects the machine working
performances from the environmental conditions, which can vary among the tests.
Applying specific coefficients, it provides the performance values of flow rate, power
and specific energy corresponding to the ambient inlet standard ISO 1217 conditions,
mentioned at the beginning of this chapter.
Power
There are two coefficients ( and ) to correct the absorbed mechanical power as
follows:
(29)
where is the correction factor for shaft speed as expressed by:
(30)
Absorbed power strictly depends on the shaft speed value. Hence, it is considered a
coefficient to correct the deviation of the operational shaft speed from the contractual
one. Then, , is the correction factor for inlet pressure, written as:
(31)
Flow rate
The corrective equation for the volume flow rate is written as:
(32)
where is the correction factor for shaft speed, defined as in the power section:
(33)
51
Specific Energy
Finally, “corrected” specific energy is written as:
(34)
Namely, specific energy is the ratio between the “corrected” mechanical power
and the “corrected” volumetric flow rate, .
3.4 Propagation of uncertainties
Before analyzing the results of the tests (Chapter 5), it is necessary to study the
accuracy of the direct measurements and their influence on the subsequent
calculation. In fact, direct measurements have an uncertainty, due to the instruments
through which they are observed, which propagates within each calculation on that
quantity.
Resolution, a, of each instrument is obtained from the correspondent technical
datasheet (reported in Table 2.3). For convenience, with the resolution value it is
possible to calculate the rectangular Type B variance for a generic quantity , as:
(35)
This can be simplified in the rectangular Type B standard uncertainty:
(36)
Afterwards, when 2 or multiple quantities are combined among themselves in order
to obtain the result, , combined standard uncertainty is calculated through the
following expression:
(37)
where f represents the function:
(38)
52
where all input quantities are independent, according to ISO IEC Guide 98-3 [13].
Direct measurement uncertainties
- Temperature
Uncertainty on temperature measurement is calculated as described in ISO
5168 (Annex G, paragraph G.1.2.3). Namely, the uncertainty is given by
combining three components of the temperature reading: thermocouple,
display reading. In particular, the expanded uncertainty, , of the
thermocouple is 1°C with a level of confidence (LC) equal to 95% (k=2).
Therefore, the standard uncertainty (LC = 68%), , is calculated
by the equation:
(39)
Additionally, it is considered the uncertainty of the scale division on the
temperature readout, , written as:
(40)
Finally, even though the thermocouple is correctly installed, it has a small
error in detecting the mean temperature value of the flow, given by the
equation:
(41)
Thus, the whole absolute standard uncertainty on the temperature values is:
(42)
- Pressure
Within this investigation are utilised three different pressure measurement
devices listed in the table in Chapter 2. It is reported their resolution, through
which it is directly calculate the uncertainty of the pressure reading.
The environmental pressure is detected by a barometer, which as resolution,
, equal to 1 mbar. Therefore, the absolute standard uncertainty of
the inlet pressure is written as:
53
(43)
Outlet pressure is given by a manometer located near the last volume of the
compression chamber. The uncertainty of its reading can be written as:
(44)
In the same way, uncertainties of the static pressure, , and differential
pressure, read by the water column are calculated as:
(45)
- Torque
As already done for the temperature measurement, ISO 5168 provides the
same method to combine the resolution of the torquemeter (0.05 %FSO, with
FSO = 1000Nm) with the readout error of the display (0.05 %FSO). So that the
standard uncertainty of the torque measurement is:
(46)
- Relative Humidity
Since the resolution of the hygrometer is known, it is possible to calculate the
standard uncertainty of the relative humidity measurement as:
(47)
- Shaft Speed
Similarly, the RPM meter has resolution of 1 rpm, from which is
possible to calculate the standard uncertainty of the shaft speed
measurement as:
(48)
Combined uncertainties
Combined uncertainties of every quantity have been calculated according to equation
(38), starting from the standard absolute uncertainty of direct measurements. These
54
latter have to be independent between themselves according to ISO IEC Guide (98)
[13].
The only quantity which does not follow the mentioned procedure is the wet mass
flow rate at the differential pressure device. In fact, the various quantities which
appear on the right-hand side of equation (11) are not independent, so that it is not
correct to compute the uncertainty of directly from the uncertainties of these
quantities. For example, is a function of , , , and , and ε is a function of ,
, , and . A practical working formula for (relative standard uncertainty a
level of confidence of 95 % ) is derived in ISO 5167 (section 8.2) [14], and it is written
as below:
(49)
where:
-
and
are taken from the applicable ISO 5167 - Part 3, ;
-
and
are adopted equal to their maximum value, determined in ISO
5167 – Part 3, which are respectively 0,4 % and 0,1 %.
- The values of
and
are determined through the combined uncertainty
equation (38).
3.5 Compatibility Check and Averaged Values
This analysis is carried as a verification of the data collection goodness. The aim of this
step is to exclude those data that for some reasons are not aligned on the average
Table 3.4 Direct measurement uncertainties
Instrumentation Measured quantity Uncertainty
Barometer p_barometric ±0.289 mbar Manometer p_chamber ±0.014 bar Hygrometer UR ±1% Thermocouple T_in ±0.5°C Thermocouple T_chamber ±0.5°C Thermocouple T_device ±0.5°C Water column p_static ±0.289mmH2O Water column p_differential ±0.289mmH2O Torque meter Torque ±0.707 Nm Digital RPM Meter Shaft speed ±0.289 rpm
55
value of the most of them. In order to do that expanded uncertainty is calculated as
written in the following equation:
(50)
In most of measurement situations where the probability distribution characterized by
and is approximately normal and the effective degrees of freedom, , of
is of significant size, one can assume that taking produces an interval
having a level of confidence of approximately 95 per cent (95% LC) [13]. At this point,
knowing the average of the data population, it is possible to understand whether or
not a single test is compatible or not with the other.
(51)
Looking at the example in Figure 3.6, the vertical segments represent the single tests
values with their expanded uncertainty value at 95% LC. If this segment does not
include the average value, indicated with a red line, that test has to be discarded. After
that, a new average value has to be calculated and the compatibility analysis has to be
repeated. When all the tests stand the check, it is possible to calculate the t-Student
distribution of that sample of tests, as follows.
First of all, starting from the number of tests which have passed the compatibility
check it is possible to calculate the degrees of freedom of the distribution. The degrees
of freedom are equal to for a single quantity estimated by the arithmetic
mean of independent observations. Then the standard deviation, , is equal to:
Figure 3.6 Examples of analysis of compatibility.
56
(52)
where:
- , is the single test value
- , is the average of the considerate population
- , is the number of compatible tests.
Then, the variance of the distribution is:
(53)
Lastly, the interval of the final value of the considerate quantity can be express as:
(54)
where is the 95th percentile of this probability distribution. If it is supposed
to be double-sided, that interval is the is a 95% confidence interval for the expected
value .
At this point it is possible to draw bar charts, where bars are function of the degree of
freedom of the population (how many tests are collected), and of its standard
deviation (dispersion of the data). This method to analyse results has been
implemented within an Excel workbook, as described in the following chapter, in order
to compare cases between themselves with a good grade of confidence.
57
Chapter 4. Software Tools
In order to execute the calculations described in the previous chapter and to apply the
compatibility analysis to the results, two main tools are used. The first one, compiled
with MATLAB®, carries out the whole calculations according to the standards ISO 5167
and ISO 1217, to get the values, and according to ISO IEC Guide (98) to get the
uncertainties. The second tool is a VBA Excel workbook which gathers all the values
(collected data and results) of every test and carries out the compatibility check of the
results. The working principle of these two tools is illustrated within the following
sections.
4.1 MATLAB Code “XisoRS”
In order to support the experimental work, the calculation procedure just described is
reported in a computational code, rewritten starting from previous versions
developed into former thesis projects, [4] and [5].
The main characteristic of the new code version are:
New implementation of all the functions, including the calculation of
uncertainties for the quantities elaborated within each of them.
New approach to the problem: all structure variables are substituted with
simpler array variables.
More flexibility due to a new fix calling scheme of the functions, composed by
the input quantities with their respective absolute standard uncertainties to
give output quantities with their respective absolute standard uncertainties.
Simplified structure: the code develops along four levels of complexity, from the
calculation of dry air and vapour properties, through wet air properties until
the calculation of the performance parameters.
New Excel Workbook in which save the collected data and the results, without
any auxiliary database or other external Excel workbooks.
The possibility to chose two different paths of proceeding (as a pair of
“scissors”) according to the flow rate measurement device: water column or
an in-line flow meter (Krohne, VPScope, …).
XisoRS is tested within MATLAB® 2014a software and Windows 7 operating system.
Different MATLAB versions and different computer configurations are not tested here.
The code structure, reported below in Figure 4.1, is a scheme of the different levels of
operations of the calculation procedure.
58
Figure 4.1 Flow chart of the “XisoRS” code structure.
59
Code Structure
This code has a pyramidal structure divided into multiple levels of functions as shown
in Figure 4.1. The higher level (black in Figure 4.1) introduces to the program,
acquiring input data with the “read” operation and returning the results through the
writing operation. This first level splits into two main functions:
call_XisoRS_excel. It is an introductory script which introduces the program. It
asks the user to write the name of the Excel workbook where are stored the
data to analyse. It also asks the name to give to the output Excel file into
which it writes the results with the correspondent uncertainties.
call_XisoRS_manual. It gives the possibility to insert data manually directly into
the command window in order to execute a quick resolution of a single test.
A second level (blue in Figure 4.1) of the code is made of a single function which can
be called by both the two previous functions or directly by the command window:
performances. It is the main body of the code. It gets all data as input and
returns all the result values with the correspondent uncertainties. The
function inside is divided into two possible paths. The first is executed when a
water column is used as flow measurement device, whereas the second path
is executed when a flow meter is used.
The third level (red in Figure 4.1) comprises four functions in this order:
flow_column. This function calculates the dry air mass flow rate (indicated as
mDa), which is constant along all the circuit and wet air mass flow rate at flow
device (mWaust) under hypotheses mentioned in section 3.1.
flow_fmeter. The other path, when no water column is used, is taken when a
flow meter inside the conduit directly detects the wet air flow rate at the
delivery section, after the condensate separator. With the same working
principle of the previous function, it calculates dry air mass flow rate
(indicated as mDa) and wet air mass flow rate (mWaust).
flow_at. It is the function that calculates mass (mWA) and volumetric flow
(qWA) at specified input conditions of temperature, pressure and relative
humidity, such as inlet operational conditions (Free Air Delivery).
ISO_1217. This function implements ISO 1217. Flow rate, power and specific
energy are computed according to ambient standard ISO 1217 conditions.
The fourth and last level (green in Figure 4.1) includes the calculation of the properties
of dry air and vapour and the calculation of the mass flow rate at operational
conditions:
60
VAPsatp. This function calculates the saturation pressure of water according to
"Revised Release on the IAPWS Industrial Formulation 1997 for the
Thermodynamic Properties of Water and Steam (August 2007)".
VAPrho. This function calculates the density of water vapor according to
"Revised Release on the IAPWS Industrial Formulation 1997 for the
Thermodynamic Properties of Water and Steam (August 2007)".
WAcpcvk. Assuming wet air to be an ideal gas mixture, this function is used to
calculate the heat capacity of wet air, according to the Langen's formulas for
ideal gases (24) and (25).
ISO 5167. Iterative mass flow rate calculation according to ISO 5167 (all parts).
All functions have a standard format (VAPsatp example is illustrated in Figure 4.2):
input data with their uncertainties and a a debugging flag (fDBG) in order to give the
possibility to the user to run a test function and check the accuracy of the code;
output data with their correspondent uncertainties and a check flag (fOK) in order to
alert the user about possible errors along the program execution. An introductory
comment describes the method, the variables and the history of the function,
suggesting a command example.
Figure 4.2 Introductory comments of the VAPsatp MATLAB function.
61
4.2 VBA Excel “XisoRS_base.xlsm”
This VBA Excel spreadsheet is the same which enters the MATLAB XisoRS code. Its
main functions are to collect, check and draw graphical trends of test results. In
particular, this workbook is divided into four worksheets:
Data, which serves as data storage of the test to analyse (Figure 4.3)
Export result, which helps the operations within the program transposing all the
cells of the results in Data sheet.
Figure 4.3 Data collected in XisoRS_base Excel file.
62
Compatibility Check, execute the compatibility verification between tests in the
same configuration of lubricant and material as shown in Figure 3.6, in the
previous chapter.
Performances Chart, after the compatibility check, the averaged performance
parameters are calculated according with t-Student distribution (Figure 4.4),
as explained in ISO IEC Guide (98) [13]. Charts can be draw in order to
compare trends.
Figure 4.4 Compatibility check and t-Student analysis.
Figure 4.5 Performances Charts worksheet in “XisoRS_base.xlsm”.
63
Chapter 5. Results and Discussion
As previously stated, the purpose of this thesis is the investigation into the effect of 2
vane materials and 4 commercial lubricants on the performance of a sliding vane
rotary mid-size compressor at 5 delivery pressures, from 6.5 to 8.5 bar(g) by a 0.5-
step. Tests are performed at the operating temperature of about 85°C and an
operating speed of 1500 RPM. A single configuration test is repeated 10 times in order
to verify the repeatability of the measurements under different inlet air conditions of
temperature, pressure and relative humidity and to increase the reliability of the
results.
Performance results are represented by means of three fundamental parameters
(volumetric flow rate, mechanical power, and mechanical specific energy) calculated
according to the international standards ISO 1217 and ISO 5167. Standard absolute
uncertainties of direct measurements are combined to eventually obtain the 68 %LC
uncertainty of the three performance parameters. Expanded uncertainty with a 95
%LC is calculated for the compatibility check within each case, and outlier tests are
then removed from the following analysis.
Therefore, average values of the remaining test results are evaluated to verify the
variation of the performances. Figure 5.1 and Figure 5.2 are representative of the
experimental approach: they illustrate the average values of the standard ISO 1217
volumetric flow rate and the standard ISO 1217 mechanical power, for each lubricant
and vane material configuration at the pressure of 7.5 bar(g).
The standard relative uncertainty (95 %LC) on the calculation of volumetric flow rate
and mechanical power values are around 1.5 % and 1%, respectively. Additionally, the
error bars represent the symmetric interval obtained by applying to each case a t-
Student distribution, with 95 %LC too. The resulting interval contains the greatest
average value (of the considered parameter) for each case population, considering the
amplitude of the population (through the standard deviation) and its mean value.
The experiment results reported in Figure 5.1 show that it is not possible to clearly
establish a correlation between the volumetric flow rate and the lubricant with the
adopted instrumentation. In fact, the relative uncertainty of ISO 1217 volume flow
rate value (1.4 %) is double the variation among the different lubricants (0.7 %).
Further, comparing volumetric flow rate among the two vane materials, the difference
between cast iron and aluminium flow rate mean values (0.4 %) is less than a third of
the standard relative uncertainty. As a result, lubricants do not affect appreciably the
64
volumetric flow rate, and the measured variations fall within the range of uncertainty
of measurement.
On the contrary, as shown in Figure 5.2, when the mechanical power is considered,
the effects of lubricants on the SVRC performance are more significant. In fact, even
though the relative uncertainty of ISO 1217 shaft power value (1 %) is double the
deviation among the different lubricants in the cast iron case (0.5 %), it has the same
order of magnitude of the deviation among lubricants within the aluminium tests (1
%). Furthermore, comparing the shaft power of the two vane materials, the difference
between cast iron and aluminium flow rate mean values (1.7 %) is almost double the
standard relative uncertainty. Thus, as reported in “Experimental investigation on
materials and lubricants for sliding-vane air compressors” [15], it is possible to
correlate the variation of the vane material with a change in the absorbed power.
Figure 5.1 Cast Iron ISO 1217 volumetric flow rate at 7.5 bar(g).
Figure 5.2 Cast Iron ISO 1217 shaft power at 7.5 bar(g).
65
In particular, when the lubricant B is used, the lowest mechanical power is achieved
with both vane materials. This is probably due to the low viscosity of lubricant B
besides its higher viscosity index. At the considered operating temperature of about
85°C the viscosities of all lubricants are comparable and cannot justify the differences
in mechanical power. On the other hand, viscosity index (VI) suggests how viscosity
characteristics of the lubricant are stable when subjected to temperature variations.
The higher the viscosity index, the lower the change of viscosity of the oil with
temperature, and vice versa. This allows for consistent compressor performance
within the normal working conditions. In fact, viscosity of liquids decreases as
temperature increases. The viscosity of a lubricant is closely related to its ability to
reduce friction. If the lubricant is too viscous, it will require a large amount of power to
move; if it is too thin, the surfaces will come in contact and friction will increase. In this
regard, the lower viscosity index of lubricant D justifies the worst performance of this
lubricant. As shown in Figure 5.2, when diester lubricants A, C and D are employed, an
increase in the absorbed mechanical power can be noted, especially in the case of
Table 5.1 Variations comparison.
Shaft Power (r.u. 1 %), kW Cast Iron Aluminium Variation
Minimum (Oil B) 20.648 20.248 1.9 % Maximum (Oil D) 20.749 20.444 1.5 %
Variation 0.5 % 1 % -
Figure 5.3 ISO 1217 shaft power at 7.5 bar(g): cast iron (CI) and aluminium (Al).
66
aluminium vanes. On the contrary, B is constantly the best-performing lubricant.
However, results show that the lubricant replacements determine only a slight
variation (up to 1%) of the mechanical power for both the considered materials.
The absorbed mechanical power decreases when aluminium vanes (AlV) are used: the
reduction of vane weight implies a reduction in the mechanical power on the order of
1.3-2.5%, depending on the pressure and the lubricant (e.g. 7.5 bar(g) case is reported
in Table 5.2). The reduction of power consumption can be attributed not only to the
lower density of the aluminium alloy but also to the improvement of its mechanical
properties due to the surface treatment. The chosen anodizing allows a very high
surface hardness to be obtained, while the homogeneous surface finishing increases
the affinity with the lubricant. Despite the reduced thickness of the superficial
treatment, compared to conventional hard oxide coating, it is highly wear resistant,
Table 5.2 Mechanical power variation comparison changing lubricants at 7.5 bar(g).
Shaft Power at 7.5 bar(g), kW
A B C D
Cast Iron 20.659 20.648 20.679 20.749 Aluminium 20.312 20.248 20.311 20.444
Variation 1.70 % 1.98 % 1.81 % 1.49 %
Figure 5.4 ISO 1217 volumetric flow rate at 7.5 bar(g): cast iron (CI) and aluminium (Al).
67
improving the mechanical performance of aluminium vanes. As shown in Figure 5.3,
the performances of AlV are better than those of cast iron vanes (CIV) in the range of
the considered pressures: the reduction of the weight and the good finishing surface
of the AlV determine an appreciable reduction of the absorbed mechanical power.
Consequently, the maximum mechanical power reduction is achieved with the AlV and
lubricant B. For both CIV and AlV the lubricants A and C are the only diester-based
lubricants able to achieve the same performance as the poly-α-olefin-based lubricant
B. The good performances of the lubricants A and C are probably due to the use of
some addictives (i.e. phosphorus, magnesium, etc.) in the formulation (Table 3.7) to
improve the affinity between the lubricant and AlV surface.
Even though volumetric flow rate is not affected by the change in configuration
(lubricant or material) as shown in Figure 5.4, looking at the shaft specific energy
trends along the pressure steps Figure 5.5, it can be noticed that there is a clear
difference between the two vane materials (around 1.5%). Probably this is because
the reduction in the absorbed mechanical power, in case of aluminium vanes, is
enough to generate a reduction in the shaft specific energy too. This means that,
independently form which kind of oil is used, the aluminium alloy grants more efficient
power consumption.
In general, results show that lubricant D offers the worst performance in the
considered operating conditions: the lower viscosity index and in all probability the
absence of additives justify the worse performances.
Figure 5.5 ISO 1217 shaft specific energy at 7.5 bar(g): cast iron (CI) and aluminium (Al).
68
Table 5.3 Performance parameters values in every configuration.
Oil Mat. bar(g) qISO , l/min powmISO, kW mseISO, kw min /m3
A
CIV
6.5 3364 ±9 0.27% 19.65 ±0.04 0.20% 5.84 ±0.02 0.34% 7 3359 ±7 0.21% 19.99 ±0.04 0.20% 5.95 ±0.02 0.34%
7.5 3345 ±10 0.30% 20.66 ±0.06 0.29% 6.18 ±0.03 0.49% 8 3329 ±11 0.33% 21.34 ±0.05 0.23% 6.41 ±0.03 0.47%
8.5 3304 ±12 0.36% 22.02 ±0.05 0.23% 6.67 ±0.03 0.45%
AlV
6.5 3342 ±12 0.36% 19.23 ±0.07 0.36% 5.75 ±0.03 0.52% 7 3334 ±11 0.33% 19.66 ±0.06 0.31% 5.90 ±0.03 0.51%
7.5 3314 ±11 0.33% 20.31 ±0.06 0.30% 6.13 ±0.03 0.49% 8 3289 ±12 0.36% 20.96 ±0.08 0.38% 6.37 ±0.04 0.63%
8.5 3258 ±11 0.34% 21.63 ±0.06 0.28% 6.64 ±0.03 0.45%
B
CIV
6.5 3355 ±4 0.12% 19.73 ±0.04 0.20% 5.88 ±0.01 0.17% 7 3347 ±5 0.15% 20.10 ±0.10 0.50% 6.01 ±0.03 0.50%
7.5 3331 ±4 0.12% 20.65 ±0.08 0.39% 6.20 ±0.02 0.32% 8 3311 ±9 0.27% 21.37 ±0.10 0.47% 6.45 ±0.04 0.62%
8.5 3292 ±6 0.18% 22.02 ±0.10 0.45% 6.69 ±0.04 0.58%
AlV
6.5 3348 ±6 0.18% 19.25 ±0.02 0.10% 5.75 ±0.01 0.17% 7 3340 ±5 0.15% 19.60 ±0.06 0.31% 5.87 ±0.02 0.34%
7.5 3321 ±5 0.15% 20.25 ±0.05 0.25% 6.10 ±0.02 0.33% 8 3295 ±6 0.18% 20.89 ±0.05 0.24% 6.34 ±0.02 0.32%
8.5 3263 ±9 0.28% 21.55 ±0.05 0.23% 6.60 ±0.03 0.45%
C
CIV
6.5 3365 ±8 0.24% 19.64 ±0.07 0.35% 5.84 ±0.02 0.34% 7 3346 ±9 0.27% 20.04 ±0.07 0.35% 5.99 ±0.03 0.50%
7.5 3338 ±7 0.21% 20.68 ±0.08 0.39% 6.20 ±0.03 0.48% 8 3316 ±6 0.18% 21.36 ±0.07 0.33% 6.44 ±0.03 0.47%
8.5 3302 ±6 0.18% 22.03 ±0.07 0.32% 6.67 ±0.03 0.45%
AlV
6.5 3343 ±14 0.42% 19.21 ±0.05 0.26% 5.75 ±0.03 0.52% 7 3337 ±11 0.33% 19.65 ±0.04 0.20% 5.89 ±0.03 0.51%
7.5 3321 ±12 0.36% 20.31 ±0.04 0.20% 6.12 ±0.03 0.49% 8 3294 ±12 0.36% 20.95 ±0.05 0.21% 6.36 ±0.04 0.63%
8.5 3260 ±13 0.40% 21.60 ±0.05 0.24% 6.62 ±0.04 0.60%
D
CIV
6.5 3343 ±9 0.27% 19.60 ±0.04 0.20% 5.86 ±0.02 0.34% 7 3323 ±12 0.36% 20.17 ±0.07 0.35% 6.07 ±0.03 0.49%
7.5 3322 ±8 0.24% 20.75 ±0.05 0.24% 6.25 ±0.03 0.48% 8 3299 ±16 0.48% 21.38 ±0.09 0.42% 6.48 ±0.04 0.62%
8.5 3279 ±14 0.43% 22.16 ±0.09 0.41% 6.76 ±0.04 0.59%
AlV
6.5 3349 ±9 0.27% 19.35 ±0.07 0.36% 5.78 ±0.03 0.52% 7 3346 ±15 0.45% 19.75 ±0.09 0.46% 5.90 ±0.05 0.85%
7.5 3325 ±13 0.39% 20.44 ±0.10 0.49% 6.15 ±0.05 0.81% 8 3294 ±14 0.43% 21.11 ±0.11 0.52% 6.41 ±0.06 0.94%
8.5 3268 ±12 0.37% 21.75 ±0.12 0.55% 6.66 ±0.06 0.90%
Mat. = Vane material qISO = volumetric flow rate ISO 1217
CIV = Cast iron vanes powmISO = mechanical power ISO 1217
AlV = Aluminium alloy vanes mseISO = mechanical specifc energy ISO1217
69
Chapter 6. Conclusions
This work experimentally examines the performances of a commercial mid-capacity
sliding-vane rotary compressor using 2 types of material vanes (cast iron, CIV, and
aluminium with anodized surface, AlV) and 4 different commercial oils (indicated as
lubricant A to D, varying in viscosity index and in additives concentration) at 5 delivery
pressures. All the 40 test combinations are replicated 10 times to verify the
repeatability of the measurements and increase the level of confidence in the results.
For each test volumetric flow rate, absorbed mechanical power and specific energy are
calculated according to the international standards ISO 1217 and ISO 5167.
Calculations of their uncertainty and of their compatibility within each test
configuration are executed according to standards ISO 5168 and ISO IEC Guide 98. A
MATLAB® program and a VBA Excel Workbook are utilised to analyse the collected
data, both of which are developed within this project. The conclusions of the work are
as follows.
• The considered vane materials and lubricants do not affect appreciably the
volumetric flow rate, as the measured variations of performances are contained within
the range of the measurement uncertainty. On the other hand, the effects of vane
materials and lubricants on absorbed mechanical power are more significant: in the
considered operating conditions, the lubricant replacement determines a slight
variation of about 0.5 % of the shaft power for cast iron and of 1% for aluminium, with
the latter aligned to the relative uncertainty of the power values.
• Lubricant B allows the lowest mechanical power within both cases of cast iron
(CIV) and aluminium vanes (AlV) to be reached. When lubricants A, C and D are used,
an increase in the absorbed mechanical power can be noted, especially in the case of
AlV. At the considered operating temperature of about 85°C the viscosities of all
lubricants are comparable and cannot justify the slight differences in mechanical
power. Performances are more probably influenced in the long term by lubricant
formulations and additive concentrations.
• The absorbed mechanical power decreases with aluminium vanes by up to 2.5%,
compared with CIV results. The reduction of power consumption can be attributed not
only to the lower density of the aluminium alloy but also to the improvement of its
mechanical properties due to the superficial treatment. The finishing surface of AlV
allows reaching a good affinity with the lubricants. Finally, results suggest that AlV is a
promising solution for the energy optimization of SVRC.
71
Chapter 7. Future work
A future work could focus on the evaluation of the effect of vane material and
lubricants on a larger capacity compressor, as for example a 75 kW SVRC. Another
aspect, consequent from this investigation, which could be studied in depth, is the
performance of AlV varying the rotation speed of the compressor.
Performances of lubricants durability have not been evaluated during this internship
but it could be a possible future work, in order to understand if long running period
with certain lubricants can affect compressors performances.
73
Bibliography
[1] R. Cipollone, G. Contaldi and D. Di Battista, “Energy Consumption in air
compression - Theoretical and experimental research activity on sliding vane
rotary compressors”.
[2] International Energy Agency, “Redrawing the Energy Climate-Map,” 11 June 2013.
[Online]. Available: http://www.iea.org.
[3] R. N. Brown, Compressors: Selection and Sizing, 3rd Edition, Houston, Texas: RNB
Engineering, 1997.
[4] M. Miggiano, Campagna sperimentale per lo sviluppo e la verifica di un
compressore a palette con geometria statorica modificata, Politecnico di Milano,
AA 2012-2013.
[5] M. Recalcati, Verifica sperimentale delle prestazioni di compressori a palette,
Politecnico di Milano, AA 2013-2014.
[6] T. P. Calvi, Verifica sperimentale degli effetti dell’iniezione di olio tramite ugelli in
compressori volumetrici a palette, Politecnico di Milano, AA 2011-2012.
[7] I. Standard, ISO 1217 - Displacement compressors - Acceptance test, 2009.
[8] M. E. Wieser, “ATOMIC WEIGHTS OF THE ELEMENTS,” INTERNATIONAL UNION OF
PURE AND APPLIED CHEMISTRY, 2005.
[9] P. T. Tsilingiris, “Thermophysical and transport properties of humid airat
temperature range between 0 and 100°C,” Energy Conversion and Management,
pp. 1098-1110, 2008.
[10] B. R. Bird, W. E. Stewart and E. N. Lightfoot, Transport Phenomena - 2nd Edition,
New York: John Wiley & Sons, Inc., 2002.
[11] J. R. Cooper, “The Internal Association for the Properties of Water and Steam,”
74
School of Engineering and Materials Science, Berlin, 2008.
[12] J. R. Cooper, “Revised Release on the IAPWS Industrial Formulation 1997 for the
Thermodynamic Properties of Water and Steam,” International Association for
the Properties of Water and Steam, Lucerne, 2007.
[13] GUM:1995, “Part3: Guide to the expression of uncertainty in measurement,”
2008.
[14] I. Standard, ISO 5167 - Measurement of fluid flow by means of pressure
differential devices inserted in circular cross-section conduits running full, 2003.
[15] S. Murgia and G. Valenti, “Experimental investigation on materials and lubricants
for sliding-vane air compressors,” 2015.
75
Appendix A
In the following pages of this appendix are reported the compatibility graphs for each
of the 40 cases. Within each case, the three performance parameters at ambient
standard ISO 1217 conditions are presented, respectively: volumetric flow rate,
mechanical (shaft) power, mechanical (shaft) specific energy.
76
Cast Iron – A – 6.5 bar(g)
77
Cast Iron – A – 7.0 bar(g)
78
Cast Iron – A– 7.5 bar(g)
79
Cast Iron – A – 8 bar(g)
80
Cast Iron – A – 8.5 bar(g)
81
Aluminium – A – 6.5 bar(g)
82
Aluminium – A – 7 bar(g)
83
Aluminium – A – 7.5 bar(g)
84
Aluminium – A – 8 bar(g)
85
Aluminium – A – 8.5 bar(g)
86
Cast Iron – B – 6.5 bar(g)
87
Cast Iron – B – 7 bar(g)
88
Cast Iron – B – 7.5 bar(g)
89
Cast Iron – B – 8 bar(g)
90
Cast Iron – B – 8.5 bar(g)
91
Aluminium – B – 6.5 bar(g)
92
Aluminium – B – 7 bar(g)
93
Aluminium – B – 7.5 bar(g)
94
Aluminium – B – 8 bar(g)
95
Aluminium – B – 8.5 bar(g)
96
Cast Iron – C – 6.5 bar(g)
97
Cast Iron – C – 7 bar(g)
98
Cast Iron – C – 7.5 bar(g)
99
Cast Iron – C – 8 bar(g)
100
Cast Iron – C – 8.5 bar(g)
101
Aluminium – C – 6.5 bar(g)
102
Aluminium –C – 7 bar(g)
103
Aluminium – C – 7.5 bar(g)
104
Aluminium – C – 8 bar(g)
105
Aluminium – C – 8.5 bar(g)
106
Cast Iron – D – 6.5 bar(g)
107
Cast Iron – D – 7 bar(g)
108
Cast Iron – D – 7.5 bar(g)
109
Cast Iron – D – 8 bar(g)
110
Cast Iron – D – 8.5 bar(g)
111
Aluminium – D – 6.5 bar(g)
112
Aluminium – D – 7 bar(g)
113
Aluminium – D – 7.5 bar(g)
114
Aluminium – D – 8 bar(g)
115
Aluminium – D – 8.5 bar(g)
![[Quarteroni, Saleri] - Introduzione Al Calcolo Scientifico](https://static.documents.pub/doc/80x56/55cf9ba8550346d033a6e6cc/quarteroni-saleri-introduzione-al-calcolo-scientifico.jpg)
