Embed Size (px)
Citation preview
Effect of inter-stage phenomena on the performance
prediction of two-stage turbocharging systems
Authors:
Calogero Avola a +, Colin Copeland a, Richard Burke a and Chris Brace a
a Department of Mechanical Engineering, University of Bath, Claverton Down, Bath BA1 2PB, UK
Corresponding Author:
Calogero Avola+ Email address: [email protected]
AbstractThe paper investigates accuracy of performance measurement in two-stage turbocharging
systems, due to aero-thermal inter-stage phenomena. A novel methodology to measure
performance of turbochargers into equivalent maps has been implemented, for mapping of
turbocharging systems in steady turbocharger gas-stands. The comparison of equivalent maps and
stand-alone high and low pressure turbochargers maps is performed, via single maps combinations.
In this scenario, two-stage system performance are calculated on the basis of single stages variables
in a simplified map-based one-dimensional code. In order to quantify the influence of heat transfer
in turbochargers on the two-stage turbocharging system, diabatic and adiabatic turbochargers maps
with heat corrections for each stage ares implemented. In conclusion, in comparison to equivalent
two-stage maps, combined stand-alone maps predict a significantly higher pressure ratio and
efficiency at compressors, due to low speed maps extrapolation. Meanwhile, the turbine net
efficiency is missed by about 10% at elevated corrected mass flow operations, due to
underestimation of swallowing capacity and isentropic expansion in the combined map approach.
Keywords: Two-stage; Turbocharger; Compressor; Turbine; Equivalent map; Heat transfer
Nomenclaturem Mass flow rate [Kg/s]
1D One-dimensional
app Apparent
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
c Compressor
CBP Compressor By-Pass
corr Corrected
eff Effective
EGR Exhaust Gas Recirculation
eq Equivalent
HP High Pressure
in Inlet
is Isentropic
LP Low Pressure
MAF Mass Air Flow
N Speed [rpm]
NA Not Available
P Pressure
PR Pressure ratio
PRT Platinum Resistance Temperature
Q Specific mass flow heat [KJ/Kg]
ref Reference
s Static
T Temperature
T Total
t Turbine
TBP Turbine By-Pass
TIT Turbine Inlet Temperature
T-T Total-to-Total
VGT Variable Geometry Turbine
WC Water-cooling
γ Ratio of specific heats
η Efficiency
1. IntroductionThe increasing demand for environmental friendly automobiles is leading the efficiency
improvement and reduction of harmful emissions in internal combustion engines, adopting of key
technologies, such as, high pressure (HP) and low pressure (LP) exhaust gas recirculation (EGR) [1, 2],
turbocharging [3] and waste heat recovery solutions [4]. Furthermore, the reduction of engine swept
30
31
32
33
34
35
volume and number of cylinders is able to improve thermal efficiency of internal combustion
engines, through the decimation of friction losses [5]. Therefore, the adoption of boosting
technologies is a necessary action to restore rated power of downsized internal combustion engines
at steady [6] and transient [7] operations. In this scenario, two-stage turbocharging systems are able
to impact positively on the engine pumping losses [8] and powertrain system flexibility [9]. In fact,
the choice of a two-stage system results in a wider flow range operation, generating high levels of
boost at every engine speed, as the maximum pressure ratio is not delivered by a single
turbocharger [10]. In this way, reduction of mechanical and thermal stresses on the single
turbocharger is reduces [11]. However, two-stage turbocharging systems lead to a rise of system
complexity in regards to powertrain control [12].
In these boosting technology, high pressure (HP) and low pressure (LP) turbochargers are
connected sequentially and regulated via by-pass valves [13]. In fact, exhaust gases of the internal
combustion engine are ingested by the HP turbine, incurring a first expansion phase, and the LP
turbine, being subjected to the last expansion phase. As well as turbines, the air is compressed
sequentially by LP (high mass flow and rotating inertia) and HP (low mass flow and rotating inertia)
compressors. In conditions of elevated engine speed, the exhaust flow is diverted away from the HP
turbine inlet via the turbine by-pass (TBP) valve, in order to expand exhaust gasses in LP turbine and
reduce engine back-pressure. Additionally, the HP compressor by-pass (CBP) valve is activated at
high mass flows, to avoid choking of the HP compressor and performance disruption of the two-
stage system. Furthermore, the sequence of HP and LP turbochargers can generate aero-thermal
effects, causing a variation of performance maps of two-stage systems [14].
Analysis of turbochargers performance in two-stage regulated systems has stated that
performance changes in high pressure (HP) compressor and low pressure (LP) turbine can occur [15].
Specifically, reduction of swallowing capacity and pressure ratio of LP turbine and HP compressor,
respectively, can be recorded in comparison to the stand-alone maps. Moreover, LP turbine and HP
compressor are seemed to deliver lower efficiencies in two-stage regulated systems [15]. In this
scenario, it is important to focus on the cause of performance distortion of the turbomachine in
sequential systems. In fact, the presence of complex ducting geometries at the inlet of HP and LP
compressors causes a variation of the performance map measured with straight ducts, as in gas-
stands [16, 17]. In the case of radial turbines, swirling flows generated at the HP turbine outlet cause
vortexes at the LP turbine inlet, resulting in a variation of LP turbine efficiency [18]. As well as
turbines, the presence of pre-whirl can distort pressure ratio and efficiency of compressors in two-
stage systems [19]. Additionally, in diabatic operations of the two-stage system, heat transfer from
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
turbines to compressors influences thermodynamic boundary conditions of the turbomachinery [20,
21].
The performance evaluation of two-stage turbocharging systems in equivalent maps has
reduced the inaccuracy of performance data, particularly, at low loads [22]. In addition, thermal
effects at inter-stage ducts due to intercooling and heat transfer can be incorporated in the
equivalent map. In this scenario, equivalent performance variables for compressor and turbine
systems would have to be considered, as well as, an equivalent two-stage speed term [23].
Equivalent maps of a regulated two-stage turbocharging system are measured in a steady
turbocharger gas-stand, in order to account for inter-stage phenomena and performance variations
of HP compressor and LP turbine. Subsequently, stand-alone maps of HP and LP turbochargers are
measured and combined in order to quantify the inter-stage effects and the influence on the two-
stage system performance. In order to diversify flow motions and heat transfer effects, internal heat
transfer in turbocharger is evaluated through adiabatic and diabatic operations of the two stand-
alone turbochargers with a turbine inlet temperature (TIT) of 773K. Furthermore, the presence of a
water-cooling loop at the LP compressor is analysed and influences on two-stage system is
investigated through the implementation of maps corrections.
This paper aims to investigate the turbochargers performance difference between stand-
alone and two-stage system configurations. In particular, the study introduces a novel mapping
approach for two-stage turbocharging systems, generating equivalent compressor and turbine maps.
In order to achieve the aim, the two HP and LP turbochargers would have to be tested in the steady
gas-stand, in order to measure performance under diabatic conditions. In fact, the investigation
tends to quantify the aero-thermal inter-stage effects on compressors and turbines performance,
comparing combined stand-alone measured maps and equivalent maps of the two-stage system. In
addition, as well as, combining diabatic maps of the two stages, the heat transfer effect is isolated
through the adoption of adiabatic maps and heat transfer corrections. Moreover, due to the
presence of water-cooling at the LP compressor housing, the two-stage turbochargers performance
are evaluated in two different conditions, evaluating the effects of the water-cooling loop. Lastly, the
paper reports the comparison of equivalent and combined maps, as diabatic and heat corrected,
evaluating the influence of inter-stage phenomena.
2. Experimental setting
2.1 Steady turbocharger gas-standIn order to quantify the performance of HP and LP turbochargers in stand-alone and
equivalent two-stage system configurations, experiments on a specifically built steady turbocharger
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
gas-stand have been performed. In this scenario, performance maps for compressors and turbines
can be generated through the monitoring of mass flow rates, pressure and temperature of air path
in compressors and turbines. In addition, the turbocharger gas-stand is equipped with eddy current
sensors for evaluating rotational speeds at the compressor casing. In order to investigate adiabatic
and diabatic operations, the turbocharger gas-stand of figure 1 is able to generate hot and
pressurised steady flows at the turbine inlet, due to the presence of two 44KW electric heaters
(element 9 of figure 1) and a 7bar pressurised air source controlled via a regulator (element 2 of
figure 1). In addition, the facility consists of separate ducting systems for compressor and turbine
sides. In order to perform experiments on the turbocharger, load on the compressor is controlled
through a back-pressure valve (element 13 of figure 1). Due to lubrication and, in the case of the LP
stage, cooling requirements of turbochargers, oil (element 15 of figure 1) and water-cooling
(element 16 of figure 1) control units are available, maintaining the loops at the desired temperature
and pressure.
Compressor and turbine performance maps are generated, as pressure, temperature, mass
flow and speed of the turbomachine are measured with sensors listed in table 1. The sensors
positioning in the turbocharger gas-stand has been performed accordingly to ASME [24] and SAE
standards [25, 26]. Specifically, pressure is measured through a transducer monitoring conditions at
four points along a radial section of the duct, as in figure 2a, in order to obtain an averaged value
across the single section of the duct. A similar approach is applied for the temperature estimation. At
compressor inlet, depths of two platinum resistance temperature (PRT) sensors are oppositely
positioned at 1/3 of the duct diameter, while, four PRTs are placed at 1/4, 1/3 and 1/2 of the
diameter at the compressor outlet, as in figure 2b. As well as the compressor, turbine inlet and
outlet temperatures are measured through four thermocouples positioned in a similar way as
represented in figure 2b. Importantly, in the case of PRT sensors, the presence of a long sensing
element at the PRT sensor tip (up to 20mm for PRTs with 150mm long steam) brings to slightly
deeper protrusion of sensor tip within the flow, when compared to K-type thermocouples with
sensing elements in the order of few millimetres.
The turbocharger speed is monitored through Eddy current sensors, counting the passing of
compressor blades. Moreover, in the experimental investigation, temperature of lubricating oil is
controlled downstream of the bearing housing at about 360K, maintaining a pressure of about 2.4-
3bar varying directly with the turbocharger speed. In case of water-cooled LP compressor housing, a
temperature of 360K is controlled downstream the compressor with a water flow of 10l/min. it is
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
important to notice that the measuring sections in the gas-stand are fully insulated to avoid heat
transfer between the flow and test cell ambient, introducing errors in the estimation of compression
and expansion efficiencies [27].
2.2 Full two-stage system and stand-alone turbochargersIn order to perform the study, a two-stage sequential turbocharging system with regulating
valves has been experimentally investigated. In this system layout, HP and LP turbochargers are
connected in series with compressor by-pass (CBP) and turbine by-pass (TBP) valves controlling the
operation across a vast range of mass flows through compressors and turbines. In fact, the two
turbochargers have different sizes in order to be able to generate elevated levels of boost at low and
high mass flow rates. Specifically, as in table 2, HP and LP compressors have wheel diameters of
approximately 40cm and 60cm, respectively. Clearly, in figure 3, pressure at point 3 is varied in order
to control the expansion ratio in the turbine stages. Meanwhile, load at HP and LP compressors is
generated though a back-pressure valve at point 2 in the turbocharger gas-stand.
In according to the definition of equivalent two-stage map [23], turbine and compressor equivalent
maps are evaluated across the two stages, as in figure 3, using the definitions of pressure ratio ( PR),
mass flow rate (m) and efficiency (η) for compressor and turbine in equations 1-6, respectively. In
this scenario, the two-stage system is treated as a single turbocharger with compression and
expansion processes. The equivalent mass flow rate of compressor and turbine are corrected for
pressure (Pref ¿ and temperature (T ref ), being equivalent to 298K and 1bar for compressor and 288K
and 1atm for turbine. In equation 1, the total-to-total (T-T) pressure ratio (PRT-T) is defined for the
compressor as the ratio between outlet (P2T) and inlet (P1T) total pressure. In regards to mass flow
rate in the compressors (mc), correction from measured total temperature (T1T) and pressure (P1T)
conditions to reference values is shown in equation 2. Furthermore, total-to-total compression
efficiency (ηT-T) is shown in equation 3, relating isentropic and adiabatic compressions, including the
ratio of specific heats (γ)
PRT−T=P2T /P1T (1)
mc corr=mc √T 1T /TrefP1T /Pref
(2)
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
ηT−T=T 1T∗(PRT−T
γ−1γ )
T2T /T 1T(3)
As well as, the definition of compressor performance, pressure ratio (PRT-s), corrected mass
flow (mtcorr) and turbine net efficiency (ηnet) are considered in equations 4-6. In particular, turbine net
efficiency replaces the total-to-static expansion efficiency, due to the dependency on adiabatic
turbine power, accounting for turbine outlet temperature. In fact, the presence of swirling flows at
the turbine outlet could introduce inaccuracy in the estimation of turbine total-to-static efficiency. In
fact, turbine net efficiency of equation 6 considers the ratio between compressor power and turbine
isentropic power. Furthermore, friction influence is removed in the definition of turbine net
efficiency.
PRT−s=P3T /P4 s (4)
mtcorr=mt √T 3T /T refP3T /Pref
(5)
ηnet=PowercPoweris t
(6)
Due to the dependency between the performance of the turbocharging system and the
speeds of HP and LP stages, an equivalent two-stage system speed [23] has been defined in equation
7 and corrected for inlet total temperature (T ¿t) of turbine or compressor in equation 8. Accordingly
to the definitions in equations 1-8, the equivalent two-stage system maps can be generated in the
steady turbocharger gas-stand, treating the turbocharging system as a single turbocharger.
N eq=N LP∗( N LP
NHP ) (7)
N eq corr=N eq √T ref /T¿T (8)
In the proposed study, in order to focus on the effect of inter-stage phenomena on two-
stage system performance prediction, the TBV valve is constrained to the fully shut position, as well
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
as, the variable geometry turbines (VGT) at both HP and LP stages fixed at 50% between minimum
and maximum allowable opening area. Due to the complexity of the turbocharging system design,
difficulties persist in the estimation of the effect of VGT on the opening area of HP and LP turbines.
Therefore, HP and LP turbochargers have been treated as fixed geometry turbines with a reduction
of the operating range, reducing the VGT to 50%. In conjunction with the testing of the full two-stage
system, stand-alone HP and LP turbochargers have been investigated in the steady turbocharger gas-
stand, as shown in figure 4.
2.3 Experimental campaignBoth HP and LP turbochargers have been mapped under adiabatic and diabatic conditions in
order to quantify the heat correction of the turbochargers. Due to the presence of a water-cooling
housing in the LP compressor, tests w/ and w/o cooling effects have been performed in both the
two-stage system and the LP turbochargers in stand-alone configurations. In details, in case of
adiabatic maps, compressor outlet and turbine inlet temperatures are matched, although the
bearing housing is controlled at an oil outlet temperature of 360K. Although, this setting may not
represent complete adiabatic conditions of the turbocharger [28], the dependency of compressor
and turbine power from friction changes between adiabatic and diabatic maps is reduced [29].
Furthermore, in case of diabatic maps, the TIT is maintained at 773K for both LP and HP
turbochargers whilst lubricating oil temperature is controlled at 360K. In summary, the experimental
campaign for the study is reported in the test matrix of table 3.
3. Experimental results
3.1 Equivalent two-stage mapsIn figure 5, equivalent two-stage maps for compressor and turbine are generated in the
turbocharger gas-stand w/ and w/o water-cooling at the LP compressor. In the case of compressor
cooling, the downstream compressor coolant temperature is maintained at 360K. It is important to
notice that the corrected speed lines in figure 5 relates to the equivalent speed term of equation 8.
As visible, the equivalent two-stage compressor and turbine maps have been limited to 46Krpm and
28.3Krpm, respectively, corresponding to 200Krpm and 100Krpm of HP and LP turbochargers,
respectively.
The adoption of a water-cooling system at the LP compressor has a negative effect on the
equivalent compressor total-to-total (T-T) efficiency, due to the downstream compressor water
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
temperature controlled at 360K. In this scenario, the water-cooling system can extract and introduce
heat to the LP compressor flow due to the constantly controlled temperature of 360K. However, as
visible at the top right corner of figure 5, compressor efficiency is lower in the presence of water-
cooling for the vast majority of speed lines. In addition, a small reduction of available pressure ratio
in the two-stage system is obtained in comparison to the case of the uncooled LP compressor. The
increase in turbine net efficiency (ηnet) for the case w/ water-cooling could be supported by an
increase in apparent compressor work due to the lower compressor efficiency. Meanwhile, the
effect of water-cooling LP compressor housing shows a small change in turbine swallowing capacity
at low pressure ratios.
3.2 Stand-alone turbochargers mapsIn order to analyse the influence of inter-stage effects on the prediction of two-stage system
performance, stand-alone turbochargers maps would have to be measured, resembling operating
conditions of the complete system. In conjunction with the equivalent two-stage maps generated in
the turbocharger gas-stand at diabatic conditions, stand-alone maps for HP and LP turbochargers
would have to be investigated at similar conditions of heat transfer. However, it is important to
consider that temperature at inlet of HP compressor would be higher than ambient, while TIT of LP
turbine could lower than 773K [14].
The investigation of adiabatic and diabatic compressor maps is able to provided correct
estimation of heat transfer and effective efficiencies of compressors and turbines [20]. For both LP
and HP turbochargers, the quantification of the heat transfer term to the compressor is possible,
assuming that the heat source is added to the flow following the adiabatic compression. Meanwhile,
the heat transfer term is included at the turbine entry, due to the temperature dependency of the
expansion processes on the turbine efficiency. In this way, heat corrected efficiencies for LP and HP
compressors and turbines can be estimated, as shown in figures 6 and 7.
In figures 6 and 7, variations in compressor efficiency can be noticed between adiabatic and
diabatic conditions for both HP and LP turbochargers. Specifically, the diabatic efficiency decreases
significantly at low corrected compressor mass flow rates in comparison to the adiabatic efficiency.
In the case of turbines, the adiabatic map could not be generated due to varying TIT and corrected
speed terms. Furthermore, heat corrected efficiencies have been calculated for both turbine and
compressor. As visible in both figures 6 and 7, compressor efficiency is closer to adiabatic operations
when heat corrections are applied. On the other side, a reduction of turbine net efficiency is
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
obtained with heat correction due to lower power required by HP and LP compressors, in figures 6
and 7, respectively.
Due to the experienced variation of equivalent two-stage compressor efficiency in figure 5,
the effect of water-cooling on LP compressor performance has been analysed. In figure 8, the
maximum speed line tested on the LP compressor has shown a rise in pressure ratio in comparison
to adiabatic conditions. In fact, a significant increase in compressor efficiency is recorded in the
presence of water-cooling. Specifically, the apparent compressor efficiency reaches values higher
than 0.8 at 133.7Krpm. Meanwhile, peak efficiency at 54.2Krpm reduces from 0.6 in diabatic
operations to 0.4 in water-cooled conditions. However, in order to evaluate the effects of heat
transfer from water-cooling on the effective compressor power (Power eff) in equation 9, heat
sources from hot turbine (Power heat) and compressor water-cooling effect (Power cool) would have
to be considered from the apparent compressor power measured in the gas-stand (Power app).
Power eff=Powerapp−Power heat−Power cool (9)
Applying the correction for heat and cooling power to the temperature related LP
compressor efficiency (apparent) in the gas-stand can lead to the analysis of water-cooling effects on
the compression process. In figure 9, the variation of compressor efficiency due to water-cooling
system controlled at a downstream water-flow temperature of 360K is shown. As visible in figure 9,
the efficiency corrected for water-cooling power (Cool Corr) is higher than adiabatic and heat
corrected efficiency values at mass flow operations lower than peak efficiency points at 116.4Krpm
and 133.7Krpm. At lower speeds of the LP turbocharger, the water-cooling system is not able to
extract heat from the compression process due to compressor outlet temperatures unable to reach
360K. In addition, owing to a more efficient compressor, the cool corrected turbine net efficiency is
affected.
3.3 Map correctionThe turbocharger gas-stand is extremely important for the performance evaluation of
compressors and turbines. However, the characteristic design of a turbocharger can cause the
measurement of diabatic compression and expansion processes, in case of heat sources introduced
at the turbine inlet [30]. In this scenario, corrections of performance would have to be performed, in
order to distinguish between heat and thermodynamic work. The evaluation of adiabatic maps in the
turbocharger gas-stand reduces the requirements for heat transfer models [31]. However, adiabatic
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
and diabatic maps would have to measured experimentally in the steady gas-stand, increasing the
testing time. It is important to state that the focus of the investigation regards the comparison of
equivalent and combined maps for two-stage turbocharging systems and the map correction
procedure is adopted to differentiate between turbocharger shaft power and heat. Therefore, after
assuming the total energy balance across the turbocharger operating in adiabatic operations, the
friction power can be calculated for HP and LP turbochargers. Specifically, the turbochargers are
non-insulated and a small amount of heat is able to escape to ambient.
The difference in turbocharger sizes has resulted in a different magnitude of friction losses,
as presented in figure 10. Specifically, although the HP turbocharger is tested at rotating speeds of
about 220Krpm, the friction losses account for about 2.5KW. In the case of LP turbocharger, the
same amount of friction power is achieved at about 100Krpm. Moreover, the relationship between
shaft speed and friction power develops differently for the two turbochargers. In this way, the heat
corrected turbine power can be calculated by joining the friction power to the heat corrected
compressor power. In this scenario, the assumption of friction independency from variation of axial
trust is stated, although, this could differ between adiabatic and diabatic operations [32].
Furthermore, adiabatic and diabatic maps have allowed for the calculation of specific heat
flow to the compressor, assuming that heat addition to the compressor is occurring after the
compressor. In figures 11 and 12, the relationship between compressor and turbine heat and mass
flow for HP and LP turbochargers is shown. It is visible that the specific heat flow is significantly
higher in magnitude in the LP compressor, achieving about twice the amount HP compressor energy
at 0.02Kg/s. Additionally, the specific heat flow consists of 2KJ/Kg at about 0.07Kg/s in the HP
compressor and 0.14Kg/s in the LP compressor. A different trend is observed for the specific heat
flow escaping the turbine. In fact, the amount of heat is similar between HP and LP turbines,
although a shift towards higher mass flows is recorded for the LP compressor.
In order to evaluate the changes in compressor efficiency with the introduction of water-
cooling systems at the LP compressor housing, the multiple effects of heat and cooling flow would
have to be subtracted from the measured compressor efficiency in the gas-stand (equation 9). In this
perspective, the cooling power is analysed from availability of friction and effective turbine powers.
Specifically, the calculation of maps correcting factors can be analysed in the appendix A.1.
Moreover, the cooling capacity of the LP compressor along the entire mapped operations is shown
in figure 13. In this graph, significant benefits on compressor efficiency from the water-cooling
system can be achieved at mass flow operations below 0.12Kg/s, as supported by results in figure 9.
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
However, at mass flow ranges between 0.06 and 0.12Kg/s, heat can be provided by the water-
cooling system in conditions of compressor outlet temperatures below 360K.
4. Two-stage performance prediction
4.1 Equivalent and combined mapsIn order to evaluate the gap in performance prediction of the two-stage turbocharging
system, analysis of equivalent two-stage and combined stand-alone HP and LP maps could provide
accurate information. In fact, inter-stage phenomena occurring between HP and LP turbochargers
could affect the performance of the entire two-stage system. In this scenario, the equivalent two-
stage maps for compressors and turbines generated under diabatic conditions should be compared
to combined maps of the two turbochargers. The process of maps combination is performed into a
1D model of the steady turbocharger gas-stand. HP and LP turbochargers are represented by the
stand-alone maps measured in the engine gas-stand at diabatic (773K) and adiabatic conditions,
including heat transfer corrections. The 1D model is adopted for the combination of maps and the
analysis of boundary conditions of each turbocharging stage. The 1D modelling approach is widely
adopted for the analysis of two-stage systems in automotive powertrains [33-35]. It is important to
consider that inter-stage components have been included in the mapping procedures of stand-alone
HP and LP turbochargers. Therefore, the inter-connecting ducts should not be represented into the
1D model as previously included in the measured maps. Specifically, measured HP and LP
turbochargers speeds are imposed and performance across the two turbochargers are calculated, as
shown in the diagram of figure 14. Due to the possibility of difference in HP and LP turbocharger
speeds in the measured equivalent map, extrapolation of unmeasured speed values is done through
quadratic fits to regions of the speed lines via least squares regression.
The process in figure 14 highlights the maps combination. Firstly, the temperature and
pressure conditions at inlet of HP turbine and LP compressor are imposed, as well as, the ambient
conditions at LP turbine and HP compressor outlets. Secondly, HP and LP turbocharger speeds are
imposed, accordingly to speed values in the measured equivalent map. Finally, in order to generate
the combined maps, the back-pressure at compressor system is imposed, matching the mass flow
through turbines and compressors. Furthermore, heat correction at HP and LP turbochargers are
included as heat sources, in the case of adiabatic maps.
The results of figure 15 show a significant different in both pressure ratio and efficiency
predictions for the compressors system. In particular, equivalent two-stage compressor map
measures a lower pressure ratio at 12.5Krpm and 18Krpm in comparison to combined maps w/ and
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
w/o heat correction, as visible at the top left corner of figure 15. The change in pressure ratio
prediction is nearly absent at equivalent speeds equal and higher than 23Krpm. The difference
between equivalent and combined pressure ratios at low speeds in the compressors could be
induced by the absence of measured data with high confidence and accuracy, as supported by the
results [22]. In fact, the equivalent mapping approach would reduce the absolute error in
performance variables. Instead of measuring performance for each of the two compressors with
combination of values and errors due to sensors accuracy, performance data result in higher
accuracy for equivalent two-stage maps with measurements at inlet and outlet of the entire
compression system. Additionally, a similar trend to pressure ratios in figure 15 is recorded in the
estimation of compressor efficiency. In these conditions, a maximum change in efficiency for about
0.05 is monitored at 12.5Krpm between equivalent and combined compressors. Furthermore,
combined map estimate a difference in the swallowing capacity of the two-stage turbines in relation
to the equivalent two-stage map measured directly in the turbocharger gas-stand. Differently from
the compressors case, the turbine net efficiency obtained by the combination of HP and LP maps is
overestimated at high equivalent speed values, as visible in figure 15. The increase in turbine
efficiency prediction could be explained in the underestimation of the isentropic turbine power. This
can be supported by the variation of the swallowing capacity between equivalent and combined
turbine maps in figure 15. In fact, the definition of corrected turbine mass flow in the stand-alone HP
and LP turbine maps could introduce source of errors due to the temperature, mass flow and
pressure sensor accuracy. Above all, the implementation of heat correction to diabatic operations of
HP and LP turbochargers is not able to significantly improve predictions of two-stage system
performance.
Moreover, the heat and cooling correction of the LP compressor map is able to improve the
prediction of the two-stage system performance, in relation to the equivalent configuration, as
visible in figure 16. In fact, in the case of equivalent compressor pressure ratio, combination of HP
and LP maps with cooling corrections can reduce the gap with the measured equivalent map at low
rotating speeds. Accordingly, compressor efficiency estimation is generally improved with the
adoption of cooling power correction at the LP stage. Specifically, higher efficiency is monitored for
the two-stage in the gas-stand at low mass flows across the analysed speed lines. However, the
absence of cooling correction factors at the LP stage is not able to predict compressor efficiency at
12.5Krpm. Furthermore, it is important to notice that the combination of stand-alone turbochargers
maps is unable to confidently predict two-stage turbocharging performance at low operating speeds
and pressure ratios.
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
In order to analyse the effect of heat and cool corrections to stand-alone maps, HP and LP
compressors powers are investigated in combined two-stage maps. Figure 16 shows the relationship
between HP and LP powers under diabatic conditions w/ and w/o water-cooling at the LP
turbocharger. Moreover, the introduction of heat and cooling power corrections in relation to the
two diabatic cases highlights variation of trends in figure 16. Accordingly to plots in figures 15 and
16, a higher influence to compressors power is achieved in the implementation of corrections of the
water-cooled LP turbocharger. In fact, in conjunction to the low speed operations analysed in figure
17, LP compressor power is reduced with correction for heating effects by the water-cooling system
controlled at 360K. Moreover, the application of correction for heat flux from the turbine to the
compressor reduces both HP and LP compressor powers, although a smaller effect is recorded in
figure 17.
5. ConclusionsIn conclusion, the main findings of this research paper can be listed below:
Firstly, the equivalent mapping approach has been able to measure complete two-
stage system performance. The equivalent speed has been defined, capturing
constant speed trends and operations of the two turbochargers. In the equivalent
compressor map, variation between diabatic operation w/ and w/o water-cooling is
visible, reaching about 0.1 efficiency points difference at 12.5Krpm.
Secondly, differences between equivalent and combined two-stage maps recorded.
In fact, compressor performance differ at low equivalent speeds, due to inaccuracies
from the extrapolation of stand-alone compressors performance in the low speed
regions. Furthermore, combined two-stage turbine map shows increased efficiency,
caused by underestimation of equivalent swallowing capacity and isentropic power.
Specifically, the equivalent two-stage mapping approach improves accuracy,
monitoring conditions upstream and downstream the turbocharging system.
Thirdly, heat corrections and adiabatic maps affect efficiencies and powers of each
turbocharging stage. In fact, in comparison to apparent efficiencies of diabatic maps,
compressor power is higher for the heat corrected case. Similarly, the variation is
visible in the presence of water-cooling at LP stage.
Lastly, the investigation is performed for constant positions of the two
turbochargers VGTs. It would necessary to expand the analysis to different VGT
positions, analysing the capabilities of the equivalent mapping approach in capturing
complete operating conditions of the turbocharging system.
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
Acknowledgement The Authors would like to acknowledge the technical staff at the Powertrain and Vehicle
Research Centre for the support received in implementing the experimental facility and researchers
Tomasz Duda and Ramkumar Vijayakumar for the support in running the experimental facility. The
Authors would like to acknowledge the University of Bath and the TurboCentre2 consortium for the
financial support.
Appendix A
A.1 Map heat transfer correctionIn the case of diabatic operations with TIT of 773K, the heat correction is performed,
considering compressor and turbine performance measured at both adiabatic and diabatic
conditions. Due to heat transfer, the compressor power at diabatic operations (Power cdia) results
higher than the adiabatic compressor power (Power cadia). Therefore, the change in power can be
estimated as in equations A.1 and A.2. The heat from turbine to compressor (Qc) is evaluated for HP
and LP turbochargers as in equation A.3.
Heatc=Power cdia−Power cadia (A.1)
Power c=mc(h2−h1) (A.2)
Qc=Heat c/mc (A.3)
In order to calculate the total heat escaping the turbine (Qt ¿, the same process of equations
A.1-3 is implemented, as reported in equations A.4-6.
Heat t=Powert dia−Power t adia (A.4)
Powert=mt (h3−h4) (A.5)
Qt=Heat t /mt (A.6)
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
References[1] G. Zamboni, M. Capobianco, Experimental study on the effects of HP and LP EGR in an
automotive turbocharged diesel engine. Applied Energy. 94 (2012) 117-128.
10.1016/j.apenergy.2012.01.046.
[2] V. Bermúdez, J.M. Lujan, B. Pla, W.G. Linares, Effects of low pressure exhaust gas
recirculation on regulated and unregulated gaseous emissions during NEDC in a light-duty diesel
engine. Energy. 36 (2011) 9 5655-5665. 10.1016/j.energy.2011.06.061.
[3] R.P. Roethlisberger, D. Favrat, Comparison between direct and indirect (prechamber)
spark ignition in the case of a cogeneration natural gas engine, part II: engine operating parameters
and turbocharger characteristics. Applied Thermal Engineering. 22 (2002) 1231-1243. 0.1016/S1359-
4311(02)00041-8.
[4] J. Galindo, S. Ruiz, V. Dolz, L. Royo-Pascual, Advanced exergy analysis for a bottoming
organic rankine cycle coupled to an internal combustion engine. Energy Conversion and
Management. 126 (2016) 217-227. 10.1016/j.enconman.2016.07.080.
[5] J.W.G. Turner, A. Popplewell, R. Patel, T.R. Johnson, N.J. Darnton, S. Richardson, S.W.
Bredda, R.J. Tudor, C.I. Bithell, R. Jackson, S.M. Remmert, R.F. Cracknell, J.X. Fernandes, A.G.J. Lewis,
S. Akehurst, C. Brace, C. Copeland, R. Martinez-Botas, A. Romagnoli, A.A. Burluka, Ultra Boost for
Economy: Extending the Limits of Extreme Engine Downsizing. SAE Technical Paper. (2014) 2014-01-
1185. 10.4271/2014-01-1185.
[6] I. Al-Hinti, M. Samhouri, A. Al-Ghandoor, A. Sakhrieh, The effect of boost pressure on the
performance characteristics of a diesel engine: A neuro-fuzzy approach. Applied Energy. 86 (2009) 1
113-121. 10.1016/j.apenergy.2008.04.015.
[7] Q. Tang, J. Fu, J. Liu, B. Boulet, L. Tan, Z. Zhao, Comparison and analysis of the effects of
various improved turbocharging approaches on gasoline engine transient performances. Applied
Thermal Engineering. 93 (2016) 797-812. 10.1016/j.applthermaleng.2015.09.063.
[8] J. Galindo, J.R. Serrano, H. Climent, O. Varnier, Impact of two-stage turbocharging
architectures on pumping losses of automotive engines based on an analytical model. Energy
Conversion and Management. 51 (2010) 10 1958-1969. DOI 10.1016/j.enconman.2010.02.028.
[9] S. Bernasconi, E. Codan, D. Yang, P. Jacoby, G. Weisser, Two-stage Turbocharging
Solutions for Tier 4 Rail Applications. ASME 2015 Internal Combustion Engine Division Fall Technical
Conference, Houston, TX, USA, November 8-11, 2015. 10.1115/ICEF2015-1076
[10] A. Grönman, P. Sallinen, J. Honkatukia, J. Backman, A. Uusitalo, Design and experiments
of two-stage intercooled electrically assisted turbocharger. Energy Conversion and Management.
111 (2016) 115-124. 10.1016/j.enconman.2015.12.055.
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
[11] R. Numakura, Performance of a small-size two-stage centrifugal compressor. 10th
International Conference on Turbochargers and Turbocharging,2012. London, UK, 307-318,
10.1533/9780857096135.6.307
[12] M.E. Emekli, B.A. Güvenç, Explicit MIMO Model Predictive Boost Pressure Control of a
Two-Stage Turbocharged Diesel Engine. IEEE Transactions on Control Systems Technology. PP (2016)
99. 10.1109/TCST.2016.2554558.
[13] N. Watson, S. Janota, Turbocharging the internal combustion engine. 1982. The
Macmillan Press Ltd, London.
[14] C. Avola, C. Copeland, T. Duda, R.D. Burke, S. Akehurst, C.J. Brace, Review of
Turbocharger Mapping and 1D Modelling Inaccuracies with specific focus on Two-Stage Systems. SAE
Technical Paper. (2015) 2015-21-2523. 10.4271/2015-24-2523.
[15] F. Westin, R. Burenius, Measurement of Interstage Losses of a Twostage Turbocharger
System in a Turbocharger Test Rig. SAE Technical Paper. (2010) 2010-01-1221. 10.4271/2010-01-
1221.
[16] J.R. Serrano, X. Margot, A. Tiseira, L.M. Garcia-Cuevas, Optimization of the inlet air line
of an automotive turbocharger. International Journal of Engine Research. 14 (2013) 1 92-104.
10.1177/1468087412449085.
[17] J. Galindo, A. Tiseira, R. Navarro, D. Tarí, C.M. Meano, Effect of the inlet geometry on
performance, surge margin and noise emission of an automotive turbocharger compressor. Applied
Thermal Engineering. 110 (2017) 875-882. 10.1016/j.applthermaleng.2016.08.099.
[18] Y.B. Liu, W.L. Zhuge, Y.J. Zhang, S.Y. Zhang, Numerical analysis of flow interaction of
turbine system in two-stage turbocharger of internal combustion engine. IOP Conference Series:
Materials Science and Engineering. 129 (2016) 012004. 10.1088/1757-899x/129/1/012004.
[19] A. Whitfield, A.H. Abdullah, The Performance of a Centrifugal Compressor With High
Inlet Prewhirl. ASME Journal of Turbomachinery. 120 (1998) July 1998 487-493.
[20] A. Romagnoli, R. Martinez-Botas, Heat transfer analysis in a turbocharger turbine: An
experimental and computational evaluation. Applied Thermal Engineering. 38 (2012) 58-77.
10.1016/j.applthermaleng.2011.12.022.
[21] R.D. Burke, Analysis and Modeling of the Transient Thermal Behavior of Automotive
Turbochargers. Journal of Engineering for Gas Turbines and Power. 136 (2014) 101511.
10.1115/1.4027290.
[22] G. Fitzky, M. Bothien, S. Zbinden, E. Codan, S. Voegeli, Testing and qualification of two-
stage turbocharging systems. 9th International Conference on Turbochargers and
Turbocharging,2010. London, 79-94, 10.1243/17547164C0012010006
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
[23] C. Avola, C. Copeland, R.D. Burke, C.J. Brace, Numerical Investigation of Two-Stage
Turbocharging Systems Performance. ASME ICEF 2016,2016. Greenville, SC, USA, ICEF2016-9449.
10.1115/ICEF2016-9449
[24] ASME, Performance Test Code on Compressors and Exhausters. ASME Standards.
(1997).
[25] SAE-International, SAE J1826 Turbocharger Gas Stand Test Code. in: Society of
Automotive Engineers, Inc. 1995.
[26] SAE-International, SAE J1723 Supercharger Testing Standard. Society of Automotive
Engineers, Inc. (1995).
[27] M. Woehr, M. Moeller, J. Leweux, Variable geometry compressors for heavy duty truchk
engine turbochargers. ASME Turbo EXPO 2017,2017. Charlotte, USA, 26-30 June 2017. GT2017-
64178.
[28] J.R. Serrano, P. Olmeda, F.J. Arnau, M.A. Reyes-Belmonte, H. Tartoussi, A study on the
internal convection in small turbochargers. Proposal of heat transfer convective coefficients. Applied
Thermal Engineering. 89 (2015) 587-599. 10.1016/j.applthermaleng.2015.06.053.
[29] M. Deligant, P. Podevin, G. Descombes, Experimental identification of turbocharger
mechanical friction losses. Energy. 39 (2012) 1 388-394. 10.1016/j.energy.2011.12.049.
[30] S. Shaaban, J. Seume, Analysis of Turbocharger Non-Adiabatic Performance. 8th
International Conference on Turbochargers and Turbocharging,2006. London, 119-130,
10.1016/B978-1-84569-174-5.50012-9
[31] R.D. Burke, C. Copeland, T. Duda, M.A. Reyes Belmonte, Lumped capacitance and 3D
CFD conjugate heat transfer modelling of an automotive turbocharger. Proceedings of the ASME
Turbo Expo 2015,2015. Montreal, 10.1115/GT2015-42612
[32] J. Scharf, T. Uhlmann, C. Schernus, D. Lueckmann, B. Hoepke, N. Schorn, Extended
Turbine Mapping and its Benefits for the Development of Turbocharged Internal Combustion
Engines. 21st Aachen Colloquium Automobile and Engine Technology,2012. Aachen, 8-10 October
2012. 449-473,
[33] S. Saulnier, S. Guilan, Computational Study of Diesel Engine Downsizing using Two-Stage
Turbocharging. SAE Technical Paper. (2004) 2004-01-0929. 10.4271/2004-01-0929.
[34] Q. Zhang, C. Brace, S. Akehurst, R. Burke, G. Capon, L. Smith, S. Garrett, K. Zhang,
Simulation Study of the Series Sequential Turbocharging for Engine Downsizing and Fuel Efficiency.
SAE Technical Paper. (2013) 2013-01-0935. 10.4271/2013-01-0935.
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
[35] M. Amann, D. Ouwenga, Engine Parameter Optimization for Improved Engine and Drive
Cycle Efficiency for Boosted, GDI Engines with Different Boosting System Architecture. SAE Technical
Paper. (2014) 2014-01-1204. 10.4271/2014-01-1204.
Figure 1. Steady turbocharger gas-stand for investigating automotive turbochargers
Table 1. List of sensors adopted in the steady turbocharger gas-stand, including measuring range and
accuracy. Response and sampling frequency is not reported due to steady state tests being
performed
SENSOR RANGE ACCURACY
PRT -50 to +200 degC ±0.3 + 0.005*T
1.5mm K type
thermocouple-200 to 1260 degC 0.0075*T
V-cone mass flow 0 to 1200 Kg/h ±0.5%
Pressure 0 barA to 6 barA 0.25%
Turbo speed 0 to 400,000 rpm 0.1%
543
544
545
546
547548
549
550
551
552
Figure 2. Pressure measurement at four radial positions of a measurement section (a) and
temperature measurement at four radial positions of a measurement section with sensors tips
placed at 1/4, 1/3 and 1/2 of the diameter (b)
Table 2. Characteristics of HP and LP turbochargers, including compressor and turbine sizes,
maximum rotating speed and VGT position
Stage Compressor size Turbine size Maximum speed VGT position
HP 40cm 36cm 260Krpm 50% shut
LP 60cm 47cm 186Krpm 50% shut
553
554555
556
557
558
559
560
561
562
Figure 3. Schematic of two-stage turbocharging system with regulating valves: turbine by-pass (TBP)
and compressor by-pass (CBP), as installed in the turbocharger gas-stand
Figure 4. On the left hand-side, the two-stage turbocharging system installed in the steady
turbocharger gas-stand. The LP turbocharger is positioned on top, while, the HP turbocharger is
connected at the bottom of the exhaust manifold. On the right hand-side, LP and HP stage in stand-
alone configuration
Table 3. Test matrix for HP and LP turbochargers and the full two-stage system. The adiabatic test is
performed matching compressor outlet and turbine inlet temperature. Lubricating oil temperature is
controlled equally across the experiments at 90degC downstream the bearing housing
AdiabaticDiabatic at 773K TIT w/o
water-cooling
Diabatic at 773K TIT w/ water-
cooling at 360K
HP stage √ √ NA
LP stage √ √ √
Two-stage NA √ √
563
564
565
566567
568
569
570
571
572
573
574
575
Figure 5. At the top, equivalent performance map for the two-stage compressors. At the bottom,
equivalent performance map for the two-stage turbines. Two cases w/ and w/o water-cooling at the
LP compressor are considered
576577
578
579
580
Figure 7. At the top, LP compressor performance maps at adiabatic and diabatic are shown. At the
bottom, LP turbine performance maps at diabatic conditions are shown. The heat correction is
applied to both compressor and turbine apparent efficiencies for the estimation of effective
efficiencies
581582
583
584
585
586
Figure 6. At the top, HP compressor performance maps at adiabatic and diabatic are shown. At the
bottom, HP turbine performance maps at diabatic condition are shown. The heat correction is
applied to both compressor and turbine apparent efficiencies for the estimation of effective
efficiencies
Figure 8. At the left hand-side, LP compressor performance map for adiabatic and water-cooled
conditions at 90degC (WC 360K). At the right hand-side, apparent LP compressor T-T efficiency for
diabatic and water-cooled conditions at 360K
587588
589
590
591
592
593594
595
596
597
Figure 9. At the left hand-side, adiabatic and heat corrected compressor T-T efficiency is compared
with effective compressor efficiency in the presence of water-cooling (Cool Corr) for the LP
compressor. At the right hand-side, heat and cool corrected turbine net efficiency for the two
highest LP turbine corrected speeds tested in gas-stand
Figure 10. Relationship between adiabatic friction power and turbo speed for HP and LP
turbochargers. Best-fit curves for HP and LP turbochargers as dashed lines
598599
600
601
602
603
604605
606
607
Figure 11. Relationship between specific compressor heat flow and compressor mass flow for HP and
LP turbochargers. Best-fit curves for HP and LP turbochargers as dashed lines. Positive specific heat
flow is transmitted to the compressor
Figure 12. Relationship between specific turbine heat flow and turbine mass flow for HP and LP
turbochargers. Best-fit curves for HP and LP turbochargers as dashed lines. Positive specific heat flow
is escaping the turbine
608609
610
611
612
613614
615
616
617
Figure 13. Relationship between compressor cooling power and compressor mass flow for LP
turbocharger. Positive power values represent cooling action of water on compressor outlet flow
Figure 14. Diagram resembling the combination process of stand-alone HP and LP maps occurring
into a steady 1D model
618619
620
621
622623
624
625
Figure 15. At the top, compressor performance maps of equivalent two-stage system at diabatic
conditions, combined stand-alone HP and LP stages at diabatic conditions w/ and w/o heat
corrections. At the bottom, turbine performance maps obtained in the same conditions
Figure 16. Compressor performance maps of equivalent two-stage system with water-cooling at LP
stage (Equivalent w/ WC), combined stand-alone HP and water-cooled LP stages w/ and w/o cooling
corrections
626627
628
629
630
631632
633
634
635
Figure 17. Relationship between HP and LP compressor power in combined stand-alone maps at
diabatic conditions, with water-cooling at LP stage (w/ WC), with heat and cooling corrections
636637
638
639
640
