Analyses guided optimization of wide range and

high efficiency turbocharger compressor

H. Sun, D. Hanna, Ford Motor Co.

L. Hu, J. Zhang, Ming-Chia Lai, Wayne State Univ.
E. Krivitzky, C. Osborne, ConceptsNREC
NETL Project Manager: Ralph Nine
DOE Contract: DE-FC26-07-NT43280
 Single turbo?

The focus is to
move high
efficiency
island on a
compressor
map to cover
customer
driving cycles
 Engine Performance
Critical Operating Points
Targets
(Indicating map width and performance targets)
18 compressor and 12 turbine

Turbine/Compressor
Conceptual Design and 1D
analytical design iterations

4 design/bench test iterations

Simulation/map generation

3D Geometric Specification, CFD

Performance Map Validation

3D CAE Structure Analyses for

HCF/LCF

Fabrication and Flow Bench Test

Engine Dyno Test

Numerical simulation is the key to guide the design iterations !! 3

 Validation of compressor model

 Compressor with arbitrary blades;

 6 main/splitter blades
 Low solidity vaned diffuser with different setting angle
 Solidity=0.7 with 7 vanes

 Test at 108krpm, vaneless diffuser also tested.

 Numerical model

 Mesh : Auto/Grid, Structure O-4H mesh, Y+≈3

 Solver: Fine/Turbo Euranus
 Turbulence Model: Spalart-Allmaras
 R/S interface: non reflecting BC
 Boundary Condition
• Inlet: Tt, Pt, flow direction
• Outlet: Ps/ target mass flow
 Test of Compressor Performance
 Two setting angle of diffuser vane and vaneless diffuser
82% 3.2

78% 3.0

74% 2.8

70%

Pressure ratio(t-t)
Efficiency(t-t)

2.6

66% 2.4
vaneless diffuser
62% 64 vaned diffuser 2.2 vaneless diffuser
67 vaned diffuser
58% 2.0 64 vaned diffuser
67 vaned diffuser
54% 1.8

50% 1.6
40% 50% 60% 70% 80% 90% 100% 40% 50% 60% 70% 80% 90% 100%
Relative mass flow rate Relative mass flow rate

 Improved efficiency at low end;

 Surge extension with increase of setting angle;
 Penalty of choke flow rate: further reduction of setting angle would possibly improve the choke flow.
 Variable vaned diffuser is the best.
 Numerical simulation of compressor performance
 Reduction of setting angle is included: 61˚ setting angle
0.82 3.0

0.78 2.8
Unsteady simulation
0.74
2.6

Pressure ratio(t-t)
0.70
Efficiency(t-t)

2.4
0.66
vaneless diffuser 2.2
0.62
67 vaned diffuser vaneless diffuser
64 vaned diffuser 2.0 67 vaned diffuser
0.58
61 vaned diffuser
64 vaned diffuser
0.54 1.8
61 vaned diffuser
0.50 1.6
40% 50% 60% 70% 80% 90% 100% 40% 50% 60% 70% 80% 90% 100%
Relative mass flow rate Relative mass flow rate

 Increase of setting angle, compressor surge could be extended;

 Improved efficiency especially at low end;
 Reduction of setting angle could improve performance near choke;

Compared with test data, CFD prediction matches well with the test data except the
near surge areas.
 Transient simulation is needed for surge prediction
 Unsteady simulation: investigate how the vaned diffuser affects the
compressor surge flow rate;
 Sliding mesh-diffuser vane was scaled to 6 vanes;
 Same BC, mesh applied @ 61˚,67 ˚vaned diffuer;
 Impeller+diffuser simulated.
 Unsteady CFD analysis -pressure oscillation
 Fluctuation of Ps at leading edge of impeller:
124
122
122
T1 121
120
Static pressure(kPa)

Static pressure(kPa)
120
T2
118
119
T3
116
118
114 T4
117

112 T5
116

110
61˚ 115
67˚
108
114
0 2 4 6 8 10 12 14 16 18
0 2 4 6 8 10 12 14 16 18
Impeller rovolutions
Impeller rovolutions

 Both Ps oscillations are visible;

 With 61˚ vaned diffuser, more fluctuation in Ps
 The impeller with 61˚ vaned diffuser is less stable
 Unsteady CFD analysis -pressure oscillation
 They differ in the frequency domain:

2000

67˚
2500
61˚
2000 1500

Magnitude
Magnitude

1500
1000

1000

500
500

0 0
0 10000 20000 30000 40000 0 10000 20000 30000 40000
Frequency Frequency

 Two frequencies: 10.8kHz(blade passing frequency)

 450Hz- related to the stall phenomenon.
 61˚ vaned diffuser: the magnitude at 450Hz is ~3 times that of 67˚ vaned diffuser, i.e. impeller matched
with 61˚ vaned diffuser is less stable.
 Unsteady CFD analysis - vorticity

 Flow field investigation@5 time instants, 95% Spanwise:

T1 T1
T2 T3 T4 T5 T1 T4

61˚ 67˚
 61˚ vaned diffuser: the high vorticity gradually spills out of the impeller passage, unstable-
indicating impeller stall;
 67˚ vaned diffuser: the high vorticity stays inside the impeller passage, relatively stable.
 Compressor with 67 ˚ vaned diffuser is more stable.
 Unsteady CFD analysis - entropy

 Flow field investigation@5 time instants, 95% Spanwise:

T1 T2 T3 T4 T5 T1 T4

61˚ 67˚

 Entropy distribution:
 61˚ vaned diffuser: high entropy area gradually spills out of the impeller passage, entropy increased
– impeller unstable;
 67 ˚ vaned diffuser: more stable – increased diffuser vane angle is able to stabilize impeller flow.
CFD to guide optimization of compressor casing treatment

No casing treatment
Lean forward

Lean backward
Lean backward w/ wider slot
 0.80

0.78

0.76

0.74
Adv comp tech
Conventional 2
Simulation
@ 120krpm
Conventional 1
Eff (t-t)

0.72 (with volute)

0.70

0.68

0.66 3.4

3.3
0.64
0.50 0.55 0.60 0.65 0.70 0.75 0.803.2 0.85 0.90 0.95 1.00
Mass flow(normalized)
3.1

3.0
Optimization of casing
Pr (t-t)

treatment can extend flow 2.9 Adv comp tech

Conventional 2

range/surge margin 2.8 Conventional 1

without compromising 2.7

efficiency 2.6

2.5

2.4
0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00
Mass flow(normalized)
 Conclusions

 CFD simulation tool (Numeca) has been demonstrated to be very

effective to guide the compressor optimization (the compressor
went through 18 design iterations and only 4 designs went to
hardware build/test);
 The accuracy of CFD tool has been validated in steady state and
transient simulations for both vaned and vaneless compressors;
 Variable geometry compressor can extend the flow range
dramatically but may have reduced peak efficiencies;
 Optimization of compressor casing treatment shows the potential
of improvement in both surge margin and flow capacity without
compromising of efficiency.