EPJ Web of Conferences 180, 02028 (2018) [Link]

1051/epjconf/201818002028
EFM 2017

CFD analysis of a twin scroll radial turbine

Jiří Fürst1 ,⋆ and Zdeněk Žák2 ,⋆⋆

1
Faculty of Mechanical Engineering, Czech Technical University in Prague, Karlovo nám. 13, 121 35 Praha 2, Czech Republic
2
Centre of Vehicles for Sustainable Mobility, VTP Roztoky, Přílepská 1920 252 63 Roztoky u Prahy, Czech Republic

Abstract. The contribution deals with the application of coupled implicit solver for compressible flows to
CFD analysis of a twin scroll radial turbine. The solver is based on the finite volume method, convective terms
are approximated using AUSM+up scheme, viscous terms use central approximation and the time evolution is
achieved with lower-upper symmetric Gauss-Seidel (LU-SGS) method. The solver allows steady simulation
with the so called frozen rotor approach as well as the fully unsteady solution. Both approaches are at first
validated for the case of ERCOFTAC pump [1]. Then the CFD analysis of the flow through a twin scroll radial
turbine and the predictions of the efficiency and turbine power is performed and the results are compared to
experimental data obtained in the framework of Josef Božek – Competence Centre for Automotive Industry.

1 Introduction especially for steady state flows at high Reynolds num-

ber. Therefore we developed in [6] an implicit coupled
CFD simulations play an important role in turbomachin- solver based on the symmetric lower-upper Gauss-Seidel
ery design and analysis. Although there are a number of procedure, see e.g. [10] or [11] and we further extend this
ready-made software packages (both commercial and in- solver for the case of turbomachinery flows using either
house), the reliable simulations remains still a challenging the frozen rotor or unsteady approaches.
task especially in the case of high rotational speed radial
turbines or low-pressure stages of axial turbines. The flow 2 LU-SGS solver for turbomachinery flows
passing through such types of turbines achieves high Mach
number and has to therefore considered as compressible. 2.1 Description of the solver
Most software packages for the simulation of com- The system of time averaged Navier-Stokes equations in
pressible flows are based on finite volume approach and conservative form are given as follows
use either a segregated or a coupled approach. When us-
∂ρ  = 0,
ing the segregated solver the equations expressing conser- + ∇ · (ρV) (1)
vation of momentum, density, and total energy are solved ∂t
independently in a loop in order to obtain steady state or ∂(ρU)
+ ∇ · (ρV  ⊗U − τ) + ∇p + ρΩ  ×U  = 0, (2)
unsteady solution. The algorithm is usually based on the ∂t
so called SIMPLE algorithm proposed originally Patankar ∂(ρE) 
 + pU  − τ · U
 − k∇T = 0,

+ ∇ · ρE V (3)
and Spalding in [2] for stationary case, or on the PISO ∂t
algorithm of Issa [3] for unsteady case. Although the seg- where ρ is the density, U  is the absolute velocity vector,
regated solvers have been extended also to compressible  =U  −Ω  × x is the velocity vector in the reference frame
V
flow problems including transonic or supersonic flows (see  p is the pres-
rotating with constant angular velocity Ω,
e.g. [4] or [5]), the efficiency of these method is in the
sure, τ is the (effective) stress tensor, E is the specific total
case of high speed flows usually worse than the efficiency
energy, k is the (effective) thermal conductivity, and T is
of implicit fully coupled solvers [6].
the temperature. The system is closed by the equation of
So far several attempts have been done in order to de-
state for ideal gas p/ρ = rT .
velop coupled solvers for OpenFOAM platform. Let us
The discretization in space is achieved using finite vol-
mention the solver based on the central-upwind scheme
ume method where convective fluxes are replaced by an
of Kurganov and Tadmor [7] or the density based solvers
approximate Riemann solver (in our case with the HLLC
on OpenFOAM-extend package based on AUSM+up [8]
flux, see [9]) and viscous fluxes are discretized using cen-
or HLLC [9] fluxes. Unfortunately all these solvers use
tral differences. The temporal discretization is achieved
explicit time-marching method which puts a strong limit
by using the implicit Euler method of first or second order,
to the time step and therefore the efficiency is very weak
thus for the first order method
⋆ e-mail: [Link]@[Link]
⋆⋆ e-mail: [Link]@[Link] W  in − ∆t R(
 in+1 = W W n+1 )i . (4)
|Ωi |

© The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution
License 4.0 ([Link]
EPJ Web of Conferences 180, 02028 (2018) [Link]
EFM 2017

Here W  = [ρ, ρu, ρE] and R( W)

 is the discretization of
spatial terms.
The resulting system of non-linear equations is then
solved by a matrix-free lower-upper symmetric Gauss-
Seidel method using the following two step procedure:

i − 1
 
Di ∆W (1) = −R ∆F (1) 
· S i j + λ ∗  (1)
i j ∆W ,
i j j
2 j<i

 (1) − 1
 
Di ∆W i = Di ∆W ∆F j · S i j + λ∗i j ∆Wj .
i
2 j>i

Here
 (1)
∆W  (1) − W
= W  in , (5)
i i
∆F(1)
i = F(W  (1) ) − F(W
i
 in ), (6)
∆Fi =  in+1 ) − F(W
F(W  in ), (7)
Fig. 1. Structured 2D multiblock mesh for ERCOFTAC centrifu-
λ∗ being the spectral radius of jacobains of fluxes and gal pump, rotating zone in red.
 
 |Ωi | 1  
∗
Di =   + λ  I. (8)
∆ti 2 j∈N i j 
i
• Flow rate Q = 0.292 m3 /s,
See [10] or [6] for details. U 2 D2
• Reynolds number Re = 2ν = 6.5 × 105 .
In order to solve turbomachinery flows the solver has
been extended to support calculations both in multiple ref-
The numerical solution has been calculated using a do-
erence frames allowing the so called frozen-rotor method
main with inlet diameter D0 = 200 mm and outlet diameter
and using the so called sliding mesh method [12].
D4 = 730 mm. The multi-block structured mesh provided
by Petit at al. [13] contains 93 886 quadrilateral cells with
2.2 Validation for the case of ERCOFTAC pump average near wall cell size about y+1 ≈ 20, see fig. 1.
In order to validate the solver and to assess suitability of
frozen rotor approach a simulation of flow through a two-
dimensional model of the ERCOFTAC radial pump has 2.2.1 Steady state frozen rotor simulation
been performed. The results are compared to experimental
data obtained by Ubaldi et al. [1]. The case was previously In order to assess the accuracy and suitability of frozen
studied by [13] assuming incompressible flow and detailed rotor approach two simulations have been run. First one
data together with numerical results were published in the using standard segregated solver based on SIMPLE loop
form of validation test case for evaluation of rotor-stator (namely simpleFoam) assuming incompressible flow and
interaction. the second one using coupled LU-SGS scheme for com-
The simplified model of the pump has 7 impeller pressible flow.
blades, 12 diffuser vanes and 6% vaneless radial gap. The
Left half of figure 2 shows the absolute velocity mag-
parameters of the model are given bellow:
nitude obtained with segregated solver and the right half
Impeller : shows results obtained with coupled LU-SGS solver. One
can see that both models give very similar flow field.
• Inlet blade diameter D1 = 240 mm,
Figures 3 and 4 show a comparison of the radial and
• Outlet blade diameter D2 = 420 mm, tangential components of the velocity at impeller outlet
• Number of blades Zi = 7, Dm /D2 = 1.02 obtained with segregated and LU-SGS
• Blade span z = 40.4 mm, solver with experimental data of [1]. Note that the velocity
is normalized by the impeller tip speed and the arc length
Diffuser : yi is normalized by the impeller pitch Gi = πD2 /Zi . These
• Inlet vane diameter D3 = 444 mm, results correspond to data obtained with another numerical
method based on segregated approach by Petit et al. [14].
• Outlet vane diameter D2 = 664 mm,
One can see that the frozen approach is capable to re-
• Number of vanes Zd = 12,
solve some basic features of the flow. Namely the radial
Operating conditions : component of the velocity corresponds well to experimen-
tal data. The tangential velocity is slightly over-estimated
• Rotational speed n = 2000 rpm, by both calculations. On the other hand the shape and po-
• Impeller tip speed U2 = 43.98 m/s, sition of wakes is completely wrong due to frozen rotor
• Flow rate coefficient φ = 4Q/(U2 πD22 ) = 0.048, approach.

2
EPJ Web of Conferences 180, 02028 (2018) [Link]
EFM 2017

Table 1. Efficiency evaluation of the ERCOFTAC pump.

Q (m3 /s) ∆p (Pa) M (Nm) η (%)

Steady state simulation

LU-SGS 0.296 931.12 1.48 88.9

simpleFoam 0.292 921.21 1.43 89.6

Unsteady simulation

LU-SGS 0.296 940.90 1.43 93.1

pimpleFoam 0.292 931.09 1.39 93.6

Fig. 2. Comparison of calculated velocity magnitude using

frozen rotor approach and segregated solver (left part) or coupled
2.2.2 Unsteady sliding mesh simulation
LU-SGS solver (right part).
In order to resolve unsteady rotor-stator interactions the
unsteady simulation with moving rotor zone had to be car-
ried out. The calculation has been run with the same mesh
as in the previous case using second order backward dif-
ferencing in time combined with dual time stepping tech-
nique in the case of LU-SGS solver. The physical time step
was set to ∆t = 3 × 10−5 s which corresponds to 1/1000 of
revolution period.
Figure 5 shows comparison of phase locked ensem-
ble averaged velocity components calculated with cou-
pled LU-SGS solver with results obtained with segregated
solver (pimpleDyMFoam) and with experimental data of
Ubaldi et al. [1]. The figures correspond to dimension-
less time t/T i = 0.126, 0.226, 0.326, and 0.426 where
T i = 2π/(ωZi) is the impeller blade passing time.
One can see that both solvers predict very similar ve-
locity at the impeller outlet. The most pronounced dif-
Fig. 3. Radial component of absolute velocity at the ERCOFTAC ference is in the tangential component of the velocity for
impeller outlet, frozen rotor approach. t/T i = 0.126, see fig. 5b. The coupled solver predicts
higher peak in the velocity component at yi /Gi ≈ 0.3
whereas the results obtained with segregated solver cor-
respond very well to experimental data.
The table 1 sumarizes results obtained with coupled
LU-SGS solver for compressible flow and with segregated
solver for incompressible flows. One can see that the LU-
SGS solver predicts both in steady and unsteady cases
slightly higher pressure rise ∆p and torque M. On the
other hand both solvers predict similar pump efficiency
η = ∆pQ/(Mω). The efficiency predicted by steady-state
frozen rotor simulations is slightly lower than the one cal-
culated with unsteady method, nevertheless the difference
isn’t excessively large and therefore the frozen rotor ap-
proach is a viable method for estimating efficiency of the
pump.

Fig. 4. Tangential component of absolute velocity at the ER- 3 Analysis of twin scroll turbine
COFTAC impeller outlet, frozen rotor approach.
Twin scroll radial centripetal turbocharger turbines are
used in the automotive industry. The twin scroll design
is beneficial in conjunction with the internal combustion

3
EPJ Web of Conferences 180, 02028 (2018) [Link]
EFM 2017

(a) Radial velocity, t/T i = 0.126. (b) Tangential velocity, t/T i = 0.126.

(c) Radial velocity, t/T i = 0.226. (d) Tangential velocity, t/T i = 0.226.

(e) Radial velocity, t/T i = 0.326. (f) Tangential velocity, t/T i = 0.326.

(g) Radial velocity, t/T i = 0.426. (h) Tangential velocity, t/T i = 0.426.

Fig. 5. Ensemble averaged velocity components at the ERCOFTAC impeller outlet.

4
EPJ Web of Conferences 180, 02028 (2018) [Link]
EFM 2017

Fig. 7. Computational mesh for twin scroll turbine simulations.

Fig. 6. Computational domain for twin scroll turbine simula-
tions.

engine with firing interval between cylinders around 240

degrees in exhaust manifold branches, i.e. six cylinder en-
gines. The advantages consist in the higher efficiency of
a turbine and better engine response during the transients
due to better utilization of exhaust gas energy. The twin
scroll turbines are also used for the four cylinder engines
to improve the engine behaviour during transients. The fir-
ing interval between cylinders in mentioned case is 360◦
in exhaust manifold branches, thus out of optimum and
the high pulsating flow upstream of a turbine cannot be
utilized as effective as in the case of six cylinder engine. Fig. 8. Surface mesh at the rotor for twin scroll turbine simula-
The tested twin scroll turbine with symmetric design of tions.
sections is used for turbocharging of a six cylinder Diesel
engine. The impeller has 12 vanes. The impeller diameter
at inlet equals to 76.6 mm and averaged outlet diameter
steady frozen rotor simulation whereas second order back-
equals to 46.7 mm.
ward Euler method combined with local pseudo-time step-
ping was used in the transient case.
3.1 Simulation setup The flow regime was characterized by the following
boundary conditions:
Due to high rotational speed together with high expan-
sion ratio, the maximal Mach number exceeds one in some • Inlets - the value of total pressures p0A , p0B and total
cases and the flow has to be definitely considered as a temperatures T 0A , T 0B was prescribed at both inlets. The
compressible gas. Simulations have been run with LU- velocity direction was normal to the inlet planes and the
SGS solver which is developed especially for dealing with inlet turbulence intensity T u = 5% together with mixing
transonic flows. The domain (see figure 6) was divided to length scale l = 2.8 mm was assumed in all cases.
three zones: the stationary inlet zone containing both inlet
• Outlet - the value of static pressure pout was prescribed
tubes and volutes, the rotating zone containing turbine ro-
according to given expansion ratio.
tor, and the stationary zone including outlet tube. Each
zone was meshed independently using snappyHexMesh • Shroud - the no-slip condition was assumed together
software and the three parts were then glued into single with ∂p/∂n and given wall temperature T wall at the
mesh with interfaces between zones. The final unstruc- shroud. Due to relatively coarse mesh the wall functions
tured hexa-dominant mesh contained 2.2 × 106 cells with were used for turbulence model.
near-wall mesh size y+1 < 160, (60 at the rotor) with aver-
• Rotor - the same set of boundary conditions as for the
age y+1 ≈ 10, see fig. 7 and 8..
shroud was assumed here with the exception of velocity
The spatial discretization was performed using second
at the rotor wall which was set to the velocity of the
order scheme with Barth-Jaspersen limiter [15] and HLLC
rotor.
flux [9] for convective terms and central scheme for diffu-
sive terms. The implicit Euler method with local time step- The gas was governed by the ideal equation of state
ping was used for advancing in pseudo-time in the case of with constant heat capacity c p corresponding to air at the

5
EPJ Web of Conferences 180, 02028 (2018) [Link]
EFM 2017

by the combination of the compressor with larger wheel as

a load and relatively high temperature upstream of the tur-
bine about 1100 K. With the temperature of exhaust gases
about 900 K upstream of the turbine, the turbine was able
to run with isentropic efficiency close to optimum. The tur-
bine, driven by cold air only, was unloaded and achieved
high blade speed ratios. Three levels of impeller admis-
sion were measured in our case - full, partial (throttling in
one section) and extreme partial with closed section.

3.3 Results

Although the experimental data covers different partial

levels of admission, the CFD study was limited to cases
Fig. 9. Twin scroll turbocharger test bed with separated sections with uniform admission only. The results of simulations
upstream of a turbine were postprocessed in order to make comparison with ex-
perimental data. First, the reduced mass flow rate defined
by √
T 0AB 5
ṁred = (ṁA + ṁB ) 10 , (9)
p0AB
was evaluated. Here ṁA and ṁB are the mass flow rates
through inlet A and B, T 0AB = (T 0A + T 0B )/2, and p0AB =
(p0A + p0B )/2. Next, the torque M has been evaluated and
used for the rotor power calculation

Protor = Mωrotor . (10)

The ventilation losses were estimated using empirical for-

mula
Pvent = Kρ2 D22 U23 Re−0.2
2 (11)
where K = 0.07 is an empirical constant, D2 is the rotor
Fig. 10. Turbocharger test bed (down from top) - uniform ad-
diameter, U2 = ωrotor D2 /2 is the rotor inlet edge speed
mission; partial admission (throttling in one turbine section); ex- and Re2 = ρ2 D2 U2 / is the Reynolds number and finally
treme partial admission with closed section; backflow the turbine efficiency was evaluated as
Protor − Pvent
η=    κ−1  · 100%. (12)
(ṁA + ṁB )c p T 0AB 1 − pp0AB
out κ

temperature T wall . The effects of turbulence were modeled

by Menter’s SST turbulence model [16].
Note that in the case of transient simulations calcula-
The table 2 sumarizes boundary conditions and gas
tions have been started from steady the initial condition
properties used for the simulations.
corresponding to steady state frozen rotor simulation and
the ten revolutions of the rotor have been calculated. The
3.2 Experimental data resulting mass flow rate and torque were then calculated
by averaging over the last revolution. Figure 11 shows a
The goal of the experiments on the steady flow test bed time dependent torque obtained by transient simulation for
(see fig. 9), with separated sections upstream of a turbine, TP1 regime. From the amplitude spectrum displayed at the
was to obtain a sufficient set of data, which describes the figure 12 one can see that time dependent signal contains
twin scroll turbine behaviour under different level of the only one important harmonics corresponding to the blade
impeller admission. The specific hot gas stand with open passage frequency. Similar results were obtained also for
loop is suitable for achievement of the arbitrary level of other regimes.
turbine impeller admission. The unequal partial level of Figures 13, 14, and 15 display the distribution of
admission is reached via the throttling in a turbine section the pressure at solid and rotating walls, the absolute
and extreme partial admission by closure of one section, Mach number and velocity vectors in the longitudinal cut
see the sketch at fig. 10. through the domain, and the relative Mach number in the
The selected twin scroll turbine was tested at different rotor zone for TP1 regime. One can see that both absolute
pressure ratios, relatively broad range of load and under and relative Mach numbers exceed unity and therefore the
different levels of the impeller admission. The turbine load compressibility plays an important role in the simulations.
is described by the blade speed ratio - BSR. Low blade The table 4 shows the comparison of experimental data
speed ratios, i.e. high load of the turbine, were achieved with the results of steady and unsteady simulations.

6
EPJ Web of Conferences 180, 02028 (2018) [Link]
EFM 2017

Table 2. Boundary conditions and gas properties for twin scroll turbine analysis

Run p0A (Pa) p0B (Pa) pout (Pa) T 0A (K) T 0B (K) T wall (K) RPM (rpm) c p (J/kg/K) κ (-)

TP 1 2.070 × 105 2.077 × 105 9.232 × 104 877.030 891.656 800 104 850 1097.997 1.354

TP 2 2.066 × 105 2.075 × 105 9.236 × 104 874.636 889.859 800 109 011 1097.997 1.354

TP 3 2.084 × 105 2.092 × 105 9.146 × 104 1077.040 1080.259 1000 88 776 1142.580 1.336

TP 4 1.544 × 105 1.550 × 105 9.512 × 104 873.127 891.125 800 85 227 1097.997 1.354

TP 5 2.063 × 105 2.067 × 105 9.333 × 104 288.478 288.191 288 73 521 1005.051 1.400

TP 6 1.545 × 105 1.550 × 105 9.296 × 104 1075.105 1080.552 1000 72 814 1142.580 1.336

TP 7 1.550 × 105 1.554 × 105 9.310 × 104 1079.464 1084.990 1000 67 489 1142.580 1.336

TP 8 1.216 × 105 1.219 × 105 9.654 × 104 881.779 888.423 800 63 430 1097.997 1.354

TP 9 1.534 × 105 1.535 × 105 9.623 × 104 289.580 289.341 288 57 048 1005.051 1.400

TP 10 1.219 × 105 1.223 × 105 9.629 × 104 874.801 888.021 800 54 418 1097.997 1.354

Fig. 11. Time dependent torque as a result of transient simulation Fig. 12. Amplitudes of FFT of time dependent torque for TP1
for TP1 regime. regime.

proximated as
Figures 16 and 17 show graphically the reduced mass
flow rates and turbine efficiency as a function of turbine ṁred ≈ a0 +a1 Π +a2 nred +a11 Π 2 +a12 Πnred +a22 n2red , (13)
expansion ratio Π = p0AB /pout . Each number represents
where the coefficients are a0 = −3.902, a1 = 8.192,
corresponding regime (see table 2) and its color corre-
a2 = −0.621 × 10−3 , a11 = −2.213, a12 = 0.473 × 10−3 ,
sponds to a method for obtaining the result: black color
and a22 = −0.105 × 10−6 . The figure 18 compares the re-
means experimental data, red color means steady state
duced mass flow rate predictions given by the model (13)
CFD, and blue color represents transient CFD. Both fig-
with experimental data and with the results of CFD sim-
ures contains also lines displaying results of steady state
ulations. One can see that the differences between the
calculation with three fixed rotation speeds 50 000 rpm,
model predictions and the results are less than 2 % with
75 000 rpm and 100 000 rpm with T 0AB = 850 K.
the exception of TP5 and TP9 regimes. In TP5 regime the
A model for reduced mass flow rate and turbine effi- unsteady simulation gives lower value of the mass flow
ciency has been built from the frozen rotor results at three rate whereas experimental data corresponds well to the
fixed rotation speeds. The reduced mass flow rate is ap- model. Similarly in the TP9 the unsteady CFD predicts

7
EPJ Web of Conferences 180, 02028 (2018) [Link]
EFM 2017

Table 3. Expansion ratio

√ Π = p0AB /pout and reduced rotation
speed nred = RPM/ T 0AB for twin scroll turbine analysis.

√
Run Π (−) nred (rpm/ K)

TP 1 2.246 3526

TP 2 2.242 3670

TP 3 2.283 2703

TP 4 1.627 2870

TP 5 2.213 4330

TP 6 1.665 2218
Fig. 13. Pressure distribution at solid walls in the twin scroll
turbine, regime TP 1.
TP 7 1.667 2052

TP 8 1.261 2132

TP 9 1.595 3353

TP 10 1.268 1833

Table 4. Comparison of reduced mass flow rate and turbine

efficiency obtained with the experiment (columns “Exp”), by the
steady state simulation (columns “Steady”) and transient
simulation (columns “Trans”).

√
ṁred (kg s−1 K/bar) η (%)
Run
Exp Steady Trans Exp Steady Trans

TP 1 3.370 3.411 3.365 64.6 65.3 65.3

Fig. 14. Mach number in the cut through twin scroll turbine,
regime TP 1. TP 2 3.303 3.371 3.309 64.4 64.2 63.9

TP 3 3.613 3.570 3.540 60.8 61.4 61.7

TP 4 3.030 3.066 3.021 68.1 62.1 62.4

TP 5 3.092 3.112 2.839 52.4 54.6 45.7

TP 6 3.321 3.329 3.300 64.8 63.4 62.9

TP 7 3.373 3.371 3.352 61.0 61.8 61.7

TP 8 2.332 2.242 2.219 70.5 58.2 57.7

Fig. 15. Relative Mach number in cuts through rotor part of twin TP 9 2.815 2.665 2.487 61.1 47.4 45.0
scroll turbine, regime TP 1.
TP 10 2.457 2.444 2.424 70.9 64.4 64.2

8
EPJ Web of Conferences 180, 02028 (2018) [Link]
EFM 2017

Fig. 16. Reduced mass flow rate for twin scroll turbine (exper- Fig. 18. Comparison of the mass flow rate model (13) with ex-
imental data depicted as black numbers, steady state simulation perimental and CFD results.
as red numbers, and transient simulation as blue numbers).

Fig. 19. Comparison of the efficiency model (14) with experi-

Fig. 17. Efficiency of twin scroll turbine (experimental data de- mental and CFD results.
picted as black numbers, steady state simulation as red numbers,
and transient simulation as blue numbers).
unsteady simulations are characterized by lower inlet tem-
perature resulting in higher Mach numbers in the turbine
lower mass flow rate whereas the experimental data gives when comparing to other regimes.
a slightly higher value.
Similarly the efficiency can be approximated as 4 Conclusion
η̇ ≈ b0 + b1 Π + b2 nred + b11 Π 2 + b12 Πnred + b22 n2red , (14) The LU-SGS solver has been extended for turbomachin-
ery flows using stationary frozen-rotor and unsteady slid-
with b0 = 53.333, b1 = 26.649, b2 = −11.255 × 10−3 , ing mesh approaches. The solver has been validated using
b11 = −38.737, b12 = 44.195 × 10−3 , and b22 = the ERCOFTAC centrifugal pump comparing the results
−12.994 × 10−6 . both with experimental data and with the pressure based
The figure 19 compares the predicted efficiency given solver from the OpenFOAM package. Finally the solver
by the model (14) with the efficiency evaluated from the has been used for the analysis of the flows through a twin-
experimental data and from the CFD simulations. One can scroll turbine. The comparison of simulation results using
see that most of the data falls into 5 % region, however both approaches with the experimental data shows that the
the experimental data give slightly higher efficiencies in mass flow rate is predicted within 2 % error by both com-
regimes TP4, TP8, TP9, and TP10 whereas the unsteady putational approaches and that a reasonable agreement be-
simulations predict lower efficiency in regimes TP5 and tween predicted and measured efficiency can be achieved.
TP9. Although the frozen rotor approach completely disregards
Note that the regimes TP5 and TP9 where the mod- the unsteady character of turbomachinery flow, the cur-
els for ṁred and η differ from the experimental data and rent study shows that it can be viable alternative to much

9
EPJ Web of Conferences 180, 02028 (2018) [Link]
EFM 2017

more computationally demanding sliding mesh method for 6. J. Fürst, in EPJ Web of Conferences (2017), Vol. 143,
flows in radial turbines. p. 02027
7. A. Kurganov, E. Tadmor, J. Comput. Phys. 160, 241
Acknowledgements. Authors acknowledge sup- (2000)
port from the EU Operational Programme Re- 8. M.S. Liou, J. Comput. Phys. 214, 137 (2006)
search, Development and Education, and from 9. E.F. Toro, in Riemann Solvers and Numerical Meth-
the Center of Advanced Aerospace Technology ods for Fluid Dynamics (Springer Berlin Heidelberg,
(CZ.02.1.01/0.0/0.0/16_019/0000826), Faculty of Berlin, Heidelberg, 1997), pp. 293–311
Mechanical Engineering, Czech Technical University in 10. J. Blazek, Computational fluid dynamics : principles
Prague and support by TACR, project no. TE01020020. and applications (Butterworth-Heinemann, 2015)
11. C. Shen, X.l. Xia, Y.z. Wang, F. Yu, Z.w. Jiao, Adv.
References Eng. Softw. 91, 80 (2016)
1. M. Ubaldi, P. Zunino, G. Barigozzi, A. Cattanei, J. 12. J. Fürst, in 12th OpenFOAM workshop (Exeter, 2017)
Turbomach. 118, 41 (1996) 13. O. Petit, M. Page, M. Beaudoin, H. Nilsson, in IAHR
2. S.V. Patankar, D.B. Spalding, Int. J. Heat Mass Transf. International Meeting of the Workgroup on Cavitation
15, 1787 (1972) and Dynamic Problems in Hydraulic Machinery and
Systems (2009), pp. 523–532
3. R. Issa, J. Comput. Phys. 62, 40 (1986)
14. O. Petit, H. Nilsson, Int. J. Rot. Mach. 2013, 1 (2013)
4. J. Benajes, J. Galindo, P. Fajardo, R. Navarro, J. Appl.
Fluid Mech. 7, 673 (2014) 15. T. Barth, D. Jespersen, in 27th Aerospace Sciences
Meeting (AIAA, Reston, Virigina, 1989)
5. L. Mangani, M. Darwish, F. Moukalled, in Proceed-
ings of ASME TURBO EXPO 2014 (ASME, Düssel- 16. F.R. Menter, AIAA Journal 32, 1598 (1994)
dorf, 2014), p. 13

10