Documente Academic
Documente Profesional
Documente Cultură
Spiral flow induced destabilizing force analysis and its reduction with
a novel helix-comb gas seal
Mingjie Zhang, Jiangang Yang ∗
National Engineering Research Center of Turbogenerator Vibration, School of Energy and Environment, Southeast University, Nanjing 210096, Jiangsu Province,
China
a r t i c l e i n f o a b s t r a c t
Article history: The circumferential spiral flow is the main cause of destabilizing tangential force. To control the
Received 15 January 2020 circumferential spiral flow, the helix-comb seal is presented in this study. With the guiding role of
Received in revised form 26 April 2020 helical teeth, the negative circumferential spiral flow can be generated. It can significantly suppress
Accepted 8 June 2020
the circumferential flow within the seal gap. A 3D computational fluid dynamics method is employed
Available online 16 June 2020
Communicated by Xinqian Zheng
to investigate the performance of helix-comb seal. From the parameter analysis, it is found that the
minimum tangential force can be obtained with the helix angle of 15 deg. Tangential force decreases
Keywords: with the increasing helical tooth number and helical tooth height. Considering the rubbing between the
Helix-comb seal helical teeth and rotor, helical tooth height should be slightly smaller than the height of its adjacent
Circumferential spiral flow comb tooth. Comparisons are conducted between the labyrinth seal (with and without swirl brakes) and
Tangential force helix-comb seal. Results reveal that the three seals have a comparable sealing performance. Considering
Leakage the destabilizing force, the helix-comb seal has the lowest tangential force and the best stability. The
Computational fluid dynamics
helix-comb seal provides a novel method to control self-excited vibration.
© 2020 Elsevier Masson SAS. All rights reserved.
https://doi.org/10.1016/j.ast.2020.105997
1270-9638/© 2020 Elsevier Masson SAS. All rights reserved.
2 M. Zhang, J. Yang / Aerospace Science and Technology 105 (2020) 105997
Nomenclature
Swirl brakes (Fig. 1(c)) are a series of vanes provided at seal en-
trance to impede or direct the circumferential flow to reduce the
destabilizing tangential force and cross-coupled stiffness. Wachter
and Benckert [11] demonstrated the effectiveness of swirl brakes in
reducing the destabilizing tangential force. Nielsen [17] addressed
swirl brakes design and flow structure description. It was shown
that the three-dimensional vortical structure of the flows could be
efficiently used to reduce the swirl at seal inlet. In order to in-
vert the circumferential direction of inlet flow, the “negative-swirl”
brakes [18] were invented. Investigations showed that the swirl
brakes reduced destabilizing cross-coupled stiffness by a factor
of 0.8, and the negative-swirl brakes changed the sign of cross-
coupled stiffness at an appreciable magnitude [18]. The deswirl
effect of swirl brakes is sometimes limited due to their location
away from seal entrance. The implantation of swirl brakes would
change the flow pattern behind the final stage impeller backwall
[15]. Hence, further researches are needed to limit the circumfer-
ential flow more effectively.
Helically grooved seals [19] have been proposed as a means of
achieving greater reduction in leakage. The sealing mechanism is
the pumping effect [20], by which working fluid is pushed back
toward the higher-pressure side along the helical grooves by the
spinning motion of rotor. The pumping effect increases with rota-
tional speed and fluid viscosity [21]. Helically grooved seals have
been widely used as liquid seals [19,20]. Zero leakage can be ob-
tained because of the significant pumping effect with high liquid
viscosity. For gas seals, the gas viscosity is so small that the pump-
Fig. 1. LS and two typical seals for circumferential spiral flow control. ing effect is not obvious. The larger flow passage cross-sectional
area of the grooves results in an increase of leakage flow rate
[22]. This disadvantage restricts its application in gas seal field.
jection. Kim [14] employed the CFD perturbation model to predict
the effects of shunt injection on the rotordynamics of LS. Predicted However, the helical groove concept can be used to control the
results revealed that drastically reduced (negative) cross-coupled destabilizing circumferential spiral flow.
stiffness could be obtained with the use of shunt injection. The dis- This study proposes a new kind of helix-comb seal (HCS) to
advantage of shunt injection is the performance penalty involved control the circumferential spiral flow and tangential force. With
in diverting flow from compressor discharge and recirculating it the guiding role, the helical teeth are designed to generate the
through the machine [13]. The use of shunt injection can result in negative circumferential spiral flow inside the seal cavities. The
a large leakage increase [13]. In addition, the shunt line from the circumferential velocity within the seal gap is reduced substan-
discharge volute to the seal front section should be designed and tially. The tangential force is also significantly reduced. To study
installed carefully to assure the pressure drop across the shunt line the dependency of the seal performance on the structure param-
is minimized [15]. These restrict the further application of shunt eters, three major parameters are investigated including helical
injection. Swirl brakes [16] is one promising alternative of shunt tooth number, helical tooth height, and helix angle. To evaluate
injection to control the circumferential spiral flow. and demonstrate the effectiveness of HCS in leakage control and
M. Zhang, J. Yang / Aerospace Science and Technology 105 (2020) 105997 3
tangential force reduction, comparisons between the LS (with and helical grooves of equal angular size. The helix direction is set
without swirl brakes) and HCS are conducted. to oppose the rotational direction of the rotor with the goal of
generating the negative circumferential spiral flow inside the heli-
2. Helix-comb seal for circumferential spiral flow control cal grooves. In order to avoid the rubbing between the rotor and
helical teeth, the helical tooth height is smaller than comb tooth
For the LS, the circumferential spiral flow is dominant for the height.
destabilizing tangential force. Retarding circumferential velocity is
a direct and effective means to reduce the destabilizing tangen- 3. Numerical approach
tial force. Inspired by the helical groove concept, HCS is proposed.
With the guiding role of helical teeth, the fluid moves along the 3.1. Seal model
helical grooves. If the helix angle is set to oppose the rotational di-
rection, the direction of groove flow becomes the opposite to that The LS, LS with swirl brakes, and HCS are used for CFD com-
of rotor spinning motion. The circumferential flow within the seal parison analysis in this study. Fig. 3 shows the seal geometries of
gap is weakened by the groove flow. The destabilizing tangential the three seals. The LS has four rectangular comb teeth on the
force is reduced. stator and a smooth rotor. For the LS with swirl brakes, a series
Fig. 2 shows the specific structure of HCS. The tooth-on-stator of vanes are placed at the entrance of LS. For the HCS, the spe-
configuration is used. Several helical teeth are inserted equally cially designed helical teeth are inserted equally spaced around
spaced around the circumferential in the annular cavities com- the circumferential in the annular cavities of LS. In order to have a
posed of comb teeth. Each annular cavity is divided into isolate straightforward comparison, the three seals share the same comb
4 M. Zhang, J. Yang / Aerospace Science and Technology 105 (2020) 105997
Table 1 Table 2
Seal dimensions. Operating and boundary conditions.
Ft
k= (1) 4. Results and discussion
e
4.1. Flow field analysis of helix-comb seal
where k is the cross-coupled stiffness, F t is the tangential force,
and e is the eccentricity.
The major feature of HCS is that the helical teeth are used to
Fig. 9 compares the calculated leakage flow rate with experi-
divide the annular cavities into several smaller helical grooves. The
mental results. Both the calculated leakage flow rate and the mea-
flow field in this section is the key to understand the working
sured reach their maximum at zero entry swirl. The maximum
mechanisms of HCS. For comparison, the flow field results of LS
difference of leakage flow rate is less than 9%. Fig. 10 compares are also plotted. The results are obtained for an operating con-
the calculated cross-coupled stiffness with experimental results. dition with a rotational speed of 6000 rpm, inlet pressure of 0.6
The cross-coupled stiffness is well predicted. The numerical results MPa, outlet pressure of 0.3 MPa, and high preswirl (λ = 1). For
capture the correct preswirl dependence of cross-coupled stiffness. the HCS, the helical tooth number (each cavity) is 8, the helical
The cross-coupled stiffness increases almost linearly with the in- tooth height is 3.45 mm, and the helix angle is 15 deg. Fig. 12
creasing preswirl velocity. The prediction difference is due to that depicts the contours of velocity for the LS and HCS. The velocity
uniform velocity and pressure boundary conditions used in CFD vectors in the second cavity of the two seals are also plotted in
analysis, which is not consistent with the practical non-uniform in Fig. 13. Working fluid is accelerated at the first comb tooth tip be-
the experiments. cause the flow area suddenly contracted at the small clearance. A
Due to that the detailed static pressure distribution was not part of the pressure energy is transformed into kinetic energy. The
reported in the reference [28], another example of [30] with the high-speed jet drives a large vortex inside the subsequent chamber
measured static pressure was adopted. The static pressure distri- because of the viscous force between the jet and cavity flows. As
bution of the CFD method was compared with the experimental a result, the kinetic energy is dissipated into heat. For the LS, the
data, as shown in Fig. 11. The calculated static pressure distribu- high-speed jet passes through the next comb tooth and repeats the
tion and the measured decrease along the axial flow direction. The previous process. For the HCS, a high-speed jet is observed at comb
maximum difference is less than 3.2%. The CFD results show a good tooth tip. The downstream helical tooth clearance faces a straight
agreement with the experimental data. jet impingement escaping from the previous comb tooth clearance.
In general, the present CFD method has reasonable accuracy to An additional flow contraction occurs at the helical tooth tip. The
predict the static pressure distribution, leakage flow rate, and tan- contraction effect is not obvious at helical tooth tip because the
gential force of annular gas seals. helical tooth clearance is larger than comb tooth clearance. Af-
6 M. Zhang, J. Yang / Aerospace Science and Technology 105 (2020) 105997
Fig. 6. Circumferential flow velocity vectors for the LS with and without FAN bound-
ary.
ter the helical tooth, the second vortex zone occurs. Then the jet
passes through the next comb tooth and repeats the previous pro-
cess. Compared with LS, which has only one recirculation zone,
HCS has a twin vortex pattern in its cavity.
The throttling process across each seal tooth and the kinetic
energy dissipation within the subsequent chamber lead to the de-
crease in pressure. Pressure gradually drops from the first to the Fig. 7. Meshes of three types of seals.
last tooth, as shown in Fig. 14. The change of cavity pressure is
not significant when the jet flows through helical tooth tip. The tion along the axial direction. In this picture, the four calculating
pressures in the chambers on both sides of a helical tooth are al- points locate at comb teeth clearance. The average circumferential
most equal. The pressure drop mainly occurs at comb tooth tip. velocities of both seals decrease along the axial direction. For the
It confirms that the helical teeth do not play the effective role in LS, this decrease is because the circumferential velocity is high at
throttling process. Comb teeth play the key role in sealing gas. high preswirl, the shear stress exerted by stator is larger than that
Fig. 15 shows the streamlines traveling through the LS and HCS. exerted by rotor, the net effect is to slow the circumferential veloc-
The streamlines start from a position near the first comb tooth. For ity. For the HCS, the negative spiral flow inside the helical grooves
the LS, circumferential spiral flow occurs inside annular cavity. It significantly increase the circumferential flow resistance. As a re-
is driven by the viscous shear stress due to rotor rotation. The cav- sult, the descending rate of average circumferential velocity of HCS
ity flow and gap flow have the same circumferential flow direction. is larger than that of LS. At seal outlet, the average circumferential
For the HCS, the flow inside helical grooves is guided by the helical velocity of HCS is less than 1 m/s, the ratio of average circumfer-
teeth. The groove flow and gap flow have opposite circumferential ential velocity to the rotor surface velocity is less than 0.005.
flow direction because the helix angle is set to oppose the rota-
tional direction. The groove flow can weaken the circumferential 4.2. Structure parameter analysis of helix-comb seal
flow within the seal gap.
In Fig. 15(a) and Fig. 15(b), the red dotted lines are parallel to The flow field inside HCS is complex due to the existence of
the rotation axis. For the LS, the angles between dotted lines and helical teeth. Helical teeth determine the flow pattern and seal
streamlines are similar in each cavity. This means that the circum- performance of HCS. Before going into details of performance com-
ferential displacements of streamlines within the LS are similar parisons with the other seals, it is needed for HCS to study the de-
in each cavity. For the HCS, the angle between dotted line and pendency of seal performance on the structure parameters. Three
streamlines decreases along the axial flow direction. The circum- major parameters are investigated including helical tooth number,
ferential displacements of streamlines in the latter cavities are helical tooth height, and helix angle.
smaller than that in the first cavity. This phenomenon demon-
strates the effectiveness of negative spiral flow in limiting the cir- 4.2.1. Helical tooth number
cumferential flow within the seal gap. To quantitatively study the In each annular cavity, three values of helical tooth number are
development of circumferential velocity from the seal inlet to the considered, corresponding to 0, 4, and 8 helical teeth. Except for
outlet, Fig. 16 shows the average circumferential velocity distribu- the helical tooth number, the helical tooth height is 3.45 mm and
M. Zhang, J. Yang / Aerospace Science and Technology 105 (2020) 105997 7
Fig. 11. Comparison of the calculated static pressure distribution with the measured.
Fig. 9. Comparison of the calculated leakage flow rate with the measured.
Fig. 12. Contours of velocity. (For interpretation of the colors in the figure(s), the Fig. 14. Contours of static pressure.
reader is referred to the web version of this article.)
Table 4
Variation in leakage flow rate with helical tooth height.
Fig. 16. Average circumferential velocity along the axial direction. Fig. 17. Variation in tangential force with helical tooth number ( P = 0.3 MPa, ω=
6000 rpm, λ = 0).
helix angle increases from 5 deg to 15 deg, the tangential force de-
creases, but as the helix angle continues to increase from 15 deg 4.3.1. Leakage
to 25 deg, the tangential force begins to increase. 15 deg is the op- Fig. 21 compares the leakage flow rates for the three seals with
timum helix angle for the HCS to gain the lowest tangential force. respect to the pressure difference across the seal. As the pressure
difference increases, the leakage flow rate shows a tendency to in-
4.3. Performance comparisons of three types of seals crease. The maximum difference of leakage flow rate between the
HCS and LS is less than 1.35%. The maximum difference of leakage
After the parameter analysis of HCS, the HCS with helical tooth flow rate among the three seals is less than 2.5%. These three seals
number (each cavity) of 8, helical tooth height of 3.45 mm, and have the similar leakage flow rates.
helix angle of 15 deg is selected to conduct performance compar- The flow through the annular seal can be considered as a series
isons with the conventional LS and LS with swirl brakes. of throttling process. The seal tooth number is the key parameter
10 M. Zhang, J. Yang / Aerospace Science and Technology 105 (2020) 105997
Fig. 18. Variation in tangential force with helical tooth height ( P = 0.3 MPa, ω=
6000 rpm, λ = 0).
Fig. 23. Variation in tangential force with pressure difference (ω = 6000 rpm, λ = 1).
Fig. 20. Variation in tangential force with helix angle ( P = 0.3 MPa, ω = 6000
rpm, λ = 0).
Fig. 21. Variation in leakage flow rate with pressure difference (ω = 6000 rpm, λ = Fig. 24. Variation in tangential force with preswirl ratio (ω = 6000 rpm, P = 0.3
1). MPa).
M. Zhang, J. Yang / Aerospace Science and Technology 105 (2020) 105997 11
Fig. 25. Circumferential pressure distribution in the middle of the second cavity (ω = 6000 rpm, P = 0.3 MPa, λ = 1).
in determining the leakage flow rate. For the HCS, the helical teeth force of HCS is desirable because it can stabilize the rotor system.
do not play a role in throttling process. The effective tooth number The HCS has the excellent ability to strengthen the rotor stability.
of HCS is the same as that of other two seals. Thus, the three seals The tangential force originates from the non-uniform circumfer-
have a comparable sealing performance. ential pressure distribution in the seal cavity. In order to study the
pressure distribution pattern on the rotor surface, take the second
cavity as an example, Fig. 25 shows the circumferential pressure
4.3.2. Tangential force
distribution on the rotor surface in the middle of the second cav-
Tangential force is the cause of rotor instability. As shown in
ity for the three seals. In this figure, θ stands for the angular
Fig. 22, tangential force is orthogonal to the displacement. When
coordinate in the direction of rotor spinning. The position θ = 0
the tangential force is in the same direction as rotor velocity, the
deg corresponds to the minimum clearance and θ = 180 deg cor-
tangential force is a positive value. The positive tangential force
responds to the maximum clearance, as shown in Fig. 4. For the
tends to destabilize the rotor because it adds energy to the for-
LS, Fig. 25(a) indicates that the maximum pressure locates at 226
ward whirl motion of the rotor. Conversely the tangential force
deg and minimum pressure locates at 61 deg. The maximum pres-
is a negative value when the force and the rotor velocity are in
sure is behind the minimum clearance within the LS. For the LS
opposite direction. The negative tangential force tends to stabilize
with swirl brakes, Fig. 25(b) indicates that the maximum pres-
the rotor because it acts to remove energy from the forward whirl sure locates at 30 deg and the minimum pressure locates at 239
motion of the rotor. The tangential force should have a low or neg- deg. The frequent pressure fluctuation is due to the existence of
ative value to keep the seal system stable. Fig. 23 compares the swirl brakes. At seal entrance, the swirl brakes divide the annular
tangential forces for the three seals with respect to the pressure channel into multiple vane-to-vane passages. In each vane-to-vane
difference across the seal. With the increase of pressure difference, passage, the flow field is complex. This results in the significant
the absolute value of tangential force increases. The HCS has the difference in velocity profile at different circumferential positions
lowest tangential force. The tangential force of HCS decreases by in each vane-to-vane passage. It would influence the downstream
236–620% relative to the conventional LS. The application of swirl flow field. Each pressure fluctuation corresponds to one vane-to-
brakes can decrease the tangential force by 64–113% compared vane passage. Fig. 25(c) indicates that the pressure of HCS takes
with the LS. The difference between the HCS and LS is about 2–10 its maximum value at 68 deg and minimum value at 287 deg. The
times as strong as that between the LS with swirl brakes and LS. rapid pressure fluctuation is due to the existence of helical teeth.
The HCS is better to reduce the tangential force compared with The pressure increase is due to the blocking effect of helical tooth.
swirl brakes. Note that the differences in tangential force between After the helical tooth, the pressure decreases. Each pressure fluc-
the HCS and other seals increase with the pressure difference. The tuation corresponds to one helical tooth.
HCS performs more efficient in reducing tangential force with high In order to study the development of circumferential pressure
pressure difference. distribution along the axial direction, Fig. 26 shows the contours
Fig. 24 shows the variation in tangential force with preswirl ra- of pressure on the cross section through the middle of three cavi-
tio. Increasing preswirl ratio increases the tangential force. At high ties for the three seals. For the LS, the positions with maximum
preswirl, the tangential force of HCS is negative while the tangen- pressure in three cavities locate at the left side. The tangential
tial forces of the other seals are positive. The negative tangential force obtained by integrating pressure circumferentially is positive.
12 M. Zhang, J. Yang / Aerospace Science and Technology 105 (2020) 105997
It tends to destabilize the rotor by adding energy to the rotor mo- tangential force. This leads to the significant negative tangential
tion. For the LS with swirl brakes, the high pressure zone locates force for the HCS.
at the left side in the first cavity, which is similar to the LS. The
effect of this pressure distribution is to generate a positive tangen- 5. Conclusions
tial force. In the subsequent two cavities, the high pressure zones
locate at the right side, which is contrary to the LS. The effect of In this article, a novel HCS is proposed to generate the negative
this pressure distribution is to generate a negative tangential force. circumferential spiral flow and reduce the destabilizing tangential
force. Its effectiveness in reducing tangential force and leakage is
Hence, the tangential force of LS with swirl brakes is lower than
studied by the comparison with the LS and LS with swirl brakes.
that of LS. For the HCS, the zone with high pressure locates at the
Conclusions are summarized as follows:
top side in the first cavity. Its contribution to the positive tangen-
tial force is small. In the subsequent two cavities, the high pressure (1) The three seals have a comparable sealing performance. The
zones locate at the right side. Its effect is to generate the negative maximum difference of leakage flow rate between the HCS and
M. Zhang, J. Yang / Aerospace Science and Technology 105 (2020) 105997 13
LS is less than 1.35%. The maximum difference of leakage flow [10] J.S. Alford, Protecting turbomachinery from self-excited rotor whirl, J. Eng.
rate among the three seals is less than 2.5%. Power 87 (4) (1965) 333–343.
[11] H. Benckert, J. Wachter, Flow induced spring coefficients of labyrinth seals for
(2) For the HCS, the negative circumferential spiral flow can be
application in rotor dynamics, NASA CP-2133, 1980.
generated with the guiding role of helical teeth. It can weaken [12] D. Sun, S. Wang, Y. Ai, H. Zhou, K. Wang, Experimental investigation of rotor-
the circumferential flow within the seal gap substantially. At dynamic coefficients for the labyrinth seals with and without shunt injection,
seal exit, the ratio of average circumferential velocity to the J. Vibroeng. 17 (8) (2015) 4289–4300.
rotor surface velocity is less than 0.005. [13] E.A. Soto, D.W. Childs, Experimental rotordynamic coefficient results for: (a) a
labyrinth seal with and without shunt injection and (b) a honeycomb seal, J.
(3) For the HCS, the tangential force decreases with the increas- Eng. Gas Turbines Power 121 (1) (1999) 153–159.
ing helical tooth number and helical tooth height. Considering [14] N. Kim, S.Y. Park, D.L. Rhode, Predicted effects of shunt injection on the ro-
the rubbing between the helical teeth and rotor, helical tooth tordynamics of gas labyrinth seals, J. Eng. Gas Turbines Power 125 (1) (2003)
height should be slightly smaller than comb tooth height. 15 167–174.
[15] J. Li, P.D. Choudhury, F. Kushner, Evaluation of centrifugal compressor stabil-
deg is the preferable value for the helix angle to gain the low-
ity margin and investigation of antiswirl mechanism, in: Proceedings of the
est tangential force. Thirsty-Second Turbomachinery Symposium, 2003, pp. 49–58.
(4) The HCS has the superior tangential force reduction capacity [16] A. Untaroiu, H. Jin, G. Fu, V. Hayrapetiau, K. Elebiary, The effects of fluid
compared with the other two seals. The tangential force of preswirl and swirl brakes design on the performance of labyrinth seals, J. Eng.
HCS decreases by 236–620% relative to the conventional LS. Gas Turbines Power 140 (8) (2018) 082503.
[17] K.K. Nielsen, R.A.V. Braembussche, C.M. Myllerup, Optimization of swirl brakes
The difference between the HCS and LS is about 2–10 times as by means of a 3D Navier-Stokes solver, ASME Paper No. 98-GT-328, 1998.
strong as that between the LS with swirl brakes and LS. [18] D.W. Childs, J.E. Mclean, M. Zhang, S.P. Arthur, Rotordynamic performance of a
negative-swirl brake for a tooth-on-stator labyrinth seal, J. Eng. Gas Turbines
Declaration of competing interest Power 138 (6) (2016) 062505.
[19] K. Nagai, S. Kaneko, H. Taura, Y. Watanabe, Numerical and experimental anal-
yses of dynamic characteristics for liquid annular seals with helical grooves in
The authors declare that they have no known competing finan- seal stator, J. Tribol. 140 (5) (2018) 052201.
cial interests or personal relationships that could have appeared to [20] K. Nagai, S. Kaneko, H. Taura, Y. Watanabe, Numerical and experimental analy-
influence the work reported in this paper. ses of static characteristics for liquid annular seals with helical grooves in seal
stator, J. Tribol. 140 (3) (2018) 032201.
[21] W.J. Anderson, L.P. Ludwig, Bearing and seal technology, NASA N69-12582
Acknowledgements
1968.
[22] D.W. Childs, A.J. Gansle, Experimental leakage and rotordynamic results for he-
This study was supported by the National Natural Science Foun- lically grooved annular gas seals, J. Eng. Gas Turbines Power 118 (2) (1996)
dation of China (Nos. 51575105). 389–393.
[23] M. Zhang, J. Yang, W. Xu, Y. Xia, Leakage and rotordynamic performance of a
mixed labyrinth seal compared with that of a staggered labyrinth seal, J. Mech.
References Sci. Technol. 31 (5) (2017) 2261–2277.
[24] Z. Li, J. Li, Z. Feng, Numerical investigations on the leakage and rotordynamic
[1] S. Kim, K. Kim, C. Son, Three-dimensional unsteady simulation of a multistage characteristics of pocket damper seals – part II: effects of partition wall type,
axial compressor with labyrinth seals and its effects on overall performance partition wall number, and cavity depth, J. Eng. Gas Turbines Power 137 (3)
and flow characteristics, Aerosp. Sci. Technol. 86 (2019) 683–693. (2015) 032504.
[2] D. Sun, N. Liu, C. Fei, G. Hu, Y. Ai, Y. Choy, Theoretical and numerical inves- [25] Z. Li, J. Li, Z. Feng, Numerical investigations on the leakage and rotordynamic
tigation on the leakage characteristics of brush seals based on fluid-structure characteristics of pocket damper seals – part I: effects of pressure ratio, ro-
interaction, Aerosp. Sci. Technol. 58 (2016) 207–216. tational speed, and inlet preswirl, J. Eng. Gas Turbines Power 137 (3) (2015)
[3] M. Zhu, J. Teng, X. Qiang, J. Feng, J. Feng, Impact of hub gap leakage on stator 032503.
endwall flow in an axial compressor stage with casing treatment, Aerosp. Sci. [26] X. Kong, G. Liu, Y. Liu, L. Zheng, Experimental testing for the influences of rota-
Technol. 94 (2019) 105399. tion and tip clearance on the labyrinth seal in a compressor stator well, Aerosp.
[4] E. Zhang, Y. Jiao, Z. Chen, Effect of radial growth on rotordynamic characteris- Sci. Technol. 71 (2017) 556–567.
tics of labyrinth seal-rotor system, J. Aerosp. Eng. 32 (4) (2019) 04019043. [27] R. Gao, G. Kirk, CFD study on stepped and drum balance labyrinth seal, Tribol.
[5] X. Jia, Q. Zheng, Y. Jiang, H. Zhang, Leakage and rotordynamic performance of Trans. 56 (4) (2013) 663–671.
T type labyrinth seal, Aerosp. Sci. Technol. 88 (2019) 22–31. [28] K. Kwanka, J. Sobotzik, R. Nordmann, Dynamic coefficients of labyrinth gas
[6] S.M. Jeon, H.D. Kwak, S.H. Yoon, J. Kim, Rotordynamic analysis of a high thrust seals: a comparison of experimental results and numerical calculations, ASME
liquid rocket engine fuel (kerosene) turbopump, Aerosp. Sci. Technol. 26 (1) Paper No. 2000-GT-0403, 2000.
(2013) 169–175. [29] T. Hirano, T. Sasaki, H. Sakakida, T. Uchida, M. Tsutsui, K. Ikeda, Evaluation of
[7] L.N. Butymova, V.Y. Modorskii, Numerical modeling of the labyrinth seal taking rotordynamic stability of a steam turbine due to labyrinth seal force, in: Inter-
into account vibrations of the gas transmittal unit rotor in aeroelastic formula- national Conference on Power Engineering, 2007, pp. 361–367.
tion, Proc. Eng. 201 (2017) 666–676. [30] Z. Li, J. Li, X. Yan, Z. Feng, Effects of pressure ratio and rotational speed on
[8] S. Li, Q. Xu, X. Zhang, Nonlinear dynamic behaviors of a rotor-labyrinth seal leakage flow and cavity pressure in the staggered labyrinth seal, J. Eng. Gas
system, Nonlinear Dyn. 47 (4) (2007) 321–329. Turbines Power 133 (11) (2011) 114503.
[9] T. Tsukuda, T. Hirano, C. Watson, N.R. Morgan, B.K. Weaver, H.G. Wood, A nu-
merical investigation of the effect of inlet preswirl ratio on rotordynamic char-
acteristics of labyrinth seal, J. Eng. Gas Turbines Power 140 (8) (2018) 082506.