World Academy of Science, Engineering and Technology

International Journal of Aerospace and Mechanical Engineering

Vol:5, No:11, 2011

Location of Vortex Formation Threshold at

Suction Inlets near Ground Planes – Ascending
and Descending Conditions
Wei Hua Ho

• Diameter of the engine inlet or suction inlet

Abstract—Vortices can develop in intakes of turbojet and turbo • Distance of the engine or suction inlet from the solid
fan aero engines during high power operation in the vicinity of solid surface or interface
surfaces. These vortices can cause catastrophic damage to the engine. The first two factors are non-geometric factors and depend
The factors determining the formation of the vortex include both
Open Science Index, Aerospace and Mechanical Engineering Vol:5, No:11, 2011 publications.waset.org/15812/pdf

geometric dimensions as well as flow parameters. It was shown that

strongly on the engine operating parameters and local flow
the threshold at which the vortex forms or disappears is also field respectively. For example, vortices may form at
dependent on the initial flow condition (i.e. whether a vortex forms configurations where previously no vortices were detected
after stabilised non vortex flow or vice-versa). A computational fluid because of a new engine setting or an unusually strong
dynamics study was conducted to determine the difference in component of crosswind on the runway or distortion in the
thresholds between the two conditions. This is the first reported JETC inlet velocity profile. This distortion in the JETC inlet
numerical investigation of the “memory effect”. The numerical
results reproduce the phenomenon reported in previous experimental
velocity profile may be a result of cell geometry, or crosswind
studies and additional factors, which had not been previously studied, at the top of the inlet stack [1].
were investigated. They are the rate at which ambient velocity These vortices concentrate ambient vorticity leading to
changes and the initial value of ambient velocity. The former was single core vortices forming at the engine or suction inlet. The
found to cause a shift in the threshold but not the later. It was also formation and ingestion of such vortices can potentially lead
found that the varying condition thresholds are not symmetrical about to catastrophic damage to the engine. In addition to physical
the neutral threshold. The vortex to no vortex threshold lie slightly
further away from the neutral threshold compared to the no vortex to
damage caused by ingestion of foreign objects, commonly
vortex threshold. The results suggests that experimental investigation termed foreign object damage (F.O.D.), the strong pressure
of vortex formation threshold performed either in vortex to no vortex variations within the vortex core can also cause the
conditions, or vice versa, solely may introduce mis-predictions compressor to stall.
greater than 10%. In the formation of such vortices, there exists a blow-away
velocity upstream of the inlet. Above this threshold, the vortex
Keywords—Jet Engine Test Cell, Unsteady flow, Inlet Vortex core is convected downstream and disconnected from the inlet
(blown-away). Conversely if the upstream air velocity is
I. INTRODUCTION below the blow-away velocity, a vortex may be formed,

V ORTICES can develop in the intakes of aero engines

during high power operation near solid surfaces such as
the runway or the walls of Jet Engine Test Cells (JETC). This
subjected to other conditions being favourable.
Equally important is the distance between the engine (or
suction inlet) and the solid surface. If this distance is too large,
phenomenon can occur during take-off, ground run of engines no stagnation point (a point with a diverging radial velocity
or during engine tests in a JETC. The structure of the vortex is profile) will form on the surface and the vortex cannot form.
very similar to vortices seen in draining basins or bath tubs. In other words, the capture stream-tube does not intersect with
The fluid streamlines spiral towards and into the suction inlet any solid surface.
with decreasing radius of gyration. The other end of the vortex The velocity condition is expressed as the ratio between the
is anchored to a nearby solid surface or an interface, such as suction inlet velocity (Vi) and the ambient velocity (Vo),
the case of draining basins where the anchor is at the liquid-air Vi/Vo. The distance condition is expressed as the ratio
interface. between the distance between the engine centreline and the
Various factors determine whether a vortex will form. These solid plane (H) to engine diameter (Di), H/Di. Sometimes D is
include the following: used instead of Di in studies. Fig shows the principal
• Thrust of the engine or suction power in the case of a parameters as described above.
generic suction inlet
• Amount of ambient vorticity

W.H. HO is with the University of South Africa, Johannesburg, Private

Bag X6 Florida 1710, RSA (phone: (27) 011 471 3293; fax (27) 011 471
2142; email: howh@unisa.ac.za).

International Scholarly and Scientific Research & Innovation 5(11) 2011 2370 ISNI:0000000091950263
 World Academy of Science, Engineering and Technology
International Journal of Aerospace and Mechanical Engineering
Vol:5, No:11, 2011

The computational results agree very well with previous

experimental results qualitatively. The quantitative difference
in the threshold could be due to various geometric and flow
conditions as reported by Ho and Jermy [6,7] or to the
memory effect that Ridder and Samuelsson [13] reported. It is
also unclear whether the experimental results had the same
level of sensitivity as the simulations. Some of the
experiments did not use actual visualisation of a vortex as a
means of detecting a vortex but used other manifestations such
as the lifting of beads instead.
To date, there have been no reports of any computational
study on the memory effect.
Fig. 1 Sketch showing the principal parameters Note that the computational studies of Ho and Jermy always
initialised the flow field before solving for a suction inlet –
At intermediate values, both conditions are dependent on ambient velocity ratio, in an intentional manner to eliminate
each other, i.e. a H/Di value that is too high at low Vi/V0 may the memory effect [6]. This is not the same as moving from
not be, at higher Vi/V0 values.
Open Science Index, Aerospace and Mechanical Engineering Vol:5, No:11, 2011 publications.waset.org/15812/pdf

conditions without a vortex to one with a vortex. In the

A number of studies [2,3,4,5] have been reported the use of initialised CFD simulations, a stable no vortex flow did not
computational fluid dynamics to investigate various aspects of exist before the formation of the vortex.
the phenomenon such as the unsteady vortex behaviour and
the ingestion of particles. II. MEMORY EFFECT OF VORTEX FORMATION PHENOMENON
Besides understanding the physical characteristics of the
Ridder and Samuelsson [13] reported that “while
vortex, understanding the conditions which encourages the
determining the blow-away height of an inlet vortex a
formation of the vortex is also very important. Both
pronounced lag in the vertical distance was found between an
computational methods [6,7,8,9] and experimental methods
ascending inlet with the vortex just having collapsed and a
[10,11,12] have been utilised to investigate the conditions
descending inlet with the vortex just being established”. Note
permitting the formation of vortices Fig and Fig shows the
that Ridder and Samuelsson’s “ascending” and “descending”
computational and experimental results.
describes physical movement of the inlet. This is opposite to
the operating condition on the Vi/Vo vs H/D graph. The
ascending inlet corresponds to moving downwards on the
graph where a prevailing vortex collapses and vice versa. The
second definition will be used from this point forward, and the
term ascending threshold would indicate moving upwards on
the Vi/Vo vs H/D graph and vice versa.
This indicates that there are potentially other “sub-
thresholds” lying on either side of the computational
thresholds. These sub-thresholds are a direct consequence of
the inertial and viscous forces present. The distance between
the “neutral threshold” in previous computational studies and
these sub-thresholds is hypothesized to be affected by the
Fig. 2 Vortex formation threshold from computational studies [6] following factors:
1. Ascending or descending scenarios
2. Rate at which Vi (or V0) changes
3. The magnitude difference between the “starting” and
threshold conditions, i.e. whether the starting conditions
lie closer or further from the neutral threshold.
4. Flow and geometry conditions such as velocity gradient
and engine diameter.
Understanding of the memory effect together with
qualitative investigations and quantitative measurement of the
two sub-thresholds is critical to the accurate prediction of the
conditions that results in a vortex forming. It is also a critical
step to bridge the gap between future computational and
Fig. 3 Vortex formation threshold from experimental studies [6] experimental studies in this area. Unlike computational studies
where each set of calculation can be initialised from a
particular value easily, experimental studies will almost
always involve some adjustments to the experimental

International Scholarly and Scientific Research & Innovation 5(11) 2011 2371 ISNI:0000000091950263
 World Academy of Science, Engineering and Technology
International Journal of Aerospace and Mechanical Engineering
Vol:5, No:11, 2011

parameters (either intentional or unintentional). These • ceiling – ≈7 x suction inlet diameter

adjustments will inevitably lead to some shift in the “observed This was found after a series of tests were conducted to
threshold” due to the memory effect. determine the optimum size.Fig below shows a pictorial view
Another important reason for the understanding of the of the model. The calculations were run on only one geometric
memory effect lies in conditions where the ambient (take-off) model with a suction inlet diameter (D) of 1m and an H/D
or JETC inlet flow have high turbulent intensity. Ho and value of 2.0.
Jermy [14] reported that an increase in turbulent intensity
lowers the threshold thus increasing the probability of a vortex
forming. It was hypothesized that this may be due to non-
symmetry of the location of the two sub-thresholds from the
neutral threshold. If either the ascending or descending
threshold is significantly further from the neutral threshold
compared to the other then a change in turbulent intensity may
shift the “threshold” in either direction. It is not clear how
long the “instantaneous” conditions need to persist for a vortex
to form or dissipate.
Open Science Index, Aerospace and Mechanical Engineering Vol:5, No:11, 2011 publications.waset.org/15812/pdf

III. SCOPE OF INVESTIGATION

The vortex investigated in previous computational and
experimental investigations referenced in this document as
well as the type that is investigated in this paper deals with
single core vortices that result from a concentration of ambient
vorticity. This is opposed to the type which does not require
ambient vorticity and results in two counter-rotating vortices
described by de Siervi et al. [15] and Murphy and MacManus
[16]. The same counter-rotating vortices were also observed
when the upstream velocity gradient is less than 0.001/s [6]. Fig. 4 Flow Model
At the time of writing, only the first three factors that were
hypothesized to have an effect on the memory effect are The solver used was ANSYS Fluent 13.0 with the following
investigated and they are: parameters:
1. Ascending or descending scenarios • Mesh density – 104087 cells
2. Rate at which Vo (or Vi) changes • Discretisation scheme – first order discretisation scheme
3. Different starting conditions • Turbulence model – SST-Kω
Only the take-off or engine ground run scenarios are being • Incompressible flow
investigated and no JETC scenarios were completed. • Initial conditions – Solution was initialised at cell inlet
plane
IV. METHODOLOGY The following boundary conditions (as they are named in
3D CFD simulations of a cylindrical inlet of negligible ANSYS Fluent) were used during the simulations.
thickness over a ground place were used in this study. The
model is the same model used in previous simulations by
Jermy and Ho [6]. The flow region was split into two regions
of differing mesh densities to reduce mesh size and hence
computation resources in regions far away from the suction
inlet. Higher mesh density was used in the region near the
suction inlet and where the vortex forms, under the
appropriate conditions. A boundary layer was modelled with 4
layers (thickness of 0.1m for the first layer), each layer
increases by a factor of 1.2 from the previous. Standard wall
functions were used. The model underwent a number of
convergence tests as detailed in [6].
The size of ambient space around the suction inlet is as
Fig. 5 Boundary Conditions
follows:
• upstream – 5 x suction inlet diameter The change in conditions was simulated through the
• upstream – 5 x suction inlet diameter increase and decrease of Vo. The model was initialised with a
• downstream – 10 x suction diameter Vo value below or above the neutral threshold value and
• sides – ≈8 x suction inlet diameter gradually increased or decreased at a fixed rate through the

International Scholarly and Scientific Research & Innovation 5(11) 2011 2372 ISNI:0000000091950263
 World Academy of Science, Engineering and Technology
International Journal of Aerospace and Mechanical Engineering
Vol:5, No:11, 2011

use of a UDF function. The neutral threshold value was taken

from [6]. Three rates of change (0.1/s, 0.05/s and 0.01/s) of V0
were solved. The upstream velocity gradient was 0.2/s and two
initial average V0 values corresponding to +/- 1m/s and +/-
0.5m/s were solved (depending on whether it was from a
scenario with no vortex to one with a vortex and vice versa.
It was anticipated that it would take a period of time for a
vortex to form in conditions favouring its formation. This
would be dependent on the geometry of the suction inlet. Thus
it was necessary to conduct an initial experiment to find out
the time needed for a vortex to form in constant favourable
conditions. For descending conditions, it will be necessary to
have an initial time (say X secs) at constant Vo in the
beginning of the simulation to allow the vortex to form before
increasing Vo. This initial condition may or may not be
necessary for ascending conditions depending on the total time
Open Science Index, Aerospace and Mechanical Engineering Vol:5, No:11, 2011 publications.waset.org/15812/pdf

before a vortex forms in each case. If the total time it takes for
a vortex to form in each simulation (without the initial time) is
much longer than the time it takes a vortex to form in constant
favourable conditions, then it is not necessary to implement
the constant Vo condition.
A vortex was deemed to have been formed when vector
plots on the ground show clear circulatory flow. Fig illustrates
this.

Fig. 6 Velocity plots of a vortex

V. RESULTS AND DISCUSSION

All time steps were solved to 3000 iterations and the
eventual residuals were less than 5 x 10-3 with a large majority
less than 5 x 10-4.
A. Duration for a vortex to form under favourable
conditions
In order to accurately simulate a descending vortex (where
a vortex dissipates after being formed), it is necessary to know
the amount of time needed for a vortex to form under
favourable conditions.
This will be dependent on the size of the suction inlet and
the distance from the solid surface. Thus the current set of Fig. 7 Velocity vector plots showing the formation of a vortex (1-4
secs)
results is only applicable to an engine size of around 1m with
a H/D ratio of 2.0.

International Scholarly and Scientific Research & Innovation 5(11) 2011 2373 ISNI:0000000091950263
 World Academy of Science, Engineering and Technology
International Journal of Aerospace and Mechanical Engineering
Vol:5, No:11, 2011
Open Science Index, Aerospace and Mechanical Engineering Vol:5, No:11, 2011 publications.waset.org/15812/pdf

Fig. 8 Velocity vector plots showing the formation of a vortex (5-6

secs)

Fig shows the formation of a vortex under favourable

conditions at time-steps of 1 sec interval. The upstream
velocity was kept constant at velocity gradient of 0.2/s with an
average velocity difference of -1m/s from the neutral threshold
value.
Thus for simulations of descending conditions, it is
necessary to include a buffer time of 5 secs for the vortex to
form before commencing the increase in Vo. For ascending
conditions, it was found that a vortex took much longer than 5
secs to form in all simulations. Thus it is not necessary to
include this buffer time.
B. Dissipating of a vortex
As mentioned earlier in this paper, as the blow-away
velocity increases in the presence of a vortex, the vortex core
gets convected downstream and eventually gets blown away.
Fig illustrates this process.

Fig. 9 Vortex being blown away

In Fig, the flow direction is from right to left in the positive

X-direction. The white rectangle is the suction tube with the
right-most edge being the inlet. The pictures from top to
bottom show the effect of increasing upstream velocity Vo.

International Scholarly and Scientific Research & Innovation 5(11) 2011 2374 ISNI:0000000091950263
 World Academy of Science, Engineering and Technology
International Journal of Aerospace and Mechanical Engineering
Vol:5, No:11, 2011

The dissipation of the vortex is shown clearly where from TABLE IV

VI/VO PREDICTION SHIFT PERCENTAGE (FROM NEUTRAL)
an almost perfect circular vortex core (top picture) shifts
downstream showing signs of elongation when Vo is Ascending
Neutral
increased (2nd and 3rd picture). Eventually the core is not Rate 0.01/s 0.05/s 0.1/s
longer able to be sustained (4th picture). Vi/Vo
0 4.4 12.1 21.95
C. Ascending and. Descending Threshold shift (%)
Descending
CFD results agree with previous findings by Ridder and Neutral
Samuelsson [13]. If a vortex has formed, it requires a higher Rate 0.01/s 0.05/s 0.1/s
blow-away velocity (compared to the neutral case) to dissipate Vi/Vo
0 11.9 20.9 24.1
the vortex and vice versa. The results are presented in TABLE . shift (%)

TABLE I The results indicate that any experiments conducted with

ASCENDING AND DESCENDING THRESHOLD
the aim of getting the vortex formation threshold should be
Neutral Ascending Descending
conducted in both an ascending as well as descending manner.
Vi (m/s) 151.18 151.48 151.97 So far only Ridder and Samuelsson have indicated clearly that
it was performed [13].
Open Science Index, Aerospace and Mechanical Engineering Vol:5, No:11, 2011 publications.waset.org/15812/pdf

Vo (m/s) 3.75 3.60 4.40

The difference between the ascending and descending
Vi/Vo 40.32 42.08 34.54 thresholds (at the same change rates) indicates that increased
The calculated values of Vi/Vo have uncertainty of +/- turbulent intensity in flows maybe lower the threshold [14]. A
1.4% and +/- 1.2% for ascending and descending respectively. highly turbulent flow, with mean conditions at the no vortex
D. Different rate of change of Vo regime, may cross the ascending threshold, thus going into the
vortex regime, but not the descending threshold. This may
Different rate of change of Vo affects the position of both
have the effect of lowering the formation threshold. Although
the ascending as well as descending threshold.
it takes between 4 – 5 secs for a vortex to form in constant
TABLE II favourable conditions, it is not clear what happens in
ASCENDING THRESHOLD FOR DIFFERENT VO CHANGE RATES fluctuating flows which encompasses the ascending threshold
Ascending and with instantaneous conditions constantly crossing one of
Neutral the sub-thresholds.
Rate 0.01/s 0.05/s 0.1/s
Fig. 10 below illustrates a scenario where the fluctuating
Vi (m/s) 151.18 151.48 151.35 151.19 conditions encompass the ascending threshold. The relative
Vo (m/s) 3.75 3.60 3.35 3.08 locations of the three thresholds are not drawn according to
scale with the double-headed arrow indicates the maximum
Vi/Vo 40.32 42.08 45.18 49.17 range of fluctuation in conditions.
Uncertainty
1.38 4.43 5.63
(+/- %)
TABLE III
DESCENDING THRESHOLD FOR DIFFERENT VO CHANGE RATES
Descending
Neutral
Rate 0.01/s 0.05/s 0.1/s
Vi (m/s) 151.18 151.97 152.21 152.29
Vo (m/s) 3.75 4.40 4.78 4.98
Vi/Vo 40.32 35.54 31.88 30.61
Uncertainty
1.14 1.55 2.48
(+/- %) Fig. 10 Highly turbulent intensity flow
The results show that a quicker ascend or descend rate
“shifts” the threshold further away from the neutral threshold.
The results also suggest that descending shifts the threshold
further from the neutral than the ascending. However, more
calculations need to be performed before a more definite
conclusion can be drawn on this.
Increasing Vo at a rate of 0.01s, from a regime with a
vortex, may result in a mis-prediction of threshold conditions
of up to 4.4%. Other conditions may lead to larger mis-
predictions. Table gives the other values for the other
calculation scenarios.

International Scholarly and Scientific Research & Innovation 5(11) 2011 2375 ISNI:0000000091950263
 World Academy of Science, Engineering and Technology
International Journal of Aerospace and Mechanical Engineering
Vol:5, No:11, 2011

E. Starting at different values of Vo Lastly, different starting velocities (as long as they are in
the same regime i.e. vortex or no vortex) do not seem to have
TABLE V
DIFFERENT STARTING VALUES OF VO (ASCENDING)
an effect on the threshold. However it is not certain if there
will be any effects when the starting velocities are very close
Ascending
to the neutral threshold values.
Neutral 0.01/s 0.05/s 0.01/s 0.05/s At the current geometry, it was found that it took around 4-
Starting 5 secs for a vortex to form.
4.75 4.25
aver. Vo The difference between the ascending and descending
Vi (m/s) 151.18 151.47 151.31 151.47 151.33 thresholds (at the same change rates) indicates that increased
Vo (m/s) 3.75 3.60 3.30 3.60 3.35 turbulent intensity in flows maybe lower the threshold [14]. A
highly turbulent flow, with mean conditions at the no vortex
Vi/Vo 40.31 42.07 45.85 42.08 45.17 region, may cross the ascending threshold, thus going into the
Uncertainty vortex region, but not the descending threshold. Although it
1.38 4.43 1.38 4.43
(+/- %) takes between 4 – 5 secs for a vortex to form in constant
favourable conditions, it is not clear what happens in
TABLE VI
DIFFERENT STARTING VALUES OF VO (DESCENDING) fluctuating flows which encompasses the ascending threshold
Open Science Index, Aerospace and Mechanical Engineering Vol:5, No:11, 2011 publications.waset.org/15812/pdf

Descending and with instantaneous conditions repeatedly crossing only

one of the sub-thresholds.
Neutral 0.01/s 0.05/s 0.01/s 0.05/s These results confirm that a “memory effect” exists for such
Starting flow situations. Thus, when conducting experiments, it is
2.75 3.25
aver. Vo critical to perform them in both ascending and descending
Vi (m/s) 151.18 151.97 152.22 151.98 152.12 conditions. Depending on the relative geometries of
Vo (m/s) 3.75 4.40 4.80 4.40 4.65 experimental set-ups as well as other flow parameters, it is
possible that mis-predictions of more than 10% may be
Vi/Vo 40.31 34.54 31.71 34.54 32.71 present if this is not performed. It is also recommended to
Uncertainty change the conditions as gradual as possible to reduce the mis-
1.14 1.55 1.14 1.59
(+/- %) predictions.

The results do not show any indication that commencing RECOMMENDATIONS FOR FUTURE WORK
with different average values of Vo have any significant effect
As a continuation of the current studies, it is recommended
on the thresholds. Both starting values are still a good
to conduct CFD simulations and experimental investigations
difference from the neutral value, it is not clear if there will be
on geometries relating to an enclosed test cell using similar
any effect if the starting value is very close to the neutral
methodologies. It will also be interesting to conduct
threshold value. However it is unlikely that the starting Vo
investigations on runway and test cell models using other
will be close to the neutral value very often in experiments.
methods of changing flow conditions such as changing Vi or
H with Vi having more foreseeable practical implications.
VI. CONCLUSIONS
Of particular interest will be the effect of fluctuating or high
The CFD-based methods used in this study successfully turbulent intensity upstream flows with conditions
reproduced the ascending and descending thresholds reported encompassing one of the sub-thresholds, as discussed in
in previous experimental studies. The trends observed in the section V.D on page 6.
numerical results are in qualitative agreement with the
experimental data. When a vortex has formed (descending), it ACKNOWLEDGEMENT
requires a higher blow-away velocity compared to the neutral
The author wishes to thank the Research Committee for the
value to dissipate the vortex. Conversely when a stable no
College of Science Engineering and Technology at the
vortex flow has been achieved (ascending), it required to
University of South Africa for providing financial support to
lower the blow-away velocity below the neutral threshold
have this work presented at this conference. Appreciation also
values for a vortex to form. This of course indicates that the
goes to Dr. Mark Jermy for his valuable comments on the
ascending threshold has higher Vi/Vo values compared to the
manuscript.
descending threshold.
Further, it was observed that different rates of change of Vo
can affect the threshold value both in ascending and [1] W.H. Ho, H. Dumbleton and M. Jermy, “Effect of Upstream Velocity
descending conditions. A quicker ascend (or descend) shifts Gradient on the Formation of Sink Vortices in a Jet Engine Test Cell”,
the threshold by a larger amount. At the same rate of change, IMECS 2008 Proceedings, pp 1767 – 1772, 2008.
[2] A. Karlsson and L. Fuchs, “Time evolution of the vortex between an air
the descending threshold is further from the neutral threshold inlet and the ground”. AIAA paper 2000-0990, 2000.
compared to the ascending threshold. However more [3] A. Secareanu, D. Morioanu, A. Karlsson and L. Fuchs, “Experimental
calculations at different geometries are necessary before this and numerical study of ground vortex interaction in an air-intake, AIAA
paper 2005-1206, 2005.
can be conclusively ascertain.

International Scholarly and Scientific Research & Innovation 5(11) 2011 2376 ISNI:0000000091950263
 World Academy of Science, Engineering and Technology
International Journal of Aerospace and Mechanical Engineering
Vol:5, No:11, 2011

[4] A. Gulia, P. Laskaridis, K.W. Ramsden and P. Pilidis, “A preliminary

investigation of thrust measurement correction in an enclosed engine test
facility, AIAA paper 2005-1128, 2005.
[5] S. Zantopp, D. Macmanus and J. Murphy, “Computational and
experimental study of intake ground vortices”, The Aeronautical Journal
Vol. 114 No. 1162, 2010.
[6] M. Jermy and W.H. Ho, “Location of the vortex formation threshold at
suction inlets near ground planes by computational fluid dynamics
simulation”, Proceedings of the I Mech E Part G Journal of Aerospace
Engineering Vol. 222 No. 3, pp. 393-402, 2008.
[7] W.H. Ho, M. Jermy and H. Dumbleton, “Formation of Sink Vortices in
a Jet Engine Test Cell”, Engineering Letters 16:3, pp. 406-411, 2008.
[8] W.H. Ho and M. Jermy, “Formation and Ingestion of Vortices into Jet
Engines during operation”, IAENG TRANSACTIONS ON
ENGINEERING TECHNOLOGIES VOL. 1 - Special Editions of the
International MultiConference of Engineers and Computer Scientists
2008, pp 132 – 143, 2008.
[9] W.H. Ho and M. Jermy, “Validated CFD simulations of vortex
formation in jet engine test cells”, 16th Australasian Fluid Mechanics
Open Science Index, Aerospace and Mechanical Engineering Vol:5, No:11, 2011 publications.waset.org/15812/pdf

Conference Proceedings, pp. 1102-1107, 2007.

[10] A. Nakayama and J.R. Jones, “Vortex formation in let flow near a wall”,
AIAA paper 96-0803, 1996.
[11] D.E. Glenny, “Ingestion of debris into intakes by vortex action,
Aeronautical Research Council CP No. 1114, 1970.
[12] W. Liu, E.M. Greitzer and C.S. Tan, “Surface static pressure in an inlet
vortex flow field”, ASME J. Engineering Gas Turbines Power 107, pp.
387-393, 1985.
[13] S.O. Ridder and I. Samuelsson, “An experimental study of strength and
existence of domain of ground-to-air inlet vortices by ground board
static pressure measurements, Stockholm Royal Institute of Technology,
KTH AERO TN 62, 1982.
[14] W.H. Ho and M. Jermy, “Effect of Turbulence Intensity on Vortex
Formation Threshold in a Jet Engine Test Cell”, 25th European
Conference on Modelling and Simulation (ECMS) Proceedings, 2011.
[15] F. De Siervi, H.C. Viguier, E.M. Greitzer and C.S. Tan, “Mechanisms of
inlet vortex formation”, J. Fluid Mech. 124, pp. 173-207, 1982.
[16] J.P. Murphy and D.G. MacManus, “Intake Ground Vortex Prediction
Methods”, Journal of Aircraft Vol 48 No. 1, 2011.

International Scholarly and Scientific Research & Innovation 5(11) 2011 2377 ISNI:0000000091950263