SAN ANDRES (2013) ENGINEERING ANALYSES FOR POCKET DAMPER SEALS (YEAR II)

1
Continuation Proposal to the TAMU Turbomachinery Research Consortium
ENGINEERING ANALYSES FOR POCKET DAMPER SEALS AND COMBINED
LABYRINTH-BRUSH SEALS
Dr. Luis San Andrs, Mast-Childs Tribology Professor
Weilian Shan, Research Assistant
May 2013 (YEAR II)

SIGNIFICANCE AND JUSTIFICATION
Seals leakage in centrifugal compressors and turbines represents a substantial loss in efficiency and
power delivery with an increase in specific fuel consumption. Labyrinth seals (LS) are the most common
and inexpensive means of reducing secondary leakage, albeit wearing out with operation and thereby
penalizing performance and even affecting rotordynamic stability. Improperly designed and operated
labyrinth seals can be the source of rotordynamic instabilities.
Pocket damper seals (PDS), adding baffles in (alternating)
circumferential cavities of a LS and engineering the inlet and
exit tip clearances, have demonstrated enormous benefits in
system stability by providing physically large damping
coefficients [1]. These seals, when properly designed, can act as
damper bearings by offering also large stiffness coefficients.
Bulk-flow analyses for pocket damper seals (PDS) are available
[2] with predictions showing moderately good correlation with
test data. A recent seal development, the fully-partitioned PDS
[3] lacks a proper physical modeling. This seal type, FPDS,
has axial baffles covering the whole seal extent and the pockets
are separated by thick small clearance regions (not sharp teeth)
which can amount to 20% of the whole sealing area. In
comparisons [4] with experimental results for a LS and a honeycomb seal, the FPDS showed the largest
destabilizing cross-coupled stiffnesses, strong functions of inlet preswirl and rotor speed. As a result, it
appears a LS with a swirl brake could still be a favorable choice.
Brush seals (BS), although costlier, are common in
specialized applications (aircraft engines). BS may
increase plant efficiency by up to 1/6 of a point and
with as little as 10% of the leakage in a similar size
labyrinth seal. In addition, some steam turbines
presently incorporate a hybrid seal composed of a BS
installed mid-way of a multiple-teeth labyrinth seal.
The hybrid BS-LS [5] affords significant sealing
improvements over conventional labyrinth seals.
Additionally, when retrofit into labyrinth seals with
radial movement, the individual labyrinth seal segments are free to move radially independently of one
another during transients. [5] Retractable gland packing and inter-stage packing seals benefit most from
the change in seal configuration. Predictive tools for evaluating BS leakage are based on empirical
models, while their force coefficients are largely ignored. A computational tool modeling the combined
LS-BS would be a welcome addition into the engineering design process of steam and gas turbines.

STATUS OF WORK 2012-2013
In 2012, the TRC funded a two-year project to develop a new computational tool for accurate
prediction of the force coefficients and leakage of pocket damper seals. Note that commercial FPDSs [4]
have thick walls rather than blades with sharp edges as in conventional LSs and early PDSs [2].
Pocket damper seal and variation [3]
Stator (segmented)
flow
Stator (segmented)
flow
Labyrinth seal
Combined brush seal -
labyrinth seal
Labyrinth seal and combination with a brush seal [4]
Labyrinth seal and combination with a brush seal [5]
SAN ANDRES (2013) ENGINEERING ANALYSES FOR POCKET DAMPER SEALS (YEAR II)

2
Weilian Shan, M.S. graduate student, studied an existing code, PDSEAL

, developed by Li et al. [2]

for prediction of leakage and force coefficients in labyrinth seals and pocket damper seals with sharp
blades, and constructed an EXCEL graphic user interface (GUI) for ready interface with the original
Fortran code. XLPDS

, the GUI for pocket damper seals, is simple to use and obtains predictions for
multiples cases varying the gas supply and exit pressures, rotor speed, and whirl excitation frequency. The
tool will be soon available in the XLTRC
2
software suite.
Technical report TRC-SEAL-0213 [6] reports the progress and delivers predictions of rotordynamic
force coefficients for a test LS and a test FPDS in Ref. [4]. The predicted leakage and dynamic force
coefficients correlate well with the test data for the labyrinth seal [4]. Predicted force coefficients for the
FPDS are in gross error when compared to the experimental coefficients. Hence, the current
computational program shows severe limitations to predict the dynamic force performance of PDSs with
thick walls. The physical model is a one-control volume, turbulent bulk flow model that includes the
effects of circumferential flow velocity within a seal pocket and uses Neumanns leakage equation across
the seal blades. The model ignores the flow resistance along the circumferential direction, badly needed
for PDS with blunt blades of sizable axial thickness. Needless to state that, it is in this region the seal
develops a cross-coupled stiffness as the gas is whirled because of shaft rotation.

PROPOSED WORK 2013-2014 (YEAR II)
To develop computational models for prediction of leakage, drag power loss and force coefficients of
fully partitioned PDSs and combined labyrinth-brush seals. During Year II, the specific tasks are:
Update the physical model for PDS [2] by replacing the empirical leakage formulas with a bulk flow model that
incudes flow conservation and circumferential and axial momenta transport equations in the flow region under a
thick blade and the spinning rotor.
Perform further calibration of the physical model predictions against test leakage and force coefficient, those in
Ref. [4] for example.
Begin extensions of the bulk-flow model to include two-component mixtures (liquid in a gas) as in Ref. [7].
Search and evaluate leakage models for brush seals; empirical models are preferred, perhaps using equivalences to
porous media flow models.
The models will integrate real gas properties, including steam and supercritical CO
2
. The analysis will be limited
to a centered rotor thus giving force coefficients of the form K
YY
=K
XX
and K
XY
=-K
YX

BUDGET FROM TRC FOR 2013-2014 YEAR I
Support for graduate student (20 h/week) x $ 1,950 x 12 months $ 23,400
Fringe benefits (0.6%) and medical insurance ($185/month) $ 2,360
Travel to (US) technical conference $ 1,200
Tuition & fees three semesters ($362 credit hour x 24 ch/year) $ 8,686
Other (Mathcadand portable data storage) $ 220
Total Cost:
$ 35,866
REFERENCES
[1] Childs, D. W., Vance, J . M., 1997, Annular Gas Seals and Rotordynamics of Compressors and Turbines, Proc. 26
th

Turbomachinery Symposium, September, Houston, TX, pp. 201-220.
[2] Li, J ., San Andrs, L., and J . Vance, 1999, "Bulk Flow Analysis of Multiple-Pocket Gas Damper Seals," ASME J ournal of
Gas Turbines and Power, Vol. 121, 2, pp. 355-362.
[3] Ertas, B., Vance, J ., 2007, "Rotordynamic Force Coefficients for a New Damper Seal Design," ASME J . Tribol., 129, pp.
365-374.
[4] Ertas, B.H., Delgado, A., and Vannini, G., 2011, Rotordynamic Force Coefficients for Three Types of Annular Gas Seals
with Inlet Preswirl and High Pressure Differential Ratio, ASME GT2011-45556.
[5] Skinner, D.R., et al., 2001, Combined Brush Seal and Labyrinth Seal Segment for Rotary Machines, US Patent 6,257,586
B1 (see also http://www.turbocare.com/retractable_brush_seals.html .
[6] Shan, W., and San Andrs, L., 2013, Predictions vs. Test Results for Leakage and Rotordynamic Force Coefficients of a
Fully Partitioned Pocket Damper Seal and a Labyrinth Seal Limitations and the Current Computational model, Progress
Report to the Turbomachinery Research Consortium, TRC-SEAL-02-2013.
[7] San Andrs, L., 2012, Rotordynamic Force Coefficients of Bubbly Mixture Annular Pressure Seals, ASME J . Eng. Gas
Turbines Power, 134, p. 022503.