This article explores a set of inner and outer brush seals capable of bidirectionally restricting flows in a wave rotor, which has been successfully fabricated and tested. Advantages cited for the wave rotor include enhanced efficiency, rotor material temperatures lower than the peak gas temperatures, lower speed rotation with reduced stress, simple robust construction, and rapid transient response. A cross-sectional view of the rotor shows the cavities and placement of the brush seals. No definitive tests were undertaken during the break in stage or during the testing; thus, one can only conclude that the bristles wore significantly, and the rotor coating showed little evidence of tracking other than being highly polished with some spottiness. Some powder debris was found in the exhaust port. The principles of bidirectional brush design can be applied to all brush configurations, providing bidirectional capability, controlled seal stiffness/damping, bristle spread, flutter, protection from foreign object damage, and a double labyrinth-orifice at essentially the same cost as a conventional brush seal construction.
If the performance goals for advanced gas turbines are to be achieved by conventional means, increasingly difficult thermal demands on materials will have to be met and high overall pressure ratios will be required. Unconventional techniques, such as pulsed combustion and wave rotor topping cycles, have been suggested. However, the achievement of high efficiency through these means will depend on the ability to control leakage losses by means of adequate sealing.
Pulsed combustion can result in combustion with a pressure gain. A wave rotor topping cycle, which can also be thought of as pressure gain combustion, appears to generate higher pressure ratios than pulsed combustion currently does, and so is more attractive. Fuel intercooling and wave rotors could be used in conjunction in the same engine to enhance performance.
A wave rotor machine uses expansion and shock waves within rotating passages on the rotor to accomplish the work typically done by an axial blade/vane or centrifugal component in expanding or compressing the working fluid. In the wave rotor machine both compression and expansion occur, and in some cases chemical reactions take place. Advantages cited for the wave rotor include enhanced efficiency, rotor material temperatures lower than the peak gas temperatures, lower speed rotation with reduced stress, simple robust construction, and rapid transient response.
Sealing can be a major factor in the wave rotor, and the seal is dynamic. The leakage takes place into or out of a passage through the gap between the passage and the end walls of the machine. The leakage flow changes direction, depending on whether the passage is at high or low pressure. These end wall losses can potentially be controlled by seals such as the compliant brush configuration or the close self-activating rim or leaf seals. Because brush seals could be incorporated easily into an existing three-port wave rotor at NASA Lewis Research Center, bidirectional seals were used. Tests were made of wave rotor performance with and without the brush seals. The results of these tests are reported here.
The wave rotor test rig at NASA Lewis has a three-port flow divider. A single inlet flow is separated into two outlet flows, one at a higher stagnation pressure than the inlet and the other at a lower stagnation pressure. The high-pressure air exhausts from the duct in the center. The rotor itself is a cylinder with axially aligned passages on its circumference. As a passage rotates, the pressure at the ends of the passage fluctuates between the high and low pressures. The cavity surrounding the rotor will be at a pressure close to the inlet pressure. Leakage will take place from the passage to the cavity when the passage is at high pressure, and from the cavity to the passage when the latter is at low pressure. Thus, the seals must be capable of withstanding pressure reversals and the corresponding flow reversals.
A cross-sectional view of the rotor (on the facing page) shows the cavities and placement of the brush seals. The rotor passage represents that portion of the rotating cylinder which contains the working fluid; the inner-and outer cavities constitute potential paths for leakage. The movable end wall can be adjusted for a desired setting for the size of the gap between the rotor and the end wall. Without seals, the leakage is proportional to the size of these gaps.
There are 60 passages, each 1.37 centimeters (0.54 inch) wide, spaced about the circumference of the rotor. As the rotor turns, the pressure at each end of the passage varies, with the highest pressure being about three to four times that of the lowest. The cavity pressure is nearly the average of the two. The function of the seals is to prevent leakage from the cavities into the rotor passage and vice versa. The leakage may be thought of as composed of radial and circumferential components. Circumferential leakage is from passage to passage where the pressure differences are not large, so this leakage is believed to be less important than radial leakage. While this circumferential leakage could have been moderated somewhat by swirl brakes, only the brush seals were considered, which had little effect on the circumferential leakage. Radial leakage is from a high-pressure passage into the cavity, and then from the cavity back into a low-pressure passage. By blocking the path from the passages into the cavity, brush seals can reduce the radial leakage.
Brush Seal Requirements
The wave rotor represents an unusual set of operating conditions for a brush seal. In addition to the usual compliance and sealing requirements, the brush must be capable of sealing bidirectional flows and of sealing pressures to ± 276 kilopascals (± 40 lbs. per square inch). The brush pair must seal the interface at both the inside and outside diameters. Both ends of the wave rotor must be sealed; that is, matching pairs of seals are required, and they must be retrofitted into the existing equipment with minimum modifications. The surface speeds, to 152 meters per second (500 feet per second), and environmental temperatures, to 177°C (350°F), are modest design requirements for a brush seal.
After some consideration, it was decided that the shielded design could be modified to provide sealing in both directions, provided a gap was introduced between each side plate and the bristles. The brush was otherwise of standard construction, by the Cross Manufacturing Co. (Devizes, England). The bristles were 0.071-millimeter--(0.0028 inch) diameter Haynes 25 (AMS 5796 28), at angles of 40 to 50 degrees to the interface and inclined in the direction of rotation with suitable antirotation pins. The rotor inner seal radius was nominally 140.2 mm (5.518 inches— 5.5167 left and 5.5192 right) after testing, and the rotor outer seal radius was nominally 162.1 mm (6.380 inches— 6.3747 left and 6.3803 right) after testing. In each case, 0.25 mm (0.010 inch) radial interference was built into the as-manufactured seal.
The rubbing interfaces were Praxair (Union Carbide) LC-1H chromium carbide coated to 0.15 to 0.25 mm (0.006 to 0.010 inch) thickness.
After the wave testing was completed, the outside-diameter and inside-diameter seals were examined carefully.
The brushes were found to be in good condition, with the exception of a tuft pullout in one location of the inside-diameter brush.
The brush seals were installed by rotating them into position (that is, in a direction opposite the rotor’s rotating direction). Suitable static O-rings provided the necessary static seals. The rotor was torqued by hand to set the bristles. In a set of break-in runs, the speed was increased in increments of 650 rpm from 1000 to 7500 rpm at nominal one- hour intervals (for 10 hours) with unheated inlet air. A similar schedule was used with the inlet air heated to 49°C (120°F), for 7.5 hours, with an additional 5 hours of preliminary testing. Once the bristles were set and rubbed into place, a borescope examination of the bristles revealed the characteristic powder debris in flow stagnation regions. The remainder was swept away with the flow. Some bristles strayed beyond the pack; those of the inner seal were most susceptible. Little evidence of unusual bristle dispersion was found, with the exception of an isolated tuft pullout.
The system was operated for a total of 7.5 hours at 7400 rpm. The rotor average temperature was approximately the inlet temperature, 49°C (120°F), with hotgas temperatures to 129°C (264°F) and cold-side temperatures to 11°C (51°F).
A measure of the performance of a three-port wave rotor is the efficiency, defined as h = (b/ (1 - b)).[ (Phi/Pin)(g-1)/g - 1)/(1 - (Plo/Pin)(g - 1)/g)] where Pin is the inlet stagnation pressure, Phi the stagnation pressure in the high-pressure outlet, Plo the stagnation pressure in the low-pressure outlet, and b the ratio of mass flow in the high-pressure outlet to total mass flow. Higher values of both Phi/Pin and Plo/Pin will result in higher efficiency. Reducing leakage will create higher values of system pressure.
The wave rotor efficiency is a function of the size of the gap between the end wall and the rotor for b = 0.37 and Plo/Pin = 0.6. The brush seals were effective in increasing efficiency. The mass leakage is approximately proportional to gap size, and in the experiment the gap size was altered and the associated wave rotor efficiencies measured. To enable a direct comparison, tests were run with and without the brush seals. On average, the brushes diminished the leakage by a factor of two. At a large gap spacing, the tor cycle and against the gap plate during other brush seals have a very pronounced effect. However, at small spacings there is less effect on efficiency and pressure.
The principles of bidirectional brush design can be applied to all brush configurations.
Although the bristles wore significantly and some tufts were disheveled to the point of permitting rivering (a condition in which bristles are pushed aside and leakage flows like a river), both sets of brushes did not appear to deteriorate further with time. But the testing lasted only for 19 hours and 54 minutes—far less than the thousands of hours necessary for an engine application. The nominal operating interferences (radial) were, 0.15 and 0.23 mm (0.006 and 0.009 inch) for the outside-diameter and inside-diameter brush seals, respectively.
The core has a surface resembling a cast-iron fracture. The core is rough, with pits and bumps. There appears to be material transfer. No metallurgical analysis has been performed, as the brush is preserved for further flow testing. Only marginally visible are grooves and ridges scratched across the tips in the direction of rotation. Typically, there will be 15 to 20 visible scars. With a diameter of 0.071 mm (0.0028 inch), each scar is perhaps 4 micrometers (150 millionths of an inch) wide with transfer material scattered over the surface.
Attempts to quantify the surface wear were unsuccessful. From surface measurements we know, however, that the chromium carbide interface coating wore less than 0.025 mm (0.001 inch) on all four interfaces. By optical inspection, however, polishing/burnishing of the interfaces could be readily observed. A closer examination of these tracks reveals “skipping,” or changes in hardness, which may represent tool marks of the parent machining operation or changes in the chromium carbide coating.
The “spottiness” of the wear scar may be related to bristle bunching combined with sharp pressure changes and a fluttering motion during the light-pressure loading portion of the cycle. The dark bands or car tracks on either side of the wear scar could be a result of bristle stiffening near the fence plate during one portion of the ro- brush portions of the rotor cycle. Bristle natural frequency affects bristle movement and fluid flows in the cavities between the bristles and the back plates; further work on defining the back plate and rotor interface geometries is required.
No definitive tests were undertaken during the breakin stage or during the testing; thus, one can only conclude that the bristles wore significantly, and the rotor coating showed little evidence of tracking other than being highly polished with some spottiness. Some powder debris was found in the exhaust port.
A set of inner and outer brush seals capable of bidirectionally restricting flows has been successfully fabricated and tested. The brush seals represented an extension of the side plates with sufficient gap to permit compliance.
The wave rotor efficiency improvement due to the seals was more pronounced for the largest end wall gap (rising from 16 percent without seals to 34 percent with seals), and to a lesser extent for the smallest end wall gaps (rising from 36 percent without seals to 45 percent with seals) where the leakages become quite small. On average, the leakages with brush seals were half those of the gap-controlled baseline configuration.
The optical microscope revealed bristle tip grooving. Material transfer was inferred from the fracture-like appearance of the bristle tips of the outside-diameter brush bristles. Bristle tips for the inside-diameter brush were similar, except that no grooving was evident. The coated rotor appeared polished without distress with the exception of car-track wear scars and bristle whipping, or bristle-bunch incursion loading marks, which may be related to residual tooling or surface hardening in the parent rotor material. Inner and outer wear tracks, like car tracks, may be the result of bristle stiffening at the fence plate during part of the rotor cycle and stiffening at the gap plate during other parts of the rotor cycle. Small pressure perturbations may cause bristle flutter during the lightly loaded portion of the cycle. Bristle natural frequency affects motion and cavity flows between the back plates and requires better definition of back plate geometries.
For the limited test time and operating conditions, the rotor surface appeared polished, the bristles did wear in, and the brush seals provided enhanced efficiencies over the base line.
The principles of bidirectional brush design can be applied to all brush configurations, providing bidirectional capability, controlled seal stiffness/damping, bristle spread, flutter, protection from foreign object damage, and a double labyrinth-orifice at essentially the same cost as a conventional brush seal construction.