The TTF operates, and was used in these experiments, in two modes. One mode is the traditional shock-tube mode as described by [21]. A second mode, which was first used while the tunnel was at Calspan on a previous research program, and which has become more widely used for aerodynamic work (as opposed to heat-transfer work) is the blowdown mode. Both modes will be described below and the major characteristics of the facility operation will be briefly touched upon, but for more detailed descriptions the reader is referred to review [21].
Figure 3.1shows a scaled layout of the TTF. The major sections are shaded differently. The driver and driven tubes together are 100 ft long. A picture of the facility looking towards the dump tank from the connection of the double-diaphragm cavity and the driver tube is shown in Figure 3.2. The Driver tube is scarlet and the driven tube is gray, the expansion nozzle is yellow and the double diaphragm cavity is white. The corner of the control room can be seen on the left
Figure 3.1 TTF Layout
18 .
Figure 3.2 Picture of TTF Looking Towards Dump Tank
The main idea behind any shock tube or shock tunnel is to use a shock wave as a compression device to create a relatively small pocket of high-temperature, high-pressure gas. The shock wave is created by releasing high-pressure gas of lower molecular weight (mixture of helium and air in this case) into a lower pressure gas (air in this case). To accomplish this, two different parts of a tube, separated by a barrier, are pressurized to different pressures using different gas mixtures. The pressure ratio across these sections (the higher pressure section is called the driver tube, and the lower pressure section the driven tube) sets the shock speed, and this in turn controls the resulting reflected-shock pressure and temperature. Usually the barrier is some sort of diaphragm system. One of the difficult problems associated with shock tube operation is controlling how and when the diaphragm(s) breaks. Sometimes explosives are used, other times a mechanical device is used to puncture the diaphragm (depending upon how thick they are), or the driver is pressurized until the diaphragm breaks. Most shock tubes used in universities are generating relatively low Mach numbers, and are done in tubes of only a few centimeters in diameter, so the diaphragm is a thin membrane. For the TTF, the inner diameter of the tube is 18.5” and pressure differences between the driver and driven tubes are on the order of 1000 psi (or about a quarter of a million lbf) on a diaphragm
separating the two tubes. Breaking the diaphragm poses a significant problem, since the
19
pressure ratio has to be controlled very accurately in order to provide repeatable
reflected-shock pressure and temperature conditions, and because the timing between the diaphragm rupture and the turbine speed is critical.
The main differences between a shock tube and a shock tunnel are how the diaphragm is controlled, and what is done with the shock-processed gas. A shock tunnel (such as the OSU TTF) uses a repeatable, clean mechanism for breaking the diaphragm, and a fast acting valve at the end of the driven tube so that the shock process gas can be allowed to flow into an experimental rig. The system used to break the diaphragms is the double diaphragm cavity, first used at the Cornell Aeronautical Laboratory. The idea behind this device is that the diaphragms are designed to break when a pressure of about 60% of the pressure difference between the driven and the driver is applied. This is accomplished by machine scoring the diaphragms to different depths to control the burst pressure. During the loading cycle the pressure in the double diaphragm cavity is maintained at about 1/2 the pressure difference. When the load conditions (both tubes pressures and rotor speed) are reached, the double diaphragm cavity is vented to atmospheric pressure via a Valcor valve, which very rapidly puts a larger pressure difference across the upstream diaphragm (number 1). This cause the diaphragm 1 to break which them put the entire load across diaphragm 2 (causing it to break). This system provides a relatively robust way of controlling the shock conditions in the facility.
The second major difference between a shock tube and a shock tunnel is how the test-gas is used. In a shock tube, the instruments are usually placed on the endwall and sidewall of the driven tube, and one obtains the high-temperature, high-pressure
condition from the time the original shock reaches the end of the tube until the time of an interface reflection (or the interface passes to the endwalls and destroys the uniform flow conditions). In a shock tunnel (such as the OSU TTF), there is a fast-acting valve placed at the end of the driven tube and the entrance to the expansion nozzle leading to the test section that is designed to open immediately after the initial shock wave compresses the gas at the end of the driven tube, and then it must close again before the interface arrives.
This last step is very critical, since with these large copper diaphragms the interface often contains little copper particles that, if allowed to pass into the test-section, will be
20
expanded to a Mach number of 6 and can then sandblast the instrumentation on the airfoils. Usually, the valve remains open for about 100 ms or less. In these types of facilities, the addition of the fast-acting valve allows the compressed gas to be used in rigs that could not be mounted inside a standard shock tube.
The basic TTF operation (in either shock-mode, or blowdown mode) starts by evacuating the dump tank. The dump tank is connected to a two-stage vacuum pump and absolute pressures on the order of 0.05 psia are achieved when a turbine rig is fully installed. The dump tank is evacuated initially for three main reasons:
1) For ease in accelerating the turbine to the proper speed. With very little air in the dump tank, there is little resistance to the turbine, and relatively small air motors can be used to bring relatively heavy rotating assemblies up to design speed in a short period of time.
2) The vacuum allows for rapid flow establishment
3) If there were air in the dump tank at a pressure level on the order of a hundred torr, the initial pulse of test gas from the driven tube would compress the existing air resulting in an upstream facing shock wave instead of an expansion fan and prevent flow establishment in the nozzle.
With a rig installed in the expansion nozzle, this event can result in a severe hammer shock with ensuing mechanical damage to the rig.
Once the dump tank is evacuated, the driver and driven tubes are brought to the desired pressure levels. In the shock-tube mode, the driven and driver tubes are pressurized to different levels and the driver tube usually has a gas mixture of air and helium (to achieve higher incident shock Mach number for the same driver to driven tube pressure ratio, one could substitute heated hydrogen for helium) to create the proper incident shock speed and hence the desired temperature in the reflected region. In the blowdown mode, the double diaphragms are removed from the system and both the driver and driven tubes are pressurized to the desired level. This level is the same pressure that would have been obtained in the reflected-shock reservoir for a shock run. Depending on timing issues, the rotor is brought up to speed (about 1% lower than the physical speed desired at the test point) in the dump-tank that is at a vacuum. Once the proper speed is reached and
21
the load conditions are set, then either the double diaphragm cavity is vented to start the experiment (shock mode), or a manual trigger (blowdown mode) is used to start the experiment. The valve stays open for the same amount of time in either case, although there is more test-gas available in the blowdown case.
During an experiment, the rotor will increase in speed by about 2% total, and the system is timed so that the proper physical speed occurs at the proper temperature during the run so that the proper corrected speed is obtained. The pressure ratio across the machine is set using an exit choke on the rig. After the main valve closes, the test section is rapidly brought up in pressure to slow the rotor as quickly as possible.
The main difference between the two operational modes is that shock mode can generate high temperatures, although the duration of high-pressure, high-temperature gas is relatively short. While the rig is designed for use in scaled conditions (discussed below), using the shock mode one can increase the pressures and temperatures to that of actual engine conditions (at the expense of test-time). In blowdown mode, the
temperature is only slightly elevated (usually due to fact that the air is compressed and thus heated during the load conditions), but there is much more test gas available at the high pressure. Thus, there is the ability to have a longer test time. The downside is that because of the lower total temperature associated with the blowdown case; the flow establishment time is longer because of the lower sonic speed. No longer is the main limiter the interface, or the supply of gas, but rather it is the increase in speed of the turbine. Using the blowdown mode, one can take data later in the testing sequence (sometimes 50 ms later than in a shock mode), which usually reduces the noise in the system leftover from the initial star-up operations.
The key to any short-duration full-scaled rotating rig is the ability to match the non-dimensional parameters of interest. Briefly, these generally are the ratio of the metal to gas temperatures (for heat transfer studies), the Reynolds number, the pressure ratio and the corrected speed, and the Rosby number (for rotating effects). The Rosby number is usually matched if the same physical equipment is used with the proper corrected speed and thus it is not usually explicitly defined. For the heat transfer studies, the metal
temperature is scaled to room temperature, which usually dictates the inlet gas
22
temperature of about 600 K. If one is interested in aerodynamic studies as opposed to heat-transfer studies, this requirement is not present and one can match the Reynolds number, pressure ratio, and corrected speed with warm gas from a blowdown experiment.