- Email: [email protected]
Thermodynamic analysis of a demonstration concept for the long-duration storage and transfer of cryogenic propellants
Cryogenics xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Cryogenics journal homepage: www.elsevier.com/locate/cryogenics
Thermodynamic analysis of a demonstration concept for the long-duration storage and transfer of cryogenic propellants Russell B. Schweickart ⇑ Mail Code CO-8, Ball Aerospace, 1600 Commerce St., Boulder, CO 80301, United States
a r t i c l e
i n f o
Article history: Available online xxxx Keywords: Cryogenic Propellant Transfer Thermodynamic Analysis
a b s t r a c t In the development of a concept for an experimental platform for demonstrating technologies associated with the on-orbit handling, transfer and storage of cryogenic propellants, credibility is enhanced with simulations of operations using thermodynamic models. Predictions have further credibility if the modeling technique is verified against simulations of actual cryogenic fuel transfers during ground testing. This paper will demonstrate the capability of simulating the transfer of liquid hydrogen as preformed at NASA’s Glenn Research Center. Results of simulations of an experimental space mission developed at Ball Aerospace will then follow. The mission concept is intended to demonstrate the technologies and storage methodologies for supporting long-term storage and transfer of cryogenic fuels in space. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction
2. Mission justification
One method NASA is considering for sending rockets on deep space missions (e.g. lunar, Mars) involves refueling with cryogenic propellants in Earth orbit. The benefit over traditional single use rockets is significant improvement in delivered payload mass. Prior to deploying such infrastructure as fuel depots and Earth Departure Stages, demonstration missions are needed to prove necessary technologies such as cryogenic liquid-only transfer and long term fuel storage. To this end, Ball Aerospace developed a concept for a Cryogenic Propellant Storage and Transfer (CPST) mission for NASA Glenn Research Center (GRC). The effort entailed conducting a detailed design of the demonstration mission concept with the goal of meeting all CPST program capabilities including remaining within a cost cap of $200M that includes launch vehicle and group operations. Along with designing the hardware needed to demonstrate cryo-fluid management (CFM) technologies, Ball Aerospace developed thermodynamic system level models to simulate cryogenic storage and transfer operations. Concurrently, similar models proved highly accurate at simulating no-vent hydrogen transfers tests that were conducted at GRC in 1991. Verification of the thermal/fluid modeling technique provides confidence for the success of the demonstration mission and its extension to fuel depots and cryogenic propulsion stages (CPS).
Numerous studies of efficient access to deep space have noted the benefits of using orbiting fuel depots [1–4]. Not only can transferring cryogenic fuels in orbit increase the delivered mass to deep space destinations, but using this refueling option provides the potential to use the same reliable and less expensive rockets that currently insert satellites into low Earth orbit for longer duration missions. With low fluid temperatures and two-phase conditions, high specific-impulse fuels present low-gravity fluid management challenges. Demonstration missions will be necessary to verify the feasibility of long term, in-space cryogenic propellant storage and zero-gravity transfer of cryogenic propellants. These missions must demonstrate the ability to repeatedly extract gas-free liquid fuel from propellant storage tanks, measure fluid quantities onorbit as well as expel liquid-free gas for thermal management. Key technologies for making these operations possible include propellant management devices (PMD) for controlling the location of liquid in zero-gravity, validated micro-gravity mass gauging devices, and correlated fluid/thermodynamic simulations of fluid transfers and storage. Once flown, the CPST mission will validate these operations for future deep space flight programs.
⇑ Tel.: +1 303 939 4138. E-mail address: [email protected]
3. Mission concept The intent of the CPST mission is to demonstrate long term cryogenic fuel storage as well as liquid transfers. Ideally, fluid would be transferred in and out of the storage tank as would be
http://dx.doi.org/10.1016/j.cryogenics.2014.03.004 0011-2275/Ó 2014 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Schweickart RB. Thermodynamic analysis of a demonstration concept for the long-duration storage and transfer of cryogenic propellants. Cryogenics (2014), http://dx.doi.org/10.1016/j.cryogenics.2014.03.004
2
R.B. Schweickart / Cryogenics xxx (2014) xxx–xxx
the case for a fuel depot. A demonstration mission with such a capability would require greater complexity, limit verification of long term storage and likely exceed the cost goal. Thus, Ball Aerospace’s CPST mission concept includes a single 1650 liter hydrogen storage tank containing a propellant management device (PMD), a single 110 liter tank to which fluid is transferred and two pressurant tanks all supported above a spacecraft bus module as shown in Fig. 1. A cross-section of the supply tank is shown in Fig. 2. Additionally, a fully functional propellant storage depot must be capable of delivering oxygen as well as hydrogen. This CPST concept demonstrates fluid transfers with LH2 alone (LH2 being the colder and more difficult to manage of the two fluids), accomplishing the vast majority of NASA’s objectives for a CPST while remaining within the cost cap goal. The complete CPST mission concept includes ground operations for loading and conditioning hydrogen prior to launch and is addressed in the concept description report [5]. The primary operation intended to be demonstrated is the reliable and repeatable transfer of vapor-free liquid propellant at cryogenic temperatures. This is accomplished in the Ball mission concept by autogenous transfers of LH2 from the storage tank to the transfer tank using higher pressure gaseous hydrogen. Once a transfer has been completed, heaters on the transfer tank drive out the liquid propellant to space so that a second transfer operation can be performed. Following these transfers, the fuel tank will demonstrate its long term storage capability using a highly efficient multi-layer insulation (MLI), a thermodynamic vent system (TVS) and a vapor cooled shield (VCS). A final fuel transfer operation occurs when only 15% of the original quantity remains in the supply tank. A fuel transfer near the end of the mission will show the capability of the transfer process regardless of storage tank fill level. Fig. 3 shows a simplified schematic of the design.
Fig. 2. Cut-away view of CPST storage tank.
Fig. 3. Simplified CPST Schematic.
3.1. Storage tank fluid management design One of the most critical design features of the CPST storage tank that has only poor ground test verification capability is the propellant management device (PMD). Successful fuel transfer requires guaranteed access to vapor-free fuel. For emergency pressure control, however, access to liquid-free vapor is also necessary, even with a tank liquid fill fraction as great as 95%. The PMD designed for this demonstration mission is show in Fig. 4 and consist of 16
Fig. 4. CPST supply tank PMD concept.
Fig. 1. Expanded view of the CPST spacecraft components.
perforated and tapered panels supported radially from a center post. A similar PMD was used successfully in the Viking Orbiter 0.91 m (36 in.) diameter, 1.37 m (54 in.) long propellant tank. The titanium panels had a thickness of 0.18 mm (0.007 in.) [6]. At a fill fraction of 95%, the corresponding vapor bubble that would exist in a microgravity environment is supported at the open end of the tank where a vent tube can extract vapor. The
Please cite this article in press as: Schweickart RB. Thermodynamic analysis of a demonstration concept for the long-duration storage and transfer of cryogenic propellants. Cryogenics (2014), http://dx.doi.org/10.1016/j.cryogenics.2014.03.004
R.B. Schweickart / Cryogenics xxx (2014) xxx–xxx
taper in the panels ensures any vapor generated within the liquid volume is forced in the direction of the vapor bubble. As liquid exits the tank from the opposite end of the tank, the liquid conforms to the shape of the panels, maintaining liquid at the vent port. Fig. 5 shows the predicted location of liquid within the tank over the duration of the mission.
3.2. Storage tank thermal management Ball Aerospace has developed a multi-layer insulation [7] that not only has greater thermal isolating capabilities than traditional MLI, but the new MLI also provides structural support for a vaporcooled shield. VCS’s that have flown on cryostats for cooling space telescopes such as SIRTF [8] and NICMOS [9] are traditionally structurally supported by the same struts that support the cryogenic fuel tank assembly within a vacuum shell. Due to the size of the tanks envisioned for fuel depots, this traditional technique will not be possible. A TVS consists of an orifice in the vent line through which liquid fuel is directed. Upon expansion through the orifice, the two-phase fluid on the downstream side of the device is colder than the contained fuel and thus provides a passive cooling capability [10]. This venting fluid can then be directed into lines attached to the support struts and a VCS located within the MLI to further limit parasitic heating of the fuel in the storage tank. More detail of TVS operation is given in Section 7.2.
4. Model description Simulation of fluid conditions within the CPST tanks and the interaction with the orbital environment were conducted with SINDA/FLUINT [11] and the corresponding graphical user interface tool Thermal Desktop. This software allows for the simultaneous simulation of conductive and radiative heat transfer between CPST assembly components and the full thermodynamic one-dimensional interaction with cryogenic fuels. This includes non-equilibrium heat and mass transfer between vapor and liquid volumes within tanks. Results of the analysis are shown in Section 7. Many analytical tools have been developed over the years to simulate cryogenic fluid flow with varying degrees of success [12–14]. SINDA/FLUINT has been used in this case because the program is widely used across the aerospace industry, its results have been verified accurate for a wide range of thermal/fluid applications, the complete thermal and fluid subassemblies are modeled together, and models can easily be modified/expanded to include any range of spacecraft designs and fluid flow configurations.
3
While Ball Aerospace has used SINDA/FLUINT extensively to successfully model cryogen tanks and other fluid flow tests [8,9], confidence in model predictions should be minimal at best until the model has been correlated to existing tests of like hardware. For this application, the most applicable tests were those conducted at NASA/Glenn Research Center involving the no-vent transfer of liquid hydrogen between insulated tanks. 5. No-vent ground test As part of the decades-long development of CPST concepts and components at NASA Glenn Research Center, a set of no-vent liquid hydrogen transfer tests was conducted in 1991 [12]. Liquid hydrogen was transferred into a 4.96 m3 test tank from a 49.2 m3 ground storage dewar at various liquid inlet temperatures, mass flow rates and initial test tank wall temperatures. Each test was considered complete when the fill level had reached 94% in the test tank. The test report shows results from 12 successful transfers including final test tank pressure. Liquid inlet temperatures varied from 21.7 to 23.4 K with mass flow rates between 234 and 535 kg/h. The range of initial wall temperatures varied from 21.7 to 126.1 K. Final tank pressures ranged from 0.177 to 0.257 MPa. Heat transfer to the test tank was limited with the use of 34 layers of MLI within a guard vacuum and a liquid hydrogen ‘‘coldguard’’ through which all instrumentation and fluid lines passed. The steady state heat load on the tank was estimated at 4.7 W. Each test started with a test tank pressure of less than 0.0138 MPa (2 psia) and the supply tank pressure at 0.31 MPa (45 psia). The supply tank incorporated a pressure recirculation loop that allowed the pressure in the supply tank to be maintained at a constant level throughout the course of each test, while the test tank pressure remained uncontrolled. The pressurization process occurred rapidly (45–90 min, depending on test conditions), while the time required for enough heat to be absorbed by the supply tank hydrogen to reach saturation conditions at the set point pressure was many hours. Thus the fluid flowing to the test tank was subcooled allowing the fluid to remain in a liquid state even after absorbing heat in the transfer lines. Once the test tank had reached a fill level of 94%, the fill line was closed. The test tank was then allowed to stay in a locked-up, quiescent state for an hour after each fill unless the tank reached its maximum operating pressure of 0.358 MPa (52 psia), at which point it was vented to atmosphere. Another variable in the test was the method liquid hydrogen entered the test tank. Liquid hydrogen flowed into the test tank through 13 spray nozzles at the top of the tank during some tests, while liquid flowed in from a single nozzle at the bottom of the tank during others. Results showed that fill through the top spray nozzles caused the tank walls to cool and thus result in lower final test tank pressures than when filled with the lower nozzle, all other parameters equal. Fill with the lower nozzle, however, was shown to be quite successful too. The amount of liquid that entered the tank before the lower nozzle was submerged was estimated at 7%. Table 1 shows the initial parameters and results for each of the tests. 6. Verification of ground test models
Fig. 5. Location of LH2 during phases of mission.
The document summarizing the GRC work presents sufficient detail of the testing to allow for a complete system level thermal/thermodynamic analysis with SINDA/FLUINT. In order to correlate analytic results to the test results a single correlation parameter is needed representing the heat transfer rate between vapor and liquid in the test tank, which is dependent on the surface area between the vapor and liquid. Figs. 6 and 7 show a
Please cite this article in press as: Schweickart RB. Thermodynamic analysis of a demonstration concept for the long-duration storage and transfer of cryogenic propellants. Cryogenics (2014), http://dx.doi.org/10.1016/j.cryogenics.2014.03.004
4
R.B. Schweickart / Cryogenics xxx (2014) xxx–xxx
Table 1 Parameters and results for successful no-vent transfer tests. Test ID
Liquid inlet
Initial wall temperature (K)
Liquid inlet temperature (K)
Inlet mass flowrate (kg/h)
Final pressure (MPa)
Final fill percentage
Fill 18 Fill 19 Fill 20 Fill 21 Fill 22 Fill 23 K2091 K251 K2631 K2227 K2731
Bottom spray Top spray Top spray Bottom spray Bottom spray Top spray Top spray Top spray Top spray Bottom spray bottom spray
21.7 104.4 126.1 101.7 66.7 70.0 18.1 23.2 86.0 17.0 87.9
22.2 23.0 22.4 22.4 21.7 22.8 23.4 21.9 22.3 22.0 21.9
495.0 421.8 337.7 448.6 505.0 419.5 442.2 234.0 477.2 534.5 510.0
0.177 0.227 0.257 0.233 0.187 0.210 0.204 0.181 0.195 0.177 0.196
94 94 94 94 94 94 94 94 94 94 94
the tank), is that the success of the transfer process is insensitive to the method of injecting liquid into the receiving tank. 7. CPST model results
Fig. 6. Receiving tank pressure during no-vent fill tests (see Table 1 for case descriptions).
With a degree of confidence in the ability of the modeling technique to simulate ground transfers, the method could then be applied to a transfer in zero-g as part of the CPST concept. While the uncertainty in the fluid conditions in the transfer tank have been greatly reduced by the GRC test results, the same cannot be said for the supply/storage tank. During ground operations, the vapor liquid interface in the storage tank is defined by gravity and the profile of the tank: a flat surface with an area defined by the horizontal cross section of the tank at a given fill level. In space, however, the interface area will be defined by the size of the vapor bubble when the tank is full, but then more and more by the shape of the PMD as the fuel depletes. Fortunately, the model can account for the change in interface area, at least to first order, based on assumption of liquid location as a function of fill level. 7.1. Initial Transfer Operations
Fig. 7. Correlated model results assuming equilibrium conditions in receiving tank.
comparison of tank pressures from the GRC test results and correlated model results from a SINDA/FLUINT simulation for three of the test cases, all when liquid entered the test tank through a single nozzle at the bottom of the tank. In each of these cases, the model used initial conditions provided in the paper, and a parametric analysis was conducted by varying the vapor-to-liquid heat transfer coefficient. Simulations matched the test data best when the heat transfer coefficient was set arbitrarily high. In fact the analysis results matched test data when the vapor and liquid in the transfer tank were assumed in an equilibrium state. Even for Fill 21 when the wall temperature started at 101.7 K and an inlet liquid temperature of 22.2 K, the fill was successful with only 1.47 K of subcooling based on the final tank pressure. The fortunate implication of this result, given that a fill level of 94% was achieved with the minimum possible interface area between the vapor and liquid (a disc representing the cross-section of the tank, as compared to that of vapor droplets from 13 nozzles spraying into
The first simulation involves two successive transfers of liquid hydrogen from the storage tank to the smaller tank as outlined above in the description of the mission. Predicted supply and transfer tank pressures are shown in Fig. 8. Initially, the transfer tank is warm. This will not likely be the case for transfers to and from fuel depots, but the situation clearly needs to be accounted for. In this case, after an initial helium gas charge (for launch) has been vented from the transfer tank (Stage 1), pressure that has been building in the storage tank since lock-up prior to launch is reduced by flowing liquid-free vapor through the transfer tank
Fig. 8. Predicted supply and transfer tank pressures during transfer operations.
Please cite this article in press as: Schweickart RB. Thermodynamic analysis of a demonstration concept for the long-duration storage and transfer of cryogenic propellants. Cryogenics (2014), http://dx.doi.org/10.1016/j.cryogenics.2014.03.004
R.B. Schweickart / Cryogenics xxx (2014) xxx–xxx
and then out to space (Stage 2). While some the of the initial hydrogen charge is lost in this process, cooling the transfer tank not only insures a successful transfer operation, but also prevents a hazardous condition which can result from over-pressurization of the transfer tank if cold liquid were to come into contact with warm tank walls. When the supply tank pressure drops to 0.034 MPa (5 psia), well above the triple point of hydrogen at 0.007 MPa (1 psia), the transfer tank vent is then closed and warm gaseous hydrogen at a regulated 0.14 MPa (20 psia) fed into the supply tank through a diffuser from the hydrogen pressurant tanks (Stage 3). The diffuser is necessary to minimize the heat transfer to the contained liquid. Liquid then flows from the supply tank to the transfer tank (Stage 4) until the tank is 95% full, when valves on the pressurant tank and in the transfer line are closed (Stage 5). The model accounts for parasitic heating to the transfer lines. The spacecraft then fires thrusters in order to perform a settled quantity gaging operation in each tank, followed by a liquid expulsion from the transfer tank with the use of heaters (Stage 6). When approximately 40% of fuel remains in the transfer tank (simulating a transfer to a partially filled tank), the vent valve is closed, and the second transfer is conducted using the same procedure (Stage 7 like Stage 3, Stage 8 line Stage 4 and so on). The model simulates each of these stages including operation of valves and regulators, and predicts 5 h will be necessary for two transfers (Stages 1–8). Fig. 9 shows the predicted fill levels within the tanks during the transfer operations, and Fig. 10 shows the predicted corresponding gaseous hydrogen mass change in the pressurant tank. The model predicts only 40 g of hydrogen gas are needed to perform the first transfer and 50 g for the second.
5
Fig. 10. Pressurant tank hydrogen mass during transfer operations.
7.2. Mission storage phase simulation During long term storage operations, tank pressure control is accomplished by cycling the vent valve. Due to the necessity of a reasonable size of the TVS orifice, the tank pressure will decrease as hydrogen vents through the TVS. One reaching a selected level, the vent valve is closed. The tank pressure rises as parasitic heat is absorbed by the contained fluid, and the cycle is repeated. For this CPST concept, supply tank pressures were cycled between 0.034 and 0.041 MPa (5 and 6 psia). The hydrogen flow rate from the tank is controlled by the size of the TVS orifice as the flow is choked. The frequency of vent valve operation is determined by the parasitic heat load. Predictions for this CPST are shown in Figs. 11–13, indicating a cycle frequency of 35 h with a depletion of about 0.9 kg during each TVS operating period. Fig. 11 shows the flow rate cycling between 0 and 100 g/h when
Fig. 11. Mass flow rate from supply tank at start of long term storage phase of mission.
Fig. 12. Heat load/sink on supply tank at start of long term storage phase of mission.
Fig. 9. Fill level in supply and transfer tanks during transfer operations.
the TVS vent valve is open. Fig. 12 shows the corresponding heat load on the supply tank (negative values indicates cooling of the supply tank when the TVS is operational). Finally, Fig. 13 shows the predicted mass of hydrogen in the supply tank during this same period. If no further transfer operations were to be conducted at this point, the model estimates the hydrogen would be depleted in 126 days as compared to 112 days if vapor were vented directly and continuously from the tank through the VCS rather than flowing liquid through the TVS.
Please cite this article in press as: Schweickart RB. Thermodynamic analysis of a demonstration concept for the long-duration storage and transfer of cryogenic propellants. Cryogenics (2014), http://dx.doi.org/10.1016/j.cryogenics.2014.03.004
6
R.B. Schweickart / Cryogenics xxx (2014) xxx–xxx
transferring hydrogen when the tank is near empty is a necessity to verifying the viability of this CPST concept. The larger surface area of the vapor/liquid interface in the supply tank increases the heat transfer rate between the two phases, and the smaller mass of remaining liquid hydrogen increases the rate that the liquid approaches the saturation conditions corresponding to the tank pressure. Thus less subcooling is available to enable liquid transfer as the tank depletes. Nevertheless, the model predicts a successful transfer when the supply tank liquid charge is only 15% by volume. Figs. 14 and 15 show predictions for transfer and supply tank pressures and fill levels during this transfer operation. The model predicts a transfer tank fill to 95% in 32 min during which 106 g of pressurant hydrogen are needed. 8. Conclusion Fig. 13. Predicted mass loss rate during long term storage.
Simulations of CPST operations including microgravity cryogenic fuel transfers and long term storage have been conducted with the use of a verified thermal/thermodynamic analysis tool. Results predict successful transfer operations even at a supply tank fill level of 15%. Correlations of similar models to test results from no-vent hydrogen transfer tests on the ground provide insight into on orbit transfers and confidence in the model results. These results should help to justify the need for a timely CPST demonstration mission and kick-start a move to using fuel depots for deep space missions. References
Fig. 14. Predicted supply and transfer tank pressures during late transfer operation.
Fig. 15. Liquid fill quantities in supply and transfer tanks during late transfer.
[1] McLean C, Mustafi S, Walls L, Pitchford B, Wollen M, Schmidt J. Simple, robust cryogenic propellant depot for near term applications. In: 2011 IEEE Aerospace Conference. Big Sky, MT, March 2011. [2] Zegler F, Kutter B. Evolving to a depot-based space transportation architecture. In: Space 2010. Anaheim, CA; September 2010 [AIAA-2010-8638]. [3] Braun RD. Investment in the Future: Overview of NASA’s Space Technology. NASA Chief Technologist. May 5, 2010. [4] McLean C, Mills G, Riesco M, Meyer M, Plachta D, Hurlbert E. Long term space storage and delivery of cryogenic propellants for exploration, In: Joint Propulsion Conference, Hartford, CT; July 2008 [AIAA-2008-4853]. [5] McLean C, et al. In-space cryogenic propellant storage and transfer (CPST) demonstration mission concept study report, In: Ball Aerospace and Technologies Corp., Boulder, CO., NASA Glenn Research Center (GRC). Contract Number NNC11CA29C. Final, Report, December 2011. [6] Tam W, Ballinger I, Jaekle Jr DE. Surface tension PMD tank for on orbit fluid transfer. Hartford, CT, July 2008 [AIAA 2008-5105]. [7] Dye S, Kopelove A, Mills GL. Load Responsive Multilayer Insulation Performance Testing. CEC, June 2013. [8] Hopkins RA, Finley PT, Schweickart RB, Volz SM. Cryogenic/thermal system for the SIRTF cryogenic telescope assembly. Proc SPIE 2003;4850:42–9. [9] Thompson RI. Near-infrared camera and multi-object spectrometer (NICMOS): the near-infrared space mission on HST. Proc SPIE 1994;2209:319–30. [10] Sterbentz WH. Liquid propellant thermal conditioning system. NASA CR72365. 1968. [11] SINDA/FLUINT User’s Manual. Version 5.5. C&R Technologies. Octeber 2011. [12] Taylor WJ, Chato DJ. Comparing the results of an analytical model of the novent fill process with no-vent fill test results for a 4.96 m3 tank. July 1992 [AIAA-92-3078]. [13] Plachta DW, Christie RJ, Carlberg E, Feller JR. Cryogenic propellant storage analyses and design tool evolved from in-space cryogenic propellant depot project. Huntsville, AL: Space Cryogenics Workshop; July 2007. [14] Chato DJ, Marchetta J, Hochstein JI, Kassemi M. Approaches to validation of models for low gravity fluid behavior. September 2005 [AIAA-2004-1150].
7.3. Low fill level transfer operation Due to the significantly different amount and location of the liquid in the supply tank from beginning to end of the mission,
Please cite this article in press as: Schweickart RB. Thermodynamic analysis of a demonstration concept for the long-duration storage and transfer of cryogenic propellants. Cryogenics (2014), http://dx.doi.org/10.1016/j.cryogenics.2014.03.004
Contents lists available at ScienceDirect
Cryogenics journal homepage: www.elsevier.com/locate/cryogenics
Thermodynamic analysis of a demonstration concept for the long-duration storage and transfer of cryogenic propellants Russell B. Schweickart ⇑ Mail Code CO-8, Ball Aerospace, 1600 Commerce St., Boulder, CO 80301, United States
a r t i c l e
i n f o
Article history: Available online xxxx Keywords: Cryogenic Propellant Transfer Thermodynamic Analysis
a b s t r a c t In the development of a concept for an experimental platform for demonstrating technologies associated with the on-orbit handling, transfer and storage of cryogenic propellants, credibility is enhanced with simulations of operations using thermodynamic models. Predictions have further credibility if the modeling technique is verified against simulations of actual cryogenic fuel transfers during ground testing. This paper will demonstrate the capability of simulating the transfer of liquid hydrogen as preformed at NASA’s Glenn Research Center. Results of simulations of an experimental space mission developed at Ball Aerospace will then follow. The mission concept is intended to demonstrate the technologies and storage methodologies for supporting long-term storage and transfer of cryogenic fuels in space. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction
2. Mission justification
One method NASA is considering for sending rockets on deep space missions (e.g. lunar, Mars) involves refueling with cryogenic propellants in Earth orbit. The benefit over traditional single use rockets is significant improvement in delivered payload mass. Prior to deploying such infrastructure as fuel depots and Earth Departure Stages, demonstration missions are needed to prove necessary technologies such as cryogenic liquid-only transfer and long term fuel storage. To this end, Ball Aerospace developed a concept for a Cryogenic Propellant Storage and Transfer (CPST) mission for NASA Glenn Research Center (GRC). The effort entailed conducting a detailed design of the demonstration mission concept with the goal of meeting all CPST program capabilities including remaining within a cost cap of $200M that includes launch vehicle and group operations. Along with designing the hardware needed to demonstrate cryo-fluid management (CFM) technologies, Ball Aerospace developed thermodynamic system level models to simulate cryogenic storage and transfer operations. Concurrently, similar models proved highly accurate at simulating no-vent hydrogen transfers tests that were conducted at GRC in 1991. Verification of the thermal/fluid modeling technique provides confidence for the success of the demonstration mission and its extension to fuel depots and cryogenic propulsion stages (CPS).
Numerous studies of efficient access to deep space have noted the benefits of using orbiting fuel depots [1–4]. Not only can transferring cryogenic fuels in orbit increase the delivered mass to deep space destinations, but using this refueling option provides the potential to use the same reliable and less expensive rockets that currently insert satellites into low Earth orbit for longer duration missions. With low fluid temperatures and two-phase conditions, high specific-impulse fuels present low-gravity fluid management challenges. Demonstration missions will be necessary to verify the feasibility of long term, in-space cryogenic propellant storage and zero-gravity transfer of cryogenic propellants. These missions must demonstrate the ability to repeatedly extract gas-free liquid fuel from propellant storage tanks, measure fluid quantities onorbit as well as expel liquid-free gas for thermal management. Key technologies for making these operations possible include propellant management devices (PMD) for controlling the location of liquid in zero-gravity, validated micro-gravity mass gauging devices, and correlated fluid/thermodynamic simulations of fluid transfers and storage. Once flown, the CPST mission will validate these operations for future deep space flight programs.
⇑ Tel.: +1 303 939 4138. E-mail address: [email protected]
3. Mission concept The intent of the CPST mission is to demonstrate long term cryogenic fuel storage as well as liquid transfers. Ideally, fluid would be transferred in and out of the storage tank as would be
http://dx.doi.org/10.1016/j.cryogenics.2014.03.004 0011-2275/Ó 2014 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Schweickart RB. Thermodynamic analysis of a demonstration concept for the long-duration storage and transfer of cryogenic propellants. Cryogenics (2014), http://dx.doi.org/10.1016/j.cryogenics.2014.03.004
2
R.B. Schweickart / Cryogenics xxx (2014) xxx–xxx
the case for a fuel depot. A demonstration mission with such a capability would require greater complexity, limit verification of long term storage and likely exceed the cost goal. Thus, Ball Aerospace’s CPST mission concept includes a single 1650 liter hydrogen storage tank containing a propellant management device (PMD), a single 110 liter tank to which fluid is transferred and two pressurant tanks all supported above a spacecraft bus module as shown in Fig. 1. A cross-section of the supply tank is shown in Fig. 2. Additionally, a fully functional propellant storage depot must be capable of delivering oxygen as well as hydrogen. This CPST concept demonstrates fluid transfers with LH2 alone (LH2 being the colder and more difficult to manage of the two fluids), accomplishing the vast majority of NASA’s objectives for a CPST while remaining within the cost cap goal. The complete CPST mission concept includes ground operations for loading and conditioning hydrogen prior to launch and is addressed in the concept description report [5]. The primary operation intended to be demonstrated is the reliable and repeatable transfer of vapor-free liquid propellant at cryogenic temperatures. This is accomplished in the Ball mission concept by autogenous transfers of LH2 from the storage tank to the transfer tank using higher pressure gaseous hydrogen. Once a transfer has been completed, heaters on the transfer tank drive out the liquid propellant to space so that a second transfer operation can be performed. Following these transfers, the fuel tank will demonstrate its long term storage capability using a highly efficient multi-layer insulation (MLI), a thermodynamic vent system (TVS) and a vapor cooled shield (VCS). A final fuel transfer operation occurs when only 15% of the original quantity remains in the supply tank. A fuel transfer near the end of the mission will show the capability of the transfer process regardless of storage tank fill level. Fig. 3 shows a simplified schematic of the design.
Fig. 2. Cut-away view of CPST storage tank.
Fig. 3. Simplified CPST Schematic.
3.1. Storage tank fluid management design One of the most critical design features of the CPST storage tank that has only poor ground test verification capability is the propellant management device (PMD). Successful fuel transfer requires guaranteed access to vapor-free fuel. For emergency pressure control, however, access to liquid-free vapor is also necessary, even with a tank liquid fill fraction as great as 95%. The PMD designed for this demonstration mission is show in Fig. 4 and consist of 16
Fig. 4. CPST supply tank PMD concept.
Fig. 1. Expanded view of the CPST spacecraft components.
perforated and tapered panels supported radially from a center post. A similar PMD was used successfully in the Viking Orbiter 0.91 m (36 in.) diameter, 1.37 m (54 in.) long propellant tank. The titanium panels had a thickness of 0.18 mm (0.007 in.) [6]. At a fill fraction of 95%, the corresponding vapor bubble that would exist in a microgravity environment is supported at the open end of the tank where a vent tube can extract vapor. The
Please cite this article in press as: Schweickart RB. Thermodynamic analysis of a demonstration concept for the long-duration storage and transfer of cryogenic propellants. Cryogenics (2014), http://dx.doi.org/10.1016/j.cryogenics.2014.03.004
R.B. Schweickart / Cryogenics xxx (2014) xxx–xxx
taper in the panels ensures any vapor generated within the liquid volume is forced in the direction of the vapor bubble. As liquid exits the tank from the opposite end of the tank, the liquid conforms to the shape of the panels, maintaining liquid at the vent port. Fig. 5 shows the predicted location of liquid within the tank over the duration of the mission.
3.2. Storage tank thermal management Ball Aerospace has developed a multi-layer insulation [7] that not only has greater thermal isolating capabilities than traditional MLI, but the new MLI also provides structural support for a vaporcooled shield. VCS’s that have flown on cryostats for cooling space telescopes such as SIRTF [8] and NICMOS [9] are traditionally structurally supported by the same struts that support the cryogenic fuel tank assembly within a vacuum shell. Due to the size of the tanks envisioned for fuel depots, this traditional technique will not be possible. A TVS consists of an orifice in the vent line through which liquid fuel is directed. Upon expansion through the orifice, the two-phase fluid on the downstream side of the device is colder than the contained fuel and thus provides a passive cooling capability [10]. This venting fluid can then be directed into lines attached to the support struts and a VCS located within the MLI to further limit parasitic heating of the fuel in the storage tank. More detail of TVS operation is given in Section 7.2.
4. Model description Simulation of fluid conditions within the CPST tanks and the interaction with the orbital environment were conducted with SINDA/FLUINT [11] and the corresponding graphical user interface tool Thermal Desktop. This software allows for the simultaneous simulation of conductive and radiative heat transfer between CPST assembly components and the full thermodynamic one-dimensional interaction with cryogenic fuels. This includes non-equilibrium heat and mass transfer between vapor and liquid volumes within tanks. Results of the analysis are shown in Section 7. Many analytical tools have been developed over the years to simulate cryogenic fluid flow with varying degrees of success [12–14]. SINDA/FLUINT has been used in this case because the program is widely used across the aerospace industry, its results have been verified accurate for a wide range of thermal/fluid applications, the complete thermal and fluid subassemblies are modeled together, and models can easily be modified/expanded to include any range of spacecraft designs and fluid flow configurations.
3
While Ball Aerospace has used SINDA/FLUINT extensively to successfully model cryogen tanks and other fluid flow tests [8,9], confidence in model predictions should be minimal at best until the model has been correlated to existing tests of like hardware. For this application, the most applicable tests were those conducted at NASA/Glenn Research Center involving the no-vent transfer of liquid hydrogen between insulated tanks. 5. No-vent ground test As part of the decades-long development of CPST concepts and components at NASA Glenn Research Center, a set of no-vent liquid hydrogen transfer tests was conducted in 1991 [12]. Liquid hydrogen was transferred into a 4.96 m3 test tank from a 49.2 m3 ground storage dewar at various liquid inlet temperatures, mass flow rates and initial test tank wall temperatures. Each test was considered complete when the fill level had reached 94% in the test tank. The test report shows results from 12 successful transfers including final test tank pressure. Liquid inlet temperatures varied from 21.7 to 23.4 K with mass flow rates between 234 and 535 kg/h. The range of initial wall temperatures varied from 21.7 to 126.1 K. Final tank pressures ranged from 0.177 to 0.257 MPa. Heat transfer to the test tank was limited with the use of 34 layers of MLI within a guard vacuum and a liquid hydrogen ‘‘coldguard’’ through which all instrumentation and fluid lines passed. The steady state heat load on the tank was estimated at 4.7 W. Each test started with a test tank pressure of less than 0.0138 MPa (2 psia) and the supply tank pressure at 0.31 MPa (45 psia). The supply tank incorporated a pressure recirculation loop that allowed the pressure in the supply tank to be maintained at a constant level throughout the course of each test, while the test tank pressure remained uncontrolled. The pressurization process occurred rapidly (45–90 min, depending on test conditions), while the time required for enough heat to be absorbed by the supply tank hydrogen to reach saturation conditions at the set point pressure was many hours. Thus the fluid flowing to the test tank was subcooled allowing the fluid to remain in a liquid state even after absorbing heat in the transfer lines. Once the test tank had reached a fill level of 94%, the fill line was closed. The test tank was then allowed to stay in a locked-up, quiescent state for an hour after each fill unless the tank reached its maximum operating pressure of 0.358 MPa (52 psia), at which point it was vented to atmosphere. Another variable in the test was the method liquid hydrogen entered the test tank. Liquid hydrogen flowed into the test tank through 13 spray nozzles at the top of the tank during some tests, while liquid flowed in from a single nozzle at the bottom of the tank during others. Results showed that fill through the top spray nozzles caused the tank walls to cool and thus result in lower final test tank pressures than when filled with the lower nozzle, all other parameters equal. Fill with the lower nozzle, however, was shown to be quite successful too. The amount of liquid that entered the tank before the lower nozzle was submerged was estimated at 7%. Table 1 shows the initial parameters and results for each of the tests. 6. Verification of ground test models
Fig. 5. Location of LH2 during phases of mission.
The document summarizing the GRC work presents sufficient detail of the testing to allow for a complete system level thermal/thermodynamic analysis with SINDA/FLUINT. In order to correlate analytic results to the test results a single correlation parameter is needed representing the heat transfer rate between vapor and liquid in the test tank, which is dependent on the surface area between the vapor and liquid. Figs. 6 and 7 show a
Please cite this article in press as: Schweickart RB. Thermodynamic analysis of a demonstration concept for the long-duration storage and transfer of cryogenic propellants. Cryogenics (2014), http://dx.doi.org/10.1016/j.cryogenics.2014.03.004
4
R.B. Schweickart / Cryogenics xxx (2014) xxx–xxx
Table 1 Parameters and results for successful no-vent transfer tests. Test ID
Liquid inlet
Initial wall temperature (K)
Liquid inlet temperature (K)
Inlet mass flowrate (kg/h)
Final pressure (MPa)
Final fill percentage
Fill 18 Fill 19 Fill 20 Fill 21 Fill 22 Fill 23 K2091 K251 K2631 K2227 K2731
Bottom spray Top spray Top spray Bottom spray Bottom spray Top spray Top spray Top spray Top spray Bottom spray bottom spray
21.7 104.4 126.1 101.7 66.7 70.0 18.1 23.2 86.0 17.0 87.9
22.2 23.0 22.4 22.4 21.7 22.8 23.4 21.9 22.3 22.0 21.9
495.0 421.8 337.7 448.6 505.0 419.5 442.2 234.0 477.2 534.5 510.0
0.177 0.227 0.257 0.233 0.187 0.210 0.204 0.181 0.195 0.177 0.196
94 94 94 94 94 94 94 94 94 94 94
the tank), is that the success of the transfer process is insensitive to the method of injecting liquid into the receiving tank. 7. CPST model results
Fig. 6. Receiving tank pressure during no-vent fill tests (see Table 1 for case descriptions).
With a degree of confidence in the ability of the modeling technique to simulate ground transfers, the method could then be applied to a transfer in zero-g as part of the CPST concept. While the uncertainty in the fluid conditions in the transfer tank have been greatly reduced by the GRC test results, the same cannot be said for the supply/storage tank. During ground operations, the vapor liquid interface in the storage tank is defined by gravity and the profile of the tank: a flat surface with an area defined by the horizontal cross section of the tank at a given fill level. In space, however, the interface area will be defined by the size of the vapor bubble when the tank is full, but then more and more by the shape of the PMD as the fuel depletes. Fortunately, the model can account for the change in interface area, at least to first order, based on assumption of liquid location as a function of fill level. 7.1. Initial Transfer Operations
Fig. 7. Correlated model results assuming equilibrium conditions in receiving tank.
comparison of tank pressures from the GRC test results and correlated model results from a SINDA/FLUINT simulation for three of the test cases, all when liquid entered the test tank through a single nozzle at the bottom of the tank. In each of these cases, the model used initial conditions provided in the paper, and a parametric analysis was conducted by varying the vapor-to-liquid heat transfer coefficient. Simulations matched the test data best when the heat transfer coefficient was set arbitrarily high. In fact the analysis results matched test data when the vapor and liquid in the transfer tank were assumed in an equilibrium state. Even for Fill 21 when the wall temperature started at 101.7 K and an inlet liquid temperature of 22.2 K, the fill was successful with only 1.47 K of subcooling based on the final tank pressure. The fortunate implication of this result, given that a fill level of 94% was achieved with the minimum possible interface area between the vapor and liquid (a disc representing the cross-section of the tank, as compared to that of vapor droplets from 13 nozzles spraying into
The first simulation involves two successive transfers of liquid hydrogen from the storage tank to the smaller tank as outlined above in the description of the mission. Predicted supply and transfer tank pressures are shown in Fig. 8. Initially, the transfer tank is warm. This will not likely be the case for transfers to and from fuel depots, but the situation clearly needs to be accounted for. In this case, after an initial helium gas charge (for launch) has been vented from the transfer tank (Stage 1), pressure that has been building in the storage tank since lock-up prior to launch is reduced by flowing liquid-free vapor through the transfer tank
Fig. 8. Predicted supply and transfer tank pressures during transfer operations.
Please cite this article in press as: Schweickart RB. Thermodynamic analysis of a demonstration concept for the long-duration storage and transfer of cryogenic propellants. Cryogenics (2014), http://dx.doi.org/10.1016/j.cryogenics.2014.03.004
R.B. Schweickart / Cryogenics xxx (2014) xxx–xxx
and then out to space (Stage 2). While some the of the initial hydrogen charge is lost in this process, cooling the transfer tank not only insures a successful transfer operation, but also prevents a hazardous condition which can result from over-pressurization of the transfer tank if cold liquid were to come into contact with warm tank walls. When the supply tank pressure drops to 0.034 MPa (5 psia), well above the triple point of hydrogen at 0.007 MPa (1 psia), the transfer tank vent is then closed and warm gaseous hydrogen at a regulated 0.14 MPa (20 psia) fed into the supply tank through a diffuser from the hydrogen pressurant tanks (Stage 3). The diffuser is necessary to minimize the heat transfer to the contained liquid. Liquid then flows from the supply tank to the transfer tank (Stage 4) until the tank is 95% full, when valves on the pressurant tank and in the transfer line are closed (Stage 5). The model accounts for parasitic heating to the transfer lines. The spacecraft then fires thrusters in order to perform a settled quantity gaging operation in each tank, followed by a liquid expulsion from the transfer tank with the use of heaters (Stage 6). When approximately 40% of fuel remains in the transfer tank (simulating a transfer to a partially filled tank), the vent valve is closed, and the second transfer is conducted using the same procedure (Stage 7 like Stage 3, Stage 8 line Stage 4 and so on). The model simulates each of these stages including operation of valves and regulators, and predicts 5 h will be necessary for two transfers (Stages 1–8). Fig. 9 shows the predicted fill levels within the tanks during the transfer operations, and Fig. 10 shows the predicted corresponding gaseous hydrogen mass change in the pressurant tank. The model predicts only 40 g of hydrogen gas are needed to perform the first transfer and 50 g for the second.
5
Fig. 10. Pressurant tank hydrogen mass during transfer operations.
7.2. Mission storage phase simulation During long term storage operations, tank pressure control is accomplished by cycling the vent valve. Due to the necessity of a reasonable size of the TVS orifice, the tank pressure will decrease as hydrogen vents through the TVS. One reaching a selected level, the vent valve is closed. The tank pressure rises as parasitic heat is absorbed by the contained fluid, and the cycle is repeated. For this CPST concept, supply tank pressures were cycled between 0.034 and 0.041 MPa (5 and 6 psia). The hydrogen flow rate from the tank is controlled by the size of the TVS orifice as the flow is choked. The frequency of vent valve operation is determined by the parasitic heat load. Predictions for this CPST are shown in Figs. 11–13, indicating a cycle frequency of 35 h with a depletion of about 0.9 kg during each TVS operating period. Fig. 11 shows the flow rate cycling between 0 and 100 g/h when
Fig. 11. Mass flow rate from supply tank at start of long term storage phase of mission.
Fig. 12. Heat load/sink on supply tank at start of long term storage phase of mission.
Fig. 9. Fill level in supply and transfer tanks during transfer operations.
the TVS vent valve is open. Fig. 12 shows the corresponding heat load on the supply tank (negative values indicates cooling of the supply tank when the TVS is operational). Finally, Fig. 13 shows the predicted mass of hydrogen in the supply tank during this same period. If no further transfer operations were to be conducted at this point, the model estimates the hydrogen would be depleted in 126 days as compared to 112 days if vapor were vented directly and continuously from the tank through the VCS rather than flowing liquid through the TVS.
Please cite this article in press as: Schweickart RB. Thermodynamic analysis of a demonstration concept for the long-duration storage and transfer of cryogenic propellants. Cryogenics (2014), http://dx.doi.org/10.1016/j.cryogenics.2014.03.004
6
R.B. Schweickart / Cryogenics xxx (2014) xxx–xxx
transferring hydrogen when the tank is near empty is a necessity to verifying the viability of this CPST concept. The larger surface area of the vapor/liquid interface in the supply tank increases the heat transfer rate between the two phases, and the smaller mass of remaining liquid hydrogen increases the rate that the liquid approaches the saturation conditions corresponding to the tank pressure. Thus less subcooling is available to enable liquid transfer as the tank depletes. Nevertheless, the model predicts a successful transfer when the supply tank liquid charge is only 15% by volume. Figs. 14 and 15 show predictions for transfer and supply tank pressures and fill levels during this transfer operation. The model predicts a transfer tank fill to 95% in 32 min during which 106 g of pressurant hydrogen are needed. 8. Conclusion Fig. 13. Predicted mass loss rate during long term storage.
Simulations of CPST operations including microgravity cryogenic fuel transfers and long term storage have been conducted with the use of a verified thermal/thermodynamic analysis tool. Results predict successful transfer operations even at a supply tank fill level of 15%. Correlations of similar models to test results from no-vent hydrogen transfer tests on the ground provide insight into on orbit transfers and confidence in the model results. These results should help to justify the need for a timely CPST demonstration mission and kick-start a move to using fuel depots for deep space missions. References
Fig. 14. Predicted supply and transfer tank pressures during late transfer operation.
Fig. 15. Liquid fill quantities in supply and transfer tanks during late transfer.
[1] McLean C, Mustafi S, Walls L, Pitchford B, Wollen M, Schmidt J. Simple, robust cryogenic propellant depot for near term applications. In: 2011 IEEE Aerospace Conference. Big Sky, MT, March 2011. [2] Zegler F, Kutter B. Evolving to a depot-based space transportation architecture. In: Space 2010. Anaheim, CA; September 2010 [AIAA-2010-8638]. [3] Braun RD. Investment in the Future: Overview of NASA’s Space Technology. NASA Chief Technologist. May 5, 2010. [4] McLean C, Mills G, Riesco M, Meyer M, Plachta D, Hurlbert E. Long term space storage and delivery of cryogenic propellants for exploration, In: Joint Propulsion Conference, Hartford, CT; July 2008 [AIAA-2008-4853]. [5] McLean C, et al. In-space cryogenic propellant storage and transfer (CPST) demonstration mission concept study report, In: Ball Aerospace and Technologies Corp., Boulder, CO., NASA Glenn Research Center (GRC). Contract Number NNC11CA29C. Final, Report, December 2011. [6] Tam W, Ballinger I, Jaekle Jr DE. Surface tension PMD tank for on orbit fluid transfer. Hartford, CT, July 2008 [AIAA 2008-5105]. [7] Dye S, Kopelove A, Mills GL. Load Responsive Multilayer Insulation Performance Testing. CEC, June 2013. [8] Hopkins RA, Finley PT, Schweickart RB, Volz SM. Cryogenic/thermal system for the SIRTF cryogenic telescope assembly. Proc SPIE 2003;4850:42–9. [9] Thompson RI. Near-infrared camera and multi-object spectrometer (NICMOS): the near-infrared space mission on HST. Proc SPIE 1994;2209:319–30. [10] Sterbentz WH. Liquid propellant thermal conditioning system. NASA CR72365. 1968. [11] SINDA/FLUINT User’s Manual. Version 5.5. C&R Technologies. Octeber 2011. [12] Taylor WJ, Chato DJ. Comparing the results of an analytical model of the novent fill process with no-vent fill test results for a 4.96 m3 tank. July 1992 [AIAA-92-3078]. [13] Plachta DW, Christie RJ, Carlberg E, Feller JR. Cryogenic propellant storage analyses and design tool evolved from in-space cryogenic propellant depot project. Huntsville, AL: Space Cryogenics Workshop; July 2007. [14] Chato DJ, Marchetta J, Hochstein JI, Kassemi M. Approaches to validation of models for low gravity fluid behavior. September 2005 [AIAA-2004-1150].
7.3. Low fill level transfer operation Due to the significantly different amount and location of the liquid in the supply tank from beginning to end of the mission,
Please cite this article in press as: Schweickart RB. Thermodynamic analysis of a demonstration concept for the long-duration storage and transfer of cryogenic propellants. Cryogenics (2014), http://dx.doi.org/10.1016/j.cryogenics.2014.03.004