1. INTRODUCTION

In early 2007, NASA initiated a study to examine various architectures for a human Mars mission campaign using the Constellation System launch architecture defined in NASA's 2005 Exploration Systems Architecture Study (NASA, 2005). The primary goals of this study were to update NASA's human Mars mission reference architecture, define the long term science goals and objectives for human exploration missions, and to define flight and surface systems for human missions and supporting infrastructure. The study focused on definition of operational concepts, identifying key challenges including risk and cost drivers, examination of development schedule options, and key trade studies for future analysis. A major element of this human Mars campaign study included the examination of various options for delivering crew and cargo to Mars orbit, and subsequently to the surface of Mars; specifically, the Entry, Descent, and Landing (EDL) challenges, technology requirements, and the system level impacts that the EDL architecture imparts on the overall human Mars mission architecture. The study concluded in late 2008 with results that subsequently became known as NASA's human Mars Design Reference Architecture 5 (DRA5) (Drake, 2009, 2010).

Prior to this effort, and over the course of the last several decades, there have been a multitude of human Mars architecture proposals, studies and reports. An excellent historical review of the various architectures is provided in (Griffin, 2004), and a description of the NASA Design Reference Mission (DRM) for a Human Mars system that was used to provide guidance and requirements to the 2005 Exploration Systems Architecture Study is available in Drake (1998). Another excellent overview and survey examination of the challenges and potential solutions to the large-scale EDL issues at Mars may be found in Manning and Braun (2004).

NASA investments in fundamental aeronautics and atmospheric flight science technologies in the 1960's and '70's enabled our current EDL capabilities of today. For missions with atmospheric flight elements requiring entry or reentry into an atmosphere, the EDL system drives the mission architecture (site accessibility, location, elevation, landed mass, etc). All of NASA's six successful Mars landers (Viking I & II, Pathfinder, MER – Spirit and Opportunity, and Phoenix) have relied largely on the Viking EDL heritage technology set developed four decades ago. If the Agency wishes to move beyond small robotic missions and the less than 1 t landed payload mass capability at Mars that will be demonstrated on MSL in 2012, sustained and coordinated investments in new atmospheric flight system technologies must be made due to the fact that current entry technologies cannot successfully and accurately slow heavier mass entry bodies to acceptable landing speeds by the time they reach the Martian surface.

In the present study, a set of ground rules and assumptions were developed to guide the overall human Mars architecture study. Several of the key ground rules and assumptions which defined the reference system architecture, are listed here:

2. HISTORICAL SIZING FOR MARS AEROCAPTURE AND EDL DESIGN

Human Mars missions will require transport of crew and very large masses and volumes of cargo to the surface of Mars. Human Mars mission studies in the late 1990's and early 2000's (Drake, 1998; Griffin et al. 2004,) were based on electric propulsion into a high energy Earth orbit followed by a trans-Mars injection using chemical propulsion. These studies hinted at the large mass payloads that would be required for Mars aerocapture and entry, on the order of 50 to 100 t. Further study efforts on Human Mars missions abated early in the 2000's, but NASA's recent exploration initiatives with human Mars missions being considered for the 2030 decade and beyond served to renew interest in this topic, and benefits from a review of the previous studies. In addition, comparison to other human and robotic missions provides insight to the trends and "rules of thumb" inherent in space exploration vehicle packaging densities, ballistic numbers, and mass ratios.

To ascertain a feasible and reasonable range of spacecraft and delivery system packaging densities, a number of historical spacecraft were reviewed. Planetary probe data from previous and current Mars robotic missions were examined, in addition to a wealth of mass properties data from human spacecraft designs (Heineman, 1994). In Figure 1 below, the packaging density versus outer mold line volume for a number of historical spacecraft, including the Apollo, Gemini, and Mercury entry capsules, the Space Shuttle Orbiter without the main engines included, and the Apollo Service Module. Also included are the Mars Viking entry vehicle, and the planned 2011 Mars Science Laboratory robotic entry vehicle, the later being designed using a 4.6 m aeroshell, similar in scale to the previous human spacecraft.

For an entry vehicle of size about 15m3, approximately Apollo command module volume, or smaller, it can be seen that a reasonable upper limit on packaging density is approximately 350 kg/m3. For much larger vehicles than the Apollo command module, such as the human Mars mission elements, an extrapolation of the packaging density curves based on empirical data can be made. The extrapolation is a power curvefit for packaging density versus OML volume, similar to the methodology used in Heineman (1994). Use of the historical human vehicles, along with the Mars Science Laboratory entry vehicle, provides a curvefit as:

This line is shown in the plot of Figure 1 and is extrapolated to the estimated volumes of human Mars entry vehicles, which are estimated to be of order 1000m³. It indicates that large human scale entry vehicles may be reasonably packaged at densities of up to 200 kg/m³. The Shuttle Orbiter point is much lower than this on the plot for two reasons: the mass of the Shuttle Orbiter main engines, which are used for ascent, were not included here, and the fact that the Shuttle uses a light weight reusable TPS as a result of the fact that it only needs to re-enter the atmosphere from low Earth orbit at moderate velocities. The Viking entry vehicle densities were similarly low, as a result of low TPS mass requirements due to the fact that the Viking entries were also from low Mars orbit at relatively low velocities (3.5 km/s).

A significant design parameter for atmospheric entry vehicles is the ballistic number, defined as: where m is the entry vehicle mass, C_D is the aerodynamic drag coefficient, and A is an aerodynamic reference area. If the packaging density were allowed to stay consistent with previous spacecraft data, i.e. along the trend line, then the ballistic number of such a large vehicle would naturally tend to be much larger than can be feasibly flown with fixed, rigid aeroshell entry technology (Manning and Braun, 2004). The problem lies in the fact that the density would allow the mass to increase with vehicle characteristic length generally as a cube, while this would drive the ballistic number exceedingly high since the area would only increase as a square of the vehicle dimension.

Further investigation of multiple previous design studies pertaining to the aerocapture, entry, descent, and landing at Mars, as well as other conceptual and flown spacecraft, does provide some beneficial insights for the human Mars mission aerocapture and EDL design trade space. For the large volumes and masses expected for the human Mars mission, packaging density may not be the limiting factor for rigid aeroshell concepts since the ballistic number requirements for aerocapture and EDL will define the limiting constraints. Further discussion on ballistic number scaling and details on sensitivities will be provided in the following sections. For extremely large-scale vehicles, packaging density will likely be much less than that of the small scale robotic missions or even human lunar spacecraft. Packaging will be dictated by c.g. limitations, and also be driven by the constraints of the launch vehicle shroud, particularly if rigid aeroshells are employed. Ballistic numbers for human Mars Aerocapture and EDL vehicles will be larger than those flown on any Mars mission to date, assuming the rigid aeroshell approach. The primary reason for this is the necessity to package into slender bodies or behind small brake diameters due to limitations of the launch vehicle shrouds. The other reason is the natural tendency for the ballistic number to grow as the scale of the vehicle increases, as explained earlier.

3. MARS ORBIT INSERTION: AEROCAPTURE VS. PROPULSIVE TRADE STUDY

In the development of the architecture for this study, all elements (crewed and cargo) are assumed to be first inserted into a Mars orbit for operational and safety reasons prior to entry, descent, and landing, as opposed to a direct entry from the Earth-Mars cruise phase. In order to place a spacecraft in orbit around a planetary body, sufficient velocity must be removed such that the gravitational field of the target body will transform the approach hyperbolic trajectory into a closed elliptical orbit. Traditionally this has been accomplished using chemical propulsion to provide deceleration forces to slow the spacecraft to the required velocity for orbit capture. For planetary bodies that possess an atmosphere, including Mars, using atmospheric drag to provide aerodynamic deceleration can result in significant mass savings over the more traditional propulsive orbit insertion methods.

Aerocapture is a method to directly capture into a planet's orbit from a hyperbolic arrival trajectory using single, atmospheric aerodynamic drag pass, thereby reducing the propellant required for orbit insertion. Over the last several decades, multiple aerocapture systems analysis studies have been conducted for multiple planetary destinations (Earth, Mars, Venus, Titan, Neptune), with a variety of aerodynamic shapes and guidance algorithms, and have all concluded that aerocapture is a moderate to relatively low risk technology (Cerimele and Gamble, 1985; Lockwood, 2004; Hall et. al, 2005; Lockwood, 2005; Wright, et. al., 2006). However, these studies were typically limited to significantly smaller payloads associated with robotic missions. This effort attempted to address aerocapture performance for the much larger 50-100 t payloads required for human class missions. The aerocapture technique requires an aeroshell with sufficient thermal protection system (TPS) to protect the payload from the aerodynamic heating encountered during the atmospheric pass. During the aeropass maneuver, an atmospheric flight guidance and control algorithm is utilized to target the trajectory to a specified condition following atmospheric exit; then an orbit periapsis raise maneuver is executed to achieve the target orbit conditions, as shown in the aerocapture flight profile schematic in Figure 2.

Although the aerocapture technique has not yet been demonstrated on an operational mission, studies have demonstrated the feasibility, and identified potential savings in the propellant and overall system mass required for orbit insertion. The primary aerocapture technology challenges are the TPS, sufficient knowledge of the atmospheric density profiles, and aerocapture guidance and control algorithms. TPS challenges are thought to be no more demanding than direct entry TPS, but are obviously configuration specific to the aeroshell geometry, and the maximum heat rate and heat pulse duration. Given that the Mars EDL system will already require a hypersonic entry aeroshell, the system can be easily modified to also serve as the aerocapture aeroshell by adding the appropriate TPS to that which already exists for the hypersonic entry phase of the mission. In fact this, and a small amount of additional propellant for the post aerocapture periapsis raise and orbit trim burn, are the only major hardware additions to the system to enable an aerocapture maneuver. The question then is what are these additional mass requirements, relative to the propellant mass requirements for the propulsive orbit capture?

In order to determine the potential mass savings, system level trades were conducted using aerocapture and both chemical and nuclear thermal rocket (NTR) propulsion options for Mars orbit insertions. First, it was important to understand the sensitivities of aerocapture performance to the possible variations in vehicle design for this mission. The key parameters of interest were ballistic number, aeroshell aerodynamic lift-to-drag ratio, target orbit (500 km circular orbit, or a 1 Mars sol period orbit), and the atmospheric entry velocity of the arriving vehicle. Early in the study, the specific EDL system definitions were not yet fully defined, so in order to investigate the scope of the problem, several initial conservative assumptions were made. The desired useful landed payload mass at the surface of Mars was assumed to be between 20 t and 80 t. These values were used to derive an entry mass using a set range of "gear ratios" obtained from historical studies previously cited (Drake, et. al., 1998; Griffin, 2004; Manning and Braun, 2006). A lower bound on the useful landed mass/entry mass gear ratio equal to 0.5 was selected based on early human Mars mission studies, along with an upper bound on this gear ratio term equal to 0.64, obtained from the more recent design studies. The entry mass from orbit thus spanned a range from 31 t to 160 t. The orbit mass was then estimated by assuming an approximate ΔV = 110 m/s requirement for a de-orbit maneuver. Using an I_sp range of 330s to 450s resulted in an orbit mass of 32 t to 166 t. The estimated aerocapture mass was then computed by adding the propellant mass required to perform the post-aerocapture circularization burn of approximately 150 m/s. Assuming the same I_sp range as above, the resultant aerocapture mass range was 33 t to 174 t.

Based on the assumption of no EVA on-orbit assembly of components, aeroshell dimensions were assumed constrained to the estimated capability of the launch vehicle. The reference Ares-V launch vehicle payload shroud provided accommodation for a 7.5 m diameter and 12 m long vehicle. Potential growth was estimated to 12 m in diameter and 35 m in length. In order to limit the scope of the study, initial geometric assumptions were based on an ellipsled configuration for the aeroshell (Edquist, et al. 2004), however, given the same dimensions small modifications to the trim angle-of-attack could be made to achieve similar ballistic numbers and lift-to-drag ratios for other shapes including biconics and triconic class of slender body mid L/D designs. Given the pre-aerocapture total vehicle mass range of 33 t to 174 t and the range of the vehicle sizes from 7.5m x 12m to 12 m x 35 m, the vehicle's packing density was determined, to compare with the results of previous configurations. A comparison of the aeroshell range estimates to those of several historical manned and robotic vehicles can be seen in Figure 3. The three red x's on the upper and lower boundaries of the blue region represent three aeroshells that were examined in further detail, the 7.5m x 12m, 10m x 25m, and 12m x 35m. The upper limit (blue line) is determined by the packaging density of the aeroshell of 174 t mass, while the lower limit is determined by the packaging density of the aeroshell of 33 t.

To understand the benefits and consequences of performing aerocapture at various L/Ds and ballistic numbers, a parametric aerocapture assessment assuming optimal apoapsis targeting performance was employed. This analysis indicated that an open-loop guidance assumption was valid, with a noted variability in minimum altitude of approximately ±2 to 3 km. The peak heat rate and total integrated heat load sensitivities were also assessed as a function of the aeroshell L/D and ballistic number for the aerocapture maneuver. These data indicate the fact that for moderate ballistic numbers (400 to 1,000 kg/m²), the peak heating and total heat loads are well bounded by the TPS performance capabilities that are being developed for the Orion CEV lunar return conditions. An aerocapture point design solution was developed for the reference 40-t useful payload on the surface of Mars. Corresponding to this was an EDL system design that resulted in a Mars arrival mass of approximately 115 t. Monte Carlo simulations were used to conduct capability verification and the results indicate that there was adequate targeting performance for this vehicle despite the moderate ballistic number (approximately 500 kg/m²) and the high-energy orbit (exit velocity representing 97% of escape velocity), which increases the difficulty of targeting the desired orbit. The overall performance of the aerocapture maneuver can be measured in terms of the postaerocapture circularization burn requirements, which indicates that the mean delta-V required is only 19 m/s, with a maximum 66 m/s case. Based on the results of the parametric and Monte Carlo performance assessments, the aerocapture maneuver was deemed to be a feasible option for the large-scale, high-mass systems that are consistent with the human-class mission set.

Both propulsive and aerocapture Mars orbit insertion were examined and compared for both the NTR and conventional chemical propulsion approaches. The aerocapture cases required slightly more mass in Mars orbit (4 t) due in large part to the additional TPS that is required to execute the aerocapture maneuver, which would then be reused during the EDL phase. Detailed TPS sizing assessments were performed for both the aerocapture followed by entry and the entry alone options to validate these data. However, the masses were small in comparison to the additional NTR and chemical propellant masses that are required to execute the all-propulsive burns for orbit insertion. Ultimately, the performance metric that was used for direct comparison was determined to be the initial mass in low Earth orbit (IMLEO). The missions that use aerocapture achieve a significant savings in IMLEO requirements. The analysis indicated a mass savings of 350 t IMLEO when using aerocapture as compared to chemical MOI, and a savings of approximately 87 t IMLEO for aerocapture vs. propulsive capture for the NTR case. Risk and cost assessments were also conducted for the aerocapture vs. propulsive MOI options, and further details are provided by Drake, et al. (2009; Drake 2010). The use of aerocapture is felt to be a relatively small incremental cost to the larger, more challenging EDL system development costs and risks. The major engineering challenges and technology risk reduction efforts that are required for EDL system development will also serve to retire many of the risks that are associated with aerocapture technology. Some incremental technology development and risk reduction efforts will be required, but these are felt to be moderate and easily manageable.

The results of the aerocapture design trades for large aeroshell, high mass systems yielded an appropriate measure of confidence such that the DRA5 architecture adopted the use of aerocapture for the cargo elements of the mission. The large volumetric requirements of the crewed mission elements, and the assumption that on-orbit assembly of large aeroshell structures to envelop the associated elements were precluded from consideration, resulted in a design that required propulsive MOI for crew assets. Future architecture studies may re-examine the on-orbit assembly options or crew size architecture trades and perhaps realize significant additional mass savings through the use of aerocapture in more elements of the overall architecture.

4. ENTRY, DESCENT, AND LANDING SYSTEM PARAMETRIC STUDY

Several EDL configuration architectures were considered for this study. They included an all propulsive entry with no aeroassist elements. Ultimately, this option was not selected because of the large orbit-to-landed payload mass fraction (on the order of 8) required for payload masses considered. Supersonic aerodynamic decelerators, including parachutes and inflatable aerodynamic devices were also considered for use in the descent phase, but the performance and mass models for the scale and dimensions required for the systems in this study were felt to be lacking in sufficient detail to be considered here. The extrapolation in performance and masses from the references available were too large for these technologies to be weighed as viable options in the trade space. However, it was strongly recommended that future development of improved models for these types of systems technologies be pursued, so that credible trades can be conducted and more optimal EDL system performance and reliability improvements realized. More discussion on this topic will be provided in the final section of this paper. The reference EDL architecture ultimately selected for this study was a hypersonic aeroassisted entry system, with a dual use mid-L/D aeroshell, that also serves as the launch vehicle shroud, ejected at low supersonic Mach numbers. A liquid oxygen and methane (LOX/CH₄) fueled propulsion system was then used for final terminal descent to the surface. Conceptually, this EDL system design is shown in Fig 4.

Similar to the aerocapture parametric trade space defined in the previous section, an EDL parametric trade space was defined to bound the estimated effective payload landed masses and architecture design options. The baseline EDL system design was developed using a 10 m x 30 m aeroshell and a reference Mars orbit with a 1 sol period and apoapsis of 33,793 km. EDL system designs were developed for both the aerocapture and propulsive Mars Orbit Insertion (MOI) from cruise cases because the capture method affects the aeroshell TPS mass. Systems were sized for three landed payload masses of 20, 40, and 70 t, which would cover the assumed range of Mars surface systems required for the human Mars architecture. In the case where aerocapture was used to achieve Mars orbit, the same aeroshell was used for both the aerocapture and EDL phase, with the appropriate additional TPS mass required to accommodate the additional heating associated with the aerocapture maneuver. A pseudo guidance methodology was developed to provide a realistic entry profile that would minimize terminal descent propulsive fuel requirements as well as the TPS mass and land the vehicle at 0 km above MOLA. A summary of the EDL system parametric space is provided in Table 1. What follows is a description of the entry aerothermal environment, TPS sizing, mass models and parametric studies that contributed to the selection of the baseline configuration.

The key to determining the EDL system mass of a human scale mission to Mars are the vehicle components considered and the fidelity of the models used to approximate the mass of each. The major components of the entry vehicle considered here include the aeroshell structure, TPS, the reaction control system (RCS), the descent stage structure, terminal propulsion system including propellant tankage and plumbing, and the useful landed payload. In all cases, a 30% margin allocation was assumed on structural mass, and a 20% margin on TPS masses (based on Orion/CEV "heritage" material and design principals). A 30% margin on terminal descent propellant was levied to account for terminal descent hazard avoidance/pinpoint landing. The entry trajectories were simulated using the Program to Optimize Simulated Trajectories (POST2) (Striepe, et al. 2004).

The aerocapture and entry aeroshell structure mass estimates were made using preliminary estimates and guidance from the Ares V launch vehicle shroud development efforts (Feldman et al., 2010). Detailed structural sizing and additional load cases were defined which resulted in estimates of the aeroshell mass of 22.5 t for the 10 m x 30 m aeroshell; mass estimates for a similar geometrically scaled 12 m x 35 m aeroshell configuration were estimated at 34.5 t.

The aerothermal environments associated with the Mars aerocapture and entry trajectories were determined using the NASA CBAERO tool (Kinney et al. 2006), modified for use in the Mars atmosphere. The database for the CBAERO code was developed from a sparse set of high fidelity, real-gas computational fluid dynamic (CFD) solutions from the DPLR code (Wright, 2005) combined with the line-by-line radiative heating code NEQAIR (Park, 1985) to provide predictions of convective and radiative heating solutions. Each solution contained the full surface aerothermal environments including surface pressure, temperature, shear and uncoupled, convective and radiative heating. While these codes represent the current state-of-the art, much uncertainty remains in understanding of large scale entry vehicle environments for Mars. In order to account for uncertainties in aerothermal environments, margins were adopted using the following factors:

Descent Stage The descent stage engines were assumed from previous large lander studies to be RL10 derivatives and assumed a thrust to weight ratio of the engines to be 40 lbf/lbm(Earth). Recognizing that the LOX/LH2 RL10 may not be the best analog for the LOX/CH4 engines currently baselined in this architecture, the parametric space was expanded to include engines based on an RD-180 derivative which has a thrust to mass ratio of 80 lbf/lbm. The mass of the engines used in the thrust to mass ratio includes all associated turbopumps and all hardware attached to the engine before installation, but did not include the pressurant or tank-to-engine transfer line masses.

The descent stage dry mass was based on mass characteristics modeled using the Envision mass sizing and simulation program (Smith, 2009). The descent stage is an all-propulsive, legged lander concept utilizing 4 pumpfed LOX/methane engines with the following reference characteristics: specific impulse of 369 sec, engine O/F of 3.5, chamber pressure of 600 psi, and a nozzle area ratio of 200. The baseline vehicle was sized to conform to a 10.0m diameter aeroshell. The descent stage thrust structure was assumed to undergo maximum loading during the descent maneuver and is sized to withstand the user-defined system T/W without the aeroshell attached as payload, assuming the aeroshell was deployed prior to terminal descent engine initiation. In addition, the tanks of the descent stage were sized to include the deorbit fuel.

Mathematical models for all masses were incorporated into the POST2 simulation. The POST2 optimization capability was utilized to converge the parameters used to generate the mass models with the optimal trajectory determined values.

The aeroshell parameter range included both the 10 x 30m and 12 x 35m aeroshell to evaluate the effect of both packing density and ballistic number. The 10 x 30m aeroshell was selected as a baseline because it corresponded to the reference Ares V launch shroud and required less TPS and structure mass and provided packing densities that were consistent with historical data for human missions as described previously. Entries from the 500 km circular and 1 sol elliptic orbits were considered; entry from a 1 sol period orbit was ultimately selected as the baseline, driven by the overall system architecture optimization (Drake 2010; Drake et al., 2009) Results for the baseline entry configuration, which consisted of the 40 t payload in a 10 m x 30 m aeroshell, and entry from a 1 sol period orbit are presented in Table 2 below. In this case, the trajectory pull-out maneuver was initiated at 2 g's, then a constant 2 g profile was maintained for 134 sec, rolled to full lift up and maintained full lift up for 168 seconds; engines were initiated at 1.3 km altitude, and held a constant 3 g's for 20 seconds until the velocity was reduced to 2.5 m/s, and then maintained until the vehicle reached 0 km MOLA. A 30% ΔV margin was allocated for precision landing and terminal descent hazard avoidance. The mass results for both the baseline aerocapture and propulsive Mars orbit capture cases are also shown in Table 2.

Additional trajectory shaping and optimization efforts were made to determine if more efficient trajectories, and thus lower mass entry systems, could be found. Various trajectory pull out g-load levels were considered in the EDL parameter trade space. The trajectories for each payload size were designed to hold a 1, 2, and 3 g constant profile upon entry. As previously described, a significant entry mass savings was realized using constant g profile over the constant bank angle profile; additional mass savings was seen with increased g loading, such that a mass optimized trajectory flown to hold a constant 3 g's during entry would have a lower entry mass than a trajectory optimized to hold 1g. However the mass saving was not found to be linear with g load and payload, and dramatically increased from 20 t to 70 t payloads (i.e. going from 1 g to a 2 g trajectory for the 20 t case only resulted in a ~400 kg mass reduction where a 70 t payload case of similar trajectory type yielded an entry mass reduction of ~10000 kg). Since there was not a significant difference between the entry mass saved going from the 2g trajectory to the 3g trajectory, the 2 g case was baselined for the study. The 2 g case provides more timeline margin, and would be more tolerant to atmospheric dust loading and seasonal variation in atmospheric density profiles. Thrust to weight of the engines was also considered as a parameter in the design space when it was recognized that previous large lander studies had assumed RL10 derivatives with an engine thrust to weight ratio of 40 lbf/lb_m. Acknowledging that a human Mars mission would likely benefit from and ultimately require more efficient engines, the parametric analysis also considered RD-180 derivatives engine with a thrust to weight ratio of 80 lbf/lb_m. The results of the parametric study indicate an approximate 6% entry mass savings for the 40 t entry mass using the larger thrust to weight. Results of the parametric trades are shown in Table 3 below, with details on system and component masses for 20, 40 and 70 t payloads (assuming no terminal descent ΔV margins). The 70 t entry mass had a ballistic coefficient that was too large to maintain a 3g entry profile, which resulted in an unconverged trajectory (i.e. the vehicle impacted the surface prior decelerating to appropriate engine initiation conditions).

Ultimately the final EDL system design selected for the DRA5 study was that which delivered a 40t payload to the surface from a 1-sol orbit, and utilized the 10 x 30m aeroshell, constrained to a 2g entry profile, with the advanced engine T/W of 80 lbf/lbm, with details and subsystem mass breakdowns as indicated in Table 3.

As part the DRA5 study close out activities, there were a number of high-level observations and recommended actions related to enabling eventual human Mars EDL capabilities. Landing large payloads on the surface of Mars, greater than our current capability of 1 t that will be demonstrated with MSL in 2012, remains a key challenge. Despite the significant efforts and detail that went into this particular DRA5 EDL system design, it was still only at the conceptual level, and much work remains to determine the true viability of this or other EDL architectures at the human mission scale. It was noted that the first use of a human scale EDL system was a key risk driver in the overall human Mars mission architecture, and that scalable partial or near-full scale robotic precursors would help retire risk. The question remains, how do we demonstrate a capability to reliably decelerate the human scale entry systems with associated high ballistic coefficient vehicles, and bring them to a safe landing on the surface? NASA clearly needs a long term sustained Mars EDL risk mitigation strategy, including robotic mission demonstration and use of EDL systems which are scalable/near-full scale to human mission needs. It was further noted that research and system studies of fundamental EDL technologies should be pursued.

5. ADVANCED MARS ENTRY, DESCENT, AND LANDING ARCHITECTURE AND TECHNOLOGY SYSTEMS ANALYSIS STUDIES

To further the insight on desirable technologies to engage, following the completion of the DRA5 study in late 2008, NASA stood up a small cross Agency team to look at advanced EDL system architectures and enabling technologies. As noted previously, a human mission capable EDL system is major system performance and mission risk driver, and will require long lead technology development and subsystem qualification in the overall human Mars system architecture. Not only does the EDL architecture define mass limitations to the surface and the site accessibility capability (elevations, surface hazards and slopes, etc.), the mass sensitivities and penalties associated with getting down into and out of the Mars gravity well and atmosphere are large and require detailed designs and understanding in order to optimize the overall system architecture.

NASA's EDL Systems Analysis (EDL-SA) project began in the fall of 2008 to advance the EDL capabilities needed to safely place large mass payloads on the surface of Mars, including both human and robotic missions. The NASA Chief Engineer chartered this team, with support from multiple Mission Directorates, to provide an integrated systems analysis capability with an appropriate level of detail in the multiple disciplines associated with the Mars EDL challenges. The team was tasked with identifying and defining a set of EDL architectures and technology options, along with the attendant technology development roadmaps and schedules, to enable eventual human Mars exploration in the 2030 decade.

Based on previous experience (Cruz et al. 2005), need for appropriate analysis fidelity, and complexity of the problems, the integrated EDL systems analysis activities were planned be conducted over a three-year period. The first phase of the project was designed to identify various architectures and technology sets technologies for the 20- 50 t surface payload capability required for a human exploration-class mission. The second phase of the project would determine the minimum technologies required to land a 1-5 t spacecraft on the surface of Mars enabling post- MSL robotic missions. The spacecraft EDL system would have no requirement for technology feed-forward to the human exploration class mission requirements. The third phase of the project will seek to identify those EDL technologies that could be used for post-MSL missions that would provide significant feed forward to the human exploration-class missions. Finally, for the technologies identified, roadmaps and detailed technology development plans will be developed to mature each technology to a Technology Readiness Level (TRL) of 6. As of the fall of 2010, the EDL Systems Analysis Project team had completed the first phase of the planned activities, and were executing the second phase of the project, focused on the small robotic mission EDL technologies. The preliminary results of the efforts have been reported in various references (Dwyer-Cianciolo et al. 2010; Shidner et al., 2010; Davis et al., 2010; Zumwalt et al., 2010; Zang et al., 2010). The project team identified a set of eight candidate EDL architectures for the human exploration class mission, using various technologies ranging from the baseline mid-L/D rigid aeroshell and supersonic retro-propulsion terminal descent stages as previously described in the DRA5 EDL architecture, to architectures that included more novel technology sets including hypersonic and supersonic deployable aerodynamic decelerators and dual use TPS options for rigid aeroshells. The eight different architecture options are shown graphically in Figure 5, along with estimates of their respective Mars arrival masses.

As part of the study, significant efforts were made to mature the element models for the various EDL technologies. The reference DRA5 EDL architecture, which was the baseline architecture 1 option in the EDL-SA study, was refined and additional detail was developed in the component models and trajectory modeling. Detailed component mass models and systems sizing estimates were also developed for the alternate candidate technology systems. The initial results of the study indicated that significant mass savings could be achieved by using deployable aerodynamic decelerator technologies over the rigid aeroshell baseline. These deployable systems may also allow for payloads whose volumes and dimensions exceed the constraints of the standard entry shroud options and certainly warrant further detailed investigations. Clearly, these are not the only technology options available, and others are currently being studied and developed. Regardless, the initial indications and results from this study have been promising, and are helping guide NASA's current investments in advanced EDL component technology to one day enable future human exploration of the planet's surface.

6. SUMMARY

Eventual human exploration and presence on the surface of Mars will require long term and sustained investments in many new technologies and engineering developments. Certainly among these large development efforts will be those that provide the new EDL systems to ultimately deliver the large cargo and crew to the surface, which will be unlike anything we have used to date in our relatively small (<1 t) robotic science payloads. The Mars DRA5 study highlighted the criticality of the EDL systems, not only in terms of system level performance requirements (approximately 50 t payloads to the surface of Mars), but also in terms of the elements of overall mission risk and architectural level design drivers. The follow-on EDL-SA project, continues to make progress in identifying and defining new and innovative EDL technology options for large mass delivery to the surface of Mars. If our collective desire is to move beyond small robotic missions and the less than 1 t landed payload mass capability at Mars, new atmospheric flight system capabilities must be defined and developed, which will form the basis of any EDL system and overall mission architecture design. The systems that will be used to land tens of tons of payload on Mars are not available or even known today, and cannot be defined through paper studies alone. The development and qualification of these systems will require multiple cycles of design, test, evaluation and flight test. This will require a long term commitment and sustained development effort, likely over multiple generations. Nevertheless, these are merely engineering challenges before us, and nothing to date has indicated that the capabilities to meet these human exploration EDL challenges are not physically realizable.