1.0 Introduction

During the 1980's there was much talk in the US, and around the world, about the prospects of exciting large scale manned space exploration missions. During that time the US was in an economic boom and it seemed very likely that we would soon have a manned orbiting space station, that we would establish a permanently manned lunar base, and that we could even send a crew to Mars and return them safely to Earth. Now, in the 1990's, in the early stages of recovery from and economic recession, we see that we can no longer afford so many large scale missions all at one time. We have learned from the Space Shuttle and Space Station Freedom that such large scale programs are slowed down, redesigned and possibly even canceled because of the extremely large amount of funds necessary over many years. By Congress reassessing the programs' budget every year the projects drag on for years, losing public support and gaining scrutiny, while draining funds from smaller less visible projects.

As a result, in an effort to reduce costs while advancing technology and building a large success rate, NASA has been directed to carry out much smaller individual faster-paced missions. Thus instead of developing many technologies under one large program, like those necessary for a manned Mars mission, much smaller programs with identifiable affordable price tags and relatively short timelines should be carried out. It is not suggested that we abandon the ideas of large scale nature, just that we abandon the idea of developing them all at the same time under one program. It is too late to apply this method to the space station since its fate will be determined shortly. However, if the US is determined to expand the presence of man in space we should plan now to refocus our efforts on how to attain our other major goals: a manned lunar base, and a manned mission to Mars.

Many technologies necessary to complete these missions efficiently and safely need much further development. These include: long-term micro-gravity effects on humans, time lag in communications, and telerobotics, just to name a few. These technologies can be advanced individually under smaller programs with better defined objectives. By developing the technologies in single steps we can more quickly reap the benefits while working toward an ultimate goal. This paper describes a mission that takes the first step in furthering manned exploration.

2.0 Lunar Robotics Testbed

2.1 Background

It has been determined that for any of the lunar base concepts proposed, robots will somehow be employed in the construction, maintenance, repair, mobility, or environmental protection of the base. The logic in this assumption is simple: the more that can be accomplished without manned presence, the cheaper the mission becomes. For example, if a robot can perform functions, such as filling bags of lunar soil for radiation protection or repair a solar array, the astronaut is not needed for these tasks thus reducing the cost of the mission by orders of magnitude. It would not be necessary to factor in special space suits, life support systems, astronaut fatigue, etc. Thus if a base was completely established and maintained roboticly the astronauts time could be used more efficiently for scientific endeavors.

The major problem with this scenario is that robotics technology has not yet advanced to the level required for such tasks.

The major function of robots today is in automation; performing tedious or repetitive tasks with a high level of precision. These tasks typically include spot welding or spray painting, as used in the manufacturing of cars, and are controlled by programming. Autonomous robots perform the same task over and over and do not allow for interactive control by humans. Although automation will certainly have a place in the establishment of a lunar base we would also like to be able to instruct the robot to perform various (possibly one-time only) tasks.

This sort of control is commonly known as telerobotics, where the device extends human senses or dexterity to a remote location. A human operator sees, feels and controls the remote task through the teleoperator thus eliminating the need of human presence at the remote location. This two-way communication merges the benefits of mechanization and human intelligence [1]. Due to the three second time lag in communications between the Earth and the Moon total reliance on telerobotics becomes impractical. Therefore for lunar applications we would like to merge these two types for optimum performance; a telerobot which would have some level of autonomy to perform various preprogrammed tasks (such as filling bags with lunar soil), as well as the flexibility to perform more detailed tasks that require human supervision (like replacing or repairing a solar array.) This leads to the definition of the proposed mission.

2.2 Proposed Mission Concept

In the spirit of advancing technology under small scale, relatively cheaper and faster programs, a series of lunar robotic missions is proposed. The overall goal of these missions is the advancement of telerobotics technology. More specific goals are to:

test precision of telerobotics versus pure automation

test various methods of control (possibly use virtual reality)

test durability of space robots in harsh space environment

gain experience in telerobotics technology (especially in space)

The concept is to send a series of relatively small payloads (200 - 300 lbs) to the lunar surface via a current launch system. Each payload would consist of two identical micro-robots attached to a common platform. The platform supplies the rockets and propellant used for lunar transfer and landing, as well as the major communications link to the earth. Shortly after landing the robots would detach from the platform and begin their tasks. The robots would be required to perform identical tasks; this provides researchers with twice as much test data as well as a redundancy should one of the robots become incapacitated.

Some future expectations on telerobots for lunar base activities are classified by the following general tasks [2]:

soil excavation/movement for site preparation

module/node/tunnel/garage transportation, handling and interconnection

structural assemble and disassembly

module protection and materials processing

remote repair and maintenance of nuclear power plants

For some of these tasks autonomous systems will be sufficient, but total reliance on autonomy is impractical due to the difficulty in providing the protocols and procedures necessary to handle all abnormal situations. Similarly the robots cannot rely on pure telerobotics due to the three second time lag in communications; the robot could cause permanent damage during fault detection and correction procedures. Thus the robot must effectively employ the advantages of both telerobotics as well as autonomy.

2.2.1 Proposed Robot Configuration

Based on the above tasks two general robot configurations were considered: a lunar roving vehicle with a manipulator arm using wheels or tracks for its mobility, and a three legged walking robot with two manipulator arms.

For this program the first concept was abandoned for two reasons. First, a similar mission including rovers is currently in the planning stages. This program, called Artemis (after the Greek Goddess of the Moon), would employ rovers that would conduct one-time exploration of a scientifically interesting or operationally challenging site, or the detailed reconnaissance of a potential lunar outpost site [3]. The goal of these missions however is not to advance robotics technology but to re-establish the expertise needed to conduct human missions beyond low-Earth orbit.

Secondly the focus of the robotics testbed is advancing robotics precision. A rover requires much less precision for its mobility than a walker. Additionally a three legged configuration has the advantages of: being able to squat down to pick up cargo, tilt for precision aiming (as for a drill bit), stabilize for digging operations, and extend its legs to provide a stable crane platform. Wheeled vehicles provide "high-speed transportation over smooth lunar surfaces, but the robot's ability to walk in any direction and turn about any point gives it the nod for construction work on rough, sloping, and boulder strewn surfaces." [4]

2.2.2 Task Definition

It is proposed that both robots will carry out the same or similar tasks so that performance reliability of the design can be measured. It could be instructive to have the robots carry out the same task employing three different control methods; purely automated, purely telerobotic and a combination of both. By completion of the same tasks using these different methods the level of precision could be compared. It is expected that pure automation will have higher precision thus this will be used to gauge the other methods. Along with precision, the time to carry out the task will also vary. It is expected that most of the test materials will be contained within the robot as opposed to relying on lunar surface materials (such as rocks, old landers from Apollo etc.).

Specific tasks have not yet been determined but they can be classified under general functions that require advancement of current technology. These functions might include:

precision pointing of the manipulator; possibly putting pegs in a

pegboard, using a screwdriver or drill, etc.

precision gripping of the manipulator; determining the amount of pressure

needed to pick up an object; as in the difference between handling an egg or

a rock.

precision pointing of the mainpulator and legged platform in concert; for

example it might be necessary to level the platform on slopped terrain while

using the manipulator for various functions.

filling bags with lunar soil and placing in a specified pattern to simulate

radiation protection procedures

2.3 Mission Justification

It is widely accepted that in order to re-establish manned presence on the Moon, permanently, or to venture further out to Mars, advanced robotics technology is essential. These advancements could significantly reduce the number of EVA's in future manned space exploration; thereby reducing risk to the astronauts and overall cost of the missions. Specifically advancements in telerobotics are of the most importance. This however is not a technology whose use is limited to space applications only. Telerobotics presently has widespread uses in private industry, research and in the home. Some of the more visible applications are the Jason Junior deep sea explorer and the fiber-optic cable giving doctors an inside view of the human body. Other application of telerobotics on Earth could include robots that: disarm bombs, fight fires in places too dangerous for humans an aid the handicapped in chores around the office or at home. These advancements could become reality much quicker through these small scale space missions.

3.0 Mission Analysis

A preliminary mission analysis was carried out to determine overall mission geometry, DV's required, propellant mass, etc. It was proposed to send 200-300 lb of payload to the lunar surface (includes platform and both robots). This was to be accomplished using a current launch system. Due to its relatively low cost and its flexibility in launch conditions, the Orbital Sciences' Pegasus launch vehicle would be an ideal selection. The following analysis compares two launch scenarios and determines the launcher and upper stage necessary to complete this mission.

3.1 Mission Scenario

Two general launch scenarios were considered. The first was to launch and circularize in 300 km altitude orbit, and the second was to launch and circularize at half-GEO altitude before translunar insertion. For each of the respective insertion altitudes delta-V calculations were obtained for translunar insertion, lunar orbit insertion and lunar landing. For the translunar insertion velocity and lunar encounter velocity calculations a Lambert targeting routine was used. Maximum time of flight was assumed for each insertion altitude and the encounter radius was set at the sphere of influence of the Moon. After the five day translunar coast and encounter at the sphere of influence, the spacecraft is placed on a Hohmann ellipse to its low lunar parking orbit. It remains in orbit until lunar lighting conditions are optimum for lunar landing. It is assumed that landing would be timed to coincide with the lunar dawn so that operations could be conducted for at least the full period of one lunar day (which is equivalent to fourteen Earth days). After landing the two robots detach from the landing/communication platform and begin operations. All communication to Earth is relayed from the robot through the platform communications system to Earth. The specific burn sequences are described in the following section.

3.2 Burn Sequence

A sequence of five burns is carried out in the analysis of the Earth-to-Moon landing scenario. The first burn occurs at a specified altitude above the Earth (ro). This is the translunar Insertion burn. Burn #2 occurs at arrival at the Moon's sphere of influence. The arrival point is determined by the angle l1 measured relative to the Earth-Moon line (Figures 1 & 2). Burn 2 is used to place the spacecraft on an elliptic Hohmann transfer to a low circular lunar parking orbit. Burn #3 occurs at periapsis of the transfer ellipse and circularizes the orbit at periapsis altitude. When conditions are set for a lunar dawn landing Burn #4 initiates the deorbit trajectory. This trajectory is assumed a Hohmann to the lunar surface where Burn #5 negates the periapsis velocity for soft landing. The following parameters describe the geometry shown in Figures 1, 2, & 3. Figure 1 shows the complete generalized mission geometry, Figure 2 shows the patch conic condition at the Moon's sphere of influence and Figure 3 shows the rocket burn locations.

ro = translunar insertion (TLI) altitude

Vo = velocity necessary for TLI

D = mean Earth-Moon distance

r1 = radius from Earth center to arrival point on Moon sphere of influence

go = angle between radius vector at TLI and Earth Moon line

Y = angular distance traveled by moon during translunar coast

(TOF*wm)

g1 = angle between Earth moon line and r1

Rs = radius of Moon's sphere of influence

b1 = flight path angle with respect to Earth at Rs arrival

V1 = s/c velocity with respect the Earth at Rs arrival

b2 = flight path angle with respect to the Moon at Rs arrival

V2 = s/c velocity with respect to Moon at Rs arrival

Vm = Moon's Earth orbit velocity

r2 = lunar parking orbit radius

rm = Moon's equatorial radius

3.3 DV/Geometry Optimization Analysis

A program employing a Lambert targeting routine was used to calculate the translunar insertion and lunar encounter velocities. The program LUNLAMB.FOR used these calculations in the mission geometry optimization described below.

The following parameters were fixed for the optimization analysis:

ro, Rs, D, TOF, Y

Using the program LUNLAMB.FOR the analysis was carried out as follows:

Geometry optimization algorithm:

Given: ro, Rs, D, TOF, r2, rm, Vm

go = 20, 180 (outer loop)

l1 = -45, 45 (inner loop)

calc: g1, r1, Dq = go + Y + g1

Call Lambert Targeting (ro, r1, TOF, DQ)

returns==> Vo, bo, V1, b1

From Geonetry at patch condition

calc : DVlht = lunar Hohmann transfer

DVclo = circularize orbit

DVdo = deorbit burn

DVsl = soft land burn

DVtot = DVlht + DVclo + DVdo + DVsl

Continue

Continue

Ultimately we require minimum DV and its corresponding geometry. This optimization analysis was carried out for both of the insertion altitudes: 300 km and .5 GEO (21181.46 km). The results of the geometry optimization for both insertion altitudes are shown graphically in Figures 4(a-b) and 5(a-b). As is seen from this analysis the choice of both l1 and go can significantly effect the DV calculations. (It should be noted that the only DV calculation effected by the geometry is DVlht that occurs at SOI encounter.) After the analysis was complete it was observed that minimum DV would probably occur at a l1 less than -45° (which was the limit on this analysis). Although this could conceivably introduce significant savings in DV it is sufficient for this preliminary analysis to demonstrate the trend.

Figure 6 uses the l1 = -45° (minimum DV case) to demonstrate the DV savings by inserting from the .5 GEO orbit. Additionally the optimum go angle for both altitudes can be determined (for 300 km go = 125°, .5 GEO go = 120 °).

Table 1 summarizes the DV comparison results. As expected it is seen that there is a significant DV savings (~1.63 km/s) by inserting from the higher Earth parking orbit. This saving is overcome however by the DV due to launch directly into that orbit. The total DV savings for the entire mission is in the favor of the LEO orbit with a savings of (~0.6 km/s). This demonstrates the tradeoff between DV savings due to insertion altitude and DV loss due to launch. It is possible that further study might determine an optimum altitude, higher than LEO, where the DV loss from launch would be offset by the DV savings in the rest of the mission.

The results of the geometry optimization are given in Table 2. As expected it is shown that by requiring a maximum time of flight trajectory to the Moon assuming an ellipse the trajectory approaches a parabola. This is shown by the calculated eccentricity (~.98) and the mission geometry sketch in Figure 7.

a = 1955.38 km l1 = -45°

e = 0.98 Y = 65.88°

go = 125° D = 384400 km

g1 = 7.91° Rs = 66300 km

3.4 Mass Estimates

The minimum requirements for the total mass delivered to the lunar surface was set at about 100 to 130 kg (200 -300 lbs). This would include the lander and communications platform as well as the two robots. A preliminary mass estimates analysis was carried out using the DV calculations from the previous sections. It was originally proposed to use an Orbital Sciences Pegasus launcher to save on launch expenses as well as the convenience of a mobile launch site. The analysis of this follows as well as the choosing of appropriate Star motors follows.

3.4.1 TK Solver Model

A simple TK Solver model was employed to estimate the various masses involved. Using the rocket equation and the DV's calculated previously, propellant, stage and lander masses were estimated. From this total delivered mass to the Moon could be determined as well as the Star motor required. The model is included in Appendix A.

3.4.2 Pegasus Analysis

Based on a due East launch performance of the Pegasus rocket [Wertz, 613], the maximum payload mass that can be carried to a 300 km altitude orbit is approximately 420 kg. Using the TK model mo was set at 420 kg and was solved for the various masses. From this analysis it was determined that the Pegasus could only deliver about 60 kg (132.65 lbs) to the Moon. This is far short of the 100 kg minimum assumed in the mission design. It was therefore determined that a larger launch vehicle would be needed. However to complete the Pegasus analysis a Star motor was chosen should such a light mass mission be possible. With some modification to decrease the propellant mass the Star 6A (TE-M-542-3) motor proved to be the best choice. This decrease is required since the propellant mass required is only about 3.95 lbm and the standard Star 6A has a propellant mass of 7.2 lbm. The Star 6A has been flown over 280 times making it a reliable choice. Although this motor would deliver 60.17 Kg to the lunar surface, it would carry alot of unnecessary structure mass with the upper stage. As a result the use of a Pegasus launch vehicle is not recommended.

3.4.3 Delta II Analysis

Since the Pegasus could not deliver the minimum payload to the Moon the next largest vehicle, the Delta II, was analyzed. The analysis was carried out the same as for the Pegasus. In this case the maximum payload mass delivered to a 300 km orbit is approximately 3800 kg [Wertz, 613]. This was used to determine the maximum deliverable payload to the Moon. A second analysis was carried out by reducing mo to 2500 kg. The results of both of these analyses are given in Table 4. It is apparent that for each of these cases the required propellant mass is less than the standard motor specifications. This will result in some dead weight from the structure.

From Table 4 we see that the maximum mass delivered to the Moon's surface is approximately 471.35 kg (1039 lb). This is well above the limits on the mission as is the case for an initial payload mass of 2500 kg. Further analysis should use the extra boost of the Delta to insert into a higher orbit thus reducing the DV on the upper stage. In addition it would prove advantageous to launch directly into the translunar trajectory from Earth. These scenarios should be considered in further studies.

3.5 Cost Analysis

Since this mission depends on the development of new technologies it is difficult to determine overall costs of the mission. However, since we are sending relatively small payloads to the Moon and since more than one mission will take place, the total cost of each mission will be reduced.

4.0 Conclusions and Recommendations

Much needs to be done to better define the proposed robotic testbed. Lander and robot masses cannot arbitrarily be defined in further studies. Since the lander, which consists of the propulsion and communication systems, comprises of a significant portion of the delivered mass better estimates should be obtained. This analysis gave the preliminary mass and DV studies based on crude calculations. However these calculations demonstrated that the mission can be completed with an existing launcher and star motor. This mission involves the development of cutting edge technology that will not only benefit space sciences but will have widespread applications on Earth. It is recommended that further study in this area be carried out.

References

1 Sheridan, Thomas B., "Merging Mind and Machine", Technology Review, v92, p135, March 27, 1992.

2 Davis, George W. and Wallace T. Fowler, "The Lunar Split Mission: A Robotically Constructed Lunar Base Scenario", The University of Texas at Austin.

3 Burnham, Darren L., "Back to the Moon with Robots?", Spaceflight v35, p 54-7, February 1993.

4 "Legs Win Over Wheels for Moon Work", Machine Design, v60, p14, February 15, 1988.

5 Wertz, James R. and Wiley J. Larson, "Space Mission Analysis and Design", Kluwer Academic Publishers, Dordrecht, the Netherlands, 1991.