ch8-5

On Mars: Exploration of the Red Planet. 1958-1978



TESTING THE LANDER


[256] Another phase of the lander's evolution was the multiplicity of tests to which the components, subassemblies, and assemblies were subjected.


[Whole page 257] (The guidance, control, and sequencing computer was the Viking lander's brain. At right, magnetic wires as fine as human hair are inserted into the computer at Honeywell Aerospace, Saint Petersbury, Florida. In testing below, the HDC-402 computer, part of the lander's computer, looks like pages of a book. At bottom, Jim Martin (second from right) on 10 January 1975 congratulates Barton Geer (left), director of system engineering and operations at Langley; R. Wigley, Honeywell's Viking program manager; and F. X. Carey, Martin Marietta resident manager at Honeywell. GCSC flight article 2 and the qualification unit are in the foreground.)

[258] Again, this was not terribly exciting work, but it was essential to producing spacecraft that could be relied on to function far from Earth.

As with the Viking orbiter, a number of simulators were developed to verify analytical predictions of lander system performance, to investigate the effects of the thermal and dynamic environment on the craft, and to permit tests of subsystems, such as the scientific experiments. The major Viking simulators included:
Lander (structural) dynamic-test model (LDTM) . A flight-style structure with partial flight-style or equivalent propulsion lines and tanks. Mass (weight) simulators were used for nonstructural hardware. The LDTM was used for structural vibration, acoustic noise, separation tests, and pyrotechnic shock evaluation.
Lander (structural static-test model (LSTM) . A flight-style structure used for qualification of the primary structure under steady-state and low-frequency loads.
Orbiter thermal effects simulator (OTES). A simulator used to study the orbiter's thermal and shadowing effects during the lander-development thermal environmental tests.
Proof-test capsule (PTC) . A complete Viking lander capsule assembly assembled from flight-style hardware, used for system-level qualification.
Structural landing test model (SLTM) . A 3/8, geometrically scaled model of the lander, dropped at various velocities and attitudes to determine landing stability boundaries. The 3/8 scale was chosen because the Martian gravity was 3/8 that of Earth's.
Thermal-effects test model (TETM) . A full-scale model incorporating developmental thermal control systems and flight cabling test harness. Flight equipment thermal effects were simulated by special equipment. The TETM was used to verify the system developed for controlling the temperature of the lander.
Electrical thermoelectric generators (ETGs) . Generators used in testing in place of the radioisotope thermoelectric generators (RTGs). ETGs had electrical heating elements that simulated the electrical and thermal characteristics without the hazard of nuclear radiation. ¹²
There were three broad categories of tests: system development, qualification, and flight acceptance. Development tests determined the levels of performance that components and subassemblies would have to meet to be acceptable. They also provided early identification of design deficiencies. These trials used primarily the dynamic-test model, the orbiter thermal-effects simulator, and the thermal-effects test model. Qualification rests used hardware attached to the proof-test capsule and the static-test model. During the "qual tests," hardware was subjected to stresses and environmental conditions that exceeded any expected during the mission. Environmental tests included heat compatibility, acoustic noise, launch sinewave vibration, landing shock (drop test), pyrotechnic shock, solar vacuum, and Mars-surface simulation. These and additional tests were performed at the [259] component and subsystem level. The flight acceptance tests were performed on flight hardware before qualification testing. Only the thermal sterilization and solar vacuum tests were made with assembled flight landers.

Environmental Tests

The proof-test capsule, encapsulated in its bioshield, was subjected to the heat compatibility test to verify that the system could withstand heat sterilization. During this test, the chamber atmosphere consisted of dry nitrogen containing about three percent oxygen and other gases. The capsule was subjected to 50 hours of 121°C heat, and flight landers were exposed to 112°C for 40 hours. Components were subjected to five 40-hour cycles and three 54-hour cycles at 121°C.
The vibrations of liftoff were computed by analysis of data from earlier Titan-Centaur flights and the February 1974 proof flight of a Titan IIIE-Centaur D-1T launch vehicle (this flight and preparation of the Viking launch vehicle are discussed in appendix E). Despite the necessary destruction of the Centaur stage on this flight after its main engine failed to start, some information was gained to help define the ground-based simulations (launch sinewave vibration tests) of the low-frequency vibrations encountered during launch, stage separation, and spacecraft separation. Through combined analysis, flight-derived data, and simulations, the engineers were able to determine if the lander components could withstand the predicted vibrations. ¹³ The acoustic noise test simulated the effects of the sounds of powered flight. Levels of the individual components were determined by earlier tests using the lander dynamic-rest model and proof-test capsule. ¹⁴
Random vibration tests were applied only at the component level, to screen out faulty workmanship and design defects. Laboratory simulations of the levels of vibration encountered during actual flight proved not to produce satisfactory data. Borrowing from procedures devised during the Apollo program, the vibration levels were raised to a level that would screen out bad components but not damage good ones. Component vibration levels were the same for both qualification and flight acceptance testing, but the latter was shorter so that multiple tests could be run without exceeding the qualification test levels. In the pyrotechnic shock tests run at the system level, a series of pyrotechnic devices was fired to simulate the effects of actual mission events and at the same time demonstrate the actual performance of the pyrotechnically actuated mechanisms. Components were subjected to vibrations similar to those expected with the Viking pyrotechnic devices and to contained explosions that replicated the impact of explosions and gas pressure buildups on specific assemblies. ¹⁵
Solar vacuum tests, held in a nearly complete vacuum in Martin Marietta's test chamber (4.5 meters in diameter and 20 meters high), simulated the worst predictions for thermal heating and cooling during the flight to Mars. Both the effects of heating and cooling and the performance of the lander's thermal control system were evaluated. Each mission phase [260] was completed twice for the qualification tests of the proof-test capsule and once for the flight acceptance tests of the flight landers. ¹⁶
Deorbit-entry-landing thermal simulation tests, conducted on a component level, duplicated the effects of entering the Martian atmosphere-pressure increase, entry heating, and the post-landing cooldown. Components were placed in the vacuum test chamber at 1/760 of an Earth atmosphere, heated to a temperature of 149°C, and held there for 530 seconds. Chamber pressure was then raised to 5/760 of an Earth atmosphere with cooled nitrogen gas, to provide an atmospheric temperature of -101°C. In this manner, the lander's passage through the Martian atmosphere with the attendant heating and cooling was duplicated. The change of 250°C represented the wide range of temperatures that the lander would be exposed to on Mars. Such extremes were part of the reason the engineering of the lander had been such a complicated task. For all components, the most critical period would be the 15 to 20 minutes after landing, since by that time all equipment would be operating and the entry heat buildup would not have had time to dissipate.
In the landing shock tests, the proof-test capsule, with landing gear extended, was dropped from a height necessary to achieve a velocity of 3.36 meters per second on impact. Each drop produced the worst possible dynamic loads on a different landing leg and footpad. In addition to these evaluated analytically and then measured during the balloon drop tests (balloon-launched decelerator tests) at the NASA White Sands Test Facility in New' Mexico in the summer of 1972. They were carried out successfully despite postponements caused by uncooperative weather. As a consequence of these tests, new techniques were developed to unfurl the parachute progressively, minimizing the deployment shocks to the lander. ¹⁷
During the Mars-surface simulation tests, the lander configuration of the proof-test capsule was subjected to thermal conditions worse than those expected on the surface of Mars. By subjecting the lander to different conditions and varying the vehicles' internal electrical power, three basic tests were performed-hot extreme, cold extreme, and the predicted norm. ¹⁸ In consultation with the Science Steering Group, the test engineers chose argon for the chamber atmosphere during the cold extreme, because preliminary data from the Soviet Mars probes had indicated that as much as 30 percent of the planet's atmosphere might be composed of this rare gas. ^* Since argon promotes electrical corona and arcing in electronic components, the test teams were to determine whether there would be any adverse effects on lander subassemblies if the concentration of argon was that high.

Science End-to-End Test

One of the most significant activities during the lander testing cycle was the science end-to-end test (SEET), conducted during the Martian....

Table 45
Mars Surface Thermal Simulation

Parameter	Hot Extreme	Cold Extreme	Nominal

Shroud temperature	-129°C ±5°C	-151°C to -81°C	-112°C ±5°C
Chamber pressure and atmosphere	2 ± 0.1 mb, CO₂	35 ± 2 mb, Argon	4 ±0.1 mb, CO₂
Solar radiation	1078 ±47 watts/m²	0 watts/m²	539± 31.5 watts/m²
Solar duration	12.33 hrs.	None	12.33 hrs.
Vehicle power	1395 watt-hrs	1345 watt-hrs	1371 watt-hrs
Ground simulator	-46°C to +24°C	Uncontrolled	Nominal
Thermal coating	Degraded	Original	Original
ETG thermal output	Maximum (680 ±2 watts)	Minimum (630 ± 2 watts)	Nominal (673 ± 2 watts)
Test duration (PTC)	3 days	4 days	3 days before hot extreme; 3 days before cold extreme


[261]....surface simulations at Martin Marietta. The two major SEET objectives were "to verify the adequacy of the implementation of the scientific investigations from sampler collection to interpretation of resulting data by the scientists'' and to "familiarize the Viking scientists and other flight operations personnel with total operation of the investigations and their respective characteristics.'' In the course of carrying out these basic objectives, any hardware or procedural problems were to be resolved, to avoid similar difficulties during the actual mission. ¹⁹

Getting the science end-to-end test started took some effort. It was postponed several times because of problems with the motor used to load samples into the oven heating assembly of the gas chromatograph-mass spectrometer. When the pumpdown of the vacuum test chamber began on 17 September 1974, the proof-test capsule lander used in the operation had a GCMS simulator aboard instead of the actual test unit. SEET was also run without the biology instrument. Despite the absence of these two major experiments, the test was useful.
The lander systems were examined rigorously. During the thermal vacuum chamber operations, a Martin Marietta computer facility sent commands via cable to the guidance, control, and sequencing computers. The plated-wire memory, once a leading top 10 problem, performed very well in the simulated Martian atmosphere. In addition, JPL processed data recorded on computer tapes from lander subsystems much as data would be during the real mission. Tests of the ultrahigh-frequency (UHF) radio link....



[262] (Viking simulators went through intensive environment tests to ensure the final spacecraft would function far from Earth. Above left, Viking program technician Alonzo McCann adjusts a cable on the proof-test capsule-decelerator assembly, as lander and capsule are prepared for January 1974 heat-verification tests. One of the lander cameras is to the right of center. Above right, a technician watches as an acoustic shroud is lowered over the proof-test capsule before acoustic tests in mid-June 1974. At lower right, the proof-test capsule is lifted out of the vacuum chamber at Martian Marietta, Denver, in October 1974 after a month-long series of rigorous tests to qualify it for operations on Mars.)

[262].....for data transmission and the lander tape recorder also indicated that those systems were ready for flight.

Other subsystems were given a thorough examination: the surface sampler, the lander's imaging system, the weather sensors, the x-ray fluorescence spectrometer, the seismometer, and the biology sample processor. [263] The multipurpose surface sampler (boom-and-scoop assembly) successfully delivered soil samples to the x-ray spectrometer and biology processor unit and to the GCMS position. The only significant problem occurred when the sampler arm snagged on the holder of a brush used to clean a magnet on the magnetic properties experiment. This problem was cleared up by minor hardware modifications and a new mission rule that prohibited cleaning the magnets until all of the biology samples had been taken. The lander facsimile cameras made nearly 100 images, including pictures of trenching exercises with the backhoe and of particles adhering to the magnets. The meteorology instrument performed well in Marslike conditions that could not be duplicated in a standard wind tunnel. Although the biology instrument was not on board, the processor containing the screens and cavities for the measurement and separation of the materials scooped up by the surface sampler was tested and proved satisfactory. ²⁰ During the seven days (18-23 September) that it took to simulate five days of experiments on the surface of Mars, many important lessons were learned as procedural and hardware "glitches" were encountered and overcome, and much needed experience was gained with the meteorology, seismology, camera, x-ray fluorescence spectrometer, and magnetic properties experiments. ²¹
Priestley Toulmin, team leader for the inorganic chemical investigation (x-ray fluorescence spectrometer) had been uncertain about the merits of SEET as it was planned, however. Toulmin's experiment, a late addition to the lander science payload, would determine the nature of inorganic compounds (minerals) in the Martian soil. As early as 1968, the Space Science Board had suggested it in recommendations to NASA for planetary explorations. But the priority given inorganic analysis was much lower than that assigned the search for biologically derived compounds-although, with the exception of this experiment, the original payload for Viking had followed the board's suggestions closely. Information gathered from the lunar samples returned by Apollo astronauts and early Mariner 9 results suggested the need to reconsider the utility of inorganic analysis. Mariner 71's findings were particularly evocative because they indicated that Mars was geologically younger and more active than had been expected. As a result, in the fall of 1971 the space science community lobbied the NASA management, especially John Naugle, associate administrator for space science and applications, to include an inorganic experiment on the lander. Of two possible investigations, the one designed by Martin Marietta and the team led by Pete Toulmin was selected. (The other instrument, designed by a team led by Anthony L. Turkevich at the Enrico Fermi Institute of the University of Chicago, had been under development for a longer time, but the XRFS was expected to cost less, be lighter, and require less space and power.) ²²
As time for SEET approached, Toulmin was concerned about the manner in which it would be conducted. Both he and Klaus Biemann, team [264] leader for the molecular organic analysis (the GCMS), had insisted strongly on the inclusion of "blind" samples in the analyses to be done by their instruments. These materials, unknown to the teams, would be identified by the results of the experiments, to simulate the interpretative work of the actual mission. In addition to making certain that this aspect of the SEET experience was carried out, Toulmin told Jerry Soffen in early September 1974 that he was concerned about the validity of the trials since the x-ray fluorescence spectrometer to be used in the test was different from the actual flight article. The test version had several shortcomings that had already been corrected in the flight units. A final reservation centered on the seeming inflexibility of the test plans. ²³
By the end of the science end-to-end test, however, Toulmin believed in its worth. He had previously discussed with Jerry Soffen "some reservations and qualifications the Inorganic Chemical Investigations Team felt were applicable to that program." In most instances, Toulmin believed that "the events proved us correct in our concerns regarding the state of the hardware, the software, and ourselves' and they had predicted several of the break downs that occurred. But in one major respect Toulmin felt he and his colleagues had misjudged the testing program: "I...grossly underestimated the tremendous value of the experience for those who participated in it. We learned things about the operation of the instrument and its relations with the rest of the lander, and about the recognition, diagnosis, and correction of problems and malfunctions that we would never have learned by any other method." Although the actual mission would differ greatly from the simulations, "it was an invaluable introduction to a whole new world." In his report to Martin, Toulmin singled out "for special mention the three unflappable controllers of the SEET data room: Henry von Struve, Frank Hitz, and Ron Frank." ²⁴
Phase B of the science end-to-end test was less satisfactory. Begun on 7 October with the reworked gas chromatograph-mass spectrometer, it had to be terminated on the 10th when additional problems were encountered with that instrument. These difficulties led to a special test of the GCMS in conjunction with the biology instrument's performance verification test in February 1975. Despite some additional functional difficulties, Klaus Biemann was able to identify from the GCMS data tapes the five compounds in the blind samples.
Whereas the mass spectrometer went through the end-to-end functional and operational exercise, the biology instrument did not. The biology instruments were delivered too late for proper testing. By the time the hardware became available, limited time, money, and manpower argued against the thorough test. To questions about the adequacy of the functional testing of the hardware on the proof-test capsule lander in Martin Marietta's thermal vacuum chamber and the biological operation of the experiments, Cal Broome told Martin on 30 June 1975, less than two [265] months before liftoff. "The current planning assumes that the testing already accomplished is adequate, i.e., the combination of Biology [performance verification] at the lander level (instrument 103) and soil biology at the instrument level (instrument 102 and 103) is adequate to provide assurance of proper operation on Mars." He added. "There is no question that this program does not provide ultimate verification, i.e., operation of a flight instrument in a lander with real flight sequences, and verification of proper results,'' but said, "Our position has been that risk of the current approach is acceptable." Broome was responding to a NASA Office of Space Science inquiry about the possibility of conducting a biology end-to- end test after the Viking spacecraft had been launched. ²⁵
Four major factors influenced the scope of the biology instrument acceptance test program. One, the introduction of soil or experiment nutrient into an instrument would render it unusable for flight. Cleaning the instrument was impossible without destructive disassembly. Thus, the functions of the flight instruments (S/N 104, 105, and 106) had to be tested only by simulating their operations on Mars. Soil testing was necessarily limited to components and units not reserved for flight use. Two, the complexity of the instruments, the multiplicity of their functions, and the operational pace (one minute between commands) meant that complete functional tests would be extremely time-consuming. The minimum time required for an entire end-to-end electrical and pneumatic checkout of a biology instrument was one month on a round-the-clock schedule. Only abbreviated functional tests could be performed. Three, given the long turnaround time required to repair and retest instruments if a component failed, all components and subassemblies had to be tested before assembly in the integrated instrument, where accessibility was a problem. And four, a substantial number of design changes were incorporated into the flight units after the manufacture and test of the qualification unit (S/N 102), requiring additional qual tests. Functional tests were then carried out to ensure that the flight instruments had not been harmed by the qualification test stress levels.
Each flight version of the biology instrument was subjected to a sequence of acceptance tests: operational system checkout, vibration test, functional verification, thermal verification, sterilization, and operational system checkout. The operational checks were computer-controlled, testing the electrical functioning of the instrument. Mechanical and structural quality was verified through vibration tests, while the functional verification tests were complete validations of all instrument systems. Computer- controlled electrical and pneumatic sequences assessed individually the functioning of each critical component or subassembly. The thermal verification tests were performed with the biology instrument in a Marslike atmosphere of carbon dioxide through a temperature range of - 18° to 30°C. Instruments were sterilized in a biologically filtered nitrogen atmosphere at....

[266] (Science end-to-end tests sought to verify complete performance of the Viking scientific instruments and familiarize scientists and flight operators with the total operation of Mars investigations. Above, a technician prepares the proof-test-capsule lander for the environment and SEET tests. At right, sample boxes are positioned for testing the lander's surface-sampler assembly.)

[266]....20°C for 54 hours. The total acceptance test spanned three to five months, depending on problems encountered during the process. ²⁶
Although this was a busy test schedule, no flight-model biology instrument had been tested as part of the total lander system, and in the fall of 1975 Harold P. Klein, leader of the biology team, and his colleagues argued for such a test. Langley and headquarters personnel resisted any lengthy additional testing. Such an examination could not take place before January 1976 and would interrupt a number of schedules. In late September, the Viking Project Office proposed a committee led by Gary Bowman, biology instrument team engineer, to take an in-depth look at the biology instrument test data from a lander systems point of view. From the team review, areas of specific concern could be identified and a decision about additional tests made. ²⁷

Klein responded on behalf of his teammates in November after Bowman's group and the biology team had looked at the testing issue again. [267] Ideally, the biology team would have liked to install the flight-model S/N 104 biology instrument its the proof-test capsule at Denver, to make biological examinations of soil samples, but the S/N 104 unit had to he kept sterile until after the mission, when additional tests might be necessary.
What the biology team could do was install the proof-test-capsule unit (S/N 103) on the proof-test-capsule lander to observe real data being processed from the biology instrument detectors through the lander system. The biology instrument simulator was not similar enough to the flight hardware to provide a meaningful test of the lander-biology instrument interface, but the test could simulate the sequence of biology instrument operations from soil collection through processing, analysis, and data return. Not only would experimenters have a chance to see if the instrument would function as planned, but they could watch their hardware in action, in preparation for the days when the instruments would be operated on Mars. ²⁸
Jim Martin and his staff on 25 November 1975 decided at least part of the tests the biology team wanted could be carried out during the flight operations-software verification tests scheduled for the proof-test-capsule lander in February 1976. Only the tests that would not require extra funds could be done. Martin told Klein: "We have neither the dollars to extend the test nor the people to analyze the data." Other aspects of the biologists' plans for testing were likewise impossible:

: ....your request for lander/biology tests with transmitters/antennae in real operational modes is also difficult to accommodate. As you know, this test would require use of an anechoic chamber (very expensive) or moving the entire lander to an outdoor location to avoid RF reflections (also expensive). We made a fundamental decision in 1973/1974 that the lander [electromagnetic compatibility] test program had to proceed without a real biology instrument because such an instrument did not exist until much too late. Instead, we have relied upon the positive results of a rigorous EMC test on the instrument at TRW. In today's dollar limited environment, the dollars to plan, set up, and conduct another radiated EMC test for biology are prohibitive. We must rely on analysis and instrument level test experience. ²⁹


While not enthusiastic about any additional biology testing, Martin informed Noel Hinners at NASA Headquarters that the "potential return from [the partial testing he had agreed to] is sufficient to incorporate it into our plans.'' He believed that the project management had "done everything reasonable to satisfy the concerns of the Biology Team as to the adequacy of the pre-landing test program." Martin wanted to turn to other more important issues: "Following the test, we must and will devote the full biology flight team resources to preparation for landed operations,. . . .including [268] training contingency analysis and preparation of pre-canned sequences to be ready for the multiplicity of possible required reactions to data from Mars." ³⁰

^* Subsequent Viking data indicated that the argon content in the Martian atmosphere was only about 1.5 percent.

[Whole page 257] (The guidance, control, and sequencing computer was the Viking lander's brain. At right, magnetic wires as fine as human hair are inserted into the computer at Honeywell Aerospace, Saint Petersbury, Florida. In testing below, the HDC-402 computer, part of the lander's computer, looks like pages of a book. At bottom, Jim Martin (second from right) on 10 January 1975 congratulates Barton Geer (left), director of system engineering and operations at Langley; R. Wigley, Honeywell's Viking program manager; and F. X. Carey, Martin Marietta resident manager at Honeywell. GCSC flight article 2 and the qualification unit are in the foreground.)


[266] (Science end-to-end tests sought to verify complete performance of the Viking scientific instruments and familiarize scientists and flight operators with the total operation of Mars investigations. Above, a technician prepares the proof-test-capsule lander for the environment and SEET tests. At right, sample boxes are positioned for testing the lander's surface-sampler assembly.)