Entering orbit around another planet from an interplanetary approach trajectory is probably one of the most critical phases in a planetary mission. Everything must come together almost perfectly the first time around in order to achieve orbit typically with no practical opportunity at a second chance if something were to go wrong. Over the last forty years of planetary exploration, there have been a number of unsuccessful orbit insertion attempts documented resulting from failures of hardware, software or navigation. But what about near failures?

Examining the history of planetary exploration, one mission stands out as probably coming the closest to failing to achieve orbit around its intended target while still actually making into orbit. The Soviet Mars 3 mission, launched on May 28, 1971 with the goal of delivering a lander on the surface of Mars and subsequently observing the Red Planet from orbit, was suppose to enter a 1,500-by-33,000 kilometer orbit about Mars with an orbital period of 25 hours when it arrived on December 2. Instead it entered a highly elongated 1,530-by-190,000 kilometer orbit with a period of 12.79 days. With an orbital eccentricity of about 0.95, it is the most eccentric orbit of any spacecraft ever to reach Mars. With this orbit, the velocity of Mars 3 at periapsis (where the orbit insertion burn normally takes place) was about 4,123 m/s or just 52 m/s shy of escape velocity at that altitude. In other words, Mars 3 was just barely bound to Mars. So what happened to Mars 3?



Mars 3 was the last of three spacecraft built and launched as part of the Soviet Union’s ambitious M-71 Mars program. The M-71 spacecraft were the first of a new and much more capable generation of Soviet planetary explorers that used the powerful Proton-D launch vehicle. Designed and built at NPO Lavochkin under the direction of Chief Designer Georgi Babakin, Mars 3 and its sister craft, Mars 2, were designated M-71P (with the “P” standing for “Posadka” or “Lander” in Russian). The M-71P was about 4.1 meters tall with a total launch mass of 4,650 kilograms – four times more massive than any previous planetary probe (in fact, no American planetary spacecraft would weigh more than the M-71P until the launch of the 5,548-kilogram Cassini Saturn mission in 1997).


Mars 3 being prepared for launch.

The core of the M-71P consisted of a 1.8-meter in diameter cylinder that housed huge tanks carrying over a metric ton of propellant needed by the craft’s KTDU-425 engine used for course corrections and orbital insertion. Mounted at the base of the spacecraft was a pressurized torroidal equipment section 2.3 meters across that housed all the critical electronic systems such as the communication equipment, computer control systems, telemetry systems and some experiment electronics. In addition to a pair of solar panels with a span of 5.9 meters and a 2.5-meter in diameter parabolic high gain antenna, various navigational sensors and experiments were mounted on the exterior of the spacecraft. At the top of the M-71P was a 1,210-kilogram entry vehicle. It consisted of a 3.2 meter in diameter aeroshell with a 635-kilogram lander nestled inside topped by a parachute and landing rocket package.

American planetary spacecraft relied on detailed instructions uplinked by controllers to enter orbit around its target based on the latest navigation and ephemeris data processed and analyzed on the ground. The M-71 spacecraft and subsequent second generation Soviet planetary probes could also operate in this mode or, as a backup, use a highly sophisticated autonomous navigation system that relied on inputs from various onboard sensors combined with the latest ephemeris data loaded into the spacecraft’s computer memory to automatically correct its approach path, eject the entry vehicle into its proper trajectory for landing and subsequently fire the KTDU-425 engine to enter the desired orbit. Not only did this sophisticated navigation system control the spacecraft, but it also controlled the Proton launch vehicle’s Blok D escape stage that sent the craft on its way to Mars from its Earth parking orbit saving 167 kilograms of mass (which could now be devoted to payload) in the process. This system was far ahead of anything the Americans had attempted with their planetary spacecraft.


Closeup of optical navigation instruments used by the autonomous navigation system on a second-generation Soviet planetary spacecraft. (A.J. LePage)

At this time, Soviet engineers and scientists did not possess an ephemeris for Mars with sufficient accuracy to ensure the success of their M-71 mission. To resolve this problem, the M-71 program also included a second spacecraft variant designated M-71S (with the “S” standing for “Sputnik” or Russian for “Satellite”). Similar in design to the M-71P, the 4,549-kilogram M-71S carried no lander. Instead it carried over two metric tons of propellant in enlarged tanks in the lengthened 3-meter long cylindrical core of the spacecraft. Its primary mission was to be launched on a high energy trajectory that would reach Mars and enter orbit several weeks ahead of the M-71P spacecraft.  The extra propellant carried by the M-71S was needed because of its faster approach trajectory compared to the M-71P orbiters. Tracking the M-71S in orbit around Mars would allow Soviet engineers and scientists to derive an improved ephemeris for Mars increasing the chances that the M-71P spacecraft would more accurately deploy their landers and enter Martian orbit. A side benefit of this approach was that the M-71S would follow a faster trajectory than the American Mariner 8 and 9 spacecraft also bound for Martian orbit at this time allowing the Soviet Union to claim another space first: the first spacecraft placed into orbit around another planet.

Unfortunately a programming error in the timer of the M-71S navigation system stranded the spacecraft and its Blok D escape stage in Earth orbit after its launch on May 10, 1971. The orbit of Kosmos 419, as it was called, decayed two days later. Without the navigation data provided by the M-71S, Mars 2 (which was successfully launched on May 19) and Mars 3 would have to rely solely on their new autonomous navigation system to precisely deploy their landers and enter Martian orbit.


Russian diagram of the M-71P approach sequence. Click on image to enlarge. (NASA)


The Automated Approach Sequence

The steps in the automated approach sequence for the M-71P spacecraft are illustrated in the Russian figure above. The first step in the sequence, “1” on the diagram, occurred about 7 hours before reaching Mars at a range of 70,000 kilometers when the spacecraft made measurements using its onboard navigational sensors to determine its position relative to Mars. At this stage, the spacecraft was aimed at a point 2,350 kilometers above the Martian surface with an accuracy of only ±1,000 kilometers. Using these navigational measurements and the latest ephemeris data stored in memory, the digital navigation computer automatically determined the final course correction (expected to have a Δv of less than about 100 m/s) that was carried out at “2”. After this course correction, the M-71P would be on course to deploy not only its lander but pass 1,500 ±200 kilometers above the Martian surface where the orbit insertion burn would occur.

About 4.5 hours before reaching Mars, the M-71P would automatically orient itself and deploy its lander at “3”. The entry package would then use its own rocket motor at “4” to perform a 100 m/s deflection burn that placed it in the proper trajectory to enter the Martian atmosphere. After the deflection maneuver, the lander would use its attitude control system to properly orient itself at “5” and set it spinning to maintain attitude for a ballistic entry. At a range of 20,000 kilometers, the M-71P would take a second navigational fix on Mars at “6” to make the final determination of what its proper attitude should be and when to start the orbit insertion burn at “7”. Using information from the navigational sensors, onboard gyroscopes and acceleration sensors, the M-71P computer would determine when to ignite the KTDU-425 engine and when to shut it down in order to achieve an orbit with a period of 25 ±2 hours.


The M-71P Missions to Mars

Mars 2 made its first course correction on ground command 17 days after launch on June 5, 1971. Because of the lack of a sufficiently accurate ephemeris for Mars, an automatic optical navigation measurement was made on November 21 while Mars 2 was still six days out from Mars followed by the second course correction commanded automatically by the onboard computer. The automatic approach sequence started as scheduled on November 27 seven hours before reaching Mars. Unfortunately an incorrect ephemeris was used by the computer in its calculations and the last course correction burn placed Mars 2 on a trajectory that brought the craft closer to Mars than intended. This same error also resulted in the lander being deployed on an incorrect approach trajectory. The Mars 2 lander entered the Martian atmosphere at too steep of an angle and was destroyed on impact. Not realizing that it was on a trajectory that brought it closer to Mars than expected, Mars 2 ended up entering a 1,380-by-24,940 kilometer orbit with a period of about 18 hours or about 7 hours shorter than intended.

Table I summarizes my calculations for the approach trajectory of Mars 2 and compares the actual and intended orbits assuming an asymptotic approach velocity of 2,880 m/s (taken from Clark et al.). It was further assumed that the orbit insertion burn occurred at periapsis with all the maneuvers, the approach trajectory and final orbit about Mars being coplanar. For the intended 25-hour orbit calculations, the same periapsis as the actual orbit was used and it was assumed that the automated navigation system knew the periapsis of its hyperbolic approach trajectory to Mars was closer than intended. As can be seen, my calculated Δv for Mars 2 for its actual orbit is 1,203 m/s which compares well with the nominal Δv of 1,190 m/s quoted in the literature. Had the Mars 2 navigational computer been aware of the lower than intended periapsis, the Δv should have been 1,139 m/s or 64 m/s less than it actually was.


Table I: Mars 2 Approach Trajectory and Orbit
Actual orbit Intended orbit

P = 25 hr

Periapsis (km, surface) 1,380 1,380
Apoapsis (km, surface) 24,940 33,122
Escape Velocity @ periapsis (m/s) 4,240 4,240
Asymptotic Approach Velocity (m/s) 2,880 2,880
Approach Velocity @ periapsis (m/s) 5,126 5,126
Orbital Velocity @ periapsis (m/s) 3,923 3,987
Δv (m/s) 1,203 1,139


Mars 3 performed its first course correction on ground command on June 8 or 11 days after its launch. With an updated ephemeris in its memory, Mars 3 began its automated approach sequence seven hours before reaching closest approach of Mars on December 2, 1971. After making a final automated course correction, Mars 3 deployed its lander on the correct approach path 4 hours and 35 minutes before reaching Mars. The Mars 3 lander entered the atmosphere of Mars and successfully landed on its surface at 13:50:35 UT. About 90 seconds after landing, Mars 3 began transmitting data to the orbiter passing overhead. Unfortunately contact with the lander was lost after only 20 seconds and never regained. The exact cause of the loss of signal was never definitely determined but may have been caused by a fatal coronal discharge resulting from the global dust storm then enshrouding Mars.


A color image of Mars acquired by Mars 3 in mid-December 1971 from its highly eccentric orbit during the global dust storm that covered Mars during the last quarter of 1971.

Meanwhile, Mars 3 entered orbit around Mars with a periapsis of 1,530 kilometers. This was very close to the nominal 1,500 kilometer periapsis that had been planned and well within the expected ±200 kilometer uncertainty. But instead of the intended 25-hour orbit with a 33,000 kilometer apoapsis, Mars 3 had entered a highly elongated 12.79-day orbit with an apoapsis of 190,000 kilometers. Table II summarizes my calculations for the approach trajectory of Mars 3 and compares the actual and intended orbits assuming an asymptotic approach velocity of 2,960 m/s (taken from Clark et al.) and coplanar maneuvers as before. The Δv for the intended orbit was 1,200 m/s which, again, is almost a perfect match for the nominal Δv of 1,190 m/s found in the available literature. But the Δv for the actual orbit was only 995 m/s or about 205 m/s short of the target.


Table II: Mars 3 Approach Trajectory and Orbit
Actual orbit Intended orbit with P = 25 hr
Periapsis (km, surface) 1,530 1,530
Apoapsis (km, surface) 190,000 32,972
Escape Velocity @ periapsis (m/s) 4,175 4,175
Asymptotic Approach Velocity (m/s) 2,960 2,960
Approach Velocity @ periapsis (m/s) 5,118 5,118
Orbital Velocity @ periapsis (m/s) 4,123 3,918
Δv (m/s) 995 1,200


Just to illustrate how close Mars 3 came to not attaining orbit, I calculated the burn time of the KTDU-425 engine used for orbit insertion. There are apparently no performance figures available for this engine but such numbers are available for the updated KTDU-425A engine used on the Soviet’s subsequent second-generation planetary spacecraft. Assuming an Isp of 315 seconds, a maximum thrust of 18.89 kN and an initial M-71P mass of 3,440 kilograms immediately before entering orbit (i.e. the 4,650 kilogram launch mass minus the 1,210 kilogram mass of the entry package), a nominal Δv of 1,200 m/s would require a 181 second burn (give or take a few seconds) to consume about 1,110 kilograms of propellant (which compares well with the stated 1,175 kilogram mass of propellant and attitude control gas carried by the M-71P). I calculate that the actual burn time for Mars 3 was 155 seconds or about 26 seconds shorter than intended. Had the orbit insertion burn been just 7 seconds shorter still, Mars 3 would have failed to slow below Mars’ escape velocity and would have flown back into solar orbit. Using these same assumptions, Mars 2 burned its KTDU-425 engine about 8 seconds too long during orbit insertion to end up in its 18-hour orbit.

So what caused the near-failure of Mars 3? Sources seem to indicate that it all boiled down to inadequately tested software used by the autonomous navigation system. Apparently if the velocity of the spacecraft was changing quickly (as would happen during a full-thrust orbit insertion burn), the control system could miscalculate the total change in spacecraft velocity based on data from onboard sensors resulting in the engine shutting down at the wrong time. In the case of Mars 3, the KTDU-425 engine had cut out about 26 seconds too early.

Although its highly eccentric orbit was ill-suited for planetary observations, Mars 3 was able to gather some useful data on Mars during the handful of close passes it made during its mission.  The flow of data was curtailed, however, by a failure of its high-bandwidth centimeter-wavelength (equivalent of C-band) telemetry system and an intermittent malfunction of its lower bandwidth decimeter-wavelength (UHF) telemetry system.  As a result of these problems, what few images of the originally planned 480 images Mars 3 did return using its film-based cameras were transmitted in a low resolution, 250-line mode.  Little useable data were returned from Mars 2 because of excessive noise in its telemetry which limited useful investigations to radio occultation observations.  Both spacecraft completed their science missions in March 1972 and finally ran out of attitude control gas that July ending their missions.

While the M-71P spacecraft were highly sophisticated engineering marvels that incorporated many new innovations, they did suffer from teething pains typical with the introduction of a new spacecraft design.  Although the scientific return from the M-71 missions was disappointing especially compared to the successful American Mariner 9, they did return valuable engineering data on this new spacecraft design.  Problems encountered during the M-71 and the subsequent M-73 missions to Mars launched in 1973 were eventually resolved and this second-generation planetary spacecraft went on to be used in the highly successful Venera program starting in 1975 (see “Venera 9 and 10 to Venus“).


Follow Drew Ex Machina on Facebook.


Related Video

Here is a short Soviet-era documentary film on the M-71 mission to Mars.  Although it is in Russian, it provides some excellent footage of the construction and testing of the M-71P spacecraft along with typical high-quality Russian animation of the mission to land on Mars.



Related Reading

“Venera 9 and 10 to Venus”, Drew Ex Machina, October 22, 2015 [Post]

“The First Mars Mission Attempts”, Drew Ex Machina, October 10, 2015 [Post]

“Planetary Orbit Insertion Failures Part II”, The Space Review, Article #2536, June 23, 2014 [Article]


General References

V.C. Clark, Jr., W.E. Bollman, R.Y. Roth and W.J. Scholey, “Design Parameters for Ballistic Interplanetary Trajectories Part I. One-way Transfers to Mars and Venus”, Technical Report No. 32-77, JPL, January 16, 1963

Wesley T. Hunter, Jr. and Mikhail Ya. Marov, Soviet Robots in the Solar System: Mission Technologies and Discoveries, Springer-Praxis, 2011

V.G. Perminov, The Difficult Road to Mars: A Brief History of Mars Exploration in the Soviet Union, Monographs in Aerospace History No. 15, NASA, July 1999