This document has been provided to the Friends and Partners in Space WWW
Server by the author for user reference only. It may not be reproduced or
re-published in any form without the direct consent of the author. E-mail 
addresses can be found in the author bibliographies if they were provided.

(note: the graphics to this paper will be added at a later date)

                                   Phillip S Clark
             The Block D rocket stage was first flown on a Soviet booster in 
        March 1967 when the Zond/L-1 test-bed Cosmos 146 was placed into 
        orbit.   It has been a standard part of the four stage Proton booster 
        since that time, although it is now known that the stage was initially 
        developed as the service module for the 1960s attempt to land a man on 
        the Moon.   Various versions of the Block D have been developed for 
        past and future missions, and this article will review the history of 
        the Block D family.
             Table 1 provides a summary of the programmes which have used the 
        four stage Proton booster (here called Proton-4) with the Block D and 
        Block DM fourth stages.
             To avoid confusion over the terms "Soviet" and "Russian" with 
        their derivatives the acronym FSU (former Soviet Union) will be used 
        in the remainder of this article.
        Naming the Stage
             FSU literature started to refer to the fourth stage of the Proton 
        booster as being the "Block D" in the early 1980s, a designator which 
        was immediately realised to be anomalous.   From descriptions of 
        labelling the Korolyov Sputnik-Vostok-Soyuz-Molniya family of launch 
        vehicles it was known that it was normal practice for the rocket 
        stages to be named in cyrillic alphabetical order and since "D" is the 
        fifth letter in the cyrillic alphabet the Proton Block D must have 
        started life as the fifth stage on another FSU launch vehicle.
             Confirmation of this hypothesis came when the first FSU 
        descriptions of the N-1 lunar booster and the FSU manned lunar 
        programme appeared.   The four stage Proton booster used a modified 
        Chelomei Proton launch vehicle with a new fourth stage added, Block D, 
        which was transferred from its original application as the fifth stage 
        of the N-1 lunar booster.
             The main engine of the Block D was designed by Mikhail Melnikov 
        of the Korolyov design bureau, and Vasili Mishin of the same bureau 
        (now NPO Energiya) has stated that he was in charge of designing the 
        rocket stage itself.   The current chief designer of the Block D 
        programme is Boris V Cherniatiev. 
        *  A much-abridged version of this paper was presented at the British 
        Interplanetary Society's Soviet Astronautics meeting on 12 June 1993.   
        In addition, the material has been used in the two editions of the 
        1993 and 1994 editions of Molniya Space Consultancy Report The Proton 
        Launch Vehicle Family
             FSU literature in the mid-1980s also noted that a modification of 
        the original Block D, designated Block DM, had been introduced in the 
        mid-1970s and was still in use.   How the Block DM differed from the 
        Block D was originally not known, but it was generally assumed in the 
        West that the Block DM had actually replaced the original Block D on 
        Proton-4 operations.
        Description of the Block D and Block DM
             The 1985 edition of Cosmomavtika Entsiklopediya edited by 
        Valentin P Glushko gave the first descriptions of the Block D, and the 
        basic numerical data are summarised in Table 2 [1].
             In addition, tables of cumulative payload and rocket body masses 
        have been published in FSU literature written or edited by Glushko 
        [2,3,4] and in 1983 an analysis of the first of these was published 
        [5,6].   Using the data from the FSU literature it was possible to 
        obtain the dry masses of the Block D stages when they were used on 
        deep space missions, the results being shown in Table 3.
             The data presented in Tables 2 and 3 can be used to obtain 
        further figures initially for the Block D and then for the Block DM.
             The Block D and DM assemblies comprise four objects: the rocket 
        stage itself, a casing which surrounds the rocket stage until just 
        after it separates from the Proton third stage plus two ullage motors.   
        The ullage motors separate as the main Block D/DM engine is ignited 
        for the final (or only) time.
             Using the figures in Table 2, the propellant mass of the Block D 
        comes to 14.8 tonnes, leaving an empty stage mass of 2.5 tonnes.   
        This is far heavier than the calculated Block D masses obtained from 
        Glushko's figures.
             However, one cannot claim that these results are accurate to two 
        or three significant figures because the total burn time of the Block 
        D is described as "600 seconds".   If this implies an accuracy of +_ 50 
        seconds then the result of the propellant mass calculation can only 
        have an accuracy of +_ 1.2 tonnes.   While it is probably safe to say 
        that the Block D propellant mass is close to 15 tonnes, it is 
        difficult to justify the accuracy in obtaining a dry structure mass of 
        2,500 kg.
             Table 3 shows that the original Block D stages had masses of 
        about 1,815 kg (+_ 40 kg) through to the end of 1970 and 1,880 kg (+_ 123 
        kg) from 1971 to 1984.   In fact, the masses of the six Block Ds used 
        for the 1971 and 1973 Mars missions were anomalously low (1,700 kg), 
        and excluding these gives a post-1971 mean value of 1,940 kg (+_ 85 kg).   
        The mass of the outer casing is said to be approximately a tonne - 
        here taken to be 800 kg which is a reasonable figure.
             The Block D assembly also carries two ullage rocket motors.   
        Each ullage motor - designated SOZ by FSU engineers - has a dry mass 
        of 56 kg and carries a similar amount of propellant [7].   The pair of 
        laden ullage motors will have a total mass of about 240 kg.
             The Block DM differs from the Block D in that it carries its own 
        control unit atop the standard stage assembly: on the Block D the 
        spacecraft control unit is linked with the rocket stage's propulsion 
        system.   Looking at the available sketch of the fourth stage casing, 
        it seems to be designed to simply cover the Block D/DM propellant 
        tanks, with the engine (simply designated 58 on the Block D, 58M on 
        the Block DM) protruding below the casing and the Block DM control 
        unit above the level of the casing.   It seems reasonable to assume 
        that the same casing was used for the Block D missions rather than a 
        shorter one.
             FSU descriptions have noted that the dry mass of the Block DM 
        assembly is 3,370 kg, and this appears to be a figure which includes 
        not simply the dry stage but also the outer casing and the two ullage 
        motors.   Since the casing and ullage motors are probably the same for 
        both the Block D and Block DM, this implies that the dry mass of the 
        rocket stage is about 2,350 kg.
             The results of these calculations are shown in Table 4.
             FSU literature has generally given the length of the Block D as 
        5.5 metres, and this length equates with that from the tip of the 
        oxidiser tank to the base of the main engine's rocket nozzle.
        Propellants for the Block D and Block DM
             FSU literature describes the propellants for the Block D and 
        Block DM as being a liquid oxygen oxidiser and a "hydrocarbon" fuel.   
        The latter term in FSU literature has normally related to kerosene 
        which is what the Block D and Block DM actually used.
             In the late 1980s the Block DM fuel was varied, one called SYNTIN 
        being used on some (but not all) missions.   SYNTIN has been described 
        as a "synthetic hydrocarbon of [the] cyclopropane row" on page 5 of 
        reference 20.   FSU references indicate that its use permits the 
        specific impulse of the Block DM's 58M engine to be increased from 352 
        seconds to 362 seconds, permitting at least 200 kg additional payload 
        to be carried to geosynchronous orbit for the same propellant load.   
        It is not known whether the use of SYNTIN requires modifications to be 
        made to the standard Block DM assembly.
             The liquid oxygen is carried in an approximately spherical 
        propellant tank at the top of the rocket stage, while the fuel 
        (kerosene or SYNTIN) is in a torus tank which surrounds the upper part 
        of the rocket engine assembly.
             The first and second engine ignitions performed on geosynchronous 
        orbit Block DM missions last for 450 seconds and 230 seconds 
        respectively [7].   If it is assumed that the Block D and Block DM 
        engines have the same thrust levels, this would imply a propellant 
        load of 16.7 tonnes and 16.3 tonnes for the LOX/kerosene and 
        LOX/SYNTIN combinations respectively.   In turn this would imply that 
        the propellant tanks would require an increase of 5%, but there is no 
        evidence of such a change from FSU scaled drawings (admittedly, such a 
        small change would be difficult to detect from the available 
        drawings).   More realistically, the Block DM burn times probably 
        include the period that the engine is being throttled up to full 
        thrust, and thus will overstate the calculated propellant load.   

             Therefore, it is reasonable to assume that the maximum propellant 
        masses for the Block D and Block DM are the same - about 15 tonnes.   
        Such an assumption permits the performance which coincides with that 
        which has been demonstrated on Earth orbit and deep space missions.
             The SOZ (stabilisation and launching provision) ullage motors use 
        nitrogen tetroxide and UDMH propellants [7].   Each assembly carries 
        five engines: two have a thrust of 5 kg provide pitch and yaw control, 
        one with a thrust of 10 kg provides yaw control and two with 2.5 kg 
        thrust produce longitudinal accelerations.   The propellant mass is 
        about the same as the dry mass of each SOZ (56 kg) and when they 
        separate from the Block DM each SOZ still carries 10-40 kg of unused 
        Block D and Block DM Launch Profiles
             It will be realised from Table 1 that the Block D and Block DM 
        have different missions applications.   Originally it was thought in 
        the West that the Block DM was introduced as a replacement for the 
        original Block D, but this is not the case.   The Block D is used for 
        deep space missions to the Moon and planets (plus some Earth orbital 
        tests in the L-1 manned lunar programme), while the Block DM was 
        introduced for the geosynchronous orbit and GLONASS missions - all 
        dedicated Earth orbit missions.
             The original Block D application as part of the L-3 manned lunar 
        programme required a multiple restart capability and this was used on 
        the standard Proton missions although from western analyses of the 
        Proton launch profile this was not realised.  Based upon the launch 
        profile of the Molniya booster, it was thought that a third stage 
        placed the combined Block D/payload assembly in a low Earth orbit and 
        the Block D made a single burn out of Earth orbit.
             In fact, this analysis is now known to be incorrect.   From the 
        very first launches the first three stages of the Proton-4 were sub-
        orbital and the Block D fired once to place itself and its payload 
        into a low parking orbit.   The two ullage motors were then discarded 
        (to be tracked and wrongly identified as discarded rocket stages, etc 
        in western catalogues) and the Block D would re-ignite to place its 
        payload into it's planned trans-lunar or interplanetary trajectory.
             On the other hand, when the Block DM was introduced the launch 
        profile became similar to that of the Molniya booster.   because the 
        planned geosynchronous orbit payloads (and, later GLONASS payloads) 
        were far lighter than the lunar and planetary probes (typically they 
        were 25% or less than the deep space probes) which had been launched, 
        the three-stage Proton was able to place the third stage in orbit 
        with a fully-laden Block DM and satellite.
             After reaching low Earth orbit the Block DM assembly and payload 
        separate from the Proton booster's third stage and 55 seconds later 
        the fourth stage casing is ejected.   During a period of about 15 
        minutes a series of pre-programmed manoeuvres is completed by the 
        Block D/DM and some 40 minutes after launch the assembly is rotated by 
        180o to compensate for the gyroscopic shift of the guidance system: 
        this takes about 2.5 minutes.

             Approximately 80 minutes after launch the fourth stage engine is 
        ignited as the assembly passes over the equator northbound.   The 
        ullage motors on Block D missions separate as the burn begins: they 
        are small enough not to be tracked before orbital decay or when they 
        are tracked they remain in USSPACECOM's catalogue of temporary orbital 
        contacts and not normally transferred to the catalogues which the 
        public sees.
             On Block DM missions to geosynchronous orbit the transfer orbit 
        inclination is shifted to 47.5o, close to the value required to 
        minimise the total manoeuvre value to geosynchronous orbit.   A series 
        of rotations about the longitudinal axis of the rocket stage and 
        payload are made during the transfer orbit coast which even out the 
        effects of solar heating on the assembly: in the Apollo era this was 
        called "barbecue mode" during trans-lunar coasts.   Generally three 
        axis stabilisation is maintained when not in this mode.   During the 
        coast to the transfer orbit's apogee another 180o rotation is made to 
        align the guidance system gyroscopes - again taking 2.5 minutes.
             After coasting half way around the transfer orbit preparations 
        are made for the second and final burn of the Block DM main engine 
        which takes place at the first transfer orbit descending node pass.   
        To settle the propellant remaining in the Block DM's tanks the two 
        ullage motors ignite 300 seconds before main engine ignition and are 
        ejected about 1 second after the main engine begins operating.   These 
        are the two objects which have always been tracked in geosynchronous 
        transfer orbits and their purpose was previously unknown.
             When the Block D and Block DM stages are shut down for the final 
        time subsequent operations are believed to be similar.   Immediately 
        after shutdown a separation command is sent to the payload.   
        Approximately 15 seconds later the payload separates.   During the 
        intervening period the assembly has been uncontrollable, with the 
        rotation about each axis possibly reaching a maximum of 5 o/sec.   On 
        Block DM missions the rocket stage then vents its unused propellant.
        Additional Comments on the Geosynchronous Mission Profile
             The standard launch technique described above means that 
        geosynchronous orbit satellites will always be initially placed in 
        orbit over the same longitude.   Given that the first burn of the 
        Block DM comes at the first equator crossing northbound which is over 
        10.3 oW, the longitude of the satellite at the time of geosynchronous 
        drift orbit injection will be:-
                  L  =  169.56 - P/8
        where L is the injection longitude (oE: 360o might have to be added to 
        the result to obtain a longitude in the range 0-360 oE) and P is the 
        transfer orbital period (minutes).   Typically the transfer orbit 
        periods are 615-635 minutes, so this means that injection takes place 
        over 91.43+_ 1.25 oE.
             This launch technique has resulted in some FSU data being 
        misunderstood in the West.   In describing the launch technique for 
        geosynchronous missions the commercial FSU literature gave the final 
        orbital period as having a range of 1,436+_ 20 minutes, and this was 
        interpreted as the error range for the launch vehicle - apparently not 
        a particularly accurate launch vehicle !   Such an assumption was 
        immediately known to be incorrect by astute observers of the FSU space 
             Geosynchronous payloads generally enter drift orbits with periods 
        of 1,400-1,480 minutes and are then allowed to drift around the 
        geosynchronous band until they approach their planned longitude: they 
        then perform a small manoeuvre to reach a near-stationary 1,436 minute 
        orbit over the required longitude.
             Therefore, the range of 1,436+_ 20 minutes was not an error range, 
        but the range of drift orbits which was being offered.   In terms of 
        orbital injections the Block DM atop a Proton booster is at least as 
        accurate as any western launch vehicle system.
        Other Block D and Block DM Mission Profiles
             While the above description gives the typical mission profile for 
        a geosynchronous mission it has been varied for a few missions.   A 
        disadvantage of the standard profile is that it limits the drift orbit 
        injection longitude.   Commercial literature has noted that the Block 
        DM assembly can be maintained in the low Earth orbit for up to a day 
        before the first firing takes place.   The previous formula for 
        orbital injection longitudes therefore requires amending to allow for 
        such a change in profile:-
                  L  =  192.010 - 22.452*N - P/8
        where N is the (integer) number passes through the parking orbit 
        ascending node before the first Block DM burn and with L and P as 
        before.   The second burn of the Block DM always takes place at the 
        first pass through the transfer orbit descending node.
             This modified launch technique was first used by Cosmos 1940 (the 
        first Prognoz early warning satellite), launched in April 1988.   
        After launch the Block DM/payload assembly remained in the low orbit 
        for four additional circuits and was then boosted to its transfer 
        orbit: this resulted in a drift orbit injection close to 0 oE.   If 
        the Block DM had remained in the parking orbit for one additional 
        circuit the injection longitude would have been about 340 oE - closer 
        to the planned initial Cosmos 1940 location over 336 oE.
             This technique has been used on two further missions - Cosmos 
        2155 (operated over 335 oE) and Cosmos 2209 (336 oE).   However, it 
        has been used sparingly and other satellites heading towards 
        approximately the same area around the geosynchronous band have not 
        used the new profile and thus can take a couple of weeks to reach 
        their stations.
             The GLONASS launch profile was originally similar to the 
        geosynchronous profile: the low parking orbit was at 51.6o but the 
        satellites were manoeuvred to 64.8o, 19,100 km orbits using two Block 
        DM manoeuvres (each involving a plane change): one manoeuvre would be 
        during the first parking orbit ascending node pass and the second 
        would be during the first transfer orbit descending node pass.

             Starting with Cosmos 1710-1712 in 1985 the initial parking orbit 
        inclination was changed to 64.8o, meaning that no orbit plane changes 
        need be conducted.   Since plane changes are no longer required the 
        initial Block DM manoeuvre can take place near the northern apex of 
        the parking orbit and the second one can come at the southern apex of 
        the transfer orbit.
             The launch technique for Astron and Granat was a throw-back to 
        the original Block D profile, although the Block DM was apparently 
        used on the missions.   For these two astronomical satellites the 
        first three stages of the Proton-4 booster were sub-orbital with the 
        combined Block DM/payload assembly being placed into a 200-2,000 km 
        initial orbit: apogee was in the southern hemisphere.   The two ullage 
        motors were tracked in this orbit, but misidentified in western 
        catalogues as what we would now call the Proton third stage and Block 
        DM casing (in reality these had been sub-orbital).   At the first pass 
        through apogee the Block DM re-ignited to place itself and the science 
        payload into an eccentric orbit reaching out to approximately 200,000 
             A modification of the standard Block D launch profile was used 
        for the two Fobos launches in 1988.   Contemporary Russian 
        descriptions of the launch and later written descriptions note that 
        the first three stages of the Proton booster were sub-orbital and the 
        Block D on each mission performed first part of the heliocentric 
        orbital injection phase.
             On Fobos 1 a single low orbit object was tracked (1988-058C) and 
        on fobos 2 two were tracked (1988-059C and D): these would be the 
        discarded ullage motors, although one was not tracked for Fobos 1.
             Heliocentric orbit injection was performed in two stages.   The 
        second burn of the Block D placed the spacecraft into an orbit 
        reaching out to more than 130,000 km (as measured for Fobos 1), and 
        after the Block D was discarded the main propulsion system of the 
        Fobos spacecraft itself was used for the final injection to 
        heliocentric orbit.
             The final unusual launch profile to be reviewed here is that for 
        Cosmos 1603 and Cosmos 1656.   This profile is unique in that it marks 
        the only one which saw the Block DM ignited three times, and it is 
        discussed in detail elsewhere [8,9].
             After launch the standard two objects were left in the initial 
        low orbit on these two missions.   At the first pass through the 
        initial orbit's descending node the Block DM ignited for the first 
        time, placing the assembly in a 51.6o, 190-835 km orbit (this was the 
        first time that the Block DM had ignited on the initial orbit 
        descending node).   At the first pass through the ascending node the 
        Block DM again ignited, raising the orbit to 815-855 km and changing 
        the orbital inclination to 66.6o.   The ullage motors were cast off as 
        the third manoeuvre began, changing the orbit to 850 km circular at 
        71o.   This final manoeuvre took place over Plesetsk, and therefore 
        was very expensive in terms of propellant (orbital inclination changes 
        should be done over the equator to minimise propellant consumption).
             This launch profile has not been seen again after Cosmos 1656 in 
        1985, and later Zenit launches to 71o, 850 km orbits went directly to 
        the required orbital inclination - no plane changes were needed.
        The Block D as Part of the Manned Lunar Programme
             Having discussed the modern missions conducted by the Block DM 
        which have permitted insights for operations of the original Block D, 
        the manned lunar applications will be reviewed.
             The Block D began its life as the fifth stage of the N-1/L-3 
        manned lunar landing complex: depending on how one wishes to 
        differentiate between a rocket stage and a spacecraft service module, 
        one could argue that the Block D was the L-3 lunar service module.
             After launch of an N-1/L-3 vehicle the first two stages (Blocks A 
        and B) would have been sub-orbital, the third stage (Block V) would 
        remain in a low Earth parking orbit and the fourth stage (Block G) 
        would perform trans-lunar injection.   After this the Block D and its 
        lunar payload separate from the Block G.
             The Block D must have been capable of performing small trajectory 
        corrections during the trans-lunar coast, but its first major ignition 
        would place the assembly in a 110 km lunar orbit, later lowering it to 
        one with a periselene of 16 km [11].   Here the lunar Soyuz with its 
        Block I* propulsion system would separate when one of the two 
        cosmonauts flying the mission had transferred by EVA to the lunar 
        lander.   The shroud covering the lunar lander would separate and the 
        Block D would ignite again to take the Block Ye lunar lander down 
        towards the Moon.   The ullage motors on the Block D would separate at 
        the beginning of the lunar descent burn.   At an altitude of 1.5-2 km 
        the dry Block D would separate and crash onto the lunar surface, while 
        the lunar lander would (hopefully !) soft land with its lone cosmonaut 
        some distance away.
             With the information known about the individual components of the 
        L-3 lunar stack it is possible to derive realistic masses and 
        performance data for the lunar orbit and surface operations, and this 
        is summarised in Table 5.   It is interesting to note that if the 
        Soviets had wanted to, they could probably have launched an Apollo 9 
        class mission using the Proton-4 booster carrying both the lunar 
        lander and the lunar Soyuz with cosmonauts (using the planned lunar 
        Block I propulsion system).
             The flight profile of Cosmos 382 appears to have been unique, and 
        it is now known that the spacecraft was designated L-1E, and carried a 
        modified Zond spacecraft on a flight to qualify the Block D for 
        orbital manoeuvres over a period of some days - as would be required 
        on the L-3 lunar mission.
             Although the flight of Cosmos 382 appears to have been unique, 
        there might have been two earlier missions which failed to reach 
        orbit.   Western literature has noted that November 1969 saw the 
        unsuccessful launch attempt of a spacecraft with "quite similar" 
        telemetry characteristics to those of Cosmos 382 [14], so this might 
        have been the first L-1E launch.   Official FSU literature released as 
        part of the Proton booster commercialisation indicates a launch 
        failure of a four stage Proton in February 1970, the payload being 
        identified as a Cosmos satellite: the only Cosmos to be launched atop 
        a Proton close to this time which was not using the name simply to 
        cover a launch vehicle or payload malfunction was Cosmos 382: 
        therefore the February 1970 failure might also have been a L-1E launch 
             The other flights of the Block D within the manned lunar 
        programme were as part of the Zond/L-1 launch vehicle.   In 1967 
        Cosmos 146 was a successful Zond/L-1 test with the Block D reportedly 
        performing two manoeuvres, while Cosmos 154 on a similar mission was 
        unsuccessful when a control system failure resulted in one of the two 
        ullage motors separating early, thus preventing Block D ignition.   
        These two spacecraft were designated L-1P, as precursors to L-1 
        missions proper.
             During 1967-1970 four unsuccessful and five successful launches 
        of Zond spacecraft were completed using the Proton-4/Block D 
        combination: of the successful launches, Zond 4 apparently travelled 
        to a lunar distance but was launched away from the Moon, while Zonds 
        5-8 performed loops around the Moon [15,16].
        The Future of the Block D/DM Propulsion System
             FSU engineers have announced plans for future uses of the Block 
        D/DM technology, although it is presently unclear how many of the 
        plans will be brought to fully operational systems.
             When the first FSU literature appeared describing the planned 
        upper stages of the Energiya shuttle booster [17] it was immediately 
        realised by western analysts that what was being described as the 
        retro- and correction stage (R&CS) was simply the Block D under 
        another name.   Drawings of the R&CS do not show the control unit of 
        the Block DM being carried, although this could simply be an omission 
        by the artist.
             The R&CS is proposed for three classes of mission.   One is 
        simply to provide the propulsion necessary to place a 90 tonnes 
        payload into a relatively low Earth orbit.   The second is an 
        application as an inter-orbit space tug.   And the third is to act as 
        a spacecraft service module for either lunar landing or Mars 
        orbiter/landing missions.
             A pair of Block D engines appears to be used as the orbital 
        manoeuvring system of the Buran space shuttle, which might still go 
        down in history as the first reusable space shuttle which only had a 
        single orbital flight.
             Finally, the Block DM has been proposed for addition to the two 
        stage Zenit (Zenit-2) booster to give the Zenit-3, capable of 
        geosynchronous orbit missions.   In terms of useful payload, the 
        Zenit-3 only seems to make sense if it is to be flown from an 
        equatorial launch site, from where it can match the current Proton 
        capability from Tyuratam to geosynchronous orbit.   Zenit-3 was 
        supposed to fly from the proposed Australian Cape York launch site, 
        but that project has been in a state of flux for a few years and its 
        future (as well as that of the Zenit-3) is extremely uncertain.
        The Block D and Block DM Flight Record
             Table 6 presents what can be considered to be a complete list of 
        the failures which involved the Block D and Block DM.   Such failures 
        are easy to detect since they normally take place after parking orbit 
        injection, and thus the debris from the launch (operational or 
        otherwise) is catalogued in the public domain.
             Failures which are more difficult to track down are those which 
        failed to reach Earth orbit, and there are two known cases of the 
        Block D preventing orbital injection (Luna probes in 1969 and 1975).
        The May 1993 Launch Failure
             A  Proton-4 was launched on May 27 1993 but failed to place a 
        Gorizont communications satellite in orbit.   While the ITAR-TASS 
        launch announcement appeared garbled, something closer to the correct 
        story was obtained by Nicholas L Johnson in conversations with Boris 
        Cherniatiev [19].   The second stage of the launch vehicle failed to 
        perform nominally, but the third stage operated as planned, although 
        it was incapable of generating the velocity required to reach orbit.   
        As a safety measure the Block DM vented all of its propellant and 
        subsequently was destroyed when it burned up in the atmosphere.
             The venting of the propellant from the Block DM is a procedure 
        which had previously been unknown from FSU literature.
             The cause of the failure was later said to have been due to too 
        high a level of copper in the propellants.
        NPO Energiya (USA) Data For the Block DM Variants
             A leaflet issued by the Energiya USA company [20] describing the 
        Block DM includes numerical data which differs from that quoted 
        elsewhere in this review.   As is often the case when data from 
        different Russian sources disagree it is not possible to be definitive 
        in saying which is correct and which is in error.
             The leaflet describes two versions of the Block DM: one with an 
        equipment bay and one without.   Other literature has indicated that 
        the fourth stage with the bay is the Block DM and that the one without 
        the bay is the older Block D.   However, it is possible that the 
        Energiya USA literature is describing an as-yet unflown variant of the 
        Block DM without the equipment bay.
             The Energiya USA leaflet notes that the Block DM propulsion 
        system permits operations over a period of up to 240 hours - more than 
        long enough for all of the Proton-4/Block DM  missions described in 
        Russian literature.
             Table 7 provides the mass details for the Block DM variants noted 
        in the Energiya USA literature.
             The Block DM engine is described as being developed by NPO 
        Energiya during 1970-1973, the first flight coming in 1974.   The 

        Block DM fuel characteristics are quoted in Table 8, again based upon 
        the Energiya USA leaflet.   It is also noted that the Block DM can be 
        fired up to seven times, although if necessary this number can be 
        increased.   There is a nozzle expansion ratio of 3,000 due to the use 
        of a "radiation-cooled" nozzle extension made of columbium alloy.
             Finally, the Energiya USA literature has included details of the 
        performance of the LOX/Kerosene and LOX/SYNTIN versions of the Block 
        DM, possibly called the DM-1 and DM-2 respectively (Table 9).   Once 
        more, Russian literature has used these designators inconsistently.   
        In 1994 all of the Proton missions have been credited as using the 
        Block DM-2 variant.
             The engine reliability is claimed to be 0.997, with a proven 
        success probability of 0.9.

        [1]   V P Glushko (editor), Cosmomavtika Entsiklopediya (published 
        1985 in Russian), p 48.
        [2]   V P Glushko, Razvitiye Raketostroeniya i Cosmonavtiki i SSSR 
        (second edition, published 1981 in Russian), pp 194-198.
        [3]   V P Glushko (editor), Cosmomavtika Entsiklopediya, p 498.
        [4]   V P Glushko, Razvitiye Raketostroeniya i Cosmonavtiki i SSSR 
        (third edition, published 1987 in Russian), pp 252-254.
        [5]   Phillip S Clark, "Soviet Spacecraft Masses for Earth Orbital 
        Programmes", Journal of the British Interplanetary Society, January 
        1985, pp 19-24.
        [6]   Phillip S Clark, "Soviet Spacecraft Masses for Deep Space 
        Missions", Journal of the British Interplanetary Society, January 
        1985, pp 25-30.
        [7]   B V Cherniatiev et al, "Identification and Resolution of an 
        Orbital Debris Problem with the Proton Launch Vehicle", presented at 
        the ESA First European Conference on Space Debris, Darmstadt (5-7 
        April 1993) and SPIE Space Debris Detection and Mitigation Conference, 
        Orlando (15-16 April 1993).
        [8]   Phillip S Clark, unpublished correspondence, 6 November 1984.
        [9]   Phillip S Clark, "Obscure Unmanned Soviet Satellite Missions", 
        submitted for publication in Journal of the British Interplanetary 
        Society (paper presented at the BIS Soviet Astronautics meeting, 12 
        June 1993).   Section 9 of the paper refers to Cosmos 1603 and Cosmos 
        [10]   V P Mishin, Pochemu Mou Ne Sletali Na Lunu ?, Znanye issue 12 
        1990, p 20.
        [11]   I B Afanase'ev, Neizvestnaye Korabli (Unknown Spacecraft), 
        Znanye issue 12 1991, p 29.
        [12]   Phillip S Clark, "Obscure Unmanned Soviet Satellite Missions", 
        ibid: section 7 discusses the lunar stack and Cosmos 382.
        [13]   V Filin, "Project N1-L3", Aviatshiya i Cosmonavtika, issue 2 
        1992, p 40.
        [14]   Anon, "Salyut Elements Separate, Signals Lost", Aviation Week & 
        Space Technology, 30 April 1973, p 21.
        [15]   I B Afanase'ev, ibid, pp 24-27.
        [16]   Phillip S Clark, "The Soviet Manned Circumlunar Program", 
        Quest, Winter 1992, pp 17-20 (reviews the material in reference 14).
        [17]   B I Gubanov, The Space Vehicle for Today and Tomorrow, Space 
        Studies Institute paper SSI BIG-2 (1990), figure 3.

        [18]   G Y Maksimov, From the History of Constructing and Testing of 
        the First Soviet Automatic Interplanetary Stations, paper presented at 
        the 1991 IAF Congress (IAA-91-690).
        [19]   Nicholas L Johnson, private conversations, 28th May 1993.
        [20]   Anon, Multipurpose Cryogenic Block DM with 11D58M Engine, paper 
        issued by Energiya USA during 1993.

        Illustration Captions
        Figure 1   The original Block D stage without its outer casing.   The 
        spherical tank at the top carries the liquid oxygen oxidiser and the 
        lower torus tank carries the kerosene fuel.   As with Figures 2, 3 and 
        4, the dimensions are given in mm on the original FSU material.   This 
        stage as flown on the four stage Proton booster is possibly little 
        changed from the original fifth stage to be carried on the N-1 lunar 
        booster.   (Copyright NPO Energiya and Kaman Sciences Company)
        Figure 2   The Block DM, shown both inside its casing and without the 
        casing.   The identically-sized case is believed to be used for the 
        Block D.   It will be noted that the only real difference between the 
        Block D and Block DM is the presence of the control unit added above 
        the oxidiser tank.   (Copyright NPO Energiya and Kaman Sciences 
        Figure 3   One of the two small ullage motors which are carried by the 
        Block D and Block DM stages.   On Block D missions they are not 
        normally tracked in orbit, but the resulting orbit would be so low 
        that the motors could decay before being catalogued.   On Block DM 
        missions the motors separate one second after the final ignition of 
        the main engine.   (Copyright NPO Energiya and Kaman Sciences Company)
        Figure 4   Russian drawing of the L-3 lunar stack carried aboard the 
        N-1 booster [10].   This Russian drawing is used rather than a western 
        re-drawn version since the majority of re-drawn versions include 
        details of what the western artists think should be shown rather than 
        what the Russian drawing actually does show.   (Originally published 
        in Znanye, issue 12 1990)
        Figure 5   The retro- and correction stage proposed for the Energiya 
        booster.   This appears to be a virtually unmodified Block D, the 
        major change being an improved and longer-life refrigeration system 
        being carried.   The two ullage motors are identified as "auxiliary 
        propulsion" in this Russian drawing.   (Copyright NPO Energiya)

        Table 1   Summary Launch Record of Proton Block D and Block DM Missions (to the end of June 1994)
        Spacecraft Series       Period               Proton-4/Block D                   Proton-4/Block DM
                                               Failures  Block  D  Block  D       Failures  Block DM  Block DM
                                                to LEO   Failures  Successes       to LEO   Failures  Successes
        Astron                  1983                                                                      1
        Cosmos (GEO/comms)      1974-Date                                                       1        12
        Cosmos (GEO/ELINT/EW)   1975-1986                                                                 4
        Cosmos (GEO/EW)         1988-Date                                                                 4
        Cosmos (GLONASS)        1982-Date                                                       2        20
        Cosmos/Zond (L-1P/L-1)  1967-1970          4         1         6
        Cosmos (L-1E)           1969-1970          2?                  1
        Cosmos 1603/1656        1984-1985                                                                 2
        Ekran                   1976-Date                                             4                  20
        Fobos                   1988                                   2
        Gals                    1994-Date                                                                 1
        Gorizont                1978-Date                                             1         1        29
        Granat                  1989                                                                      1
        Luna                    1969-1976          2?        4        10
        Mars                    1969-1973          2         1         6
        Molniya-1S (GEO)        1974                                                                      1
        Raduga                  1975-Date                                             1                  31
        Raduga-1                1989-Date                                                                 3
        Unknown (GEO)           1988+1990                                             2
        Venera/VEGA             1975-1984                             10
                                Totals            10?        6        35              8         4       129
        Notes   Launches are noted in the alphabetical order of satellite names.   For each programme the 
        launches are summarised as failures to reach orbit, Block D/DM failures once in orbit and launch 
        successes.   Launch failures to reach orbit are based upon FSU literature together with some western 
        rumours: the failures of two Luna probes (February and April 1969) and one possible Cosmos (L-1E)) to 
        reach orbit in 1969 have not been confirmed in FSU literature.   In addition, the FSU launch records of 
        the Proton booster indicate launches of geosynchronous orbit payloads in 1988 and 1990 which failed to 
        reach orbit: the intended payload names for these missions are unknown.   Cosmos missions have their 
        programmes indicated: GEO - geosynchronous orbit, comms - communications, EW - early warning.
        Table 2   Basic Data for the Block D
        Length                         5.5 metres
        Diameter                       4   metres
        Total mass                    17.3 tonnes
        Engine thrust                 85   kN
        Engine specific impulse      350   sec
        Burn time                    600   sec
        Notes   These data are extracted from reference 3, page 48 (the "Block 
        D" entry).

        Table 3   Dry Masses for Various Block Ds (derived from FSU Literature)
        Mission(s)         Dry Mass            Mission(s)         Dry Mass
                              kg                                     kg
        Cosmos 146           1,800             Luna 15              1,785
        Luna 16-17           1,888             Luna 18-19           1,937
        Luna 20              2,160             Luna 21              1,843
        Luna 22-23           1,826             Luna 24              1,830
        Mars 2-7             1,714             Venera 9-10          2,018
        Venera 11-12         2,020             Venera 13-14         1,925
        Venera 15-16         1,925             VEGA 1-2             1,920             
        Zond 4-6             1,792             Zond 7-8             1,796             
        Notes   These figures are derived from references 2, 3 and 4 with the 
        derivation of pre-1981 masses discussed in reference 6.   For the 32 
        Block Ds listed above the mean mass is 1,862 kg with a standard 
        deviation of 111 kg.   However, it is reasonable to split the data 
        into three groups.   The first represents the Block D missions through 
        to the end of 1970, and are thus stages close to the original N-1 
        variants: the mean mass and standard deviation comes to 1,814 kg +_  40 
        kg.   The second are the Mars 2-7 Block Ds which had unusually low 
        mean masses of 1,714 kg.   Finally, the remainder of the Block Ds 
        flown during 1971-1984 have a mean and standard deviation of 1,940 +_  
        85 kg.

        Table 4   Actual and Derived Block D and Block DM Data
        Item                  Block D Assembly                                 Block DM Assembly
                      Length  Diameter  Dry     Propellant  Total    Length  Diameter  Dry     Propellant  Total
                                        Mass       Mass      Mass                      Mass       Mass      Mass
                         m        m      kg         kg        kg        m        m      kg         kg        kg
        Stage Casing   3.996    3.700     800       --        800     3.996    3.700     800       --        800
        Stage          5.366    3.700*  1,860    14,800    16,660     6.218    3.700*  2,350    14,800    17,150
        Ullage Motor   1.000    0.600      56        60       240**   1.000    0.600      56        60       240**
        Notes   The derivation of the above figures is discussed in the text.   The rocket stage diameters marked 
        * refer to the maximum values, while the total mass for the ullage rockets marked ** refers to the two 
        motors together which are carried by each rocket stage.   The Block D dry mass which is quoted is the 
        overall mean for all the stages noted in Table 3.The lengths and diameters are quoted above to the same 
        (apparent) accuracy as the FSU literature permits.

        Table 5   Mass Model of the L-3 Lunar Stack
        Spacecraft/Component         Total     Propellant    Delta-V        Comment
                                      Mass         Mass
        Block I/Soyuz                9,400 kg      400 kg      120 m/s      LO manoeuvres: isp = 280 sec
                                                 2,700 kg    1,100 m/s      TEI manoeuvre: isp = 315 sec [?]
        Block Ye/Lunar Lander        5,500 kg                  100 m/s      Cushion landing: isp = 315 sec [?]
                                                             1,900 m/s      Lunar ascent: isp = 315 sec [?]
        Lunar Lander shroud          1,000 kg                               Surrounds lunar lander until Soyuz
        Block D                     16,580 kg   14,780 kg    1,000 m/s      LOI manoeuvre: isp = 352 sec
                                                             2,100 m/s      Lunar descent: isp = 352 sec
        Ullage Motors                  240 kg      120 kg                   Two motors, cast off as lunar
                                                                            descent begins
        Block D casing                 800 kg
        Total L-3 Stack             33,520 kg
        Notes   The masses of most of the above components are from Soviet literature or derived from material 
        given in Soviet literature: the only exception is the lunar lander shroud.   The dry mass of the Block D 
        is taken to be 1,800 kg, close to the mean value derived for the 1967-1970 period (see Table 3).   The 
        lunar lander comprises two sections: the landing frame, mass 575 kg (?), and the ascent stage, mass 4,925 
        kg (?).   Abbreviations used above are: LO - lunar orbit, TEI - trans-Earth injection, LOI - lunar orbit 
        injection.   The magnitudes of the manoeuvres are based upon requirements from Apollo 11 and taken from 
        the Apollo 11 Press Kit.   Estimated specific impulses (Isp) are noted by "[?]", and are based upon 
        typical specific impulse levels of engines being developed by the Isayev bureau during the mid-late 1960s 
        (although the Block Ye was actually the product of the Yangel bureau).

        Table 6   Summary of Block D and Block DM Failures
        Launch Date        Payload             Block     Comment
        1967 Apr  8        Cosmos 154            D       Zond/L-1P test.   Control failure resulted in one of the
                                                         ullage motors separating early, thus preventing Block D
        1969 Jun 14            --                D       Luna probe.   First Block D burn intended to place the
                                                         rocket stage and spacecraft in low Earth orbit failed.
        1969 Sep 23        Cosmos 300            D       Luna probe.   Cause of failure not known.   Only two
                                                         objects catalogued from launch, possibly indicating
                                                         that not even the Block D casing separated.
        1969 Oct 22        Cosmos 305            D       Luna probe.   As Cosmos 300.   Launch announcement
                                                         gave no orbital period, suggesting that less than one
                                                         orbit was completed: Block D mis-aligned at ignition,
                                                         thus driving the Luna probe back into the atmosphere
                                                         instead of out towards the Moon ?
        1971 May 10        Cosmos 419            D       Mars probe.   On-board flight sequencer mis-programmed,
                                                         preventing even separation from Proton third stage after
                                                         reaching parking orbit.
        1975 Oct 16            --                D       Luna probe.   First Block D burn intended to place the
                                                         rocket stage and spacecraft in low Earth orbit failed.
        1978 Dec 19        Gorizont 1            DM      Communications.   Borderline case of failure.   Block DM
                                                         mis-aligned for second manoeuvre, leaving satellite in
                                                         11.3o, 22,580-48,365 km orbit instead of approximately
                                                         1.5o, 36,000 km circular.
        1987 Jan 30        Cosmos 1817           DM      Communications (could have been an intended Raduga,
                                                         Ekran or Gorizont satellite).   Block DM failed to
                                                         ignite following casing separation.   Glavcosmos stated
                                                         that this Block DM had some experimental modifications
                                                         which caused the failure.
        Table 7, cont
        Launch Date        Payload             Block     Comment
        1987 Apr 24        Cosmos 1838-1840      DM      GLONASS (navigation).   Block DM shut down early during
                                                         first manoeuvre, attaining a 64.8o, 190-17,500 km orbit,
                                                         apogee some 2,000 km lower than intended.   Glavcosmos
                                                         stated that same modification as on Cosmos 1817 caused
                                                         the failure.
        1988 Feb 17        Cosmos 1917-1919      DM      GLONASS (navigation).   Unspecified failure prevented
                                                         separation of Proton third stage and Block DM.   Flight
                                                         sequencer deployed satellites in low parking orbit.

        Table 7   Block DM Variants Described by Energiya USA Leaflet
                                                         With Equipment      Without Equipment
                                                               Bay                 Bay
        Mass of Block DM, completely integrated
           and loaded with propellants                   18,400 kg           17,650 kg
        Propellant load:                                 15,050 kg           15,050 kg
           oxidiser                                      10,600 kg           10,600 kg
           fuel                                           4,300 kg            4,300 kg
        Mass of dry Block DM, fully intergrated           3,350 kg            2,600 kg
        Dry Block DM at staging                           2,500 kg approx     1,700 kg approx
        Notes   TYhese numbers are taken exactkly from the tablwe on page 3 of [20].   It should be noted that 
        the propellant masses as quoted do not agree: in each case the sum of the constituent oxidiser and fuel 
        is 150 kg less than the propellant load.   It is possible that this is because the latter figures include 
        the ullage motor propellants while the former do not.

        Table 8   Block DM Fuel Characteristics
        Fuel                         Density at               Viscosity, centistoke
                                     at 20 oC                 at 20 oC    at -40 oC
        Kerosene (RP-1 type)         833 kg/m3                >2.5        <25
        SYNTIN                       851 kg/m3                1.437       5.44
        Notes   This is reproduced  from the table on page 5 of [20].

        Table 9   Comparison of the Block DM Engine Data for Kerosene and SYNTIN Applications
        Fuel                         Kerosene            SYNTIN
        Thrust, kgf                  8,500               8,800
        Thrust, kN                   85*                 88*
        Thrust,tonnes                8.7*                9.0*
        Chamber pressure, kg/cm2     79                  81
        Exhaust velocity, m/s        3,452*              3,540*
        Specific Impulse, sec        352                 361
        Notes   Reference 20 gives the thrust values in kgf: the values marked * are converted from these values.   
        Similarly, the document quotes the specific impulses and again the exhaust velocities are simply 
        conversions of these.