Space Launch Report
Home    On the Pad      Space Logs     Library       Links
On the Pad:  Space Launch Site Data 

Questions/Comments to

Ed Kyle

delta31s.jpg (14566 bytes)THOR:  FLOWN VARIANTS
by Ed Kyle

Delta 31, a Sleek Thor-Delta C, Orbited NASA's Explorer 28 (IMP C) from Cape Kennedy LC 17B on May 29, 1965.

"Baseball Card" type details of most of the flown Thor family variants are provided in the following links, listed in approximate chronological order.

1.  Thor IRBM

In 1955 the U.S. began Thor IRBM development as a stop-gap until Atlas could enter service. Thor would use Atlas-derived propulson and warhead and an existing inertial guidance system. It would have 2,400 km range to reach the Soviet Union from Western Europe. Douglas Aircraft Company won the SM-75 Thor contract in December 1955. The company delivered its first Thor from Santa Monica, California to Cape Canaveral on October 26, 1956.

The early R&D Thors, named Douglas Missile 18, or DM-18, were powered by 135,000 lbf Rockeydyne MB-1 engines and topped by dummy nose cones. Thor 101 blew up on its LC 17B launch pad on January 25, 1957. This was followed by three more failures. Finally, on September 29, 1957, still less than two years after the program began, Thor 105 flew a successful long-range flight from the Cape. One month later, Thor 109 flew the first full-range flight down the Atlantic Missile Range.

All-inertial guidance system test flights began in December, 1957. Reentry vehicle test flights began in February 1958 with Thor 120, which also debuted a new, less-tapered guidance section.

Thor DM-18A, the operational variant without fins, numbered Thor 138 and higher, began flying in 1958. The first crew training launch from Vandenberg, which was also the first long-range launch from that West Coast base, took place on December 16, 1958, when Thor 151 flew successfully over the Pacific Ocean.

Thor became operational in Great Britain during December 1959. There, 60 Thors were deployed at four former airfield bases. A total of 1,000 personnel manned each base. Thors were stored horizontally in retractable steel shelters on an erector-launcher mount. Missiles could be launched within 15 minutes of an order, in theory.

Thors were retired from IRBM duty in August 1963 and returned to the United States where nearly all would be assigned to other duties.

2. Thor Able

"Thor-Able" (orginally proposed as "Thor-Vanguard") was a Thor first stage topped by Vanguard second, and in some cases, third stages. The second ("Able") stage was a pressure-fed nitric acid/UDMH stage powered by an Aerojet AJ-10 series engine. The third stage was a spin-stabilized Allegany Ballistics Lab X-248 "Altair" series solid motor.

The rocket, in two-stage form, was created by Ramo-Wooldridge Space Technology Laboratories (as the prime contractor) to test ICBM ablative reentry vehicle techology before Atlas and Titan were ready. In 1958 it flung GE Advanced Reentry Test Vehicles more than 10,000 km to prove the ablative technology. It was the first time that a big USAF liquid fueled rocket had staged successfully. The two-stage rocket was controlled by an autopilot rather than an active guidance system.

Someone at STL quickly realized that Thor-Able could be turned into an ICBM (briefly named "Thoric") - that it would be able to carry the lighter RVs and warheads soon to appear. The idea was rapidly quashed by the chain of command, which also removed STL from the Thor-Able prime contractor role in favor of Douglas.

Thor's first orbital attempts were part of ARPA's "Operation Mona", better known as "Pioneer" - the name assigned by NASA when it assumed control after the first launch. Pioneer was the first U.S. attempt to reach the Moon. Adding radio guidance to the Able second stage and an ABL X-248 third stage created the "Thor-Able 1" variant to do the attempt. Three attempts in 1958 failed to reach the moon, but did manage to loft Pioneer 1 to record altitude where it collected data on the Van Allen belts.

"Thor-Able 2" was a two-stage variant with BTL guidance in the Able stage that boosted GE "RVX-1" (or "Precisely Guided Reentry Test Vehicle") scaled ablative reentry vehicles on multiple ICBM-distance flights in 1959. RVX-1 looked a lot like the Mark 3 and 4 RVs that later topped Atlas and Titan ICBMs. The Navy managed to recover some of these, proving the design.

Thor-Able 2 with a third stage orbited NASA TIROS 1, the first weather satellite, in 1960. Thor-Able's 3 and 4 orbited NASA's Explorer 6 and Pioneer 5 in 1959 and 1960, respectfully. These used an improved AJ10-101A second stage engine.

Thor-Able served as the starting point for NASA's highly successful Thor-Delta, which first flew in 1960.

3.  Thor-Agena A

At the dawn of the Space Age the highest priority U.S. space program was U.S. Air Force Weapons System 117L, also known as the Advanced Reconnaissance System. One component of WS-117L was the "Corona" photographic "spy" satellite, which returned exposed film in small "satellite recovery vehicles".

For the program, Lockheed Missiles and Space Company developed a new upper stage named "Hustler", later renamed "Agena". Agena A was more than a rocket stage, it was an orbiting platform for Corona - a spacecraft in its own right. It was powered by a Bell 8048 UDMH/IRFNA turbopump-fed engine. The engine was not restartable, so after separating from Thor the stage would coast for a couple of minutes to apogee before beginning its burn. Agena's forward section housed an inertial guidance system that used horizon sensors to provide updates. Cold-gas thrusters located in its aft section provided flight control.

Fifteen Thor-Agena A launches occurred between February 28, 1959 and September 13, 1960. All launched toward near-polar orbits from converted Thor IRBM pads 75-3-4 and 75-3-5 at Vandenberg AFB. The missions were given the "Discoverer" cover name. Discoverer was said to be a scientific research effort, but it was actually a Corona development program.

Development was hard-won. One Agena A stage was destroyed even before the first flight in the "Discoverer Zero" pad accident on January 21, 1959.  A sneak circuit triggered during a pre-launch test caused the stage to fire its ullage rockets. Agena 1019 was subsequently scrapped, but the Thor 160 booster was refurbished and used on the Discoverer 12 flight in 1960. 

Six of the 15 Thor-Agena A launches failed to reach orbit, and a seventh launch failed to achieve the proper orbit due to a guidance system failure. Brand-new Agena was most-often the culprit. Not until Discoverer 13, flown in August 1960, would a capsule be orbited and successfully recovered - the first man-made object recovered from space. Discoverer 14 returned film containing images taken by the spacecraft's Keyhole camera. The images covered 1.5 million square miles of Warsaw Pact territory and revealed the presence of 64 previously unknown airfields, 26 surface to air missile sites, and a previously unknown launch center at Plesetsk.

4. Thor DM-18C

Thor DM-18C was a special test vehicle that was powered by an upgraded Rocketdyne MB-3 Block 2 (LR-79-NA-11) engine that produced 165,000 pounds thrust. A GE low drag fairing topped the missile, covering the standard GE Mark 2 reentry vehicle. DM-18C tested the improved, higher thrust engine and demonstrated modest range improvement. Three launches, all from Cape Canaveral LC 18B, took place during January-February, 1960 (Thors 256, 259, and 263). All were successful.

MB-3 Block 2 would soon begin to power Thor space launch variants.

5.  Thor Able-Star

In 1960, ARPA and the U.S. Air Force began flying Thor-Able-Star, originally named "Thor-Epsilon" which used the Aerojet General "Able-Star" pressure-fed hypergolic second stage powered by an AJ10-104 series engine. Developed for primary customer the U.S. Navy, Able-Star was the world’s first re-startable stage. It was a fat version of the Able stage, carrying more than twice as much propellant.

Thor's tapered guidance section was replaced by a stepped interstage adapter. MB-3 Block 1 150 Klbf engines apparently powered the initial Thor boosters, with MB-3-2 165 Klbf engines replacing them at some point. A lightweight STL guidance system topped the second stage. Nitrogen jets provided three-axis flight control of the stage during coast periods.

Thor-Able-Star flew 19 times during 1960-65, including 11 launches from the Cape and 8 from Vandenberg AFB. It performed the first in-space stage restart during its first flight on April 13, 1960. After the stage completed its initial 258 second burn, it and its Transit 1B payload coasted for 19 minutes before the stage performed a second, 13 second long burn to raise the orbit.

Early flights orbited the U.S. Navy's initial Transit (navigation) and U.S. Army's Courier (communications) satellites, along with Solrad/GRAB electronic intelligence radar signal "spy" satellites flown piggyback with Transit. The final two of six Cape-launched Transit satellites were powered by SNAP 3B nuclear power sources (radio-isotope thermoelectric generators or "RTGs") - the first time that RTGs were launched into orbit. Thor-Able-Star also orbited ANNA 1B (Army, Navy, NASA, Air Force) a satellite that carried beacons for use in ground surveying.

Thor-Able-Star flew from Vandenberg AFB Complexes 75-1-1 and 75-1-2 during 1963-65. The sites were later renamed Space Launch Complex (SLC ) 2 East and 2 West, respectively. All eight Vandenberg launches carried Transit navigation satellites, aimed toward near-polar orbits. They used SNAP 9A RTGs, which were loaded with more Plutonium 238 than the SNAP 3B RTGs. The SNAP 9A design was discontinued after the Transit 5-BN-3 flight failed to reach orbit due to an Able-Star stage failure. The RTG disintegrated in the atmosphere releasing about 1 kg of Plutonium 238. The final five Transit-O ("Oscar") missions, all successful from a launch vehicle perspective, used solar powered satellites.

Thor Able-Star's early flight record was spotty, with five launch vehicle failures, two by Thor and three by Able-Star, during the first 10 flights. But only one of the final nine Thor Able-Stars, the one that carried the last nuclear powered Transit, failed, and the Thor first stage itself flew successfully during the final 14 flights.

Although August 13, 1965 saw the last flight of Thor-Able-Star, the basic stage structure would subsequently migrate to NASA's ever-improving Delta launch vehicle.

6. Thor-Delta

NASA’s Milton Rosen coined the rocket name "Delta". He chose “Delta” because it would be the fourth Thor-based space launch vehicle after Thor Able, Thor Able-Star, and Thor Agena. Rosen had led the Naval Research Laboratory’s Viking sounding rocket and Vanguard satellite programs. He transferred to NASA in October 1958 along with 157 NRL Vanguard program employees to form the Agency’s new Goddard Space Flight Center.

Thor-Delta was an adaptation of Thor-Able 2 (originally "Thor-Vanguard"). It was meant to serve as an "interim" launch vehicle for NASA, until larger Atlas launch vehicles (Atlas Vega and Atlas Centaur were the plan at the time) were brought on line. A cold gas-jet attitude control system was added to the previous Able second stage to create Delta. With this system, Delta could coast and reorient itself in space after its pressure-fed AJ-10-118 engine had performed its burn. This improved the accuracy of solid fuel third stage spin-up insertions. Both upper stages were tweaked and weight was shaved from Thor itself. A Bell Telephone Laboratories BTL-300 radio guidance system was added in an equipment compartment atop the second stage to control the vehicle. Typical missions saw the second stage coast for several minutes after its burn before aiming and spin/separating the ABL X-248-A5 third stage.

Goddard ordered 12 Thor-Deltas from Douglas Aircraft and the other stage contractors in April 1959. The first, carrying Echo 1 from Cape Canaveral’s Complex 17A on May 13, 1960, failed. A re-try in August with Echo 1A succeeded. So did the third, and fourth, and so on until all of the final 11 were successful. They orbited five Tiros, two Explorer, one Orbiting Solar Observatory, Ariel 1 the first U.K./U.S. satellite, Echo 1A, and Telstar 1 the famous experimental AT&T active repeater communications satellite. With such unprecedented success, NASA removed the "interim" label and ordered more, improve Thor-Deltas, about which more later.

7. Thor-Agena B

More-capable Thor Agena B began flying on October 26, 1960. It launched 43 times, failing eight times, during its five-years of service. It used an upgraded DM-21 Thor first stage powered by an MB-3 Block 2 (initially) or Block 3 engine that produced 165 to 170 Klbs of liftoff thrust. The original Thor guidance section was replaced by a shorter, lighter adapter section. The Agena B second stage, which was powered by the restartable 16 Klbf thrust Bell 8081 (initially) or 8096 engine engine, was 60 inches in diameter and weighed about 14,770 lbs fueled. It carried two times more propellant than Agena A and used a more powerful, more efficient engine that also gained payload capability by being able to restart. Like Agena A, Agena B was a spacecraft in its own right, able to maintain attitude while operating on batteries for up to two weeks.

The improved launch vehicle orbited Keyhole 2 through 5 (mostly) film return spysats and three "Ferret" electronic intelligence satellites. It also flew several times for NASA, orbiting second generation weathersat Nimbus 1, Canadian ionespheric satellites Alouette 1 and 2, and NASA's Echo 2 and Explorer 31. Alouette 1, the first non-U.S. or U.S.S.R. built satellite, operated for a then-unheard-of decade.

Most Thor Agena B missions, which either went to LEO or to slightly elliptical orbits with low perigees, used a single restart for a total of two Agena burns. The first burn was usually something like 234-ish seconds duration, while the second burn at first apogee would only last a few seconds. Agena B was the second stage to restart in orbit (Able Star was first), but it was the first turbopump powered stage to restart.

Thor Agena B flew from three VAFB pads, flying 17 times in 1961 and 18 times in 1962.

Most of the NASA launches took place a year or two after Thor Agena B was done with its Pentagon work. The NASA launchers appeared to use Agenas that were similar to the three Ferret launchers, with longer cylindrical equipment sections rather than the partially tapered sections used by the KH Agenas.

8. Thor DSV-2D

Two Thor DSV-2D vehicles launched suborbital "Big Shot" missions from Cape Canaveral in 1962. They boosted 135 foot diameter Echo balloons above the atmosphere to test balloon deployment methods for subsequent planned orbital launches. This was NASA's Applications Vertical Test Program (AVT), better known as "Big Shot". A cylindrical Equipment Section, topped by a DA-92 shroud built by Douglas, were added to the now-standard DM-21 Thor, which was also used as the first stage for Thor Ablestar and Thor Agena B. After the boost phase, the shroud and a balloon deployment canister were jettisoned at an altitude of about 250 nmi. During the coast to 1,000 nmi, the balloon inflation experiment was performed while TV and film camera's in the Equipment Section recorded the result. During the first launch, the balloon ruptured during inflation. The second launch produced a successful inflation.

9. Thor DSV-2E

DSV-2E Thors were DM-18A IRBM Thors modified to perform live exoatmospheric thermonuclear warhead launches from Johnston Island in the Pacific Ocean as part of Operation Fishbowl during 1962. The Thors carried instrumented "pods" attached to the side of their propulsion sections. The pods were released at intervals during the boost phase to gain differential separation from the exploding W49 or W50 warhead. After reentering, the contaminated pods floated beneath parachutes to the Pacific and were recovered. The Thors flew from a pair of tactical launchers (Launch Emplacements 1 and 2) set up near one corner of the tiny island. During launches, most Johnston Island personnel had to be evacuated to ships standing offshore.

The tests taught the U.S. about the effects of exoatmospheric nuclear explosions, both on orbiting satellites and on ground-based communications and power systems. The project produced dazzling nuclear effects, but it also suffered a series of disastrous failures. There were eight DSV-2E launches, seven with live warheads. Four of the seven "live" launches failed.

The first Fishbowl launch was a successful R&D flight with no warhead. The second launch, carrying an active warhead, was "lost" by a defective range safety tracking radar and had to be destroyed 10 minutes after liftoff. Three subsequent Thors, all carrying nuclear warheads, suffered propulsion system failures and had to be destroyed by range safety. Two of those destructions occurred downrange, a minute or more into flight, dropping some radioactive contamination on and near Johnston Island. The third failure, on July 25, 1962, was a true Cold War disaster.

Thor 180, the missile for that "Bluegill Prime" shot attempt, was fitted with a W50 thermonuclear warhead capable of producing a 400 kiloton explosion. A propellant valve stuck at ignition, causing a leak that fed a rapidly expanding fireball that enveloped Thor on its launch pad. The range safety officer fired the destruct system, destroying the Thor, the warhead, and the launch emplacement, which burned for some time, contaminating the island. Despite several subsequent cleanup efforts, Johnston Atoll, managed by the U.S. Fish and Wildlife Service since 2000, is still affected.

In the end, Operation Fishbowl only produced three successful high altitude explosions. One of these, Starfish Prime on July 9, 1962, was a 1.4 megaton explosion, created by a W49 warhead at an altitude of 400 kilometers. It created a fireball and artificial aurora visible in Hawaii, along with an electromagnetic pulse that disrupted power and communications. It also pumped enough radiation into the Van Allen belts to destroy or seriously degrade seven orbiting satellites.

Two attempts took place during the midst of the Cuban Missile Crises, on October 16 and 26, 1962. The latter Bluegill Triple Prime shot, which detonated a W50 warhead at 48 km, almost unbelievably took place while SAC was at DEFCON 2.

The final Fishbowl launch carried the "Kingfish" 400 kiloton warhead up to its 98 km detonation altitude. Kingfish was one of the last above-ground U.S. nuclear tests, because the U.S. and the Soviet Union signed an atmospheric test ban treaty shortly thereafter.

10. Thor-Agena D

Thor-Agena D used a "standardized" Agena D upper stage that was designed to fly atop Thor, Atlas, and Titan with minimal changes.

Agena D used an improved Bell 8096 restartable engine, still producing 7.26 tonnes thrust but now with higher specific impulse. It also used a Bell Telephone Laboratories (BTL) 600 radio guidance system.

Thor-Agena D flew 21 times during 1962 to 1967 from Vandenberg AFB pads 75-1-1, 75-1-2, 75-3-4 and 75-3-5, carrying Keyhole 4 and 5 film return spysats, DSAP Block 1 military weather satellites, Poppy electronic intelligence satellites, and experimental U.S.Navy satellites among others. Four launches failed, and a fifth placed Poppy 1A/1B in a too-high orbit when Agena failed to cut off as planned. The satellites still functioned, but the orbit limited data collection.

11.  Delta A, B, C, C1

NASA/Goddard couldn't ignore the success of the original twelve Thor-Deltas, so it ordered more. These featured incremental improvements, introduced pretty much "one at a time". The method worked. Only 2 of these 24 Deltas would fail outright.

Delta A introduced MB-3 Block 2 engines that produced 170 klbf liftoff thrust. Delta A, which flew twice in 1962, also featured a shorter interstage between Thor and Able to shave weight. The improved Delta actually stood 4 to 5 feet shorter than the original Thor-Delta.

Delta B, which flew nine times during 1962-64, introduced a 36 inch second stage stretch and improved AJ-10 engine performance. This variant launched Explorer 17, TIROS 7 and 8, Relay 1 and 2, Telstar 2, and Syncom 1 and 2. Syncom 1 was the first launch to GTO, though the satellite was lost during its apogee motor firing. Syncom 2 succeeded, reaching an inclined geosynchronous orbit.

Delta C added a more powerful ABL X-258 “Altair 2” third stage during its 11 flight, 1963-67 run. It orbited more Explorers and a string of early solar observatories and weather satellites. Remarkably, Delta boosted the TIROS 9 and 10 and ESSA 1 weathersats into near sun synchronous orbits - from Cape Canaveral, Florida! The flight paths doglegged south, crossing Cuba and Panama before the third stage fired over the equator just northwest of South America to complete the insertion. (Delta would not fly from Vandenberg AFB until 1966.)

Delta C1 added an even more potent United Technologies FW-4D third stage motor during its two launches. Delta 64, the 38th and final Thor-Delta with the 32 inch diameter Vanguard-derived second stage, orbited NASA’s fifth Orbiting Solar Observatory (OSO 5) on January 22, 1969. This was the final Thor-Delta to fly without strap-on solid motors. The old Vangaurd stage scored its 37th consecutive "Delta" orbital success on that flight, a relic of the early Space Age surviving to fly even after NASA had launched astronauts atop massive Saturn V boosters a few miles up the Florida coast.

12.  TAT Agena

Thor-Agena was beefed up with the addition of three Thiokol Castor 1 solid rocket motors beginning in 1963. These "Thrust Augmented Thor" (or TAT) boosters lifted two Agena B, and 61 Agena D, stages with payloads toward orbit from 1963 until 1968. The Castor 1 motors, derived from the Sergeant missile motor, nearly doubled the liftoff thrust compared to Thor-Agena. The solids burned for about 40 seconds, with the last dozen seconds comprising a tailoff. They were jettisoned at T+65 seconds to reach a safe drop zone. The boosters augmented the upgraded MB-3 Block 3 Thor first stage engine, which itself burned for nearly 150 seconds.

TAT-Agena D would become the most-oft flown U.S. Air Force Thor space launch vehicle. It could lift roughly 1.5 tonnes to polar orbit, including the Agena stage. The vehicle's busiest year was 1964, when 20 launches occurred from Vandenberg.

Five pads, 75-1-1, 75-1-2, 75-3-4, 75-3-5, and PA-1-1, a former Atlas Agena pad on the U.S. Navy test facility at Point Arguello (incorporated into Vandenberg as South Vandenberg after 1964), handled TAT Agena D launches.

For the first time, Keyhole 4A imaging satellites equipped with two film return "buckets" were flown, accounting for the majority of launches. TAT-Agena D also orbited electronic intelligence satellites, the Quill 1 experimental radar mapping satellite, and three NASA payloads (OGO 2 and 4 and PAGEOS). TAT-Agena B orbited NASA's NIMBUS 2.

13. Thor ASSET

Thors lofted six ASSET (Aerothermodynamic/elastic Structural Systems Environmental Tests) lifting body reentry experiments on suborbital flights from Cape Canaveral/Cape Kennedy during 1963-65. These were U.S. Air Force missions that evaluated reusable, maneuverable, re-entry vehicle designs that might be able to fly to a precise landing point on earth. McDonnell Aircraft of St. Louis built the ASSET vehicles. The original reason for the program was to support X-20 development, but ASSET continued after X-20 was cancelled. One of the ASSET reentry vehicles was recovered after parachuting to an ocean landing. Two other recoveries were attempted. Significant telemetry hauls were made even when the vehicles were not recovered.

Single-stage Thors, retired UK IRBMs returned to Tulsa, Oklahoma for refurbishment, performed three of the flights as "DSV-2F" variants. Two-stage vehicles that also used retired Thor first stages topped by what were essentially Delta B type second stages performed the other three flights, as "DSV-2G/Delta" vehicles. The Delta second stage was only partially loaded with propellant for a relatively short 50-60 second burn. The only mission failure occurred during the first DSV-2G launch when the second stage failed to ignite. The final ASSET launch on February 22, 1965 was the last U.S. Air Force Thor flown from Cape Canaveral.

14. Thor DSV-2J ASAT

Thor DSV-2J vehicles were retired Thor DM-18A IRBMs that were refurbished for use in ASAT Program 437 and its follow-ons. A total of 17 of these Thors performed suborbital flights between 1964 and 1975. The launches were from two Johnston Island launch emplacements originally built for Operation Fishbowl.

Program 437 Thors were designed to pass within 3 nmi of an orbiting satellite where the missile's 1.44 mT W49 warhead would destroy the satellite. Test launches were performed with dummy warheads against orbiting U.S. upper stages and satellites. A total of nine launches took place. During the initial tests, two Thors would be counted down simultaneously to ensure that at least one would meet the short launch window. (A sizable staff was required. Personnel rotated between Johnston Island and a training pad at Vandenberg AFB.) The ASAT system stood active watch from 1964 until 1970 (longer than Thors had stood IRBM duty), then was placed on disassembled standby until 1975.

Program 437AP (Alternate Payload) carried a camera system rather than a warhead, to photograph orbiting satellites. The camera and film reentry vehicle was adapted from Corona KH-4. Only four launches took place during 1965-66, two of which apparently succeeded in photographing orbiting U.S. objects.

After the ASAT program stood down, two one-off launches took place in 1970. The first, for Program 922 (former 437Y) was an ABM sensor test against a Minuteman 2 RV launched toward Kwajalein. The Thor shut down 6 seconds early, preventing an intercept, which would not have occurred regardless because, first, the payload and Thor collided after separation and, second, the Minuteman RV failed to separate from its upper stage! I haven't seen any photos of this Thor, but it was said to have a separable maneuvering payload.

The second launch, for the High Altitude Program (HAP), was completely successful. This Thor's heavy payload, which created a simulated nuclear explosion for an on-board X-ray detector to sense, was housed in an Agena-like shroud, which required installation of a special service tower at LE-2. (I wonder if the tower might also have supported the 922 launch.) A reentry pod returned film or data or both.

After four years of non activity, Johnston Island hosted two final Thor launches in 1975. These BMDTTP (Ballistic Missile Defense Test Target Program) launches served as targets for Kwajalein ABM radars. I haven't seen any photos of these vehicles, but I have read one report that they were in their IRBM configuration. They were the final suborbital Thor launches.

15. Delta D (TAD)

Goddard Center adopted solid motor thrust augmentation to Delta about 1.5 years after it was proven by the U.S. Air Force TAT-Agena D. Adding three Castor 1 SRMs to the DSV-3C Delta C model created "Delta D" (DSV-3D), also known as "Thrust Augmented Delta" (TAD). The engine skirt and engine section were slightly modified to support the SRM loads. TAD jumped off its pad with a nearly 2.37 thrust to weight ratio. This variant only flew twice, but both launches were historic.

Delta 25 boosted Syncom 3 to GTO on August 19, 1964. The extra boost allowed for a 16.5 deg GTO. Syncom 3's own apogee kick motor and thrusters then had enough energy to boost itself into the first-ever geostationary orbit. The satellite was positioned above the International Date Line, where it relayed coverage of the 1964 Tokyo Olympics.

Delta 30 performed a similar launch, on April 6, 1965, for "Early Bird", the first commercial communications satellite. Early Bird, or Intelsat 1, was similar to the first three Syncoms, but provided 240 telephone circuit equivalent service (or one TV channel). This garbage-can size satellite substantially increased trans-Atlantic circuit capacity, decisively changing how long-range telecommunication service would subsequently be provided. Early Bird functioned for four years before being retired.

16. Thor-Burner

Thor Burner 1 (originally "Thor Altair"), was developed to orbit Defense Meterological Satellite Program missions (originally Defense Satellite or Systems Application Program) to support NRO Corona missions. The early DSAP satellites, built by RCA, were derived from the original TIROS spinners. Scout, intended to launch these satellites, was still struggling at the time (three of five DSAP launches failed), so plans were made to top refurbished retired Thor IRBMs with Scout "Altair" fourth stages and payload "heat shields". The move was costly for LTV and NASA, which saw eight planned Scout launches canceled.

Two types of motors and heat shields ended up flying atop Thor Burner 1. The first two, in 1965, used Lockheed/Grand Central Rocket Co. MG-18 motors and 25.7 inch diameter heat shields from the two already built Scouts assigned to the program. The final four in 1965-66 used UTC FW-4S Altair 3 motors and 34 inch diameter heat shields.

Douglas Aircraft replaced the Thor guidance section with a shorter, lighter set of adapters and swapped inertial for lighter BTL radio guidance. The company also added a cold gas attitude control system atop Thor to provide stability during the coast to apogee and the proper attitude for second stage spin-up and separation. I'm looking for details on this system, which might only have been used by the FW-4S vehicles that went to higher orbits.

Launches took place from VAFB 4300 B6 (former 75-2-6, later SLC 10W) to boost the tiny satellites into sun synchronous orbits. SAC's 4300 support squadron performed the launches. Four of the six launches were successful. The payload heat shield failed to separate during the first launch and the second stage motor failed to start during the fifth flight.

Subsequent "Thor Burner 2" types continued the DMSP launches until 1980, as we shall see.

17. Delta E-H (TAID)

After Thrust Augmented Delta, the next logical step was to take advantage of all that thrust by increasing the second stage mass. That step, taken in 1965, created Thrust Augmented Improved Delta (TAID). Douglas borrowed the Able Star tanks, stretched them, and added an improved AJ10-118E engine to create the second stage. This roughly doubled the burn time compared to the previous Vanguard-based Delta stage, increasing payload to orbit. The TAID second stages performed a single burn, then provided attitude control during a coast prior to spin-up and separation of the third stage.

Both Castor 1 and Castor 2 motors could be used with the TAID series. Castor 2 provided slightly less liftoff thrust but burned slightly longer. An MB-3 Block 3 Rocketdyne engine powered Improved Delta's Thor first stage. MB-3-3 improved reliability and a bit more thrust.

Third stage options included the ABL-258 and FW-4D spin-stabilized solid motors. With ABL-258, Improved Delta was called "Delta E". With FW-4D it was "Delta E1". The rocket was identified as "Delta G" when no third stage was used. "Delta J", with a Star 37D third stage motor, flew once. (Unflown types included "Delta F/F1", which was Delta E/E1 without SRMs, and "Delta H", which was Delta G without SRMs.) An Agena-style 65 inch diameter payload fairing topped the rocket, providing much more internal volume than earlier Delta shrouds.

Twenty six TAID launches took place during 1965-71. Every single one succeeded. Launches took place from Cape Canaveral and Vandenberg AFB, marking Delta's introduction to the West Coast.

Payloads included Intelsat 2 communication satellites, NASA Explorer and Pioneer satellites, ESSA weather satellites, HEOS 1, ISIS 1-2, and Biosat 1-2. The Intelsats separated into GTO and raised themselves into GEO. The Pioneers went into solar orbits. Most of the Explorers were launched into highly elliptical Earth orbits. Explorer 35 (IMP-E) inserted itself into lunar orbit after a precise TAID launch. Delta boosted HEOS 1 into a 440 x 230,000 km x 28.3 deg orbit. The satellite's apogee kick motor subsequently raised its perigee to 6,800 km.

TAID was the first Delta to sport the soon-familiar "Delta" triangle logo.

18. Thorad Agena D

In 1961, Douglas Aircraft proposed a series of Thor upgrades for space launch. The "Thor Advanced" (Thorad) concept called for a constant diameter airframe that would eliminate the tapered IRBM LOX tank and increase propellant capacity. Three Sergeant solid motors would augment thrust. Before Thorad would fly, Douglas would add the solids to a standard Thor to create "Thorad Junior" (better known as TAT Agena-D). Replacing the MB-3-3 engine with a higher thrust Saturn H-1 engine would create "Thorad B". These proposals were eventually realized.

Thorad-Agena D (also Long Tank Thrust Augmented Thor-Agena D), had the constant 96 inch diameter airframe and a 14 foot stretch. Stage weight increased 45% to nearly 70 tonnes and burn time increased to 218 seconds MECO/227 seconds VECO. Three Castor 2 SRMs roughly doubled the liftoff thrust provided by the first stage engine. A new adapter section topped the stage, allowing continued use of the existing transition section and Agena D adapter. This resulted in a distinctive three-step taper between the first and second stages.

There were 43 Thorad Agena D launches, with three failures, from Vandenberg AFB between 1966 and 1972. Launches took place from SLC 1W, 1E, 2E, and 3W, the former 75-3-4, 75-3-5, 75-1-1, and PA-1-1. Payloads included double bucket Keyhole 4A and 4B satellites, Poppy and Strawman signals intelligence satellites, Nimbus weather satellites, OGO 8, and the remarkable SERT 2.

The Nimbus satellites were powered by SNAP-19 RTGs. On May 18, 1968, Nimbus B was lost when Thorad's control system failed about two minutes into the flight. Unlike earlier SNAP RTGs, SNAP-19 was designed to survive launch vehicle failure, and it did. The RTG was eventually salvaged from the Pacific and its nuclear material reused.

Two Thorad-Agena D variants are listed. SLV-2G (1966-71) used a DSV-2L first stage. SLV-2H (1969-72) used a DSV-2L-1A first stage. The stages were identical in external appearance. It is possible that the "1A" was only added to differentiate Thorad-Agena D from Thorad-Delta (which used the DSV-2L-1B stage). At the time, newly merged McDonnell Douglas was shifting manufacturing from the original Santa Monica Thor factory to Huntington Beach, where the Thorad stages came off a common production line.

Thor 571, launched May 25, 1972, was the final Thorad-Agena D and the last U.S. Air Force Thor-based stage manufactured, although space launches of converted Thor IRBM missiles would continue for several more years.

19. Thor-Burner 2(A)

In March 1964, the DMSP program office approved plans to develop a more powerful Thor Burner 2 launch vehicle that still used repatriated Thor IRBMs.

Burner 2 used a Thiokol Star 37B motor (TE-M-364-2, a modified Surveyor retro-rocket motor) to power the "Burner 2" second stage. Boeing's Burner 2 stage was built around the Star 37B. It had a strap-down inertial guidance system and a 3-axis reaction control system, allowing it to coast without spin stabilization. Four 10 kgf hot-gas hydrogen peroxide thrusters performed stage separation, provided pitch and yaw reaction control thrust during the Star 37B motor firing, and completed a vernier maneuver immediately after the Star 37B burn. Eight 1 kgf gaseous nitrogen cold-gas thrusters on the stage provided pitch-yaw-roll attitude control during coast and performed spacecraft spinup and post-spacecraft separation maneuvers. Thor Burner 2 was topped by a new Goodyear conical phenolic shroud that enclosed the upper stage and payload. The unpainted fairing was distinctively vermilion (orange-red) in color.

Thor Burner 2 flew 12 times from September 16, 1966 to June 8, 1971, carrying 10 DMSP Block 4 and 5A satellites and performing a pair of U.S. Air Force Space Test Program (STP) missions. All 12 launches were successful. One of the STP missions, flown on June 29, 1967, used a Star 13A apogee kick motor to insert Aurora 1 and SECOR 9 into a 3,792 x 3,947 km x 90.1 deg polar orbit.

Thor Burner 2A added a third stage and a modified fairing to the Thor Burner 2 design. A Star 26B motor served as the third stage motor. The 3-axis Burner 2A control bus was built around the Star 26B. A Star 37B motor served as the second stage, with the 3rd stage bus providing guidance and control during its burn. The shroud was extended by the addition of a cylindrical section. Thor-Burner 2A performed eight launches with DMSP Block 5B and 5C satellites between October 14, 1971 and February 19, 1976.

The final launch failed because the Thor was not loaded with enough kerosene fuel, causing Thor to burn out a few seconds early. The upper stages performed their burns, but the end result was insufficient velocity to maintain a stable orbit. A old mixture ratio typographical error on the LR79 main engine certification testing data sheet was deemed responsible for the improper fuel load.

20. Long Tank Delta

In 1968, NASA gained access to the stretched Long Tank Thor stage already proven as an Agena booster. Long Tank Thor served as the first stage for Delta Models L, M, N, M6, and N6, which all used the TAID second stage. L, M, and N used a trio of Castor 2 strap-on solid motors to augment liftoff thrust. M6 and N6 used six SRMs with three ground-lit and three air-lit. All motors jettisonned in sets of three beginning after T+90 seconds.

Delta L used the FW-4D third stage motor. Delta M, the most often-flown Long Tank Thor Delta model, used the more powerful Star 37D third stage motor. Delta N did not have a third stage. Deltas M6 and N6 were the six SRM versions (also called "Super Six") of Deltas M and N.

Delta 58, an "N" carrying the Tiros 17 weather satellite from Vandenberg on August 16, 1968, was the first Long Tank Thor Delta (also called "Thorad Delta" or "Long Tank Thrust Augmented Thor Delta"). Altogether, there were twenty-four L, M, N, M6, and N6 flights, with four failures, during 1968-72. They launched eight Intelsat 3, two Skynet 1, two NATO 2, two ESSA, two OSO, and three ITOS/NOAA satellites. Single launches included HEOS 2, IMP I (Explorer 43), Biosat 3, and TD-1A.

Delta 59, the first "M" with Intelsat 3-1, failed on September 18, 1968 from Cape Kennedy. The rocket suffered a pitch rate gyro failure that became noticeable about 20 seconds after liftoff. It began to break up at T+102 seconds. The range safety officer sent a destruct command 6 seconds later.

Delta 71, another M, left Intelsat 3-5 in a useless orbit on July 25, 1969 when its Star 37D third stage motor either suffered a motor case rupture or a nozzle failure during its burn.

Delta 73, the first Delta L, failed on August 27, 1969 when it attempted to launch Pioneer E from the Cape. This time the culprit was an unstable high pressure relief valve in the MB-3-3 first stage power pack. Pressure fluctuations caused a line to rupture and leak hydraulic oil. First stage main engine gimbal control was lost 213 seconds after liftoff, during the latter portion of the first stage burn. The second stage separated and ignited, but was too far off course to make up the lost velocity. Range safety sent a destruct command at T+ 8 minutes 3 seconds.

Delta 86, an N6, failed on October 21, 1971 when its second stage suffered an oxidizer leak. The stage tumbled out of control after its attitude control system fought the side thrust from the leak. The control system finally used up its supply of control gas.

Delta 85, an N, nearly failed after launched from the Cape on September 29, 1971 with OSO 7. During the AJ10-118E second stage engine's second burn, the stage suffered a control system failure, caused by a nitrogen pressure leak that cascaded into a main engine gimbal thrust vector control (TVC) hydraulic pressure decay. (The hydraulic pump was run by pressurized nitrogen gas during the coast phase prior to the burn.) The stage tumbled, but it and OSO-7 still managed to achieve a usable low earth orbit. Ground crews stablized OSO 7 after it separated, a "save" that allowed it to perform its mission.

Delta 88, launched on March 12, 1972 with Europe's TD-1A science satellite, was the last "N" and the final launch from Vandenberg AFB SLC 2E. Delta 88 used a transitional Long Tank stage equipped with the first "Universal Boat Tail" - a beefed up aft thrust structure equipped with mounting points for nine solid motors. The change was part of the transition from Long Tank to Extended Long Tank Delta that began in 1972.

Two transitional Long Tank Thor Delta models flew in 1972-73. Delta 300 used three Castor 2 strap on boosters and a modified second stage powered by a more-powerful AJ10-118F engine derived from the Titan 3 Aerojet Improved Transtage Engine Program (ITIP). It burned nitrogen tetroxide and Aerozine 50 (a 50-50 mix of UDMH and hydrazine) rather than the previous nitric acid/UDMH. Delta 900 used nine Castor 2 boosters (six ground-lit) and the same second stage. Both models used the Universal Boat Tail. Neither flew with a third stage. Both used the Delta Inertial Guidance System (DIGS). Prior Deltas had used radio-inertial guidance.

Three Delta 300 and two Delta 900 launches took place from Vandenberg AFB. One (Delta 96) failed to orbit ITOS E from Vandenberg on July 16, 1973 when a hydraulic pump failed 270 seconds after the second stage ignited. The pump failure led to loss of thrust vector control. Successes included NOAA 2 and 3, ERTS 1, and Nimbus 5.

Long Tank Thor ended service as an Air Force Agena launcher on May 25, 1972. The final Long Tank Thor Delta launch occurred several months later on November 6, 1973 when Delta 98, a Delta 300 model, orbited NOAA 3 from Vandenberg's still-active SLC 2W.

During seven years of service, 72 Long Tank Thor launches occurred, boosting 29 Delta and 43 Agena missions. Long Tank served as the basis for the follow-on Extended Long Tank Delta stage.

21.  Extended Long Tank Delta

Extended Long Tank (ELT) Delta flew 93 times during 1972-1990, succeeding 89 times and orbiting all manner of important payloads.

Extended Long Tank included a roughly 120 inch first stage tank stretch compared to the Long Tank stage. It flew with both MB-3-3 and RS-27 engines. Rocketdyne's RS-27 was essentially a repackaged H-1 engine salvaged from the large inventory of unflown Saturn IB engines. A total of 83 RS-27 engines flew.

Three different second stages flew atop ELT Deltas.

The first was the AJ10-118F powered stage that was a holdover from Long Tank Delta. It was topped by the 65 inch diameter Agena shroud.

The second was powered by the TR-201 TRW Lunar Module descent engine derivative. This stage was the first to be suspended within an extended interstage cylinder, through the use of a "Miniskirt". The design allowed use of a 96 inch diameter payload fairing. It was informally named "Straight Eight" because, for the first time, the entire launch vehicle had a constant eight foot diameter.

The third was the AJ10-118K powered "ITIP" (Improved Transtage Injector Program) stage that began flying in 1982. This stage used fatter tanks originally developed for Japan's N-2 launch vehicle. Today's Delta 2 second stage is similar.

Spin-stablized Star 37D, 37E, and 48B third stage motors flew atop ELT Deltas aimed beyond LEO. During the Shuttle era when Delta served as an STS backup, NASA carded Delta 3910/PAM-D and Delta 3920/PAM-D variants. The PAM stage, a Star 48B spin-stable solid motor, was considered part of the "payload" just as it was on Shuttle.

Delta 147, launched on December 17, 1978, used the first Delta Redundant Inertial Measurement System (DRIMS). DRIMS improved the inertial measurement unit introduced with DIGS, but kept the DIGS guidance computer. DRIMS added redundancy on all axes of motion.

ELT Delta's used the four-number model identification system.

Delta Model Numbers

First Digit: First Stage and Strap on Motor Types

0: Long Tank, MB-3-3 engine, Castor 2 motors (1968)
1: Extended Long Tank, MB-3-3 engine, Castor 2 motors (1972)
2: Extended Long Tank, RS-27 engine, Castor 2 motors (1974)
3: Extended Long Tank, RS-27 engine, Castor 4 motors (1975)
4: Extended Long Tank, MB-3-3 engine, Castor 4A motors (1989)
5: Extended Long Tank, RS-27 engine, Castor 4A motors (1989)

Second Digit: Number of Strap on Motors

Third Digit: Second Stage Type

0: AJ10-118F (Aerojet Transtage derivative, 1972)
1: TR-201 (TRW LM Descent Engine derivative, 1972)
2: AJ10-118K (Aerojet ITIP engine, 1982)

Fourth Digit: Third Stage Type

0: No third stage
3: Star 37D (TE-364-3, 1968)
4: Star 37E (TE-364-4, 1972)
5: Star 48B (TE-M-799, 1989)

22.  Thor Star 37/Star 37/ISS

Thor Burner 2A could not lift DMSP 5D1, a much upgraded, heavier military weather satellite. Two Star 37 upper stage motors were needed to reach orbit. Thus, the final five Thor LV-2F flights, launched between September 11, 1976 and July 14, 1980, flew as Thor Star 37/Star 37/ISS launch vehicles. "ISS" stood for "Integrated Stage System", a hydrazine-based propulsion system on the satellite that provided 3-axis control during the solid motor burns and a final trim burn. A Star 37XE motor served as the second stage while a Star 37S-ISS acted as the third stage. A longer payload fairing with a blunter nose housed both stages and the payload. The upgraded launch vehicle could lift roughly 500 kg to the DMSP sun synchronous orbit.

The first four launches were good, but the July 14, 1980 finale was a disheartening failure. Refurbished IRBM Thor 304 flew true, and the first Star 37 burn looked good, but at Stage 3 startup all telemetry was lost. It was subsequently determined that connectors between the second and third stages had not disconnected due to a misalignment. When the Star 37S motor ignited, the wiring harness was jerked out of the third stage and satellite, killing the flight control system. The stage pitched down and failed to generate sufficient orbital velocity. It turned out that an incident during launch vehicle erection - a broken pin that caused the rocket to suddenly drop a few centimeters - had most likely caused the connector misalignment.

Thor 304 was the final Thor IRBM to fly, and the final launch from Space Launch Complex 10 West.

23.  N-1, N-2, H-1

From 1975 through 1992, it was possible to see "Delta" lookalike rockets liftoff from Tanegashima, Japan. A total of 24 Thor-based rockets, assembled in Japan under license from the U.S., flew for the National Space Development Agency of Japan (NASDA). There were three variants, all unique to Japan.

N-1 ("N" stood for "Nippon"), launched from 1975 through 1982, was an MB-3-3 powered Long Tank Thor Delta with three Castor 2 boosters. An MHI built LE-3 pressure-fed hypergolic engine powered the second stage. A Star 37N solid motor served as the third stage. An Agena shroud topped the rocket. N-1 could lift 1.2 tonnes to LEO or 0.36 tonnes to GTO. It flew 7 times with one failure. Notable successes included Japan's first geostationary orbit launch, of Kiku 2 (ETS-2) on February 23, 1977. The lone failure occurred on February 6, 1979 when the fifth N-1's Star 37N third stage collided with its Experimental Communications Satellite (ECS-A) satellite payload shortly after spacecraft separation.

N-2, which flew 8 times during 1981-87, used an MB-3-3 powered Extended Long Tank Thor stage augmented by nine Castor 2 strap on motors. The second stage was powered by a restartable Aerojet AJ10-118FJ pressure-fed engine (after NASDA's planned LE-4 engine stumbled during development). The stage used new fatter tanks that would later be adopted by NASA's 3920 series and later Deltas, including today's Delta 2. Star 37E served as a third stage motor. N-2 used DIGS inertial guidance. It could lift 2 tonnes to LEO or 0.73 tonnes to GTO.

H-1, which flew 9 times during 1986-92, introduced a new NASDA-developed common bulkhead liquid hydrogen fueled second stage that was powered by a brand new NASDA-developed LE-5 engine built by MHI and IHI. The rocket was controlled, for the first time, by an inertial guidance system developed in Japan. H-1 could lift 2.25 tonnes to LEO or 1.1 tonnes to GTO.

For a time during the post-Challenger accident period, McDonnell Douglas considered adopting Japan's upper stage, or at least the LE-5 upper stage engine, for use on U.S. Delta launch vehicles.

24.  Delta 2

During the mid-1980s, McDonnell Douglas shut down the long-running Delta production line as NASA moved payloads to Space Shuttle. The 1986 Challenger disaster changed everything. Soon, the U.S. Air Force was asking for an expendable launcher that could orbit the GPS satellites originally slated for STS. McDonnell Douglas won the resulting Medium Launch Vehicle competition over General Dynamics (Atlas K), Martin Marietta (Titan 3 Commercial), and Hughes (Jarvis).

After pondering use of Japan's H-1 liquid hydrogen upper stage, McDonnell Douglas pulled an early 1980s proposal off the shelf to create "Delta 2". The rocket was built around a stretched "Extra Extended Long Tank" first stage that was 148 inches longer than the "Extended Long Tank" version. It carried 96 tonnes of propellant, a 16 tonne increase.

The first, interim Delta 6000 series vehicles used RS-27 engines and upgraded steel-case Thiokol Castor 4A strap on motors. The ultimate Delta 7000 series rockets used new RS-27A engines that were more efficient in vacuum and new GEM-40 Graphite Epoxy Motors developed by Hercules. All versions used the existing Aerojet AJ10-118K ITIP-powered second stage. Star 48B served as the third stage for GPS missions. A new 9.5 foot diameter standard fairing housed most payloads. The old 8-foot fairing flew a few times. McDonnell-Douglas also developed a 10 foot diameter metal shroud based on the company's Titan 3C fairing.

Delta 2 initially used the existing DRIMS guidance and control system, but on December 30, 1995 Delta 230 became the first to use the Redundant Inertial Flight Control Assembly (RIFCA). RIFCA, built around six ring laser gyroscopes and six accelerometers, provided triple redundant guidance, flight control and mission sequencing functions.

The first Delta 2, a 6925 identified as Delta 184, orbited GPS-2 1 from Cape Canveral LC 17A on February 14, 1989. Seventeen 6000-series Deltas flew. Delta 212, the last in 1992, was the final flight of a Saturn H-1 derived RS-27 engine. Delta 201, the first 7925, orbited the first of the heavier GPS-2A satellites on November 26, 1990. All told, Delta 2 performed 49 GPS launches with one failure.

In December 1994, NASA requested bids for a Medium Light Expendable Launch Vehicle (Med-Lite). McDonnell-Douglas's offered Delta 732X and 742X, which used three and four GEM-40 boosters respectfully. In 1997 a new 10 foot diameter composite payload fairing (10C) began flying for Iridium. A stretched 10L version was developed for NASA beginning in 2002, ending use of the Titan-derived fairing.

From 1989 through 2011, Delta 2 was the most often-flown, versatile, productive, and reliable U.S. launch vehicle. It became the longest-lived essentially unchanged U.S. launcher, even though its ownership and production sites moved twice. It flew from three launch pads at two launch sites. It was common to see multiple Deltas stacked simultaneously. 151 Delta 2 rockets flew by the end of 2011 with two failures, making Delta 2 one of the most successful orbital launchers in history. In addition to its bread-and-butter GPS work, Delta 2 orbited commercial, non-U.S. government, and NASA satellites.

For NASA, Delta 2 did something no previous Thor/Delta had done - it reached into deep space, to Mars, and to asteriods. The list of payloads includes Mars Pathfinder, Odyssey, Spirit, Deep Impact, and many others. Commercial payloads included Iridium and Globalstar "little LEO" satellites.

In 2003, Delta 7920H and 7925H began flying. They used powerful GEM-46 solid rocket motors reassigned from the terminated Delta 3 program (about which more soon). They launched the Opportunity Mars rover, the SIRTF/Spitzer space telescope, the planet Mercury orbiter MESSENGER, asteroid Vesta and Ceres orbiter Dawn, gamma ray telescope GLAST, and lunar orbiters GRAIL A and B.

Although Delta 2's Cape launch site was closed after 2011, the Delta 2 story is not over. Two more launches from Vandeberg AFB SLC 2W remain on tap. If both succeed - never a given - the rocket's consecutive success string would reach 100.

25.  Delta 3

During the post-Challenger era, McDonnell Douglas studied liquid hydrogen fueled upper stages for its Delta launch vehicle. In 1986, it briefly considered Japan's H-1 LH2/LOX upper stage for the U.S. Air Force Medium Launch Vehicle (MLV) program before deciding on the "Delta 2" approach. Two years later, it proposed a new LH2/LOX upper stage to be built by Martin Marietta for the MLV-2 program (won by GD Atlas 2). The company continued to study the idea until, on May 10, 1995, it announced that it would develop "Delta 3" using more than $200 million of its own funds, with a planned first launch in 1998.

Soon, Delta 3 held contracts for 18 launches through 2002, including NASA/NOAA GOES N, O, and P and five ICO Global Communications launches.

Delta 3 used a 4 meter diameter "Delta Cryogenic Upper Stage" (DCUS) and more-powerful Alliant 46 inch diameter Graphite Epoxy Motors (GEM-46) to lift the heavier stage and payload. The first stage had a shorter but fatter kerosene fuel tank (4 meters rather than 2.4 meters diameter) so that Delta 3 would fit within the Delta 2 service tower. A Pratt & Whitney RL10B-2 engine with a large Snecma-built extendible nozzle powered DCUS. At liftoff, the boosters and RS-27A main engine would together produce more than 1 million pounds of thrust. The rocket could lift 3.8 tonnes to GTO.

Delta 3 was assembled in Pueblo, Colorado. Japan's Mitsubishi Heavy Industries made the 4 meter diameter second stage liquid hydrogen tank and first stage kerosene tank using tank tooling from its H-2 stage.

On December 15, 1996, McDonnell Douglas and Boeing announced their intention to merge under the Boeing name. The merger was consummated on July 1, 1997. The merger would have decisive consequences for the Delta 3 program, though that was not initially apparent.

A relatively smooth development program was followed by a troubled flight program. Delta 259, the first Delta 3, launched from Cape Canaveral SLC 17B (the only pad rebuilt for the type) on August 27, 1998 with the Galaxy 10 communications satellite. At about T+50 seconds, the rocket began to suffer 4 Hertz roll oscillations, using up the GEM-46 TVC hydraulic fluid. The rocket pitched over and broke apart at T+72 seconds. Flawed roll control equations were found to be the cause.

The second Delta 3, Delta 269, launched with the Orion 3 communications satellite on May 5, 1999. This time the flight proceeded flawlessly through the first RL10B-2 burn, pushing the stage and payload into a parking orbit. After a coast period, the RL10B-2 engine restarted for a planned 162 second burn, but it shut down after only 3.4 seconds, stranding Orion 3 in LEO. An investigation found that the RL10B-2 engine's combustion chamber had burst during the restart due to defective brazing of a welded reinforcing strip. Pratt & Whitney subsequently modified its brazing process and its inspection methods.

Delta 3 finally succeeded on August 23, 2000 when Delta 280 launched the 4,348 kg DM-F3 mass simulator to subsynchronous transfer orbit. By then, however, Boeing was committed to its Sea Launch commercial launch partnership and had begun to develop Delta 4 for the EELV program. In 2000, Boeing bought Hughes Space & Communications, the satellite builder that held the bulk of the Delta 3 backlog. Then the commercial satellite market collapsed, leaving Boeing deeply overextended. Something had to give, and the first of those somethings was Delta 3, which was quietly shut down after Delta 280.

In the end, Delta 3's primary accomplishment was to prove RL10B-2 in flight for Delta 4, and to prove DCUS, which was the first all-new high energy upper stage developed in the U.S. since the 1960s. The Delta 4 Medium "Delta Cryogenic Second Stage" was largely derived from DCUS.