Atlas Chronology

[Spaceman Spiff]

A lot has been written about the Atlas Missile en Atlas Launch Vehicle, including in these forums. And it deserves it, having played such an important role in the history of rockets and space travel.

There are different ways to look at history. In narrative form, telling the tale from its beginning, its origins, its evolution. Through the eyes of participants, who tell their own stories as they lived them. Through tables of launches and tests. Through technical specifications and their evolution and adaptation.

I have been working lately on presenting the history of the Atlas rocket (or at least a part of it) through chronology, a visual timeline. It puts the events and facts in a different format and, while limiting in certain areas, it gives a different perspective on what happened when. For example, through the timeline I noticed that  from 12 to 17 May 1959 all Cape Canaveral Atlas launch pads (LC-11, LC-12, LC-13 and LC-14) as well as launch pad 576A-2 at Vandenberg Air Force Base had Atlas missiles on them at the same time. Quite impressive.

It is of course a work in progress. Currently I have entered events roughly through April 1961. I will continue to add to it when time allows.

You can find the timeline here :http://www.wayo.be/atlas-rocket-history/

Of course I welcome all corrections, additions, references or other information.

- Michel -

[Spaceman Spiff]

Great info! That second half of 1961 was interesting chronology wise too. No less than 4 days with two launches per day : Aug 23 (101D and 111D), Sep 9 (26E and 106D), Nov 22 (108D and 4F) and Nov 29 (93D and 53D). The year closes out with a salvo of 3 consecutive days with launches : Dec 20 (36E), 21 (6F) and 22 (114D). (All assuming UTC launch times).
That period will be added to the chronology shortly.

[Spaceman Spiff]

After the 48D failure, for the following nine launches, a 4.25 second holddown delay was incorporated in order to verify engine stability before liftoff. Also a second redundant accelerometer was added to the Rough Combustion Cutoff system to increase reliability in case an accelerometer would fail. No booster instabilities were detected in the nine following flight although in the attempted launch of 32D the RCC in the sustainer engine was triggered, the engine was subsequently replaced and 32D had a successful flight. Of course 27E would show that the problem still existed.

I guess the 48D footage would resemble that of 9C which also burned on the pad for some time before exploding. Of course this was during a Flight Readiness Firing and not an actual launch.

[Jim]

9C also produced an incredibly powerful explosion that leveled the entire service tower on LC-12

Umbilical tower.  The Service tower was unaffected.

[Spaceman Spiff]

I read the postflight report for 48D and it gives a fairly detailed description of the events during the attempted launch.

The postflight reports contain a lot of interesting information. Unfortunately I haven't been able to find many of them. DTIC has a few but some are 'released to the public' but the text is not available on-line. Does anybody know another source for them ?
For the Atlas Timeline, after about 95D they are for the moment my only source for operational movements of the missiles (to/from the pad, hangars, etc.)

51D actually should not have lifted from the pad, but the RCC sensor in the B-1 engine was not working so it allowed the missile to be released anyway. The B-1 RCC sensor on 48D wasn't working either but the malfunction occurred in the B-2 so that engine's RCC operated correctly and terminated thrust before lifoff could be achieved.

That explains the reason why they added redundancy to the RCC accelerometers after 48D.

[RIB]

Interestingly enough, Big Joe, an unmanned Mercury prototype, made it through MAX-Q even though, the Atlas failed to stage. No escape tower on that Mercury-Atlas launch either.

[RIB]

 If poor  aerodynamics were the suspected cause of the MA-1 failure, then why did they Big Joe capsule, essentially the same shape as the MA-1 capsule (though lighter) make it through Max-Q? I understand they reshaped the trajectory after MA-1 but weren't the aerodynamic forces on the Big Joe shot essentially the same as the ones on the MA-1 shot. I don't understand how weight would affect the aerodynamics.  I understand the shape of the Mercury Capsule was different than the "standard" Atlas warhead, but I'm assuming that the weight of the Mercury capsule was LESS than the Atlas warhead, given the velocity requirements for an orbital vs a ballistic shot.

[Jim]

Ultimately the blame for the failure lay on Flight Director Walter Williams both for ordering the launch to take place without the LES and for launching it into cloudy weather where the booster couldn't be filmed after the first 20 or so seconds of flight (as well as depriving us of footage of what was probably a pretty cool explosion).

He didn't have the authority to make either of those calls.
A.  Flight Director is not responsible for the vehicle configuration.
b.  Flight Director does not make launch weather calls.  That is the launch director's responsibility.

[Jim]

Also on the topic of MA-3, I can think of at least three other flights that went straight up and didn't pitch over. These were a pair of Thor launches from VAFB in 1959 and an Atlas R/V test in 1968. The Thors failed because some technician forgot to cut a wire holding the programmer tape in place so they never executed the pitch and roll sequence. The Atlas (95F specifically, it had an ABRES TVX vehicle) I'm not sure of the exact reason for the failure, but probably similar circumstances to MA-3.

Delta 59 Intelsat III-1

[WallE]

On the night of June 22, 1961, Cape Canaveral was treated to a spectacular fireworks display as Atlas 17E launched from Pad 11 on an R&D test. The pitch and roll program resulted in a pitchover rate that was 56% too high. At T+79 seconds pitch oscillations caused an increase in LOX tank pressure and at T+97 seconds aerodynamic forces caused collapse of the vernier fairing. This ruptured a hydraulic line and caused loss of hydraulic pressure to the sustainer and verniers. The missile broke up at T+100 seconds due to either structural failure of the propellant tanks caused by loads or aerodynamic heating.

Investigation into the failure found that the pitch gyro motor was running at 84% speed. With this being the fourth Atlas lost to a gyroscope malfunction, it was decided to immediately phase in the Spin Motor Rotation Detection System in all Atlas vehicles which would generate a No Go signal and prevent launch if the gyro motors were not operating properly. One more Atlas failed this way (Missile 102D in 1963) but it was using the old-style "round" autopilot without the SMRD. Three other Atlas missiles in VAFB's inventory were found to be using the same outdated hardware so their gyro canisters were replaced with Project Mercury spares.

No videos of this launch online unfortunately.

[WallE]

Another flight control mishap. Atlas 23D (code name "Lucky Dragon") lifted from Site B-2 at VAFB on May 6, 1960 as an IOC training flight. After a normal engine start and rise-off, the Atlas abruptly flipped upside down at T+21 seconds and did a couple of cartwheels before the RSO sent the destruct command at T+26 seconds. The booster section landed near the Southern Pacific railroad line but missed actually hitting it. Initial postflight analysis found that the pitch gyro motor was non-operational so that missile stability was lost the moment the programmer entered the pitch and roll sequence. The guidance decoder had also failed during the prelaunch countdown so had the flight continued no guidance commands could have been received and the Atlas would have missed its planned target point in the South Pacific.

The ultimate cause of the pitch gyro failure was a flexible cable lead that contacted the metal gyro case and caused a short that burned out the motor. Afterwards the cable leads would be properly insulated. Another case for the SMRD system as prelaunch checks had no way of detecting an nonfunctional gyroscope.

The cause of the guidance failure was thought to be the result of a faulty vacuum tube; after the pulse beacon was recovered, replacement of one tube in it got it working again. The failed tube had cracked and degassed and this was thought to have occurred sometime during ejection of the pulse beacon canister from the missile or on ground impact and there was no indication of degassing while it was powered on.

Video here. That booster section really goes up like a napalm strike when it impacts the ground, there must have been a lot of fuel still in there. One thing that puzzles me is why there's no engine cutoff prior to the RSO destruct. Isn't there normally supposed to be a cutoff and one second delay until destruct?

Edit: My speculation is the missile never got the engine cutoff command due to tumbling; perhaps the RSC receiver antenna was unable to lock onto it

[WallE]

Flight control issues struck again when Atlas 74D (code name "Tiger Skin") lifted from Site B-1 on July 22, 1960 on another IOC test. The pitch and roll program resulted in a pitch angle that was 69% too high. The Atlas continued downrange until missile self-destruction at T+69 seconds. Film coverage of the flight was lost after T+30 seconds due to overcast skies and tracking cameras were unable to record the latter portion of the flight, however it was believed that pitch oscillations during Max Q overstressed the airframe. The failure was attributed either to an improper motor speed of the pitch gyro or excessive gain in the torquer amplifier. Efforts were made to inspect the Atlas inventory at VAFB for proper gyroscope operation and to get the supplier of the gyros (Kearfott) to overview their manufacturing and testing processes.

Video of 74D does not capture the missile failure as noted above but does show the abnormal pitch angle. The initial round of Atlas IOC tests in 1960 went poorly; out of seven launches only one was considered entirely satisfactory, the rest ending in explosions or missing the target by a long distance.

[WallE]

Atlas 5F launched from Pad 11 at Cape Canaveral on the afternoon of December 12, 1961. It was the third F-series test flight. At T+7 seconds the ARMA guidance system issued erroneous SECO/VECO commands but they were not acted upon as the programmer was blocked from receiving a BECO signal until 120 seconds after liftoff and SECO at 200 seconds to prevent a fallback on or around the pad. The missile took off and flew without incident until the SECO/VECO command was unblocked at T+200 seconds and terminated thrust, causing the Mk IV R/V to only impact about 400 miles downrange. A side pod carried by the missile was released and deployed several flares which were sighted.

The accidental engine cutoff commands were due to a malfunction of the ARMA circuitry due to in-flight vibration levels and it would eventually be solved by adding acoustic liner insulation. A similar malfunction had been noted on 27E during its brief flight six months earlier. Atlas 48E over two years later experienced similar accidental cutoff discretes at launch that resulted in premature thrust termination but that was an early production E-series missile that predated the introduction of the acoustic liner.

This newspaper clipping mentions the release of the flares but calls the flight a success and claims it traveled the full 5,000 mile range which it could not have done anyway as 5F was an R&D missile with three telemetry packages and thus too heavy to manage that kind of distance. There's a lot of lies in that press release.

[WallE]

Atlas 7D was the second D-series Atlas test, launched from Pad 14 on the night of May 18-19, 1959 and the second attempt at launching an RVX-2 R/V after the first attempt on 7C miscarried two months ago. It was the first launch from Pad 14 after conversion for Atlas D, which included the construction of a large umbilical tower to support space launches. In addition, the Mercury astronauts, recruited just a little over a month earlier, were invited to watch the vehicle that would eventually take them into orbit. The Atlas's engines lit up and the missile rose into the night sky and steered downrange. Just about a minute into launch and with no prior warning it disappeared into a fireball. Gus Grissom remarked "Are we really gonna get on top of one of those things?" Alan Shepard asked the launch crews what happened. They replied that they didn't know but would study the data to find out.

As it turned out, the failure was not caused by the Atlas itself but the launcher mechanism--study of liftoff film showed that at first missile motion the B-2 hold-down pin did not retract and as the missile lifted it remained in the hold-down position, yanking at the airframe like someone using a crowbar. This caused a four inch gap between the booster thrust barrel and the airframe above it. The airborne helium fill duct was also ruptured, causing loss of tank pressurization gas during ascent. Fuel tank pressure eventually became too low to maintain structural integrity; at T+62 seconds the intermediate bulkhead collapsed from the weight of the LOX tank above it, and the missile exploded two seconds later.

The hold-down pin failure was due to excessive play in the pulley system designed to retract it at launch and failure of the bell crank retaining bolt. Improved maintenance procedures for the launcher were imposed and the retaining bolts replaced with ones made of higher heat treat steel. The problems with the launcher were evidently quickly fixed since Atlas 10D was erected on Pad 14 just fourteen days later. The launcher mechanisms on Atlas pads were normally very reliable and just about never failed on a launch, making this almost a freak occurrence.

The RVX-2 would fly successfully on its third attempt three months later.

[Kyra's kosmos]

Atlas ICBM artist's conception. What can I say, it's my desktop background.

[WallE]

Atlas 51D launched from Pad 13 at Cape Canaveral on the evening of March 10, 1960 on an R&D flight and dramatically ended a streak of almost 20 consecutive successful or mostly successful Atlas launches. The B-1 engine lost thrust almost immediately at liftoff followed by the missile toppling and an explosion that among other things appears to have caused the structural failure of the propellant tanks, which split open and collapsed (in the second photo the Mk. III R/V can be seen falling to the ground after the tank structure crumpled, one of the more bizarre aspects of this failure). The nearly full propellant load spilled out, mixed, and went up in a gigantic fireball that left Pad 13 a charred mess.

The disaster was found to be the result of rough combustion that caused an explosion in the B-1 engine injector head, badly damaging it and rupturing the LOX dome above. A major fire and explosion then caused the loss of the missile only three seconds after launch. Pad 13 was put out of use for six months, after which it was converted for Atlas E tests and not used for D-series vehicles again (until 1963 when it was converted to an Atlas-Agena pad). Nobody predicted the same failure mode would strike again in just a few weeks and put another launch complex out of use for a while.

The booster engines did have RCC (Rough Combustion Cutoff) sensors designed to shut them down prior to liftoff if this happened, but as it turned out the B-1 RCC on 51D was not working and so the missile was allowed to lift and produce a pad fallback.

The cause of the rough combustion was never precisely nailed down, but several fixes were implemented including installing copper baffles in the booster injector heads to prevent rough combustion and the lessons learned here were later applied to the Saturn F-1 engines.

[WallE]

SAMOS 4 launched on Atlas 108D and Agena 2202 at slightly past noon on November 22, 1961 from PALC 1-1 at VAFB. The pad was quickly restored to use from the SAMOS 3 explosion and back online by October 29, after which 108D was erected. Despite extensive efforts to ensure mission success such as checks for faulty transistors and making sure all plumbing was clean and free of contamination, the result was another (less spectacular) failure. The Atlas rose into a clear, sunny autumn sky and steered downrange--the first problem happened with a BECO at T+133 seconds, eight seconds earlier than planned. Pitch control was lost at T+244 seconds and the Atlas pitched up about 4.7 degrees from nominal. This prevented the guidance antenna from locking onto ground guidance and the programmed SECO signal at T+259 seconds was not received. Sustainer shutdown at T+289 seconds occurred as a result of LOX depletion. At Agena staging the Atlas had pitched up about 160 degrees and as a result of the improper flight trajectory the Agena was pointed downward and backward from nominal and orbit was not achieved. The Agena and satellite impacted at an unknown point in the Pacific.

The premature BECO was found to be the result of locating the staging backup accelerometer (the purpose of it was to send a backup BECO if the programmer failed to do so) on the side of the LOX tank where it malfunctioned from the extremely cold environment there. This same malfunction had occurred during the launch of Ranger 2 a few days earlier and was solved by moving the accelerometer to the fuel tank. The main problem that resulted in loss of the mission was the pitch control issue which was found to be caused by in-flight loss of the heat shield over the retrorockets, exposing the gyroscope package to aerodynamic heating. The heat shield covers were redesigned afterward as well as replacement of transistors in the gyroscope signal amplifier with a type that was not as prone to overheating.

Tracking film showed an object falling from the booster at T+45 seconds which was thought to be the retrorocket heat shield breaking off.

In addition, SAMOS 4 was the first top secret DoD space launch while prior ones were publicly acknowledged. From this point onward the Air Force said nothing in press releases outside a satellite being launched into orbit, which was all that was announced for SAMOS 4 but since the press release also left out the "launched into orbit" part, it was quickly apparent that the launch had not been a success.

Should add that this goes back to a point I made in the Ferret thread about the lack of standardized booster hardware prior to the SLV cores and the reliability issues caused by that. For example it seems only 108D and 111D had a staging accelerometer put on the LOX tank where it would malfunction in this way, I have not heard of this happening on other Atlas vehicles. Same also seems to apply to the retrorocket heat shields.

[WallE]

RM-20/P72-2 aka another unlucky package of scientific test equipment much like the MPRV lost in the 9-65 Thor mishap launched on Atlas 71F on April 13, 1975 from PALC 1-1 at VAFB--after a decade as a TAT Thor-Agena pad, the facility was once again hosting Atlas launches and 71F, once an on-duty ICBM at Dyess AFB in Texas, was mated with a Thiokol Star-17A solid upper stage for this mission, and it was in trouble almost immediately at liftoff as a malfunction of a water valve prevented the flame bucket from being hosed down with deluge water. Propellant overflow prior to engine start pooled in the flame bucket and formed a kerolox glob, causing an explosion. This was registered on the vertical accerometer as a shock. In addition an electrical umbilical supplying power to the booster T/M pulled out prematurely and before it could be cycled to internal power, causing an instant loss of power to the T/M system. As a result no data was returned during the launch and an indeterminate degree of thrust section damage resulted from the kerolox explosion.

The Atlas took off and flew and booster jettison was completed properly but the sustainer and verniers shut down some time later and the Range Safety officer sent the destruct command at T+303 seconds. Owing to the T/M loss it was not possible to determine exactly what caused the propulsion shutdown, but it was probably a leak in some system, perhaps a loss of control gas to the sustainer propellant feed system.

Several changes made as a result of the mishap included re-coating the flame bucket on PALC 1-1 to remove cracks and pits where propellant could pool in, an improved water deluge system, eliminating the practice of draining propellant overflow into the flame bucket, and adding a nitrogen purge system to the booster thrust section to prevent fire damage at liftoff.



Film footage of 71F's launch. The normal speed film at the start shows that it was a quite violent explosion, it actually caused the camera to shake and the pad is on fire for a bit afterward. One is almost surprised the booster was able to lift and fly after that.

[WallE]

Atlas 102F takes off from 576-A1 on a successful launch of a RADCAT radar calibration satellite and a Radsat scientific satellite  for the Air Force Space Experiment Support Program, October 2, 1972. The Atlas, formerly an ICBM stationed at Walker AFB in New Mexico, was topped with a Thiokol/Boeing Burner-2 solid stage. Many refurbed Atlas E/F missiles would fly over the 70s-80s launching payloads of various kinds, including a few spectacular failures as well as the final Agena stage ever flown. Among the modifications made to the vehicles included replacing the ARMA inertial guidance system with an Atlas-Agena Mod III-G radio guidance system and standardized range safety and electrical systems--there had been a lot of E/F missiles built and over the production run there were many changes to the missile hardware as the design evolved or due to findings from failed launches so they could often have considerably different hardware in them; it was considered pertinent that each vehicle have a single, standardized electrical system.

This was the second attempt at launching a RADCAT and it remained in orbit until 2012; the first had been on an Atlas SLV vehicle with a Burner-2 in 1968. That launch was part of a "grab bag" that had 13 different payloads on it all scientific or calibration targets of various types; the payload fairing did not separate and the satellites fell back into the atmosphere and broke up.

[WallE]

Atlas 26E lifted from Pad 13 at Cape Canaveral on the night of September 10, 1961 as part of an R&D launch carrying a Mk III R/V. The flight was entirely normal up until booster jettison at T+129 seconds when the sustainer engine inexplicably shut down. With nothing but the two small verniers for power, the Atlas tumbled its way on down to an impact point about 500 nm downrange instead of the planned lob into the South Atlantic over 3,000 miles downrange and separation of the R/V was not accomplished. What caused the loss of sustainer thrust was not clear as the shutdown occurred during the normal two second data blackout caused by exhaust gases impinging on the T/M antenna during the staging sequence. When the data resumed, the sustainer was not only not operating, but its gas generator was recording temperatures in excess of 1000°F, the result of an apparent LOX-rich shutdown.

GD/A engineer Ed Hujsak suggested that the propellant disconnect lines in the booster section of E/F series missiles could allow propellant to pool in them and be ignited by the sustainer exhaust at booster jettison, possibly damaging components in the missile. He noted that on prior E/F launches there had often been a momentary pitching motion of the missile at staging that could be oscillation caused by exploding propellant and suggested adding cutoff valves to the booster section to prevent leftover propellant from escaping and being ignited.

During booster phase there was an unusual rise in thrust section temperatures beginning at T+77 seconds that eventually reached 115°F at T+115 seconds and stabilized there. The cause of this was unknown and it was not thought to have any connection with the subsequent loss of the missile.

The Air Force in their infinite wisdom then decreed that only cutoff valves for the LOX lines would be added as it was impossible for leaking RP-1 to explode in the absence of LOX.  On December 21, Atlas 6F launched from Pad 11 in an attempt to finish the mission that Atlas 32E had failed to do a month earlier, which was fly a monkey in a pressurized capsule on a suborbital lob. Once again the Atlas performed normally through booster phase, after staging the sustainer hydraulic system developed a leak and eventually engine gimbaling control was lost. The sustainer shut down from fuel starvation caused by missile tumbling and the capsule, containing a monkey named Scatback, separated and landed in the South Atlantic but could not be found as the tracking beacon failed to operate.

The Air Force then finally accepted that the fuel lines on the E/F missiles needed cutoff valves as well and no more mysterious system failures during booster jettison occurred after this modification was done.