gandalf@pro-canaveral.cts.com (Ken Hollis) (11/17/90)
Greetings and Salutations: >From: gsh7w@astsun.astro.Virginia.EDU (Greg Hennessy) >Subject: Re: Causes for Columbia leaks ?? >In article <1475@eastman.UUCP> gerwitz@Kodak.com writes: >#Now that Columbia seems to be back together and ready to launch... ># >#I don't recall any mention of the actual causes of the hydrogen leaks that >#plagued it over the summer. Maybe I just wasn't paying real close >#attention this time around. Any quick explanations ?? >The problem over the summer was in the 17 inch disconnect valve, which >had scratches in the teflon seals. The scratches were caused by small >glass beads. How the beads got there is unknown, but I have hard at >least two theories. When they were putting back together the engines >after fixing the disconnect valve, two seals were crushed, causing >leaks. In fixing one leak they introduced two more. The procedures >were changed to prevent a re-occurrence of this event. >-Greg Hennessy, University of Virginia Let's start off correctly. Columbia (OV-102) STS - 32 was stacked on MLP-3 (Mobile Launch Platform) in the VAB (Vehicle Assembly Building), the first use of MLP-3 for the shuttle program. It was then transferred to Pad - A for launch. Pad-A had just finished a major modifications program to get it upgraded to the same extent as Pad-B had been earlier. This did include some work on the LH2 system. The external tank had also been modified for the ECO (Engine Cut-Off) sensors. All of these operations require sand paper of some sort . Launch was on 0735 a.m. EST 1/9/1990. The payload was the SYNCOM IV-5 geostationary communications satellite, also known as LEASAT. The LDEF (Long Duration Exposure Facility) was also retrieved (the great killer tomato seed experiments...). OV-102 was returned to the OPF (Orbiter Processing Facility) for standard LH2 system maintenance. Upon removing a component (PV14, E-1 recirc valve) therein was discovered small pieces of what was later analyzed to be corundum (a type of sandpaper). The entire LH2 system was then thoroughly inspected, and all traces of the corundum removed. One of the inspections required access to a screen that catches all particles going to the engines, the prevalve screen. To boroscope this screen a small part, the detent (a retaining device for the prevalve), was removed. Because of the placement of this part, re-installation (with its teflon (naflex) seal) was with the tech laying on his back. The seal slipped upon reinstallation (unknown to all) on PV5 & PV6 (E-2 & E-3 LH2 prevalve respectively) causing the seal not to seat at cryo conditions (but it sealed very nicely, thank-you, at ambient temperatures). I will give you a hint, Cause #1. On the ET (External Tank) there is a 17" flapper valve to match the 17" flapper valve on the vehicle side. These valves allow the flow of LH2 from the ET to the SSME's (Space Shuttle Main Engines) during flight. The actuation of both of these valves comes from the vehicle. There is an arm that rotates a shaft hooked up to the flapper, and the flapper opens. There is one teflon seal on each end of that shaft that rotates the flapper. Sometime after manufacture, due to improper storage at the vendor, some glass beads got imbedded into these seals. Cause #2. When OV-102 was tanked, there were leaks from both inside & outside the aft (there had not been any leakage of this magnitude before on any vehicle). Isolation tests were not fully successful because of the two different areas of leaks. OV-102 was rolled back to the OPF and the 17" disconnect from OV-105 was placed on OV-102 in hopes that it wouldn't leak. All joints were re-leak checked with the detent cover seal on PV6 found crushed. This seal was R&R'ed. Tests with the ET 17" disconnect at that time were inconclusive, but later it was found that the ET half of the disconnect leaked. In the meantime OV-104 was rolled out, tanked & it leaked. See below for that story. OV-102 was restacked, and rolled to Pad-A, where it was tanked & it leaked. Leakage, however, was noted only inside the aft this time (at a lower level). The assumption from before was that the leakage was large enough to also enter the aft. Intensive leak checks of all joints at ambient temperature followed. Many bolts had their torque checked. The only out of spec conditions found were a bolt with an upside down washer & a very small leak at the PV5 detent seal. PV5 detent seal was then found crushed, and R&R'ed. The tanking test was run & successful. Scheduled for this weekend are SSME FRT's & next week preps for launch. OV-104 The DOD mission that was launched yesterday (11/15/90) was originally stacked & ready to go earlier this year. As a check of the system to make sure that OV-104 didn't have the same problems as OV103, NASA decided to run a tanking test. It leaked. Extensive testing showed that a small amount of hydrogen (1 SCFM, Standard Cubic Feet per Minute) could show up (on a calm day, like when they tanked) as a large leak (6% --> 10%). Hydrogen expands about 700 times when going from liquid to gas. It also showed that OV-104's problem was on the ET 17" side, not the orbiter. OV-104 was rolled back to the OPF & the ET 17" disconnect was changed out. OV-103 was launched with Ulysses. OV-104 was rolled to Pad-A, and OV-102 To Pad-B, hence the shuttle shuffle. The disconnect that leaked also had the same problem as Columbia, ie the teflon seals were damaged (all ET halves of that lot number have been flagged & inspected last I heard). OV-104 was tanked with LH2 & LO2 as a total system check & it didn't leak. >From: gregc@cimage.com (Greg Cronau) >Subject: Re: Causes for Columbia leaks ?? >The result has been the bane of diagnosticians for a long time: >When they set up a system to allow them to watch it real carefully, >It didn't leak! >gregc@cimage.com As per above, all discrepant conditions were found and corrected. Ken Hollis ProLine: gandalf@pro-canaveral Internet: gandalf@pro-canaveral.cts.com UUCP: crash!pro-canaveral!gandalf
gandalf@pro-canaveral.cts.com (Ken Hollis) (11/21/90)
Greetings and Salutations: >From: Wales.Larrison@ofa123.fidonet.org (Wales Larrison) >Subject: Re: Causes for Columbia leaks ?? > The scuttlebutt in the industry was that the Columbia problems >were caused by the new MLP used, and some mis-judgements and >untrained technicans. Now this is "third bar seat on the left" >unoffical rumor data, but I've heard it from a couple of sources. > The Mobile Launch Platform (MLP) used for the initial tanking >test was a new one - and the cleaning procedures used on the cryo >lines through the MLP through the aft end of the orbiter and into ET Note: The MLP lines and/or the ET had the corundum, not the MPS lines. The most likeley source is the MLP GSE (Ground Support Equipment) lines. >weren't really up to snuff. When they did the first tanking test, ...If you wanna call "launch" a tanking test, then so be it... >they found cleaning material grit in the fuel system filters (used to >clean the MLP cryo lines before use...) - and since that is very After the launch & after the corundum was found in the MPS lines... >highly abrasive, and not something you want to run through high >speed turbopumps, and shouldn't have been there in the first place - The corundum was found in the OPF (Orbiter Processing Facility) upon returning from landing (Reference LDEF retrieval). >they decided to dissamble and clean out the MPS (Main Propulsion >System) feedlines in the aft end of the orbiter. > The MPS system is pretty complex, since the systems are very >complex physically, pretty sizable (each line is about 24" across), 17" across (Internal diameter) >and are vacuum jacketed and foamed. There are about 83 pieces which >have to be fitted back together and tested. Disassembly and 83 major components (depending on how you count major). Hundreds of joints alone were leak checked. >reassembly has been done several times - each system is assembled >and tested at the Rockwell manufacturing plant in Palmdale, Assembled for acceptance testing by NASA, and most of the tech's at Palmdale have less experience than the tech's at KSC. >California when each shuttle is built, and at the Arrowhead Products >plant in Southern California, when they make the parts before >shipping them to Rockwell Palmdale. Nobody knew of one being >stripped and reassembled at KSC. It's not thier problem when it is at KSC. They don't have control over it anymore. > The rumor goes that Rockwell Palmdale and Arrowhead Products >quickly put together a team to go down to KSC and perform the work Sure, "At a price..." >needed - figuring the most experienced team would do the job the All of the tech's at KSC were hired from Rockwell when LSOC (Lockheed Space Operations) got the contract. Alot of them worked on STS-1. >quickest and best. However, NASA KSC and Lockheed, the Shuttle >operations contractor to NASA KSC, have been pushing for more >maintenance work to be done at KSC. (They figure after the >Endeavour is built, all of the Rockwell technicans will be out of a >job, so they'll have to fix it themselves...) So they decided that All of the regular maintenance and testing and testing is done at KSC. It is rare when Rockwell actually does work on the bird. >Lockheed shuttle techs would strip down, clean, and reassemble the >MPS at KSC. > However, in putting the system back together, several teflon >seals were "crushed", which allowed the leaks. This required The "Crushed Teflon Seals" were installed by a Rockwell subcontractor, not by Lockheed. >dissasembling and retesting the system, yet again. > To their credit, the Lockheed techs did manage to get it put >back together the right way, and the tested leak rate is the lowest >of any shuttle. > > Remember - this is rumor - and unconfirmed by any published >reports. > >Wales Larrison Ken Hollis ProLine: gandalf@pro-canaveral Internet: gandalf@pro-canaveral.cts.com UUCP: crash!pro-canaveral!gandalf