CHAIRMAN ROGERS: Will you please give your name and present assignment and a little bit about your background and what you're presently doing here at NASA.
COLONEL O'CONNOR: Edward O'Connor. I'm a Colonel in the United States Air Force. I'm assigned to the Eastern Space and Missile Center. Currently I'm acting as director of search and recovery operations for NASA on the 51-L accident.
I've been involved in this activity since the day of the accident. Prior to that, I was director of operations in the 6555th Aerospace Task Group, where we put together the payloads that fly on the shuttle. And I've worked in different shuttle missions.
CHAIRMAN ROGERS: All right, you may proceed, please.
COLONEL O'CONNOR: First chart, please.
(Viewgraph.) [Ref. 3/7-1]
Immediately after the accident, we went about establishing the search blocks so we could go out and recover components. We were able to use our radar
optical data that's available from the range instrumentation cameras and the range radars. We were also able to use visual sightings obtained by some of the aircraft in the area to identify those areas where the majority of the shuttle components impacted the water.
Since that time we've also used a trajectory calculation from our radars and optical sensors to better improve our location of debris on the ocean bottom. Since that time, we've also had to modify our search areas based on the located components that we have recovered, as well as additional sonar contacts that we have obtained through the Navy assets.
The second chart, please.
(Viewgraph.) [Ref. 3/7-2]
The first box in our search area I would like to point out to you is a ten nautical mile by 25 nautical mile search box, the lower side of the chart. That particular box was initially established in the days immediately following the accident. Since that time, we have added that top section, 8.5 nautical miles on the eastern edge, three nautical miles on the western edge.
That was added because it was felt that the majority of the right-hand SRB components should be
located in that area based upon trajectory calculations and optical data.
Next chart, please.
[1071] (Viewgraph.) [Ref. 3/7-3]
In order to indicate the difficulty of the recovery activity we are engaged in, I would like to go over very quickly the oceanographic conditions that we were experiencing in the area. We are quite close to the Gulf Stream axis in the search box and we are experiencing
CHAIRMAN ROGERS: Could you move your microphone up just a little bit closer?
COLONEL O'CONNOR: At this time we have a search box that is very close to the axis of the Gulf Stream. Therefore, we are experiencing currents of the level of four to five nautical miles. We are experiencing a little seasonal variation as we progress toward the summer in the position or the speed of the Gulf Stream.
So this is a constant factor we're going to have to deal with during the whole recovery. The depth in the vicinity of the SRB's we feel is between 220 feet and 1200 feet. This has been established by some of the videos we have taken on the ocean bottom and some of the recovery action we have taken to date.
The remainder of the search area, the area where we expect to find the majority of the orbiter, as well as the orbiter payloads, is in a search depth of approximately 90 to 220 feet.
Next chart, please.
(Viewgraph.) [Ref. 3/7-4]
The at sea operations are being directed by the United States Navy Superintendent of Salvage. The most experienced salvage team that has been put together in a long time. The assets we have employed at this time in the search and recovery activity are eleven ships, one manned submersible, two remotely operated vehicles, seven sonars, and 41 divers.
This is a fairly constant number of ships and assets that we are involving every day, seven days a week, in the recovery.
Next chart, please.
(Viewgraph.) [Ref. 3/7-5]
In order to maximize the utility of the use of these assets, we have broken them into three basic tasks, the first being the SRB recovery as being the highest priority recovery. We are using right now the submarine NR-1, which is a small nuclear-powered submersible that the Navy uses for underwater submarine search and rescue. We are using this as a sonar
platform to map the SRB area.
We have also rented the STENA Workhorse, which is a dynamically stabilized oil support ship that has been used in the Gulf during high winds and during high currents. It is very suitable working close to the Gulf Stream for these types of recoveries.
We are also using the Seward Johnson, which is a support ship for the submersible Johnson Sea Link 11. This is a small four-man submarine that lets us get outstanding video and photography of pieces on the bottom, and in a few moments I will show you a piece of video from that particular submersible.
The shallow water recovery, we're using basically diving teams. We're diving off the United States Navy vessel Preserver and the United States naval vessel Sunbird.
We're also using the Independence, which is one of the Thiokol SRB recovery vessels, which has a sonar platform.
We are also engaged in a wide area sonar search. We're going to search with sonar, sidescan sonar, the entire search areas that were indicated in that initial chart. That is approximately 350 nautical square miles, a significant piece of ocean to be covered, and it is going to be a laborious process to
recover that.
Next chart, please.
(Viewgraph.) (Ref. 3/7-6]
Of the 350 square nautical miles that we have to cover, as of today we have covered 190 nautical miles. In this 190 nautical miles that we have searched, we have had 227 sonar contacts that have to be verified.
The ones that we have verified to date give us 17 shuttle components, definite pieces of the 51-L mission. We have 25 non-shuttle components. These are things such as geology on the bottom, rocks, oil drums that have been discarded by ships passing through the area previous to the accident.
We have 185 sonar contacts that have not been characterized as shuttle or non-shuttle at this time. We are working as quickly as possible to characterize those, to complete our search and start the recovery of those items.
CHAIRMAN ROGERS: Colonel, when you say 227 contacts made, what does that mean?
COLONEL O'CONNOR: That means that as our sonar ships were going through the search area they had hard contacts, acoustic returns off of objects on the bottom. The objects, we don't know what it is. We know
it is large enough in all cases, probably something the size of a 55 gallon drum or larger, that we need to go out and look at and characterize visually.
CHAIRMAN ROGERS: But you have identified 17 shuttle components that you feel confident are part of the shuttle system?
COLONEL O'CONNOR: Yes. We have those components. We have videotapes and still photography of those components. We take that verification photography and video from some of our remotely operated vehicles.
CHAIRMAN ROGERS: Have you any way of knowing which of those 25-or which of those 17 might involve the SRB?
COLONEL O'CONNOR: Yes. A few charts down, I will have a chart that identifies the components on the bottom that we will attempt to recover shortly.
CHAIRMAN ROGERS: Fine.
COLONEL O'CONNOR: Next chart, please.
(Viewgraph.) (Ref. 3/7-7]
We have established a recovery priority list of the things that we feel that are important to the analysis of the failure of 51-L. Naturally, the most important item is the right SRB aft components, those components in the vicinity of where the suspected failure occurred.
We also would like to obtain significant portions of the left SRB aft component so that we can use that, if you would, as a witness plate, so that we can compare the effects of the range destruct action and the impact with the water and also be able to look at other components to see if they experienced similar types of failure mechanisms that we may find on the right-hand components.
We also feel it is important to obtain some external tank to SRB attach struts. Part of the breakup mechanism had to involve failure of some of these structural elements, and we want to recover those so that we can better understand that failure mechanism. And the crew compartment is also a priority recovery.
[1073] Next chart, please.
(Viewgraph.) (Ref-3/7-8]
As of today, we have located small portions of the right-hand SRB. We have the aft skirt assembly, which contains the thrust vector control systems, the nozzle, and a few other components from the aft of the SRB.
We also have some portions of the aft segment, some case components. In the case of the left SRB, we have the aft skirt assembly, almost the entire skirt assembly, and a larger number of aft segment
components.
We also have a forward dome igniter from the top of the SRB stack. We have not been able to recover that at this time, nor can we characterize whether it be right or left.
I would like to show you some video now. This particular video that you will see was taken by the Sea Link II on a dive in the area that we have now identified as the right-hand SRB.
Roll the video.
(A videotape was shown.) [Not published]
COLONEL O'CONNOR: There we are looking at part of the aft skirt assembly, and also the rear portion of the aft segment. As you can see in the video, here are some particles drifting by. That is characteristic of the water conditions we have in 1200 feet of water.
We're experiencing about a half a knot of current on the bottom at this time. As we go up through the water column, the current begins to increase up to about four and a half knots. It makes it a difficult recovery activity because we have to work all of our tools down through that varying current.
I also have some photographs which show more clearly the components that we have located.
(Slide.) [Ref. 3/7-9]
These photographs were taken at the same time that the video was taken, also by the Sea Link 11. As you can see, we have broken-up aft components. Most of this damage we feel probably occurred at the time of water impact of the components.
We have multitudes of parts on the bottom that must be recovered,
Next picture, please.
(Slide.) [Ref. 3/7-10]
In this picture you can see part of the clevis joint. This is not the joint that is suspect at this time. It is what is referred to as a factory joint, something that was fabricated at the manufacturing plant.
There is no indications of failure on this particular joint, although you can see towards the center right of the photograph some O-rings that are out of the groove and hanging down in the water. Naturally, it is important that we bring all of these O-rings and this clevis material up to the surface.
Next chart, please.
(Viewgraph.) (Ref. 3/7-11]
This sketch is indicative of the right-hand SRM hardware that we have identified to date. At the
left, upper left, is the top of the case, the joint area that we are trying to locate at this time. The shaded portions indicate those portions of SRB case that we have located and feel that we can recover in the near future.
As you can see from this chart, we are not very close yet to that particular field joint. We have been impeded by some severe weather conditions, as many of these fronts have been going through the area, which impeded dive operations, and we hope to have improving weather conditions shortly so we can make
DR. FEYNMAN: Could you explain a little bit better which end is which and how big this is? This isn't the whole booster, it's just a piece?
COLONEL O'CONNOR: No, this is the aft segment. This is the segment as it is brought from the factory to Kennedy for assembly with the rest of the SRB segments.
DR. FEYNMAN: And it has a skirt around it?
COLONEL O'CONNOR: It has a skirt at the bottom, at the lower right-hand corner, and the nozzle is attached to the circular opening you see also in that lower right-hand corner.
DR. COVERT: Colonel, do you think that that mid-section forward of the critical joint is going to be
in roughly the same neighborhood as the after section?
COLONEL O'CONNOR: We had initially hoped we would find it very quickly after we found these components. So far we have been unsuccessful in locating it. We have used the NR-1 to do wide-area mapping around this particular point.
At this time, we're also trying to improve our radar and ballistic studies, so that we can have a better idea of where to locate these components. We are hoping that it is in the near vicinity and we just haven't found it yet.
DR. COVERT: We wish you good luck in this enterprise.
COLONEL O'CONNOR: Thank you.
Next chart, please.
(Viewgraph.) [Ref. 3/7-12]
In view of the critical nature of the SRB components to the investigation, we have put together a team to go and develop detailed procedures for the recovery, so that we can maximize the benefit of the recovered material.
We're going through a very methodical process of getting the engineering people involved, both from the manufacturers as well as from NASA design centers. We're getting metallurgists involved so that we can do
on-site examination of the components before we attempt any lift, and they are also characterizing those components we're finding as to whether they be critical to the investigation or not. Therefore, we can prioritize the recovery.
Guidelines have been also thoroughly established and briefed to the task force on what recovery priorities we need to have, what other guidelines they feel are important to the investigative process.
Next chart, please.
(Viewgraph.) [Ref. 3/7-13]
[1075] I would like to now briefly cover some of the components we have recovered. These components were recovered both from the surface immediately following the accident, as well as some few parts that we have already recovered from the ocean bottom.
In the case of the orbiter, the significant items that we have found are a multitude of pieces of exterior skin with thermal tile. We now have the majority of the main propulsion system, the power heads, expansion nozzles, and turbo pumps.
We have located on the ocean floor but have not recovered major portions of the thrust structure of the aft part of the vehicle.
In the case of the SRB, we have both
frustums, the forward part of the SRB, the left drogue chute, some gyros, a hydraulic reservoir, and some systems from the forward skirt, as well as those components that I have previously shown you on the other charts.
On the external tank, we have 60 percent of the inner tank skin. This is of high significance, showing burn patterns and some of the failure mechanisms.
Next chart, please.
(Viewgraph.) (Ref. 3/7-14]
In order to maximize the benefit associated with the recovered components, we are directing a detailed development implementation of handling documents, procedures, for all of the components that we have in hand.
We are taking these components, running them through metallurgical analysis, chemical analysis, and reconstructing as best we can the failure mechanism and the breakup modes that the vehicle experienced at the time of the accident.
After we have correlated the thermal and blast effects and determined probable breakup patterns from the metallurgists, we will use these particular scenarios and this evidence to go back and attempt to validate the
conclusions that the rest of the NASA task force teams are putting together as far as what they feel from telemetry and optical data was the breakup failure mechanism.
In order to maximize the benefit of this process, we have brought an individual aboard of great experience.
Next chart, please.
(Viewgraph.) ( 3/7-15]
Shortly after the accident occurred, the National Transportation Safety Board was contacted so that they could provide us some of their expertise. They have released to us Mr. Terry Armentrout, who is the director, Bureau of Accident Investigation, of the National Transportation Safety Board.
He is giving us his expert assistance in directing the activity that will reconstruct the flight components that we have recovered and direct the analysis of the failure mechanisms. He has been temporarily assigned to NASA and released from his other duties with the National Transportation Safety Board. He has also been granted permission to bring whatever other assets he feels necessary from National Transportation Safety Board.
He has been of great benefit to us.
Next photograph, please.
(Slide.) [Ref. 3/7-16]
[1076] This is a photograph of the orbiter components that are now located in the NASA logistics building. Mr. Armentrout has directed the development of a grid system for placing all of the components in their proper geometric relationship, so that we can best understand the breakup patterns, the flame patterns, the other failure mechanisms that might have been incident and collateral to the breakup of the vehicle.
Next chart, please.
(Slide.) (Ref. 3/7-17]
This photograph indicates where we are mocking up the external tank components. As you can see, we're using the same technique, gridding the floors, arranging the components so they can be available for review on both sides, and putting them in the proper geometric relationships to fully understand the breakup mechanisms.
Next chart, please.
(Viewgraph.) (Ref. 3/7-18]
As a result of the review of the components we have to date, we have some preliminary estimates of what we feel the orbiter experienced and how it broke up. With approximately ten percent of the orbiter now
located, the metallurgical and chemical teams feel that we had a catastrophic in-flight breakup due to a combination of blast effects and aerodynamic overloads.
Basically, there is no evidence of anything exploding or igniting within the orbiter itself, from the evidence that we see. The external tank mockup, where we have approximately eight percent, also indicates a complete in-flight catastrophic breakup due to structural overload forces applied to it at angles and at velocities and at levels that are not normally experienced by an external tank.
We have no preliminary identification of the SRB components. We just don't have enough at this time.
I would like to also restate that these are preliminary estimates, based upon a relatively small amount of debris that we have recovered to date.
DR. COVERT: Colonel, are these loads on the external tank primarily internal loads or external loads?
COLONEL O'CONNOR: There appears to be a combination of both.
DR. COVERT: Are there places where you can identify differences?
COLONEL O'CONNOR: Yes, there are, and there
are places on the external tank where we can detect impacts from the right SRB.
DR. COVERT: So these are consistent with the pictures that we have seen, then?
COLONEL O'CONNOR: Yes, they are.
DR. COVERT: Thank you.
CHAIRMAN ROGERS: Do you have any idea of how long the process may take? And I realize it's difficult to predict, but can you give us some idea of how you think it may progress?
COLONEL O'CONNOR: Seeing that we're putting the maximum available assets to work on it, we're hopeful that within another few weeks we will have completed our mapping effort, hopefully by the 1st of April.
And at that time we will be able to characterize all of the sonar contacts we have on the ocean floor. I would like to be able to say that we would have all of the parts of interest up within a week. I think a more likely case would be it would take us at least a month or two additional to locate and recover significant portions of the right-hand SRB.
[1077] CHAIRMAN ROGERS: So that you would estimate that-and I realize that this is just your best guess. I gather you would think, then, that within
three months you might have completed your work?
COLONEL O'CONNOR: I would hope within three months that we would have recovered all necessary components to let us complete the analysis, and that the analysis at that time would be fairly far along, and that we would be doing an extensive set of cross-correlation with the other NASA design centers to understand and validate their conclusions.
CHAIRMAN ROGERS: Commission members were shown some of the reconstruction yesterday, and it really is truly remarkable the work that has been done, and I congratulate you and all of the team that have worked on it.
I notice that-what is the name for the SRB rocket, the frustum? I notice that you found both the right and the left.
COLONEL O'CONNOR: Yes.
CHAIRMAN ROGERS: Are you in a position to make any comments about the fact that the right one is quite severely damaged and the left one is not?
COLONEL O'CONNOR: Not at this time. We really need to have some more of the right-hand SRB to more fully understand that mechanism. We're also taking some photography and we're enhancing that photography at this time.
It is quite likely that some of the damage to the right-hand frustum might have been the impact damage on the external tank, but we really can't confirm that at this time.
CHAIRMAN ROGERS: Well, I guess the other possibility would be on the ocean, when it struck.
COLONEL O'CONNOR: Yes, that's also true. It could be on the ocean, and we need to do quite a bit more work in that area before we can define that.
CHAIRMAN ROGERS: Thank you. Thank you very much, Colonel. We appreciate it.
DR. KEEL: Mr. Lang, Mr. Kennedy, and Mr. Barsh.
(Witnesses sworn.)
[Please note that some of the titles to the references listed below
do not appear in the original text. Titles are included to identify
and clarify the linked references- Chris Gamble, html editor]
[1078] [Ref. 3/7-1] Search, Recovery, Reconstruction. [Ref. 3/7-2] Reference chart.
[1079] [Ref. 3/7-3] Oceanographic Conditions.
[1080] [Ref. 3/7-4] Recovery Operations.
[1081] [Ref. 3/7-5] Allocation of Assets. [Ref. 3/7-6] ?
[1082] [Ref. 3/7-7] Recovery Priority [Ref. 3/7-8] Solid Rocket Motor Components Located.
[1083] [Ref. 3/7-9] Not Reproducible. [Ref. 3/7-10] Kick Ring.
[1084] [Ref. 3/7-11] R.H. SRM Hardware identified.
[1085] [Ref. 3/7-12] Solid Rocket Motor Recovery Plan. [Ref. 3/7-13] 51-L Components Recovered.
[1086] [Ref. 3/7-14] Reconstruction and Analysis Effort.
[1087] [Ref. 3/7-15] Note.
[1088] [Ref. 3/7-16] ET Wreckage?. [see Volume 3, appendix O, page O383- Chris Gamble, html editor]
[1089] [Ref. 3/7-17] Orbiter wreckage?. [see Volume 3, appendix O, page
O382- Chris Gamble, html editor]. [Ref. 3/7-18] Preliminary estimates.
CHAIRMAN ROGERS: Gentlemen, will you identify yourselves and make reference to your present assignment and anything else you care to tell us about yourself before you start the presentation.
MR. LANG: My name is Bob Lang. I'm with NASA here at Kennedy Space Center. I have been here for the past 19 years.
My present position is chief of the mechanical systems division and shuttle engineering. I have been in that position for the last eight months. Prior to that I was in the fluids division of the same directorate.
MR. KENNEDY: My name is Carver Kennedy. I'm employed by Morton-Thiokol. My present position is as director, vehicle assembly building operations, for the shuttle-processing contractor. We perform this work under
subcontract to Lockheed Space Operations, who is the SPC.
MR. BARSH: My name is Bill Barsh. I work for Lockheed Space Operations Company at KSC. I am the engineering manager for the ET/SRB operations. I have been involved in SRB stacking since 1977.
MR. LANG: Mr. Chairman, what I would like to do is have Mr. Kennedy present a brief overview of an SRB segment stack operation, and I will follow that with a discussion of this particular joint on the right SRB to point out any anomalous conditions that we noted during the processing, if that is okay.
CHAIRMAN ROGERS: Yes that is fine. Thank you.
MR. KENNEDY: What I'm going to do is, utilizing charts, is take you through a standard flow from the time the segment arrives here at KSC and is turned over to the shuttle processing contractor until the completion of the making of the joint between the aft segment and the aft center segment, which is the one in question.
(Viewgraph.) (Ref. 3/7-19]
The first chart here, reading from left to right, the segments arrive by rail car. They weigh approximately 250,000 pounds
CHAIRMAN ROGERS: Can I interrupt to say, is this just a general description or are you talking about 51-L now?
MR. KENNEDY: Both. This is the process that 51-L followed and it's also the general pattern that we follow throughout.
CHAIRMAN ROGERS: Okay. If you can, when you present it, make reference to 51-L so that we know the difference between the general procedure and what happened in that case.
MR. KENNEDY: All right, sir.
The segments arrive by rail car. They are brought into the building, which we call the rotation processing and surge facility, or the RPSF. The rail car cover is removed and, utilizing two cranes in the building, we remove the segment from the rail car, rotate it from a horizontal position to a vertical, and place it on a support stand.
At that point, there are two heavy metal rings, one on each end, which completely enclose the joint, the clevis and the tang ends of the segment, and those heavy rings are used to tie the segment to the rail car during shipment.
After putting the segment vertical in the stand, we remove the bottom ring. In the meantime, we
bring in an aft skirt assembly, which is furnished by USBI, completely assembled and delivered to us. We place it in what we call the buildup stand.
We then pick up this aft segment and mate it to the aft skirt assembly, as shown in section 6 on this chart. While we have the aft segment suspended, we do an inspection of the tang, which is the lower end, before we mate that to the aft skirt assembly.
Following the mating to the aft skirt assembly
CHAIRMAN ROGERS: Can I ask, who is responsible for that mating? Who certifies that it is properly done?
MR. KENNEDY: That is performed by the Morton Thiokol technicians. It is signed off by Morton Thiokol quality, and in some cases, depending on the operation, there is a NASA quality buy in addition.
This particular operation, this joint right here is not a pressure vessel/joint. This is a structural joint between the aft skirt assembly and the aft segment. It does not see the internal gases from the rocket motor. It is strictly a structural attach between the skirt and the segment itself, the first aft segment.
CHAIRMAN ROGERS: But the ultimate
responsibility for seeing that that is done properly is with Thiokol?
MR. KENNEDY: Yes, that is correct.
CHAIRMAN ROGERS: Thank you.
DR. COVERT: Mr. Kennedy, is this joint usually quite concentric and the mating proceeds without any necessary procedural modifications to ensure a fit?
MR. KENNEDY: Generally. Sometimes there are problems, but generally it proceeds fairly routinely.
DR. COVERT: What fraction of the time are there problems?
MR. KENNEDY: I cannot answer that without reviewing the data, sir. I'm sorry. This particular joint being a mechanical structural joint, we do not have those data.
DR. COVERT: Is it a normal procedure on 51-L, with no exceptions?
[1092] MR. KENNEDY: Yes.
DR. COVERT: No exceptions?
MR. KENNEDY: We had no what we call problem reports written against that one.
DR. COVERT: Thank you.
MR. KENNEDY: Following the completion of that structural joint, the components that are attached to the aft segment of the solid rocket motors, which
consist of some electronics components, structural external tanks attach ring, and struts-and at this point there is a nozzle extension cone installed. It is shipped separately from the plant in Utah because of its length and size.
The nozzle itself is attached to the aft segment at the factory. We receive that here already attached. We install an extension cone after it arrives here, which is shipped separately.
Following the completion of this work, the completed aft assembly is picked up from the buildup stand and placed on what we call a transportation pallet. At that point, the top handling ring, which is still in place, having been on there since we came off the rail car, the top handling ring is removed and the clevis portion of that segment is inspected. That is the first inspection it receives.
It is inspected for any shipping damage, pitting, scratches, so forth, particularly in the 0-ring areas. This is the clevis area that has the two O-ring grooves.
That completes the processing
CHAIRMAN ROGERS: Who does that inspection.
MR. KENNEDY: We also do that. Morton Thiokol does the inspection, and I believe it is signed off also
by a NASA inspector.
CHAIRMAN ROGERS: Were any anomalies found in that process?
MR. KENNEDY: No, sir.
CHAIRMAN ROGERS: As far as 51-L is concerned?
MR. KENNEDY: No, sir.
MR. LANG: Mr. Chairman, let me restate, if you would. What Carver was doing was basically running through a standard type process. I was going to follow that up with a detailed discussion of any anomalies we found on the 51-L process.
CHAIRMAN ROGERS: Okay, fine.
MR. KENNEDY: The answer is still correct. We did not find any anomalies on the O-ring grooves.
May I have the next chart, please.
DR. FEYNMAN: These rings you're talking about, are they very strong? Do they keep the shape of the thing while it is on the car?
MR. KENNEDY: They are a weldment of steel, a one-piece weldment. They interface with the joint configuration of the end of the segment. They would offer some support, because they are strong enough to tie the segment to the rail car, and so they take the normal shipping loads and restrain it.
And of course, they are not flight weight thickness. They are heavy sections.
(Viewgraph.) [Ref. 3/7-20]
[1093] This chart shows a similar process, again in the same building, for the center segments and the forward segments that are received. The main difference here is that we attach no hardware, nor do we attach any other parts to the center or forward segments.
It is unloaded in a similar manner. The rail car cover is removed, the segment is removed, it is placed on a stand. The bottom ring is removed, it is moved to a transportation pallet. The top ring is removed.
At both of these times when the shipping ring is removed, there is an inspection made of that particular end of the segment, looking for shipping damage, pitting, rust, et cetera. We do that with the segment suspended from the crane for the bottom part or the tang end of the segments. We do it after we remove the ring on the top end, just prior to moving the segment into the surge building.
At the completion of those inspections, the segments are moved into what we call the surge facility, which is
DR. FEYNMAN: What do you inspect for? What
do you look for? Scratches, broken pieces?
MR. KENNEDY: We have a criteria which looks for pitting, scratches, contamination, dirt, rust.
DR. FEYNMAN: How much do you allow?
MR. KENNEDY: We have a specification, which allows visual blemishes, which are not detectable by a five-thousandths piece of shim stock or fingernail check. If it will not hang up a five-thousandths shim stock-now, I'm being very general, because there are various and sundry inspections for various parts. That is the most critical one. That is the sealing surface on the tang.
MR. WALKER: Is there any grease on the tang at this point in time?
MR. KENNEDY: We remove the grease prior to inspection and re-apply it afterward.
MR. WALKER: It is shipped with grease and then that's cleaned off and inspected?
MR. KENNEDY: Yes.
MR. WALKER: Are there photographs made at this point in time?
MR. KENNEDY: Not at this point.
MR. WALKER: Thank you.
CHAIRMAN ROGERS: Could I ask a question about who does these inspections. Are these long-time?
employees of yours, and do the same people do the same work all the time or do you change them around often?
MR. KENNEDY: Generally, the inspectors are assigned to SRB, and they will do the inspections on any of the SRB activities.
They are our employees. They have been employed either by Morton Thiokol or by the predecessor USBI in most cases.
CHAIRMAN ROGERS: The Commission has heard reports that some of these people are overworked because of the pressure of getting these launches off, and that therefore they may not be as capable of doing their inspecting jobs as they might otherwise be.
Have you any records of whether they are overworked or not, or whether they have normal days of activity?
MR. KENNEDY: Well, they have normal days, and they do work some overtime. It depends on the activity. For instance, when we-and we're getting ahead of the presentation, but when [1094] we do our stacking operation we do that around the clock. We have propellant exposed. It is a hazardous operation and we continue around the clock until we complete the job.
And both the personnel, the supervisory and inspection and technical personnel, do work around the
clock, and they will work some overtime depending on what the particular stack requires.
CHAIRMAN ROGERS: Do you have any reason to believe that any of these inspectors were not capable and not able to perform their duties efficiently?
MR. KENNEDY: No, sir, I do not.
CHAIRMAN ROGERS: Thank you.
MR. RUMMEL: Can you explain the fingernail check? Is this a process where the inspector is supposed to feel an imperfection?
MR. KENNEDY: The normal technique is to use a piece of a five thousandths brass shim stock. The specification which is placed on us here permits inspection by either method to determine whether there is a detectable imperfection in the surface.
CHAIRMAN ROGERS: Would you provide the Commission with a report of the inspectors who did this work on 51-L and how long they have been employed and whether they were overworked or not, how much time they had been given to do this responsible work?
MR. KENNEDY: Yes, sir, we will do that.
CHAIRMAN ROGERS: Thank you.
MR. RUMMEL: I'm still not clear on the fingernail check. How does that enter into this inspection procedure? Is the surface examined 100
percent with fingernails, or is it visual and when he finds something he sees with his eyes then he scratches with his finger?
I just don't understand.
MR. KENNEDY: It is a visual inspection. Any visual blemish is then inspected with shim stock or fingernail or both, the visual inspection first.
MR. RUMMEL: Fingernail sounds unscientific. That is why I was asking.
MR. ACHESON: Mr. Kennedy, are you going to discuss the frequent loss of roundness phenomenon in your presentation?
MR. LANG: I was going to cover that.
MR. ACHESON: Very well.
DR. FEYNMAN: This surface where, on the tang, where the O-ring is in contact with the tang, the question is how flat, how smooth that is, how deep. Is this the last test that we make? What experience do we have on that?
MR. KENNEDY: I'm sorry, I missed your question.
DR. FEYNMAN: This isn't the last inspection?
MR. KENNEDY: No, this is the receiving inspection as it arrives on rail car, to detect any anomalies that have occurred during shipment or that may
have been shipped from the plant, to detect-this is the first inspection. They are inspected again before stack operations.
VICE CHAIRMAN ARMSTRONG: What is the experience on those inspections? Have you found rejections?
[1095] MR. KENNEDY: On reused hardware, which is the majority of the hardware we see now, we do see some what I would call arrested corrosion. That is, it has been immersed in salt water, but it has been arrested, detailed, and used and sent back. It has been inspected at the plant and approved for use and sent back.
The inspection here, if there is any question about any of them, we take dental mold impressions and have it inspected. There is a criteria which comes, which is applied to us here at Kennedy, that we're not allowed to have any imperfections of certain dimensions and certain sizes, depending upon locations.
If there is any question about that, we take dental impressions and have it inspected and measured to see if it meets the criteria or not.
CHAIRMAN ROGERS: You may proceed.
MR. KENNEDY: May I have the next chart, please.
(Viewgraph.) [Ref 3/7-21]
This chart shows the process used in the vehicle assembly building. This is the large building where we actually assemble the solid rocket boosters. The segment, the aft segment, is brought in first by a transporter on a transportation pallet in the vertical orientation.
It is brought in and placed in place in front of the particular high bay that is to be used in assembling the vehicle. The lifting beam is attached. The segment is picked up, transported up into the high bay, and placed in position on a mobile launch platform.
At that point we install four holddown studs, which is the stud used to retain the entire assembly and to hold it to the launch pad until the time for launch. And we conduct a tensioning operation, which is an operation where we tension the stud and then run a large nut down to retain it, so we will retain the tension on the stud that is a compressive load between the skirt and the mobile launch platform.
This is typical of both sides, this operation. That is called placing the aft-placing the aft booster and tensioning.
May I have the next chart, please.
(Viewgraph.) [Rev 3/7-22]
The aft center segment is brought in from surge in a vertical orientation, placed in position, lifting beam is installed. That segment is lifted off the pallet, and at this point again the lower end or the tang end of the joint is cleaned. It has been previously greased. It has been in storage with grease.
It is cleaned, it is inspected again for pitting, corrosion, scratch, nicks, damage of any sort, contamination. It is an inspection by a company inspector, followed by a NASA inspector.
DR. FEYNMAN: How irregular is the surface of the tang?. How many wiggles and how many scratches? How deep are they?
MR. KENNEDY: I'm sorry, sir?
DR. FEYNMAN: How good is the surface of the tang? How deep are scratches and so on?
MR. KENNEDY: Well, they are controlled by specification, and there is a myriad of those. I would be glad to furnish the exact specification, depending on the location.
For instance, in the pinholes where the retaining pins go through there is a criteria. There is a criteria on the O-ring sealing surface. There is a criteria on the non-sealing surfaces at those points,
and that is a fairly detailed set of specifications.
DR. FEYNMAN: Are all of these criteria checked when you mate these?
MR. KENNEDY: Yes, they are inspected for that series of criteria.
MR. ACHESON: How frequent is it to find imperfections that are serious enough so that that particular segment is recycled or sent back to the factory for further processing.?
MR. KENNEDY: I don't believe we have sent one back to the factory. There is a procedure by which blemishes or pits or scratches can be blended out and re-inspected, and that happens occasionally.
That requires what we call a problem report and what we call a material review board action if one of those is developed, which means it has to go back to the development center and the manufacturing plant for approval. And they provide the information on how they want it repaired or done.
MR. RUMMEL: Have you had any such cases in the ring grooves that you can recall?
MR. KENNEDY: Where we had to rework the O-ring grooves?
MR. RUMMEL: Yes.
MR. KENNEDY: I cannot answer precisely, but I
suspect there may have been one during the program at least. I cannot recall precisely.
MR. RUMMEL: I would be quite interested in knowing that if you could find it out.
MR. BARSH: We have had some scratches in the O-ring grooves that we got concurrence with the design agency to go in there and polish them out using a fine-grit sandpaper.
MR. RUMMEL: I assume there are specific tolerances?
MR. BARSH: Yes, there is.
DR. COVERT: Are there detailed instructions on how you remove the results of the fine-grit sanding? Do you blast it and then inspect it with a magnifying glass?
MR. BARSH: After we use the sandpaper, we then go in with a cleaning solvent and take away all of the grit and wipe it clean, and regrease it again.
DR. COVERT: How far on either side along the length do the instructions ask you to clean it?
MR. BARSH: I would say that the groove is very small, and you have got to try and put a piece of sandpaper in there, and you sand it in a circumferential motion. You probably go about a foot on each side.
DR. COVERT: Thank you.
CHAIRMAN ROGERS: Thank you.
MR. KENNEDY: Following the inspection of the tang-and the segment is hanging from a crane in what we call transfer aisle-measurements are taken at six locations around the tang end 30 degrees apart, to determine the diameters in those planes.
In the meantime, the segment which is on the MLP-that is, the aft segment with the clevis end up-is similarly inspected again and similarly measured in the same planes.
These measurements are compared and, assuming that we are within our criteria or our own self-applied criteria for successful mate, the segment is lifted over and placed. The putty is installed on the aft segment end or the clevis end.
[1097] The O-rings are emplaced, the segment is mated, and the pins are installed. The leak test is conducted and the closeout or the installation of the retainer band, and the external insulation is applied over that, on that particular joint.
And that is the standard flow. That is the flow that the 51-L right aft segment and the right aft center segment followed.
MR. LANG: Now I would like to back up, if you will, and address each of the general areas that Carver
talked about, and starting with the aft segment operations in the processing facility.
(Viewgraph.) [Ref. 3/7-23]
As he stated, we didn't have any anomalies of any kind on that particular segment in that facility, but I did want to mention
CHAIRMAN ROGERS: Are all of your remarks going to be directed to 51-L?
MR. LANG: Yes, sir. In fact, all will be directed to the two segments on the right side booster, the aft and the aft center.
On the aft center or in the aft booster assembly, part of that operation involves, like Carver said, lifting the handling ring off the segment. That went nominally on this segment. And I do want to point out, originally in the processing of the segment intended for 51-L, the original booster-excuse me. The original segment intended for the left side forward center segment, there was an incident where the procedure was not properly followed.
The clevis end was damaged in the lifting operation. We took steps to correct the procedures to clarify all of the operations.
CHAIRMAN ROGERS: Could you be a little more specific? We have heard about this, but now it would be
helpful if you could be real specific about this particular incident: What happened, who did it, what was the result, and so forth?
MR. LANG: I didn't bring a lot of information with me, but basically what happened, in the handling ring as delivered from Thiokol is in two pieces, two general pieces. One is the solid ring that the crane hooks actually attach to. That is bolted to what we call a segmented ring that actually is-it is like an elbow shape, that fits down inside the clevis of the segment.
The procedure is intended to loosen all of the bolts between the solid ring and the segmented ring, and then to slightly lift up on that solid ring to relieve the load so the pins can be removed from the segmented ring. The segmented ring is the part that fits down in the clevis.
For whatever reason-and I can't give you the answer right now, but I can provide you a copy of the report-the bolts between the solid ring and the segmented ring were not loosened, so therefore there was no slop or tolerance in between those two particular rings.
We had a second thing happen to us, naturally. The lifting operation, the guys who were
running the procedure, running the crane, were using a load cell to determine how much lift they were lifting up on, how much load they were lifting up on.
As it turned out, the particular load cell had failed. The operator was trying to lift a certain weight, thinking he was just lifting the ring. The load cell wasn't giving him indication, so he kept lifting. And at that time we had all but 31 of the pins removed, and he kept lifting.
[1098] And what happened was, he wound up trying to lift the total weight of the segment on 31 pins, and most of those pins were in the same quadrant. So it was just an unbalanced load, trying to lift too much weight on too few pins.
The damaged area was restricted to the segmented ring. One of the segments actually broke, physically broke, and I think it was two or three-and Bill, correct me if I'm wrong-of the clevis pinholes were slightly elongated, therefore damaged.
That segment was moved aside and another segment put in its place. All of the procedures, by the way, were modified such that all of the segments that flew on 51-L had had their rings removed with the modified, updated procedure. And we will be glad to provide you whatever details you would like.
VICE CHAIRMAN ARMSTRONG: What happened to the rejected segment?
MR. LANG: I think it's still here. We have a team from Thiokol coming down to do a full inspection to determine flightworthiness. I don't know the results of that inspection. Bill or Carver, if you know
MR. KENNEDY: I don't think we know here. The segment is still here in storage.
CHAIRMAN ROGERS: Is there any possibility that that incident contributed at all to the accident of 51-L?
MR. LANG: No, sir. We took special steps to go back and review that incident, the corrections to the procedure to prevent that kind of an incident, and all of the segments that flew on 51-L had been handled with the modified procedure. And we are very confident that that did not contribute at all.
CHAIRMAN ROGERS: So none of the damage that was involved in that incident could possibly have affected the 51-L launch?
MR. LANG: That is correct.
DR. RIDE: Just to be perfectly clear, the segment of the solid rocket that was damaged was not used on 51-L, so no damaged hardware from this incident flew?
MR. LANG: That is correct.
DR. COVERT: Mr. Lang, are you going to talk about whether or not this had any influence on the stacking of the right-hand booster?
MR. LANG: I can mention that. It had an influence in this manner: The booster that was damaged, like I said earlier, was the left side forward center segment. And I don't understand the propulsion match from left side to right side. It is a performance matching that maybe Carver can address.
But to match the performance characteristics of the left side booster to the right side, the right side aft center segment, which is one of the segments that form the aft joint, was swapped out. And we did at that time bring in another booster, another segment, to be used for the right aft center segment for 51-L.
DR. COVERT: Would this require unstacking anything?
MR. LANG: No, sir. We had not at that time started the stacking operation. We were still just handling the segments off of the rail car and preparing them for transfer into the VAB.
DR. COVERT: Thank you.
MR. RUMMEL: When you speak of matching, are you speaking of burn rate?
MR. LANG: Like I said, I really don't understand the performance matching, but I was told that was the reason for swapping out a right side segment.
DR. FEYNMAN: The variations in the propellant from different batches of making explosives is relatively large, and so what happens is two segments are poured from the same batch so that one will go on the right booster and one will go on the left booster, and they will burn at the same rate.
Then when the next pair of segments is put in they may burn at a different rate, but they will be the same on the left and right, so that both rockets will go always the same.
So if you have some trouble with one of the segments on the left side, it isn't right to leave the right one in because it may burn at a different rate, and it is better to take the other segment which was poured from the same explosive batch as the one that you just spoiled and use that on the other side with the new one that you put on the right side. And that is why it is done that way.
MR. LANG: That was the extent. There were no other anomalies, then, on the handling of the aft segment.
The aft center segment that we wound up moving
into stack for 51-L, in its process, like Carver explained the inspection, we found two things in the detailed inspection.
One, we found a de-bond between the rubber insulation and the metal case on the lower end, the tang end. At that time we picked up a problem report and measured the depth of that de-bond area. The criteria is we are allowed to have de-bonds, by the way, up to five thousandths deep-excuse me, 50 thousandths deep.
We measured that particular de-bond and it was greater that 50 thousandths, and so we followed a standard repair procedure for that type of anomaly. We packed that de-bond area with an epoxy sealant, covered it with putty to protect it. And in parallel with that, we took a sample of that epoxy sealant, set it aside for cure test later on.
And I might note that that type of de-bond anomaly is fairly common. We have seen several on individual segments before. I think probably every stack has had one or two of those type anomalies. We have seen depth of de-bond areas much greater than I think this particular one was, 109 inches deep. And we have seen de-bonds much greater than that.
So we followed the standard repair procedure
and packed that de-bond area with sealant and continued on with the rest of the inspection. The discussion that we just had about the tang inspection, there was in the area of between circumferentially 252 degrees to a little more than 300 degrees, there were pitting. There was pitting noted in that area of the tang at the O-ring sealing surface.
We did apply the criteria of using a five thousandths shim stock to see if we could tell any feel at all. Any feel at all would have driven us to, number one, a problem report condition; number two, mold impression to actually measure that depth.
In this case, there was absolutely no feel with the shim stock, with the five thousandths shim stock, and so we noted it in notes and comments in the processing log as an observation that met the criteria. It was not a problem report condition, and so we again pressed on.
MR. RUMMEL: Was the de-bond and the repair in the area where the smoke was seen to come out or some other location?
[1100] MR. BARSH: The de-bond was located between 165 degrees and 168 degrees.
MR. RUMMEL: And how does that relate?
MR. BARSH: Pardon me?
MR. RUMMEL: How does that relate to the area where the initial smoke emission was observed?
MR. LANG: We think the area of concern is in the area of 300 degrees.
MR. RUMMEL: Thank you.
VICE CHAIRMAN ARMSTRONG: Does that mean that there was no burnishing or other finishing of the imperfections in the tang in this case?
MR. LANG: That is correct.
VICE CHAIRMAN ARMSTRONG: Could there have been on prior occasions, prior inspections or prior uses of this segment?
MR. LANG: Of this same segment? We researched that and we couldn't find any evidence of any documentation showing any anomaly in that or any other area on that tang. We went back through the data packs.
VICE CHAIRMAN ARMSTRONG: If there was finishing or burnishing, it would be recorded in the paperwork?
MR. LANG: I can't be sure of that. We think that it should be if they did any kind of what I would consider rework, but I am unsure of this. I probably shouldn't say it, but we think that at Thiokol they are allowed to do certain minor polishing, cleanup, if you
will, without calling it rework. And therefore, if that was the case it would not appear in their rework records.
DR. FEYNMAN: You say you took an impression of this apparent place?
MR. LANG: No, sir, we did not. I didn't mean to imply that.
What we would have done had the shim stock caught or given us an indication of feeling something, we then at that time would have taken a mold impression to actually measure the depth of the anomaly. But we did not feel anything and so we did not take a mold impression.
DR. RIDE: Do you know how common it is to note pitting like this, that is less than your criteria? Has that happened on previous flights, or have you had a chance to go back to the logs?
MR. BARSH: Yes, it happens quite often that it is less than-I mean, it is an acceptable criteria. We had pitting in six locations, six degree locations on that tang.
DR. FEYNMAN: In the handbooks about O-rings and the conditions in which they're supposed to be used, they talk about surfaces having something like, they say, 64 RMS. I don't know what that means. And I would
like to know how much RMS, or however you describe the character of the surfaces that you have on the tang that the O-ring is in contact with, so I can compare it to normal practice.
But I need to have some way of converting I don't know what the "64 RMS" means, and I don't know what it means that you scratch with a fingernail or a piece of shim stock and find no five thousandths. I can't convert one number to the other. Can you help me?
MR. BARSH: Well, our inspection criteria doesn't require us to check the surface finish. We have to look at for visual imperfections, visually only. And if we-see something like this, in the case of the pitting then we will use the shim stock method to try and determine if that is a new pit or if it is a pit that has been there and reworked.
[1101] DR. FEYNMAN: Okay. Now, when you rework it and you have some sandpaper, which, as you said, very fine or whatever it is, and you polish it away, that five thousandths thing that you discovered, then you've created a surface with some kind of an RMS. What is it?
MR. BARSH: I don't know, sir.
DR. COVERT: Mr. Lang, could you describe
you cut a little piece of shim stock. At the time you use it, is it sort of pointed or rounded, just so I can have a feel?
MR. LANG: I'll let Mr. Barsh describe that. He's witnessed that before.
MR. BARSH: It is a piece of shim stock, probably about a quarter of an inch wide, and it has a rounded tip on the end. It is not a pointy tip, but it is a rounded tip, and
DR. COVERT: How long is it?
MR. BARSH: Probably four or five inches long; just something to be able to hold in your hands. The inspector can hold it in his hands and also to get down into the O-ring grooves if they see something in that area.
DR. FEYNMAN: What do you do, move it along and see if it gets caught?
MR. BARSH: That's right, see if you snag. If you have a pit, that is like a little crater, and if you snag the shim stock against the crater then you have a problem. But if it runs across there smoothly, you don't have a problem and it is an acceptable condition.
DR. COVERT: Is there a relation of the roundness of the edge to the maximum acceptable pit size?
MR. BARSH: I don't know, sir.
CHAIRMAN ROGERS: Okay. Go ahead.
MR. LANG: Okay. The rest of the handling in the processing facility of that segment went well. We completed-we moved it to what Carver described as the surge facility, which is the storage facility, on November 26th.
On December 5th, we were then moving it into the VAB to begin the stacking operation. We have a safety requirement that says we can't have more than two segments in the VAB at one time unstacked-only one allowed in the transfer out, excuse me. We had one in there, and so we left that segment outside for a day.
What we were doing-and I will get to this in a little bit-we had a problem with the right side aft segment, mounting that on the mobile launch platform. We had a problem with the holddown post, and I will describe that in just a minute.
But what we did, we left that right aft center outside for a day. It got caught in a very heavy rainstorm, and we noticed water coming out from underneath the plastic cover on the lower end of the segment.
At that time we picked up a problem report, removed the cover, cleaned up all of the water from all
of the visible accessible areas, cleaned up the tang so that it was perfectly dry, did the same thing to the top end, cleaned any water off the insulation and the clevis, cleaned it out, and closed the problem report as cleaning up the water.
DR. COVERT: What day was that?
MR. LANG: I think that was December 5th, was the day we moved to the VAB. I think the next day was the 6th, was the day we discovered the water and cleaned it up.
[1102] DR. COVERT: Thank you.
MR. LANG: Okay. Now, getting back to putting the aft segment on the MLP, typically what we do, we alternate left to right side each segment, aft, aft, aft center, aft center, and on up the stack. We had a problem in this case.
When we mounted the right side aft segment to the holddown posts-those are the four posts that the booster sits on and actually gets bolted in place-the holddown post number one is the post that is most inboard and on the orbiter side. It is inboards toward the tank, and on the orbiter side, on the right side booster, we had a problem.
We have to tension the bolts so that the bolts are loaded such that they will not relax, even through
all of the orbiter mating operations, the external tank servicing, and even through SSME ignition. You've heard of the twang concept. When the main engines start before booster ignition, the bolts-the load on those bolts, holddown posts, will relax as the vehicle rotates.
And the requirement is to make sure the tension bolts, the stud tensioning bolts, and the holddown posts never slack, so there's always a positive tension in those bolts under all those conditions.
We measure the tension of those bolts with an ultrasonic transducer. The number one holddown post had a transducer that failed, and so we had to go and replace that booster-or that holddown post assembly or that portion of the holddown post assembly, the stud, the washers, the nuts, or the nut, that whole assembly.
Because of that work, we went ahead and stacked the entire left booster. It doesn't have any bearing, but it is just something that is different in the process. We normally rotate. In this case we stacked the whole left booster and then came back to the right side.
DR. RIDE: Had you ever done that before?
MR. BARSH: Yes, we have done that before.
DR. RIDE: Which flight? Do you know
offhand?
MR. BARSH: Not offhand, I don't.
MR. LANG: Okay. We went ahead and put the right side, replaced the number one holddown post hardware, had no other problems with holddown post numbers one, three, or four.
Holddown post number two, however, we couldn't achieve the total desired tension on the bolt. The process is, we put a tensioner device, hydraulically operated tensioner device, on the end of the bolt, onto the stud, and pull it. And I don't know the forces off the top of my head. I'm sure Bill has them here, but pull it such that we get the maximum load in the bolt, and then torque a nut down to hold that bolt, and then relax the tensioner, and the tension between the head of the stud and the nut remains at a high load.
Well, it appeared we couldn't-then, by the way, we measured the transducer to see what that result in tension is. We couldn't, after removing the tensioner from the end of the stud, read the kind of tension we were shooting for.
We did, however, have a load in that stud that was more than adequate for actual stacking the booster. But we felt we were marginal for the rest of the loads on through SSME ignition.
What we decided to do, however, was go ahead and stack that booster all the way up with that bolt tensioned to a good enough load for that stack, and after the stack was over we came back [1103] and replaced that bolt. We relaxed the tension on it, replaced the bolt, the stud, the nut, the washer assembly, and re-tensioned it.
MR. RUMMEL: When you replaced the bolt, didn't that place a sizable eccentric load in the stack?
MR. LANG: No, sir. The load in the stack stays the same. The load in the bolt-the stud is there just to squeeze the skirt of the booster to the holddown post.
MR. RUMMEL: In other words, it's sitting on the surface and the bolt simply holds it down?
MR. LANG: Yes, that is correct. And again, we've done this operation, either replaced or at least de-tensioned bolts, I think on four previous occasions.
Okay. Let me then go through the rest of this, the mating. We went through the putty application and I think all of you saw what that putty application looks like on the aft segment. We-this time frame now is December 7th.
We lifted the right aft center segment in the
VAB in the transfer aisle, and there was a picture on the screen of that lifting operation.
(Viewgraph.) [Ref. 3/7-24]
The initial lift, as you can see, is what we call a four-point lift. There are four-just that, four points on the top of the segment that equally carry the load. It is at this point we do our initial what we call rounding measurement, the circumferential checks. And we have a criteria. We have a criteria that says the outer diameter of the tang cannot exceed the inner diameter of the outside leg of the clevis by more than .25 inches.
I think at this point if I could get chart 7 up, I would like to show that before I go into the use of the rounding tool, just in case there is any-and zero in, if you would, please, on the top half of that.
(Viewgraph.) [Ref. 3/7-25 & 26]
What I've shown here is the tang OD with relationship to the clevis, what I will just call the clevis ID, and you can see it's the outer leg of the clevis. That particular measurement we would call zero. The tang OD and the clevis ID are the same dimension and therefore that would certainly fit our criteria and we would go ahead and mate that.
And by the way, we take this dimension, this
measurement, six places, six locations around the circumference of the segment.
If you would go to the lower one of that page, please. Here I tried to show what we would call plus. Again, if you look at the right side, the tang OD and the clevis ID are coincident, because we measure from the same point every time. But in this case the tang ID is slightly greater than the clevis ID, and that would be a plus and our criteria-let me back up.
I said our criteria is 25. That is our goal. Our requirement criteria really is to find that there is no flat metal on flat metal. And the way I drew this, ironically, it turns out I showed it flat on flat. If you can look at the left side, the bottom horizontal face of the tang, if you brought that straight down, would be flat on top of the upper flat level of the outer clevis leg.
That would violate our criteria. We would not mate that in that condition. As it turns out, that is a condition we had on this particular segment. We want to-now could you go back to number 5, please.
(Viewgraph.) [Ref. 3/7-25 & 26]
MR. SUTTER: How far out was it?
MR. LANG: We had a delta, if you will, of .512 inches of tang exceeding the clevis ID.
MR. SUTTER: And the limit was .25?
MR. LANG: That is correct.
So we went to an option that we have used many times, what we call the two-point lift. If you can see-how to describe it? The lower right and the upper left legs of the four-point attach, those are hydraulically controlled such that we can put a man in a boatswain's chair, or I guess we use a High-Ranger or a Condor, and relieve the hydraulic pressure at those two points, so that we are now hanging the total load by the other two points.
And what that does for us, experience shows that as the booster is hanging, all of the propellant weight and mass tends to sag the segment, and the experience shows that-you can see the left side of the booster, the line running up and down, is the 90 degree point. That is the cable tray.
When the booster is shipped, that portion is up. The 90 degrees from that seems to be, in our experience base, the elliptical part that the booster seems to come together, form somewhat of an ellipse. And so what we do, we hang at the zero and 90 degree points-I'm sorry, zero and 180. And the other two sides tend to sag in and come in together.
Now, with the two-point lift we took three
sets of measurements, and the best we could do is improve to a .334 positive difference. At that point, if you go to the next chart, please
(Viewgraph.) [Ref. 3/7-27]
-we installed what we call a rounding tool or what I call a rounding tool. The guys call it a circumferential alignment tool. But the intent of this tool is to put a squeeze on the segment itself and try to deflect, bring the segment back into a dimension that will match up with the clevis.
We've used this type of tool six previous times, and the first time we used it we had a design that was strictly a manual mechanical design. It didn't have-it was not a hydraulic system. It was just a rod with nuts on both ends, and we would twist the nuts and just measure the deflection.
We had a criteria not to exceed a half-inch deflection using the rounding tool, and in this case we, like I said, we started off with the dimension of .334, and with the rounding tool in place we were able to deflect an additional .236 by squeezing in with the rounding tool.
MR. SUTTER: So in effect you changed from .51 down to almost zero?
MR. LANG: Actually, we changed from .51 down
to .098.
MR. SUTTER: So about .4, roughly?
MR. LANG: That is correct, a total deflection of about .414, 1 think it is.
MR. SUTTER: What could this do to the bonding and the insulation? You had some de-bond. Would this bring on any more de-bonding?.
MR. LANG: Our experience has shown that it has not, and the Marshall designers and Thiokol designers agree that that is not going to be, that is not a concern.
MR. SUTTER: What does cause de-bond? Is it transportation loads, or does anybody know?
MR. LANG: I don't know.
MR. KENNEDY: I think you would probably best ask that of the Morton Thiokol plant design folks. It is not an uncommon phenomenon in these segments to see. And these are minor.
[1105] These are two or three degrees in arc, 100 thousandths or so, and it is probably related to the method of manufacture of the internal insulation in the case.
But I'm not qualified to answer that question.
MR. SUTTER: I was wondering, though. Something starts it. There must be some kind of a load,
and would that load change if you take it out of round from .5 to 4. 1 was just curious. You started out with some de-bond, and then you had to move the joint roughly four tenths of an inch, and that has got to put some kind of a load, and could it effect-could it bring on more debonding?
It builds a load in there that wasn't there. It has to.
MR. KENNEDY: I don't feel, since I'm not familiar with the design of that rocket motor, that I'm qualified to answer. I could get the answer for you on that particular subject. That requires some propellant mechanics and some bonding expertise that we don't have here at Kennedy.
MR. SUTTER: To change that joint four tenths of an inch and do it all with, say, the hydraulic unit, if you did it all with the hydraulic unit, what would be the load pushing on the side of the case? Do you know that?
MR. LANG: We have a conversion, reading the hydraulic pressure, a calculation that says 1200 psi hydraulics on that, on the piston, would equate to 3,000 pounds force into the segment. Thiokol has stated we could go as high as 5,000 pounds force on the segment with no problems.
MR. RUMMEL: For clarification, on the quarter-inch limit, does that mean that if you have a differential dimension of a quarter inch between segments that you go ahead and mate? Is that the correct interpretation?
MR. KENNEDY: No, that's not exactly correct. A quarter inch is a self-imposed criteria. Based on experience, we know that if we are within a quarter of an inch interference fit, if you will, that the chances are we can successfully mate that segment to a clevis without reaching the, if you will, the flat on flat, which is the actual controlling criteria.
The criteria that are placed on us in all of these cases we keep talking about are placed on us by a design agency, and we're not allowed to have the end of the tang, the flat end metal end of the tang, contact the flat end of the outer clevis leg. And that is the controlling criteria.
The quarter inch is a factor we have developed here with experience, saying that if we have no more than a quarter, an apparent quarter inch interference, we can in fact avoid a flat on flat condition and attain a successful mate.
MR. RUMMEL: One more question on this line. The one unit was 512 thousandths out of round. Is there
a limit as to the maximum out of roundness that would be acceptable, beyond which you would not try to squeeze it back into shape? Is there no limit?
MR. KENNEDY: No, sir.
DR. COVERT: Mr. Kennedy, if you have this thing within a quarter of an inch and then the beveled part of the tang slides into the clevis, there is a metal on metal action at the time that this mating takes place.
MR. KENNEDY: Yes.
[1106] DR. COVERT: Is it possible that chips or cracks or scratches result from this, and that ultimately might affect the surface smoothness so that it would exceed the 46 micro-inches?
MR. KENNEDY: The section we're talking about right now, of course, is the outer clevis leg and the tang, which is
DR. COVERT: That gets covered up by that shim anyway?
MR. KENNEDY: Yes, that has a shim applied there. And keep in mind that both components are greased when they are assembled. They have a coating of grease of both clevis, and tang components.
DR. FEYNMAN: How much is that-when I look at the picture of the tang, I see there's a kind of a
beveled edge. That is not a sharp corner of the tang.
MR. KENNEDY: It does have a chamfer on it, yes, sir.
DR. FEYNMAN: How deep is that? That is, how far does it come in from the wall? Do you know?
MR. KENNEDY: The length, as I remember from the drawing, is .34, the length. That is up the length of the tang, and the angle called out on the drawing is 25 to 35 degrees. And so I believe mathematically
MR. LANG: We calculate it comes out to .196, if you assume a 30 degree angle.
VICE CHAIRMAN ARMSTRONG: Has your experience been that the major axis of the ovality tends to be predictable based upon shipping considerations and so on?
MR. LANG: Yes, it does.
VICE CHAIRMAN ARMSTRONG: Can you identify what the tendency is?
MR. LANG: The tendency-if we could go back to number five, please.
(Viewgraph.) [ 3/7-28]
MR. LANG: The tendency is that, since the cable tray, as you can see on the left side of that photograph there or sketch-the vertical line is the cable tunnel. It is shipped up, and therefore that is
at 90 degrees. And our experience is at zero and 180 is the major axis.
And so while it's on the rail car with the 90 degrees up, it tends to flatten somewhat, if you will, and that has been our experience.
MR. SUTTER: How many times have you had to use a load of as high as 3,000 pounds to get your roundness criteria? Do you have the records on previous segments?
MR. BARSH: We have always had to go up to the maximum 1200 psi on the pressure gauge, each time. The deflection may be different, depending upon that particular case and propellant, but we pump it up to 1200 psi.
MR. SUTTER: And how many times has that happened? How many cases have you handled, and what percent of the time do you go that high?
MR. BARSH: Well, we've used the rounding tool six times, and every time we've gone up to 1200 psi.
MR. LANG: Let me back up. The first I think three or four, we didn't even-we weren't able to measure anything except deflection. It was all twist the nut and squeeze the case. So we actually couldn't measure any force at all.
And the criteria, the only criteria we had at
that time was the maximum allowable deflection of a half an inch.
MR. SUTTER: Well, what percent of the joints have you had to use this technique?
MR. LANG: It's six out of how many? It's a lot.
MR. KENNEDY: I would like to calculate that, because there are six joints per assembly, and I need to go back and see the first assembly that we used it on. We have not used it since the beginning of the program. It has been in use since approximately December 1984. The first use of the rounding tool as such occurred in December 1984.
MR. SUTTER: It's a little bit unusual, then. That is what I am asking: Is it unusual or not?
MR. KENNEDY: We don't use it every stack, that is correct.
MR. LANG: Bill has got some experience earlier in the flow, before we had a rounding tool, where we would hang a two-point configuration, trying to let the propellant weight and mass bring it into round. And didn't we have some fairly lengthy two-point hang times?
MR. BARSH: In the earlier part of the program, we had a different lifting beam than what you
see on the monitor there. It interfaced with the handling ring, which stayed on the segment until after we stacked that segment, and then we removed the handling ring.
We had four lifting eyes on the handling ring, where we could take this two-point lift and switch axis in order to make the segment round in the way we needed to mate it. That was quite frequent, that we had to do that switch from a zero-180 axis over to a 90-270 axis in order to get the segment to egg out.
MR. SUTTER: As you started the program, you had new units, and now the units have been used one or two times. Have you noticed any difference between new and used units on having to use this technique?
MR. BARSH: I really can't answer that.
MR. KENNEDY: We would have to go back and look at the pedigree of the segments and which ones had new and which ones had used parts and which ones we used the rounding system on. It is not uncommon. As Bill said, it is not uncommon to have to shift from a fourpoint lift to a two-point lift and remain suspended from the crane for some number of hours to allow the mass of the segment to realign and bring it closer into the conformance to the clevis we're trying to put it in. That is, just about every flow we do that.
MR. ACHESON: What is the longest it has taken, that you can recall, to achieve enough roundness to assemble?
MR. KENNEDY: That is probably before the rounding tool was developed, and it may be that the rounding tool was developed as a result of a situation like that, where it had hung a number of days and would not round up. And I would have to say, that would probably be in the neighborhood of days.
MR. LANG: We will have to find a good answer for you, because I think that data is available.
DR. FEYNMAN: You've explained that you can't have the tang OD bigger that about .25 in your experience, to avoid metal contact. On the other side, suppose the tang OD is too small. Is there some criteria about how much too small it is before you get into trouble again on the other side?
MR. LANG: Could we have chart number 8, it is please.
[1108] (Viewgraph.) [Ref. 3/7-29]
I have already showed you the zero and the positive. Your question would be the negative side of that same measurement. The top half there shows a case where the tang OD is in fact smaller, like your described, than the outer leg of the clevis ID.
Now if you would go to the lower, what I showed here is what we would consider a case where the negative number would be such that the inside, if you will, of the tang, it actually comes down in the same line, the same surface of the outside diameter of the inner clevis leg.
DR. FEYNMAN: What minus is that?
MR. LANG: Well, that minus, that depends, of course, on the actual dimensions of the tang and clevis in question. What I did on my own, I took the numbers off the drawing, took the worst case tang, the thickest tang by design allowance, and the narrowest clevis opening.
And I don't remember those precise numbers, but that dimension as shown there with the worst case tang and clevis would be a .320 negative difference.
DR. FEYNMAN: I understand it was minus .393, actually, when you went to mate these.
MR. LANG: That was our largest negative number recorded, that is correct.
DR. FEYNMAN: In other words, the situation that you have drawn, it is even worse.
MR. LANG: Well, it may be. But again, 320 is based on the worst case tang and clevis dimensions. I don't know the exact dimensions of this particular
joint, although if you assume they are the worst case tolerances then that .393 would have been this case, that is right.
DR. FEYNMAN: In other words, this piece will come down and push directly on the O-ring as it is coming in.
MR. LANG: And it will squeeze the O-ring as it comes in. The chamfer of the tang would in fact ride the O-ring side and push it into the O-ring groove.
DR. RIDE: Is there a chance it could have come down metal on metal on that side?
MR. LANG: We don't think so, although it is hard to see, because you can see it is on the inside as the segment comes down. To actually get metal on metal, again, if you will take the worst case, the 320 worst case of those surfaces coincide, the chamfer on the tang, if we say .2 is the correct dimension, it is certainly close to that.
Add that to the 320. What is that? Five something, and all that says is you're going to ride down at minus 5 something, you will ride down the chamfer, just like we do on the outside. Metal on metal I think would be a larger number, and I don't know what that would be. We could probably calculate it or something close.
DR. RIDE: Could you tell us where roughly circumferentially the minus .393 occurred?
MR. BARSH: It was at 120 degrees, 120 and 300.
MR. KENNEDY: I don't think we answered all the questions. I think the first question that started this discussion was do we have any criteria placed on us on a negative. We do not.
And this segment, by the way, was not the maximum negative we have assembled.
DR. FEYNMAN: It's possible we would have some rubbing and produce some metal shavings in there by the rubbing of one piece against another?
MR. LANG: I don't think we can say that it is impossible, but the surfaces are inspected for any kind of metal deformation, raised metal that might cause scraping. And certainly, like Carver stated, the surfaces are greased, and our experience has been, taking them apart-and [1109] we had-like Carver said, we know we've had several cases bigger than this, and we have had probably quite a few that had metal on metal, not quite this bad, if we assume 393 was bad.
But we still haven't seen, taking the boosters apart, that kind of evidence or that kind of damage.
MR. RUMMEL: Do you know how much out of
roundness occurred in this particular joint in 51-L after the shims were applied?
MR. LANG: I'm sorry? Say again, please?
MR. RUMMEL: Do you know how much out of roundness existed after the shims were applied to this joint?
MR. LANG: No, sir, we don't have any way to measure that.
MR. RUMMEL: One of the purposes of the shims is to help to keep it round, is it not?
MR. LANG: It's my understanding it is to help the squeeze on the O-rings, that is right, that is its primary purpose.
MR. ACHESON: When you have this condition, have you ever, after seating the upper segment, have you ever lifted it off again to see what the effect was on metal to metal contact or on the O-rings?
MR. LANG: The only time we have pulled a segment apart is if we had-if we failed a leak check of the O-rings, and I think that has happened on three different occasions. And as far as I know-in fact, I can state that there was no metal damage.
MR. ACHESON: Have you examined-where you had these extreme out of roundness problems, have you examined the segments that have been recovered to
ascertain metal to metal damage or damage to the O-rings by the tight fit?
MR. KENNEDY: The inspection of the recovered hardware is all done at the Morton Thiokol plant in Utah. We don't do any of that inspection here, and that data would have been recorded and examined there.
I am not aware of any indication of any damage. But their post-flight reports would have indicated that if they found it.
MR. ACHESON: During the stacking process, the inspectors who initially inspect the work, J take it these are Thiokol direct employees?
MR. KENNEDY: Inspect the tang and the clevis, yes.
MR. ACHESON: And in fact inspect each stage of the assembly process?
MR. KENNEDY: That is correct, until the joint is made.
MR. ACHESON: Now, Thiokol being responsible to Lockheed, Mr. Barsh, perhaps you can tell me, does Lockheed then have its inspectors inspect that work, or do they accept the Thiokol inspection?
MR. BARSH: They accept the Thiokol inspectors, and there are certain places in the procedures where a NASA inspector has also got to buy
that step.
CHAIRMAN ROGERS: Dr. Ride.
DR. RIDE: Have you had a chance to go back and talk to the technicians who were watching the mating process, watching the tang go into the clevis, to find out whether they have-actually were able to witness any anomalies?
MR. LANG: No, we have not.
[1110] DR. RIDE: How many people do you have stationed around the circumference during that process?
MR. LANG: I think we have a requirement that says four minimum. Normally there are six or seven. I think in this case, Carver, you said six?
MR. KENNEDY: I believe that there would be at least six technicians alone. There would be an inspector. There would be probably some representative of what we call the LSS, which is an oversight group from Morton Thiokol plant. And a NASA inspector would be there.
DR. RIDE: Do you think that they would have been able to see around that area at either 120 degrees or 300 degrees, to watch the tang as it went in and see whether it affected the O-ring or the inner surface, the sealing surface?
MR. KENNEDY: I would have to say that, the
emphasis being on not reaching a flat on flat condition on the outer leg, the concentration on watching the outer leg is focused there. And that is where most of the attention is.
However, since the inner leg is higher than the outer leg, there is a step in the procedure that is, as you come down you must center the tang, so that it does not-so that it is not eccentric to the clevis. That is, you look-it is implicit that you look to see that you're not going to hit the tang on the end of the inner clevis leg.
And if you are, in fact there are instructions in the procedure of how to direct the crane to move the segment so that it is centered over the clevis. But insofar as being a step in the procedure to look for flat on flat on the inner leg and the clevis, there is no step in the procedure.
The requirement it places on us is on the outer leg.
DR. RIDE: Is that area around 300 degrees, is that an easy area to see? Is that something that, if they were looking there, they probably would have noticed that?
MR. KENNEDY: You can walk completely around the clevis. There is nothing else there except the
platform. You can walk around the platform 360 degrees.
DR. RIDE: Do you have plans to go back and talk to those technicians?
MR. LANG: Yes, we do.
DR. FEYNMAN: Technicians have a certain amount of trouble with a particular fact that the procedure is unsatisfactory. If you have decided that six diameters are correct, you still don't know that it will fit.
Let me suppose, for example, that the bottom is an exact circle, so the six diameters of the bottom of the clevis are all equal, that then the tang could be a figure something like this, which has six diameters, if you figure it out, all equal; and yet, it wouldn't fit in because it is more triangular.
And they find this from time to time, and have given it a name and tried to fix it. But there is no fix-up up and down between the guys who are doing the work and the people who have written the procedure to discover that they have this difficulty from time to time.
MR. LANG: Well, again, that case, I guess, that you drew could be true some time, some day. But the criteria, again, is no flat on flat, and that is
carefully inspected all the way around.
MR. KENNEDY: That is a 360 degree requirement, not just at six locations.
[1111] VICE CHAIRMAN ARMSTRONG: When you have a difficult fit, do you forward that information back to the plant so that on their post-recovery inspection they can correlate their inspections to those pre-flight difficulties, stacking difficulties?
MR. KENNEDY: The information is forwarded to the plant. I do not know what use is made of that, but the information is forwarded. In fact, as I said, they have a group here called LSS, a representative group here, and they actually prepare that report. They prepare a summary of each stack and furnish that information back to the plant in Utah.
And it covers other things other than this dimension. It covers all of the activities on the SRB's.
CHAIRMAN ROGERS: Mr. Lang, does that complete your presentation?
MR. LANG: I had a couple of more points to go over.
CHAIRMAN ROGERS: Why don't you go ahead.
MR. LANG: We, as discussed, wound up meeting the criteria of no flat on flat, and-well, let me
back up a minute.
While we were in the process of those final dimensions, we proceeded with the O-ring installation. I wanted to mention particularly the O-ring installation and the O-ring inspection criteria because it has become obvious to most people we do not inspect the O-rings per se here at Kennedy Space Center.
At one time in the program, we did. That inspection took place prior to installation up on the platform, up on the level. And subsequently in the greasing process, it was a long-term-not a long-term, but a handling process.
It was decided that the inspections would take place back in the Thiokol plant. The O-rings would be greased there, packaged very carefully, and all we would do here at Kennedy Space Center is inspect the bag for any damage and then remove it from the bag and install it.
CHAIRMAN ROGERS: When was that change made?
MR. LANG: That was-I don't have the date for this. I promised I would do that and I didn't. That change was made for the STS-13 stack, and it has been that way ever since.
CHAIRMAN ROGERS: So you've had a lot of launches since then.
MR. LANG: Yes, we have.
April 1984, I'm told. In this case, in this particular segment, the first two O-rings that we took out of the box to install we rejected because the inner bag had some problem with it. I'm not sure-we have been trying to find the exact nature of that. We think it was merely the bag was not thermally sealed; it was folded over and stapled.
But at any rate, they were rejected. We got two new O-rings out. The bags were very carefully inspected and opened and installed.
We then went through the process of then lowering, after the rounding tool was removed, of course, then lowering the tang into the clevis. That operation was completed, the pins installed, and of course the pin retainer clips, as they're called, were installed, and we performed the joint leak check.
Now, that leak check procedure is an initial pressurization to 200 psi between-in the cavity between the two O-rings, and followed by-and that 200 psi, by the way, is maintained throughout a 15-minute time period, constantly maintained such that if there was a leak of any kind we wouldn't see it, because we maintain the source.
After the 15-minute wait, we vent that
pressure down, let everything settle for a 15 minute stabilization period, come back up to 50 psi, isolate the source at this time, and do a 50 psi pressure check for ten minutes. And our criteria is we are allowed one psi decay from the 50 in ten minutes, and this particular joint, it didn't indicate any leakage at all.
Now, that leak check procedure, by the way, has had some modifications, too. We started off initially in the program through STS-7 just pressurizing it to 50 and doing a leak check. For STS-8 and 9, the procedure was modified to pressurize to 100 psi for-just momentarily come up to 100, and I think some number of seconds-was it eight seconds?
And we were told that the rationale there was to make sure that if there was a problem with the O-ring the flow, the GN-2 flow would go through that O-ring, through the putty, so that when we turn around and pressurize to 50 to do the leak check putty would not mask an O-ring leak.
STS-11, that procedure was again modified, to increase that initial pressure to 200 and to maintain it for 15 minutes, like we did on this joint. And again, the primary reason was to make sure that the putty did not mask a faulty O-ring.
That is the history of the O-ring leak check procedure. And I was just going to say, that is all we had planned.
CHAIRMAN ROGERS: Mr. Lang, I gather you've been head of the SRB joint group investigating?
MR. LANG: Yes, I have been, that's right.
CHAIRMAN ROGERS: And how large is that group?
MR. LANG: About 12 people, roughly.
CHAIRMAN ROGERS: Can you give us some idea when that-when your investigation will be completed?
MR. LANG: Actually, we are fairly close, I think, to completing our portion of it. We directly looked at this one joint to make sure, to go find out everything we did to it, to uncover anything at all that may indicate that there was something in our process that may have been a problem.
CHAIRMAN ROGERS: Do you plan to have a written report?
MR. LANG: Yes, sir, and I would hope to have that report completed in maybe a couple of weeks.
CHAIRMAN ROGERS: As long as that, two weeks?
MR. LANG: A week.
[Laughter.]
CHAIRMAN ROGERS: Good. We're making
progress.
[Laughter.]
CHAIRMAN ROGERS: Well, the reason I ask, of course, this Commission only lasts another 90 days, and we're anxious to move ahead as quickly as possible. And we would like as soon as you can to get a copy of your report, because we want to work with you, and we're going to have to rely in large measure on the work that you do.
[1113] We will be free, of course, to make suggestions or to make any further investigation. But by and large, we appreciate the speed with which you are operating, and we hope that you will give us the report as soon as you can.
Thank you very much for your presentation. We will take a ten minute recess now, please.
(Recess.)
CHAIRMAN ROGERS: The Commission will come to order, please.
DR. KEEL: Mr. Moser, Mr. Littles, and Mr. Lee.
(Witnesses sworn.)
[Please note that some of the titles to the references listed below
do not appear in the original text. Titles are included to identify
and clarify the linked references- Chris Gamble, html editor]
[1114] [Ref. 3/7-19] Aft Segment Pre-Stack Processing Flow Rotating,
Processing & Surge Facility.
[1115] [Ref. 3/7-20] Center & Forward Segment Processing Flow in Rotating, Processing & Surge Facility (RPSF).
[1116] [Ref. 3/7-21] Vehicle Assemby Building - Aft Segment Assembly Installation on MLP. [Ref. 3/7-22] Vehicle Assemby Building - Aft Segment to Aft Center Segment Stack.
[1117] [Ref. 3/7-23 1 of 3] KSC Shuttle Operations- SRB processing. [Ref. 3/7-23 2 of 3] KSC Shuttle Operations- SRB processing.
[1118] [Ref. 3/7-23 3 of 3] Summary of KSC Segment Assembly Operations- RH Aft (SRM-25) to RH Aft Center (SRM-26).
[1119] [Ref. 3/7-24] SRM Assembly.
[1120] [Ref. 3/7-25] + [Ref. 3/7-26] SRM Assembly.
[1121] [Ref. 3/7-27] SRB Circumference Alignment Tool.
[1122] [Ref. 3/7-28] SRM Assembly.
[1123] [Ref. 3/7-29] SRM Assembly.