A 2.371Mb PDF of this article as it appeared in the magazine—complete with images—is available by clicking HERE
United Launch Alliance (ULA) operates its Atlas 5 Launch Vehicle program at Cape Canaveral Air Force Station, Florida and Vandenberg Air Force Base (VAFB), California. At VAFB, the vehicle is launched from Space Launch Complex 3 (SLC3) near the City of Lompoc and a few miles inland of the Pacific Ocean. The bright white Mobile Service Tower (MST) that encloses the Atlas 5 as it is prepared for launch is visible from all over the City of Lompoc and its surrounding hills and agricultural fields. The location provides an excellent opportunity to safely view a launch for many people in the small community.
The Atlas 5 operating procedures require that a technician from the Lockheed Martin Precision Alignment Lab (PAL) in Denver check certain components of the launch vehicle (a.k.a. "rocket") alignment during assembly on the launch pad. Due to circumstances that prevented PAL from performing the alignment survey, ULA had to call in a private surveyor at the last minute. Penfield & Smith is well known on VAFB and had previously provided surveying services to ULA. In addition, the Professional Land Surveyor (PLS) in responsible charge of this survey provided construction surveying services to the contractor in charge of the renovation of SLC3 in 2004, a project to retrofit the complex for the Atlas 5 program. The PLS was involved with site survey control, including the monuments set by the Contractor to mark the center of launch.
The first of the three survey components is the booster stage vertical alignment to verify the first stage of the launch vehicle is sitting plumb when installed on the launch heads. Second is the rotational alignment of the Centaur stage with respect to the booster stage. The Centaur stage contains the guidance system and directional rocket engines it uses to maneuver the payload into orbit. The stages share the same vertical axis, but have different horizontal coordinate systems. Third is the rotational alignment of the "Centaur Forward Load Reactor" (CFLR) ring with respect to the "Boat-tail" mounted at the bottom of the Centaur stage. The Payload Fairing, which houses and protects the Centaur and payload, is attached to the launch vehicle at the Boat-tail and CFLR, which must line up precisely.
Once the tasks were identified and the tolerances confirmed by ULA engineers, it was decided that Penfield & Smith would perform the survey for the booster vertical alignment and the Centaur rotational alignment. The tolerances for the CFLR alignment were beyond the limits of conventional equipment available to P&S, so ULA brought in a consultant with specialized equipment and software for the task.
Field surveying for flight AV-025 began in early May 2010 with the establishment of a site survey control to be used to make measurements of the booster and centaur bodies. Unlike the PAL approach, which utilized close range scanning instruments and independent, localized coordinate systems, P&S designed a survey method to create an inclusive coordinate system which capitalized on the strengths of our conventional survey instruments and our measurement analysis and adjustment software.
Determining the actual center of launch was necessary for determining the vertical alignment of the booster body. It also was interesting to compare the actual center of launch to the one defined by the existing site survey control network. After discussing the survey requirements with PAL personnel from Denver, it became clear that the physical location of the three launch heads (retractable arms upon which the rocket rests in launch position) located in the Fixed Launch Pad (FLP) defined the center of launch. The actual center of launch was found to be slightly more than ¼ inch from the center of launch defined by the survey control monuments, almost exactly the difference the ULA engineer informed us about the day before.
With site control and the center of launch defined, we returned to SLC3 with survey equipment and a laptop computer for immediate processing of survey observation data. We arrived as the AV-025 booster body was being raised off its transport trailer and several hours before the booster was ready to be lowered onto the launch heads. While waiting, we checked our instrument, set up the laptop and spent a few hours ensuring we could work remotely. The PLS went to the Level 8 platform of the MST to get a close look at the machined survey targets created by the PAL for specific use in an Atlas 5 vertical alignment survey. Seven targets were placed around half of the circumference of the booster, facing out of the MST. Our survey instrument was set up on a centerline of launch monument, which provided a perfectly clear line of sight to make multiple observations to each target. The data was downloaded, adjusted and imported into AutoCAD for a graphical analysis to determine the center of the booster stage at the forward end. Seven different circles were circumscribed using seven different target point combinations. Rejecting two outlier solutions, the average resulting center fell within the 0.625 inch tolerance limit (compared to actual center of launch 100 feet below) for the alignment of the booster stage. The survey successfully confirmed that the alignment of the AV-025 booster stage was within the acceptable tolerance limit.
Ten days later, Penfield & Smith returned to SLC3 to perform the Centaur survey. While the exact purpose was still not entirely understood, we decided that once we were able to have physical contact with the Centaur, inspect it, and discuss the requirements with a few ULA people, we would be able to deduce the specific objective of the survey. We checked our instrument and reported to Level 14, approximately 140 feet above the pad deck, to assess how to proceed with transferring survey control to the platform. We also requested that the plastic covering be removed from the top of the Centaur so we could determine the specific feature to be surveyed. A phone conversation with PAL personnel confirmed we were to survey the payload interface ring, a 12-foot diameter ring with a series of bolt holes to which the payload is fastened. The center of the circle would define the center of the Centaur body and the direction to axis marks on the circle would define the Centaur body coordinate system. We set two control stations on the platform that had a good view of the payload ring and direct line of sight to three control stations on the ground. On the ground, our survey instrument was set up and we made multiple observations to the platform control points. The instrument was then transferred to the platform, and from each platform station we measured back to the ground stations and to multiple bolt holes around the circumference of the payload ring.
Two of the holes straddled an axis marker so we could determine the Centaur body orientation, and we counted the total number of holes to determine the nominal angular spacing between each, noting the position of axis markers and quad labels on the Centaur body. The relationship between the Y axis of the Booster Body coordinate system and the Y axis the Centaur body coordinate system was calculated to be slightly different than the nominal value of 135 degrees. The survey successfully confirmed the rotational alignment of the Centaur Body.
In December 2010, ULA requested that P&S perform the same alignment checks on flight AV-027. We anticipated and achieved a much more efficient work flow since we already had survey control set up on site. Physical features of the Atlas 5 for flight AV-027 differed in a few ways that res
ulted in a slight twist to the survey procedures.
First, the booster as fitted with a single Solid Rocket Booster (SRB), which required a vertical alignment survey before and after installation of the SRB. The initial check was to confirm the booster body was leaning at the predetermined amount. After installation, we confirmed that the weight of the SRB pulled the booster back into plumb within a specified tolerance. Second, the bolt pattern on the payload interface ring was not the same because of the specific payload for the flight and the fact that the payload fairing was smaller than the AV-025 payload fairing. Lessons learned and procedures developed in the AV-025 surveys allowed the AV-027 survey to proceed very smoothly and without problems.
The precision of measurements can be greatly enhanced by eliminating or compensating for systematic errors by following procedures to identify and eliminate gross errors, and by making redundant measurements to minimize the effects of random errors. Techniques to minimize error of measurement are included in every surveyor’s "toolbox", and been developed over centuries of practical application. The idea is the same with any type of equipment, from steel tape, to total station, to GPS receivers.
Penfield & Smith surveyors routinely check and adjust equipment, design network surveys with redundancy, and follow specific field procedures for consistent results. The AV-025 and AV-027 surveys required extra care since we were tasked with determining alignments within tolerance ranges considerably smaller than the requirements for our typical work. For example, we needed to determine that the forward end (top) of the booster was within 0.625 inches (0.05 feet) of being vertical, which is about the amount of error that we might accept on a typical topographic survey feature point. We needed to achieve error ellipses less than 0.01 feet at the 95% confidence level to be sure that the top of the booster would be within the tolerance. An error ellipse describes the area within which the determined position should fall with a confidence level of 95%, i.e. there is a 5% chance that it does not fall within the error ellipse. We methodically set up a system of survey control on the ground with the specific intention of achieving excellent results.
When we surveyed the location of the booster targets on AV-025, the 95% error ellipses were nearly circular, with a diameter of 0.007 feet, or about 0.084 inches. Our method allowed us to report with confidence that the booster center was within the 0.625 inch tolerance.
The rotational alignment survey required even greater care, and was more difficult because we had to transfer survey control to Level 14 of the MST, about 140 feet above the pad deck. Careful planning and measurements helped us determine the position of control points on the platform with nearly circular error ellipses of 0.003 feet. In turn, these precise control points allowed us to obtain nearly circular error ellipses on the payload ring bolt holes with a diameter of 1/16 inch and allowed us to report with confidence that resulting rotational relationship between the booster stage and the Centaur stage was within the nominal range.
Our equipment maintenance and field procedures virtually eliminate systematic and gross errors, leaving only random errors to detect and distribute throughout the survey network. We utilize a least squares adjustment software that allows weighting of each individual observation, provides a detailed statistical report on how a set of data relates within itself, and clearly highlights any measurements that are suspect. When the data set is "cleaned up", the software further reports on the reliability of the remaining measurements, positions of points, and the overall integrity of the network. The results of these surveys would not have been possible, or provable, without the application of the redundant measurement technique, or the use of a least squares adjustment software program.
Penfield & Smith performed alignment surveys for flight AV-025 in April and May 2010, and then again for flight AV-027 in December 2010 and January 2011. We responded to the initial April 2010 request the same day, and maintained this level of responsiveness throughout the AV-025 and AV-027 surveys. ULA clearly communicated their need for us to complete surveys quickly to avoid any deviation in their launch preparation schedule. According to the chief engineer in charge of launch preparations, vehicle assembly could not proceed until we confirmed that the booster stage was plumb within specified limits. To meet this need, we worked several 12 hour days in a row on the AV-025 survey to make certain that solid survey control was in place when the launch vehicle was lowered onto the launch heads and ready for the first vertical alignment check.
In addition to long days, P&S was flexible and arrived on site at the times and on the dates requested to work around the demanding assembly and preparation schedule. In December, we arrived early in the morning for the AV-027 vertical alignment survey to find that the booster had arrived late and was still on its transport trailer. It was a beautiful clear morning, but with a thick marine layer hanging just at the edge of the ocean a few miles away. By the time the booster was in place and ready for measurements, the fog had rolled in and significantly impaired line of sight. Knowing the results were critical, we informed our clients of the situation and arranged to stay until the fog cleared and allowed a clear view and solid measurements.
For both the AV-025 and AV-027 surveys, we came prepared for onsite data download, processing and analysis to provide immediate results on the vertical alignment check. Our results were ready about 45 minutes after measurements were completed. ULA engineers and crews waited nervously to see if the launch vehicle needed to be lifted for launch head adjustments, or bolted down to enable the next step of launch preparation. Our positive results met with sighs of relief, broad smiles, and many expressions of thanks in the launch crew ready room.
Justin Height is a Principal Surveyor and Regional Survey Manager for Penfield & Smith Engineers in California, where he has been employed since 1986. He graduated from California State University Fresno with a degree in Surveying & Photogrammetry, and is licensed in California and Nevada.
A 2.371Mb PDF of this article as it appeared in the magazine—complete with images—is available by clicking HERE