Replacement of the Green Bank Telescope azimuth track

Replacement of the Green Bank Telescope azimuth track Robert Anderson*, Arthur Symmes, Dennis Egan National Radio Astronomy Observatory (NRAO) P.O. Box 2, Green Bank, WV USA 24944 ABSTRACT The azimuth track of the Green Bank Telescope did not perform as designed. Relative movement of components was noted during construction; in addition, fretting of the base plate and wear plate faying surfaces, fatigue cracking of the wear plates, fatigue failure of wear plate fasteners, and deterioration of the cementitous grout layer occurred at a rapid pace during the first few years of operation. After extensive failure analysis, a new system of components was designed and fabricated, and installation of the components was performed during 2007 (Symmes, Anderson, and Egan, “Improving the service life of the 100m Green Bank Telescope azimuth track”, SPIE 7012-121). The highlights and lessons learned during the fabrication and installation phases are described herein. This information will benefit any organization performing a similar replacement, and may be helpful in new installations as well. Keywords: Green Bank Telescope, Azimuth track replacement 1.0 Fabrication 1.1 Base Plates A contract was awarded by NRAO to Continental Field Systems (CFS), Savannah, GA, to provide the 9-inch deep base plates and weld them together in a manner similar to that developed by CFS for the fabrication of the one piece Large Millimeter Telescope track. The design specified that the base plates be fabricated from a material that has 50,000 psi yield strength. ASTM A572 Grade 50 material was chosen for its commercial availability. As the plates were specified to be nine inches thick, the specification was modified for gauge. The 105 foot radius track consists of 48 base plate sections. Fifty such sections were ordered by CFS from Mittal Steel’s Coatesville, PA facility. Each mill slab was piece tested for chemistry, yield strength, tensile strength, and Charpy v-notch energy. Yield strengths ranged from 57,400 psi to 61,500 psi. After rolling, two base plate sections were obtained from each mill slab. All test results were reviewed and the plates themselves were inspected by representatives from NRAO and CFS prior to shipment. In order to save costs, the inside and outside radial surfaces were flame cut. Some pieces shifted during the flame cutting process when large sections were dropped. This resulted in at least one plate being too narrow, and it was replaced. Variation in the width of the plates also had to be compensated for later during CFS’s machining and welding fit-up processes. Close attention was also given to plate flatness. As the plates only had a one-half inch allowance on thickness for machining, each piece was checked to verify that a flat, nine inch thick section could be obtained. Machining of the base plate sections began upon arrival in CFS’s Savannah, GA facility. Cutting of the bolt holes, top and bottom surfaces, plate ends, and weld preps were all done on a new Johnford five-axis machine bought for this contract. As expected, there were some commissioning problems associated with this new five-axis machine that had to be overcome. It is difficult to check dimensions of pieces this large to the required tolerance. Overall length was being checked by use of a jig that had been manufactured on the same machine as the production parts. For quality assurance of this method, NRAO and CFS agreed to verify a machined base plate and the inspection jig with a coordinate measurement machine (CMM) mounted on a CNC machine (similar to CFS’s Johnford mill) at a nearby company. This gave independent verification of CFS’s quality control jig. The measurements made with the CMM were used to adjust overall part length to tolerance. *[email protected]; USA telephone (304) 456-2205; fax (304) 456-2170. www.nrao.edu Ground-based and Airborne Telescopes II, edited by Larry M. Stepp, Roberto Gilmozzi, Proc. of SPIE Vol. 7012, 701237, (2008) · 0277-786X/08/$18 · doi: 10.1117/12.784807 Proc. of SPIE Vol. 7012 701237-1 2008 SPIE Digital Library -- Subscriber Archive Copy Following machining, the sections were fit-up in pairs for welding in the shop on a mock-up of the GBT track foundation. Alignment was checked by optical surveying, dial indicator measurements, and check templates. The contract with CFS specified that they develop a qualified welding procedure to use both in the shop and in the field. A welding procedure was developed, utilizing a 10 degree included angle groove. This groove was designed to provide a minimum depth of three and a half inches of weld metal. The first few passes would be welded by shielded metal arc process, and the body of the weld would be completed by submerged arc welding. This welding procedure provided satisfactory tensile and hardness results. /;( Figure 1. Fit-up of two base plates for welding in CFS’s facility. A field test of the welding process was performed in 2003 on the existing GBT track. The weld was made with the pieces flat and restrained. The new track was to be welded in the same fashion. The first production weld resulted in an unacceptable upward draw, or gull-wing. Subsequent welds were made with the pieces lifted along the weld (cambered) to compensate for this draw. After several trials, a camber of one and a quarter inches was found to give an acceptable flatness. Any shrinkage of the top surface adjacent to the weld was restored by weld build-up prior to final machining. Welding shrinkage also caused distortion in the horizontal plane, either opening or closing the chord of the arc composed of the two pieces. The optical measurement processes used in the shop were too coarse to determine the exact amount of this distortion. The order of magnitude of the distortion was indicated when a total station instrument was used to check the pieces. Some measures were implemented to correct this problem, but the effect of the remaining distortion was not realized until fit-up of the two-piece sections in the field. The root of each weld was drilled out by following a pilot hole machined into the weld prep. After drilling, the hole was inspected using a bore scope to verify root removal, and then the welds were ultrasonically examined. All of the shop welds passed inspection with no defects. A tie plate was welded on the bottom of the section sides, across the unwelded portion of the plate to stabilize the joint and prevent loading of the weld from the back side during handling and shipping. Proc. of SPIE Vol. 7012 701237-2 Following the weld inspection, the welded sections were moved to a planer mill to cut the finished surfaces on the plate tops and the bottoms. This planing ensured, as much as possible, full contact between the base plates top surfaces with the wear plates and on the bottom surfaces with the supporting splice plates. The bottoms of the plates were then coated with a primer compatible with the epoxy grout, and the sides were coated with a zinc-rich primer. The tops of the plates received an oil preservative to protect them during shipping and staging at the site. n _ L - __ Figure 2. The weld groove and the pilot hole used to drill out the weld root. The rectangular holes at the bottom of the plate are for access to the nuts on the bottom of the threaded fasteners, which fasten the wear plates onto the base plates. 1.2 Wear Plates A contract was awarded to Gadsden Tool Company (GTC), Gadsden, AL, to provide the wear plates. The plate material was specified to be AISI 4340, manufactured in accordance with AMS 6359G. The plate material was heat treated to give a Brinell Hardness Number between 360 and 390. The thickness was specified to be three and a half inches. GTC subcontracted with Oliver Steel Plate Corporation to provide the material in the heat treated form, perform rough machining, and surface grinding. Oliver obtained the steel from LeTourneau Steel Group, Longview, TX. GTC performed the final machining and inspections. The actual hardness of the plates ranged from 40 to 42 Rockwell C, which is comparable to a Brinell Hardness of 375 to 385. As the plates were to be installed over the welded base plates, the thickness of the plates was tightly controlled and the step between plates could not exceed 0.005 inch. The flatness of the plates after grinding did not exceed 0.005 inch anywhere on their length and steps between plates as manufactured ranged from 0.000 to 0.002 inch. The ambient design temperature range of the GBT is minus 30o to plus 90o F. NRAO’s failure analysis of the original track found that the original wear plates were relatively brittle, as inferred from Charpy Impact tests. Charpy values for the original AISI 4140 plates were a very poor 0 to 2 foot-pounds at 32o F. Charpy values for the new AISI 4340 plates were specified to be a minimum of 10 foot-pounds at 0 degrees F, and test values ranged from 10 to 16 footpounds. Each end of the wear plates was designed to be 0.010 inch undersized to provide a 0.020” clearance in the fit of one plate to the next. This was extremely helpful during the field installation. Proc. of SPIE Vol. 7012 701237-3 1.3 Threaded Connectors and Nuts A contract was awarded to FEMCO Machine Company, Punxsutawney, PA for the connectors and nuts. The threaded connectors were specified to be made from AISI 4340 material, heat treated to give a minimum yield strength of 125,000 psi; the actual yield strength was 140,000 to 145,000 psi. The thread quality and finish of the threaded connectors was excellent. The end of each connector was given a 32 rms finish to enable the use of an ultrasonic test device to measure elongation during tensioning. The heavy hex nuts were specified as 2 inch diameter, grade 2H, to be obtained commercially. The thread quality was poor: there was a significant amount of tool chatter, and some nuts had surface cracks at the first thread on the faces. About 20% of the nuts had variations in thread pitch, to the extent that they would not thread onto the connectors, and had to be replaced. These defects are likely the result of dull tooling, pushed too hard during the threading operation. We traced the manufacture of the nuts to an overseas vendor. While smaller heavy hex nuts, up to 1-1/2 inch nominal size, can be manufactured in the United States, we could not find a domestic manufacturer with the capabilities to manufacture 2 inch nuts. Many suppliers told us that quality varies among different nationalities, with most nuts of this size being manufactured in either South Korea or China. Close attention should be given to the specification of nuts, provisions for receipt inspection, and rejection of poor quality. 2.0 DEMOLITION AND INSTALLATION 2.1 Pre-outage activities NRAO staff began physical preparations for the project in the summer of 2006. A system of survey monuments was installed between the pintle and the track foundation. This system included four original construction monuments, located on the foundation ribs adjacent to the pintle structure and a series of new monuments extending outward to the track. These monuments were then used to measure the locations of all of the track jack bolts. The locations of the bolts varied as much as one inch from their design positions. The data were converted to local coordinate systems for each of the 48 base plates and specified to CFS to drill the bolt holes in the base plate lugs. The effort taken to make these measurements, while painstaking, was very beneficial to the placement of the base plates during the outage. The survey monuments were surveyed for stability several times over the intervening year. Temporary electrical boards were constructed to supply power for the welding operations as well as demolition and construction; and areas at the site were laid out for office and tool trailers, fabrication areas, and lay down areas. A set of tent frames were fabricated to shelter work activities as well and the field contractor also fabricated a set of these for their use. Contracts were awarded to General Dynamics SATCOM Technologies (GDST) for the field demolition and erection and to Continental Field Systems to perform the field welding of the base plate sections. A Project Schedule was developed integrating the activities of these two contractors and NRAO activities. The baseline duration for demolition and installation project was 88 days. We specified in all of our procurements that all components be on site 30 days prior to start of the outage. The outage was planned to start April 30, 2007. All suppliers were successful in meeting this date, except for the last truck load of base plates which arrived April 5th. This was still acceptable and a relief to the project team. The wear plates were placed in indoor storage. The threaded connectors and nuts were stored in shipping containers. These components fared well, even though they were in uncontrolled temperature or humidity environments. The base plates were stored outdoors on 6 inch thick timbers. Condensation developed under the plastic wrap on all of the sections. Light rust formed and had to be removed by wire brush prior to placement. These sections were placed in a circle around the telescope, so that they could be picked up and placed in location without a crane move. Lifting beams specifically designed and fabricated for both the new and old base plate sections and the wear plates were also employed. Proc. of SPIE Vol. 7012 701237-4 2.2 Demolition Demolition began on April 30, 2007 in four symmetric arcs around the track. Because the telescope was not removed from the track and foundation, work was performed no closer than 20 feet from the telescope wheels; our measurements determined that the track was in a flat and level position (within 0.005 inch) beyond this distance. Demolition activities were completed relatively easily and ahead of the baseline schedule. Combined staffing by the contractors and NRAO persons averaged around 40 persons during the first two weeks of demolition. The original base plates had been welded to the splice plates, and these welds were removed by arc-gouging. The splice plates were then refurbished, removing a grout anchoring structure form the bottom of the plate and re-milling the surface flat in preparation for reinstallation. Seven new splice plates were purchased prior to the outage and installed during the first phase to prevent a schedule slip. The base plates were easily removed. The original jack bolts were to be cleaned and reused, so care was taken not to break them or damage the threads. The original cementitous dry-packed grout had disintegrated into almost rubble under some sections, but larger intact sections were broken up by jack hammers and removed. NRAO had replaced the original grout in some sections with a high strength epoxy grout and it was found intact and more difficult to break up. The reinforced concrete foundation was in very good shape. A few shrinkage cracks were found, and cores were taken to verify their nature. Compression tests of these cores exceeded the original design minimum strength of 4000 psi. After removal of all of the grout and other materials, the foundation top surface was scarified and then pressure washed. 2.3 Foundation preparation Figure 3. Placement of the Structural Concrete in a splice plate pocket. Berkley Specialty Products, as a subcontractor to GDST, supplied all of the grout products, much of the tooling and equipment used during grout installation, and provided technical oversight of grout-related activities. Five Star Structural Concrete was used to build up the top of the foundation to a smooth surface where needed. After the components were installed, Five Star HP Epoxy Grout was used to connect the components to the foundation. Proc. of SPIE Vol. 7012 701237-5 The original splice plate design included a grout anchor on the bottom. This piece was cruciform in shape and extended into a pocket in the top of the foundation. The cruciforms would impede the flow of epoxy grout, resulting in voids under the plate, and, with the all welded construction, weren’t necessary, so they were removed. The foundation pockets were also chipped out, and the structural concrete product was used to fill the resulting void. The remaining space under the splice plates, about two and a half inches in height, was filled with the epoxy grout. To maximize productivity, all of the structural concrete in the splice plate pockets in an arc was placed as one evolution. This product was set within two hours, so clean up of the jacking studs and form stripping could begin quickly, followed by installation and leveling of splice plates. The concrete surface under the base plates was in very good condition, and after removal of the old grout, was pressure washed. An area of the foundation shown on the right side of Figure 3 shows this surface, which was roughed under the plates during the original construction. 2.4 Base plate alignment, leveling, and welding The 27-1/2 foot, 12 ton base plate sections were easily handled and placed with the lifting beam developed by CFS and NRAO. Three such sections were set together in each arc, along with an old base plate, modified on one end to fit the new geometry. This modified transition section was necessary to allow the telescope to be moved from the old track onto the new track sections for Phase 2 of the demolition and installation. The section ends were jacked into contact by use of temporarily attached nuts and studs. Rough leveling and x-y positioning was accomplished through the use of laser trackers and final level was determined by optical levels. GDST rented laser trackers manufactured by FARO Technologies, Inc., and subcontracted operator/technicians to perform the measuring. GDST then reviewed the data and adjusted the locations of components as needed. Laser trackers can provide excellent data repeatability and precision given an indoor, relatively stable temperature environment over a few meters. Obtaining the same level of performance outdoors, with varying solar heating, humidity, construction dust, vibration, ground stability, diurnal temperature ranges, and surveying over 100 meter distances is another matter which required a good bit of learning to accomplish the accuracy desired for our track replacement. After a few frustrating attempts, rough alignments and levels were made during the day, and verified in the late evenings and after midnight. The temperature drop during many nights caused unacceptable discrepancies between measurements made in late evening and before sunrise. This effect was particularly noticeable in the welded base plate arcs prior to grouting. We therefore limited measurements to the late evening period. These differences were much smaller after grout was installed, allowing heat conduction through the foundation. That said, the laser trackers were vital for fitting the welded sections into a best fit radial arc, and finally closing the circle. We recommend to anyone using laser trackers for a project of this magnitude that they perform identical measurements at the construction site prior to the start of production activities, under identical conditions, using the same equipment and staff that will perform the production work. Each jack bolt was straightened into its previously measured position in the global x-y coordinate frame. The splice plates in each arc were then positioned and leveled, followed in sequence by the base plates and wear plates. The centerline of the track is on a 1260 inch radius from the pintle, with a specified tolerance of 0.1 inch. The welding distortion, first noted in fabrication, had an immediate impact on installation of the first base plates. The distortion of the first section was such that the center of the arc was in spec, but the ends were out 0.5 inch. A second laser tracker and operator were hired and each base plate section was measured. NRAO staff determined a best fit of the geometry, calculated the amount of error on each end, and CFS machined the error off of each end in the field. An additional amount was removed to provide for further growth during each field weld. Each base plate section was machined by a unique amount, varying from 0.005 inch to 0.020 inch. This effort resulted in a schedule slip of several days. We recommend that parts be measured during fabrication with the same equipment, in our case, laser trackers, that will be used to install them in the field. Proc. of SPIE Vol. 7012 701237-6 The first field weld also resulted in unacceptable distortion. In this case, distortion was not limited to the ends of the arc, but the top surface of the plates twisted from inside radius to outside radius causing a 0.060 inch difference. This weld was cut apart and the ends re-machined. The cause was traced to the initial alignment of the plates for welding. The amount of gap in the root of the weld was not uniform and caused the distortion in the arc of the base plates. The field welds also required cambering to minimize distortion in final flatness and the optical method used to set the weld preps level was not adequate. It was determined that the bars placed under the weld joint to set the proper camber needed to be aligned perfectly along the centerline of the weld gap. This was true even if the gap was deliberately opened slightly on one end in order to compensate for distortion caused by other weld joints. After conferring with both contractors, the alignment of the weld preps was assigned to GDST and their contract was revised accordingly. 'ii W7. Figure 4. Drilling of weld root by CFS employee. This pre-welding alignment was costly in terms of schedule, with alignment taking four to eight hours. Alignment also required a large crew, involving a laser tracker technician, a crew of three or four to adjust the plates, and the GDST field supervisor to review the data and direct changes. An NRAO engineer also reviewed the data. GDST learned how the weld prep gap affected the arc of the plates and they became very adept at adjusting these gaps, not only to minimize distortions, but also to correct the shape of the arc. We recommend that anyone planning a welded track construction discuss methods with GDST. Preheating and welding operations went smoothly and followed the planned activity durations very closely. All welds were inspected by ultrasonic procedures. CFS had to remove indications from only one weld out of all 48 shop and field welds and the repair time was minimal. The roots of each weld were removed by drilling a ½ inch hole, following the pilot holes machined into each weld. The resulting holes were inspected using a boroscope with any remaining indications removed by a second, 5/8 inch diameter drill (Figure 4). Following welding, the base plate’s overall flatness was adjusted and the welds were ground flat by use of a heavy grinder mounted on a radial arm. This method resulted in each joint being within the specified 0.005 inch flatness. Proc. of SPIE Vol. 7012 701237-7 2.5 Wear plate installation and bolt tensioning When all base plates in an arc were placed, the wear plates were installed and properly spaced. A 0.020 inch gap had been designed between each wear plate that was achieved by laser tracker measurements and moving the plates inward and outward. Following the initial fitting, the plates were removed and a corrosion inhibitor, Cortec VpCl-378, was sprayed onto the faying surfaces of the base plates and wear plates. A 0.015 inch thick, full width, Teflon/bronze/molybdenum anti-fretting layer was then placed between the plates. The wear plates were reinstalled followed by the installation of the threaded connectors and nuts. Hydraulic tensioning equipment was used to apply 90,000 pounds of tension in each connector. Our original plan was to tension four connectors at a time; however, this resulted in too much variation in tension, which had to be corrected with repeated tensioning. We attribute this variation to flexure of the long lengths of hydraulic hose and variations in the friction of each tensioning head. We reduced the number of connectors being tensioned to two at a time. This change reduced the amount of rework. Tension was measured ultrasonically with an instrument made by Dakota Ultrasonics. Pretension and post-tensioning readings were taken, and the elongation was calculated immediately by the hand held unit. The elongations were temperature compensated by the use of an optional RTD. The elongation data is stored, and can be checked during telescope operation for any relaxation in the connectors. Figure 5. Pre-outage trial of hydraulic stud tensioning equipment. 2.6 Transition between Old and New Track Sections The new track wear plates were 3.5 inches thick and 27 inches wide with a v-shaped, or “chevron” end design. The old wear plates were 2.25 inches thick and 19 inches wide and their ends had a 45 degree angle. In order to maintain the same track top surface elevation, the new base plates were installed 1.25 inches lower and the thickness of the grout layer was correspondingly reduced. The new base plates also differed from the old design. The new plates have an approximately 22 degree angle on the ends, while the old sections had a 45 degree angle. With the purpose of fitting the two sections together so that the telescope could be driven from the old track to the new one, we developed four transition sections. These transitions consisted of four old style base plates with one end modified, two wedge-shaped pieces fabricated to bridge gaps Proc. of SPIE Vol. 7012 701237-8 between the base plates, and four specially designed wear plates with the old and new geometries on the ends. Grout was installed under the base plates to transfer the load to the foundation, but plastic sheet was used to prevent the grout from adhering to the base plates or foundation. After the telescope was driven onto the new track sections, the transition sections were removed and scrapped. Figure 6. Transition section. The new base plate and splice plate are on the right; the old base plate and wear plate are on the left. The wedge shaped shim is at bottom center. A shim plate is positioned between the old base plate and splice plate. 2.7 Grouting After final alignment checks, the high strength epoxy grout was installed. All of the splice plates in an arc were grouted first and then the base plates were grouted. GDST became very proficient at this activity, making process improvements that resulted in completion in less time than allocated in the schedule. Forming for the splice plate pockets was performed on one day, the splice plate pockets were grouted the next morning, the base plates were formed that afternoon, and then the base plates grouted the following day. The grout was allowed to cure for at least 24 hours before being loaded. Test samples (cylinders) were made from each batch; subsequent lab testing indicated that compressive strengths for the grout ranged from 15,780 to 17,240 psi (109 to 119 MPa). Five Star Products (and other suppliers as well) strongly recommends that their epoxy grout components be 80o F. or more when mixed and placed to assure good flow and curing. NRAO and GDST were prepared to provide heat during the curing phase and the materials were stored in a controlled environment prior to mixing. Grouting activities were scheduled for the warmer periods of the day. What was not anticipated was that ambient temperatures in the mid to high 90 degree range caused the product to begin curing faster which, as a consequence, slows down the flow of the product when being placed. This degraded flow rate on hot days significantly slowed the overall grouting process at times; degraded flow ability also resulted in two sections of base plates having voids at the ends of the pours. These void regions required additional pours to achieve satisfactory grout volume. Unlike concrete, the epoxy grout is almost impossible to remove from mixers, chutes, and buckets after curing. All tools and equipment used in grouting activities such as this must be cleaned immediately and thoroughly after each use to prevent build-ups and loss of performance. Proc. of SPIE Vol. 7012 701237-9 '4 Figure 7. Placement of grout below a splice plate. The concrete bucket is being positioned to deposit grout into the head box at the center of the picture, and two others are ready to the left side. 2.8 Phase 2 Installation Following a 24 hour cure period for the last phase 1 grout pours, the telescope was rotated onto the new track sections on June 29 which allowed phase 2 demolition and installation activities on the remaining four track sections to proceed incorporating a number of process improvements that recovered much of the schedule slip. An area of questionable concrete was encountered during shortly after phase 2 demolition began and a volume of about 2 cubic yards of the concrete foundation was removed and replaced. The proposed phase 1 and 2 tie-in welds were reviewed and work activities were re-sequenced so that an additional four of the eight welds originally planned to be flat welds could be performed by cambering the welds in the same manner as all of the other welds. The last four welds, completing the circle, were performed with the pieces flat. This resulted in a significant bending of the joined pieces, as much as 0.060 inches above the specified top surface elevation, several feet from the weld joint. The top surfaces of the base plates on either side of the welds, out to about 12 inches circumferentially from the weld center lines, shrank in elevation because of the weld pull. In addition, the contractor’s approach to limiting this distortion was to increase the weld preheat. This resulted in higher than anticipated temperatures in the steel and the grout. The overheating caused the bond between the grout and plates to break. Some areas of the grout were crushed through the excessive expansion of the heated steel. After extensive measurements with the laser tracker systems, a corrective action plan was implemented, consisting of building up the low areas over and adjacent to the welds, and machining the high areas so that the top surface blended with the rest of the track. Measuring and machining were performed on night shifts to avoid daytime distortions. In addition, repairs were made to the epoxy grout underneath the final four welds by removal of damaged grout and repouring with new grout as well as injection of epoxy into the areas where the initial bond had been broken. Altogether, this corrective work resulted in a three week schedule slip, primarily because of the required field machining. We recommend that anyone undertaking a track replacement such as ours sequence work activities so that all welds can be cambered. This may mean that grout be temporarily installed and removed prior to welding, and then be reinstalled afterwards. New track installations can be cambered and welded fully prior to grout placement. Proc. of SPIE Vol. 7012 701237-10 3.0 THE FINISHED TRACK The GBT azimuth track machining and welding activities finished on August 31 and the grout was allowed to cure until September 3, 2007. The total duration of the track outage was therefore 95 days. GDST remained on site for an additional ten days to dress the grout and demobilize. NRAO employees applied a waterproof coating, Miracote 7011 (Crossfield Products Corporation) to the concrete as well as a rustproof sealer to the studs and nuts ( Loctite 51213). The sides of the base plates and splice plates were caulked and painted with a structural paint. Figure 8. The finished track. A track characterization period of several weeks began on September 3rd and the telescope resumed high frequency observations on October 1st. Overall, the telescope pointing is very satisfactory and has been relatively stable. Figure 9 is a graph showing the local tilt of the elevation axis as a function of the azimuth angle of the telescope, as measured by a pair of gas-damped capacitive-readout inclinometers mounted on the elevation axle bearing housings. The measurement was performed upon completion of the track refurbishment (black curve) and was repeated three months later (red curve). The overall pattern is consistent and the changes are quite small (subarcsecond). Similar small changes were seen with the old track and are believed to be caused by water pressure variations in the underlying rock strata, which is composed of shale. The summer of 2007 was very dry (which benefited the track project), but precipitation in subsequent months has caused the water table to rise. We will continue to monitor track performance and investigate causes further. Proc. of SPIE Vol. 7012 701237-11 U— I — P P C1U — C) - O.5*(X 1 -fX2) [orcsec] (or5r.) —. r Figure 9. The local tilt of the elevation axis as a function of the azimuth angle of the telescope. The measurement was performed upon completion of the track refurbishment (black curve) and was repeated three months later (red curve). The overall pattern is consistent and the changes are quite small (subarcsecond). (Courtesy of T.R. Hunter and K.T. Constantikes – NRAO). NRAO is very satisfied with the performance of the track to date and hopes that it will continue for many years. We are very appreciative of the effort put forth by many NRAO employees as well as by the contractors and suppliers involved in the track project. Without their input and dedication to the project, it would not have been accomplished so successfully. Proc. of SPIE Vol. 7012 701237-12 View publication stats