Fire and Reconstruction at Lobato Bridge in

FREDERICK R. RUTZ, TODD M. RILEY, AND PETER FOSTER

The oldest bridge in New Mexico was Introduction resulted in the D&RG abandoning the goal of building south to El Paso and devastated by fire, leaving an operat- Fire broke out at the Lobato Bridge, a focusing instead on building west to the ing railroad with difficult choices for vital feature of the Cumbres and Toltec mining camps of Leadville, , Scenic Railroad (CTSRR), a National balancing its reconstruction with and the of south - Historic Landmark, on the night of historic preservation. west Colorado. 4 June 23, 2010. 1 By morning it was clear In 1879 reconnaissance surveying that the bridge had been severely dam - was conducted, and contracts were let aged and that operations of historic for the construction of the San Juan excursion trains would be interrupted, Extension. By April 1880 service was for how long no one knew. 2 The cause extended from Alamosa, Colorado, to was not clear — the railroad had oper - Antonito, Colorado, and by February 1, ated over 120 years without this kind of 1881, service was extended from An - disaster. tonito over Cumbres Pass to Chama, New Mexico. Building the section of The Railroad, Past and Present line between Antonito and Chama was a The Lobato Bridge, originally known as formidable task that included construct - the Wolf Creek Trestle, is located on the ing two large bridges and two tunnels, San Juan Extension of the and dealing with grades exceeding four Rio Grande Railroad (D&RG). percent, maneuvering across the sides of Gen. William Jackson Palmer founded steep canyons, and crossing Cumbres the Denver and Rio Grande Railway Pass at an elevation of more than Company in 1870. The original plan for 10,000 feet. The line was completed to the railroad was to build a line from Durango, Colorado, by July 1881 and Denver, Colorado, to El Paso, Texas, a to Silverton, Colorado, in July 1882. 5 distance of 870 miles. A narrow-gauge The two large bridges of the Cum - system with a 3-foot width was selected, bres and Toltec Scenic Railroad are the based on reduced cost compared with Cascade Bridge, with a span of more larger gauges and more favorable opera - than 400 feet at a height of more than tions in mountainous terrain. 3 The 3- 130 feet, and the Lobato Bridge, with a foot narrow gauge was more nimble, span of more than 300 feet at a height allowing trains to execute tighter turns of more than 100 feet. Originally, both and climb steeper grades; it could also bridges were constructed of wood. accommodate smaller locomotives and Noted nineteenth-century engineer C. rolling stock. Shaler Smith was retained by the D&RG The advent of the San Juan Extension to design both bridges. The earliest date is based on several important occur - of Smith’s design on surviving drawings rences: the Brunot Treaty and subse - for the Lobato Bridge is November quent treaties and agreements with the 1880, although records indicate the Fig. 1. K-36 Class engine #488 crossing Lobato crossing was built in wrought iron in Bridge with passenger train. View from below Ute Indians, which led to opening of the 1883. 6 prior to the fire. Photographer and date un - San Juan region to mining and agricul - known. From Richard L. Dorman Collection, ture; and the railroad wars between The San Juan Extension between Catalog No. RD012-122, of Historic Narrow D&RG and the Atchison, Topeka and Alamosa and Silverton continued in Gauge Railroad Photographs located at the Santa Fe Railway that culminated in the operation into the mid-1960s. In 1967 Friends of the Cumbres and Toltec Scenic Treaty of Boston, which restricted the the Denver and Rio Grande Western Railroad Library in Albuquerque, N.M., and Railroad (D&RGW), successor to the Historic American Engineering Record, HAER D&RG development in New Mexico. No. NM-16, Lobato Bridge, Wolf Creek. Both of these occurrences ultimately D&RG, applied to the Interstate Com -

37 38 APT BULLETIN: JOURNAL OF PRESERVATION TECHNOLOGY / 45:1, 2014

Some drawings of the original 1881 design have been preserved and were made available. These proved valuable in evaluating the post-fire condition and reconstruction decisions.

The Fire Nearly all of the timber ties were con - sumed in the fire, which created great heat at the top of the structure. The rails buckled. The upper flanges of the girders became distorted, buckled, and cracked in numerous places. The upper parts of the girders were raised to higher temperatures than their lower Fig. 2. Bridge arrangement, Lobato Bridge. Sketch by J. R. Harris & Co., Denver. parts, evidenced by discoloration of the paint on the heated surfaces. It is be - lieved that the lower portions of the merce Commission to cease operation feet above the ground. The built-up girders, which were not distressed, and the supporting bents below, sustained on the railroad, except for the Silverton plate girders, which were 49 1/2 inches Branch that runs from Durango to deep, were fabricated from metal plate no permanent damage. However, the Silverton. Operations ceased, and the with continuous angle connections upper portions of the plate girders, the line was abandoned from Antonito to riveted between the web and flange X-bracing, and the top-flange lateral Durango on December 29, 1969. 7 plates. Cover plates were used in the bracing all sustained significant distress On July 1, 1970, the states of Col - mid-span portion of each girder, and X- (Figs. 3 and 4). orado and New Mexico purchased the bracing served as cross bracing between At the north abutment, the bearing line and its associated right-of-way, the girders, located 10 feet apart. Addi - separated, and the girder was observed rolling stock, and real property from the tionally, lateral bracing of steel angles to be about 7 inches above the masonry D&RGW. Currently the line between provided support of the top compression plate. The top of that girder as it cooled Antonito and Chama is operated as the flange. The ties served a two purposes: after the fire contracted sufficiently to Cumbres and Toltec Scenic Railroad to support rails, as with all railroads, develop a negative camber, raising it off based on the Cumbres and Toltec Scenic and, at 15 inches deep, to span the 10 of its bearing. That meant the girder was Railroad Compact that was authorized feet between the parallel girders. Post- now cantilevered from its interior sup - by an act of Congress in 1974 and is still fire examination revealed that the rails port, 54 feet away, significantly stressing its continuity plate in tension. jointly owned by the states of Colorado were offset 2 1/2 inches from the center - and New Mexico. 8 The Cumbres and line between the girders, resulting in The east girder of each span suffered Toltec Scenic Railroad Commission slightly greater loading applied to one more damage than the west girder, manages the railroad. The commission is girder. suggesting that prevailing winds from comprised of four commissioners, two The iron bents, also referred to as the west may have concentrated more from each state. The governors of each trestles, columns, or piers, were fabri - heat on the east side. Spans 1 through 4 state appoint the commissioners. cated from built-up columns of rolled suffered irreparable damage; span 5 channels, connected with lacing bars suffered damage to the top lateral brac - ing but was considered repairable; and The Bridge and batten plates, with two such columns cross-connected with transverse struts span 6 had no apparent damage. At Lobato Bridge was fabricated by the spaced vertically every 30 feet. The bents number 2 and 3, the bolts that Andrew Carnegie’s prolific Keystone bents are not cross-connected to each connect the girders to the bearings were Bridge Company, headquartered in other; rather, the girders of the bridge either sheared off or distorted. Pittsburgh. The D&RG installed the provide all the longitudinal support for Having lost the service of Lobato bridge over Wolf Creek a few miles the structure. At each interior bent and Bridge, CTSRR was forced to temporar - north of Chama, as part of their line at each column, the two adjacent girders ily cease operations from Cumbres Pass across the mountains of northern New use one bearing assembly, which rests on to Chama. Service from Antonito to Mexico and southern Colorado (Figs. 1 a pin. The pin allows for the rotation Cumbres Pass was continued. As the and 2). effects caused by the deflections of the bridge is located only three miles north The six spans — riveted plate girders, girders. The girders have continuity of the Chama terminal, buses were used five measuring 54 feet and one measur - plates at their top flanges that connect to transport passengers from the termi - ing 40 feet totaling 310 feet in length — each span to the next. Thus, the anchor nal at Chama to Cumbres Pass. Still, were supported on stone-masonry abut - bearing at the south abutment offers the publicity affected tourist numbers; ments and iron bents as much as 100 only longitudinal restraint. business and revenues decreased at a FIRE AND RESTORATION AT LOBATO BRIDGE 39

as other elements were found in very small to trace quantities, but the table does report on the elements discussed in this paper. Physical properties were also deter - mined by tensile testing (Table 2). As shown in Table 2, samples 1 and 2 were taken from portions relatively unaf - fected by heat from the fire. Samples 3 and 4, taken from webs of girders, pre- sumably were moderately exposed to heat. Samples 5 and 6, from upper flanges, were greatly affected by heat. Elongation, an indicator of ductility, was measured across gauge marks 2 inches apart on the tensile specimens. Because the bridge was so seriously damaged and not in service, no mea - Fig. 3. Fire damage at a girder. Distortion at the Fig. 4. Fire damage. The riveted plate girder is sures were taken to compensate for the upper portions of the girder flange can be seen. distorted, ties are burned, and track displaced. Paint has been burned off the upper flange and Photograph by Roger Hogan for CTSRR. holes created where samples had been has been discolored at the upper part of the removed. web. Photograph by Roger Hogan for CTSRR. Bridge Analysis and Rating time when significant extra capital was Colorado and New Mexico; the historic needed, as insurance did not fully cover nature of the railroad itself; and limited A rating analysis was performed to the necessary bridge repairs. finances. While this paper primarily determine the structural capacity of the addresses technical issues, these adminis - spans and bents, using the as-built plans, field measurements and section proper - Studies trative concerns weighed heavily on the management. ties, and a material yield strength of Fy An initial inspection and assessment Thorough visual inspection by spe - = 29,400 psi for wrought-iron mem - resulted in a recommendation for total cialists in engineering railroad bridges bers. The rating analysis was performed replacement of Lobato Bridge. How - was initiated. Material samples were in accordance with Chapter 15, Steel ever, the railroad called in specialists to obtained, initially from the web of span Structures; Part 7, Existing Bridges, of review the conditions and appraise the 6, where the heat from the fire was the 2010 edition of the AREMA Man - damage, in part because of the historic presumed to be minimal, and from a ual for Railway Engineering. nature of the entire railroad. CTSRR lacing bar almost certainly unaffected by Information provided by CTSRR posed the following questions: heat. Later specimens came from areas indicated that two trains per day, one in 1. Could the damaged metal be repaired susceptible to heat, with those samples each direction, pass across the bridge at in place? from the flanges of spans 1 and 2, di - a maximum speed of 8 mph. The train consists of a K36 or K37 coal-fired 2. If not, could replacement be limited rectly below the burned ties, which were steam-engine locomotive; a tender; and to damaged members only? clearly affected by heat. Chemical tests revealed that the 1883 eight viewing cars with a maximum 3. If total replacement were necessary, capacity of 44 passengers per car. Rating should the historic appearance be metal had a relatively low carbon con - tent but relatively high silicon content, calculations for normal and maximum retained, or should a modern railroad loads under pre-fire, as-built conditions bridge be constructed? typical of wrought iron, which was com- monly used in this period for railroad were completed. The normal ratings are 4. If repaired in place, could the re - bridges. 11 The results shown in Table 1 for loads that can be carried by the paired structure meet modern struc - are not a complete metallurgical listing, structure for its expected service life at a tural requirements of American Rail - way Engineering and Maintenance- of-Way Association (AREMA)? 9 Table 1. Chemical properties from original metal Further, could it meet the require - Sample No. Location Carbon (%) Silicon (%) Phosphorus (%) ments of the Federal Railroad Ad - 1 Web, span 6 0.02 0.17 0.19 ministration (FRA)? 10 2 Lacing bar, pier 6 0.01 0.22 0.11 3 Web, span 1 <0.01 0.06 0.20 The choices had to be weighed 4 Web, span 2 <0.01 0.21 0.21 against the very real concerns regarding 5 Flange, span 1 0.01 0.22 0.14 reestablishment of service in a timely 6 Flange, span 2 0.01 0.28 0.18 fashion; responsibilities to the joint The list is not complete as other elements were found in very small to trace quantities, but these are owners of the railroad, the states of the constituents discussed in the paper. 40 APT BULLETIN: JOURNAL OF PRESERVATION TECHNOLOGY / 45:1, 2014

deposits oriented in the direction of tural strength. Later structural analysis rolling. While the ferrous component of revealed that the bridge was originally wrought iron has a crystalline structure designed for narrow-gauge rolling stock similar to that of steel, it is the slag that that weighed less than the 1926 engines gives wrought iron its fibrous appear - and tenders currently in use. Thus a ance (Fig. 5). 16 Slag may be present in need for strengthening the superstruc - concentrations of 1 to 3 percent by ture was identified. weight or more. 17 A typical chemical While such strengthening might be analysis of historic wrought iron with a done along with repairs in place, later slag content of 3 percent by weight was physical analyses eventually eliminated Fig. 5. Specimen, lacing bar after testing in 18 tension failure. Photograph by the author. found to contain 0.15 percent silicon. this approach as an option. Samples 3 While the slag percentage was not and 4 and particularly samples 5 and 6, specifically determined from the Lobato taken from heat-affected regions, re - standard speed. The intent of a normal samples, the silicon contents were, and vealed higher yield strength than sam - rating is to limit the stresses in the struc - they compare well to the findings from ples 1 and 2, but they also indicated ture to those that it would have typically typical historic wrought iron (Table 1). drastically reduced ductility, which can been designed for and to determine the For the Lobato samples, the carbon, be seen in Table 2 in the columns la - Cooper’s equivalent load that it could silicon, and phosphorus contents all beled “Elongation” and “ F /F .” (For carry on a daily basis while providing a u y indicate wrought iron. comparison, the ratio F /F using mini - consistent factor of safety. 12 Maximum u y Similarly, from physical testing at the mum values for ASTM A36, a ductile ratings are for loads that can be carried University of Colorado Denver, a fibrous structural steel, is 1.61.) These tests at infrequent intervals. The maximum appearance is readily apparent along the suggested that the metallurgical proper - rating provides a reduced factor of fracture surfaces (Fig. 5). The dark spots ties of material subject to the fire’s heat safety and, if more frequent maximum are slag inclusions. Both the fibrous had been affected, resulting in an in - loads are allowed, a reduced structure appearance on fracture surfaces and slag crease in strength but a decrease in life. Results of the rating analysis deter - inclusions are characteristic features of ductility. It was the reduced ductility mined that the bridge in its pre-fire historic wrought iron. The conclusion that eliminated further consideration for condition was overstressed from 20 to from the chemical tests, physical tests, repairs in place. Modern structural 30 percent when loaded by the current and visible appearance of post-test design methodologies, including the trains at the bridge spans and at two of fracture surfaces is that the material at AREMA standard for railroad-bridge the tallest bridge bents. Lobato Bridge is wrought iron. design, are based on the use of ductile metals because of the serious conse - Material Design Decisions quences of brittle failures. The loss of Bridges of the early 1880s were typi - ductility in the primary structural mem - The initial physical test results from cally fabricated from wrought iron, as bers rendered them unsuitable for repair. samples 1 and 2 suggested that repair in bridge engineers were reluctant to spec - Thus, the decision was made to re- place might be possible (Table 2). This ify the new material, steel, in their place the superstructure (i.e., the girders approach would have had the advan - designs. Most railroad bridges built and associated bracing that were af - tages of probable lowest cost and least prior to 1890 are of wrought iron, fected by heat) but to allow the sub - removal of historic fabric; the disadvan - while steel was phased in over a transi - structure (the support bents that were tages would have included a probable tion period from about 1890 to about not affected by heat) to remain in place. maximum engineering effort to design 1894, after which steel was typically Several minor repairs, as well as many specific details and probable used. 13 strength ening of the bent towers based inability to improve the overall struc - Wrought iron manufactured by the on condition assessments and structural puddling process that replaced earlier production by bloom smelting or fining Table 2. Physical properties from original metal of pig iron tended to produce ferrous Sample Location Exposure to Yield Ultimate Elongation at Fu/F y metal with a carbon content to con- No. Heat Strength (psi) Strength (psi) gauge marks (%) centrations less than 0.1 percent by 1 Web, span 6 low 29,400 45,100 40 1.53 weight. 14 Phosphorus was sometimes 2 Lacing bar, pier 6 nil 29,400 45,100 40 1.53 added to improve strength, although, 3 Web, span 1 moderate 37,000 49,600 18 1.34 except for wrought iron of very high 4 Web, span 2 moderate 34,600 49,700 18 1.43 purity, it would decrease ductility. 15 5 Flange, span 1 high 32,500 37,300 4 1.15 Slag is a characteristic feature of 6 Flange, span 2 high 37,900 40,300 3.5 1.06 wrought iron manufactured by the Samples 1 and 2 were taken from portions relatively unaffected by heat from the fire. Samples 3 and nineteenth-century puddling process, 4, from webs of girders, presumably were moderately exposed to heat. Samples 5 and 6, from upper flanges, were greatly affected by heat. Elongation, an indicator of ductility, was measured across gauge unlike modern wrought iron. Slag marks 2 inches apart on the tensile specimens. Fy = yield strength, Fu = ultimate strength, psi = pounds formed in small (mostly microscopic, per square inch. The ratio Fu/F y is an indication of ductility, with a high ratio indicative of ductile but sometimes visible), often thread-like material and a low ratio indicative of brittle material. FIRE AND RESTORATION AT LOBATO BRIDGE 41

Fig. 6. Lobato Bridge, detail from the design drawings of the end of a Fig. 7. A temporary gantry was used by the contractor for removing girder girder. Efforts were made to replicate the historic appearance as much as segments and installing new girder segments. The gantry rolls on a tempo - possible within the confines of the project’s design criteria. Compare the rary beam that also served to provide longitudinal bracing for the structure connections using tension controlled bolts to the original arrangement of during construction. Photograph by Roger Hogan for CTSRR. rivets shown in Figure 3. Drawing by HDR. analyses, were also made to the bents. of all six spans were to be removed and material of former ties, which were not Because structural wrought iron is no replaced with new plate girders of struc - original. The ties here, as at all rail - longer manufactured and because mod - tural steel, connected with TCBs. Re - roads, had been replaced as part of ern bridge designers are familiar with pairs were made to the supporting bents, ongoing maintenance. They were of structural carbon steel, Grade 50 (mini - and some miscellaneous repairs were Douglas fir and creosote-coated, similar mum yield strength of 50,000 psi) struc - made at base plates on the original stone to the ties lost to the fire. During the tural carbon steel was the final choice piers and footings. investigation stage, it was learned that for the girders and their associated the original track had been installed elements. In the analytical modeling, Construction slightly off the bridge centerline; this each member material — either iron or fault was corrected during refurbish - steel — was assigned its appropriate While the investigations, planning, and ment, with the new track centered on modulus of elasticity and yield strength. design took the better part of a year, the bridge. The original 1883 structure had been construction was able to be completed in approximately one month in the connected with rivets (Fig. 4). While the Conclusion eventual decision was to utilize high- spring of 2011. A fundamental tech - strength bolts, the railroad management nique designed by the contractor was a Laboratory testing of specimens taken gave consideration to using rivets to gantry supported on temporary beams from both the heat-affected part of the help retain the historic appearance. 19 A that were connected to the original sup- girders and from portions not affected search was undertaken to find fabrica - port bents (Fig. 7). The gantry could be by heat revealed that the heated tors with capabilities in rivets, and six rolled over all spans and was used to lift wrought iron no longer possessed sig - such fabricators were located. Unfortu - and deliver the damaged girder segments nificant ductility and thus was unsuit - nately, none possessed American Insti - to one end, where they were lifted by a able for an active railroad bridge. This tute of Steel Construction (AISC) certifi - crane for transport off site. Likewise, the discovery led to the decision to remove cation for bridges, while fabricators of new girder segments were transported and replace the girders with new gird - structural steel did. Use of tension- across the bridge to their desired loca - ers. As structural wrought iron is no controlled bolts (TCBs) was proposed in tions via the gantry. The temporary longer produced, structural steel be - a layout to match the historic rivet beams also provided lateral support for came the logical choice. Making con - patterns. While not rivets, TCBs do have the overall structure in the longitudinal nections with bolts instead of rivets was a rounded head on one side, suggestive direction while the girders were not in a more difficult decision. Rivets are of rivets from a distance. The heads place. The girders were fabricated off available, as are all the necessary ham - were placed on the viewable side of the site and assembled into segments of mers and forges. The difficulty was in girders: compare the design drawing of pairs of girders with associated bracing. finding qualified riveters and, in particu - the girder ends per Figure 6 with the The segments were transported to the lar, a fabricator that possessed AISC rivet patterns in Figure 3. bridge site for installation. New ties and certification for such bridges using The structural design was prepared in new rails were installed. The new ties rivets and the establishment of satisfac - accordance with AREMA Manual for replicated the dimensions (nominal 8 tory inspection criteria. If rivets are to Railway Engineering . The plate girders inches wide by 15 inches deep) and be used on a significant project such as 42 APT BULLETIN: JOURNAL OF PRESERVATION TECHNOLOGY / 45:1, 2014 this one, the issues of certification and 8. Ibid., 8. 15. Gordon and Kompf, 393. inspection will need to be addressed. 9. American Railway Engineering and Mainte - 16. M. O. Withey and James Aston, Johnson’s The management and design team nance-of-Way Association, AREMA Manual Materials of Construction , fifth ed. (London: struggled with weighing preservation for Railway Engineering (Lanham, Md.: John Wiley and Sons, 1919), 599. AREMA, 2010). measures against very real public-safety, 17. James Aston and Edward Story, Wrought 10. Federal Railroad Administration, 40 CFR Iron: Its Manufacture, Characteristics and financial, and schedule constraints. Parts 213 and 237, Bridge Safety Standards, Applications , second ed. (Pittsburgh, Pa.: A. M. Historic features were evaluated to Final Rule, Nov. 22, 2010. Byers Company, 1949), 2-3. determine the appropriate level of inter - 11. Robert Gordon and Robert Knopf, “Evalu - 18. Aston and Story, 42. vention. The final decision was to recon - ation of Wrought Iron for Continued Service in 19. Riveted connections were used in most iron struct the superstructure while the sub - Historic Bridges,” Journal of Materials in Civil or steel structures until about 1960, when bolts structure was largely preserved, al- Engineering 17, no. 4, (July/August 1995): 396. became more economical. Rivets are still used though some rehabilitation took place. 12. “Cooper’s Equivalent Load” is a design live in certain applications such as railroad rolling load using the Cooper E series–loading axle stock, but they have given way to bolts in While the reconstructed girders were configuration adopted by the railroad industry. general, and high-strength bolts in particular, in steel instead of wrought iron, the dis - Each axle is given a prescriptive axle loading. building and bridge structures. tinctive features defining the historic Higher loadings are obtained by multiplying character were largely retained, the only each load by the same constant; wheel spacings do not change. Many different axle configura - visible difference being use of bolts in tions and loadings currently exist, and the The APT Bulletin is published by the place of rivets. The bridge is used as it Cooper E loading represents an industry- Association of Preservation Technol - was historically, supporting loaded adopted mechanism to designate a rating. ogy International, an interdisciplinary organization dedicated to the prac - trains daily. 13. Charles C. Schneider, “The Evolution of the tical applic ation of the principles and Practice of American Bridge Building,” Trans - techniques necessary for the care and wise use of FREDERICK R. RUTZ is a structural engineer actions of the American Society of Civil Engi - the built environment. A subscription to the and principal with J. R. Harris & Company in neers, v. 54 (New York: American Society of Bulletin and free online access to past articles are Denver and an assistant professor at the Uni - Civil Engineers, 1905): 222. member benefits. For more information, visit www.apti.org. versity of Colorado Denver. He served as a 14. Francis E. Griggs, “Restoration of Cast and historic metal-bridge consultant on the project. Wrought Iron Bridges,” Structure 8, no. 7 He can be reached at fred.rutz@jrharrisandco (Sept. 2001): 18. .com.

TODD M. RILEY is a structural engineer with HDR in Jacksonville, Florida, where he is a specialist in the design of railroad bridges. He helpinghelping architectsarchitects & engineersengineers deldeliveriver superiorsuperior dedesignsign ddocumentsocuments

served as the project manager for the project, C i v i

which included the inspection, rating, and final c

hehelpinglping F s s design for the Lobato Trestle. He can be PRESERVEPRESERVE a m

historichistoric e

reached at [email protected]. ( landmarkslandmarks A d

sincesince 19921992 o e e l p

PETER FOSTER is a professional civil engineer h

W

and vice president of Wright Water Engineers, e i n Inc., specializing in water resources, where he m a n , manages the office in Durango, Colorado. He 1 m m 9 has served as a commissioner of the Cumbres 1 3 )

and Toltec Scenic Railroad for three terms, as a t o e e appointed by three Colorado governors. He p

M can be reached at [email protected]. a n h r r a t t a n t t t

Notes M u n i c

1. Jeff Brown, “Scenic Railroad to Repair Fire- i p x x x a l

Damaged Crossing,” Civil Engineering 81, no. B u i l

4 (April 2011): 30. d i n e e g

2. A television news report can be found at /

V i

http://www.krqe.com/dpp/news/local/ d e o

northwest/train-trestle-burning-outside-of s t o o i l

-chama. l :

V e r t t 3. Herbert Danneman, “A Ticket to Ride the t i c a

Colorado Rail Annual l Narrow Gauge,” no. 24, A c

Colorado Railroad Museum, Golden, Colo., c e s (2000). s o o 4. Vernon Glover, Denver & Rio Grande Railroad, San Juan Extension, Wolf Creek g g g Trestle (Cumbres and Toltec Scenic Railroad, g Lobato Trestle), HAER No. NM-16, Historic American Engineering Record 2011: 4-6. e e e 5. Ibid., 6. INSPECTIONNSPECTTIION & TESTINGTTEESTTIING SERVICESSERVICES ININ EXTREMEEEXXTTRREME LOCATIONSLOCCAATTIIONS 6. Ibid. KentKKeent DieboltDiebolt • 607-257-4049607-257-4049 IIthacathaca • NNewew YorkYork CityCity • WWashingtonashington D.C.D.C. 7. Ibid., 7. W W wwww.vertical-access.comww.vertical-access.com