Engineers Officers Quarters Final Report.pdf

FORT JEFFERSON

Engineers’ Quarters Ruin

Dry Tortugas National Park, Florida

August, 2017

Acknowledgments All research, documents, and drawings were produced through a generous donation from the

HTR Foundation in an effort to preserve Civil War era structures at Dry Tortugas National Park.

Contents

ADMINISTRATIVE DATA

GENERAL bACkGROuND

RESOuRCE bACkGROuND

PHYSICAL DESCRIPTION

TREATMENT RECOMMENDATIONS

APPENDICES

APPENDIx A. DRAWINGS

APPENDIx b. SELECT HISTORIC IMAGES

APPENDIx C. MATERIALS ANALYSIS

Engineers’ Quarters ruin 2016, Image courtesy of HAbS, Historic American building Survey

FORT JEFFERSON, DRY TORTuGAS NATIONAL PARk1

Administrative Data

RESOuRCE NAME AND NuMbERS building Name: Engineers’ Quarters

Other Names: EOQ

Commanding Officers’ Quarters

Engineering Officers’ Quarters

Structure No.: HS-08

LCS ID No.: 011921

RESOuRCE LOCATION

The Engineers’ Quarters Ruin is located along the western edge of the parade ground near bastion 3. Fort Jefferson is located on Garden key, within Dry Tortugas National Park (DRTO). The Park is located approximately 67 miles west of key West Florida in the Gulf of Mexico and only accessible by boat or seaplane. The Park encompasses an area of approximately 100 square miles and contains seven small sand and coral keys (islands) and the surrounding shoals and water.

Garden key contains the Park’s central cultural feature, Fort Jefferson. The Fort is occupied by Park staff and is the center of Park operations. The visiting public generally travels to the Park by commercial ferry or seaplane service operated out of key West.

Location: Parade Ground Fort Jefferson, Garden key, Dry Tortugas National Park

Coordinates: 24° 37’ 42” N Latitude 82° 52’ 28” W Longitude

County: Monroe

State: Florida

CuLTuRAL RESOuRCE DATA

The Engineers’’ Quarters Ruin was listed on the National Register of Historic Places in 1976 as a contributing element of the Fort Jefferson National Monument Historic District.

In 1992, Dry Tortugas National Park was established by Public Law 102-525 to “preserve and protect for the education, inspiration and enjoyment of present and future generations nationally significant natural, historic, scenic, marine and scientific values in South Florida.” under [36 CFR 60.1

(b) (1)], historic units of the National Park Service are automatically given National Register of Historic Places status by virtue of their incorporation into the park system.

Dry Tortugas National Park is currently in the process of updating its National Register documentation.

SIGNIFICANCE

The National Register of Historic Places Nomination Form completed in 1976 includes the following Statement of Significance for Fort Jefferson:

During the first half of the nineteenth century the u.S. began a chain of seacost defenses from Maine to Texas. Though not as extensive in area as Fort Monroe in Virginia, Fort Jefferson, seventy miles west of key West was the largest of all Third System Forts in terms of armament for which it was designed, about 450 guns. built to cover a strategic anchorage in the Gulf of Mexico, it was one of only three forts in the South to remain in Federal hands throughout the Civil War. It was never fully armed or even completed, though construction proceeded well beyond the war and into the 1870s.

During the post-war years the fort became a prison, housing among others some of the alleged Lincoln assassination conspirators, the most famous being Dr. Samuel Mudd. After serving nearly four years and having helped save many prisoners from yellow fever, Dr. Mudd received a full Presidential pardon. The Fort was virtually deserted in 1874, and it was not until 1900 that it was transferred to the Navy Department. It was later made available to the Dept.

of Agriculture who used it as a bird sanctuary until if finally became a national monument in 1935.

ENGINEERS’ QuARTERS RuIN 2

Tortugas Harbor

Engineers’ Quarters Ruin

Fort Jefferson

FORT JEFFERSON, DRY TORTuGAS NATIONAL PARk3

The rich history of the area of Fort Jefferson goes back as far as 1513 when it was discovered by Ponce de Leon and named the Tortugas Islands, Tortugas being the Spanish word for turtle. Since the eight low-lying islands had no fresh water, sailors called them Dry Tortugas.

Pirates operated in and around the Dry Tortugas preying upon ships of all flags. Not until 1821 when Florida became part of the united States were the pirates successfully driven out.

The importance of the Dry Tortugas to military men grew as westward expansion of the country brought more and more commerce to Gulf Coast cities. To prevent enemy seizure of the islands and to insure the control of the Gulf of Mexico shipping, a decision was made to erect a Fort somewhere on the islands Fort Jefferson served this purpose by being the key to control of the Gulf of Mexico because of its strategic location. Commerce from the Mississippi Valley to the Atlantic coast passed through this vital area. The Fort was begun on Garden key in 1846and was later named Fort Jefferson in honor of the former u.S.

President.

During the 30 years of construction, Fort Jefferson was a quiet place with the Civil War the only event to bring major changes in building or life.

Federal troops hurriedly occupied the half-completed, unarmed fortress in January 1864. The development of the rifled cannon rendered Fort Jefferson obsolete long before construction was stopped.

Hurricanes and yellow fever, coupled with the obsolescence of the fortification led to the abandonment of the fort by the Army in 1874 with the work still underway. During the 1880s and 90s, the Navy looked upon Fort Jefferson as a recoaling station for its Caribbean fleet. From Tortugas Harbor, the battleship Maine weighed anchor for Cuba, where she was blown up in Havana.

One of the first naval wireless stations was built at the Fort early in the 20th century and, during World War I, the area was equipped as a seaplane base. As the military moved out again, fires, storms and salvagers took their toll, leaving the “Gibralter of the Gulf” the vast ruin it is today.

The Nomination discusses the Engineers’ Quarters as a three building complex. The narrative focuses on those portions that were initially rehabilitated and are currently being used for housing.

Section 12 of the National Register nomination establishes the Level of Significance of Fort Jefferson as National.

The nomination does not specify a Period of Significance, however a “Record of Decision” issued in March 2004 established the Period of Significance for Fort Jefferson and its associated resources as spanning from 1846 to 1876.

RELATED STuDIES byrd, beth W. and Susan L. Hitchcock, Garden key, Dry Tortugas National Park, Cultural Landscape Report, 2011.

bears, Edwin C., Fort Jefferson National Monument, Historic Structure Report, Historical Data Section, 1983.

Lord Aeck Sargent Architects. Fort Jefferson National Monument, Historic Structure Report Amendment, 2004.

Manucy, Albert, A Constructional History of Fort Jefferson , 1846 - 1864, National Park Service, 1981.

Morrison, George T., Phillips, John Wesley and Revised by Richard Rasp. National Register of Historic Places Inventory- Nomination Form for Fort Jefferson National Monument, February, 1976.

ENGINEERS’ QuARTERS RuIN 4

DRTO 301522 - Engineers’ Quarters ruin, June 1966.

FORT JEFFERSON, DRY TORTuGAS NATIONAL PARk5

General background

From the start of construction on Fort Jefferson in late 1840s there were challenges securing adequately durable and strong brick. When construction appropriations were approved in 1844, most Florida brick manufacturers were still making brick using hand-presses. Due to unreliable local transportation and the slower pace of local brick manufacturing, additional bricks for the Fort were shipped from manufacturers as far away as New York and Maine. However, northern bricks were not suited for the tropical marine environment of the Fort and decayed quickly.

The outbreak of the Mexican-American War in 1846 delayed any substantial progress on the Fort, so it was not until 1850 that the Army began to find sources for the massive quantity of bricks the fort’s construction would require. In December of 1850, Captain Wright, the Fort’s superintending engineer, visited several existing Gulf Coast forts to observe the condition of their bricks and ascertain where those bricks were made.

Wright identified brick made in the Pensacola area as preferable for its ability to withstand the salt and high humidity of the Tortugas, but brick manufacturers and shipping companies in the area were still lacking in their ability to produce adequate bricks and ship them to the construction site. Maine bricks were a close second, having already withstood several years in place in piers on Garden key.

Failures by Congress to approve appropriations for the Fort led to more delays in securing and transporting building materials. So, it was not until

1854, that Phillip H. Raiford and General Anderson Abercrombie, influential Alabama businessmen, plantation-owners, and politicians, proposed that they would establish a brick manufacturing business in the Pensacola area to supply bricks for the construction of Fort Jefferson and Fort Taylor.

The War Department quickly approved, despite Raiford and Abercrombie’s inexperience in brick-making. A contract dated August 24, 1854 shows that Raiford and Abercrombie agreed to supply 3,000,000 bricks of specific dimensions and quality, at $21 per thousand, to each fort.

Raiford and Abercrombie’s inexperience led to continued delays in establishing their plant on Escambia bay in Pensacola. While they set up their facility, they secured bricks from other manufacturers to supply to the forts. In 1855, when Raiford and Abercrombie began producing and supplying its first bricks, they proved to be substandard, breaking and crumbling easily. The company was at risk of losing their contract to provide bricks for the forts. Financial stresses led Phillip Raiford to sell his interest in the company to John E. bacon in 1857. During this time, bricks and other building supplies for the fort were being transported from the northeast, a more expensive, slower, and difficult endeavor – shipwrecks were not uncommon.

A decision by the newly incorporated bacon and Abercrombie to hire master brick-maker, John W. Crary, saved the company from demise. Crary immediately improved the hand manufacturing process to a point that the bricks made by the company now satisfied inspectors at the forts. Shortly thereafter, Crary began inventing and implementing a mechanized process for brickmaking that automated and shortened many manufacturing steps while producing stronger and more durable brick. The reliability of shipping bricks to the forts also improved with the securing of a contract in 1858 with the key West shipping firm, Tift and Company. The supply of high quality bricks to the forts was secured; by 1861, Crary claimed to have supplied over 16,000,000 bricks to the federal government.

In the decade leading up to the Civil War, Florida experienced massive growth in both population and economy. State leaders pushed for expansion of railroads, while landowners expanded their plantations and their slaveholdings – at a time when the national debate over slavery was reaching its peak. During this time, political parties in Florida were in a state of flux; the Whig party had disbanded and two new political parties were challenging a radicalized Democratic Party. Despite their differences, however;

Floridians were mostly united in their dislike of the Republican Party, predominantly due to the Republican platform against slavery in the new territories. Not one vote was cast in Florida on November 7, 1860 for Republican presidential candidate Abraham Lincoln.

Following Lincoln’s election, a convention was held in which delegates voted on whether to immediately secede from the union or to ‘cooperatively’ secede along with the other southern states. Florida elected to secede immediately and passed an Ordinance of Secession on January 10, 1861. Many of the other Confederate states

Challenges with logistics, building materials, and appropriations create significant delays in construction at the fort

ENGINEERS’ QuARTERS RuIN 6 were also scrambling to pass Secession legislation. With secession imminent, Florida state leaders quickly organized state militias, supplying them with arms and ammunition.

Progress was slow, however, as Florida’s limited transportation networks and still sparse population made communication and organization challenging.

Florida’s senators were also preparing for the state’s inevitable secession, using their status in Washington DC to procure information on united States military activity in Florida, specifically regarding the strength of the garrisons at forts located within Florida.1 It was clear to them that one of the first actions following secession would have to be the taking of all Federal positions in Florida. In fact, an arsenal near Chattahoochee as well as Fort Marion in St. Augustine were taken by state militia just days prior to secession.

In early January, 1861, Fort Marion and the arsenal near Chattahoochee were surrendered without conflict as they were both only guarded by a handful of federal soldiers. by mid-January, 1861, combined forces of southern state militias took control of the naval yard in Pensacola and threatened to take nearby Fort Pickens.2 1 Johns, John E., Florida During the Civil War, university of Florida

Press, 1963, p. 23

2 Nulty, William H., Confederate Florida: The Road to Olustee, The university of Alabama Press, 1990, p.

On February 26, 1861, bacon and Abercrombie, halted all production of bricks for the forts – having sided with the secessionists. In March 1862 the Confederate army ordered all Pensacola industries destroyed out of concern that they might be used by federal forces. bacon and Abercrombie’s brickworks was subsequently burned to the ground and never rebuilt.

Meanwhile, the Federal military, fully aware of Florida’s intention to secede, was strategically reinforcing its southern Florida forts, namely Fort Taylor in key West and Fort Jefferson on the Tortugas. both forts were extremely important strategic posts for naval and commercial activities in the Gulf of Mexico, with construction beginning roughly 15 years prior as part of the united States’ Third System of coastal defenses.

However, both forts required significant work to be adequately prepared not only for any possible attack, but also to accommodate the increase in the size of their garrisons.

Captain M. C. Meig, Corps of Engineers, arrived at Fort Jefferson on November 8, 1861 and noted that the fort had “not a single gun”.

No attack ever came, but construction on the forts continued throughout the Civil War and in to the following decade.

Despite the logistical challenges of securing and transporting material, the fiscal shortfalls from insufficient appropriations, and two wars, some progress was made at the Fort.

by the end of the Civil War, the Fort’s walls were nearly complete; there were several sections of barracks and officer’s quarters, kitchens, a hospital, a chapel, and several wood buildings; and, foundations for several more buildings had been laid.

Over the next decade to 1876 when the fort was finally abandoned, the officer’s quarters and barracks, and their associated kitchens, were completed; the large powder magazine had walls and its barrel vaulted arch; the small magazine’s foundations were set; and many of the embrasures, casemates, and barbettes were mounted with guns.

However, in 1889, George Phillips, the Fort’s overseer and employed at the Fort since 1846, noted that most of the buildings at Fort Jefferson were in a state of dilapidation.

“Everything here is going to destruction… The once fine quarters of the Officers and Soldiers are now but a little more than a wreck.” Phillips also noted widespread vandalism by salvagers taking brass hardware, windows, and doors.

Drawer 74, Sheet 80 - Fort Jefferson, Dry Tortugas, Fla, Sketch showing Condition of Works June 30th, 1867, to accompany annual Report.

Engineers’ Quarters

FORT JEFFERSON, DRY TORTuGAS NATIONAL PARk7

Early on in the planning of the fortification’s parade buildings, Captain Wright saw the need for special accommodations for the senior members of the labor force and as well later on, for senior members of the engineering force. Rather than build temporary wood framed accommodations, Wright sought to have several of the Officers’ Quarters brick kitchens erected first, to be used to house the clerk, master mason, master carpenter, and superintending engineer. The kitchens, which were mirrored pairs sharing a party wall, allowed for the opening of doors between the two spaces, creating a large enough footprint for the resident and his family (HSR p.35).

by mid-October of 1847, work had begun on the Officers’ Quarters and three of the detached kitchen buildings (HSR p. 46). Prior to the arrival of a garrison, the first finished spaces of the Officer’s Quarters were used as project offices, a hospital, and several other non-residential temporary uses. Two of the kitchens were complete by the summer of 1848, with a third only lacking doors and windows (HSR p.47). Sometime later, though no date is given, Captain Woodbury had a kitchen built on Front 6 to accommodate the overseer and his family (Manucy 135).

These kitchen accommodations served until the garrison arrived in 1861. The Officers’ Quarters were taken over by the commanding officer, meanwhile the engineer offices and quarters were moved several times as the commanding officer ordered.

Another three kitchens, these located adjacent to frontt 5 and away from the Officers’ Quarters kitchens, were constructed for the physician, clerk, and engineer’s office and residence (Manucy 135, HSR 258).

by the end of the reporting year 1865, two Officers’ Quarters double kitchens were completed and the foundations for four more laid. One of these Officer’s Quarters kitchens was intended to serve the post quartermaster

(HSR 256)

In 1866, a storm knocked over the rear wall of the 3rd story of the first section of the unfinished range of Officers’ Quarters. While this section of wall fell inwards, just a few hours later, the upper story of the south section of the officers’ quarters toppled outwards, crushing the kitchen to its rear and killing Lt. John W. Sterling in his bed (HSR 289). The remainder of that year was spent repairing the storm’s damage.

The dispute between the garrison and the engineers’ employees resurfaces in 1867, when the clerk was again told to relocate his quarters in one of the kitchens (HSR 295).

Captain Simpson, superintending engineer for that period, argued that the principal employees of the engineering department deserved “treatment as a gentleman, and not to be stowed or crammed in anywhere, as under workmen, day laborers, or prisoners.”

Post Commander McConnell responded by evicting the clerk’s family and tossing the family’s furnishings out of the building.

The issue eventually went all the way to the Secretary of War, who decided that the engineers were to have a “proper allowance of quarters” and set aside for exclusive use by the Corps, three of the detached kitchens, the western rooms of the recently finished Officers’ Quarters and the associated kitchen

(HSR 296-297).

Only maintenance work was reported for the kitchen quarters in 1878, 1879, and 1883. A hurricane in 1886 left the kitchen quarters in a “dilapidated” condition (HSR 372).

In 1891, one of the few remaining personnel at the fort, aside from the lighthouse keeper and his family and those working at the quarantine station, was the ordnance-sergeant, who had taken over three kitchens and a room in the officers’ quarters (HSR 382)

During fiscal year 1896, more maintenance work was reported on the kitchens, including installing new roofs and gutters and minor window sash repair. by 1916 it was reported that several thousand bricks had been removed from the kitchens connected to the barracks and Officer’s Quarters and used at the key West Station (HSR 424).

Resource background

ENGINEERS’ QuARTERS RuIN 8

DRTO 301230 , Engineers’ Quarters Ruin Ca. 1937.

FORT JEFFERSON, DRY TORTuGAS NATIONAL PARk9

Timeline

ARMY

April 12, 1861 , Southern forces fire upon Fort Sumter, South Carolina.

April 9th, 1865, General Robert E. Lee surrenders at Appomattox, VA

1861-1862 - Morton has two kitchen buildings constructed along front 5 (present day Front 3) (including the subject structure) for the physician and clerk. Later a third structure was built to the south, in line with the others, to serve as housing for the engineers staff. (CLR , p. 34 and Manucy p. 135)

June 1863 - Plan of Fort showing “condition of the work” as of June

30th 1863, shows first two kitchen buildings constructed along Front 5 (present day Front 3) (including subject structure)

THE CIVIL WAR

1860 - Montgomery C. Meigs arrives at Garden key replacing

Captain Daniel Woodberry as Superintending Engineer in charge of construction of Fort Jefferson (CLR p. 8).

April 1861 - Lieutenant James St.

Clair Morton replaces Meigs as

Superintending Engineer.

December 29, 1845, The united States annexes Texas

February 2, 1848, Treaty of Guadalupe signed

THE MExICAN-

AMERICAN WAR

1874 - Army garrison leaves Fort Jefferson and two of the three kitchen structures are occupied by the Ordnance-Sergeant and Fort keeper (CLR p. 34)

ENGINEERS’ QuARTERS RuIN 10

Au gu st

, 1

8, G ar de n, b ird a nd L og ge rh ea d ke y tr an sf er re d to

M ar in e

H os pi ta l S er vi ce

A pr il 7, 0, M ili ta ry re se rv at io n at

D ry

To rt ug as tr an sf er re d to D ep ar tm en t o f t he N av y

Se pt em be r 1

6, u

S N av y ab an do ns

F or t J eff er so n

NAVY NATIONAL PARk SERVICEMARINE HOSPITAL

SERVICE

ARMY

1992 - Dry Tortugas National Park established.

TIME SCALE COMPRESSED

1916 - Two of the quarters are described as being in “very poor condition.” (CLR

p. 34) by 1937 the three strctures had become “brick shells” with no roofs, windows, doors, porches or interior floors. In 1938-1940 - The two kitchen buildings to the south of ruin (8A and 8b) are rehabilitated for use as housing for the Park Superintendent. (CLR, p. 34)

January 1, 1935 - Fort Jefferson designated a National Monument by President Franklin Delano Roosevelt

W W

I

W W

I I

September, 1939

September, 1945

July, 1914

November, 1918

1887 - It is reported that two sets of the quarters were habitable

(CLR p. 34)

1890 - Ordinance-Sergeant using three of the kitchens as quarters.

(CLR p. 34)

1910 - Structures reported to be “leaking badly.” (CLR p. 34)

1940 - present - Periodic improvements and updates made to rehabilitated quarters while third structure remains in a ruinous state.

FORT JEFFERSON, DRY TORTuGAS NATIONAL PARk11

Physical Description

Though Corps personnel occupied several different spaces throughout the fort’s history, what we now call the Engineers’ Quarters is one of three buildings, initially planned as kitchens, adjacent to Front 3 (formerly Front 5). The two southernmost kitchen buildings, were reconstructed in the 1930s and are currently used for staff or visitor housing.

The Engineers’ Quarters ruin is a two-story, 20-feet by 31-feet, brick masonry structure located in the northwest portion of the fort’s parade ground, within an area roped off from the public. The ruin is divided into two bays by a central chimney. The bays are mirrors of each other; 15-feet wide, a fireplace at the party wall, two windows in the opposite wall, and one window per bay along the longer walls.

Today, the walls of the Engineers’ Quarters ruin rise just above the second floor granite window sills (two windows on the south elevation feature segmental brick arches), with scant bits of plaster still clinging to some of the inside corners.

Remnants of joist pockets, first and second floor granite window sills, and two brick fireplace surrounds remain. A coral and cement foundation is visible within the interrior of the ruin.

The similar adjacent structures were reconstructed in the late1930s as part of a WPA program project.

The reconstructions added porches to the buildings as well as a wood frame hyphen between them.

No improvements were made to the Engineers’ Quarters and the structure has remained in a ruinous state since around the turn of the century.

ENGINEERS’ QuARTERS RuIN 12

Plaster remnant

FLOOR PLAN

Granite lintel/threshold

Two-wythe brick masonry wall

AxONOMETRIC VIEW

31’ - 1”

FORT JEFFERSON, DRY TORTuGAS NATIONAL PARk13

View of Engineers’ Quarters Ruin looking west.

View of interior of Engineers’ Quarters Ruin.

ENGINEERS’ QuARTERS RuIN 14

View of Engineers’ Quarters Ruin looking north.

View of fireplace in east half of Engineers’ Quarters Ruin.

FORT JEFFERSON, DRY TORTuGAS NATIONAL PARk15

Treatment Recommendations

Given the existing condition and character of the Engineers’ Quarters ruin, stabilization will require significant intervention beyond simple repointing of the structure’s masonry. In its current condition, the structure is stable but vulnerable to failure in the extreme conditions of the region and therefore measures must be taken to secure the standing walls to ensure preservation. (See Structural Recommendations by Silman) A second condition contributing to the instability of the structure is the presence of corroding metals, specifically the lintels supporting the masonry over the fireboxes. The steel lintels have corroded, deformed and are no longer able to support the masonry above. Cracking and displacement of the brickwork above the lintels has occurred. If left unaddressed the lintels will soon fail leaving the chimney masonry unsupported and vulnerable to collapse.

To ensure the long term preservation of the Engineers’ Quarters, stabilization measures must be coupled with a program of ongoing observation to monitor conditions and the rate of material deterioration along with cyclical maintenance to repair areas of deterioration or failure as they are identified.

Pockets of mortar deterioration appear to be concentrated where the wall is effected by moisture . In general mortar deterioration is occurring in pockets along the base of the structure, at wall openings, at the exposed upper walls and at building corners. Where deterioration is most severe, the mortar has significantly weathered within the joint or has lost is structural qualities and capacity to bind the bricks together. The latter has resulted in several areas where bricks have shifted or fallen from the wall.

Recommendations for treatment include the following:

• Ensure limbs of adjacent tree are trimmed so as not to contact ruin during wind events.

• Repoint areas were mortar has deteriorated or lost it binding qualities.

Develop compatible repair mortars based on results of materials analysis.

The pointing mortar used at the Engineers’ Quarters was determined to be a lime and natural cement mix (See Appendix C).

• Repair/reconstruct minor areas where bricks have collapsed or fallen from the walls. There are several instances of this condition at building corners. Repair using salvaged brick that matches the existing in dimension, color and physical characteristics. Maintain coursing, joint width and tooling.

• Implement structural repairs outlined below.

STRuCTuRAL RECOMMENDATIONS bY SILMAN

The exterior walls of the Engineers’ Quarters Ruin (EOQ) are in-tact and stable overall.

There are pockets in the exterior and interior walls, indicating that there was a diaphragm level at the top of the walls framed with wood joists. Silman has performed structural stress analysis for the chimneys and exterior walls of the EOQ under lateral wind loading to determine if the missing floor diaphragm requires replacement to brace the masonry walls.

The design assumptions for this analysis are as follows:

building Codes:

Florida building Code 2010

ASCE 7-10

Code Wind Loading:

Risk Category II

Exposure b

Partially Enclosed building (Note: Code defined enclosure categories assume there is a roof on the structure, and the level of enclosure is dependent on percentage area of walls that are open.

While the EOQ structure does not have a roof, the wall opening area places it in the Partially Enclosed category, rather than open.)

Vult = 175 mph - ultimate wind speed

Vasd = 136 mph - Allowable wind speed q = 43 psf - Design wind load based on Code wind loading, we determined that framing at the original joist level as well areas of plan bracing is required to recreate diaphragm action, bracing the walls and providing a lateral load path. The Code Wind loading for Category 4 or 5 level storms, is higher than the typical storm

ENGINEERS’ QuARTERS RuIN 16 that occurs at the Fort. The EOQ building is also well sheltered by the Front 3 of the Fort structure, which is not accounted for in the Partially Enclosed category. The wall capacity is controlled by assumed mortar tensile capacity of 15 psi.

based on discussion with NPS, it was agreed that a lower level wind loading could be used for the analysis. Historic data for average and hurricane wind speeds were not available from local sources, so Category 2 to 3 typical wind speeds were assumed.

Category 2/3 Storm Wind Loading:

Vult = 140 mph (assumed) - ultimate wind speed.

Vasd = 100 mph - Allowable wind speed q = 20 psf - Design wind load

Following review of preliminary results with NPS and LAS, NPS expressed a preference to provide new joists, as interpretive elements to reflect original construction as well as provide supplemental support to the remnants of the original building walls. The end detailing at the joist pocket into the masonry wall and choice of materials is critical, to prevent moisture infiltration and subsequent deterioration of the new joists.

Silman and NPS have discussed using a tinted precast joists reinforced with carbon fiber bars, similar to the material used at the Front 3 Totten Shutter repairs.

At the door opening in the interior EOC wall, there is an arched brick lintel and bearing pockets for a missing flat timber lintel. The arched bricks are loose; Silman recommends repointing and resetting the bricks here. Silman will provide a detail for a replacement lintel over the opening

At the door and window openings at the exterior walls, there are stone lintels at the exterior and pockets for missing timber lintels at the interior.

Above the fireplace opening there is a 3”x3/8” thick iron bar “lintel” that is heavily corroded and is causing the outer wythe of brick above to push out upwards of 2 inches. The iron piece appears to have been originally curved upward as permanent form and protection of brick from heat at fire box.

Silman posits that the header courses arch over the opening and that the iron bar may have been upturned to create a curve at the arch, such that there was always a gap between the iron bar and brick above. In any case, the iron bar and brick above are to be removed and rebuilt from lintel level to +/- 12 inches.

Silman has designed new non-ferrous lintels at the fireplace and over doors and windows, and provided details for repair of any broken stone lintels.

[This page intentionally left blank]

APPENDIX A. DRAWINGS

Drawings obtained from the South Florida Collections Center

Everglades National Park

Drawer 74, Sheet 80. Fort Jefferson, Dry Tortugas, FLA, Sketch showing Conditions of Works June 30th, 1867 to accompany annual report.

Images obtained from the South Florida Collections Center

Everglades National Park

APPENDIX B. SELECT HISTORIC IMAGES

DRTO 301230 - Date unknown.

EVER 7136 - Ca. 1937

DRTO 301393 - Date unknown.

DRTO 301440 - Date unknown.

DRTO 301520 - June 1966.

DRTO 301522 - June 1966

DRTO 301523 - June 1966

DRTO 301530 - June 1966.

APPENDIX C. MATERIALS ANALYSIS

Petrographic Examination of Mortar (ASTM C856 and ASTM C1723)

Acid Digestion Analysis of Mortar (ASTM C136)

Water Vapor Transmission and Coefficients of Hygric and Thermal Expansion of brick

Wood Species Identification

Prepared for: Mr. David B. Woodham, P.E.

Atkinson-Noland & Associates, Inc.

Boulder, Colorado

Prepared by: David Rothstein, Ph.D., P.G., FACI Report No.: DRP16.1504

9 DECEMBER 2016

Petrographic Examination of Mortar Samples from Fort Jefferson, Dry Tortugas National Park, Florida

DRP Consulting, Inc. 3200 Carbon Place #104 Boulder, CO 80301 www.drpcinc.com

Fort Jefferson Mortar Petrography Report No. DRP16.1504 Summary Report 9 December 2016

EXECUTIVE SUMMARY

Three (3) mortar samples from Fort Jefferson located in Dry Tortugas National Park, Florida are subjects of petrographic examination to characterize the general composition and condition of the materials represented by the samples. The findings from this scope of work indicate that the three mortars have similar aggregates and distinctly different binder compositions. The sand is carbonate in composition and has a nominal top size of 1.18 mm (#16 sieve) in the Small and Large Powder Magazine mortars and a nominal top size of 2.36 mm (#8 sieve) in the Engineers Officers Quarters mortar. The sand consists of a variety of limestones and shell fragments. The sand content appears higher in the Engineers Officers Quarters mortar with tighter particle packing than was observed in the other samples

The Small Powder Magazine mortar has a lime-gypsum binder. Occasional lime lumps were observed and rare particles of anhydrite rimmed by gypsum and ettringite were observed.

Relatively large, tabular aggregations of material rich in ettringite with minor anhydrite were also observed commonly. These may represent fragments of a material rich in calcium and aluminum (such as clay) that were used production of the mortar and are now pseudomorphed by ettringite.

The Large Powder Magazine mortar is a lime mortar with no evidence of gypsum and no evidence of other cementitious materials observed. Numerous nodules of burnt limestone and burnt lime were observed as were partially burnt fragments of limestone.

The Engineers Officers Quarters mortar has a binder phase that consists of lime and a cementitious phase typical of a historical natural cement such as Rosendale cement.

! drpcinc.com i

1.0 INTRODUCTION

Mr. David B. Woodham, P.E. of Atkinson-Noland & Associates, Inc. (ANA) located in Boulder, Colorado requested DRP Consulting, Inc. (DRP) to perform petrographic examinations of mortar samples collected from Fort Jefferson located in Dry Tortugas National Park, Florida. The purpose of the investigation is to characterize the general composition and condition of the mortars represented by the samples. On 10 November 2016 Mr. Woodham delivered three (3) mortar samples to DRP. Table 1 summarizes information regarding the identification of the samples.

Fort Jefferson was constructed between 1845 and 1876 such that no information regarding the original project materials are available. The location of the structure on a remote island in the Gulf of Mexico provides exposure to heat, salt, destructive weather systems and water. No information was provided regarding the field conditions attending the areas that provided the samples listed above.

2.0 SCOPE OF WORK

The testing involved petrographic analysis of each sample according to ASTM C856 [‑ ] 1 supplemented with scanning electron microscopy (SEM) and energy-dispersive x-ray spectrometry (EDX) as described in ASTM C1723 [‑ ]. This report summarizes the findings of 2 this scope of work; Appendix A-Appendix C contains the notes, photographs and micrographs from the petrographic examination and Appendix D describes the procedures used to perform the tests.

Table1. Summary of samples

DRP No. Sample IdentiÞcation & Location

20YD8374 Small powder magazine, North wall at entrance

20YD8375 Large powder magazine, Southeast corner

20YD8376 Engineers ofÞcers quarters

! Standard Practice for Petrographic Examination of Hardened Concrete. Annual Book of ASTM Standards, Vol. 4.02., ASTM 1 C856-14.

! Standard Guide for Examination of Hardened Concrete Using Scanning Electron Microscopy, Annual Book of ASTM 2 Standards, Vol. 4.02, ASTM C1723-10.

! drpcinc.com Page ! of !1 3

3.0 FINDINGS

The following findings are relevant to the mortars represented by the samples. The referenced figures are in Appendix A, Appendix B and Appendix C.

3.1 Small Powder Magazine Mortar The sample consists of a mortar fragment that is 55 mm (2.2 in.) long, 35 mm (1.3 in.) wide and 7 mm (0.3 in.) thick and weighs ~ 17g (0.6 oz.; Figure A1, Figure A2). No embedded objects, cracks or significant microcracks were observed. The mortar is non-air entrained. The aggregate is a natural sand with a nominal top size of 1.18 mm (#16) that consists of limestone and occasional shell fragments (Figure A3). The paste is light gray (Figure A4) and the binder is a very fine-grained mixture of lime and gypsum. Lumps or nodules of hydrated lime and burnt lime were observed along with occasional grains of anhydrite rimmed by ettringite (Figure A5-Figure A7). Aggregations of material rich in ettringite and anhydrite were also observed (Figure A8). These aggregations are tabular in shape and the size of large sand fragments. The entire surface stained purple from phenolphthalein, indicating a lack of carbonation within the mortar paste (Figure A9).

3.2 Large Powder Magazine Mortar The sample consists of two mortar fragments. One fragment is 45 mm (1.7 in.) long, 35 mm (1.3 in.) wide and 7 mm (0.3 in.) thick and the other is 35 mm (1.3 in.) long, 25 mm (1 in.) wide and 7 mm (.3 in.) thick. The combined weight of the samples is ~ 15 g (0.5 oz.; Figure B1, Figure B2). No embedded objects were observed; a few hairline cracks are present. The mortar is non-air entrained. The aggregate is a natural sand with a nominal top size of 1.18 mm (#16) that consists of limestone and occasional shell fragments (Figure B3). The paste is light gray to white with a granular texture and dull luster (Figure B4). The binder is a very fine-grained lime-based mortar with no evidence of anhydrite, gypsum or ettringite observed (Figure B5). The paste contains calcium silicate phases that are carbonated (Figure B6, Figure B7). No portion of a freshly saw-cut surface stained purple from phenolphthalein, indicating carbonation of the mortar paste (Figure B8).

3.3 Engineers Officers Quarters Mortar The sample consists of a mortar fragment that is 60 mm (2.3 in.) long, 40 mm (1.6 in.) wide and 27 mm (1.1 in.) thick and weighs ~ 95 g (3.2 oz.; Figure C1, Figure C2). No embedded objects were observed; a few hairline cracks are present.The mortar is non-air entrained. The aggregate is a natural sand with a nominal top size of 2.36 mm (#8) that consists of limestone and occasional shell fragments (Figure C3). The paste is pale brown gray (Figure C4) and the binder is a very fine-grained mixture of lime and a cementitious phase that is similar to natural cement such as the historic Rosendale natural cement (Figure C5). Carbonated calcium-magnesium silicate phases were observe in the paste (Figure C6, Figure C7). No evidence of sulfate-bearing phases. No portion of a freshly saw-cut surface stained purple from phenolphthalein, indicating carbonation of the mortar paste (Figure C8).

! drpcinc.com Page ! of !2 3

4.0 CONCLUSIONS

The findings described above indicate that the three mortars have similar aggregates and distinctly different binder compositions. The sand is carbonate in composition and has a nominal top size of 1.18 mm (#16 sieve) in the Small and Large Powder Magazine mortars and a nominal top size of 2.36 mm (#8 sieve) in the Engineers Officers Quarters mortar. The sand consists of a variety of limestones and shell fragments. The sand content appears higher in the Engineers Officers Quarters mortar with tighter particle packing than was observed in the other samples

The Small Powder Magazine mortar has a lime-gypsum binder. Occasional lime lumps were observed and rare particles of anhydrite rimmed by gypsum and ettringite were observed.

Relatively large, tabular aggregations of material rich in ettringite with minor anhydrite were also observed commonly. These may represent fragments of a material rich in calcium and aluminum (such as clay) that were used production of the mortar and are now pseudomorphed by ettringite.

The Large Powder Magazine mortar is a lime mortar with no evidence of gypsum and no evidence of other cementitious materials observed. Numerous nodules of burnt limestone and burnt lime were observed as were partially burnt fragments of limestone.

The Engineers Officers Quarters mortar has a binder phase that consists of lime and a cementitious phase typical of a historical natural cement such as Rosendale cement.

This concludes work performed on this project to date.

David Rothstein, Ph.D., P.G., FACI

! drpcinc.com Page ! of !3 3

Fort Jefferson Mortar Petrography

Appendices

Appendix A Small Powder Magazine Mortar Petrography (ASTM C856/C1723) Appendix B Large Powder Magazine Mortar Petrography (ASTM C856/C1723) Appendix C Engineers OfÞcers Quarters Mortar Petrography (ASTM C856/C1723) Appendix D Procedures

! drpcinc.com

Appendix A: Fort Jefferson Mortar Petrography Report No.: DRP16.1504 Sample ID: Small Powder Magazine Mortar (20YD8374) Date: 5 December 2016

1. RECEIVED CONDITION

ORIENTATION &

DIMENSIONS

Fragment of mortar measures 55 mm (2.2 in.) long, 35 mm (1.3 in.) wide and 7 mm (0.28 in.)

thick. The sample weighs ~ 17 g (0.6 oz.; Figure A1, Figure A2).

SURFACES

One side of the fragment is relatively flat and medium gray in color; the other side is more uneven and white.

GENERAL

CONDITION

The mortar is hard and compact.

2. EMBEDDED OBJECTS

GENERAL None observed.

3. CRACKING

MACROSCOPIC None observed.

MICROSCOPIC None observed.

4. VOIDS

VOID SYSTEM

Mortar is non air-entrained and contains less than 1% total air as estimated from visual and microscopical observations (not determined in accordance with ASTM C457).

VOID FILLINGS None observed.

5. AGGREGATE

PHYSICAL

PROPERTIES

The aggregate is a natural sand with a nominal top size of 1.18 mm (#16 sieve; Figure A3).

The rocks are hard and competent; the grading and distribution are relatively even.

ROCK TYPES

The aggregate consists of fragments of limestone and occasional shell fragments. The limestones range from very fine grained (micritic) to medium-grained; some fossiliferous rocks were also observed. Numerous sand-sized particles that consist of ettringite and anhydrite were also observed.

OTHER

FEATURES

No deleterious coatings or incrustations observed. No low w/c mortar coatings observed.

! drpcinc.com A!1

6. BINDER OBSERVATIONS

POLISHED

SURFACE

Paste is light gray (Munsell 10YR/7/1) and has a smooth texture (Figure A4). The sample was impregnated with epoxy which tends to alter the texture, luster and hardness of the paste.

THIN SECTION/

SEM*

The binder phase is very fine-grained and consist of a mixture of lime and gypsum, based on the observation of lumps of hydrated lime, burnt lime and occasional grains of anhydrite (Figure A5). SEM/EDS analysis detected the presence of calcium silicate phases in the paste that may be from burnt siliceous limestone (Figure A6, Figure A7). Some of these phases are somewhat carbonated, based primarily on SEM observations. As mentioned above, large lumps of material rich in ettringite with vestiges of anhydrite were observed as well (Figure A8).

* Abbreviations as follows: RRCG = relict and residual cement grains; SCM = supplemental cementitious materials; CH = calcium hydroxide; ITZ = interfacial transition zone. Modal abundances are based on visual estimations.

7. SECONDARY DEPOSITS

PHENOLPHTHALEIN Entire surface stains purple (Figure A9).

DEPOSITS

Deposits of ettringite observed in large lumps in the paste; gypsum observed around some anhydrite grains.

! drpcinc.com A!2

FIGURES

(a) !

(b) !

Figure A1. Photographs showing oblique views of the sample in as-received condition. The red and blue dots show the orientation of the saw cuts used to prepare the sample. The scale is in millimeters.

! drpcinc.com A!3

Figure A2. Photograph showing the polished surface of the sample. The scale is in millimeters.

(a) !

Figure A3. (a) Reflected light photomicrograph of the polished surface showing overview of aggregate.

! drpcinc.com A!4

(b) !

(c) !

Figure A3 (cont’d). Transmitted light photomicrographs of thin section showing detail of aggregate in (b) plane-polarized and (c) cross-polarized light. The red arrows indicate particles of limestone and shell fragments.

! drpcinc.com A!5

(a) !

(b) !

Figure A4. (a) Photograph and (b) reflected light photomicrograph of polished surface showing overview and detail, respectively of the binder phase of the mortar. Scale in millimeters in (a).

! drpcinc.com A!6

(a) !

(b) !

Figure A5. Transmitted light photomicrographs of thin section showing components in the mortar in (a) plane-polarized and (b) cross-polarizedlight. The red arrow indicates a particle typical of burnt lime, the yellow arrow indicates a limestone particle and the green arrow indicates a phase that consists of ettringite and anhydrite.

! drpcinc.com A!7

(a) !

(b) !

Figure A6. Backscatter electron micrographs of the polished surface showing detail of paste. The green box in

(a) shows the area of (b) where the red box and green, red and yellow dots designate areas that provided the EDS spectra shown in Figure A7 (a), (b), (c) and (d), respectively.

! drpcinc.com A!8

(a) !

(b) !

(c) !

(d) !

Figure A7. EDS spectra indicating elemental compositions of areas corresponding to the (a) red box, (b) green dot dot, (c) red dot and (d) yellow dot in Figure A6. The spectrum and BSE image shown in Figure A6 are indicative of (a) ettringite, (b) limestone that contains minor silica, (c) carbonated calcium silicate hydration product and (c) carbon-rich residue in burnt lime or limestone.

! drpcinc.com A!9

(a) !

(b) !

Figure A8. Transmitted light photomicrographs of thin section showing detail of mortar components in (a) plane-polarized and (b) cross-polarized light. The red arrows indicate grains of anhydrite; the green arrows indicate ettringite.

! drpcinc.com A!10

(a) !

(b) !

Figure A9. Photographs showing (a) overview and (b) detail of the phenolphthalein stained surface. The scale is in millimeters in both photos.

! drpcinc.com A!11

Appendix B: Fort Jefferson Mortar Petrography Report No.: DRP16.1504 Sample ID: Large Powder Magazine Mortar (20YD8375) Date: 6 December 2016

DIMENSIONS

The sample comprises two chunks of mortar. One measures 45 mm (1.7 in.) long, 35 mm (1.3 in.) wide and 7 mm (0.3 in.) thick and the other measured 35 mm (1.3 in.) long, 25 mm (1 in.)

wide and 7 mm (0.3 in.) thick. The total weight of the sample is ~ 15 g (0.5 oz.; Figure B1, Figure B2).

SURFACES

One side of the fragment is relatively flat and medium gray in color; the other side is more uneven and white.

GENERAL

CONDITION

The mortar is hard and compact.

2. EMBEDDED OBJECTS

GENERAL None observed.

3. CRACKING

MACROSCOPIC

Occasional sub-vertical hairline cracks (~ 100 µm or 4 mil wide) up to 2 mm (80 mil) long observed.

MICROSCOPIC None observed.

4. VOIDS

VOID SYSTEM

Mortar is non air-entrained and contains less than 2% total air as estimated from visual and microscopical observations (not determined in accordance with ASTM C457).

VOID FILLINGS None observed.

5. AGGREGATE

PHYSICAL

PROPERTIES

The aggregate is a natural sand with a nominal top size of 1.18 mm (#16 sieve; Figure B3).

The rocks are hard and competent; the grading and distribution are relatively even.

ROCK TYPES

The aggregate consists of fragments of limestone and occasional shell fragments. The limestones range from very fine grained (micritic) to medium-grained; some fossiliferous rocks were also observed. Numerous sand-sized particles that consist of burnt limestone were also observed.

OTHER

FEATURES

! drpcinc.com B!1

6. BINDER OBSERVATIONS

POLISHED

SURFACE

Paste is light gray (Munsell 10YR/7/1) to white (10YR/8/1), has a granular texture and dull luster (Figure B4). The paste is moderately soft (Mohs 2.5-3).

THIN SECTION/

SEM*

The binder phase is very fine-grained and is a lime-based mortar, based on the observation of lumps of hydrated lime and burnt lime (Figure B5). SEM/EDS analysis detected the presence of calcium silicate phases in the paste (Figure B6, Figure B7). Some of these phases are somewhat carbonated, based primarily on SEM observations. No evidence of anhydrite, gypsum or other sulfate-bearing phases was observed.

* Abbreviations as follows: RRCG = relict and residual cement grains; SCM = supplemental cementitious materials; CH = calcium hydroxide; ITZ = interfacial transition zone. Modal abundances are based on visual estimations.

7. SECONDARY DEPOSITS

PHENOLPHTHALEIN No staining observed (Figure B8).

DEPOSITS Paste is carbonated.

! drpcinc.com B!2

FIGURES

(a) !

(b) !

Figure B1. Photographs showing oblique views of the sample in as-received condition. The red and blue dots show the orientation of the saw cuts used to prepare the sample. The scale is in millimeters in both photos.

! drpcinc.com B!3

Figure B2. Photograph showing the polished surface of the sample. The scale is in millimeters.

(a) !

Figure B3. (a) Reflected light photomicrograph of the polished surface showing overview of aggregate; scale in millimeters.

! drpcinc.com B!4

(b) !

(c) !

Figure B3 (cont’d). Transmitted light photomicrographs of thin section showing detail of aggregate in (b) plane-polarized and (c) cross-polarized light. The red arrows indicate particles of limestone and the green arrows show burnt lime.

! drpcinc.com B!5

(a) !

(b) !

Figure B4. (a) Photograph and (b) reflected light photomicrograph of polished surface showing overview and

! drpcinc.com B!6

(a) !

(b) !

Figure B5. Transmitted light photomicrographs of thin section showing components of the mortar in (a) plane-polarized and (b) cross-polarized light. The red arrow in each image indicates a particle typical of burnt lime.

! drpcinc.com B!7

(a) !

(b) !

Figure B6. Backscatter electron micrographs of the polished surface showing detail of the binder phase. The green box in (a) shows the area of (b) where the red, green and yellow boxes show areas that provided the EDS spectra shown in Figure B7 (a), (b) and (c), respectively.

! drpcinc.com B!8

(a) !

(b) !

(c) !

Figure B7. EDS spectra…