Site Conservation 2014

Introduction

The primary focus of architectural conservation works during the 2014 field season were the Terrace Building Complex, the Early Phrygian Gate, the Visitor Circuit, and the pebble mosaic in display at the Gordion Museum.

The Terrace Building (9th century B.C.), located on the southwest side of the eastern citadel mound, is composed of eight large rooms that functioned as workshops for textile production and food processing. It has been excavated for its entire length of over 100 meters, thus making it one of the largest buildings of ancient Anatolia. The Terrace Building is among the most prominent ruins of the Early Phrygian citadel, and is visible from almost every corner of the visitors’ circuit. Conservation in 2014 continued with the structural stabilization of the walls inside Building 5 (TB5) and documentation of the construction and conditions of the wall in Building 7 (TB7).

The other significant conservation project that started in 2014 is the stabilization of the Early Phrygian Gate. This is the best-preserved city gate in Iron Age Asia Minor and it has remained intact for nearly three millennia. A significant bulge has recently developed in the limestone masonry of the South Court, however, and it is increasing each year, thereby jeopardizing its stability. In 2014 scaffolding to provide access for stabilization works on the north and east walls of the gate’s South Court was installed. Works included ground leveling of the central ramp, construction of a timber mat foundation for the heavy-duty scaffolding along the north wall, and access scaffolding.

New galvanized cable railings and stone steps have been installed over the last three years to improve and extend the Visitor Circuit in the citadel as requested by the local authorities. In 2014, stabilization of the scarp slopes and replacement of the barbed wire fence with new galvanized railing were completed along the area between visitor’s information panel 4 and 6 on the north side of the Citadel.

In 2013, a selected panel from Megaron 2 pebble mosaic currently housed in the Gordion Museum was temporarily faced, then lifted and stored at the Gordion Museum. Discovered in the 1950s, the Megaron 2 pebble mosaic floor is the earliest known mosaic to have been found (9th century B.C.) and features elaborate polychromatic geometric designs. Shortly after discovery the pavement was cut into sections for lifting and transported to the Gordion Museum. In 2014, following on three years of research and planning, conservation treatment started on the lifted panel by removing the old reinforced concrete backing using a custom-made grinding system.

Terrace Building Complex

The Terrace Building Complex is an industrial zone of Early Phrygian date (9th century B.C.) on the southwest side of the eastern citadel mound. In 2009 the site program explored new methods of conservation and preservation of the Terrace Building Complex to maximize the archaeological value of the existing walls and floors with no or limited rebuilding and to improve legibility with no or limited reburial. A pilot conservation program of the walls in TB2 was established according to objectives and intervention strategies to minimize visual impact, reduce maintenance requirements and limit the need for future interventions. These resulted in the development of a set of guidelines for works at the TB Complex.

The focus of masonry conservation works during the field seasons of 2014 was Terrace Building 5. All wall treatments were conducted following guidelines established in 2009. Works were carried out under the direction of Angelo Lanza assisted by Giuseppe Bomba and a team of trained local laborers. Intervention was limited to five weeks (June 18 – July 23) because the group was engaged in setting up conservation works at the museum at the beginning of the season and in masonry works at the gate after scaffolding/shoring installation.

Site photography and survey of existing conditions of TB walls continued during 2014 along the walls of Terrace Building 7. Works were carried out by the documentation team from the University of Pennsylvania under the direction of Meredith Keller assisted by Nityaa Iyer and Jocelyn Chan.

As every season, maintenance of the soft caps installed on the treated walls of TB1, TB2 and TB3 was also completed. Removal of unwanted vegetation from the caps was coordinated and carried out by Naomi Miller. A description of her work is included in the appendix section of the present report. Unfortunately, installation of new soft caps along the newly treated walls during fieldwork was interrupted this season due to the scarcity of Poa grasses in the area surrounding the mound. In fact, the winter and early spring of 2013/2014 were extraordinarily dry, with fairly heavy late rains at the end of May and beginning of June. This regime was unfavorable to annuals and perennial seed production (especially for the grasses) was remarkably low.

A summary of works completed during the 2014 season include:

• Removal of temporary reburial (applied in 2013) along TB5-e-1, TB5-e-5, TB5-e-7, TB5-e- 8, TB5-e-13 and TB5-e-14 walls and complete masonry consolidation and treatments;
• Treatment survey of TB5-e-1, TB5-e-5, TB5-e-7, TB5-e-8, TB5-e-13 and TB5-e-14;
• Installation of a temporary wall capping along TBR-25, TBR-26 and TBR-29;
• Exposure and reburial of TB6-e-3, TB7-e-6, TBR-32, TB7-e-7, TB7-e-8, TBR-34, TB7-e-13, TB7-e-14, TBR-35, TB7-e-12, TB8-e-12, and the west section of TBR-36 for wall documentation;
• Rectified photography and documentation of conditions along TB6-e-3, TB7-e-6, TBR-32, TB7-e-7, TB7-e-8, TBR-34, TB7-e-13, TB7-e-14, TBR-35, TB7-e-12, TB8-e-12, and the west section of TBR-36.

Masonry consolidation of TB5-e-7, TB5-e-8, TB5-e-13 and TB5-e-14

These walls represent the elevations of the interior walls that separate the ante-room from the rest of the unit inside TB5. All walls showed traces of the original doorjambs located at the center of the room in the area where probably a door was placed to connect the ante-room with the rest of the unit During the 2013 field season, the walls were exposed, surveyed and covered again with reburial material bagged in burlap or plastic bags placed along both sides of the masonry and a temporary capping system of plastic sheeting, stone edging and a mixture of sand and clay placed on top. Unstable areas of the wall with great splay along TB5-e-8 and TB5-e-13 were first shored with wooden posts and planks and then covered with reburial material. During 2014, the bagged backfilled material covering the walls was removed and cleaning of the masonry was carried out by hand and with great attention in order to avoid collapse. Where required, shoring was temporary left in place.

TB5-e-7 and TB5-e-8 are the east and west elevations of the wall located at the insertion of TB5-e- 6. The wall is approximately 4,60 meters long with a variable height of 1,30 meters at the intersection with TB5-e-6 and 0,50 meters at the other end of the wall in the area of the doorjamb toward the middle of the room. The masonry at the intersection with TB5-e-6 is a restoration using small blocks of reddish stone set in mortar. The work was conducted between the late 1990s and beginning of 2000 by Mark Goodman during works inside TB4. The original portion of masonry is splayed toward the exterior of the wall with most of the large blocks fractured in place. The blocks in the area of the doorjamb present great deterioration and disaggregation. As with other walls, dismantling of all original blocks along both TB5-e-7 and TB5-e-8 elevations was necessary because the fractures in the blocks made impracticable the installation of stainless-steel cable reinforcement inside the core. As for the interior walls inside TB1 and TB2, during the dismantling process of this wall it came to light that it was originally constructed with large blocks, some of them placed across like big header stones. All of the large blocks were found fractured still in position or splayed from the wall, which probably occurred during the fire and collapse of the roof. These large stones were carefully dismantled piece by piece and adhered together with epoxy resins or grouted piece by piece in place.

TB5-e-13 and TB5-e-14 are the west and east elevations of the other single wall at the intersection with TB5-e-4 in the interior of TB5. The approximate length of this wall is 4,55 meters with a maximum height of 1,10 meters at the intersection with TB5-e-4 and a minimum of 0,40 meters at the doorjamb. The remaining masonry along both elevations of the wall is all original and formed by large blocks, some showing fractures. As described for the other interior wall, the masonry of this wall is splayed toward the outer side and this deformation is more evident along TB5-e-13 near the area of the doorjamb. The blocks along both elevations of the wall were more intact and only a few of them fractured. Therefore, dismantling of masonry was very limited because it was possible to keep most of the masonry in its splayed position by inserting a stainless-steel cable system in the wall core to retrofit the masonry.

Masonry consolidation works were carried out as follows:

• Weeding and cleaning inside TB5 was completed by village women. The vegetation consisted of short and dense grasses, much less than in other seasons due to lack of rain during the spring;
• Hand removal of burial material. This was carried out with extreme care and organized simultaneously along both sides of a single wall;
• Masonry consolidation works were also organized simultaneously along both sides of a single wall. As needed, displaced fractured blocks were numbered in the rectified elevations of the walls and then removed from their original position to allow epoxy treatments. Original core material was removed simultaneously during dismantling and saved for future use. Epoxy treatments of individual stones were not always possible due to the deteriorated condition of some fractured blocks, especially of the reddish burnt stone. In this case, the fragments of a block were returned back in position and then grouted in place. Where required, dismantled original blocks were set in place using a hydraulic lime-sand-brick dust mortar (2.5:1:0.5) prepared on site. Original core material was used to refill the cavity between both stone veneers while reconstructing the wall. In order to reuse as much of the original masonry as possible and limit the use of new blocks, some original blocks were turned in place to have the inner face exposed to the outside of the wall while the deteriorated face was placed toward the wall core.
• The use of new stone infill blocks was very limited because most of the deteriorated masonry was repaired and reused. Only the bottom courses along TB5-e-7 and TB5-e-8 were cut and shaped by hand from rough limestone stored on site or from recycled stone blocks found during the on-going archaeological excavation in TB6. The blocks of the bottom course were all set in place in lime mortar. Dutchman replacement with new blocks was used at the lower stone courses near the wall corners along TB5-e-13 and TB5-e-14.
• Across walls TB5-e-13 and TB5-e-14 it was possible to place a retrofitting system with stainless-steel cables that allowed leaving most of the masonry along both these elevation in their deformed (outer splayed) position avoiding full dismantling.
• In the areas of the walls corresponding to the doorjambs, all original fractured stones were grouted in place.
• Masonry stabilization was completed by the insertion of chinking stones in lime mortar in the open joints of the masonry.

Masonry consolidation of TB5-e-1 and TB5-e-5

TB5-e-1 and TB5-e-5 represent the exterior and interior elevations respectively of the northeastern wall of TB5. Unlike the other TB5 walls described above, these elevations were first exposed during 2014 fieldworks and stabilized immediately after completing site photography and survey of conditions.

Most of the masonry along TB5-e-1 (exterior elevation) was found exposed with the exception of the northern side by TB6 where the wall was buttressed with soil. Once cleaned, the wall is three stone courses high with a fourth course showing at the intersection with TB4 on the south side and at TB6 on the north. The total length of the wall is almost 11 meters and the width of the wall is approximately 1,40 meters. All the original masonry along the southern 1/3 of the wall was found in good condition without major structural problems. Most of the blocks of this section presented fractures so they were removed and repaired with epoxy and then used during reconstruction. On the other hand, the withstanding masonry along the middle and northern end of the wall was formed by highly deteriorated blocks showing fractures, loss, flaking and disaggregation so new stone replacement was the only possible remedy during the stabilization of the wall. Due to the unstable conditions of the northern portion of masonry by the corner with TB5-e-4, masonry stabilization of this section will be completed next season together with the walls of TB6.

TB5-e-5 is the interior elevation of the northeastern wall of TB5 that is visible from the interior of the unit. Most of the masonry along this elevation was found covered by soil. The south side by TB4 and the north by TB6 were covered by backfilling material placed by Mark Goodman in the late 1990’s while the soil covering the center of the wall was placed in the late 1950’s or early 1960’s to backfill the archaeological remains of the grinding stations found along the wall. The stabilization approach selected for this elevation was to remove the portions of backfilled material applied by Goodman and to limit exposure of the masonry by leaving the soil in the center of the wall in order to preserve the archaeological remains. Only the top stone course was exposed and treated in place. Masonry stabilization was completed along the south portion of masonry by TB4 and along the top masonry course in the middle of the wall. The northern portion of the wall by TB6 was temporary backfilled and will be treated together with TB5-e-4 next season.

Masonry consolidation works were carried out as follows:

• Vegetation reduction inside and outside of TB5 was done by village women;
• Hand removal of burial material along the north side of TB5-e-1 and along both ends of TB5-e-5. Only the top masonry course was carefully exposed and cleaned to leave the archaeological remains of the grinding stations protected by soil. All burial material was bagged and used to buttress unstable areas of the ground along TB5-e-1;
• Dismantling along TB5-e-1 was limited to few fractured blocks along the top rows on the north portion. On the contrary, dismantling of masonry was necessary along TB5-e-1. All fractured blocks along the north section were removed to be repaired and all severe cracked blocks along the center were also removed and replaced by new or recycled masonry units due to the poor conditions of the masonry;
• All dismantled fractured blocks were repaired with epoxy resin and returned to their position along the wall. The upper course along the middle of the wall in TB5-e-5 was grouted in place;
• Reconstruction along the center portion of TB5-e-1 was carried out following the same procedure used for other walls. In this case, it was not possible to save any original blocks for the reconstruction so newly cut and recycled blocks were used instead. Original core
  material was used to refill the wall cavity while wall reconstruction;
• Chinking stones and lime mortar where inserted in the gaps of the original masonry and also used in the areas of reconstruction with new masonry.

Exposure and reburial of TB6-e-3, TB7-e-6, TBR-32, TB7-e-7, TB7-e-8, TBR-34, TB7-e-13, TB7-e-14, TBR-35, TB7-e-12, TB8-e-12, and the west section of TBR-36

Like every field season, the wall documentation team from the University of Pennsylvania returned to the Citadel during 2014. The focus of the team was the photography and documentation of the existing conditions found in the walls of TB7. Removal of the temporary reburial material to expose the masonry was carried out by the labor team along TB6-e-3, TB7-e-6, TBR-32, TB7-e-7, TB7-e- 8, TBR-34, TB7-e-13, TB7-e-14, TBR-35, TB7-e-12, TB8-e-12, and the west section of TBR-36. While exposing the walls portions of original mud-plaster covering large areas of the walls came to light in very poor and fragile conditions. Once all walls and wall tops were fully documented they were covered again with backfilling material that will stay in place until final stabilization works are carried out in the future. For recovering the walls, all material removed during clearance works was saved and bagged in plastic or burlap bags and then reused during backfilling. This system of wall protection with bags filled with debris has proved quite successful because it makes the process of reburial and exposure more efficient in terms of time and labor.

Works were carried out as follows:

• Vegetation reduction inside and outside of TB7 was completed by village women;
• Manual removal of burial material was carried out with the help of trained laborers. Additional care during clearance was necessary due to the presence of mud-plaster remains on the masonry. In addition, wooden plank and post shoring was installed along unstable areas of masonry, in particular along TB7-e-6 where the original blocks were not stable due to a great splay toward the outside of the wall. By buttressing the deformed wall with wooden elements it was possible to keep the original masonry in place during the time the wall was exposed;
• Manual cleaning to remove soil debris from the masonry and plaster remains was carried out with the aid of various brushes and small trowels. In addition, all the plaster found was temporary covered with geofabric to minimize deterioration due to exposure;
• Rectified photography and full documentation of existing conditions of the masonry and plaster remains was completed by the documentation team from the University of Pennsylvania;
• After documentation all areas covered by plaster were first faced with burlap or geofabric to help create a separation layer between the plaster surface and the backfilling material. Then, bags filled with debris were stacked along both sides of a wall like buttresses, plastic sheeting or burlap was placed on top across the wall and finally a fill of stone rubble, clay, sand and soil were placed on top of the wall to avoid water penetration from rain or snow.

Early Phrygian Gate

First excavated in the 1950’s, the Early Phrygian Gate is the most complete extant gate to survive from the Iron Age in Asia Minor. However, its dry laid construction leaves it vulnerable to the region’s high seismic activity. Constructed around 900 BCE, the gate served only briefly as the main entryway to the Citadel. Successive periods of occupation within the Citadel mound resulted in further building campaigns which incorporated the earlier structure as the foundation for new construction. The resulting changes from subsequent loading caused a series of visible conditions, most notably cracking and displacement. Although cracking occurred historically from the additional loads of the later city walls, displacement continues to be an active condition. From 2006 to 2010, the Architectural Conservation Laboratory of the University of Pennsylvania documented, monitored and assessed the gate’s overall structural stability to determine the condition of its limestone and rhyolite walls (from “Gordion Awakened. Conserving a Phrygian Landscape”, 2011).

In 2012 David Biggs from Biggs Consulting Engineering, joined the Gordion conservation team to study the walls of the Early Phrygian gate. After significant site investigations and observations of the gate walls during his first visit to Gordion, Mr. Biggs put together a number of recommendations for the structural strengthening and stabilization of the gate. The second visit to Gordion in 2013 involved additional investigations on the gate. These included excavations at the South Court walls which provided additional data to structurally analyze the court walls for seismic effects and make long-term recommendations regarding the Middle Phrygian fill around the court walls. In addition, non-destructive scanning of the North Court walls provided essential data for evaluating the structural condition of the walls and estimate the amount of voids within the masonry. Ayşem Kιlιnç, architect and conservation graduate from the University of Pennsylvania was hired to help David Biggs with preparation of preliminary works at the gate. She also played a key role in contacting local Turkish firms and contractors for the manufacturing of scaffolding and design of a gantry crane to be used in masonry dismantling.

During the field 2014 season, David Biggs and Ayşem Kιlιnç returned to Gordion to organize preliminary works at the gate that included the installation of a soil sub-base, a timber mat as scaffolding foundation and a metal scaffolding along the north and east walls of the south court. Ayşem also completed construction documents necessary for masonry intervention. In addition, masonry repointing works started along the east elevation of the south court. The following is a summary of field activities:

Soil Leveling along the center pathway of the gate:

• Scaffolding installed to temporary shore the unstable masonry at the southwest corner on the south side of the gate was removed. The lower masonry along the base of the south ledge was consolidated by inserting chinking stones and lime mortar in the open joints among the stone blocks. This work was necessary to strengthen the masonry before installing the scaffolding on top of the ledge;
• Soil from the on-going archaeological excavation southeast of the Citadel mound was brought with a backhoe to the area of the gate and used for leveling the ramp along the central area;
• A non-woven geo-fabric was placed as a separation layer on the existing ramp before installing the soil sub-base;
• Dry soil was spread manually on top of the geo-fabric in a layer of 15-20 cm approximately making sure to fill any low spots. The soil was first compacted by hand, then wetted several times and finally compacted with a tractor. A second layer of soil was added on top and wetted and compacted in the same way previously described. The soil material was sloped to match the ramp.

Timber Mat Foundation:

• Used railroad ties for the timber mat were provided through an agreement with the Ministry of Culture and Tourism, Kultepe Excavation, Kayseri Museum and Polatlı Municipality;
• The timber mat foundation was designed by David Biggs using the railroad ties in a grillage pattern of cross ties with an upper mat resting on the lower mat with a short crosspiece of timber as base for each scaffolding post;
• Installation of the timber mat started from the southwest corner at the south ledge. The timber elements were placed directly on the geofabric over the slopped sub-base and iron rebar pieces were used to nail together the timber. Where necessary, sieved soil was added below the ties to help leveling the piece in place;
• Installation of the timber mat was carried out by A-R YAPI, a local contractor from Ankara that was also responsible for scaffolding installation. This way it was easier to align the position of the top crosspiece with the vertical elements of the scaffolding;
• The empty spaces in between the ties at the bottom of the mat were filled with compacted soil and gravel before laying the next row above;
• A second grillage row was added on top and then a short crosspiece of timber was placed as base for each scaffolding post;
• Some modifications to the original timber grillage design were necessary while installation in order to compensate some irregularities of the slopped ground. These modifications were sent to David Biggs, who was not on site when installation took place.

Scaffolding:

• The heavy duty scaffolding was manufactured by Tamer Kalip ve Iskele in Ankara and installed by a local contractor (A-R YAPI) along the north wall of the south court. It has two levels of support for the stones of the north face of the south court that will be removed in the area of the bulge. It was designed for the weight of the stones at approximately 1200kg/m2 (240psf) on each of the top two levels plus the hoist system;
• The scaffolding was placed over the timber mat foundation and installation started from the southwest corner along the north wall;
• The two levels with metal decking were erected and then bracing was installed. A stair for personnel access was placed at the west side of the scaffolding;
• A frame for the access scaffolding along the east wall of the south court was also installed by the contractor. Parts of the old scaffolding system used by Mark Goodman at the gate in early 2000 were reused along this area of wall;
• Angelo Lanza completed the installation of the access scaffolding along the east wall and added a system of metal ramps and platforms that will be used for the removal of the wall core material while dismantling takes place along the north wall.

Masonry Works:

• Once the access scaffolding was installed, masonry treatments started at the lower section of the east wall above the area grouted by Mark Goodman in the beginning of 2000. The open joints of the masonry were repointed with chinking stones and lime mortar. Voids behind the masonry were filled with lime mortar or grout and rubble. Detached stone spalls from the original blocks were re-adhered with micro-injections of a lime based grout or where possible with an epoxy resin. Some large spalls at the north corner were re-adhered with epoxy and reinforced by a stainless steel rod. These treatments will continue next season along the upper portions of the wall while masonry works take place along the north wall;

The team contacted Schilling Germany through the local firm of Mr. Ahmet Kadioglu in Ankara for the manufacturing and purchasing of an aluminum gantry-crane to be used in the partial dismantling of stone blocks along the bulged area of the gate. A final design and price from Schilling were received in September 2014 and enough funds where left unspent after fieldworks to be used in purchasing the gantry crane.

Visitor Circuit

Since 2009, the conservation team of the University of Pennsylvania has been making incremental improvements to the visitor path of the Gordion Citadel. Since then, various prototypes of canopied pavilions and new informative signs have been proposed in addition to construction of new stone stairs and paths, and installation of new galvanized cable fencing.

During the field season of 2014, installation of new fencing along the visitor path continued along the northwest side of the Citadel between visitor information panels 4 and 6. The escarpments along this area contained several unstable soil portions protruding from the upper areas of the scarps with basal erosion and undercuts due to the effect of erosion. The approach used was to pull back the position of the fence and the visitor information panels, cut back the protruding areas of soil and use the removed soil in filling and leveling the areas of undercuts.

A condition assessment of the escarpments around the perimeter of the excavation area was completed. This information was used to develop a catalog of ways to cut or fill the escarpments in order to create a stabilization plan. For more information refer to Angelina Jones’s field report 2014.

The following is a summary of works:

Consolidation of unstable scarps

• Overgrown weeds along the visitor path and removal of the existing barbed-wire fence were carried out manually by the laborers. In areas of danger due to the proximity of the path to the edge of the scarp, protective bands in red and white with warning signs were installed to alert visitors of the danger;
• The location of visitor panel #5 was temporary removed in order to shape the escarpment;
• The area of the visitor circuit along the top of the scarps was leveled and all soil heaps were removed with a backhoe. The soil was used to fill undercuts present in the area of intervention;
• Friable soil portions protruding from the upper areas of the scarps were removed manually. A team of four to five workmen with shovels and picks removed first the material from the top letting the soil fall to the base of the scarp. This work proved to be quite arduous due to the instability of the soil to be removed that created a dangerous situation for the labor crew. It was necessary to work very slowly in order to avoid any accidents;
• All loose material accumulated at the base of the scarp was moved in wheelbarrows and used to fill areas of undercuts;
• Some sections of soil were very hard to be removed using the system described above. Also, it was too dangerous for the workmen to be working so close to the edge of the scarp. In this case, the soil was first manually shaped into steps or terraces and then shaped into a slope close to the angle of repose;
• All ceramic material found during this intervention was collected, carefully washed and organized so it could be inspected by Ken Sams and Shannan Stewart. All ceramic remains were then reburied under the soil added along the undercuts;
• To speed up the process, a backhoe was brought to clean all loose soil from the base of the scarp and to fill the areas with undercuts created by soil erosion. The backhoe was also used to press the soil in place along the base and top of the scarps to level the slope.
• Finally, all shaped slopes were wetted and the soil was pressed with the aid of the backhoe.

New Fence

Since 2009, the old rusted barbed-wire fencing has been gradually replaced with a new system of galvanized piping and cable to allow for maximum visibility into the Citadel Mound. The new fence follows the design proposal from Lindsay Falck. It consists of 3 mm galvanized pipes of 48 mm diameter placed at approximately 3 meter intervals, with drilled holes for five rows of galvanized cable at 25 centimeter intervals. Two pieces of galvanized cables across each panel block the entrance of large animals to the interior of the citadel. The galvanized cable is fixed to the pipes with stainless steel anchors and turnbuckles serve to tense the cables in place.

In 2014, approximately 28 galvanized pipes of the new fence were installed along the visitor path on the north section between information panels 5 and 6. The new fence was assembled and installed by a special team of local laborers under the direction of Őmür Atιtιğ, site foreman trained by Angelo Lanza. New fence installation works included:

• On site, the location and alignment of the new fence was established with strings and iron bars or wooden stakes placed in the ground at intervals of about 3.00 meters to match the spacing of the vertical pipes;
• Foundations for the fence were dug by hand. The new fence was positioned inward with respect to the old barbed-wire fence so the unstable top edge of the scarps could be reshaped and stabilized;
• The galvanized pipes of the new fence were positioned dry inside the foundations so it was possible to control and adjust the right alignment and uniform height of all pipes before pouring concrete. In order to provide more support and block the pipe in place inside the foundation, two pieces of 8mm-diameter rebar were inserted into the pipe through drilled holes opened at the bottom.
• Stone rubble was packed around each pipe to block in place the iron rebar at the base;
• The foundations were filled with ready-mix concrete alternated with more stone rubble until almost reaching the ground level. This operation was carried out with great care to avoid displacement of the pipe inside the foundation. While pouring concrete, a bubble level was used to control the verticality of the pipes and a mason string was used to make sure that all pipes were aligned at the right height;
• Once the concrete was completely set the base was buried with soil. In between the pipes, 5 rows of 4mm galvanized cable were stretched. Galvanized cable proved more flexible and less likely to deform over time than galvanized metal wire used in the original design. The cable was attached to the pole using stainless steel anchor bolts (manufactured by HILTI) inserted in holes in the pipes and the tightness of the cable was achieved by using galvanized turnbuckles.

During the 2015 spring an herbicide adapted to the vegetation of the mound will be tested to prevent weeds from growing in the joints of the masonry in the new stairs and paths. The application of the herbicide will only be limited to the top and sides of steps and paths. The efficacy of this treatment will be evaluated during next field season of 2015 in order to decide if this treatment could be part of the seasonal maintenance activities along the visitor circuit.

Pebble Mosaic Panel

Four panels from the Megaron 2 mosaic pavement in display at the Gordion museum were identified for conservation treatments during the 2011 field season. In August 2013 one of these four panels, panel 7, was lifted from display and stored at the museum during the off-season months. In June 2014 the panel was removed from storage and conservation treatments begun following a program of research and treatment testing at the University of Pennsylvania September 2013 – May 2014.

The main scope of the intervention was the removal of the heavy iron reinforced concrete from the back of the panel and replacement with a new backing system for greater stability and lighter weight. In addition, the decrease of weight will also help in the transportation of the mosaic panel to the United States for the Gordion exhibition at the Museum of the University of Pennsylvania scheduled for 2015.

A system using a Bosch router adapted for use as grinding tool to remove the concrete backing of the lifted panel was designed on site with the help of Mitchell Bring, SUNY Buffalo. A working table and scaffolding frame for the grinding operation were custom made on site by Angelo Lanza with the help of the site laborers. Grinding of the concrete backing and conservation treatments were implemented by Kevin Wohlgemuth, Meredith Keller and Jocelyn Chan under the supervision of Frank Matero. In addition, Sema Kurekci¹, Jessica Johnson, William Shelley and Cricket Harbeck helped the team with testing of different treatments on the pebbles of the mosaics in display and finishing the concrete backing of the lifted panel.

The following steps summarize the conservation intervention of the lifted mosaic panel during the field season of 2014:

• Step 1: Construction of a working table.
  A sturdy working table was needed for grinding the concrete from the back of the mosaic panel. As per the few instructions received from Mitchell Bring, the table was to be flexible enough to regulate the working height and strong enough to support the weight of the panel. Few local craftsmen were contacted to build a table but the cost of manufacture was rather expensive. In the end, it was best and cheaper to build the table on site using a custom made base of metal pipes hold together with scaffolding clamps and a top made of wooden planks manufactured by a carpenter in Polatlı. The working height of the table could be adjusted by screwing the scaffolding jacks in or out inserted in the six vertical pipes at the four corners and middle of the table. This system also proved efficient for leveling the table with great accuracy as required for the grinding operation. Additional diagonal pipes were included in the design of the base in order to guarantee no movement due to the vibrations while the grinding of the concrete took place. The wooden top was fixed to the bases of the jacks with nails and then blocked in place with wooden wedges
• Step 2: Removal of the mosaic panel from storage and transportation to the working table.The cover of the wooden crate containing the mosaic was unscrewed and all the foam padding around the mosaic was removed. All four sides of the crate were also unscrewed to complete the exposure of the mosaic leaving only the bottom of the crate sitting on a wooden structure that served to carry the mosaic last year and that was left in place. The metal pipes were added to reinforce the carrier and increase safety during transportation of the mosaic to the working table. Due to the heavy weight it was necessary to use six persons for lifting, one man at each side holding the wooden carrier, and four men to hold the metal pipes at the front and the back. Lifting and transportation of the mosaic to the area of the table was carried out synchronizing the movement of the six persons in order to avoid any unbalanced movement that could result in displacement of the mosaic. Before placing the mosaic on the table, it was first necessary to turn it upside down so to have the concrete at the top and the pebbles at the bottom. To aid in the lifting, the ropes left in place last year were wrapped around two pipes placed across the panel and four persons handled each of the four ends of the pipes. Once in the air the mosaic was slowly moved out the bottom of the crate and delicately lowered to foam padding placed on the floor. Turning of the panel upside down was done on the floor by four persons and once the concrete back was placed at the top a crack across the concrete was first noticed². Failure at the middle of the long panels is to be expected given the length to width ratio of many of the panels and their relative thinness. Many panels cracked during their initial lifting and after weathering outdoors at the Dig House before their transfer to the museum. Their dimensions and extreme weight argue now for a stronger, lighter support system in order to facilitate their display. Two metal pipes were slid under the panel across the crack to help support better the panel in case the crack opened during maneuvering the mosaic to the top of the table. Once the mosaic was in place, the foam padding and the pipes were removed from underneath the panel to leave it flat on the wooden table top.
• Step 3: Building a frame and carriage for the router.
  A special frame built with metal pipes, scaffolding elements and pipe clamps was built to hold the router according to Mitchell Bring’s instructions. The frame was placed on both sides of the table and secured to the table base using additional pipes and clamps. A Bosch router purchased in Ankara was fitted with a diamond bit and secured in a custom carriage to be adapted for use as a grinding tool to remove the concrete backing. The router and its carriage were fastened to metal pipes positioned on the metal frame and manual guided across the panel to create consistent, even cuts to remove approximately three millimeters of concrete during each pass. This process was implemented to test the potential for using a CNC³ cutting system for grinding the concrete of other mosaic panels in successive seasons. This method is crucial to the task of removing/reducing the concrete backing without the use of percussive action which will dislodge the poorly bonded pebbles in the concrete matrix.
• Step 4: Grinding the concrete backing and treatments of the concrete.
  For details on grinding the concrete backing and treatments of the concrete refer to Kevin Wohlgemuth’s field report 2014.
• Step 5: Protection of the concrete backing and storage of the mosaic panel.
  The grinding method of the concrete backing proved successful and allowed the removal of approximately half of the concrete until partial exposure of the rebar. Following grinding, different options were discussed to remove the rebar so grinding of the concrete backing could be completed. Unfortunately, several portions of the rebar were located in great proximity to the pebbles and removal of the rebar could have resulted in damage or detachment of the pebbles. As the field season was almost over, it was best to discontinue grinding the concrete backing until next season. The remaining days of fieldwork were used to apply a treatment to the back of the panel to prevent corrosion of the metal reinforcement and level and stabilize the remaining concrete left in place⁴. Following the direction of Melek Yildizturan⁵, and Sema Kurekci, the mosaic panel was left without receiving the epoxy adhesive and lightweight honeycomb backing to allow for consultation. Storage of the mosaic for the next season presented a new challenge to the team as the process used to maneuver the mosaic out of storage could not be used again. In order to prevent any damage to the pebbles, it was best to move the panel and the wooden table top together. Thus, the wooden wedges and nails connecting the top of the table to its metal base were removed and six persons carried the mosaic panel lying on the wooden top to the storage room of the museum. Once on the floor, the four sides of the old crate were screwed to the wooden table top around the panel to complete building a new crate for storage. Foam sheeting was used to cover the top of the mosaic panel and rolls of foam were fitted in the spaces between the crate sides and the mosaic to form a protective padding. Boxing of the mosaic panel was completed by screwing a top to the four sides of the crate. The metal pipe base of the working table was kept and stored so it could be reused during future treatment of the mosaic panel.

Other treatment options were also explored on several mosaic panels on display at the museum. One test treatment was the removal of the cementitious overgrout, a prevalent condition affecting many of the panels that conceals the faces of the panels and distorts the display’s design and integrity. Additionally, various pebble compensation methods were explored in areas of both historic and recent loss. For detailed information of these treatments refer to the 2014 mosaic field reports.

Footnotes

¹Senior Conservator from the Museum of Anatolian Civilizations (Ankara)

²The concrete side of the panel was exposed for the first time to the team after flipping the mosaic.

³CNC stands for Computer Numerical Control.

⁴For details on the treatment of the concrete backing and protection of the metal reinforcement please refer to Kevin Wohlgemuth’s field report 2014.

⁵Archaeologist and Ankara Representative of the Ministry of Culture and Tourism

Credits

Project Director: Frank G. Matero

Project Supervising Conservator: Elisa Del Bono

Project Team: Angelo Lanza, Meredith Keller, Ayşem Kιlιnç, David Biggs, Mitchell Bring, Sema Kurekci, Giuseppe Bomba, Angelina Jones, Kevin Wohlgemuth, Nityaa Iyer, Jocelyn Chan