Condition Assessment of Ancient Architectural Monuments in the Perga

Berliner Beiträge zur Archäometrie Seite 23-32 2008

Condition Assessment of Ancient Architectural Monuments in the Perga mon Museum- The Combined Application of NDT-Methods for Struc tural lnvestigations fRANK WEISE BERND REDMER URS MüLLER Federal Institute for Materials Research and Testing (BAM), Germany

M. H üBNER Soc. OfGeophysical lnvestigation (GGU), Germany

Abstract ln this paper the possibilities of the combined application of nondestructive testing (NDT) methods for structural investigation of ancient architectural monuments is prescnted using the examplc of the columnar portal of the holy hall at thc market gate of Pricne in the Pcrgamon Museum, Berlin. Due to missing documentation andin thc context of a structural condition assessmcnt, it was nccessary to detenninc the rear anchoragc of the entablature and to verify the structural bond bctwccn the arti ficial stonc and the original material in the architrave of the monument. Tbc combined use of radar and vidcoscopy proved very effic1ent in conjunction with the detcrmination of the type and posi tion of the meta I componcnts of thc rear anchorage of thc entab lature. Thc structural investigation of the architravc was accomplished by radar and radiography. The radar method was utilized to determine the overall position of the mctal special fittings. Additional investigations by radiography using a ncw genera tion of digital imaging platcs in the region between original marble and artificial stone showed an air gap, which was not visible from the outside. Furthermorc, it was observed that two dowels from the marble reach into this gap, but arc bent upwards and have thercforc no static function. This led to thc important concl usion that the Ioad bcaring capacity of the architrave and thus of thc whole cntablature is establis hed exclusivcly by a continuous meta! girder with a T-shaped profilc.

lntroduction

Beginning in 1875 the Berlin museums carried out several excavation campaigns near the Turkish coast, mainly in Pergamon, Priene and Magnesia, which Iasted until the beginning of thc 20'' cenrury. To exhibit the excavatcd objects in an appropriate envi ronment, a new musewn building in Berlin was planncd and opened in 1930 as the Pergamon Museum. From the beginning the museum was designed to feature the full sized architcctural objects. However, in recent years many defects of the museum building were recorded and rcpair was dcemed neccssary. Since many of the rccords

23 documenting the objects in the museum wcrc missing, or dcstroyed as a rcsult of World War 11 , it was decided to carry out an assessment of the materials and condi tion ofthe objects bcfore undertaking any intervention.

The investigation focused on a reconstruc tcd portion of thc holy hall at the market gate of Priene. This acted as a pilot study to prove the suitability of NDT-methods forthis type of construction. The main task focused on the entablature, which is pri marily reconstructed in reinforced artifici al stone (Fig. I).

The only originalmaterial in the architra ve is one !arge marble block, complemen ted by arti ficial stone. The structural bond between the original material and the artificial stone was initially unknown, and the construction method for ancho ring the cantilevered sima to the wall behind was unclear.

Since both data are necessary to reconfirm the structural stability ofthe monument, it was necessary to apply non-destructive inspection techniques in order to solve the Figure 1. View ofthe columnar portal of fo llowing investigative problems: the holy hall o fthc market gate of Pricne.

Determination of the type o f rear anchorage of the heavi ly cantilevered cornice. Verification of the type of structural bonding between the arti ficial stone and the original marble in the architrave.

2 Investigative concept

To solve the investigative problems the non-destructive testing techniques reprcsent ed in Figure 2 were applied.

2.1 Radar method

The radar method applied to both investigative problems cnabled a first and fast glo bal survey of the relevant parts of the building. lt was successfully utilized fo r detec tion of meta! reinforcements in the entablature. The measurement principle is based on the reflection of short electromagnetic impulses at material boundary layers with

24 strongly different dielectric Determination of the I characteristi cs, as found in \., V this case at the intcrface bet constructional connection of the ween marble/artificial stone rear anchorage of original marble with the natural cantilevered cornice stone substitute and the steel reinforcements. J \., The measuring setup consi Global aurvey sted of an intcgrated pair of 11 l V I ll 1.5 GHz antennas (including Videoscopy Radar~ethod ~l l Stereoradiography ll(rcfcxion tt~rr:~~ngomont ) I transmitter and receiver). The V setup was chosen considering I Dlrected detail I I investigations I thc thickness of the construc tion parts, the very low moi Figure 2. Jnvcstigative concept. srure content of the material, accessibility from three sides and the rcquired high local rcsolution. For rhc actual mcasurement the antennas werc guided by hand along a measuring grid on the surface ofthe specific parts. Thc results werc then representcd as radargrams, which coincided with the vertical and horizon tal lines of the surface grid (Fig. 3a). In such aradargram the intensity of the retlec ted signals was represented ovcr the moving path of the antcnna versus the depth of the construction part (Fig. 3b). A cylindrical retlector (e.g. a round stecl bar) appears, duc to the opening angle of the antenna, as a hyperbolically shapcd diffraction in the radargram.

The information of the radargrams can be substantially impaired because o f Superpo sition of many retlections, due to various inhomogencities in the structural elemcnts (e.g. cavitics, change of material such as marblc/ artificial stone, special meta I rein forcements). According to cxpcrience, every now and thcn clear statemcnts can not be made by using radar as a structural investi gation method exclusively (DGZfP 200 I). lt is thereforc recommended to amencl the information gathered by radar by cmploying other non-destructive testing mcthods.

moving path moving path II 111 IV V

·...... ~ :/· · 01 c:- zyl indrical ·:=!E- .. I reflector ·:;; I .Dc: I -"'o.:. I ; I :\II I ..... I I I I Cl I I I I I II 111 IV V VI radargram with hyperbolic diffraktion

b. ~ h.'3 ">llrcmcnt arram.!.t-'m~nt and 111\.'a:,un.:mcm rc:,ult '' ith ~~ lindricul rt.>Ot: ~tol

Figure 3. Mcasurement principle and practical application of the radar method 111 reflection arrangement.

25 2.2 Videoscopy

Interconnected cavities and voids behind the tryglyph frieze, the dentil course and the gaison allowed the use of videoscopy for analyzing the rear anchorage of the sima (Fi g. 4 ).

The cavities were accessed by drilling a small hole into the concrete slab covering the entablature. Figure 5 provides a visual impression of the videoscopic Observation on site.

2.3 Stereo radiography

The analysis of the marble I artificial stone interface delivered an initial overall pic ture concerning the position of steel reinforcements in the architrave. In addition to radar, stereo radiography was applied in order to acquirc detailed information about the exact position, dimension and shape of these reinforcements.

The technique of radiography is based on thc attenuation (absorption) of ionizing radiation by differences in material density and the thickness of a construction ele ment. Two radiographic images from different positions are needed (Fig. 6a). The reconstruction of the position and size of the meta! parts is real ized by a graphic or

concrete slab

"'....(J)

"'N.... L =-=:J - 1\Ti."lfT ~ al~r'!'=i'f{ 5 57 234 66

a. Front 'i~" Figure 4. Detail design of the entablature of thc columnar portal.

Figure 5. Real ization of the videoscopical observation of the anchoragc at the canti levered sima.

26 computer aided back projecti on techniquc. Figures 6b and 6c show a typical arrange ment of thc radiographic mcthod.

Considering the material th ickness, a y-cmitter, the radio nuclide Co-60 with an acti vity of 35 Ci, was used as a primary radiation source. Thc detector systcm consi stcd of digital imaging plates (Ewer! et al. 1996) witb a Iead-tin intermediate filtcr for the reduction of thc cantrast reducing scatter radiation. The exposurc time is comparable to a fluorescent screen film system and amounted up to 2.5 h depending on the wall thickncss.

3 Results and their cvaluation

3.1 Rear anchorage of the cantilevered cornice

Figurc 7 shows an exampl c of a radargram, obtained by moving the radar antcnna along a vcrticalline ofthc mcasuring grid on th c side elcvation ofthe entablaturc. The two upper diffractions werc causcd by hori zontal running tie rods and th e lowe r by horizontal running reinforcing bars.

with special metal source reinforcement position I

Source position 2 digital imaging plate

J_ P nn~.:1pl~ '""'Ich ~tccordmg w h Po'>~llonmg of thc c Computcd r.tc.hogr.tph) tcd1 IXo/11' I'I'Jo ;.u.holiUHl o.,nurcc ({"n-(•0) noqur (( R I)

Figure 6. Measurcment principle and practical application of the stereo radiograpby

runtime [ns] 7 6~ 0246810

--···-· ;:·t-:: · ··········· ·i~~~~~~:~f·J •.·· tie rod · - 1~· -~ ., , ""'< . --':.: ... ., : I •) .~•\.' .. .~; .... +:·...... rr · ~~- . \ . ~':i: : .. .,:· I L...... M..... Ii ~I ' 1 • • re·bar i . ' : ...... • ~ -,--.-..,c---..., F====:!:j 0 20 40 60 80 struture depth [cm]

a. Im ölig.ll~d >tdc h. '>tdc 1 '~" ofth~ entahlatur~ c. Radargram ofa \Crltcal ofth.: cmablalllrc "11h mca!'tun:mcnt grid mca\un:mcnt lim:

Figure 7. Sclccted results ofthc radar invcstigations with a 1.5 GHz-antenna pa ir.

27 The videoscopy permitted an exact verification of the cross sectional shape of all tie rods. Altogerher steel profiles with L- and U-shapes as weil as flat steel bars wcrc used. (Fig. 8).

3.2 Structural bonding of the original marble in the architrave

Figure 9 illustrates an exemplary radargram with two different diffractions. The radar gram has been measured by moving the antenna along a verticalline ofthe grid app lied to the front of the entablature. The upper diffraction suggests a !arge surfacc parallel reflector (back wall ofthe frieze plate) and the lower a geometrically clearly smaller reflector (meta! reinforcing bar). Since all the radargrams along the vertical grid lines on the front ofthe entablature differ only slightly from each other (with thc exception ofthe boundary region), it can be concluded that the meta! profile must run continuously through the architrave. This is also confirmed by the radargrams from the underside of the architrave (Fig. I 0). Herc meta! dowels can bc rccognizecl in the area between original marble ancl artificial stone. However, their exact position, shape and dimensions cannot be determined by radar only. The same applies to the shape

~ ' ---l .. f ~ =:: b. ~ R r------iF9- f=----1 --t

I ~ ~ A - >!

a. Position ufthc dcu.. ·c ted stc~l scclions inthe b. Classilictllion o fthc cross longitudinal cross section A·A ~ct.: t io n J\.J\ in sidc 'ie''

Figure 8. Result of the videoscopic investigations.

runtime [ns] 0 2 4 6 8 10 12 14 = - -=-!.. ~. ~ backside 1 ~?;tf . . ~f.1 ' 1 of the frieze ,--- -; II. • metalgirder · ·· .. ,..~. ttI I ·'Jiirf ' " I ~ .j . ,.,..._.,___ 020406080 structure depth [cm]

a. Cros~ ~c<.:ti on of b. Front 'il"'' of thl.!' ~.:ntahlalun: '' ith c. Radar!.!ram of i.l 'crtical t h t~ ~ntablatun: '' ith m ~asurc m cnt raslcr anJ mca;urcm~nt linc int~rprcta tion Interpretation Figure 9. Selected result ofthe radar investi gations at the front ofthe entablature with the 1.5 GHz-antenna pair.

28 and dimensions ofthc continuous meta] girders in thc architrave. The results from ste reo radiography of selected areas provided more in depth information concernmg these issues.

These are described by the examples provided by Figures II and 12. In the graysca le images bright areas indicate high material density and/or greater thickness of a structural element. Dark areas represent the opposite. The grayscale image of figure II, which was obtained by the external arrangement o f the radiation source, illustra tes thc presence of a gap between the orig inal marble and the artificial stone. ln addi tion, parts ofthc meta! bars and girder passing thro ugh the architrave could clearly be recognized. It was evident that the dowels extended only into the original marble and not into the artificial stone. This lcd to thc important conclusion that the dowels do not tie the marble to the arti ficial stone.

_: ...~ +~~-~---.:4_::,:. :": ~r,;ry,~..;," .I ' .

. ·l run time [ns) -· 2 4 6 8 10 12 14 "':"~...... ·~ " dowel :~?4 metal girder !I:f- ..,_, \ • I dowel ...... 1 ·.JIJJ;:-;. :io 4b so 6b structure depth [cm)

u. cro...... ')t..:clion of h. hont \it..:l.l. and, isiblc tuukr~r< k nf thc c. Rada r·gram of a thc cntablaturc '' ith cntabhlturc '' ith mc~Hwrcmcnt gri d. 'ic" l'rJ mrasun:mcnt linc th~ Interpretation meas ur~mem lme. as \\eil as thc mtcrprctatton

Figure J0. Se Ieeted result o f thc radar investigations at the underside of the architra vc with the 1.5 GHz antenna pair.

-..:...... · - : .. - ~ 'Ii. .. ·.::~· '

~-CJ-90 cm,

ll ~- I ~--1 + I '\ I conlinous C tr' I metal profile Lcgends ------L-.-- Digital imaging plate ...... (40 x 30 c m') airgap - Filter: 0.5 mm Fe·sheet a. Pb-Sn-foit comblnat. benl dowet

Figure II. Result of the radiography on a selected measurement point with external positioning of the radiation source.

29 The grayscale image ofFigure 12 was acquired by positioning the radiation sourcc in the entablature internally. It shows the exact position and shape of the dowel in the interface ofmarble and artificial stonc: the dowels are embedded into the marble with mortar and are bent outward, not locking into the artificial stone. The geometry ofthe measurement arrangement and the resulting radiograms also allowed for detection of a T-shaped profile for the girder, which runs through the architrave.

Figure 13 shows in summary the exact spatial position of all non-destructively deter mined reinforcements in the architrave with their shapes and dimensions.

~ ~ ' 4 ' • - ~ .:. ..;.'• ~- · :i• ,_:,_ ·.~ • •

90 _, - .-

stone substitute

bent dowel in mortar bed mortar bed

marble

continous metal T-profile

Figure 12. Result of the radiography on a selected measurement point with internal positioning of thc radiation source.

·--~·

D-. ~

22 11 13 20

Figure 13. Reconstructed position of all meta! reinforcement in the architrave as a summary of the application of non-destructive testing methods.

30 4 Conclusions and Summary

8oth, the type of rear anehorage ofthe heavily eantilevered sima and the nature ofthe mechanical bonding between the artificial stone and the original marble in the archi trave, could be determined in detail by the combined application of different non destructive testing methods. The following test methodology worked satisfactorily in praetice:

Preliminary global investigation by radar 2 Purposeful in depth analysis, depending on the investigative problem, by: Videoscopy (In the case study: type of the rear anchorage of the cantilevered sima). - Radiography (In the ease study: mechanical bonding between artificial stone I original marble in the architrave).

The following useful information for the struetural integrity of the monument eould be communicated to the structural engineer:

- Type and position of the steel sections for the rear anchorage of the sima. - Confirmation of the Ioad bearing capacity of the architrave, which is established primarily by a continuous T-shaped steel girder (dimensions were determined). - The presence of an air gap in the architrave at the interface of marble and artifi eial stone. - No actual eonnection between the original marble and the artifieial stone, since the dowels protruding from the marble were bent upwards into the gap.

5 Acknowledgement

We express our thanks to the ARGE Pfannerat the Pergamon Museum for the inter esring task assignment, and to the Fecleral Office for Building ancl Regional Planning for the financial Support of the study. Furthermore we would like to thank the colle agues ofthe ,Non-destructive Testing ancl Characterization ', ,Radiographie Methods' and ,Building Materials' divisions at the Federal InstituteforMaterials Research and Testing (BAM) for the effort expended and partieipation during nightly measurement sessions and Gabi Patitz from the Engineering Office for Building Diagnostics, Damage Analysis ancl Structural Design (IGP). Last but not least we want to thank Mary Hardy from the Getty Conservation Institute in Los Angeles, for proof reading thc manuscript.

References

DANIELS, D. J. 1996. Surface-Penetrating Radar. Institution of Electrical Engincers. London.

31 DEUTSCH E GESELLSCHAFT FÜR ZERSTÖRUNGSFREIE PRÜ FUNG (DGZfP), 200 I. Merkblatt über das Radarverfahren zur zerstörungsfreien Prüfung im Bauwesen (ß I 0). ßerlin. DEUTSCHE GESELLSCHAFT FÜR ZERSTÖRUNGSFRE IE PRÜFUNG (DGZfP), 1990. Merkblatt für Durchstrahlungsprüfung von Stahl- und Spannbeton (B I ). ßerlin. EWERT, U., STADE, J. , ZscHERPEL, U. & KALI NG, K., Luminescence imaging plates in industrial radiography - first experience and comparison to the film; Trends in NDE Science & Technology, Proceedings of the 14th World Conference on Non Destructive Testing, New Delhi, 8-13 December 1996. Vol. 3: 134 7 - 1350.