The above assumption doesn't absolutely mean that it is at the same time possible to follow the ancient constructive methodologies and techniques to gain this objective: it has to be clearly underlined that we are not repeating the same constructive experience that took place in the past, mostly because this way it wouldn't be possible to respect fully the objective of the work, this even if sometimes it is of great use to repeat some steps and to study and make researches on the hypothetical methods followed at the time.

The stone cut of the arch voussoirs is one of the works that may be not repeated following the ancient methods but following another technique that may guarantee a geometrical final correspondence to the ancient structure. The way this design of stone cut has been conceived and performed wouldn't have been possible without a computer management of all the data representing the 3d co-ordinates of every single stone joint of the arch. Stone wasting optimisation, management of irregularities, anomalies and documented imperfections concerning raising and heights of rows have been planned trough the use of purpose built software calculation routines and every single stone has been represented in 3d digital drawings and defined in its final co-ordinates and dimensions: stone cut design has soon brought to manage an electronic chart of many sheets with a global amount of elements that, including the analysis stage, counts about 200.000 data.

Nevertheless beyond this exact method for geometry control, it has been carried out a special methodology to manage and repeat smaller ordinary constructive imperfections trough special devices which will follow some procedures and steps of the ancient way of assembling the arch but in predefined ranges and to be performed among points of known positions.

Stone to be used and cut is Tenelija of the same type, colour and resistance of the ancient one originally used for the bridge construction, quarried in the same quarry where the ancient stones where presumably originally taken. Detail specification to guarantee the above requirements will be given by LGA company (which is in charge of all the laboratory tests).

Stone quality and correct procedures for all the quarrying phases should be guaranteed by a supervisor that is supposed to be on site during the works and that has the necessary expertise to make the checks. The supervisor should also check that the bedding is respected coherently to the warnings here next reported and that the numbering is performed as well as here next explained.

For what concern the lowest required resistance to compression stress and for the mechanical characteristic of the material it should be referred to the structural design report.

Each block of construction material should be checked with ultrasound tests at least once, before installation, and if necessary, more than once to establish the responsibility of the different involved companies; this will be up to the PCU TA. Ultrasound checks should be performed with the presence of a supervisor, and acceptable limits will be defined on the basis of the results of LGA company (which is in charge of all the laboratory tests).

Main assumption of this work is that all the measures of all the stones of the bridge are worked out considering the joints equal to zero. This assumption has been agreed with PCU TA and is aimed at a global simplification of the work mainly in the management of the co-ordinates which are therefore referred to the center line between two adjacent stones or voussoirs. Moreover this approach was necessary since the average dimension of the joint was not yet defined, at the current design stage, depending on many matters and being in the ancient construction quite variable.

On the other side this choice has been found favourable since joint dimensions will be used during the final cut as an additional margin to be managed in the assemblage phase, coherently with the global design approach which foresees this method for the sub-centimetre dimensions in order to perform a randomly distributed amount of ordinary construction imperfections.

Another assumption adopted for this work, agreed with PCU TA as well, was to work out dimensions for the stone blocks to be quarried increasing them of a quantity of three centimetres per side. Being the voussoirs of irregular shapes, the above mentioned device wouldn’t have guarantee a feasible procedure for the quarrying work and for this reasons it has been provided an additional calculation aimed at working out the smallest parallelepipeds dimensions as possible that could contain each stone to be cut. In other words, trough a purpose built software routine, three dimensions have been worked out for each stone block to be quarried, representing the base, height and thickness of a parallelepiped that is big enough to contain the final voussoir solid shape with a tolerance of a minimum of three additional centimetres for each side. Note that "for each side" means that if a base dimension is taken, for instance, this will be longer of a quantity of three centimetres on the left and of three centimetres on the right which in total brings six centimetres referred to that single measure.

To minimise the stone to be quarried these parallelepipeds are calculated along the average center line of the thickness of the voussoir to be cut and along the center line of the voussoir face, and this method will lead to a slightly variation of the ideal orientation of the natural stone bedding; mentioned variation is anyhow negligible.

The terms "stone natural bedding" are referred to a natural characteristic of the lime stones (like tenelija). Lime stone has an orientation, (bedding), due to different layers that during time have settled forming the stone vein, this natural orientation has an important structural value for stones used in constructions: stress should be orthogonal to the natural bedding.

For the above reason attention should be put during quarrying and during final cut in order to prepare stones with the correct orientation and not up side down.

In detail, during quarrying, the parallelepipeds should be taken out in a way that dimensions are compatible with the final assembling position: this means that depending on the type of block (voussoir, spandrels, parapets and cornices) a different approach should be followed as here next resumed (better explained in following paragraphs and in reconstruction works drawings):

Immediately after the quarrying a clear and undeletable tag (sign) should be stuck on the side of the stone which is parallel to the stone bedding, this to avoid misunderstandings with the following phase of final stone cutting (also because the bedding may be not clearly visible after stone drying).

During stone final cut the care for the natural stone bedding will be important as well, but it will be of help either the mentioned tag (sign) either the dimensions of the quarried block.

Numbering and coding of quarrying blocks is essential, if this operation is not correctly performed the bridge may be not easily built following the design indications. Immediately after the quarrying a clear and undeletable tag (sign) should be stuck on the stone (may be together with the indication of the bedding) with code and number that has been proposed in the classification system.

For instance signs like the following:

SMS-A103/5 means: arch stone located in row 103 position 5^th (starting from north)

SMS-Bsp_ne79 means: bridge spandrel stone located on the north-east side position 79 (ref. drw.)

Also after the stone final cut the tags (signs) should be repeated on the stones to avoid position exchanging during the delicate phase of the assembling.

Stone final cut should be performed following strictly the specifications exposed in this chapter and in the one about the centering; as it will be explained, final stone cut will be performed on a wooden structure model 1:1 identical to the one that will be settled on the load bearing steel centering.

Final stone cut will have to be performed with the control of a supervisor that is totally informed of the architectural design results and objectives and well aware of the warnings and instructions here reported. A 1:1 test before the beginning of this work is required in order to obtain the approval of the ICE commission.

Stone final cut should be performed following architectural design, either for what concern the final geometry, either for what concern the different steps and procedures of preparation: it should be completed off site on a model wooden deck in order to be assembled and finalised in a short time and with no incoherences. Any other different approach is most likely to be incompatible with the planned dimensions either for the stone final cut, either for the stone rough cut, and may lead to the following consequences:

• quite different geometry and final appearance of the bridge arch
• different number of rows and different shape of voussoirs
• incoherence and unmatched of joints (specially inner joints among rows)
• rough stone blocks and final stone blocks may not be adequate or enough dimensioned
• time for assembling may get too long and bring to the bad season with high risk for the whole structure

General Engineering will not be responsible of any of the above matters if design is not fully and strictly followed.

Stone finishing should follow specifications given by Mr. Bessac (who is in charge of all the archaeological research as a stone expert). LGA company findings about the different type of surface working should also be taken into consideration for the mentioned purpose.

A supervisor with the expertise on stone working should be present in order to check the correct performance of the required specifications. A 1:1 test before the beginning of this work is required in order to obtain the approval of the ICE commission.

The design of the stone cut involves all the bridge stones, even those that are still built-in (SMS-BS) and those that have been recovered (SMS-RS). As far as the arch is concerned, all the voussoirs have been designed from the arch springer.

Stone cut design of all the stones has been necessary and required for different reasons:

• recovered stones (SMS-RS) won't be reused to avoid additional cut of ancient stones and to avoid repositioning in different locations;
• it is not possible to foresee, at this stage, how many of the built-in stones (SMS-BS) will be remounted in the original locations (depending on integrity and on disassembling);
• for what concern the load bearing arch, only built-in stones in perfect conditions might be left on-site (to be decided while works are ongoing);
• for the reconstruction work some extra stones are required.

The delicate type of work and the amount of design documentation and investigations produced requires a permanent presence of a supervisor with the necessary expertise, in order to face any difficulty that may arise while works are ongoing.