English

5. Introduction

The architectural main works held during the PHASE A of the assignment may be subdivided in groups as follow:

• photogrammetry elaboration
• direct survey
• photogrammetry of the current condition
• photogrammetry of the previous condition
• geometrical analysis
• analysis of: ancient and current surveys, documents, photogrammetry results
• numerical management of the analysis and of the results
• graphic representation of the analysis and of the results

Photogrammetry elaboration have been an important and preliminary step for the representation of all the involved elevations. A survey of the ruined portions of the bridge and of the abutments is a base for any future design proposal, and this survey has been held documenting all the elevations stone by stone. Performing this step, either on the current situation, either on the previous one, has allowed important comparing results that have helped strongly the geometrical analysis: in fact, to work out the reliability and the accuracy of a survey, one of the test was to compare measures with the direct survey dimensions.

Geometrical analysis has been another important step towards the design phase: the documentation of the Old Bridge of Mostar was critical and there were different and contrasting data that didn’t allow an easy determination of what was the monument like.

The most important data sources, for geometry determination of the arch of the bridge, were the drawings of the ancient surveys. These drawings are important documents and they gave the possibility of working out the previous shape of the bridge stone by stone.

It is self evident that these technical drawings of the bridge have been verified to check how much precise they were; for this reason many different tests have been performed looking for eventual differences and incoherences. What has appeared to a first examination of the drawings, is that parameters like quality, condition and copy passages were very much affected by heterogeneity; this means that we did not have the same level of reliability for each view of the same survey, and this of course has led to more complex evaluation systems.

Everything produced during this work phase has been basically managed in a numeric way, and the results have been set either in a numeric layout connected with automatic recalculation system (in case of changes during the elaboration), either in a graphic layout that is in a CAD digital system, (and therefore has been produced numerically as well, and may be printed in any scale with high level of accuracy).

The way it has been studied the bridge geometry was not deterministic: at the present moment the vault of the bridge doesn't exist anymore, and it is not possible to survey it with a predefined tolerance, it has only been possible to gathered the original shape from a mediation of the different data sources we now have; this was not an easy step since the differences from the sources had to be considered one by one and analysed so that to locate which was the source that gave a wrong metric information. This procedure had to be followed for both sides of the bridge, and these two curvatures were related so that to locate any deformity in the thickness of the bridge itself.

So the methodology or theory followed for this task is a "probabilistic strategy for geometry determination" and may be resumed as "the seek for the most likely bridge", which is reliable, scientific and has given important results compared to the existing data.

The most delicate phase of the whole operation has been the determination of a criteria for evaluating data and their reliability. Even if it can be found an accurate description of the matter in the following paragraphs of this chapter (from §5.4 to §5.8), it may be briefly summarised the main principles adopted:

Since we didn't have so many different sources (1955 and 1982 surveys) for every single data to work with statistic methods, it has been given difference importance and weight to the surveys. The 1955 survey, as long as we know, looks like a traditional survey with little or no use of special measure instruments. This may be confirmed by the type of dimensions that have been taken and by the way they have been graphically represented. For the above reason, measures reported in the 1955 survey may be reliable mostly for what concerns local dimensions, short distances, and easy-to take dimensions. There are, of course, exceptions to this statement that have been verified: like the arch span which was dimensioned perfectly despite it were a long distance (but it was an easy-to take measure). Dimensions, like the ones of the X and Y co-ordinates of the intrados line, being the most difficult to survey and being at the same time quite long distances, are mostly uncertain. While local dimensions of every single stone are much more reliable.

The 1982 survey seems to be a more reliable survey, since it has been held with purpose built instruments, and with the photogrammetric technique. Nevertheless, from special inquiries held, it seems that the shoots taken for the task were not so close to the monument (this also for technical requirements) and this may have brought to a difficult interpretation of details like the stone layout of the elevations being the natural colour of the stones also very much light.

For the above reasons any difference found has been analysed in terms of type of measure and source data: before any change could take place it has been verified that there weren't any incoherences in the adopted survey for that dimension. The above mentioned procedure has led to work out the "most likely Bridge of Mostar", which has been called "2000 survey".

The exposition of the phase A work was already started in the Inception Report (see chapter 2), and here next is developed and enlarged with the most recent developments.

The on site survey has been held with instrumental systems: digital total station (electronic theodolite) and laser measuring kit. Measures have been taken all over the ruined portions and all over the elevations involved in the present assignment. The origin of the system has been set on a local reference point to allow any future development and integration of the work, and data has been put in cad system in 3d co-ordinates: this file has been the first one that has been classified as a source file, and with this some other files have been generated (like the one of the bridge springers in 3d-coordinates).

The 3d direct survey has allowed the generation of a series of source files that have the same origin point, and that may be all loaded one over the other allowing any future study on the bridge and on the surroundings. The three dimensional elaboration and design control is a special processing that has been started to put the basis of the design and to manage the complex geometry of all the points, the stones and the cornices of the bridge.

The direct survey has allowed a detailed measuring of the remaining portions of the bridge arch and of its stones; moreover control points have been taken suitably to foreseen photogrammetric and stereometric processing.

During the on site survey it was held the photographic documentation campaign that was co-ordinated with the control points survey.

Photogrammetry of the current condition has been performed basically in two dimensions, as required by the ToR, each elevation has been reduced to a plan and has been represented in an orthogonal projection.

The basic input data for the photogrammetric elaboration were: control points taken on the elevations and suitable pictures of them (taken from a good point of view and with special cameras). The photographs, after been developed, were scanned and converted in a digital format in a computer system for final rectification processing. For this final task it has used a purpose built software package produced by Siscam - Galileo Siscam Technology - for photogrammetric surveys of historical buildings; it is named ARCHIS 32. This software package allows to transform a perspective view (photograph) in an orthogonal view through a rectification procedure using control points. Rectified images can be joined one to the other to create a single one mosaic. The final image is metrically correct, so that it is possible to use it for measurements of all details that appears on the photo.

To perform the above described processing it was necessary, first of all, to convert the co-ordinate system of the control points trough a rotation and translation of the reference axis on a new vertical plan chosen as "best fit" of the points belonging to one elevation at the time; this has been done in CAD and has given, as final results, control points referred to the related plan, as if they were taken in two dimensions.

Subsequently the processed images have been imported in a CAD file, they have been scaled, and they have been vectorialised (digital drawing in overlay). Drawings have been performed reproducing every single visible detail stone by stone, moreover drawings have been organised with different layers so that to allow any future elaboration or tematics.

Checks have been performed in order to correct any eventual incoherence among different but adjacent plans; and a refining work has been done in order to obtain a correct and understandable architectonic representation of the ruined portions of the bridge.

Drawings have been divided in two categories: towers-buildings and west-east abutments; for more details on the subject refer to chapter 10 of this report (§10.1.1).

Some application of three dimensional photogrammetry have been performed in addition to the foreseen duties of the ToR this because some controls were necessary to investigate the 3d co-ordinates of the ruins of the bridge, and some additional dimensions and checks nearby the springers of the bridge.

Moreover for what concern the survey round elevations profiles of the Towers, it has been necessary to use 3d photogrammetry devices to gather a correct representation of the masonry layout, even if this methodology was absolutely not foreseen by GE’s assignment: while, in the preliminary drawings, (phase A of the assignment), the towers have been represented trough the architectural main outer lines, in the finalisation of this section of the survey, (phase B – final drawings), additional detailed surveys of the elevation surfaces have been provided.

The above mentioned 3d photogrammetry process requires double photographic shots, (stereo pairs), and a predefined number of reference points that should be clearly visible on both the shots taken. Aforesaid pictures, taken with purposely built cameras, have been inserted in a soft copy workstation that, with the aid of special visualisation device, allows the 3d digitalisation of a stereo-model of the architectural object included in the shots. For this final task it has been used a purposely built software package produced by Siscam - Galileo Siscam Technology - for photogrammetric surveys of historical buildings; it is named STEREOMETRIC-PRO.

To perform the above described processing it was necessary, first of all, to convert the co-ordinate system of the control points trough a rotation and translation of the reference axis on a new 3d co-ordinates system suitable for each representation view. The 3d digital drawing is, while work is ongoing, related to a portion of the involved architectural object (elevation); at the end of the procedure, when all the different stereo-pairs have been digitised, all vectorial data files are imported in a single CAD file, being the co-ordinate systems coherent to the original reference survey system.

Final stage of the work foresees the selection of the representation plans, (according to assignment requests), on which all the data are projected in a planar representation. To have the final layout, matching all the commonly known architectural representation requirements, the digital drawings are finalised gaining the requested views.

Following the same method exposed in the above paragraph, photogrammetric applications have been performed also for the previous condition, but instead of using purpose taken pictures, have been used ancient ones. This of course has generated technical and accuracy problems and moreover has brought to the research of reference points that could be easily visible on the ancient pictures and, at the same time, could still be recognisable on the ruined portions of the Old Bridge. Sometimes it has been necessary the use of stereo-photogrammetry elaboration to gather from the surveys other control points more suitable for the specified elaboration. Control points had to be rotated and translated following the procedure described in the above paragraph.

Special attention has been put on the bridge elevations because the processing could bring to interesting data to be compared with other sources like the ancient drawings. Unlikely the available ancient pictures were of a very low quality and the ones, in which the elevations were taken in one shot, were not detailed enough to document the stone joints and connections.

But, among the ancient pictures, there were some in which, even if it was not possible to gather information about the arch curvature, it was clearly readable the stones layout of the spandrels with all the joints. Those pictures after being processed have been compared with the ancient surveys, and they have denounced some inaccuracies in the 1982 survey of the spandrels.

Photogrammetric elaboration of ancient pictures is, anyhow, an interesting and useful work method for cases of this type, it allows metric evaluation and verifies on structures and buildings that do not exist anymore. But for this task a wide and high quality photo documentation is required, otherwise there is no use for the studies and researches. Nevertheless General Engineering has proceeded in processing all the available ancient pictures that could give even a small contribute to the reconstruction design of the Monumental Complex of the Old Bridge of Mostar. For more details on the subject it may be referred to chapter 4 of this report where drawings are exposed and commented.

CREDITS:

Intellectual property of this report and of the design drawings is owned by General Engineering s.r.l.

© - General Engineering Workgroup -

SOURCE:

Final Design Report