The constructive techniques of the moorish roofing frameworks: The case of the Mirador of the Reales Alcázares of the Catholic Monarchs in Seville

The constructive techniques of the moorish roofing frameworks: The case of the Mirador of the Reales Alcázares of the Catholic Monarchs in Seville In recent times the roofing frameworks which cover the Mirador (observation pavillion or covered area) atop the Reales Alcázares (Royal Palace) in Seville have been restored. Thanks to this restoration it has been possible to examine first-hand and close-up this magnificent example of Moorish architecture, constructed during the reign of the Catholic Monarchs (Ferdinand and Isabella), which transport us from today to a very special epoch, and which speak to us of six long centuries that have left their mark on each of the basic elements. The Mirador of the Catholic Monarchs is located on the top of the high Palace of Don Pedro above what is today the Hall of King Carlos V. Its construction caused a partial destruction of the upper room of the prince, which today is called Don Pedro' s Bedchamber, cutting off the arched entryway to that room, due to the difference in height between the two. It is open on the one side looking over the garden and on the other to the high gallery of the Patio de Las Doncellas, or Maidens' Patio, and it is located between the aforementioned Bedchamber of Don Pedro and the Family Dining Room. According to Ana María Fidalgo, we had no knowledge of it until the year 1977, when the architect Rafael Manzano Martos discovered it during the restoration process being performed on the Palace of Don Pedro. This same author (Ana María Fidalgo) asserts that during these recent restorations, the ceilings were returned to their original state, since they had undergone multiple remodeling projects up until then. Cecilia Cañas Palop Nevertheless, this is not entirely the case, since in more recent investigation and studies carried out by Inmaculada Ramírez López as the Restorer of the roofing framework, paintings and inscriptions have been discovered that had been unknown of to date, and that could very well belong to the original roofing framework. The first restoration of which we have information is the one carried out by Juan Fernández and Juan de Simancas, since in a report given by the Masters of the reconstruction, there is an indication that in the new room above the garden, all the deteriorated wood of the roofing had to be repaired, being substituted by new wood and then being re-roofed.] Posterior restorations of the roofing framework have been numerous but not much information exists about them. However, they are visible in the multitude of elements that have been added throughout the centuries, with better or worse results. A BRIEF PRESENTATION OF THE FRAMEWORK Its form is a square framework made up of four aguas and three roof slopes. Due to the constructive solution of the pares, we are analysing a framework of par and nudillo which uses limas mohamares2 in the intersection of the different roof slopes. All these have been made using the apeinazado technique, and the final decoration was resolved with a tie-of-eight and mocárabe squares. The four faldones were Proceedings of the First International Congress on Construction History, Madrid, 20th-24th January 2003, ed. S. Huerta, Madrid: I. Juan de Herrera, SEdHC, ETSAM, A. E. Benvenuto, COAM, F. Dragados, 2003. 514 decorated alternating rows of eight-pointed stars and little tieso ELEMENTS AND METHOD OF CONSTRUCTION In order to describe the method of construction of the roofing/ceiling, the study was based upon an exhaustive, detailed analysis of each and every basic element and the relation between each element and the others. In the first place, we present and place each of the elements of the roofing framework, beginning with those elements that have to guarantee its placement above the walls on which it rests. This perhaps is the least known area, because of being hidden from sight, and yet it is al so the most important factor in guaranteeing the general stability of the structure. We will see later on how the greatest part of the problems that the framework has, have their origin in the supportive elements. THE HIDDEN STRUCTURE. THE FOUNDATION OF THE FRAMEWORK A very important fact worth bringing to light in regards to the foundations of this roofing framework, and also a few other roofings analysed (such as the Antecapilla or the Bedchamber of Don Pedro), is the lack of basic elements to guarantee its stabiJity in the place upon which it rests. Rstribo Jongitlldinal ] Unión 2 I). de ti rant<:'8 1 Comedor de familia &tribo transversall Uniiml J:LCffi;;tribo longitlldínal a Figure 1 Location of the estribos and the tirantes in the framework C. Cañas The Mirador framework does not have elements such as nudillos or solera. The estribos, which are basic elements of structural stability, rest directly atop the walls of the room, without the minimum presence of some other element in the form of a layer of cement between the estribos and the wall, to somehow ensure the plane level of the receiving elemen!. We will differentiate between the transversal and longitudinal estribos, since there have been substantial differences detected between them in their geometric, supportive and conservational characteristics. The former are shorter and of greater edge than their perpendicular counterparts, and they are partially embedded into the wall, which makes it impossible to know the geometric characteristic of its total thickness; whereas the longitudinal estribos sit directly atop the crowning of the wall. The longitudinal estribos are formed by two others, united in a lengthening to approximately the point of junction of the braces (tirantes). The length of the room (14 metres) is too long to support one sole element in the crowning of the wall. If we consult the «Ordenanzas de Sevilla» we see how long the different beams of wood had to be as a general rule, and these oscillate between the 25 feet of the viga de acarro (7,5 m.) and the 12 feet of the medio ponton (3,60 m.). So the Iongest beam of the Mirador's framework (9,13 m.) is quite a bit longer than the vigas de acarro specified in the «Ordenanzas»; and the rest, 4,9 and 5,2 metres, are located between the media viga (4,2-4,5 m.) and the terciada (5,32-5,70 m.) specified in the same te x!. Estriho longitudill(! 1 2 Dormitorio de Don Pedro T:niónH ['nÍón 4 17nión 3 P. t'nión 6 Estribo longitudinal 4 The constructive techniques al' the moorish roafing framewarks In the corners, the estribos belonging to different roof slopes formalise their union halfway up the wood, having a small notch which enables both pieces to not get displaced by the effect of the traction of the rest of the structures; nevertheless the difference of the section of the estribos makes it not possible for the longitudinal elements to have a cut halfway up the section. The junction of the estribos in their prolongation does not work as such. The elements which come together on them work independently. Another important element in this section which carne into being basically for structural necessity, is the tirante (brace). The size of the room makes these elements absolutely necessary. When the relation between the height and the width of the room is reduced, there is a danger of collapse in the crowning of the walls, since the horizontal pressure upon the estribos is very pronounced. The tirante collaborates in the absorption of part of the pressure and therefore gives equilibrium to the structure. A total of two pairs of tirantes have been provided in the length of the structure upon which the estribos sil. The perimetral wall of the room, upon which the framework sits, is crowned by the estribo, as we have already described, which rests directly upon it; I~ Figure 2 Embedding al' the tirante inta the wall and the unian with the longitudinal estribo 515 therefore since there is neither solera nor canes, the tirantes have to be embedded into the walls. The embedding measures about 22 cm. If we stop to observe the junction with the estribo, we see that a 7 cm tall cajeado has been performed in the tirante, leaving a 2 cm tall espiga (see figure 2) to be placed into the estribo. We do not know exactly how much of this union is embedded into the wall; however constructive logic and the load that the tirantes have to absorb lead us to think that it must be as visualised in the representation. At this time it is necessary to pause and make the observation that this is one of the points which has brought up more questions and doubts in regards to the analysis of the origins of the tirantes, since most of the data analysed lead one to think that the ones we see now do not pertain to the original framework but to the reformations carried out in 1908 in which the need to replace four wooden tirantes, replacing the original s, was noted. «To reinforce the wooden bracing, put into place four new wooden braces instead 01' the defective anes and to zancar their alfardas». (from the «Budget Boak») The tirantes in a framework of a certain size or importance, have always been ideal elements for carrying out di verse decorative craftings, mainly because of their position, up front in the visual plane of the spectator. And, what is more, there has always been a decorative hypothesis about the use of double tirantes, since they made a good support for the ties of lacework (lacería) which covered the framework. It does not seem 10gical that in a framework of this kind, the finishing would be so rough and undelicate. All of this data, together with those derived from the embedding into the wall and the junctions with the rest of the elements, bring one to consider that we are seeing elements that date back to the beginning of the 20'h Century, new tirantes without any decoration (unlike the deteriorated originals which possibly had a tied crossover lacing effect, according to the standards of the epoch). This doubt gathers force when we analyse the pictorial decoration which covers the sides of the tirantes, since according to the criteria of the modern restoration experts, these belong to periods prior to the reformation of 1908; to be exact, they are similar to others located in the frieze of the framework which 516 date back 10 the 16th century. Let us leave our «doubt» open then to the posterior development 01' the research being done in this field. Finally, we will describe the part 01'the framework setting that is not hidden, which has the assigned job precisely to leave «unseen»behind them, each 01' the previously mentioned structural elements, and to serve as a pictorial support and visual transition between the wall and the framework 01' the roofing/ceiling. THE VISIBLE ELEMENTS The superficial elements 01' the arrocabe 01' the Mirador's framework are formed by a double alicer, in which the first plank is nailed into a mold or tocadura, which is also nailed onto a type 01' wooden bracket which becomes the perimetral zuncho at the sides 01' the room. This first plank, or lower alicer, is finished at the top with another one which serves as the mold or tocadura for the second one, right in the bottom coat 01' the braceo Behind this second alicer is hidden the junction between the tirante and the estribos with the wall, as well as the bottom part 01' the pares where they meet the estribo. The former estribo merely hides the crowning 01' the pcrimetral wall. Tirante Figure 3 Section of the arrocabe C. Cañas Figure 4 Union at the comer of the bottom parts of the planks (aliceres) THE INCLINED ROOF SLOPES Before beginning an analysis 01' each 01' the constituting elements of the roof slopes, it is necessary to make mention of the set-square of the framework used (the inclination of the alfardas). Measuring directly upon the alfardas we obtain Par Alicer superior Refuerzos de lIUldera 'rocadurn Alicer inferior TO<'Mum Can The constructive techniques of the moorish roofing frameworks 517 figures that range from 35,8 to 42.67 degrees. This fact confirms that the deformation of the structure ís significant, since the theoretical set -square of the framework is at 37,6 degrees. Structural composition In the first place, as definitory elements of the inclined roof slopes, we must analyse the pares or altardas and the péndolas or manguetas which necessarily appear in the frameworks that have limas. There are a total of six pares in the transversal roof slopes and forty-four in the longitudínal ones, plus a total of five manguetas which accede to each of the limas that pertain to the faldones. All of the pares are <<iive» since they form part of the lacework of the framework. Since this is a framework of mohamar lima, each of the faldones has two limas at its laterallimit, with two from each of the contiguous faldones coming together at the corners. Let us recall one of the fundamental rules for the procedure of making this type of structures upon whích posteriorly tied lacework and stars wilJ be crafted; the rule denominated by López de Arenas the <<Law of calle y cuerda «(aisle and cord).3 In the case of the Mirador, the planning of the ties is precise, nevertheless the thickness of both pares and nudilJos is of 7,5 cm, whereas the width of the aisles vary: in the testeros of the aisles they measure from 16 to 16,5 cm, and in the rest of the framework they measure from 14,75 to 17,5 cm (the majority being between 16 to 16,5 cm). This statístic reveals to us that the differences measured are due to the movement experienced by the framework. It is produced more intensely by the longitudinal faldones, not so in the testeros, obviously the deformation increases in the extent that the length also increases. We can affirm that the real separation is close to 16 centimetres. We are speaking, then, of margins of error of around 0,25 and 0,5 cm, which in wood are not that significant, and even les s so given the procedures used in that epoch. This difference can be due to several different factors : on the one hand, the movements that the structure has had to support in general have been important, whereby the spaces between the alfardas could have grown (as we wilJ later see, many of the pieces have become unjointed) and also, the passage of time can cause minor losses of parts of the wood. All this together with the assumption of possible smaJ] errors in the procedures of its construction make us maintain the validity of the aforementioned rule, to a certain extent at least. In any case, we can find no explanation that justifies making the aísle Figure 5 Location and cut of the pares and the manguetas 518 (calle) 0,5 to 1,0 cm wider, measurements that are insignificant. After having made this important clarification, we now will discuss how the different junctions between the aforementioned elements are formalised. The junctions in the estribos of the pares and manguetas are formalised from the patillas to the barbillas, with the perpendicular cut done on the horizontal plane. Usually in order to fix the element securely onto the estribo this type of union was reinforced by nails, however in this case, no reinforcement has been has been detected whatsoever. The patilla measures 3,5 cm and the barbilla 10 cm. The cut of the barbilla is done at approximately 1/3 of the height of the par, conforming to what was the standard of carpentry. The limas have the same kind of union, however their state of conservation does not permit us to precise their geometrical characteristics. It is also usual to find in this kind of structure that the pares culminate in the hilera which closes or culminates the framework at the topo The original roofing of the Mirador did not have an hilera and the pares were united by means of ensemble of garganta and quijera. This type of ensemble diminishes the resistent section of the pares at this point, therefore it is usually reinforced by means of some metallic element. But the junctions analysed lack any type of C. Cañas reinforcement, thereby they are in a very weakened condition. On the other hand, the manguetas reach the limas at a cut perfectly parallel to the contact surface,and are joined by nails to the main structure. The limas are joined to the pares by the same means. The rest of the structural elements pertaining to the faldones are the peinazos and the arrocabas,410cated in respect to the pares and limas as seen in figure 6. We will differentiate between the orthogonal peinazos and the inclined ones. The former are the ones that are located perpindicular to the pares. The junction with the pares is produced almost certainly by means of «a romo y agudo» with an espiga or wooden nail. Although this method has not been officially contirmed, in some ofthe pares the boxed space formed for the insertion of the espiga from its corresponding peinazo has been observed. The inclined peinazos are fixed both to the pares and to each other by nails. The junction of the arrocabas and limas is also done this way. Non-structural Elements In Figure 7 each and every element pertaining to the inclined roof slopes is shown, both structural and non-structural, which has the job of conforming definitvely the decorative layout of the framcwork. Figure 6 Situation and cut of the peinazos and arrocabas The constructive techniques of the moorish roofing frameworks 519 Figure 7 In thc upper part the resistent structure and its respective unions are indicated, as well as the taujeles and the bases where the supports, wedges, and surface boards all appear. In the lower par! of the figure, the inclined roof slope is shown with all of its elements in position, including the «tablazón de relleno» The tauje]es form part of the decorative layout of the lacework ties of the framework, without belonging to the structura] plot. They are panels of ],5 cm thickness which complete the layout, usually nailed to a tablazón superficial, which appear at each of the aisles contiguous to the eight-pointed stars. We must mention the union of these pane]s into the structure as a special characteristic. The panel has the same thickness as the rebaje (diminishing) done on the pares and the peinazos, coinciding with four protrusiones carried out upon the panel, so that they coincide and fit together perfectly. In the pares the protrusions are trapezoidal in shape, thereby increasing the supportive base. There are also three distinct types of wedges that are nai]ed to the structura] elements to fulfill basically a decorative mission. A first group forms part of the layout of the eight-pointed stars of the framework. Those of a second group serve as a base to support some of the taujeles, and a third group of wedges are placed in the upper part of the roof slopes, nailed to the peinazos, making the visual bypass from the inclined roof slopes and the almizate, in such a way that the eight-pointed stars located in these points, and the other decorative elements that are necessarily broken, all have their faces parallel. Instead of making the ang]e by doing a direct cut into the corresponding peinazos. this type of wedge is just added on. There is another type of panel that directly forms part of the decorative figures of the framework called the tablazón de relleno (filler panel). There are areas on the faldones that require the location of these panels. which hide behind them part of the structure, which otherwise would be seen. These panel s are an important part of the decoration of the framework since they are situated in the p]ane that is nearest in sight (on the same p]ane as the main pieces), and sometimes occupy a great part of the surface serving as a base for perfected pictorial decoration. And lastly, we have the tablazón del trasdós. ]t is part of the fian] covering of the framework and its presence is taken advantage of to give a pictorial background to the framework. The panel s are nailed to the resistent structure of the framework, pares and limas. and cover practically the entire exterior surface ofit. THE PLANS OF THE NUDILLOS. THE ALMIZATE Structural composition In the first place, we shall locate the definitive elements of these plan s, which are the nudillos. We must mark a distinction between two different types: the long nudillos, which are devloped along the almizate (types A,B,D and F), and the short nudillos, which are interrupted by different peinazos, which are necessary for the configuration of crossing and tieing ofthe lacing (typcs e and G). AII ofthese are situated below the terciary. 520 C. Cañas .\ B(; E FB e[) eE Figure 8 Types of nudillos of the Mirador The nudillos are joined to the pares by «a garganta y quijada», following the standards and mIes of carpentry. «Echando cabeza de armadura» in the alfarda we make the corresponding cut in the garganta. The cornezuelos of the nudillos are of approximately onefourth the thickness of this element. The type of junction between the longitudinal nudillos and peinazos is unknown, although we would dare to say that the most probable resolution is the one that traditionally has been used to unite peinazos and nudillos; that is, the joining «a romo y agudo» with the wooden nail, or espiga. Next we will analyse the peinazos, which usually are located and joined perpendicular to the nudillos, nF E EeBB(' E E F u thereby contributing in an important way to the stability of the almizate. he peinazos and the nudillos together transmit the pressure to the alfardas. There are two fundamental groups of peinazos in the almizate. On the one hand, those that are situated in an orthogonal way to the nudillos (types a, b, c, d, e, f and g), and those that are placed in an oblique way, which contribute to giving a greater stability to the entire set, while at the same time serve as a base for the layout of the design of the lacework ties (types i, j, and k). In figure 9 all the peinazos that make up the almizate are determined. We have mentioned basically two groups of peinazos, and of these we only know for surre about AB D FGD eEE e Figure 9 Types of peinazos that form part of the al mizate oEe]) eE l<~ F G F E D The constructive techniques of the moorish roofing frameworks 521 the unions of the types «i» and <~» to the nudillos. These materialise by means of clavazón to the resistent structure. The rest of the unions are done through different types of junctions, of which none has been accurately confirmed. The types «a, b, e» and «f» are almost certain to be the junction of «a romo y agudo con espiga», whereas the types «d» and «h» would have this type of union at one end, while the opposite would ha ve a «caja y espiga», with a boxed area made in the panel of the nudillo. The junctions between the type «k» peinazos and the masters (which will be either peinazos or nudillos, depending on the position that they occupy in the almizate), have not been discovered, neither do we dare to suggest a possible form of this union. The oblique placement of the wooden pieces and the great number of elements that all come together in the same area make any supposition about the junction vain. In any case, we can atfirm that for the union of the main elements that make up this pair of peinazos, a boxed area would have been hollowed out in the panel of the peinazo or nudillo and in it would have been inserted some kind of wooden peg (espiga), that will have to remain unspecified since it is hidden inside the boxed area. This peg could adopt any of many forms within the boxed area, so we will leave this matter open to the further developments of this investigation. Finally remains to be discussed the solution adopted in the union between the peinazos. Except for the union of one of the components of the pair of peinazos of type «k» that rests on the resistent masters to open the mocárabe cube, by means of chiseling the peinazo, the rest make their unions by means of clavazón. In figure 10 all the structural elements of the almizate are shown and the respective unions between them. Except for the junctions that are nailed and those ofthe entire element with the pares, the rest are all theoretical, due to the impossibility of confirming these facts, save disassembling the structure. The junctions between the peinazos that form the central octogon and the masters are the only ones that have not been defined. Non-structural elements It is logical that in the almizate the same elements that made Up the inclined roof slopes are repeated, so we Figure 10 The structural composition of the almizate and the relation between each of its elements to the others will not go into this part more profoundly than we need to. Again three types of wedges appear, and the different types of tablazones, of trasdós, of relleno, and superficial. In this case we will indicate the exact situation of each of the taujeles that make up part of the almizate in relation to the general structure. The same procedure is used to make the taujeles as to make the tie is the nudillos and peinazos, with the difference that in this case the little panels are cut to 1,5 cm thickness and in the former , diminishment of the same measurement has been made where part of these pieces have to fit in. To make the different cuts, a series of set -squares is needed, used as the only plotting instrument by the carpenters «de lo blanco». These set-squares are defined by the wheel of ties or crossover lacing that they are going to plot as the design of decoration for the framework. In this case, the wheels are of eight,