Frontiers of Architectural Research (2016) 5, 265–275

Available online at www.sciencedirect.com

www.elsevier.com/locate/foar

Frontiers of Architectural Research

http://dx.doi.2095-2635/& 2(http://creati

nCorrespondE-mail add

andygpb@liverPeer review

RESEARCH ARTICLE

Digital re-analysis of lost architectureand the particular case of Lutyens' LiverpoolMetropolitan Cathedral

Nicholas Webba,n, André Browna

aSchool of Architecture, University of Liverpool, Liverpool L69 7ZN, United Kingdom

Received 28 August 2015; received in revised form 5 January 2016; accepted 20 January 2016

KEYWORDSUnbuilt architecture;Digital critique;Forensic analysis;Digital modelling

org/10.1016/j.foar.2016016 The Authors. Produvecommons.org/licenses

ing author. Tel.: +44 15resses: njwebb@liverpoopool.ac.uk (A. Brown).under responsibility of

AbstractResearch and critique of unbuilt or destroyed works of architecture is traditionally carried outthrough the examination of surviving information such as drawings, models, photographs,biographies and monographs. The case study presented here demonstrates that this approachcannot always give a full-picture of the architect or designer's intentions, and may missinconsistencies in the design and links to other precedents or antecedents in such schemes.Here, we employ strategic contemporary digital representation techniques to re-present andre-analyse the evidence available for a particular architectural project. We describe asystematic methodology, which shows that these techniques can challenge or enhance currentunderstanding. The focus therefore is on the digital re-analysis process that has been devised.Sir Edwin Lutyens' unbuilt Liverpool Metropolitan Cathedral design, that would have deliveredone of the largest cathedrals in the world, is used as a case study. The findings reveal newinformation about the cathedral by following structured lines of enquiry generated from thestudy of primary and secondary source data, as well as serendipitous results that occur as apotential by-product of the methodological process.& 2016 The Authors. Production and hosting by Elsevier B.V. This is an open access article underthe CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

.01.004ction and hosting by Elsevier B.V./by-nc-nd/4.0/).

1 794 2483.l.ac.uk (N. Webb),

Southeast University.

1. Introduction

Mediating devices are essential tools to describe architec-tural designs and traditionally these have been physicallybased, for example scale drawings and physical models. Suchdevices have been extended through the digital realm withthe widespread adoption of computer aided drafting anddesign. Beyond their use as tools for architects to construct

This is an open access article under the CC BY-NC-ND license

N. Webb, A. Brown266

graphic and textual descriptions of schemes that are yet tobe built, digital representation techniques have beenexploited to visualise damaged, destroyed and unbuilt worksof architecture (Forte and Siliotti, 1997; Novitski, 1998).Such investigations have become significant in producingenhanced understanding of what these buildings and designswould have looked like. However, debate continues into theuse of digital tools to preserve architectural and culturalheritage, as it is potentially biased on the interpretation ofthe creator of the digital representation (Kalay, 2007). Morerecently, such research has begun to focus on the capabilitiesthat digital techniques can provide as analysis tools, ratherthan focusing primarily on the representations created(Brown and Webb, 2010; Mark, 2011). The research pre-sented here extends these previous studies as it specificallylooks at the process of digitally creating an architecturaldesign and the re-analysis this can provide. Also, it exploitsthe possibilities that become available by utilising digitalsoftware to analyse the resulting representations.

Many architectural designs are not built. Also, works ofarchitecture may have been constructed and subsequentlydamaged or destroyed. In these cases, representation docu-ments may still remain, offering an insight into what couldhave been or once was. Such unbuilt works are increasinglyacknowledged for their contribution to cultural heritage;Wilson, for instance, suggests that ‘the built environment weinhabit is just the residue of a much greater imaginativeworld that never saw the light of day, evoking what mighthave been or still could be – the unbuilt, the lost’ (Wilson,2004). This paper focuses on Lutyens' unbuilt design for theLiverpool Metropolitan Cathedral, and the possibilities thatdigital techniques can bring in enhancing our understandingof it. The methodology proposed can also be utilised forexisting, damaged and destroyed works of architecture;however, the case study discussed here is an unbuilt design.

Fig. 1 Methodology developed to re-analyse damaged,destroyed or unbuilt works of architecture using digitaltechniques.

2. Methodology

The nature of unbuilt, damaged or destroyed works ofarchitecture results in the information available for inves-tigation almost always being incomplete. This can be seen indigital reconstructions such as the destroyed synagogues ofVienna project (Martens and Peter, 2002). Therefore inter-pretation of material requires parallel study into thearchitect or designer, their influences and the contemporarycontext they operated within. This research can then beused to make inferences in order to fill gaps in an informedway. It is therefore crucial to make clear the interpretivenature of such decisions, as any representations createdhave the potential to mislead the viewer. It is importantthen to re-iterate that the resulting representations aresecondary in terms of the research goals. Consequently, thefocus of this study firstly places emphasis on the reading ofsource data and how it can be utilised to pose specificquestions about architectural designs in which knowledgecan be enhanced using digital techniques that would nothave been available for the designers to make use of at thetime and secondly, emphasis is placed on the process ofconstructing the digital representations and what can belearned from this.

The methodology displayed in Fig. 1 represents theprocess of selecting a case study, researching it, generatinglines of enquiry, creating appropriate digital representa-tions, investigating the lines of enquiry using the represen-tations created, and finally analysing and reviewing themagainst identified gaps in knowledge and questions posed inthe lines of enquiry. This methodology is demonstrated inthe following sections.

3. Lutyens' Liverpool Metropolitan Cathedral

After a stalled attempt to construct a catholic cathedral inLiverpool in the mid nineteenth century the idea was raisedagain in the 1920s with the Archbishop, Dr. Downey,becoming a key figure in progressing efforts to build acathedral. Downey was introduced to Sir Edwin Lutyens inLondon in 1929, in which a rough sketch was drawn andLutyens was consequently confirmed as the architect.

Lutyens was a leading English architect of the latenineteenth and early twentieth century with architecturalhistorians going so far as stating he was ‘the greatest artistin building that this country has produced’ (Hussey, 1984).He initially gained attention designing country houses in the‘Surrey Vernacular’ style followed by Neo-Georgian houses.Around the time of the First World War his work developedbeyond housing into more abstract forms such as war gravesand monuments, the most well-known being the Cenotaphin London (1919–1920), a design that was replicated all overthe world. Another important phase of his career was hiswork in New Delhi, particularly the All India Gate andRashtrapati Bhavan. His influence there was so great thatthe area is known to this day as Lutyens' Delhi.

In 1929 Lutyens began the Liverpool Metropolitan Cathe-dral design which would arguably have been his mostsignificant building had it been completed. It is of particularinterest in his body of work as it overlaps the periods inwhich he was designing projects in New Delhi as well as warmemorials such as Thiepval in France; both of which can beseen in elements of the cathedral design (Butler, 1984).Although Lutyens worked on other projects after the initia-tion of the scheme, the cathedral became his primary focusfor the rest of his life.

Fig. 2 Gibberd's built design on top of the Lutyens crypt (left). Renovated physical model of Lutyens' unbuilt Liverpool MetropolitanCathedral (right).

267Digital re-analysis of lost architecture and the particular case of Lutyens' Liverpool Metropolitan Cathedral

The Neo-Classical cathedral design was vast in scale,second only to St Peter's Basilica in Rome, formed ofinterlocking arches and topped with a dome. Constructionbegan in 1933 but was abandoned in 1941 as wartimerestrictions resulted in a lack of labour and materials.Post-war restrictions halted building work further, and afterboth the architect and Archbishop who commissioned thebuilding died, the decision was taken to complete the cryptonly. In 1959 an architectural competition was announcedlooking for a new design incorporating the crypt; which waswon by Sir Frederick Gibberd and was subsequently built onthe site (Fig. 2). Although the crypt of Lutyens' cathedralwas constructed and therefore the scheme could be classedas partially built, it is referred to as unbuilt in this study asthe majority the design was not completed.

The design was chosen for this research as it holdssignificance in the history of Liverpool, which can be seenin its discussion as a major work of architecture alongsidebuilt designs in the city such as the Royal Liver Building andLiverpool Anglican Cathedral (Sharples, 2007). Its impor-tance as a heritage asset is also apparent as the physicalscale model of the design, completed in 1934, forms acentral exhibit in the Museum of Liverpool (Fig. 2). Further-more, monographs and biographies of Lutyens' career assignparticular importance to the design as part of his body ofwork. Summerson (1981) stated that ‘the question whethera building can assume a place of authority in the world ofarchitecture without actually being built is a curious one;but the answer is not in doubt… It will survive as anarchitectural creation of the highest order’.

In addition to the physical model and secondary sourcedata already discussed, a wealth of archival informationrelating to Lutyens' design still exists including architecturaldrawings, perspective images, newspaper articles and cor-respondence between various parties. Crucially for themethodological process proposed, these sources offer thepotential to generate several lines of enquiry to augmentcritique of the cathedral design using digital techniques.

4. Creating the digital representations

Once investigating the entire primary and secondary sourcedata available formed lines of enquiry, digital representations

could be created to answer these. This primarily took the formof three dimensional digital models of the cathedral, havingseveral iterations showing the development of the design from1929 onwards. They were constructed primarily using anextensive set of drawings available in the Lutyens Memorial(Butler, 1984); a series of books published following thearchitect's death that includes drawing plates, images, essaysand an extensive biography by Hussey (1984). The memorialset forms the most complete account of the design in terms ofboth quality and quantity of drawings of the 1934 design. Keychanges between the initial design and final design were notedduring the study of source data, and then modelled in reversechronological order back to the initial 1929 design (Fig. 3).This begins to highlight the advantages of using digitaltechniques, as the process of creating multiple versions ofthe same design is relatively quick in comparison to physicalmethods such as models, sketches and drawings. Clarifying thedesign process through multiple models also addresses cur-atorial issues, for example, the Liverpool Metropolitan Cathe-dral website displays Cecil Farey's render of the 1930scheme (Liverpool Metropolitan Cathedral, 2015), which hasa significantly different dome and buttress appearance com-pared to the final 1934 scheme and could therefore lead toconfusion regarding the design (Fig. 3). The digital investiga-tion enables the different iterations to be clearly shown.

SketchUp was chosen to model the cathedral based on itssimple interface with tools such as ‘push/pull’ and ‘followme’. These are particularly suited for the cathedral designas it is primarily formed of simple geometry. The softwarealso uses plug-ins to enhance its capabilities, such as raytracing and solid modelling. A disadvantage of SketchUp isits poor capabilities when working with freeform geometryor double curved surfaces, however, these shapes occurlittle in the cathedral design. The only aspect of the designfeaturing complex freeform shapes were the statues placedwithin the design; objects that do not have a high-impact onthe lines of enquiry pursued, and are therefore not mod-elled. The straightforward nature of the software meant thedesign could be produced quickly and efficiently given thevast size of the cathedral; a process that would still takemore than twelve weeks.

Materiality formed a key aspect of the study; particularly fora line of enquiry investigating the interior lighting of thecathedral. The process of applying materials to the model was

Fig. 3 Digital models representing the development of the design. 1929: smaller stainless steel dome with bell towers locatedabove the porches. 1930: stainless steel dome made larger with bell towers moved above the lateral transepts. 1934: domeredesigned and changed to granite, buttressing and bell towers significantly redesigned.

Fig. 4 The Thiepval Memorial to the Missing of the Somme (left) was a key reference point informing the materiality of thecathedral digital model. An ortho-rectified photograph of the ‘pinky-brown’ Roman bricks at Thiepval was used as a material swatch(right).

N. Webb, A. Brown268

relatively simple as Lutyens specifies the type of brick usedwould be a ‘pinky-brown’ Roman design, of which he alsospecifies at the Thiepval Memorial to the Missing of theSomme. Therefore, photographs were taken of the brick atThiepval, ortho-rectified digitally and then added to the digitalcathedral model as a texture (Fig. 4). It must be noted that thebricks used for the cathedral may have differed if the designhad been built; however, the description of the proposedbricks matches those used at Thiepval closely, especiallyconsidering it was designed by Lutyens at a similar time tothe cathedral. This procedure of deduction was repeated forthe granite aspects of the design and was a relatively simpleprocess as the material had already been used for the crypt;the only built part of the cathedral, which therefore offereddirect primary evidence in terms of materiality.

The previous paragraph demonstrates that in constructinga digital representation of an unbuilt design, it is highlylikely that source material used as a reference for itsconstruction will be incomplete. This can lead to ‘infer-ences, educated guesses, and just plain wild speculation’making their way into a digital representation in order to fillin these missing elements (Kensek, 2005). Although thecathedral model required some inferences to be made,ambiguity was largely avoided due to the Lutyens Memorialproviding a near complete set of architectural drawings,with additional sources filling in many of the gaps. Thecathedral design is systematically laid out; granite footingson top of which brick piers sit, followed by another band of

granite, then another band of brick and finally the vaultedarch granite-clad ceilings in the higher levels of the build-ing. This pattern, including the cornice and architravedetails, repeats throughout the design so a general rulecould be deduced that any unknown elements should adoptthese basic principles without any additional detail beingadded based on assumptions and analogies made. Thisavoids the situation of relying on educated guesses whilstbeing confident that the basic design elements are repre-sented as correctly as possible.

As a point of reference, a colour-coded version of thedigital model was created suggesting various levels of con-fidence based on the source data available. This enables theviewer to gain a clearer view of exactly what they are seeingand acts as an additional reference point when trying toenhance understanding of the cathedral design. For example,the west entrance of the cathedral required several infer-ences to be made due to a lack of source data, and thereforeit was colour coded red in the reference model, compared tothe nave of the design which had sufficient informationavailable which was consequently coloured green.

5. Lines of enquiry investigated

Upon researching the source information, specific lines ofenquiry emerged. These were formulated by ascertainingwhether digital techniques could be used to enhance

269Digital re-analysis of lost architecture and the particular case of Lutyens' Liverpool Metropolitan Cathedral

knowledge and explanation of particular aspects of thedesign. Of several lines of enquiry followed, three will bediscussed in detail here; the internal geometry of thecathedral, the lighting of the cathedral and comparisonsto Thiepval Memorial to the Missing of the Somme.

5.1. Internal geometry of the cathedral

Summerson (1981) wrote a key article on the cathedraldesign, describing its interlocking arches geometry in detail.He acknowledges that ‘to set all this up in one's mind as areal-life experience is far from easy’. What is particularlyrevealing about the text is he discusses the arches as aseries of tunnels, which gives the impression they arecarved from a solid mass giving the cathedral a cavernousfeel. This analogy suggests that the geometry of the designcould be further explained with Boolean operations as usedin solid digital modelling. This process works by taking anumber of solid models and applying operations to them tochange their form. For example, a sphere and a cube couldbe joined into one object using a Boolean union, or a tunnelcould be formed by removing the volume of a cylinder froma square using a Boolean subtraction. Further to the Booleanoperations analogy, Butler (1984) stated that the cathedraldesign should be thought out in solid geometrical relation-ships using a model ‘which could be taken to pieces andbuilt up again’. This seems to pre-empt the use of digitaltechniques to enhance understanding of the design.

Digital representations are used to precisely explain howthe arches are seemingly carved from a large mass to form‘tunnels’ within the design. Each of the different size archesdescribed by Lutyens are modelled digitally, colour coded andpositioned in relation to the plan of the cathedral. Thisenables a build-up of the definitive elements of the cathedraldesign from the interlocking arches to the dome (Fig. 5).

By way of a subtraction, these solid elements are then takenaway from a basic massing model of the cathedral design togive a simple geometric model of the interior. The modelexplains the basic geometric principles of the design and showsgraphically how it would have looked in its most primitive form

Fig. 5 Digital solid modelling is used to describe the internal geom4), followed by the dome (5). The geometry is then subtracted fro

along the length of the nave (Fig. 6). Digital techniques incomparison to manual techniques are far quicker to produceand analyse, for example you can easily edit the digital modelsvia Boolean operations as discussed, and analyse the resultsthrough a review of the solid models produced as well as takingdifferent viewpoints within the design. The digital model couldthen be developed to a higher level of detail to assist withadditional lines of enquiry into the design.

5.2. Lighting of the cathedral

The previous section demonstrates how the design can beseen as interior spaces carved from a solid, giving thecathedral a cavernous feel. Summerson (1981) discusses indetail how natural light would affect the interior, statingthat openings in the domed space would be the main sourceof light, with small amounts of light also coming fromopenings high in the nave and narthex. He also states thatthis lack of illumination would result in the interior beingevocative and mysterious, especially in comparison to StPaul's Cathedral in London.

Existing sources of information do not give a clearindication of how the interior would have been lit. This isdue to the representation methods available at the time; aseries of pencil sketches by Lutyens indicate the geometryof the space but do not demonstrate the lighting. Theclosest indication of how the interior may have been litcomes from an endoscopic film of the physical model of thedesign (Mezzo Films, 2007) however, the film shows that themodel lets in light through the saucer domes along theceiling of the nave, which does not occur in Lutyens' design,therefore the lighting displayed in the interior does notaccurately visualise what would have been.

Therefore, the digitally created model can be used tomore accurately simulate the lighting predictions with raytracing. As a starting point, a view was taken from the mainentrance porch along the nave towards the dome, then raytraced to get a near photo real image of what the scenewould have looked like. This meant that materials had to beapplied to the model as discussed in section four.

etry of the cathedral from the smallest to the largest arches (1–m a massing model of the cathedral to form the interior (6).

N. Webb, A. Brown270

The interior natural lighting conditions would haverequired supplementary lighting, an issue which was a pointof debate between Lutyens and the cathedral committee,with Lutyens stating; ‘I want my Cathedral to be lit entirelyby candles. You need wondrous few. The big Nave at St.John's College is lit by four candles and isn't it glorious andmysterious! The choir is alone well in that every choristerhas a candle. But they want electric light, and flood-lightingat that’ (Hussey, 1984).

The interior lighting is simulated to show the differencebetween what the design would have looked like in bothscenarios; candlelit and floodlit. Kerkythea, the renderingsoftware used to create the images, has an extensionenabling lighting to be placed directly in the SketchUpmodel, then exported back to Kerkythea and finally ray

Fig. 6 By subtracting the geometry produced in Fig. 5 from amassing model of the cathedral, the simplified geometry of theinternal space can be visualised.

Fig. 7 Ray traced view of the cathedral interior along thenave with natural lighting and candle lighting.

traced. Firstly, candlelight was placed throughout themodel, followed by a repeated render using floodlighting.For the candlelit render, point lights with a Kerkythea lightpower value of 0.32, which equates to 12.5 lm, were placedthroughout the model. For the floodlit render the processwas repeated using a light power equating to floodlighting(1200 lm) with the lights placed on the shelf createdbetween the bases of the piers and brickwork above. Thisseemed like an obvious area to place the lights due to itsease of access for maintenance purposes had the designbeen built, as well as the shelf hiding the lighting unit itself,therefore being advantageous aesthetically.

The ray traced renders of the two alternative lightingconditions contrast each other greatly (Figs. 7 and 8). Theimages indicate that Lutyens' preference for a candlelitinterior would have been evocative and mysterious, espe-cially as the upper levels of the design have very lowlighting levels. In contrast, the floodlit version makes thearchitecture of the cathedral interior very obvious but losessome of the mystery of the candlelit render. Crucially, thedigitally created images provide visualisations of how thecathedral could have been lit in a much more accurate waythan has been achieved in previous representations of thedesign, such as hand drawn perspectives and the physicalmodel. However, although the digitally created viewsprovide more accurate representations, they are still sub-jective in the sense that it is the decision of the viewerwhich lighting scenario they prefer. Further studies wereinitiated to ascertain a more objective investigation usingAutodesk Ecotect Analysis, enabling illuminance levels ofthe cathedral interior to be simulated. A simplified digitalmodel was produced as the software and hardware used wasnot sufficient to deal with the large file size the detaileddigital model produced. An analysis grid placed on thecathedral nave floor demonstrated both the candlelit(2 lx) and floodlit (41 lx) illuminance levels would havebeen well below the now current required level of 100 lxfor reading a hymn book (Chartered Institute of BuildingService Engineers, 2003). This supports the cathedral com-mittee's request for floodlighting, as it clearly shows howmuch higher the lux levels would have been. Using Ecotect

Fig. 8 Ray traced view of the cathedral interior along thenave with natural lighting and flood lighting.

271Digital re-analysis of lost architecture and the particular case of Lutyens' Liverpool Metropolitan Cathedral

was a pilot study that requires further investigation toproduce detailed results. However, it does demonstratethe advantages of using such software, and how the resultscould have informed lighting decisions based on the con-flicting requests of Lutyens and the cathedral authorities.

5.3. Comparisons to Thiepval Memorial to theMissing of the Somme

Lutyens' Memorial to the Missing of the Somme at Thiepval,France, has often been compared to the Liverpool Metro-politan Cathedral design in terms of its geometry and

Fig. 10 Digital line drawing of the cathedral (left) taken from th(32 ft) wide arch.

Fig. 9 Digital overlay showing a section of the cathedral navecompared to the Thiepval Memorial at the same scale. Theimage demonstrates how the arches of the two designs inter-sect at the same point.

materiality; it includes the same barrel arches that Lutyensdescribes in the design for the cathedral (Fig. 4). The maindifference between the designs is the arches at Thiepvalhave a width to height ratio of 2:5, whereas the cathedralhas a width to height ratio of 1:3 to give a loftier feel. Infact, Summerson (1981) states that the main arch ofThiepval is exactly the same width as the transept archesin the cathedral; overlaying the two designs digitally coulddemonstrate this effectively. The materiality of the twoschemes are also very similar; red brick and grey stone.Additionally, Thiepval was designed and constructedbetween 1927 and 1932, which directly overlaps the periodwhen Lutyens was commissioned to design the cathedral.Therefore it is very likely that his ideas in France influencedhis design in Liverpool.

Questioning the extent of similarities between the geo-metry of the cathedral in comparison to Thiepval can bedemonstrated using a series of comparative snapshots takenfrom digital representations of the two designs. Thisinvolved creating a digital model of Thiepval in order todirectly compare it to that already created of the cathedralmodel. The digital representation of Thiepval was producedusing a set of drawings published in the Lutyens Memorial(Butler, 1984), as well as visiting the monument itself whichenabled key features to be checked and material snapshotsto be taken. Once complete, the digital representationswere digitally overlaid at the same scale (Fig. 9).

The overlay shows how the arches connect at the samepoint along the nave transept and aisles. In addition to this,the geometry created in Section 5.1 to explain the caver-nous shapes of the cathedral interior are used to makeexplicit how the arches cross at exactly the same point.Fig. 9 clearly demonstrates the main difference betweenthe width to height ratios of the two designs; for example atThiepval the largest arch is 9.75 m wide by 24.38 m high(32 ft by 80 ft) whereas at the cathedral the equivalent archis 9.75 m wide by 29.26 m high (32 ft by 96 ft). Using digitaloverlays of the representations enables digital images to becaptured of both designs from the same point of perspec-tive; something that would be difficult to achieve if physicalversions of both schemes existed. Views of each designtaken from a human perspective in the same position arecompared in Fig. 10.

e same position as Thiepval (right) looking through the 9.75 m

Fig. 12 Assumed section cut (left) compared to digital section cuts taken (centre) revealed that the actual section cut (right) isstaggered.

Fig. 11 The Thiepval memorial model is overlaid on a half model of the cathedral five times; twice in the lateral transept, twice inthe nave and once in the sanctuary. This can then be mirrored on the other half of the cathedral design giving a total of tenrepetitions.

N. Webb, A. Brown272

273Digital re-analysis of lost architecture and the particular case of Lutyens' Liverpool Metropolitan Cathedral

The comparative images show the similarities in the basicgeometry of the two designs. On the other hand, the imagesalso reveal how different the two designs are in their levelof detail; Thiepval is simple and restrained in terms ofdecoration in comparison to the cathedral which has a high-level of detail especially in areas such as caps of columnsand spandrels. The evidence discussed here clearly supportsprevious critique noting the similarities between the twoschemes. The digital representations show that the naveaisles and transepts of the cathedral are in essence theThiepval design repeated a total of ten times; five on eitherside of the centre line along the nave (Fig. 11).

6. Serendipitous discoveries

The previous section reveals that constructing a digitalrepresentation is the chief method of answering lines ofenquiry using the methodology. This process of investigatingspecific questions using modelling as the primary techniquealso has the advantage of providing results that are unex-pected, and in addition to the lines of enquiry followedbased on the tectonic nature of constructing the models.Our previous work in this area has led us to be alert andopen to the potential for such unexpected outcomes. Thisfirst arose as an issue in the digital re-construction of theLords Court by Connell, Ward and Lucas (Brown, 2001), aswell as investigations into lost projects by Auguste Perret

Fig. 13 A section taken through the assumed cut demonstrates hocut rather than along the centre of it.

(Brown and Webb, 2010) and Sir James Stirling (Webb andBrown, 2011). These unexpected results are in part due tothe nature of architectural models, whether physical ordigital, which are required to co-ordinate the plan, sectionand elevation. When these drawings are viewed as separateentities, the margin for error increases.

During the process of constructing the digital model ofthe Lutyens cathedral design, it became apparent that anoriginal cross section drawn through the east of thesanctuary was not a straight section cut as assumed, butwas in fact staggered (Fig. 12). Two sections were takenthrough the digital model with straight cuts; the first alongthe assumed straight line (Fig. 13). This indicated that thecentre of the chapter house does not line up with the centreof the spires in the two adjacent chapels, as drawn onLutyens' original drawing and replicated in Fig. 12. Thesecond straight cut was taken through the centre of thechapter house, confirming the presence of the stagger(Fig. 14).

This lack of clarity over the precise location of thesection cut is significant as it resulted in an error in themore detailed drawings that followed. It is likely thatLutyens or one of his employees used the original crosssection as a basis for more detailed sections drawn throughthe chapels. On these more detailed drawings, which can befound in the Lutyens Memorial (Butler, 1984), a window isclearly shown at high level on either side of the room which,according to the original cross section, opens out to an

w the chapter house in the middle of the image sits behind the

N. Webb, A. Brown274

external area around the dome of the chapter house. Thesewindows can also be seen in the centre of Fig. 13. Theprocess of constructing the digital model revealed that thiswindow scenario shown on the detailed sections is incor-rect. This is because the windows that open out onto thedome of the chapter house intersect the line of thesanctuary wall; a fact that is difficult to ascertain fromthe drawings alone. This intersecting wall results in half ofthe window being obscured (Fig. 15).

It must be stated that this discrepancy would have beenfound had the cathedral been built. However, it provides aninsight into Lutyens' office and how they designed thespaces; starting from the overall design working to a greaterlevel of detail as the drawings progressed. This discrepancyre-iterates that working from two-dimensional drawingsalone can be misleading.

Fig. 14 A section cut through the centre of the chapter house doriginal section cut is staggered.

Fig. 15 The wall forming the back of the sanctuary intersects withbeing obscured by the wall. The un-obscured half of the window is hit.

7. Conclusion

The case study described here establishes a systematic andmethodological investigation into an unbuilt work of archi-tecture with great significance to the heritage of Liverpoolas well as Lutyens' architectural portfolio. When a state-ment or observation is made regarding such designs; digitaltechniques provide enhanced evidence to support theextent of whether or not previous research is still reliable.For example, the relationship between the cathedral andThiepval Memorial to the Missing of the Somme is clearlyrevealed using digital techniques to support the statementmade previously by Summerson (1981). Again, this showsthat the role of the research is to augment and enhanceunderstanding, rather than focusing on the resulting repre-sentations created.

emonstrates, in conjunction with the previous image, that the

the centre line of the chapels, resulting in half of the windowighlighted in red with the obscured half indicated to the left of

275Digital re-analysis of lost architecture and the particular case of Lutyens' Liverpool Metropolitan Cathedral

The research is also important as the methodology ofusing digital tools to detect inconsistencies in the construc-tion process of building such schemes is a procedure thattraditional architectural critique is unlikely to identify. Thiscan be seen in the serendipitous result that was found whilstconstructing the model, and revealed that the windowdesign of the sanctuary chapels would have needed arevision had the cathedral been built.

The methodological process discussed in section twoprovides a template for future studies into such works, aswell as demonstrating the enhanced knowledge that can begained when researching a particular case study of historicalor cultural importance using contemporary digital tools; inthis case providing new information on Lutyens' LiverpoolMetropolitan Cathedral design. It is also hoped that thestudy will highlight the importance of unbuilt, intangibleheritage assets, and their role alongside existing, damagedand destroyed works of architecture.

References

Brown, A., 2001. Architectural critique through digital scenariobuilding: augmenting architectural criticism and narrative. In:Proceedings of the Ninth International Conference on ComputerAided Design Futures. Eindhoven, pp. 697–709.

Brown, A., Webb, N., 2010. Examination of the designs by AugustePerret using digitally-enabled forensic techniques. Int. J. Archit.Comput. 8 (4), 537–555.

Butler, A., 1984. The Architecture of Sir Edwin Lutyens; Volume III:Town and Public Buildings: Memorials: The Metropolitan Cathe-dral, Reprint edn. Antique Collectors' Club Ltd. London.

Chartered Institute of Building Service Engineers, 2003. Code forLighting. Butterworth-Heinemann, Oxford.

Forte, M., Siliotti, A. (Eds.), 1997. Virtual Archaeology: GreatDiscoveries Brought to Life Through Virtual Reality. Thamesand Hudson Ltd., London.

Hussey, C., 1984. The Life of Sir Edwin Lutyens, Reprint edn.Antique Collectors' Club Ltd., London.

Kalay, Y., 2007. Introduction. In: Kalay, Y., Kvan, T., Affleck, J.(Eds.), New heritage: New Media and Cultural Heritage. Routle-dge, Oxon, pp. 1–10.

Kensek, K., 2005. Digital Reconstructions: Confidence and Ambi-guity. In: Proceedings of the 9th Iberoamerican Congress ofDigital Graphics. SIGraDi, pp. 640–644.

Liverpool Metropolitan Cathedral, 2015. Lutyens' Dream. ⟨http://www.liverpoolmetrocathedral.org.uk/the-second-cathedral/⟩.

Mark, E., 2011. Visualizing the Unknown in Historical VernacularArchitecture. In: Zupancic, T., Juvancic, M., Verovsek, S.,Jutraz, A., (Eds.), Respecting Fragile Places: 29th eCAADeConference Proceedings. eCAADe and UNI Ljubljana, pp. 868–874.

Martens, B., Peter, H., 2002. Developing systematics regardingvirtual reconstruction of synagogues. In: Proceedings of theThresholds-Design, Research, Education and Practice, in theSpace Between the Physical and the Virtual. ACADIA, 24–27October, pp. 349–356.

Mezzo Films, 2007. Inside Lutyens Model of Liverpool Cathedral inHD. ⟨http://www.veoh.com/watch/v341098p5fStbEh⟩.

Novitski, B., 1998. Rendering Real and Imagined Buildings. RockwellPublishers, Inc., Massachusetts.

Sharples, J., 2007. Pevsner Architectural Guides: Liverpool. YaleUniversity Press, London, 84–86.

Summerson, J., 1981. Arches of Triumph: the design for LiverpoolCathedral. In: Amery, C., Richardson, M., Stamp, G. (Eds.),Lutyens: The Work of the English Architect Sir Edwin Lutyens(1868–1944). Arts Council of Great Britain, London, pp. 45–52.

Webb, N., Brown, A., 2011. Digital forensics as a tool for augment-ing historical architectural analysis: Case study: the studentwork of James Stirling. In: Proceedings of the 16th InternationalConference on Computer Aided Architectural Design Research inAsia. Newcastle, Australia, pp. 505–514.

Wilson, R., 2004. Fighting the banalities of the built. In: Bingham,N., Carolin, C., Cook, P., Wilson, R. (Eds.), Fantasy Architecture1500-2036. Hayward Gallery and Royal Institute of BritishArchitects, London, pp. 18–23.