http://www.calameo.com/read/000202204155d7c8d7d38?authid=13Qij6QEiX1q

Un programmatore, Scott Kim, ha detto che “non c’è una differenza fondamentale fra programmare un computer e utilizzarlo”. Per lui si tratta di strati levigati e continui, dal microcodice incorporato nel firmware fino ai comandi di menu di un ATM, e il suo lavoro si colloca da qualche parte tra questi strati. (p.101)

Se queste sequenze venivano perforate su pacchi di schede, si poteva preparare un programma selezionando i pacchi opportuni, scrivendo e perforando su schede codici di collegamento e raggruppando i risultati in un nuovo pacco di schede. Questa attività venne chiamata “compilazione”. […] Grace Hopper svolse un ruolo cruciale nel trasferire quel concetto dal laboratorio di Howard Aiken ad Harvard al mondo commerciale. […] Per Hopper, un “compilatore” era “una routine generatrice di programmi, che produceva un programma specifico per un problema particolare” e chiamava “Programmazione Automatica” l’insieme delle attività legate all’utilizzo dei compilatori.

[…] I fautori della Programmazione Automatica si proponevano di sviluppare per il software quello che Henry Ford aveva concepito per la produzione delle automobili, un sistema basato su parti intercambiabili. Ma proprio perché il sistema di Ford funzionò al meglio quando venne impostato per produrre un solo modello di automobile, questi primi sistemi erano altrettanto rigidi, tentavano di standardizzare prematuramente e a un livello di astrazione sbagliato. Fa fu soltanto quando provarono a farlo che se ne resero conto.

Vedi Wilkes, Wheeler e Gill “The preparation of Programs fora n Electronic Digital Computer”, 26-37 e Martin Campbell-Kelly “Programming the EDSAC: Early Programming Activity at the University of Cambridge”, Annals of the History of Computing 2, 1980, 7-36.

Il successo iniziale di FORTRAN illustra con quale prontezza gli utenti accettarono un sistema che nascondeva I dettagli dei meccanismi interni della macchina, lasciandoli liberi di concentrarsi sulla risoluzione dei loro problemi e non di quelli della macchina. Nello stesso tempo, il fatto che si continui a usarlo ben addentro agli anni Novanta, in un’epoca nella quale sono disponibili linguaggi più nuovi, che nascondono molti altri strati di complessità, rivela i limiti di questa filosofia. Il linguaggio C, sviluppato presso i Bell Labs e uno dei più diffusi dopo il 1980, consente, come il FROTRAN, di accedere a funzionalità di basso libello quando si desidera farlo. I linguaggi per computer che hanno avuto successo e sono durati a lungo, ben pochi, sembrano avere in comune la caratteristica di nascondere ai programmatori alcuni dei meccanismi interni del computer, ma non tutti. (p.113)

ABSTRACT

This paper presents a critical study of the church of Longuelo, Italy, designed in 1966 by Giuseppe (Pino) Pizzigoni (1901–1967), an Italian architect who lived and worked in the city of Bergamo. He began his studies on shell structures in the Fifties and many of his buildings show outstanding skills in conceiving and handling complex structures. The church is one of his most interesting works: it is divided in four identical free parts, each composed by four shells joined by a fifth one, supported by twenty one bars which realize an statically-determinate spatial frame resulting in an outstanding inner space.

All during the XX Century, several Italian architects devoted their efforts in finding a synergy between structure and formal character, economically reasonable while rich of new architectural opportunities. Among the most notable there are personalities like Pier Luigi Nervi, Luigi Moretti and Sergio Musmeci, but also a lot of almost unheard-of architects, whose works are of utter interest. Here a very singular building of one of these designers, namely Pino Pizzigoni, is presented: the church of Longuelo. The main information related to this building are to be found directly in his archive, located in the public library of Bergamo “Angelo Mai [9], from his own sketches and written records, which often are all but clear and marked by consistency.
The second section describes the multifaceted personality of Pizzigoni, with a special attention to the last years of his life. In the third one we present an overview of the church to understand its genesis and evolution. The fourth section elaborates on the structural behaviour of the church: as the description by the architect are often confused, different suggestions and considerations about built structure are pieced together. Finally, before the conclusion section, the fifth one looks at the real church, its actual state and its aging.

Picture of the Church of Longuelo, front view.

Picture of the Church of Longuelo, back view.

Picture of the Church of Longuelo, lateral view 1.

Picture of the Church of Longuelo, lateral view 2.

Another experimentation with a spatial structure, Zandobbio.

Other experimentations.

This contribute paper will be published on the “Proceedings of the International Association for Shell and Spatial Structures (IASS) Symposium 2009″, Valencia. Evolution and Trends in Design, Analysis and Construction of Shell and Spatial Structures, 28 September – 2 October 2009, Universidad Politecnica de Valencia, Spain.

First of all, we have to transform a NURBS surface  in a MESH object in this way:

‘The number of elements in U and V direction

elemsNumU = 10

elemsNumV = 10

nodesNumU = elemsNumU + 1

nodesNumV = elemsNumV + 1

ReDim Nodes(nodesNumU * nodesNumV – 1)

Dim domU, domV, stepSizeU, stepSizeV

Rhino.AddLayer “Mesh”,  RGB(255, 0, 0), 0

Rhino.CurrentLayer(“Mesh”)

‘Get the domain of the surface

domU = Rhino.SurfaceDomain(surface(i), 0)

domV = Rhino.SurfaceDomain(surface(i), 1)

‘If Not IsArray(domU) Or Not IsArray(domV) Then Exit Sub

stepSizeU = (domU(1) – domU(0)) / elemsNumU

stepSizeV = (domV(1) – domV(0)) / elemsNumV

For j = 0 To elemsNumV

‘Set the y coordinate

coord(1) = domV(0) + stepSizeV * j

For k = 0 To elemsNumU

‘Set the x coordinate

coord(0) = domU(0) + stepSizeU * k     

‘Generates the z coordinate and stores the points in the A array

nodes(nodesNumU*j+k) = Rhino.EvaluateSurface(surface(i), coord)

Next

Next

ReDim topology(elemsNumU * elemsNumV – 1)

For j=0 To elemsNumV-1

For k=0 To elemsNumU-1

topology(elemsNumU*j+k) = Array((elemsNumU+1)*(j+0)+(k+0), (elemsNumU+1)*(j+0)+(k+1), (elemsNumU+1)*(j+1)+(k+1), (elemsNumU+1)*(j+1)+(k+0))

Next

Next

Rhino.AddMesh nodes, topology

Secondly, we have to translate our Rhino-MESH into a valid input file for Ansys:

‘The input file for the FEM analysis

Dim fsObject, fp, node

outPath = “c:\RhinoTemp\InputAnsys.txt”

Set fsObject = CreateObject(“Scripting.FileSystemObject”)

Set fp = fsObject.CreateTextFile(outPath, True) 

fp.Writeline “/PREP7″

fp.Writeline “!* …Static analysis”

fp.Writeline “ANTYPE,STATIC”

fp.Writeline “ET,1,SHELL63,,1″

fp.Writeline “R,1,0.3 !* …Shell thickness”      

fp.Writeline “MP,EX,1,3E+7″ ‘Units: m and KN

fp.Writeline “MP,NUXY,1,0.15″

fp.Writeline “!* …”

fp.Writeline “TYPE,1″

fp.Writeline “MAT,1″

fp.Writeline “REAL,1″

‘Mesh nodes

fp.Writeline “!*”

fp.Writeline “!* …Nodes”

For j = 0 To nodesNumV – 1

For k = 0 To nodesNumU – 1

node = nodes(nodesNumU*j+k)

fp.Writeline “N,” & nodesNumU*j+k+1 & “,” & node(0) & “,” & node(1) & “,” & node(2)

Next

Next

‘Shell elements (only geometry)

fp.Writeline “!*”

fp.Writeline “!* …Shell elements”

For j=0 To elemsNumV-1

For k=1 To elemsNumU

‘Ansys Shell 63 elements

nodes(nodesNumU*j+k) = Array((nodesNumU)*(j+0)+(k+0), (nodesNumU)*(j+0)+(k+1), (nodesNumU)*(j+1)+(k+1), (nodesNumU)*(j+1)+(k+0))

fp.Writeline “E,” & nodes(nodesNumU*j+k)(0) & “,” & nodes(nodesNumU*j+k)(1) & “,” & nodes(nodesNumU*j+k)(2) & “,” & nodes(nodesNumU*j+k)(3)

Next

Next

fp.Writeline “!* …”

fp.Writeline “FINISH”

fp.Writeline “/SOLU”

fp.Writeline “!* …Constrains”

‘Single Node constrain

fp.Writeline “D,” & CStr(1) & “,UX” ‘— 1 is the number of the node to constrain

fp.Writeline “D,” & CStr(1) & “,UY”

fp.Writeline “D,” & CStr(1) & “,UZ”

‘External constrains

For m = 0 To elemsNumU

fp.Writeline “D,” & CStr(m+1) & “,UX”

fp.Writeline “D,” & CStr(m+1) & “,UY”

fp.Writeline “D,” & CStr(m+1) & “,UZ”

Next

For m = 0 To nodesNumV-2

fp.Writeline “D,” & CStr(m*nodesNumU)+1 & “,UX”

fp.Writeline “D,” & CStr(m*nodesNumU)+1 & “,UY”

fp.Writeline “D,” & CStr(m*nodesNumU)+1 & “,UZ”

Next

fp.Writeline “!* …Forces”

‘Nodal forces

For k=0 To (nodesNumU * nodesNumV -1)

fp.Writeline “F, ” & (k+1) & “, FZ, 1″

Next

‘Gravity ACEL and MPDATA, DENS

fp.Writeline “MPTEMP,,,,,,,,” 

fp.Writeline “MPTEMP,1,0  “

fp.Writeline “MPDATA,DENS,1,,25″

fp.Writeline “ACEL,0,0,-1,” 

fp.Writeline “SOLVE”

fp.Writeline “!* …Postprocessor”

fp.Writeline “/POST1″

‘fp.Writeline “PRNSOL, U, X”

‘fp.Writeline “PRNSOL, U, Y”

fp.Writeline “PRNSOL, U, Z” ‘Z direction displacements

fp.Writeline “FINISH”

fp.close

Finally, we have to launch the analysis and to read the results (e.g. the maximum displacement). It is better to separate this operation in a private function to recall:

Dim applicPath, applic, maxDisplacement

Dim fileExe, fileInput, fileOutput

fileExe = “C:\Programmi\Ansys Inc\v100\ANSYS\bin\intel\ansys100.exe” ‘Programs for ENG WIN versions

fileInput = “c:\RhinoTemp\InputAnsys.txt”

fileOutput = “c:\RhinoTemp\OutputAnsys.txt”

applicPath = fileExe & ” -b -i ” & fileInput & ” -o ” & fileOutput

Set applic = CreateObject(“WScript.Shell”)

applic.exec(applicPath)

Dim datiOut, out, rigaDati, risultDati, fileOutput, fileCompleto, counter

out = False

counter = 1

Set datiOut = CreateObject(“Scripting.FileSystemObject”)

‘— File exists?

Do Until out = True

out = datiOut.FileExists(fileOutput)

Rhino.sleep 1000

Counter = counter + 1

Loop

Rhino.Print “Now the output file exists”

Rhino.sleep 1000

Set out = datiOut.OpenTextFile(fileOutput, 1)

risultDati = 0

‘— Max displ value exists?

contatore = 1

fileCompleto = 0

Do Until fileCompleto = 1

While Not Out.AtEndOfStream

rigaDati = Out.readLine

If rigaDati = ” |                            ANSYS RUN COMPLETED                            |” Then

Rhino.Print “Analysis complete!”

fileCompleto = 1

End If

Wend

Rhino.sleep 1000

counter = counter + 1

If contatore = 600 Then

risultDati = 100

Rhino.Print “Error”

Exit Do

End If

Loop

‘— Read Max displacement       

Set out = datiOut.OpenTextFile(fileOutput, 1)

Do While Not out.AtEndOfStream

rigaDati = out.readLine  

If Left(RigaDati, 12)     = ” *** ERROR *” Then

risultDati = 100

Exit Do

ElseIf Left(RigaDati, 12) = ” MAXIMUM ABS” Then

out.SkipLine ’1

rigaDati = out.readLine

risultDati = Mid(rigaDati, 9, 12)

Exit Do

End If

Loop

‘RESULT

maxDisplacement = risultDati

Out.Close

‘Delete the output file

‘Out = DatiOut.DeleteFile(fileOutput, True)

That’s an automatic procedure to evaluate the structural performance of a Rhino MESH object with the ANSYS FEM solver.

Reference: Pugnale A., Sassone M., Morphogenesis and Structural Optimization of Shell Structures with the Aid of a Genetic Algorithm, in “Journal of the International Association for Shell and Spatial Structures”, Vol. 48, n. 155, Dicembre 2007.

Mario SASSONE, Alberto PUGNALE

Abstract
In this paper a constructional problem related to grid-shells design is approached as an optimisation problem and an evolutionary solution technique is proposed. The construction of glass grid-shells, when only four sides cladding elements are used, requires to check the planarity of each element, it means that the four corner points have to belong to the same plane. This requirement can be satisfied by generating the grid surface in specific ways (Schlaich and Bergermann), but when the design involves truly free form shapes, a procedure of optimisation able to reduce the non planarity of each element can be applied. The proposed one is based on the use of a genetic algorithm and has been applied to a benchmark and to a real case, in order to evaluate the efficiency of the procedure and the goodness of the solution.

Mario SASSONE, Tomàs MENDEZ, Alberto PUGNALE

Abstract
In this paper a computational morphogenesis process is applied to the design and the optimization of the shell roof of a large music hall. The architectural concept is based on the folded plate technique: a set of reinforced concrete plates, connected to each other on the edges, forms the global shape of the roof. The dimensions of each plate are different, so that the structure is not regular; moreover each plate is not plane but slightly curved to form hypars. Starting from the geometric concept of the roof, a process of optimization and form finding has been applied, in order to obtain the best acoustic performance, in terms of distribution of the acoustic pressure level in the hall. The folded plate surface has been described mathematically by using a NURBS representation and the control point positions has been assumed as the design variables. The acoustic performance has been evaluated by means of an algorithm able to calculate the acoustic pressure level in all the points of the hall, considering the reflection of the roof. The uniformity of acoustic pressure level in all the points of the hall has been assumed as the target of the optimization process. A genetic algorithm, i.e. an evolutionary iterative population-based scheme, has been adopted in the optimization procedure. The evolution of the shape towards the optimal solution is controlled interactively by the designer, in order to avoid the convergence to unexpected configurations.

“Graham promoted the creation of a computerized program called Building Optimization Program (BOP), aimed to cut down the construction costs. This program was designed as a diagnostic tool by which architects and engineers could identify the points where constructive costs could be cut down, but, of course, it was accused on its the main priority, that was the economy rather than the project.
[…] Also the introduction of Computer-Aided Design and of its possibilities was the result of a collaboration between Graham and Khan. The use of computers by SOM dates back to the early fifties and to the construction of the Academy of Aeronautics. In the following years, with the development of research, computers entered in SOM, as happened in many companies, initially through the administrative sector. In 1963 SOM installed an IBM-1620 that was used for the study of complex structural systems, energy requirements and for financial planning. Khan, Graham and a third partner, Richard Keating, were convinced that computers would have also other applications, thus decided to invest in updating the equipment of SOM. When, in the mid-seventies, were first CAD programs was developed, SOM began to engage programmers and computer experts. In 1980 Douglas Stoker made a presentation to the partners including a spectacular fly-thought, on three screens, of the city of Chicago, that led to the preparation for the partners of a booklet based on the potential of informatics. Stoker became a partner in 1984 and SOM formed a joint venture with IBM for the development and sale of software and services. Later, there was a transformation of the market in which this initiative was moving. From a recollection of John Zils, associate partner and currently in charge for the group of structural engineers SOM Chicago: “many people began to realize that computer software would have produced a lot of money and they threw themselves into this market [investing in research] with a force that for us became impossible to compete with them.” As a result, SOM sold its system to IBM. However, says Zils: “this sale still remains something difficult to stand. We were used to create our own software customized for what we wanted to do… And now we are dependent of others who do things for us and that, of course, are not in the way in which we want them. We are always having to evaluate different software to find the one that better fits to our needs.”

From ADAMS N., Skidmore, Owings & Merrill: SOM since 1936, Phaidon Press, 2007.
(Extracted from the Italian publication, from which derives my English translation)

Adapting software to the specific needs of the designer is clearly a necessity that dates back to the birth of computers. Perhaps, even more than today, these technologies were considered as something that is not a constrain for design but rather something that allows to expand design tools, even leaving the designer totally free to create them by himself.
A correct approach in the use of computer is the first step to take towards a constructive critical debate on the use of new technologies within the architectural design. I note that several of the most recent publications (For instance Architettura e cultura digitale, DeLanda, etc…) forget the arguments of Frazer, Graham and Khan. They are only the first of many designers, including the Italian Luigi Moretti, who almost half a century ago have reflected about using new tools. In conclusion, why we do not start from this theoretical bases rather than digress completely from the main topic, as often happens today?
 
PERSONALIZZATORI PRECOCI, Nuovi strumenti (2)

“Graham fu promotore della creazione di un programma computerizzato denominato Building Optimization Program (BOP), finalizzato all’abbattimento dei costi di costruzione. Questo programma era studiato come uno strumento diagnostico mediante il quale architetti e ingegneri potevano individuare i punti nei quali i costi esecutivi potevano essere tagliati, ma, naturalmente, attirava l’accusa che la priorità principale fosse l’economia più che il progetto.
[…] Anche l’introduzione delle possibilità insite nella progettazione computerizzata fu il risultato della collaborazione tra Graham e Khan. L’uso dei computer da parte di SOM risale ai primi anni cinquanta e alla costruzione dell’Accademia dell’Aeronautica. Negli anni successivi, con lo sviluppo della ricerca, i computer entrarono da SOM, come accadde in molte aziende, inizialmente attraverso il settore amministrativo. Nel 1963 SOM installò un IBM-1620 che veniva utilizzato per lo studio di sistemi strutturali complessi, dei fabbisogni energetici e per la gestione progettuale e finanziaria. Khan, Graham e un terzo partner, Richard Keating, erano convinti che i computer avrebbero avuto anche altre applicazioni, quindi decisero di investire nell’aggiornamento della dotazione informatica di SOM. Quando, a metà degli anni settanta, furono elaborati i primi programmi CAD, SOM cominciò ad assumere programmatori ed esperti di computer. Nel 1980 Douglas Stoker fece una presentazione ai partner comprendente uno spettacolare fly-throught, su tre schermi, della città di Chicago, che portò alla preparazione ai partner di una brochure sulle virtù della tecnologia informatica. Stoker divenne partner nel 1984 e SOM formò una joint venture con IBM per lo sviluppo e la vendita di servizi e software informatici. Successivamente, intervenne però una trasformazione del mercato nel quale questa iniziativa si muoveva. Come ricorda John Zils, associate partner e attualmente responsabile del gruppo strutturisti di SOM Chicago, “molta gente cominciava a rendersi conto che i software per computer avrebbero fatto moltissimi soldi e ci si buttarono [investendo denaro e ricerca] con una forza tale che per noi diventò impossibile competere”. Di conseguenza, SOM vendette il proprio sistema a IBM. Tuttavia, dice Zils: “ancora oggi la vendita rimane una cosa difficile da digerire. Eravamo abituati a crearci da soli il software su misura per quello che volevamo fare… E adesso ci troviamo a dipendere da altri che fanno le cose per noi e che, naturalmente, non le fanno nel modo, in cui noi vogliamo farle. Ci troviamo sempre a dover valutare i diversi software per trovare quello che si avvicina di più alle nostre esigenze.”

Tratto da: ADAMS N., Skidmore, Owings & Merrill: SOM dal 1936, Electa, Milano 2006, pp. 34-36.

L’adattamento del software alle specifiche esigenze del progettista è una necessità che evidentemente risale alla nascita dei computer. Forse ancora più che oggi, queste tecnologie erano viste come un qualcosa che non vincolava la progettazione ma piuttosto permetteva di ampliare gli strumenti progettuali, addirittura lasciando la totale libertà di crearseli su misura.
Un corretto approccio nell’uso dei computer è il primo passo da compiere verso un dibattito critico costruttivo sull’uso delle nuove tecnologie all’interno del progetto architettonico. Noto invece che parecchie tra le più recenti pubblicazioni (vedi Architettura e cultura digitale, vedi DeLanda, ecc…) dimenticano i ragionamenti di Frazer, di Graham e Khan. Questi ultimi addirittura progettisti che non si occupavano nello specifico di ricerca informatica. Sono solo i primi di tanti nomi, tra cui anche l’italiano Luigi Moretti, che quasi mezzo secolo addietro hanno ragionato sull’uso di nuovi strumenti informatici. A questo punto, perché non partire da queste basi invece di uscire totalmente fuori tema, come spesso accade oggi?

[…] the computer is a device with the power and speed to meet the requirements of the limits of our imaginations. We need this power to compress evolutionary time and space so that results can be achieved more realistically in our life-times. The emphasis, however, rests in the techniques, in the demonstration of the theoretical model and technical feasibility, and in the workings of the thought experiment.
Perhaps this computing without computers is the most important lesson to be learned from designing these tools. The real benefits are found in having to rethink explicitly and clearly the way in which we habitually do things. […] if an appropriate toll doesn’t exist, we design and construct one ourselves.
It could further be argued that we do not need to build these tools, but could simulate their behaviour in the computer. The point is that by externalizing and materializing the inner processes of the computer, our physical models act like any architectural model by assisting understanding and visualization. Our models are not just tools which assist with the formative process, but tools of explanation.

This passage has been extracted from the book of Frazer, An Evolutionary Architecture, Architectural Publications Association, London, 1995, (free for download at this web address: http://www.aaschool.ac.uk/publications/ea/intro.html) written after many years of experience in teaching and experimenting with students at Architectural Association in London.
The most important lesson we can infer from his words is the need of a real deep understanding of the main reason for which certain tools are used, on the basis of their potential and their operation way. For this reason, a manual creation of both hardware and software parts of a computer, as well as a manual simulation of computing processes, is helpful in this (still current) phase of learning with respect to these “new tools”.

Nuovi Strumenti (1)

Questo pezzo è tratto dal testo di Frazer, An Evolutionary Architecture, Architectural Publications Association, London, 1995, (liberamente scaricabile da questo indirizzo: http://www.aaschool.ac.uk/publications/ea/intro.html) scritto dopo l’esperienza già di parecchi anni d’insegnamento e sperimentazione con gli studenti presso l’Architectural Association di Londra.
Il maggior insegnamento che si può dedurre dalle sue parole sta nel cercare veramente di capire a fondo il motivo per il quale utilizziamo certi strumenti, sulla base delle loro potenzialità e del loro funzionamento. Per questo, creare manualmente l’hardware ed il software di un computer, oltre a simulare manualmente processi di computazione, aiuta in questa fase (ancora attuale) di apprendimento nei confronti di questi “nuovi strumenti”.