Andrei Nejur | Université de Montréal (original) (raw)
Papers by Andrei Nejur
De Gruyter eBooks, Sep 17, 2023
Proceedings of the 40th International Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe) [Volume 2], 2022
Proceedings of the 39th International Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe)
Proceedings of the 40th International Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe) [Volume 1]
Author(s): Steinfeld, K | Abstract: Mesh segmentation has become an important and well-researched... more Author(s): Steinfeld, K | Abstract: Mesh segmentation has become an important and well-researched topic in computational geometry in recent years (Agathos et al 2008) . As a result, a number of new approaches have been developed that have led to innovations in a diverse set of problems in computer graphics (CG) (Shamir 2008). Specifically, a range of effective methods for the division of a mesh have recently been proposed, including by K-means (Shlafman et al. 2002), graph cuts (Golovinskiy and Funkhouser 2008; Katz and Tal 2003 ), hierarchical clustering (Garland et al 2001; Gelfand and Guibas 2004 ; Golovinskiy and Funkhouser 2008 ), primitive fitting (Athene et al 2006), random walks(Lai et al), core extraction(Katz et al), tubular multi-scale analysis(Mortara et al. 2004), spectral clustering(Liu and Zhang 2004), and critical point analysis(Lin et al 2007), all of which depend upon a weighted graph representation, typically the dual of the given mesh (Shamir 2008). While these a...
This paper presents progress in the development of practical applications for graph representatio... more This paper presents progress in the development of practical applications for graph representations of meshes for a variety of problems relevant to generative architectural design (GAD). In previous work (Nejur and Steinfeld 2016), the authors demonstrated that while approaches to marrying mesh and graph representations drawn from computer graphics (CG) can be effective within the domains of applications for which they have been developed, they have not adequately addressed wider classes of problems in GAD. There, the authors asserted that a generalized framework for working with graph representations of meshes can effectively bring recent advances in mesh segmentation to bear on GAD problems, a utility demonstrated through the development of a plug-in for the visual programming environment Grasshopper. Here, we describe a number of implemented solutions to mesh segmentation and transformation problems, articulated as a series of additional features developed as a part of this same ...
This paper presents a method for the manipulation of groups of polyhedral cells that allows geome... more This paper presents a method for the manipulation of groups of polyhedral cells that allows geometric transformation while preserving the planarity constraints of the cells and maintaining the equilibrium direction of the edges for the reciprocity of the form and force diagrams. The paper expands on previously investigated single-cell manipulations and considers the effects of these transformations in adjacent cells and the whole system. All the transformations discussed in this paper maintain the initial topology of the input system. The result of this research can be applied to both form and force diagrams to investigate various geometric transformations resulting in convex or complex (self-intersecting) polyhedra as a group. The product of this research allows intuitive user interaction in working with form and force diagrams in the early stages of geometric structural design in 3D.
This paper presents progress in the development of practical applications for graph representatio... more This paper presents progress in the development of practical applications for graph representations of meshes for a variety of problems relevant to generative architectural design (GAD). In previous work (Nejur and Steinfeld 2016), the authors demonstrated that while approaches to marrying mesh and graph representations drawn from computer graphics (CG) can be effective within the domains of applications for which they have been developed, they have not adequately addressed wider classes of problems in GAD. There, the authors asserted that a generalized framework for working with graph representations of meshes can effectively bring recent advances in mesh segmentation to bear on GAD problems, a utility demonstrated through the development of a plug-in for the visual programming environment Grasshopper. Here, we describe a number of implemented solutions to mesh segmentation and transformation problems, articulated as a series of additional features developed as a part of this same ...
In this paper, we introduce a structural form finding plugin called PolyFrame for the Rhinoceros ... more In this paper, we introduce a structural form finding plugin called PolyFrame for the Rhinoceros software. This plugin is developed based on the methods of 3D Graphic Statics and Polyhedral Reciprocal Diagrams. The computational framework of this plugin uses new robust and efficient algorithms for the creation and modification of complex funicular, compression-only structural forms and is freely available for students, designers, researchers, and practitioners in the fields of architecture, structural engineering, mechanical engineering, and material science. The geometry-based structural design methods are one of the most intuitive yet powerful structural design methods that have recently been extended to 3D based on the Principles of the Equilibrium of Polyhedral Frames. Still, the increased geometrical complexities of the polyhedral diagrams hinder more in-depth practical applications and the research in this field. The framework proposed in this paper can manage, in near real-ti...
Although geometry-based structural design methods like 3D Graphic Statics (3DGS) allow for explor... more Although geometry-based structural design methods like 3D Graphic Statics (3DGS) allow for exploring a variety of spatial funicular geometry and their force equilibria. However, the material properties are not involved in the geometric form finding and there is no experimental data on the actual mechanical behavior of such systems. This paper will explore the structural performance of a funicular polyhedral geometry using experimental testing. The geometry of the physical prototype for the presented study is designed using 3DGS method. The specimen is constructed as a cast-in-place concrete structure, and the geometry of the sample is comparable to the standard concrete cylindrical test. High-performance, self-consolidating concrete is used for casting. Experimental results validated the 3DGS force distribution in the structure and showed that the magnitude of internal force in the members of the sample can be accurately predicted by 3DGS as long as the ultimate strength of the spec...
This paper presents progress in the development of practical applications for graph representatio... more This paper presents progress in the development of practical applications for graph representations of meshes for a variety of problems relevant to generative architectural design (GAD). In previous work (Nejur and Steinfeld 2016), the authors demonstrated that while approaches to marrying
mesh and graph representations drawn from computer graphics (CG) can be effective within the domains of applications for which they have been developed, they have not adequately addressed wider classes of problems in GAD. There, the authors asserted that a generalized framework for working with graph representations of meshes can effectively bring recent advances in mesh segmentation to bear on GAD problems, a utility demonstrated through the development of a plug-in for the visual programming environment Grasshopper. Here, we describe a number of implemented solutions to mesh segmentation and transformation problems, articulated as a series of additional features developed as a part of this same software. Included are problems of mesh segmentation approached through the creation of acyclic connected graphs (trees); problems of mesh transformations, such as those that unfold a segmented mesh in anticipation of fabrication; and problems of geometry generation in relation to a segmented mesh, as demonstrated through a generalized approach to mesh weaving. We present these features in the context of their potential applications in GAD and provide a limited set of examples for their use.
Mesh segmentation has become an important and well-researched topic in computational geometry in ... more Mesh segmentation has become an important and well-researched topic in computational geometry in recent years (Agathos et al. 2008). As a result, a number of new approaches have been developed that have led to innovations in a diverse set of problems in computer graphics (CG) (Shamir 2008). Specifically, a range of effective methods for the division of a mesh have recently been proposed, including by K-means (Shlafman et al. 2002), graph cuts (Golovinskiy and Funkhouser 2008; Katz and Tal 2003), hierarchical clustering (Garland et al. 2001; Gelfand and Guibas 2004; Golovinskiy and Funkhouser 2008), primitive fitti ng (Athene et al. 2006), random walks (Lai et al.), core extraction (Katz et al.), tubular multi-scale analysis (Mortara et al. 2004), spectral clustering (Liu and Zhang 2004), and critical point analysis (Lin et al. 2007), all of which depend upon a weighted graph representation, typically the dual of the given mesh (Shamir 2008). While these approaches have been proven effective within the narrowly defined domains of application for which they have been developed (Chen 2009), they have not been brought to bear on wider classes of problems in fields outside of CG, specifically on problems relevant to generative architectural design (GAD).
Given the widespread use of meshes and the utility of segmentation in GAD, by surveying the relevant and recently matured approaches to mesh segmentation in CG that share a common representation of the mesh dual, this paper identifies and takes steps to address a heretofore unrealized transfer of technology that would resolve a missed opportunity for both subject areas. Meshes are often employed by architectural designers for purposes that are distinct from and present a uniqueset of requirements in relation to similar applications that have
enjoyed more focused study in computer science. This paper presents a survey of similar applications, including thin-sheet fabrication (Mitani and Suzuki 2004), rendering optimization (Garland et al. 2001), 3D mesh compression (Taubin et al. 1998), morphing (Shapira et al. 2008) and mesh simplification (Kalvin
and Taylor 1996), and distinguish the requirements of these applications from those presented by GAD, including non-refinement in advance of the constraining of mesh geometry to planar-quad faces, and the ability to address a diversity of mesh features that may or may not be preserved. Following this survey of existing approaches and unmet needs, the authors assert that if a generalized framework for working with graph representations of meshes is developed, allowing for the interactive adjustment of edge weights, then the recent developments in mesh segmentation may be better brought to bear on GAD problems. This paper presents recent work toward the development of just such a framework, implemented as a plug-in for the visual programming environment Grasshopper.
The paper presents a possible solution for iterative optimization of a complex architectural form... more The paper presents a possible solution for iterative optimization of a complex architectural form for solar performance. The proposed work flow makes use of the new directions of evolution of digital design, namely parametric or otherwise known. algorithmic architecture. The case study presented demonstrates the simultaneous linked use of two CAD programs and several plug-ins that connect form generation with environmental simulation in an iterative loop that improves the design with each step. In order to tackle the extraordinary number of variants resulted form the large number of variables a search algorithm engine is used.
The particulars of the case study presented in the paper include an 119 X 67 quadrilateral surface to be covered by a solar canopy. The canopy needs to allow good indirect light in and also harvest as much as it can from the direct sunlight falling on it. The height of the covered space is about seven meters and the site is located in Leipzig in a relatively densely built neighborhood.
Translated into the particulars of the design, the covering would need to have both windows, to allow indirect light in, and also photovoltaic panels, to create electricity from direct sunlight. The windows had to be facing north as much as possible to stop any direct sunlight from going inside the covered space. The photovoltaic panels had to be placed so as to face directly towards the sun for as long as possible throughout the year.
The workflow employed to find the best shape started with the assumption that the covering was going to be represented as a polygonal mesh in the virtual space anyway. This is why it was decided that the triangles of the mesh will be the geometrical support for the two kind of surfaces to make up the canopy (transparent windows and opaque pv panels). So optimizing the base shape will mean having the triangles of the mesh facing the sun, half of them, and away from the direct sunlight, the other half.
For this purpose a generative algorithmic model was built in Rhinoceros-Grasshopper [1] with a number of free moving points (on the z axis) that through a Delaunay algorithm were interpolated in a mesh surface. The surface was variable in the sense that 145 of the points were free to move vertically and were each controlled through a variable. This allowed for an immense number of possible variants, 4145 to be exact. The variants produced were tested, on each triangular face, for average daily solar radiation. The testing was carried out in Autodesk Ecotect [2] and each variant was iteratively sent from Grasshopper to Ecotect using a connection plug-in called GECO [3]. From the values computed through solar insolation simulation in Ecotect a fitness value was computed for the whole covering. The value was calculated by splitting the sorted list of insolation values for each triangle into two lists and computing the division result of the larger average to the smaller one. The higher the fitness value the better the variant was. Using a stochastic search, driven by a meta-heuristic algorithm based on the simulated annealing process in cooling metals, the computer started to find the best variant from the solution space. After 8 hours of searching, the algorithm was stopped and the best variant so far was selected. In the end the fitness value was raised from 1,17 to almost 3, meaning a three fold increase in the performance of the final solution compared to the base one (which was considered a flat covering of the space).
Although manual labor and intensive human testing can be used in conjunction with standard CAD solutions to implement similar optimization cycles, the sheer number of possible variants, coupled with the amount of time necessary to modify and test each solution makes this an absolutely impossible endeavor. This is why formal optimization using generative techniques, simulation and search algorithms should always be considered. This is especially true if a reasonable time and energy amount can be invested in the algorithmic definition of an architectural problem.
De Gruyter eBooks, Sep 17, 2023
Proceedings of the 40th International Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe) [Volume 2], 2022
Proceedings of the 39th International Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe)
Proceedings of the 40th International Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe) [Volume 1]
Author(s): Steinfeld, K | Abstract: Mesh segmentation has become an important and well-researched... more Author(s): Steinfeld, K | Abstract: Mesh segmentation has become an important and well-researched topic in computational geometry in recent years (Agathos et al 2008) . As a result, a number of new approaches have been developed that have led to innovations in a diverse set of problems in computer graphics (CG) (Shamir 2008). Specifically, a range of effective methods for the division of a mesh have recently been proposed, including by K-means (Shlafman et al. 2002), graph cuts (Golovinskiy and Funkhouser 2008; Katz and Tal 2003 ), hierarchical clustering (Garland et al 2001; Gelfand and Guibas 2004 ; Golovinskiy and Funkhouser 2008 ), primitive fitting (Athene et al 2006), random walks(Lai et al), core extraction(Katz et al), tubular multi-scale analysis(Mortara et al. 2004), spectral clustering(Liu and Zhang 2004), and critical point analysis(Lin et al 2007), all of which depend upon a weighted graph representation, typically the dual of the given mesh (Shamir 2008). While these a...
This paper presents progress in the development of practical applications for graph representatio... more This paper presents progress in the development of practical applications for graph representations of meshes for a variety of problems relevant to generative architectural design (GAD). In previous work (Nejur and Steinfeld 2016), the authors demonstrated that while approaches to marrying mesh and graph representations drawn from computer graphics (CG) can be effective within the domains of applications for which they have been developed, they have not adequately addressed wider classes of problems in GAD. There, the authors asserted that a generalized framework for working with graph representations of meshes can effectively bring recent advances in mesh segmentation to bear on GAD problems, a utility demonstrated through the development of a plug-in for the visual programming environment Grasshopper. Here, we describe a number of implemented solutions to mesh segmentation and transformation problems, articulated as a series of additional features developed as a part of this same ...
This paper presents a method for the manipulation of groups of polyhedral cells that allows geome... more This paper presents a method for the manipulation of groups of polyhedral cells that allows geometric transformation while preserving the planarity constraints of the cells and maintaining the equilibrium direction of the edges for the reciprocity of the form and force diagrams. The paper expands on previously investigated single-cell manipulations and considers the effects of these transformations in adjacent cells and the whole system. All the transformations discussed in this paper maintain the initial topology of the input system. The result of this research can be applied to both form and force diagrams to investigate various geometric transformations resulting in convex or complex (self-intersecting) polyhedra as a group. The product of this research allows intuitive user interaction in working with form and force diagrams in the early stages of geometric structural design in 3D.
This paper presents progress in the development of practical applications for graph representatio... more This paper presents progress in the development of practical applications for graph representations of meshes for a variety of problems relevant to generative architectural design (GAD). In previous work (Nejur and Steinfeld 2016), the authors demonstrated that while approaches to marrying mesh and graph representations drawn from computer graphics (CG) can be effective within the domains of applications for which they have been developed, they have not adequately addressed wider classes of problems in GAD. There, the authors asserted that a generalized framework for working with graph representations of meshes can effectively bring recent advances in mesh segmentation to bear on GAD problems, a utility demonstrated through the development of a plug-in for the visual programming environment Grasshopper. Here, we describe a number of implemented solutions to mesh segmentation and transformation problems, articulated as a series of additional features developed as a part of this same ...
In this paper, we introduce a structural form finding plugin called PolyFrame for the Rhinoceros ... more In this paper, we introduce a structural form finding plugin called PolyFrame for the Rhinoceros software. This plugin is developed based on the methods of 3D Graphic Statics and Polyhedral Reciprocal Diagrams. The computational framework of this plugin uses new robust and efficient algorithms for the creation and modification of complex funicular, compression-only structural forms and is freely available for students, designers, researchers, and practitioners in the fields of architecture, structural engineering, mechanical engineering, and material science. The geometry-based structural design methods are one of the most intuitive yet powerful structural design methods that have recently been extended to 3D based on the Principles of the Equilibrium of Polyhedral Frames. Still, the increased geometrical complexities of the polyhedral diagrams hinder more in-depth practical applications and the research in this field. The framework proposed in this paper can manage, in near real-ti...
Although geometry-based structural design methods like 3D Graphic Statics (3DGS) allow for explor... more Although geometry-based structural design methods like 3D Graphic Statics (3DGS) allow for exploring a variety of spatial funicular geometry and their force equilibria. However, the material properties are not involved in the geometric form finding and there is no experimental data on the actual mechanical behavior of such systems. This paper will explore the structural performance of a funicular polyhedral geometry using experimental testing. The geometry of the physical prototype for the presented study is designed using 3DGS method. The specimen is constructed as a cast-in-place concrete structure, and the geometry of the sample is comparable to the standard concrete cylindrical test. High-performance, self-consolidating concrete is used for casting. Experimental results validated the 3DGS force distribution in the structure and showed that the magnitude of internal force in the members of the sample can be accurately predicted by 3DGS as long as the ultimate strength of the spec...
This paper presents progress in the development of practical applications for graph representatio... more This paper presents progress in the development of practical applications for graph representations of meshes for a variety of problems relevant to generative architectural design (GAD). In previous work (Nejur and Steinfeld 2016), the authors demonstrated that while approaches to marrying
mesh and graph representations drawn from computer graphics (CG) can be effective within the domains of applications for which they have been developed, they have not adequately addressed wider classes of problems in GAD. There, the authors asserted that a generalized framework for working with graph representations of meshes can effectively bring recent advances in mesh segmentation to bear on GAD problems, a utility demonstrated through the development of a plug-in for the visual programming environment Grasshopper. Here, we describe a number of implemented solutions to mesh segmentation and transformation problems, articulated as a series of additional features developed as a part of this same software. Included are problems of mesh segmentation approached through the creation of acyclic connected graphs (trees); problems of mesh transformations, such as those that unfold a segmented mesh in anticipation of fabrication; and problems of geometry generation in relation to a segmented mesh, as demonstrated through a generalized approach to mesh weaving. We present these features in the context of their potential applications in GAD and provide a limited set of examples for their use.
Mesh segmentation has become an important and well-researched topic in computational geometry in ... more Mesh segmentation has become an important and well-researched topic in computational geometry in recent years (Agathos et al. 2008). As a result, a number of new approaches have been developed that have led to innovations in a diverse set of problems in computer graphics (CG) (Shamir 2008). Specifically, a range of effective methods for the division of a mesh have recently been proposed, including by K-means (Shlafman et al. 2002), graph cuts (Golovinskiy and Funkhouser 2008; Katz and Tal 2003), hierarchical clustering (Garland et al. 2001; Gelfand and Guibas 2004; Golovinskiy and Funkhouser 2008), primitive fitti ng (Athene et al. 2006), random walks (Lai et al.), core extraction (Katz et al.), tubular multi-scale analysis (Mortara et al. 2004), spectral clustering (Liu and Zhang 2004), and critical point analysis (Lin et al. 2007), all of which depend upon a weighted graph representation, typically the dual of the given mesh (Shamir 2008). While these approaches have been proven effective within the narrowly defined domains of application for which they have been developed (Chen 2009), they have not been brought to bear on wider classes of problems in fields outside of CG, specifically on problems relevant to generative architectural design (GAD).
Given the widespread use of meshes and the utility of segmentation in GAD, by surveying the relevant and recently matured approaches to mesh segmentation in CG that share a common representation of the mesh dual, this paper identifies and takes steps to address a heretofore unrealized transfer of technology that would resolve a missed opportunity for both subject areas. Meshes are often employed by architectural designers for purposes that are distinct from and present a uniqueset of requirements in relation to similar applications that have
enjoyed more focused study in computer science. This paper presents a survey of similar applications, including thin-sheet fabrication (Mitani and Suzuki 2004), rendering optimization (Garland et al. 2001), 3D mesh compression (Taubin et al. 1998), morphing (Shapira et al. 2008) and mesh simplification (Kalvin
and Taylor 1996), and distinguish the requirements of these applications from those presented by GAD, including non-refinement in advance of the constraining of mesh geometry to planar-quad faces, and the ability to address a diversity of mesh features that may or may not be preserved. Following this survey of existing approaches and unmet needs, the authors assert that if a generalized framework for working with graph representations of meshes is developed, allowing for the interactive adjustment of edge weights, then the recent developments in mesh segmentation may be better brought to bear on GAD problems. This paper presents recent work toward the development of just such a framework, implemented as a plug-in for the visual programming environment Grasshopper.
The paper presents a possible solution for iterative optimization of a complex architectural form... more The paper presents a possible solution for iterative optimization of a complex architectural form for solar performance. The proposed work flow makes use of the new directions of evolution of digital design, namely parametric or otherwise known. algorithmic architecture. The case study presented demonstrates the simultaneous linked use of two CAD programs and several plug-ins that connect form generation with environmental simulation in an iterative loop that improves the design with each step. In order to tackle the extraordinary number of variants resulted form the large number of variables a search algorithm engine is used.
The particulars of the case study presented in the paper include an 119 X 67 quadrilateral surface to be covered by a solar canopy. The canopy needs to allow good indirect light in and also harvest as much as it can from the direct sunlight falling on it. The height of the covered space is about seven meters and the site is located in Leipzig in a relatively densely built neighborhood.
Translated into the particulars of the design, the covering would need to have both windows, to allow indirect light in, and also photovoltaic panels, to create electricity from direct sunlight. The windows had to be facing north as much as possible to stop any direct sunlight from going inside the covered space. The photovoltaic panels had to be placed so as to face directly towards the sun for as long as possible throughout the year.
The workflow employed to find the best shape started with the assumption that the covering was going to be represented as a polygonal mesh in the virtual space anyway. This is why it was decided that the triangles of the mesh will be the geometrical support for the two kind of surfaces to make up the canopy (transparent windows and opaque pv panels). So optimizing the base shape will mean having the triangles of the mesh facing the sun, half of them, and away from the direct sunlight, the other half.
For this purpose a generative algorithmic model was built in Rhinoceros-Grasshopper [1] with a number of free moving points (on the z axis) that through a Delaunay algorithm were interpolated in a mesh surface. The surface was variable in the sense that 145 of the points were free to move vertically and were each controlled through a variable. This allowed for an immense number of possible variants, 4145 to be exact. The variants produced were tested, on each triangular face, for average daily solar radiation. The testing was carried out in Autodesk Ecotect [2] and each variant was iteratively sent from Grasshopper to Ecotect using a connection plug-in called GECO [3]. From the values computed through solar insolation simulation in Ecotect a fitness value was computed for the whole covering. The value was calculated by splitting the sorted list of insolation values for each triangle into two lists and computing the division result of the larger average to the smaller one. The higher the fitness value the better the variant was. Using a stochastic search, driven by a meta-heuristic algorithm based on the simulated annealing process in cooling metals, the computer started to find the best variant from the solution space. After 8 hours of searching, the algorithm was stopped and the best variant so far was selected. In the end the fitness value was raised from 1,17 to almost 3, meaning a three fold increase in the performance of the final solution compared to the base one (which was considered a flat covering of the space).
Although manual labor and intensive human testing can be used in conjunction with standard CAD solutions to implement similar optimization cycles, the sheer number of possible variants, coupled with the amount of time necessary to modify and test each solution makes this an absolutely impossible endeavor. This is why formal optimization using generative techniques, simulation and search algorithms should always be considered. This is especially true if a reasonable time and energy amount can be invested in the algorithmic definition of an architectural problem.
De Luca, F., Nejur, A. and Dogan, T., Computing for a better tomorrow. 36th eCAADe Conference, Lodz University of Technology, Lodz, Poland, 19-21 September, Vol. 2, pp. 585-594, 2018
Daylight standards are one of the main factors for the shape and image of cities. With urbanizati... more Daylight standards are one of the main factors for the shape and image of cities.
With urbanization and ongoing densification of cities, new planning regulations
are emerging in order to manage access to sun light. In Estonia a daylight
standard defines the rights of light for existing buildings and the direct solar
access requirement for new premises. The solar envelope method and
environmental simulations to compute direct sun light hours on building façades
can be used to design buildings that respect both daylight requirements. However,
no existing tool integrates both methods in an easy to use manner. Further, the
assessment of façade performance needs to be related to the design of interior
layouts and of building clusters to be meaningful to architects. Hence, the present
work presents a computational design workflow for the evaluation and
optimisation of high density building clusters in urban environments in relation to
direct solar access requirements and selected types of floor plans.