Monday, April 27, 2015

Yas Viceroy (BIM): Detailing

1. Introduction:

In this post I will be focusing on regenerating the different parts of Yas hotel's canopy using Dynamo (visual programming for Autodesk Revit). The process of - digitally - regenerating the canopy's different elements is divided into the following (points A-E will be using Mass Model templates. As for point F, the modeling process will be performed using an Architectural Project template):

A. Main supporting structure (Generated in Dynamo)
B. Glass panel's structural steel frame support:(generic adaptive component)
C. Glass panel's steel frame(generic adaptive component)
D. Glass panel (generic adaptive component)
E. Steel Joint (generic adaptive component)
F. Panel Deviation (Architectural Project Template)

The modeling process of the sections above is an approximation of the actual design (by Asymptote architecture), and are recreated to address the following project objectives:

A. Demonstration of a fully functional parametric BIM-based model.
B. Demonstration of the detailing process, and data accessibility and managing using visual programing in BIM applications.
C. Address issues noted in the previous post through data restructuring.
D. The benefits of using visual programming tools - such as Dynamo - in BIM-based parametric modeling applications.

2. Modeling:

Before proceeding into the modeling process of the different architectural and structural parts of the canopy, I have included a general overview of the definition developed in Dynamo (below). Each of the different node clusters shown are designed to perform a specific task and are all linked to each other to create a complete functional parametric system.

Image 1: shows the definition in Dynamo. Each cluster of nodes is design for a specific task. The goal of this graph is not to show the programming details, its purpose is to provide an overview of the modeling process and data flow.

Image 2: Shows the tasks performed at each part. Magenta, referencing Revit geometry, surface selection, and UV quad divisions. Navy blue, referencing glass panel generic adaptive component parts and parameters. Cyan, process of generating the canopy's main steel structure (Dynamo geometry). Green, referencing joint generic adaptive component geometry and parameters, and the process of finding their specific location on the canopy's surface. Lime, surface panel deviation. As shown all parts are connected and dependent on each other for the system to be fully - and parametrically - functional.

The modeling process first started by referencing the canopy's Revit geometry in Dynamo. For the purpose of this study, a section of the surface was extracted to generate the deferent architectural and structural parts mentioned above.

Image 3: shows the canopy's original model generated in Revit.

Image 4: Shows the referenced section of the canopy's surface and the U and V quad points in Dynamo.

Following this step is generating the canopy's main architectural and structural parts.

A. Main Supporting Structure:

The process of generating the canopy's structure was made in two steps. The first step was extracting the surface boundary lines (top and below), create the structural member profile, position the profile (in this case a circle) in a perpendicular manner to the surface's punditries, and finally using the "sweep" node to create the solid form.

Image 5: Shows the process of generating the first part of the canopy's main structural support.

From the image above we notice that, opposed to the original Revit model, the structural support in this case can be parametrically repositioned (control its distance from the surface) while maintaining its relationship with the other structural parts.

Furthermore, to ensure that the sweeping operation is done correctly, the profile (circle) had to be perpendicular to the surface's boundary lines. Therefore a point at parameter was created on these surface's curves and were used to reorient the profile's normal direction. Furthermore, a number slider is used here to control the profile's normal angle.

Afterwards, the solid (structural member) was repositioned with a specific distance from the surface to avoid any geometry overlap. This was achieved by reversing the curves' (surface boundary acting as the sweeping path) vector direction.

As for the second step in the process of modeling the surface's supporting structure, the surface's UV quads (points) were utilized to guide the process. However, the structural members had to respond to any changes made to the glass panels (adaptive components); they basically support these panels. So, the graph below shows how this step was achieved through connecting specific glass panel data, and surface UV points to guide and generate these structural members. Noticeable is the extensive use of data restructuring and list management.

Image 6: shows the second step in generating the supporting structural members.

After completing both steps, the generated geometry is then imported into the Revit file.

Image 7: Shows the complete steel structure (highlighted) and the Dynamo definition for importing the geometry into Revit.

Image 8: Shows how the curves guiding the structural support is controlled (distance from the original surface) to avoid geometry overlap. This was done through reversing the glass panel's surface points (point at parameter) normal direction.

B. Glass Panel's Structural Steel Frame Support:

In this process of regenerating the glass panels, the approach was to deconstruct the unit into three parts: the structural steel frame (placed directly onto the main steel structure), glass steel frame (placed directly onto the glass panel and tied to the structural steel frame through joints), and finally the glass panel. The glass steel frame and glass panel are linked together and are parametrically controlled (rotation angle) with the same variables.

Following the main structural support is the glass panels' structural steel frame. These frames will be located at each quad on the surface and fixed directly to the surface's main structural steel frame through steel joints.

Images 9: Shows the structural steel frame (generic adaptive component) with adjustable frame thickness.

Image 10: Shows the definition for creating the structural steel frame. Noticeable here is the ability to extract the adaptive component's parameters; the goal of this operation is to create a fully functional parametric model controlled in Dynamo.

Image 11: Shows the structural steel frame placed directly on the canopy's main steel structure.

C. Glass panel's steel frame: (generic adaptive component)

The steel frame utilized for this application is exactly the same as the previous generic model (structural steel frame), however, its role and behavior in the model differs. The previous structural frame was guided by the surface's UV quads, as for this one, it's guided by the glass panel's geometry and parameters.

Image 12: Shows an isolated view of the glass panel and its steel frame. 

Image 13: shows how both the glass panel and glass panel's steel frame were referenced in Dynamo along with their parameters.

D. Glass panel:

The purpose of deconstructing the adaptive component into three distinct elements was for data structuring and managing purposes, which helped in providing further control over the model.

The glass panel in this case served as a guide for the placement of the steel joints. First, the panels' "face" was extracted. Afterwards a point (at parameter) was placed on its surface to get the normal direction at that location and to mark the position of the steel joints.

Image 14: Shows how the glass panel was referenced, its face extracted, and generating a surface normal to guide both the main structural frame (previously explained) and steel joints afterwards. Furthermore, one of the advantages of this approach is the ability to maintain the panel's parametric features, such as rotation. Unlike the previous post, where the panels' parametric ability were disabled for the model to properly function.

Image 15: Shows all three parts of the glass panel in their proper location on the canopy's surface (Revit model).

C. Steel Joint (generic adaptive component)

Following is generating the steel joints. These generic adaptive components are meant to tie the glass panels (complete set) with the surface's main structural frame. The joints were first designed in Revit and then referenced in Dynamo. Their location, as mentioned previously, are based on the glass panel's surface normal. However, the normal - which was converted into a point by using (Vector. AsPoint node) - will only provide a single point for the adaptive component to snap onto. Therefore, the vector points were translated (copied and moved) with a specific distance from the original surface to create two points for the adaptive component to snap to. Each joint is perpendicular to one of the glass panel's faces when populated onto the canopy's surface. The result of this process is having joints pointing outwards in a uniform manner based on the panel's normal direction at these specific locations and properly positioned.

Image 16: Shows the steel joint (generic adaptive component).

Image 17: Shows the Dynamo definition for creating the steel joints and marking their location on the surface with proper normal orientation.

Image 18: Shows the surface and the two points marking the location of each of the steel joints.

Image 19: Shows the joints in the Revit model.

F. Panel Deviation for fabrication:

After the model was complete, the geometry was imported - from a Mass Model template - into an Architectural Project template for visualizing panel deviation. The purpose of this step, as part of the detailing process, is to extract all the relevant information that would help in creating the proper construction and fabrication documents.

Image 20: Shows the panel deviation analysis definition.

Image 21: Shows panel deviation color range. The green values represent planer surfaces, however the panels near the end of the surface (panels on the left) show a different range of colors noting that they require additional modification.

 Image 22: Shows an isolated view of the panels mentioned above. Noticeable is that each of the panels at this location include a different color range (value between 0 and 1). This helps in visualizing the design for any inconsistencies prior to construction document preparation or direct digital files transfer for fabrication.

The following images illustrate the behavior of the elements mentioned above when their parameter values are changed. The purpose is to illustrate the system's flexibility and functionality ranging from the overall form to the smallest details.

Images 23: Show how the change of two single values affect the overall system - and all its minor parts. Values controlled here are the UV divisions, and they start from 10 divisions in the U direction and increase to 70, and 6 in the V direction and increase to 9 divisions. The end values produce 630 panels (complete glass panel set) and joints (one joint per panel).

Image 24: Shows the complete model with 700 panels and joints (70 divisions in the U direction and 10 divisions in the V direction).

3. Further Investigation (future work):

Regardless of the system's functionality, some specific parts have to be further investigated. During this process of detailing previous issues in the Revit model were addressed. However, with further detailing other issues start appearing, they do not interfere with the model's parametric behavior, but will affect the proper construction and fabrication documents.

With Dynamo, and with a level of digital geometry, computational logic, and data flow understanding these problems can be traced and resolved. The example below shows one of the areas for future investigation.

Image 25: Shows geometry overlap due to panel tilting, specifically, the glass surface and panel structural steel frame. This does not cause any issues related to the functionality of the parametric model. However, the model shows this issue that needs to be resolved prior to construction or digital fabrication.

The sewing panels on the surface appear once the glass panel's tilting angle is increased. The reason is that the projected points (in the generic adaptive model and control the ends of the glass panel) are following the adaptive points' normal direction. To further explain, if the adaptive component is placed on a flat surface the glass panel's geometry will not be affected, because the adaptive points' normal is facing directly upwards. However, if the adaptive component is placed on a double curved surface, then the normal direction of the adaptive points will follow the surface's curvature, thus affecting its original normal direction that is guiding the projected points that control the glass panel. This change in normal direction will not consider any adjacent geometry and will case the panel to skew and overlap other geometry. Images below explain the issue.

Image 26: Shows the glass panel on a flat surface, notice the projected points are directly moving upwards, parallel to the adaptive point's normal direction.

Image 27: Shows another view of the glass panel (elevation). Notice the movement of the projected points, vertical along the adaptive point's normal.

Image 28: Shows the glass panel on the surface with tilting angle at zero.

Image 29: Shows the glass panel when the tilting angle is increased. The highlighted lines shows the angle difference between the adaptive point's original normal and its change once populated on the surface. The reason is that the adaptive points are following the surface's curvature, which affects their normal direction, thus skewing the glass panel.

The solution for this issue is not be addressed by fixing the adaptive component, but by creating a guide that is generated on the original surface to control the direction of the projected points.

4. Conclusion:

The detailing process shown above illustrates greater control and understanding of the geometry, however it included intensive data managing and restructuring to achieve the desired outcome. Furthermore, understanding of digital geometry and computational logic will greatly help in the process of tracing specific issues and propose proper solutions. The solutions and design decisions made here are a few of many possibilities. My goal was to minimize the use of nodes to achieve the required end result to save computing time and power.

The modeling process using Dynamo helped in achieving a fully functional parametric model. Each of the elements are dynamic and adaptive, and are linked to each other, so that a single value change will insure that all other parts respond accordingly to the change.

Visual programming and parametric design using BIM exceeds conventional geometric manipulation to include material properties, which adds an additional feature to the whole parametric modeling process and experience.

5. Project Video:


6. Acknowledgment:

I would like to thank once more Dr. Wei Yan for this exceptional learning experience.

7. Resources

Monday, March 30, 2015

Yas Viceroy (BIM)



1. Introduction:

In this report I will digitally reconstruct Yas Viceroy hotel (Abu Dhabi, UAE) using Revit (Autodesk 2014) as part of the requirement for the Building Information Modeling class at the College of Architecture, Texas A&M University (ARCH 653).
The study will aim to explore and investigate the parametric capabilities of the tool in addition to imbedded information (BIM), and how both will contribute to the learning experience of using advanced digital modeling tools while highlighting some of the technical limitation and operation difficulties that would help in future investigations.  
Furthermore, some of the processes in the sections below will address similarities and differences, advantages, and limitations of using different programs (Rhino, GH; and Revit) in achieving similar tasks.
1.1.Project Background:
In previous work (ARCH 655, College of Architecture, Texas A&M University), a general description of the project (Yas Viceroy) was added
However, and in brief, the hotel is designed by Asymptote Architecture (Hani Rashid and Anne Couture), and was completed in 2010. The hotel, as mentioned is located in the Marina are, Abu Dhabi, UAE.  The project consists of two building (hotel) and the connecting bridge, canopy (curved skin), and the F1 track.  


  Yas Viceroy by Asymptote Architecture (


In addition, and more specific, one of the main features to be explored with BIM are the canopy’s materials and structure, which are mainly steel and glass. The canopy also has advanced LED lighting system integrated in its design, however, the lighting system is not included in the study.  

 1.2.Previous Work:


Image 1: shows the complete parametric model of Yas Viceroy using Rhino and Grasshopper by Emad Al-Qattan (ARCH 655).


Previously in ARCH 655, Yas was reconstructed using Rhino (generative modeling environment) and its plug-in Grasshopper (visual programming). However, in this course (ARCH 653), as mentioned, the aim is to explore similar modeling techniques with a different digital environment and by adding building information (materials, building components, and structure). This process also helps in revealing the technical capabilities of the tools used in managing data and streamlining the design process to construction.

To further understand the process, in the following sections the digital reconstruction of the project is broken down to different parts, mainly the hotels major parts (mentioned above) using different templates in Revit such as: mass models, families (patters and generic models), and architecture projects. Furthermore, each template includes specific operational parameters and In-place models, modeling techniques (approach, options, and problem solving), and specific geometry to achieve the desired outcome.

The modeling process here is based on the gathered information (sections, plans, elevations, 3D images, and photos) from online resources, and due to their resolution (low), the final outcome represents an approximation of the actual building. Yet, the importance and goal of the project, to be restated is, exploring and investigating the tool and compare the process of reconstructing the building to other digital environments.

2.    Modeling:

In this section I have broken down the modeling process of Yas hotel into two main categories: the parametric model, using a Conceptual Mass modeling template; and building information modeling, using an Architecture Project template. Each of these categories will be shown in detail regarding the modeling approach, techniques (geometry generation), and the relationship between the individual parts as a whole. Furthermore, I will identify the similarities and difference between the use of Revit and Rhino at specific modeling stages.   

2.1.Parametric and Generative Models: Conceptual Mass and Families


2.1.1.   Conceptual Mass: Building Mass

First, the digital reconstruction here started by creating 2D geometrical forms (two perpendicular ellipses) using a Conceptual Mass family (Mass Model). And afterwards, duplicated vertically to create the level for each of the hotel’s floors.

The two main hotel buildings were completed at this stage. The two ellipses mentioned above are parametrically controlled in terms of their width and length with a Length parameter. The same two parameters are used for both ellipses in every Sketch (level). Therefore, any change of a single parameter value will affect the overall model in length or width - or changing both values.


Image 2: a parametric mass model of the two main hotel buildings.

The repeated use of the same parameter for all eight levels was to have a uniform extrusion of the building mass, yet each level could have had its own set of parameters, but that was not the intension.

Prior to further explaining the parametric relationship of the massing model, I have included the diagram below to illustrate the different geometrical forms at each level, and the fixed dimensions that have imbedded parametric information.


Image 3: shows mass model framework.

The type of line used here to create the ellipses is a Reference line. The reason for using this type opposed to Model lines is that it keeps the solid mass updating - and intact - when the parameters' values are changed.

Image 4: shows a random selection of the ellipses in an isolated view. Notice the color of the geometry is different (usually black) because it is changed from a Model to a Reference line

Image 5: shows the Length parameter (type instance) used to control the elliptical masses.


Image 6: shows the two main ellipses and the two types of parameters used for each one; Length and Width instant parameters.

The parameters of the two masses are controlled by changing three values, two on the level of the sketch (length and width) and one for the masses' total height in 3D. As for the distance value between the two main masses, in the model it is kept fixed. Later on, the distance parameter will be added between the geometric forms when creating the canopy.

Furthermore, in floors 2 to 10, a geometrical offset is made of each of their ellipses to create an outer protruding mass for the hotel's balconies, and their width and length values are parameterized. The balconies also include a height parameter to control the thickness of their floor slabs. As for the first two levels (ground floor and mezzanine) their height geometrically constrained to the first level.


Image 7: shows parameters to control the width of the ellipse and its offset.

The height of each individual floor is parameterized by linking them to the total building height through a formula. Therefore, any change in the overall height of the mass will automatically adjust the height of each individual floor. Images below show the use of these parameters.


Image 8: shows a side view of the two building masses, their levels, and height parameters (total building height, level height, and balcony slab height).

Image 9: shows the parameters' dialog box. The highlighted value in both the model and box is of the total building height.

Image 10: shows how the floor height and balcony location are adjusted when the total building height is changed.

Cutting geometry, in the design of the mass there are three areas that were shaped using void forms. The reason for using void forms (extrusions) in this case was to maintain the geometrical attributes of the sketches at different levels when reconfiguring the overall mass.

Image 11: shows the void extrusions selected (blue highlight).

At this stage of the process, what is noticeable in the parametric modeling interface of Revit (Conceptual Mass template), it follows a similar logic to Grasshopper when constructing a mass. However, in Revit, the ability to automatically generate a number of floors is not direct as using a "number slider" In Grasshopper. If that was the intent then a formula - with a conditional statement - is required.

Another important point here is the proper use of geometrical constraints. In general, Revit allows a number of constraints (three types), and here an alignment constraint is used to connect the balconies to their proper floor height (level).

Image 12: shows the Alignment constraint between the roof of the Mezzanine and first floor level.

2.1.2.   Conceptual Mass: Canopy

In this section the canopy (curved surface covering the two main hotel buildings) is designed through following a similar logic to the previous mass modeling process (geometry and parameters), yet with a greater level of geometrical relationship and parameters. The canopy also includes a double parametric feature; the level of the skin and on the level of the panels (parametric component and host).

Before explaining the process, I would like to mention the goal of paying close attention to the parametric nature of the canopy is to compare its operability compared to the one developed in ARCH 655. In the previous reconstruction of the canopy, the model’s framework was created in Rhino then referenced in Grasshopper for further development. Regardless of Grasshoppers parametric ability, the skin's framework was static, therefore the extents of manipulating the form were limited.

However, in Revit, the model was generated from scratch (framework) to be dynamic and adaptive to any possible changes in its parameters' values.

Image 13: shows the canopy's framework created in Rhino (by Emad Al-Qattan, ARCH 655).




Image 14: shows the lofted and paneled canopy in Grasshopper (by Emad Al-Qattan, ARCH


The recreation of the canopy in Revit first started by using a Conceptual Mass template adopting the same geometry (ellipses) from the previous model along with its associated parameters. After creating the first sketch, the base, it was then duplicated to create two additional levels (total of 3 levels). The geometry on each of these levels is controlled with a set of parameters (width, length, and distance all of type instance) that are all linked to a single parameter in the base sketch through a formula. In this approach we obtain an intricate model where its parts at each level are geometrically different in size, but respond to the change of a single value.

Image 15: shows the canopy's framework (geometry). Information details are further explained below. However, the aim here is give a general overview and a sense of the relationship between the parts and parameters.

Image 16: another view of the framework; showing the three levels, control points (guiding the spline to follow the elliptical geometry on each level, and to determine the lower curve's position and form), and the void extrusion geometry at the far top of the model.

Image 17: shows both geometry and use of Length parameters to establish the overall relationship between geometry (curves, splines, and points) and parameters in the base geometry at the first level.

The set of parameters used here (base geometry, first level) vary in type depending on their performance, and the two main types are: Length and Number instance parameters. The Length parameters control the width and length of each ellipse and the distance between them, and the parameters’ values are different for each ellipse on the same level and different on each of the following levels. The distance parameter between the two ellipses on the same level maintains their relationship and proportion when their length and width values are changed. 

Image 18: shows the distance control parameter between the ellipses on each level, marked by the letter "C".


Image 19: shows how the overall framework changes respectively and the different parameters update when the Length parameter of the base ellipse is changed (the ellipse's length value highlighted in blue in the parameter dialog box). 

Furthermore, in elevation, the distance between the different levels is parameterized. Therefore the skin's framework is adjustable in all axis (X, Y, and Z).


Image 20: shows the change in level height. The height between the second and third levels is different from the parameter used for the first and the second, but are linked through a formula. The reason for linking them and not using the same parameter is to provide the additional level of control over the geometry.

As for the control points (seen in the previous images) they are equal in number on all sides of the ellipse and are placed specific distances along the ellipse’s curve. They help control the spline’s shape when the parameters’ values are changed. 


Image 21: shows the placement of a single point on the ellipse and to act as an anchor for the spline later on. The point’s position on the curve and its relationship between the points preceding and following are is parameterized through a formula.  

As for the duplicated points (from the original points on the curve), as shown in the image above, they determine the shape and distance of the skin's "draping" form. The points are maintained in position and respond to the changes of both the curve and original points by a Number parameter. This parameter the unitized curve’s value (value from zero to one) to determine the points position so that the point will always remain in the same position. 

Image 22: shows the method of creating the drape form. First, the work plane on the original control point is set, to guide the duplicated point that controls the shape of the drape. The offset distance here is manually entered to provide the designer to interact with the form.

Image 23: shows the types of parameters used and their relationships (formulas) for creating the canopy’s framework.

After flexing (testing) the model to ensure it is properly working, then what follows is the lofting process to create the canopy’s mass. Once the mass is completed then void extrusions are used to cut and shape the canopy’s detailed features.

Image 24: shows the canopy's mass.

Image 25: a side view of the canopy's mass. What is noticeable here is the lower part of the mass where the control points have helped in creating the drape like shape that resembles the actual built structure.


Image 26: shows the testing of the model to ensure it is working properly. Noticeable is the duplicated point’s position on the lower spline where they have maintained their location and relationship between each other and the original points, and the solid mass geometry remains attached to the framework. The solid remains attached to the splines, due to the use of Reference lines.

Following in the process is preparing the skin's mass to house the desired pattern. In the actual built canopy the panels are diamond shaped, and are made of glass and a steel frame. Furthermore, the canopy includes a two layered skin; the first, is the steel structure, and the second is the framed glass panels.

The first step here is to divide the surface along its U and V directions (21 divisions across the U direction and 70 divisions across the V direction). The reason for choosing these values is for proportional reasons when applying the panels later on. The level of resolution (number of divisions) here is determined by the hardware (computing power), so an adequate number of panels are used here to illustrate the approach.

Image 27: shows the first half of the surface divided along the U and V directions.

Following this step is first choosing a predefined pattern to be loaded on the divided surface and test. In this case a Rhomboid pattern is used to panel the canopy.

Image 28: shows the use of a preset pattern (Rhomboid) to create the paneled surface. The model's geometry here is isolated from the original mass for better visual evaluation.

Image 29: here the model illustrates one of the major issues, which is the location of the seam and the behavior of the panels at that location. 

In this section I have found that modeling the canopy using Revit illustrated a greater level of control and accessibility to the geometry defining it. However, in terms of paneling tools, Grasshopper produced a uniform and seamless paneled surface with less of the computational power required to generate a similar number of panels.


Image 30: shows a seamless paneled surface using Grasshopper (Emad AL-Qattan, Arch 655). 

However, both Revit and Grasshopper illustrate a similar issue, which is closing the top part of the canopy. In Grasshopper, the gap, was minimized as much as possible. In Revit the gap is closed, but the skin’s curvature does not run smoothly from top to bottom, shown in the following image.

Image 31: shows the top of the canopy where its curvature is disrupted by the flat surface geometry.


 Image 32: shows the Grasshopper version of the canopy where the top is completely open.

2.1.3.   Curtain Panel by Pattern:  

In this stage of the process the model is working properly; the predefined panel system is not generating any error message. Therefore, the following step is to customize a panel similar in terms of geometry (rhomboid) and parameters, and assign material properties to it that resemble the actual building. This process introduces the first steps towards building information modeling (BIM), which will be used extensively later in the development of this project.

The process of constructing the panels first started by choosing a preset panel template from the Curtain Panel family of type rhomboid opposed to using the Curtain Panel Pattern Based template, which produced modeling errors. To further explain, the paneled surface is segmented by a Rhomboid geometry where its data is organized in a specific manner to house panels with similar characteristics, otherwise Revit will signal error messages.

Image 33: shows the Rhomboid template used to create the panel, which will be loaded onto the canopy.

Image 34: shows the developed framework for the final panel using a Rhomboid panel template. From the dialog box what is noticed is that, the parameters assigned to control the geometry are set, however, they are deleted prior to loading the panel onto the surface. This point will be further explained later on. 

Image 35: shows the panel (Revit realistic rendering) with materials applied; steel and glass.

Images 36: show the process of duplicating a Revit material to create a specific and customized material for the project; in this case glass (image above) and steel (image below).

The issue of loading panels with parameters onto the paneled surface, although they are working properly, is that the panel’s geometry deforms. One possible explanation and based on my personal approach to the project is that, because the panel is tilted in both the negative and positive directions and the parameters are active (values updating to context conditions) once it is loaded onto a surface with extreme curvature (with surface normal not uniformly projecting in the same direction), it is possible that the negative and positive values shift directions to follow the surface Normal direction.

Image 37: shows panel deformation when loaded onto the surface.


A similar issue was found during the modeling process of the same surface in Grasshopper, yet the problem was solved through fixing the direction of each panel's Normal.

Image 38: shows each panel's Normal direction reoriented outwards. Model done in Grasshopper by Emad Al-Qattan, Arch 655.
 Image 39: shows the panel in Revit when applied onto the surface after deactivating the parameters.


Image 40: shows a detailed view of the panel where the deformation is fixed.

Image 41: shows the tilted panels and the supporting steel frame working properly.


Image 42: shows the surface and the applied panel. Notice the lower end of the surface is open with no supporting steel frame and some panels are missing. This is resolved later on, by modeling an In-place "structural" Steel Frame in the final model using the Architectural Project template.  


3.    Building Information Model:


3.1.Exterior Modeling:


3.1.1.   Assembly: Architectural and Structural elements  

Once all the main building parts are modeled - using separate family templates - they were afterwards assembled in an Architectural Project template, where further building information is added. The type of data and modeling process in this stage will include material properties and In-place components (masses and generic models) of different type (architecture and structure).

The first step in this phase was to import the main Conceptual Mass models and afterwards create architectural components, such as: floors, walls, roofs, and curtain wall systems.

Image 43: shows a side view of the Conceptual Mass model imported in the Architectural Project template. The process of creating building information starts by generating floor slabs at each level. The floor slabs are created by having the level references intersect the model at the required height (height of each floor from finish to finish slab). The dialog box shows the reference levels that will create the floor slabs. The levels here are for 10 floors, pavement, F 1 asphalt track, and the floor for the connecting bridge.

The second and third step in this process are creating the roofs, walls, and curtain systems at the different levels. Some of the roof structures, as well be seen later, were reconfigured to create ones that resemble the actual building. At this point all of the architectural elements created are utilizing default Revit materials, such as generic walls and roofs, the same is for the glass type used for the curtain system.

Image 44: shows the process of creating roofs for the different building masses.

Image 45: shows the process of creating walls and curtain systems.

Once the basic elements are created. In the Architectural Project template each material has to be defined so that it could be recognized by Revit. Otherwise, and once the mass model is hidden, these elements will disappear. The predefined material properties for the panels on the curved skin helped shorten the process at this stage. However, the lower supporting steel structure had to be defined. In this case, the approach was to use a Model In-place Structural Frame to create the supporting element. Furthermore, the same process was used to create the supporting structure that lifts the canopy - from the ground.



 Image 46: shows the glass and steel canopy imported into the Architectural Project.


Image 47: shows the type of Model In-place family used for creating the canopy's supporting structure.


Image 48: shows the highlighted reference line that is used to create the canopy's structural frame.


Image 49: shows the modeling process for the supporting structure lifting the canopy.

Once all the elements have been defined and the parts are assembled, the following step is to customize materials, add wall openings and windows, and add Model In-place components (generic and mass models) as required. 

3.1.2.   Model In-place Components: Bridge (Mass)

The first in In-place model of type Mass is the bridge that connects the two hotel buildings. Several attempts were made to create the dynamic structure of the bridge prior to the example shown below. The first attempts were made using generic adaptive components and Generic model templates. However, due to the bridge's complex geometry and the extensive use of parameters, the model could not be completed. Through using the In-place modeling capabilities the process was more manageable.

The bridge consists of four layers of sketches each has its unique profile and are attached at both ends to the hotel's exterior walls. Each sketch is controlled by a number of points and projecting points (similar to the canopy's framework) to guide each other (levels). The points were then connected using defined geometry such as arcs and straight lines. Splines are not used here as they raise issues when converting any extruded geometry into an architectural element. This point will be further explained later on. Each of the connecting lines here are made as references to keep the lofted mass adaptable to any changes in its base geometry. The lofting process was made into two phase to overcome some of the unpredictable and uncontrolled mass bending profile



Image 50: shows the four levels of sketches used to create the connecting bridge. The first two sketches on the left (level 1 and level 2) were first lofted, then the remaining were lofted in the second step (level 2, level 3, and level 4).


Image 51: shows the complete mass of the bridge.

Image 52: shows the bridge's profiles. As for the side opening and windows, they are shown below. 

Images 53 (above) and 54 (below): show the process of creating the side opening for the window. Image above shows the void extrusion, Image below shows the final bridge mass with the window opening.

When modeling the window opening, the bridge's complex geometry (bending in both direction) made it difficult to set the Work Plane for modeling the curtain system. The solution was to set a point was on the outer geometry of the opening to define a new Work Plane. Afterwards, an equal number of points were placed on the edges of the opening to guide the spline. The spline in this case is not continues it was made of two spline segments to avoid massing issues. Afterwards a mass was created from the profile, then divided, and finally paneled to create the curtain system for the bridge.


Image 54: shows the modeling of the bridge's curtain system. The Work plan - generated by the point - is visible here along with the points placed on the edge of the opening and the spline creating the glass's profile.


Image 55: shows the generated surface being divided for paneling purposes.


Image 56: shows the final curtain system for the bridge.

3.1.3.   Model In-place Components: Fountain (Generic)

The outdoor fountain creating the roundabout in the project's landscape was modeled using an In-place generic model and used a similar material to the outdoor pavement. The goal here is to illustrate the different uses of components in the project and how each contributes to its completion.



Image 57: shows the outdoor fountain in Revit's shaded render view. A water body is also created.

3.1.4.   Model In-place Components: Mass and Generic Model solutions 

In some cases there was a conflict between the Conceptual Mass and Architecture Project files when converting specific types of geometry to building elements. The reason is that when the material is applied to it the material attributes and geometry tool functions are disabled if Revit, because it could not identify the type of geometry. Some examples are shown below. One solution to this problem was to use and replace the mass element with an In-place component that has a clear defined geometry such as: circles, arcs, etc.


Image 58: shows the connection between two walls. The walls originally (Conceptual Mass file) were generated using splines, however, when the geometry was converted into building elements they could not be joined as a single wall. This is not the case with all splines, only the issues that I have personally encountered.


Image 59: shows how two straight walls were fixed by joining them with an arc.

3.1.5.   Materials:

In this section I will discuss the process of adding material to building elements. Each of the previously mentioned elements first were assigned default material such as Generic walls or roofs. However, these materials were then duplicated to create a separate material library specific for the project to be used for exterior walls, roofs, curtain systems, asphalt, and railings. The material properties were adjusted in terms of color, pattern, and thickness. Some additional changes were made to some walls’ material structure. This new library will help for further development later on to include additional material features such as thermal qualities.  




Image 60: shows material properties of handrails used for the project.


Image 61: shows asphalt structure modifications dialog box.

Image 62: shows wall material editing (type and pattern).
Image 63: shows glass unit editing.

Image 64: shows the complete building exterior (shaded render view, other rendered view are added below). Here the basic architectural elements, materials, and In-place models and components are added and defined, and all the building parts are put together. Following this step is defining interior components and materials.


Image 65: shows the complete building with Revit's Realistic rendering.



Image 66: shows the complete building using Revit's Ray Trace rendering.

All the renderings are set to use sun lighting and shadow settings based on the projects location (Abu Dhabi, UAE) date (August 10th, 2015) and time (10:00 am). As for adjusting the solar angle and project orientation, this information is discussed below.

Image 67: shows the sun's settings for project rendering purposes.

3.1.6.   Project orientation:
The original architectural model prior to rendering was facing Project North, therefore to obtain better rendering, shade and shadows, and for solar analysis purposes the model was reoriented towards based on True North.


Image 68: shows the project's new orientation based on True North, with the sun's location as previously mentioned.


Image 69: shows a site view with the project's adjusted orientation.

3.2.      Interior Modeling:

The interiors spaces are created through using two types of walls (Generic 8'' and Interior Partition 61/8'' [2 hr]). The doors (5 type) differ based on the type of space, and include: Single-Flush, Double-Flush, Single Double Acting, and Sliding doors. As for the exterior, three types of doors are used: Double-Curtain, Single-Curtain, and Single-Flush (aluminum) doors. Furthermore, the interior furnishing process also included adding some Specialty Equipment such as elevators, and circulation components such as stairs.  

The floors furnished in this project are the repeated floors (First to Ninth). To accelerate this process four types of door groups were created. Each of these groups include a specific number and type of doors that will snap into location when loaded to each floor. The plans below and the 3D section box further explain these details.



Image 7-: shows the Second floor plan with interior walls (Generic and Partition walls), doors, and elevators.

Image 71: shows a close-up view of the Second floor plan layout.


mage 72: shows one of the door group created for furnishing the repeated floors.


Image 73: shows a 3D section through the rear hotel building (marina bay). The section shows the interiors of each floor, including walls, doors, and windows.


Image 74: shows another 3D section view of the floor interior elements.
3.3.      Furniture:
The bedrooms in the rear hotel building are furnished, and the goal is to give both a sense of scale and type of function. The furnishing process was similar to the way doors were placed throughout the buildings, by creating groups. The furniture group includes a number of furniture types and is duplicated at each level. If the group is edited (furniture pieces were added or removed) the furniture layout in the whole building updates.
Image 75: shows a 3D section of the bedrooms and furniture layout at each floor of the hotel.
Image 76: shows the furniture layout on the 1st floor.
4.    Documentation:

One of the crucial phases of any project is proper documentation such as drawings (architectural, structural, and MEP) for tendering and construction purposes. BIM software's ability to generate such documents goes even further beyond generating drawings to also include BOQ schedules. In this section I will only post a selected number of architectural drawings (Plans, sections, and elevations) to illustrate the ease of generating these documents using BIM (in this project Revit).

4.1. Architectural Plans


 Image 78: section A-A

 Image 79: section B-B

Image 80: section C-C



Image 81: South Elevation (based on True North).


 Image 82: North Elevation (based on True North).

Image 83: West Elevation (based on True North).


Image 84: East Elevation (based on True North).


Image 85: shows a rendered view of the hotel and bridge. The image was done through using Autodesk 360 cloud rendering service.

Image 86: Another rendered view. The image was done through using Autodesk 360 cloud rendering service.

Image 87: Perspective showing the project using Revit's Ray Trace.


This project was a great opportunity to revisit and explore Yas Viceroy once again using a different digital design tool. The first version of reconstructing the hotel as mentioned was through Rhino and grasshopper, which mainly dealt with geometry modeling. In this second attempt, modeling process was further enhanced through the use of Building Information Modeling; the level of accuracy it provided, variety of work templates, tool functionality, and overall workflow greatly reshaped my understanding of the project.
In addition, the main objective of this study was to understand the limitations, workflow, and opportunities when comparing digital tools; these tools are different, but some of the applications are the same. I have managed to explore with this project, as much as possible a wide range of the tools that the software had to offer to develop my understanding and knowledge when it comes to utilizing BIM.
Revit offered a wide range of templates (generic, parametric, and building information modeling) that are interconnected with each other. This is a great advantage to maintain consistency of the workflow across building design and construction disciplines. Furthermore, another great feature was the ease of project documentation, drawings, details, schedules etc. As mentioned producing these documents was done in a matter of minutes opposed to the conventional practice, which can also help in optimizing project time.
However, regardless of the extraordinary features and potential of Revit, there were some technical issues and limitation that the tool failed to overcome. Throughout the development of the project, I have mentioned during my process these issues, and compared some of them to Grasshopper, especially if they share a similar modeling approach. However, here I would like to mention in general the ability to conceptualize, realize, and communicate a design. Revit is very technical and systematic and the ability to overcome some of the issues associated with these features requires an in-depth understanding of the tools operation. The debugging process once the model becomes complicated with different types of data and geometry, across a range of templates, makes this process almost impossible. Some of these issues some were resolved and others will be addressed in later versions of this project. Furthermore, these limitations also provide an opportunity to investigate the possibility to enhance specific functionalities of the digital tool. After completing this phase, for future work, I will be considering environmental factors and constructability aspects related to specific architectural elements in the hotel. It is an opportunity to look into these topics with BIM. 

6.Project Video:



I would like to take this opportunity to once again show my deepest gratitude to Dr. Wei Yan ( for this exceptional learning experience and his great effort, and to also thank my dear friend and colleague Jawad Altabtabai for his help and support. (