BIM in Bridge Design - 638318 - FULLTEXT01 PDF
BIM in Bridge Design - 638318 - FULLTEXT01 PDF
BIM in Bridge Design - 638318 - FULLTEXT01 PDF
Jørund Johansen
1
Abstract
Building Information Modeling (BIM) is a relatively new technology and work
process. The main idea is to model 3D objects and store the information in a
database such that this information can easily be extracted and viewed from
these objects. The construction sector has more or less employed full BIM,
however, the infrastructure sector has not incorporated the new technology in
the same fashion. In this thesis, modeling of a specific bridge structure using
the BIM tool Autodesk
R RevitTM 2013, and the current bridge design process
used at Norconsult were analyzed in order to evaluate the advantages and
challenges of using BIM in bridge design. In order for Revit to be applicable
it has to be able to accurately design the structure and provide the necessary
documentation, as well as incorporate the steps leading up to, and following
the bridge design, satisfactory. The main conclusions are:
• The collaboration of the Revit model with road and underlay models (2D
models displaying the projection of a structure on the horizontal plane,
used as aid for the modeling of the actual 3D model) for reference is
satisfactory, however the road geometry has to be appropriate for BIM
purposes.
• The use of Robot was not studied extensively in this thesis; however,
non-standard geometry makes the incorporation of Robot difficult.
• The reinforcement tool included in Revit works well when modeling stan-
dard objects, but needs improvements in terms of modeling curved ob-
jects.
• In terms of being able to effectively utilize BIM, the skill level of the user
is of major importance; in the most extreme cases it can dictate whether
or not a task can be done.
2
Sammendrag
Bygningsinformasjonsmodellering (BIM) er en relativt ny teknologi og arbei-
dsprosess. Hovedideen er å modellere 3D-objekter og lagre informasjonen i
en database slik at denne informasjonen enkelt kan hentes ut og vises. Bygn-
ingssektoren benytter mer eller mindre BIM i hele arbeidsprosessen. Infras-
truktursektoren har derimot ikke inkorporert den nye teknologien i samme
grad. I denne rapporten har modelleringen av en brukonstruksjon i BIM-
verktøyet Autodesk
R RevitTM 2013, i tillegg til den nåværende bruprosjek-
teringsprosessen i Norconsult, blitt analysert for å evaluere fordelene og ut-
fordringene BIM står ovenfor med tanke på bruprosjektering. Dersom Revit
skal kunne benyttes som et verktøy i bruprosjektering, må programmet være
i stand til å nøyaktig representere brukonstruksjonen og produsere nødvendig
dokumentasjon, i tillegg til å inkorporere både filmateriale som fungerer som
en basis for modelleringen, og filmateriale som skal eksporteres til videre bruk
etter modelleringen. Hovedkonklusjonene er:
• Bruk av Robot ble ikke studert nøye. Allikevel ble det erfart at ikke-
standard geometri skaper problemer ved overføring av modell fra Revit
til Robot.
3
Contents
Abstract 2
Sammendrag 3
Table of Contents 4
Preface 7
1 Introduction 8
1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.2 The idea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.3 The thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2 Background 9
2.1 Building Information Modeling (BIM) . . . . . . . . . . . . . . . 9
2.2 Bridge design today . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.3 The role of BIM in bridge design . . . . . . . . . . . . . . . . . 12
2.4 Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.4.1 Revit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.4.2 Robot . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.4.3 Microstation . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.4.4 Civil 3D . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.5 Summary of the project assignment . . . . . . . . . . . . . . . . 19
2.5.1 The first model . . . . . . . . . . . . . . . . . . . . . . . 19
2.5.2 The second model . . . . . . . . . . . . . . . . . . . . . . 20
4 Modeling in Revit 26
4.1 Revit User Interface . . . . . . . . . . . . . . . . . . . . . . . . 26
4.2 Modeling Basics . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.2.1 Object Types . . . . . . . . . . . . . . . . . . . . . . . . 27
4.2.2 Views . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.2.3 Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.2.4 Reference Planes . . . . . . . . . . . . . . . . . . . . . . 31
4.3 The Bridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.4 Importing the centerline from Civil 3D into Revit . . . . . . . . 35
4.5 Modeling the bridge . . . . . . . . . . . . . . . . . . . . . . . . 39
4.5.1 Bridge deck and abutment back walls . . . . . . . . . . . 40
4.5.2 Columns and Foundations . . . . . . . . . . . . . . . . . 41
4
4.5.3 Abutment wing walls . . . . . . . . . . . . . . . . . . . . 42
4.5.4 Transition Plate . . . . . . . . . . . . . . . . . . . . . . . 43
4.6 Importing the model to Robot . . . . . . . . . . . . . . . . . . . 44
5 Design Changes 47
5.1 Length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
5.2 Position . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
5.3 Slope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
5.4 Radius . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
5.5 Depth of the piers . . . . . . . . . . . . . . . . . . . . . . . . . . 58
5.6 Orientation of the piers . . . . . . . . . . . . . . . . . . . . . . . 60
5.7 Summary of design changes . . . . . . . . . . . . . . . . . . . . 61
8 References 94
5
Acronyms, Abbreviations and Terms
Term Explanation
Civil 3DAutoCAD
R Civil 3DTM 2013
Revit Autodesk
R RevitTM 2013
Microstation Bentley
R MicrostationTM
Robot Autodesk
R Robot Structural AnalysisTM
Tekla Tekla
R StructuresTM
Robustness The objects’ resilience towards unexpected
and unwanted change when modeling other
objects
Flexibility The objects’ ability to be changed according
to user preference
Standard geometry Geometry already defined through the exist-
ing objects in the Revit library. This includes
slopes and curvature in one plane
Non-standard geometry Curved and/or irregular geometry (lacking
uniformity or symmetry; uneven in shape,
position, arrangement, etc.) not already de-
fined in the Revit library
PT Post-tensioning
PTR Post-tensioning reinforcement
Rebar Reinforcement bar
6
Preface
Through summer internships at Focus Software and project assignment, I
learned about BIM (Building Information Modeling) and tools used to per-
form BIM. I find the field interesting, and am intrigued by the potential BIM
has to increase the effectiveness of construction projects. The objective of this
thesis is to gain an understanding of the possibilities and challenges connected
to representing bridge design with BIM, especially focusing on the reinforce-
ment design. This report is based on:
Other companies may use alternative programs, and have a different ap-
proach to their work.
I would like to thank my advisor Tor Guttorm Syvertsen for valuable assistance
throughout the thesis work. I would also like to thank Focus Software and
Norconsult for their cooperation, and in particular Daniel Aase and Christer
Wolden, my contact persons in Focus Software and Norconsult, respectively.
7
1 Introduction
1.1 Motivation
Since its origin in the 1980s, Building Information Modeling (BIM) has been
used more frequently as the process itself and the tools used to perform BIM
have matured and also as a result of increased computing power. Where pure
computer performance previously was the threshold, software and information
transfer between different platforms and experts in various domains have now
taken its place as the main concerns.
Focus Software uses Revit and Robot for generating BIM-models and struc-
tural analysis models, respectively. There are still problems related to design-
ing bridges when BIM is concerned, and in this case Revit. Revit has a bridge
construction extension; yet it does not incorporate sufficient functionality for
cast-in-place concrete bridges with curvatures and slopes. There is a need to
assess the applicability of Revit for bridge design where the geometry of the
bridge is non-standard.
Reinforcement of bridges in Norway today is done mostly using 2D draw-
ings, which means that in order to get an overview of the full reinforcement of
a bridge one has to combine several separate drawings. With a 3D representa-
tion it will be easier and quicker to gain that overview, especially in terms of
interference of reinforcement bars. There is a need to evaluate Revit’s ability
to represent the reinforcement of bridges.
8
including the reinforcement of the bridge I will be able to gain a more de-
tailed understanding of the issues related to BIM in bridge design, for further
evaluation.
2 Background
One can discuss whether or not the term shared knowledge resource is a
correct description, as a BIM contains information from which people can gain
knowledge. One may argue there is no knowledge contained in the BIM per
se; however it is a resource for potential knowledge. The two most important
letters in BIM are I and M; a model is made where the information is available
for all the people working on the project. By using BIM, architects, engineers,
contractors and owners can create digital information and documentation (e.g.
cross sections, elevations and material lists). It is a powerful tool which enables
all the actors related to the project to extract the same information and connect
to the same platform.
By creating 3D-models and adding information related to all the differ-
ent disciplines involved (Plumbing, Electrical etc.), all the information which
previously had to be coordinated after they were made (sheets describing the
bridge used in the construction phase) can be incorporated in the model im-
mediately. This way, one obtains an overview of the project early, which in
turn means that problems or improvements can be identified early in the de-
sign process. In contrast to traditional drawing and the use of CAD-tools,
where the drawings are independent of each other, the BIM is a 3D-model
with 3D-objects and elements with parameterized properties and relations to
each other. Parameterized properties means that the objects are assigned with
certain parameters (e.g. dimensions, offset from original placement, slope etc.)
easily altered through a property palette. Once an object is altered, the object
will be altered everywhere in the BIM, which renders tedious changes in the
9
documentation obsolete.
• Zoning: the area assigned for the project describing structures already
situated in the area.
• Basic data: data of the area influencing the project (height of existing
terrain, vegetation, existing technical infrastructure etc.).
10
Figure 1: Bridge design workflow
Often, the engineer starts by creating the cross section which in most cases
follows the centerline. Then, either the bridge geometry in the horizontal
plane or a 3D model is established. Based on that decision, either the 3D
model or the bridge’s projection on the horizontal plane from the 3D model is
drawn, before elevations, foundation and pier details etc. are created. These
drawings are so called layers of information, and in Microstation, CAD layers
are displayed based on the wanted representation. This way the user is able
to, for instance, create an elevation showing only the bridge itself, and another
layer showing the terrain surrounding the bridge, and choose whether to show
the bridge with or without the terrain, or the terrain on its own.
The required dimensions of the bridge and the reinforcement need are spec-
ified by structural analysis, where reaction forces, stresses and strains are eval-
uated based on pertinent load combinations, both through hand calculations
and with the aid of softwares such as NovaFrame and NovaDesign. NovaFrame
is used for global analysis of the bridge, while NovaDesign focuses on the forces
at given sections of the bridge, which in turn dictates the reinforcement need
and the dimensioning of reinforced concrete cross sections. 2D reinforcement
drawings are made based on the structural analysis.
Designing a bridge is an iterative process, where changes in the basic data
and data received from other disciplines results in a need to update the bridge
design. The better the basic data, the fewer changes need to be made through-
11
out the design process.
12
Figure 2: Chart of the BIM stages [5]
There are several reasons why BIM is not utilized in the infrastructure
sector:
Benefits of utilizing BIM have been proved by pilot projects using programs
like Tekla and Revit. In these projects accurate 3D-models were created,
showing that it is possible to model more difficult geometry with BIM. Laser
scanning of existing structures have also been incorporated where additions to
the existing structures were to be created.
13
2.4 Software
The modeling reported here was done in Revit. Revit and Robot are two of the
programs Focus Software has specialized in. Microstation is the program used
by Norconsult for bridge design. Civil 3D was used to extract the centerline
of the bridge from a road model and import this centerline into Revit.
2.4.1 Revit
Revit is a BIM tool which focuses on the design part of the construction pro-
cess. It has an intuitive and clean user interface, and enables the user to model
everything from basic structures to specifically defined geometries. Previously
Revit was divided into an Architecture application, an MEP (Mechanical, Elec-
trical and Plumbing) application and a Structure application. In Revit 2013
these applications are combined in one application. Revit’s file format is .RVT,
but it can also export to and import from the file format IFC, which is the
standard format for BIM.
14
2.4.2 Robot
15
2.4.3 Microstation
16
2.4.4 Civil 3D
17
2.5.1 The first model
The first bridge model was made with the objective of quickly gaining an
overview of the complications one would encounter during the modeling of the
bridge. As an effect, it consisted of poorly defined objects and solutions, mak-
ing the BIM lack robustness and flexibility in terms of adapting to changes
made later on. It became apparent already in the early stages that the choice
of objects in the BIM and modeling procedure were important, and should
be carefully considered. The 3D model of the bridge looked correct, however
when scrutinizing the model there were multiple errors of unacceptable mag-
nitude (several millimeters). Because of the lacking robustness, remedying
errors in one place of the BIM often caused errors elsewhere, making the BIM
unpredictable. An example is shown in figure 7, where the vertical dimensions
are correct whereas the horizontal are not. If the horizontal dimensions were
changed to their appropriate values (200mm and 300mm), the vertical dimen-
sions would also change as an effect of how the object was modeled, leaving
the geometry incorrect.
Figure 7: Detail showing errors (dimensions in the red circle) caused by lack
of robustness
With the knowledge gained through modeling the first model, it was realized
that before starting the actual modeling of the bridge in the second model,
preparations had to be made. New object families were created to attain
robustness in the more difficult objects in the BIM. Objects were attached to
reference planes and levels rather than to each other in order to avoid having
18
objects’ dimensions and attributes suddenly change when changing those of
another object, which was an issue in the first model.
With a moderate amount of skill one was able to create a basic bridge
model and accompanying drawings, with the exception of the following issues:
• When creating the bridge deck and barriers, there was no convenient way
of specifying that they should have a slope while at the same time retain
vertical ends. The result was a slight angle at the ends which had to
be manually cut, leaving the dimensions of the deck at the ends slightly
wrong because of the slope. This problem needed to be circumvented by
choosing which sections to annotate; when annotating the abutments in
a vertical section one had to leave out the annotation of the bridge deck
because of the wrong dimensions.
• In order to view the cross section of the sloped bridge deck and also
be able to add annotations to this cross section, the sections which cut
the bridge deck had to be perfectly normal to the bridge deck’s x-axis
(the axis running along the deck’s length). This was experienced as a
sometimes frustrating task. Focus Software has, after the project was
concluded, made a tool for Revit used to automatically create a cross
sectional view of an object regardless of orientation. This tool has made
the operation a lot easier.
• Railings can only be attached to stairs and ramps, or placed in the hori-
zontal plane. The intention was to attach railings to the barriers, but in
19
order to do that, new specialized families needed to be created. The re-
sult was that attention was placed elsewhere, leaving the issue unsolved.
It should be noted that the level of detail was not the highest in neither of
the bridge models, and reinforcement was not studied extensively, because the
main goal was to get an overview of the big picture, and also because trivial,
yet tedious operations were neglected. However, the bridge as a whole was
relatively accurately represented. Slightly inaccurate dimensions in the second
model arose mostly due to rounding errors or minor modeling inaccuracies,
and were in the range of a few millimeters at most. When operating with
millimeter precision, this is also unacceptable. However, the errors were easier
to fix than in the first model due to superior robustness.
An important thing to mention is the fact that the ways of representing
the model have to be reconsidered when operating with BIM compared to
traditional 2D drawings. The BIM consists of actual objects, and one cannot
simply alter the representation of the objects in a 2D view without altering the
whole model. In traditional design using multiple 2D drawings one sometimes
change the position or altitude of certain elements in the drawing to make it
look cleaner and more legible. Also, for the same reason and also for dimension
purposes, cross sections are sometimes shown as vertical in a drawing when
they are in fact sloped. This is not possible, nor is it desirable, when working
with BIM. Thus, the representation of the model in the drawings needs to be
reviewed and the methods used have to be changed. This is a part of the
transfer from traditional methods to BIM.
20
vast amounts of reinforcement bars, making the reinforcement bars placement
difficult.
The project was split into two stages; the first stage included a new span
of 450 meters, and in the second stage an existing span of 550 meters was
brought down in order to create a new one. The first stage was designed and
reinforced entirely by traditional methods, along with the first bridge deck
(the distance between two piers) of the second span. The result was 20-25
out of 100 reinforcement bars that had to be reshaped at the construction
site for each deck, as a consequence of the difficulties arising from using and
interpreting 2D drawings. The rest of the bridge was modeled in Tekla as an
addition to the 2D drawings, and the error in reinforcement bars was reduced
to 0-2 per 100 on each deck. In addition, on the first bridge deck Pihl spent 23
days on the reinforcement bars work alone, in comparison to approximately 8
on the second deck. Even though some of the increase in productivity can be
explained by the season in which the two bridge decks were reinforced. The
first deck was created during the winter, and thus naturally becomes more
cumbersome. Nonetheless, Pihl states in their report, most of the explanation
lies within the incorporation of BIM.
Among the advantages of using BIM in the reinforcement of the bridge
deck, there is the possibility of printing out the reinforcement bar schematics
and hang them up on the construction site for the people working with the
reinforcement bars to see. This lets them get a much better understanding of
the greater picture, the 3D picture, instead of having to visualize the reinforce-
ment bars placement based on 2D drawings. Less time is spent discussing the
placement of the bars, and more time can be spent actually inserting them.
Another very useful tool Tekla developed for this project, based on specifica-
tions from Pihl, was a function which enabled reinforcement bars to be ordered
to the construction site by selecting and ordering them through the BIM. All
the specifications of the reinforcement bars were sent to the manufacturer. Be-
cause of this, and by having an improved overview of the reinforcement bar
schematics, Pihl could very efficiently plan the ordering of reinforcement bars
in a way that left fewer excess reinforcement bars on the construction site.
21
Navisworks Manage is a project review solution that supports coordination,
analysis, and communication of design intent and constructability. Information
about Sundt Construction and the use of BIM in their bridge design projects
was gained through a skype conversation with one of their modeling engineers,
Eric Cylwik [11], and from their webpages [12].
In one of their design projects, the renovation of the 7th street bridge in
Texas [13], Sundt Construction used Civil 3D and Revit to model the existing
bridge with its 12 new precast post-tensioned concrete arches that run the
length of the bridge. The arches included complex post-tensioning tendons
and lighting elements which needed careful planning. Being able to model
these elements in 3D helped Sundt visualize how all these components fit
together. The post-tensioning tendons, which curved in all three dimensions,
were modeled in Civil 3D before imported to Revit, where the rest of the bridge,
including reinforcement bars, was modeled. According to Eric Cylwik, Sundt
usually model the bridge superstructure (i.e. the bridge deck) in Civil 3D,
which holds superior functionality compared to Revit in terms of the creation of
non-standard geometry. The bridge superstructure is then imported to Revit,
where the substructure (i.e. columns, foundations etc.) is included. There
is not much interplay between the objects imported from Civil 3D to Revit
and the objects created in Revit. The user can neither attach to the imported
objects nor dimension to them in Revit. However, curvature in general poses
problems in Revit, and Cylwik did not consider the lack of interplay as a
major concern. He mentioned that, as of Civil 3D verison 2014, solid objects
can be shown in Civil 3D when imported from Revit to Civil 3D. This way, one
is able to bring a section from Revit into Civil 3D, obtain a solid object and
document it there. According to Cylwik, Civil 3D is superior to Revit in terms
of dimensioning objects with curved geometry. He stressed that both Civil 3D
and Revit should be utilized in bridge modeling to effectively represent a bridge
structure correctly and include the information needed.
Eric Cylwik was not thrilled with how Revit worked in terms of rebar mod-
eling. It was not very flexible or efficient in general, and in particular, Revit
had a difficult time with reinforcing objects that were curved, which caused is-
sues when modeling the concrete arches. Another reinforcement problem with
Revit experienced by Cylwik, was the fact that Revit has a concept where re-
bar can only be created within a concrete object, resulting in an error message
when attempting to add rebar outside of an object. However they found an ad
hoc solution; if the user groups the rebars in Revit, the rule Revit generates
for the rebars is removed, allowing for rebars outside of concrete objects.
The biggest advantage with BIM, according to Eric, is the better com-
munication between all parties. During planning sessions they could visually
communicate with their subcontractors to develop a more realistic construc-
tion schedule, and everyone better understood both their own task, and the
tasks of others.
22
Another great advantage with using BIM is the more efficient sharing of
information. As an example, if the elevations of specific points on the bridge
which are not included in the construction documents need to be determined,
previously one would have to figure out the elevations and include them on new
spreadsheets. Thus, new spreadsheets would need to be created when updates
and changes were made. Now, these elevations can be entered into the model
and automatically be included in drawings extracted from the model, reducing
the amount of spreadsheets.
4 Modeling in Revit
• The tools enable the user to perform different tasks in order to create a
model which will be visualized in the Model Window.
• The Quick Access Toolbar allows for prompt access to frequently used
tools, and the tools displayed in the Quick Access Toolbar can be per-
sonalized based on user preferences.
• The Type Selector displays the object currently selected or chosen for
creation with customizable properties displayed in the Properties Palette.
23
• The Project Browser contains all the objects created as well as all the
different views contained in the model.
• The View Control bar includes tools for customizing how the model is
displayed in the Model Window, including hiding or isolating elements,
and adjusting the level of detail of the objects.
• Object Types
• Views
• Levels
• Reference Planes
When describing the modeling process, I will distinguish between three ways
of defining the objects used in the model:
• Standard objects: These are the built-in objects that are already defined
in the Revit library, accessible through the Revit tools. The objects are
categorized in families (e.g. an IPE 220 beam is a member of the beam
family). In object modeling terms, a family is a generic class/superclass,
and an object is a class/subclass.
• User defined objects: These are objects defined through the families
already in Revit. As an example, the user is able to create a new beam
family with a different profile than the ones already in the Revit library.
The new beam will have beam properties and will be available when
selecting the beam tool. The user defined family can be loaded into and
used in any project.
• “In-Place Mass” objects: These are objects defined in the model window
and are specific to the current model. The “In-Place Masses” are also
categorized in families, but they lack some of the properties and modeling
operations otherwise attainable using objects already available in the
Revit library.
24
Revit works very well when using the standard objects; these are the objects
Revit is mainly programmed to deal with, and the framework of operations
available for the standard objects makes for a robust model if created smartly.
How well the incorporation of user defined objects in the model works, com-
pared to standard objects, depends greatly on how well the families are mod-
eled. A user defined family without flaws is just as good as a standard family
with the exception of the non-standard geometry it may comprise. However, a
poorly designed family may cause unexpected behavior in the objects placed
in the model, especially when making changes (altering dimensions, aligning
it with other objects etc.) to the object.
When modeling user defined geometry, it is very useful to create well func-
tioning families. Not only does this ensure robust and customizable objects,
the created families can be saved and used in other projects as well. The im-
portance of reuse of families cannot be emphasized enough, and is especially
valid when creating objects that are used often and across multiple projects.
The families are then simply loaded into the desired project, where they may
be customized according to the needs of that project.
The ”In-Place Masses” are the least robust objects; however, they can be
useful when objects do not need to be included in a family, and the user wants
to create them quickly. In fact, sometimes an ”In-Place Mass” is the only
object suitable. An example is a void form used to cut another object in order
for that object to attain the desired shape when other built-in methods fail to
provide the results the user wants. In figure 10, one of the void forms at the
abutment wing walls is shown.
25
Figure 10: Void form cutting an abutment wing wall
4.2.2 Views
As mentioned in 4.1, Revit consists of different views in which the user is able
to work on the model. The default views are 3D-view, elevations, and plan
views. The plan views are aligned to the default levels in the model, allowing
the user to evaluate the model either looking upwards or downwards from those
levels. The user can define multiple 3D-views displaying the model differently
or create more levels from which plan views are automatically included. In
addition to these views, other views can be created, displaying the model
differently or focusing on different parts of the model. There are two types of
views used in this regard:
26
too much detail, making it hard and cumbersome to work with.
4.2.3 Levels
Figure 11: Elevation view including levels and three columns attached to the
different levels
Reference planes are, as the name indicates, a plane used as a reference for
objects. They can only be placed in 2D-views as planes parallel to the view
direction of that 2D-view, however, as long as they are parallel, the reference
planes may be attributed any orientation. Furthermore, if the view direction
is defined as the x-direction of the plane, and the normal vector defines the
z-direction of the plane, the user can also alter the orientation of the plane in
a view with a view direction in the y-direction. Thus, a reference plane can
in practice be given any orientation in 3D. In figure 12 two reference planes
including dimensions are placed in the same elevations view as in figure 11.
27
Figure 12: Reference planes
28
Figure 14: Plan view of the bridge
29
this thesis consisted of multiple layers which comprised the whole road model.
One of these layers contained the centerline. The task of isolating this layer
for import to Revit is fairly simple as long as the user makes sure the units
are specified to be the same in Civil 3D and Revit. When extracting the cen-
terline from the road model one has to either alter the units within the road
model in Civil 3D, or specify the right units during the process of importing
the centerline into Revit. The centerline included in the road model received
was, in contrast to the one received in the project 2.5, a continuous line seg-
ment. However, Revit did not respond well to the very large coordinates of
the road model. Norconsult uses the map projection EUREF 89 NTM [14],
which ensures that the road model is correctly situated geographically. This
resulted in an X-value (North position in the Revit grid) of 6679346005mm.
Revit truncates imported coordinate values higher than E9, which resulted in
an X-value of about 1000km when acquiring the centerline’s coordinates in
Revit.
Furthermore, the large coordinates created inaccuracies in the centerline
object in Revit. When moving the line to the origin in Revit, it behaved
unpredictably, not letting itself attach completely to the origin but instead
shifting from either side of the origin when zooming in and out about the
origin.
30
Figure 17: Perturbation when zoomed in
Specifying the correct large coordinates at a point in the model after im-
porting the centerline worked properly, yet it is desirable to be able to acquire
large coordinates directly from the imported model without manually altering
them after import. However, the problem with the centerline’s behavior re-
mained. After some troubleshooting, it was realized that the problem could be
circumvented by altering the coordinates in the road model in Civil 3D to 0,0
before importing the centerline to Revit and specifying the coordinates there.
This operation removed the perturbation problem of the centerline in Revit.
Thus, the centerline could be correctly placed with the right geographical po-
sition.
31
Figure 18: Correct position
The first modeling step included creating the bridge deck profile and using
this profile as the foundation for a beam cross section which in turn was used
to create a new beam family. The beam family was then used to insert a beam
object of that family along the centerline, simply by clicking the centerline
while having the beam selected. The centerline imported from Civil 3D was
somewhat longer than the bridge itself, and known elevations along the center-
line were used to identify the correct start and end point of the bridge in the
model. A 2D .dwg file containing the outlining of the bridge including founda-
tions, columns, coordinates etc. was then imported to the model and moved
to the correct spot for further detailing of the bridge. The coordinates of this
file did not match the centerline’s coordinates. This was caused by slight in-
accuracies at the ends of the centerline when creating the bridge beam, which
resulted in the beam starting with an offset of approximately 15mm in the
X-direction and 70mm in the Y-direction (see figure 19) from the start of the
centerline. The reason for this was probably a slightly wrong shape of the
centerline.
32
Figure 19: Wrong coordinates, N=X, Ø=Y
Although this effect is not desirable, the problem was easily fixed by simply
altering the coordinates of a known point in the model to the ones specified
in the 2D-file. In the end it was concluded that the process of importing the
bridge centerline was successful, despite the hiccups encountered along the
way.
33
4.5.1 Bridge deck and abutment back walls
As mentioned in 4.4, the bridge deck was created as a beam along the cen-
terline. After some consideration, the abutment back walls were created the
same way. To separate the bridge deck from the abutment back walls, refer-
ence planes were used at the locations where the bridge deck ended and the
back walls started and ended. These reference planes defined the positions of
vertical cuts such that the extents of the visibility objects were contained by
the reference planes. The objects themselves still had the original extents, but
with the exception of some behavior issues (e.g. 5.3), they responded as if
they had the correct extents. By creating reference planes one obtained a way
to control the extents of the objects without altering them directly, and they
could be used as references for other objects to attach to or be dimensioned
with respect to. Another reason to model the bridge deck and abutment back
walls in this way was that the ends were supposed to be vertical. If vertical
reference planes were not used, the objects would have a slight slope at the
ends relative to the vertical plane.
Figure 21: Reference planes cutting the bridge deck and abutment back wall
Mostly standard objects were used in the modeling of the columns and foun-
dations. A new column family was made to model the columns at piers 1 and
4 (i.e. the piers at each end of the bridge), however, this family was quickly
34
establishable. What posed the most problems were the connections between
the columns and the bridge deck. One had the option of creating a new column
family with the connection attached, a new family including only the connec-
tion, or to create the connections through objects already available. In this
bridge model, the connections were simply modeled as a new family consisting
of two square cubes, seeing as a correct representation was not important for
the model as a whole. If the connections were to be modeled properly, a more
detailed family would be needed to display the connection. The question of
whether or not to include the connection in a column family would remain.
The connection family in this bridge model was not included in a column fam-
ily. If it was included, one would have to customize the family such that the
top elevation of the column could be specified while at the same time leave the
connection free to be attached to the bridge deck. This is possible; however by
letting the connection out of the column family, the extra work was avoided.
Additionally, by modeling the connection separately, it could be added to new
columns in a quicker fashion if needed.
There are four wing walls at each end of the bridge, reaching out about 2
meters from the abutment back walls. They were all created as individual wall
objects without a direct connection to the centerline, but instead attached to
the reference plane defining the end of the abutment back wall. Based on
the drawings received from Norconsult, the wing walls curve along with the
centerline of the bridge and are supposed to create a smooth transition from
the back wall. This caused some modeling issues:
• If a wall object curves, one cannot make use of the “edit profile” tool,
which lets the user customize the contour of the wall object as seen from
35
an elevation view (e.g. 21). The shape then needs to be created through
the use of void forms. Void forms are separate geometrical objects hidden
in the view unless selected. By letting a void form intersect another
object, one can choose to cut the area of intersection from the object.
This process is more cumbersome than the “edit profile” approach, and
it is harder to create exact geometry.
• The curved wall objects did not always attach to the reference plane
properly, resulting in a gap between the back wall and the wing wall.
Using straight wall segments solved these problems, and also made the
model more flexible. The question is whether this is an acceptable simplifica-
tion. The effect of the curvature of an object with a length of 2 meters and
a radius of about 600 meters is very small. However, formwork for the wing
walls is usually created as an extension of the bridge deck, such that it is ac-
tually easier to continue the curvature. Also, the aesthetics of the bridge are
improved by keeping the wing walls curved. The user has the option of creat-
ing a comment describing that the walls are supposed to curve, but it is not
preferable to resort to that solution as the model should be as visually correct
as possible. The creation of the correct geometry at the top of the wing walls
in order to obtain a smooth transition from the back wall was particularly
difficult. As well as having a slope along the length of the bridge, the bridge
deck and abutment back walls have a slope at the top in the cross sectional
direction. Regardless of how the walls were modeled, void forms were needed
to display this correctly. These void forms were not attached to the wing walls
or the bridge, rendering them inflexible.
The transition plate was perhaps the easiest component to model. It was
created by adding a wall object with a width of 3200mm, which was the width
of the transition plate, to the basic wall family. The wall object could then be
placed at the correct position, and with the use of the ”edit profile” tool (see
4.5.3), one was able to provide the correct geometry of the transition plate.
The ledge on which the transition plate was to be placed was modeled using an
”In-Place Mass” created by the sweep tool. The sweep tool lets the user pick
or define a 3D path for a 2D profile to follow, thus creating a 3D shape. By
defining the bottom edge of the transition plate situated next to the abutment
back wall as the path, and creating the appropriate profile, one was able to
easily model the ledge object.
36
Figure 23: Transition plate and ledge
37
Figure 24: Profile selection when transferring the model to Robot
The consequence of this was that the model in Robot was left without
sections for the user defined objects (figure 25), which in turn rendered the
model useless for structural analysis (figure 26).
38
Figure 26: ”Section not defined” error when attempting to analyze the model
5 Design Changes
An important aspect of BIM is the relations between objects; by defining
connections and constraints, objects behave in a certain way when altering
objects they are related to. Which objects are related and how the relationships
are defined influences the model greatly when modifying the model. In a bridge
project there will be changes throughout the design, and thus a need to make
changes in the model. Efficient change management is a highly appreciated
feature in engineering projects. A central question is how well Revit is able
to deal with changes. Are the objects related to the change flexible enough to
behave appropriately, do they behave incorrectly, do they need to be adjusted
or do they need to be recreated altogether?
Is Revit flexible enough, and how does different modeling approaches affect
the objects’ response to the changes? The ideal situation would be having a
model where all the objects behaved perfectly to every change according to the
user’s preferences. This is not possible, both because of lacking constraining
possibilities and the fact that while constraining the objects to behave cor-
rectly for one change, the very same constraint may prevent the objects from
behaving correctly for another change.
Furthermore, there are almost always several ways to perform a task/create
the geometry of a model, and how the model behaves when changes are made
39
may vary depending on modeling choices made before the stage in which the
user performs the changes. Thus, in order to achieve a flexible model, the user
needs to keep in mind the most probable changes while modeling. In Revit,
this is usually approached by attaching objects to either levels, reference planes
or to each other. If done correctly, the result is flexible objects that respond to
some of the foreseen changes appropriately. This becomes increasingly difficult
when dealing with non-standard geometry, especially curved geometry. In
order to align objects lines or faces to each other, those objects need to be able
to align completely. When operating with curved lines/faces, there are very
often slight differences in the orientation of the lines/faces, rendering alignment
impossible or causing unwanted results. Sometimes one has to sacrifice some
of the authenticity of the model in order to attain flexibility, which is not
desirable. The question needed to be asked in this regard is how much the
model can deviate from the correct geometry while still being valid as a basis
for the actual construction.
Revit may parameterize objects and create smart interplay between ob-
jects, but this does not mean adjustments can always be made easily and
automatically. One also has to consider whether or not it is appropriate to
model for automaticity in each case, since it often requires a lot of time. How-
ever, once the changes have been made, the model is updated in all the views in
the project, removing the cumbersome process of redesigning the model wher-
ever the change have an effect. Combined with the fact that material usage
and other useful information are automatically updated and can be extracted
from the model, this provides a great advantage compared to traditional design
methods in which 3D and 2D views are created separately.
This thesis includes the testing of the following design changes:
1. Length
2. Position
3. Slope
4. Radius
For each of the design changes one has the possibility of reworking the
centerline in Civil 3D before importing it into Revit yet again and create the
bridge deck beam along this line. The surrounding objects will need to be
reattached, but in some cases this provides less work than customizing them
to an altered bridge deck within Revit.
40
5.1 Length
A change in the bridge length means a change in the bridge deck and abutment
back walls objects(1), a positional and rotational change of the piers(2) and
the abutment wing walls(3).
1. As mentioned in 4.5.1, both the bridge deck and the abutment back
walls were created as beams with the bridge centerline as a reference,
and cut at the endpoints using reference planes, removing the extents of
the bridge beyond the reference planes. These reference planes can be
moved and rotated, resulting in a longer/shorter bridge deck.
In figure 27 the top and the bottom wing wall were created without cur-
vature whereas the middle two were. All of the wing walls are aligned with
the reference plane in the first state. Both the top and the bottom wing walls
were set to have a length of 2000mm, and the top wing wall was constrained
angularly to always remain perpendicular to the reference plane. The reference
plane was first moved in order to elongate the bridge before rotated.
41
Figure 27: Wingwalls and reference plane at the low end of the bridge
When moving the reference plane the top and bottom wing wall behaved
appropriately, while the constraints of the curved middle wing walls were vio-
lated.
42
Figure 28: Moving the reference plane
If the wing walls were created without curvature, aligned with the reference
plane and constrained angularly, they were able to follow the reference plane
appropriately both when moved and rotated. This can be seen in figure 29;
after the rotation, only the top wing wall, which was constrained angularly,
followed the rotation of the reference plane. The rotation is exaggerated in
order to display the effects. An interesting feature discovered in this regard
was the fact that if only the top wing wall was constrained angularly, the center
of rotation for the reference plane used by Revit was not the one specified by
the user. The reason for this remains unclear, but it is not a desirable feature.
However, if both the top and the bottom wing wall were constrained, the center
of rotation specified coincided with the one used by Revit.
43
Figure 29: Rotating the reference plane. The top wing wall is rotated
Again the option of selecting the objects and moving them separately is
available.
5.2 Position
Intuitively, one might consider locking the bridge in position relative to the
bridge centerline file imported from Civil 3D, and use the centerline to move the
entire bridge. However, after the bridge deck was created with the centerline
as a reference, there was no way of constraining the bridge deck to follow
the centerline. As an effect, the imported centerline had served its purpose
and was obsolete in terms of usage in further modeling processes. A way of
somewhat constraining the bridge deck to the centerline was to create a model
line (a line only available to be created in the horizontal and vertical plane
which is visible in all views and can be used as a reference for other objects)
and constraining parts of the bridge to it. The main problem with this was
the fact that one could not align and lock the bridge deck’s actual centerline
with the model line since the bridge deck had a slope. Thus, the easiest and
most logical way to change the position or rotation of the centerline was to
simply select the whole bridge with all of its components and move/rotate it.
In figures 30, 31 and 32, the blue and yellow lines are the ones included in the
44
2D underlay file showing the outlining of the bridge, while the bridge itself has
black lines.
45
Figure 31: The bridge moved north along the road
In this case two dimensions became invalid, meaning one or more of the
rules which the dimensions need to follow were violated as a result of the
change. The user is able to view which dimensions that have become invalid
46
and decide whether or not to go through with the operation (i.e. deleting the
dimensions). Often, the dimensions are not critical to the model, and can be
recreated if so desired.
5.3 Slope
A change in the slope of the bridge results in an elevation change at the
high end of the bridge. The slope change of the bridge deck itself can be
done by altering the elevation offset of the high end, which is a simple task.
The problem in this context was that even if the bridge deck’s visibility was
constrained by the vertical cuts at the ends, the object itself still remained
beyond these cuts. The result was that the offset values chosen were related
to the ends of the object, not the ends of the bridge deck as it was seen in the
model window. Thus, the bridge deck rose at the low end as well. In figure
33 the blue line represents the length of the bridge deck object, which is cut
where the abutment backwall starts.
47
each wing wall. The abutment back walls will need to have altered elevation
offsets at their ends too. If objects are connected to certain levels or reference
planes, these can be changed to adapt to the change in elevation of the high
end. Since the abutment back walls were modeled the same way as the bridge
deck, the extents of the abutment backwall objects needed to be set equal to
the extent of the bridge deck object for coherency between the bridge deck
and the abutment back walls. If not, slight inaccuracies were seen as a result
of this difference. A drawback by this approach was that the two abutment
backwall objects had the same extents, thus overlapping, which might cause
complications when interacting with objects at a later stage. In figure 34 the
inaccuracy effect is presented; the left side of the backwall is slightly higher
than the bridge deck at the interface between them, whereas the right side is
slightly lower.
The operation of changing the slope of the centerline might have been done
more efficiently by changing the centerline in Civil 3D. This would remove the
need for offset changes of the objects in Revit.
5.4 Radius
This is also a case where changing the centerline in Civil 3D could be a good
choice. The abutment wing walls will only need to be rotated in order to
remain perpendicular to the ends of the abutment back walls, and the piers
will need to be reworked and/or moved slightly no matter how the bridge
48
deck attains its new radius. Another approach is to create a model line in
the horizontal plane with the new radius as a reference for a new bridge deck
beam, before setting the correct height offsets at the ends of the bridge deck
beam. This is a very quick operation and gives the same result as importing
the centerline again. If the height at each end of the bridge deck remains the
same, the slope along the centerline will receive a minor change. Depending
on the desired slope, the heights might need to be changed as well.
Figure 35: Piers at the high end before the change in depth
49
Figure 36: Piers with changed depth as displayed in the field Elevation at the
top left corner
50
• Good modeling solutions take time to establish, but it may be worth
it in the long run. The knowledge gained can be transferred to other
situations and projects.
• Crude solutions like simply moving and rotating objects manually are
often just as good as creating rules for the objects to follow given a certain
change, especially if there are few objects that need to be manually
altered.
• In the case of a new slope and a new radius, returning to Civil 3D and
modify the centerline there could be a good idea. After the center-
line has been placed and given the correct topographical placement, the
bridge deck and abutment back walls are then reestablished by simply
clicking the objects onto the line, and new cuts to these objects are rela-
tively quickly made. This ensures the centerline position is correct, and
the same operations would have to be made to the surrounding objects
regardless of whether the change was made through Civil 3D or in Re-
vit. As mentioned in 3.2, Sundt Construction usually makes the whole
bridge superstructure in Civil 3D. If modeled that way most of the design
changes would have to be done in Civil 3D.
• Even if some of the design changes may require as much or even more
work in Revit than they presently require in programs like Microstation,
the benefit from the automatic update of the changed objects in all views
of the model makes the workload less in almost any case.
51
the need to add several rebars in addition to the ones placed with the other
reinforcement tools.
52
Figure 38: Area reinforcement tag
53
Figure 39: Revit reinforcement extension, beam
A lot of options, explained with figures, are available for quick establish-
ment of rebars. The user is able to enter this menu again and make changes
to the reinforcement.
54
Figure 40: Rebar options
To place rebars in an object, the user has to create a section that cuts
through the object in which the rebars should be inserted. The plane the
section in which the rebars were created defines represents the work plane for
the rebars inserted while in that section. The rebars can be inserted parallel
to that work plane, parallel to the cover, or perpendicular to the cover (again,
see figure 40). Several sections in different directions and positions are often
needed in order to reinforce an object, and the user needs to be careful where
the sections are placed related to rebars already placed in order to see them
in that section.
55
Figure 41: Rebar with shape handles
With the diameters and lengths of the rebars shown properly, the user is
able to effectively move the rebars around and fit them next to other rebars
appropriately. When rebars are inserted using one view, those rebars are also
visible in the other views given the view range spans the area where the rebars
are situated. The user also has the possibility of changing the position, length,
spacing etc. of the rebars from other views, where the need for changes might
be more apparent. If the rebars are selected in a view before switching to
another view, the rebars are still displayed as selected in the second view.
56
Figure 42: Rebars, elevation view
The rebars are also included in the 3D view, where the user can zoom in
and inspect the very details of the reinforcement to verify the correct position
of every bar.
57
Figure 44: Rebars, 3D view
This is very useful, not only to spot errors that would be difficult to discover
by looking at 2D-drawings; a 3D view of the rebars provides a good overview,
which makes the people involved in the reinforcement work quickly get an idea
of the reinforcement placement.
Rebars bent in one plane are easily modeled with the rebar tool. The only
real issue with the reinforcement of the substructure was the rebar sets that
needed to curve. The rebars anchored in the foundation slabs and going up
into the columns, following the curvature of the columns were such rebar sets
(Numbers K24328 and K24329 in figure 45).
These rebars could not be modeled with the use of the Rebar Tool and the
Rebar Set Menu, as the spacing cannot follow a curve. Revit has a tool called
“Array Group” which is suitable for this operation. By creating a bar at the
right position, the user can then click the “Array Group” tool, choose a center
of rotation around which rebars are to be placed. Then the user specifies the
58
span of the rebar set and the number of rebars in the group, creating the
appropriate distance between the rebars. Figure 46 shows the array group of
rebars from a top view of the column.
The number of rebars as well as the span range can be modified at any
time. Each rebar in the rebar group can also be modified separately. Some
customization is needed in order to get the right spacing between the rebars,
as the spacing depends on the span range and number or rebars. The process
would have been made easier by allowing the user to select the spacing between
the objects in the array group, which at this time has to be determined through
extra calculations.
In Revit, the rebar cover (the distance from the edges of the concrete elements
that defines a portion in which there should be no rebar) is defined in the
concrete object itself, and the rebar cover area will be outlined with a green
dashed line once a rebar object is selected, or a rebar object is to be created.
Upon creation, the rebar objects snap to the dashed line and cannot be placed
in a way resulting in a violation of the rebar cover distance. However, when
created, the user is able to move the rebars around regardless of the rebar cover,
yet the objects still snap to the dashed line for the user’s convenience. One
thing to note about the rebar cover distance is that the rebar objects in Revit
59
do not include the ribs, thus only the nominal diameter defines the rebar’s
extents. In Norway, the rebars are named after their nominal diameter, yet
the real extent of the rebar is taken into consideration when inserting rebars.
Figure 47 shows the nominal diameter and the full diameter including the ribs.
The x in Øk x represents the nominal diameter, and the number below is the
real extent of the rebar.
When inserting rebars snapped to the rebar cover, they will be placed with a
distance according to the nominal diameter, in practice resulting in a violation
of the rebar cover distance. In figure 48 a rebar with a nominal diameter of
16mm is snapped to the rebar cover, which is set to be 75mm. Revit calculates
the placement of the center of the rebar to be 83mm (75mm + 16mm 2
) from the
edge of the concrete object, when it in fact should be 85mm (75mm + 20mm 2
).
The user is able to manually change the distance of the rebars after insertion
in the model, but this is cumbersome. The rebar cover distance can only be
specified to certain preset values in the properties palette (every 5mm ranging
from 10-100mm), thus on is not able to specify these values accordingly for
the concrete object. Also, the rebar cover distance needs to be different for a
rebar with a nominal diameter of 16 and a rebar with a nominal diameter of
20. It is possible to edit the diameter of the rebar objects in order to represent
the correct real extent in the model. However, by doing this, the calculation
of the amount of rebar will be false. Revit keeps track of each rebar in the
60
model, and the user is able to quickly establish a sheet of the material usage,
reinforcement area of the bars etc., which is a really useful tool and renders
the tedious manual calculation obsolete. If the rebar area is not represented
correctly, the calculations will not be correct.
61
Figure 49: The original rebar drawing
The drawing is as clean as possible, usually only showing single rebars in the
rebar sets. The extents of the rebar sets are indicated by lines, and the amount
of rebars in each rebar set is shown connected to the rebars that are displayed
in the drawing. The region consisting of diagonal lines represents an area with
five rebar sets going into the plane downwards through the foundation.
All the rebars created are included in Revit, which means that the same
section as shown in figure 49 will include an abundance of rebars, making it
hard to get an idea of where each rebar is situated. Also, lines showing which
rebars the rebar tags are referring to will virtually be unspottable. Figure 50
shows the section in Revit corresponding to the drawing in figure 49.
62
Figure 50: Top view of the pier in Revit
63
and these are the objects used in this thesis. After some cleanup work, the
section provides a better overview of the reinforcement of the pier.
Figure 51: Top view of the pier in Revit, after cleanup and annotation
There are still some adjustments that could be made to this view; the rebars
going downwards from the column anchored in the foundation are all displayed
in the view on the right hand side of the column. These could be displayed as
done in figure ??, and on the left hand side of the column. However, the rebars
are already created in Revit, and the procedure of representing them this way
only provides extra work hiding them in the view and inserting an additional
detail component. When the rebars do not coincide with other objects in the
view one might as well keep them as long as the correct tag is placed. There
are a couple of other things to note about this view. The rebars anchored in
the foundation, which are situated at the ends of the column, following the
curvature, could not be represented with the rebar tool alone (as described in
6.4). The rebar set had to be created using the “Array Group” tool, which has
a drawback with regard to the rebar drawing. When attempting to tag the
rebar set, the rebars were treated as separate. As a work around, one has the
possibility of manually inserting a text annotating the rebar set properly. The
64
rebars are included in the overall schedule of rebars, such that calculations will
be correct. Additionally, they have the same schedule mark/position number.
As a result, the only part of the model affected by this work around is this
very view, as the text will not update automatically if changes are made to the
rebar set. Nevertheless, the best case scenario would be that the rebar tool
could successfully represent rebar sets following a specific curve. Furthermore,
the array group should be able to be tagged as a whole. The other thing
realized through tagging the rebars was the fact that several rebar sets could
not be tagged together. The region represented by the diagonal lines includes
five identical rebar sets. In the original rebar drawing (49), these are combined
and shown as 35 [5ø16-24210]c500. A solution is to tag all of the rebar sets
individually; however there should be a way to include all of the rebar sets in
one tag. Additionally, the spacing in Revit (51), where the spacing listed is
the one between each rebar in the rebar set, differs from the one displayed in
the original drawing.
65
Figure 52: Rebar list sorted by schedule mark
The top row shows a problem related to the rebars created as array groups
in order to represent rebar sets (see section 6.4). The groups are not included
correctly in the rebar list, nor are they assigned with a correct schedule mark
by Focus Software’s schedule mark function. In the list in figure 52, all the
rebars in the array group are included, with the right bar diameter and shape.
However, the bar length and volume fields are blank. The reason was discov-
ered to be that there were two array groups, both having the same schedule
mark but different lengths of the rebars. When assigning the rebars in one of
the array groups with a new schedule mark, the rebars were displayed correctly
in the rebar list.
66
Figure 53: Array group rebars displayed correctly in the Rebar list
The solution to the problem was simple, however, if the user frequently
needs to resort to array groups instead of using the tools already provided for
rebar by Revit, the result is a lot of additional work in order to both represent
the rebars and keep track of the schedule marks.
In addition to this list, Focus Software has developed a tool which estab-
lishes a rebar bending schedule showing information about the rebars including
geometry. The rebar bending schedule automatically gathers the information,
and a simple click on the tool generates the whole schedule. If rebars are
selected, only the selected rebars are included in the schedule. The rebar
bending schedule shown in figure 54 included 60 different schedule marks, and
was created within seconds. The reduction in the amount of work compared
to traditional methods, where this would have to be done manually, is huge.
67
Figure 54: Excerpt of a rebar bending schedule
68
Figure 55: Rebar set does not curve with the bridge deck
69
in this regard is if the representation of sets of 12000mm straight rebars is
satisfactory. At the ends of each rebar set, there will be an inaccuracy because
the next set will have a different orientation in the horizontal plane.
An issue arose when trying to add rebar to the bridge deck; the rebar
was not placed onto the work plane defined by the section created for the
insertion of the rebar, but ended up at a different location seemingly without
any connection to the section. In figure 57 the section and the added rebar are
displayed. The section is indicated by the two arrows and the blue dashed line,
and the rebar is the blue line in the bottom section of the picture (the rebar
is hovered over with the mouse, which reveals the object name and type).
70
Figure 57: Section and rebar not coinciding
The solution was to create a reference plane where the rebar was supposed
to be inserted and set that reference plane as the work plane before entering
the section and placing the rebar.
Figure 58: Overview of the PTR running along the length of the bridge
71
Standard rebar shapes are not applicable for this reinforcement, and ten-
sioned reinforcement behaves differently and redistributes stresses and strains.
The sketch tool does not work, because it is confined to a plane and has limi-
tations determined by the section in which you draw the rebar, and there is no
possibility to attribute a rebar curvature in two planes with this tool. Thus,
other methods need to be considered. It is possible to create this kind of geom-
etry using the ”In-Place Mass” tool which is mentioned in 4.2.1. A portion of
one of the PT tendons was modeled using this tool. Points were created along
the length of the rebar with a spacing of 1000mm and given approximately the
right elevation. For each point, both a reference plane and a section had to
be created in order to add a circular profile to each point. The sections were
created to view the specific point, and the corresponding reference planes were
needed for the profiles to be placed on. This way one was able to create a path
of circular profiles for an ”In-Place Mass” to follow.
72
be adjusted to create the desired geometry. Once the reference planes and sec-
tions are placed in the model, they can be used for each of the post tensioning
reinforcement objects since they follow the same curvature, which means they
only need to be created once for all of the post tensioning reinforcement.
73
Figure 61: ”In-Place Mass” points and profiles
It was concluded that this method was not feasible for the modeling of the
PTR. However, there are modeling engineers who have worked with PTR in
Revit, and they have managed to come up with families from which objects
can be inserted [15]. These families are not yet in the standard Revit library,
and require more than a moderate amount of skill to model. The PT tendons
in the examples shown of these families are limited to one or two spans with a
smooth curve from top to bottom, which require 3 and 5 points respectively.
The PTR in the bridge modeled in this thesis has area where the reinforcement
flattens out altogether before rising towards the columns, thus more points are
needed in order to represent this correctly. Since the geometry of the PT
tendons differs from bridge to bridge, making a family customizable to each
reinforcement case is a difficult challenge. Sundt Construction modeled the
PTR in Civil 3D (3.2), where geometry curving in several planes is more easily
represented. Even though the objects are not parameterized to the same extent
as in Revit, and does not hold the same interplay with Revit objects, this might
be a good alternative.
74
lists and schedules, renders several cumbersome manual operations obsolete.
The collaboration between other software and Revit works satisfactorily.
Road and underlay models can be imported as aid for continued modeling in
Revit; however, they need to be appropriate for BIM purposes (i.e. consist
of continuous line segments). This thesis does not explicitly investigate the
representation of the model created in Revit with other disciplines, however
the export of Revit models to the .dwg file format for usage in for example
Novapoint
R Virtual MapTM or Autodesk
R NavisworksTM (programs for 3D
representation of the collaboration of different disciplines like structural, ar-
chitectural and MEP) works well. Norconsult is using Revit files exported to
.dwg in Novapoint Virtual Map, and Navisworks can even handle Revit files
(.rvt) [16].
In terms of the modeling in Revit, there are still issues; non-standard ge-
ometry is hard to represent, work with, dimension and reinforce. These issues
need to be remedied, especially if Revit is to be used for bridge design without
aid from other programs. Companies that use BIM extensively in their bridge
design tend to not only model in one program, but utilize the different soft-
wares’ advantages. Sundt Construction uses Civil 3D for the modeling of the
bridge superstructure, while the substructure along with different documenta-
tion are covered in Revit.
The use of Robot was not studied extensively in this thesis; however, it
seems that non-standard geometry also makes the incorporation of Robot dif-
ficult. If the interaction between Revit and Robot was investigated further,
solutions might have been found, and a skilled user may be able to perform
analysis of bridges with non-standard geometry. Nonetheless, there are com-
plications related to this process. Not being able to transfer the bridge model
made in Revit for structural analysis in Robot effectively is a serious drawback.
If the solution is to resort to other programs, extra time will need to be spent
creating a new model for structural analysis in those programs. Furthermore,
one will not gain the ability to automatically alter the Revit model based on
proposed changes by Robot after the structural analysis.
75
the bridge. The point elevations are in fact both exactly at the edge of the
wing walls when looking perpendicularly at the bridge at that point, but since
the bridge curves and the centerline lies beyond the wing wall it looks like the
point lies to the left of the wing wall in the BIM.
76
sections. That way the user would be able to add a section with the same
curvature as the bridge with an offset and thus attain the same elevation view
as in the drawing. The question in this regard is whether or not it is necessary;
BIM brings new challenges to the table, but also possible improvements and
opportunities to think differently. In this particular problem one could agree
that it would be useful to be able to display the point elevation exactly at the
end of the wing wall, and therefore the traditional representation has some
positives. However, all of the information is available in the model! It is just a
matter of how it should be presented. For instance, the point elevation might
just as well be added in the plan view, where the length of the centerline
between specified points could also be added. This would leave the correct
horizontal length of the bridge to be displayed in the elevation view.
• The line was not properly shaped for the use of the arc length tool (i.e.
that its radius varied).
• The intersecting lines or objects might not have been appropriate for the
arc length tool.
• The elevation change throughout the centerline might have caused issues
for the tool.
The next step in the investigation was the creation of a model line (a line
that is visible in all views, and assigned to a given plane) as a projection
of the centerline on the horizontal plane. This did not provide the wanted
77
results. Finally, a model line was created in the horizontal plane by manually
typing in the radius of the arc and adjusting it to fit the centerline of the
bridge, or rather the projection of the centerline. This line would let itself be
measured by the arc length tool, thus most probable cause of the problem was
the centerline itself not being a proper arc. However, the centerline was able
to be clicked while having the arc length tool selected, which in theory should
only be possible for arcs.
7.4 Reinforcement
The reinforcement of standard objects that are not curved works really well in
Revit; rebars are quickly established, and the rebar functions Revit holds are
sufficient. However, there is a need to improve the flexibility of the rebar tools
for objects with curved geometry. Several operations attempted in this thesis
were rendered impossible or cumbersome as an effect of the lack of functions
provided by the tools.
• Curved rebar sets: rebars created through the tools provided by Revit
can be curved in one plane. However, when creating a rebar set based
on a rebar, the spacing between the rebars can only have an orientation
directly outwards from the plane in which the original rebar was placed.
This caused issues when creating the reinforcement for the columns (6.4)
and the bridge deck (6.7). The solution was to either split up the re-
bar sets, creating several straight segments to represent a curve, or to
represent the rebar sets with array groups. Neither of these options is
desirable, as splitting up the rebar sets results in wrong rebar placement,
and the array groups are harder to work with and makes the model slow
to work with.
78
8 References
[1] National BIM Standard Project Comitte-United States, 2012
The National BIM Standard-United StatesTM ,
Available at:
http://www.nationalbimstandard.org/about.php
[Accessed 2013.01.27]
[6] Bentley
,
R 2012.
LEAP bridge in Microstation. [digital print]
Available at:
http://communities.bentley.com/other/old_site_member_
blogs/bentley_employees/b/skans_blog/archive/2009/02/05/
leap-bridge-images.aspx
[Accessed 2013.01.27]
[7] Bentley
,
R 2012.
AECO Information Modeling and CAD Production Software,
Available at:
http://www.bentley.com/en-US/Products/MicroStation/
[Accessed 2013.01.27]
79
[8] Tekla
,
R 2012
Bridge Design by WSP Finland,
Available at:
http://www.tekla.com/international/solutions/references/
Pages/bridge-design.aspx
[Accessed 2013.04.23]
[10] Det Digitale Byggeri, 2010 Pihl bygger bro med BIM i Sverige,
Available at:
http://www.detdigitalebyggeri.dk/case/
pihl-bygger-bro-med-bim-i-sverige
[Accessed 2013.02.04]
80
Available at:
http://vasshaug.net/2013/02/27/post-tension-structures-in-revit/
[Accessed 2013.04.27]
81