Efficient Steel Bridge Design & Construction Using Collaborative Fabrication Models 

WSP is using OpenBrIM to model steel girder bridges with improved efficiency and accuracy.

Blueprints, like cassette tapes and Blockbuster Video are now relics of the past. The transportation infrastructure industry must undergo a similar paradigm shift to make use Building Information Models (BIM) that have the potential to make bridge design more accurate, fast and dynamic.The Federal Highway Administration (FHWA) is leading an effort entitled BIM for Infrastructure that enables users to exchange data from one discipline to the next, indicating who is building what, when each part will be built, the materials to be used, and how it will be constructed.

Typical transportation infrastructure projects use roadway design software such as Bentley OpenRoads Designer or Autodesk Civil 3D for creating the roadway geometry. Recently, WSP began using the software platform OpenBrIM for the digital deliver these bridges with a dynamic link to the roadway geometrics. This model-centric workflow using OpenBrIM will provide dramatic improvements in both efficiency and consistency in delivering the traditional deliverables of plans and quantities, but this is only the beginning of the advantages of the 3D bridge models. This article is focused on the potential advantages of using collaborative design and fabrication models in lieu of the traditional workflow process, as shown in Figure 1. 

Figure 1  Traditional vs collaborative shop drawing workflow.

https://www.fhwa.dot.gov/construction/bim/pdfs/integrated_digital_project_delivery_fhwa_hif_20_021.pdf

The traditional shop drawing workflow involves two independent data sets for design and fabrication with coordination between the two sets relying on 2D shop drawings. This process is rooted in mid-20th-century technology when paper drawings were the necessary tool for communication. The collaborative model-centric work-flow is not new, and it has a proven track record, and WSP has groups that are using it with great success. However, there are significant differences in both the nature of the contracts and the type of structures as compared to the traditional infrastructure projects. Both the challenges and the potential advantages of using collaborative design and fabrication models for the anticipated curved steel bridges are discussed below.

LOD 400 Fabrication Models

The in-model review process is not new or untested, and two case studies will be presented. Both of these models are considered Level of Development (LOD) 400 models as defined by the American Institute of Architects (AIA).

“An LOD 400 element is modeled at sufficient detail and accuracy for fabrication of the represented component. The quantity, size, shape, location, and orientation of the element as designed can be measured directly from the model without referring to non-modeled information such as notes or dimension callouts.”

https://bimforum.org/resource/level-of-development-specification/

Both of the case study models were developed using Tekla Structures. The success of the LOD 400 modeling is hard to deny. The model provided numerous benefits such as — improved collaboration, enhanced visualization, and error reduction.

A. 66 Hudson Yard

WSP Hagerstown developed the fully connected Tekla model for the owner on the 66 Hudson Yard project (Figure 2). Also known as the Spiral, this landmark project presented many challenges due to the unique geometry around the perimeter of the building and the thin interior core walls. Through the modeling, WSP worked directly with the designers and connection engineers to provide solutions for complex nodes, sloping columns, and unique truss layouts. As the project developed and a fabricator was chosen, our office customized and modeled connections to the fabricator’s preferences to speed up production and the approval process.

Figure 2  Model and photo simulation of 66 Hudson Yard.

With Trimble Connect – markups, to-dos, 2D drawing review, clash sets, and other coordination were done using color representation for an overall visual effect. The steps in the coordination process included: 

  1. Cloud mark-ups in the model
  2. To-Do Lists with assignees and status columns
  3. 2D drawing review similar to the traditional style 
  4. Clash sets, which identify conflicts that are much better resolved in the design phase

The 66 Hudson Yard model was also used for production control during fabrication, and the Numerical Control (NC1) files exported from the model were used to run the CNC equipment. This provided tremendous value in both efficiency and quality.

  • B. Chenab Bridge in India

A Tekla Structures model was also used for the Chenab Bridge as part of a new railway line between Udhampur, Jammu and Baramulla, Kashmir in northern India (Figure 3). Fabrication drawings of all the structures were extracted directly from the model. Whenever the model changed, the drawings changed as well. The bridge BIM models were a great advantage on site as well. The site crew was able to get additional manufacturing information from the models. Additional details on the use of the models for this project can be found in the e-brim article. BIM-for-Bridges-May-2023.pdf (e-brim.com)

Figure 3  Tekla model of the Chenab Bridge.

We are seeking to gain similar advantages for typical curved steel bridges by modeling them in OpenBrIM. The questions that arise in comparing the 66 Hudson Yard project and the Chenab Bridge to the land bridges are: 

  1. Can models created in the design phase deliver similar advantages to models created for construction and fabrication?
  2. Does this process apply to typical highway bridges, or is it only justified on complex projects?

Bridge Fabrication models

In order to evaluate the applicability of collaborative design and fabrication models, we need to examine the additional factors that steel bridge fabricators considered in their models, which designers may not traditionally include. These adjustments are honed by years of experience and are often unique to particular fabrication shops.

A. Vertical and Horizontal (Camber and Sweep) Geometry

The IBR project has many curved ramps with curved horizontal alignments along with parabolic vertical profiles. These girders also need to be cambered so that the final configuration with a dead load matches the design profile. This is critical for fabrication since bolted connections require tight tolerances for “fit-up” at specific loadings and associated deflections.

B. Fabrication Adjustments Made to Design Data

The fabrication models for steel plate girders also need to include adjustments for fabrication that are not typically included in design models. These include: 

  1. Welded Plate Girder Web Camber, which requires added flange length for the milling caused by the weld shrinkage. 
  2. Bearing Stiffeners Milled to Bear with added stiffener length to be milled to fit between the flanges. 
  3. Weld Joint Details, which involve chamfering plates to accommodate the welding process.

C. Multiple Models Based on Designed Fit Conditions

Steel bridges have multiple deflections, as shown in Figure 6, where two stages are shown superimposed to highlight the difference.

Figure 4  Stupp Bridge example.
  • Red steel is the Total Dead Load representation (Final Position)
  • Green steel is the Steel Dead Load (SDL) Fit (Cambered) representation

Figure 5a gives a view of a girder splice. You can see the splice plates are only modeled in the cambered position (green); this way they show up on our girder drawings. Notice that there is some overlap of the bolts on the stiffener as well.

Figure 5  Final vs. cambered model position.

Figure 5b shows a head on view of the girders and crossframes. It’s visible from the picture that the crossframes have been detailed in SDL fit condition and do not line up with the cambered stiffeners. This is a good view to note the extra holes (bolts) that had to be added to the stiffeners on the cambered girders. We need to model these so that the holes come in correctly on drawings and bolt lists.  

It may seem like finding Shangri-La to get design models capable of including the needed parameters to account for the additional needs of fabrication models. In the next section, we will discuss the capabilities of OpenBrIM and how realistic it is to account for these details.

OpenBrIM Data-Centric and Parametric Bridge Workflow

OpenBrIM is web-based software that you can access in a web browser. It includes 3D modeling, Finite Element Analysis (FEA), AAHTO Code Check, CAD and Quantities, Load Rating, Inspection and Health Monitoring (see Figure 6).

Figure 6  Openbrim data-centric & parametric bridge workflow.

A. 3D Modeling and FEA

The 3D Detailing Model and the FEA Analytical Model must have distinct characteristics. In a traditional bridge workflow, the analytical model and the 2D CAD details are in separate data repositories and require two sets of data input, introducing the potential for inconsistencies. OpenBrIM effectively shares data between the two 3D models and eliminates the possibility of inconsistencies between the models.

The FEA includes moving loads based on influence surfaces and transfers the vehicle loads to the substructure for complete bridge design in one summary report (see Figure 7). The analytical model can be exported to various software packages, including LARSA, Csi Bridge, Midas, STADD Pro, and OpenSees, among others, for additional analysis as well. 

Figure 7  Finite element analysis.

B. Code Check

The engineer of record on bridge projects has the responsibility to ensure that all the applicable code requirements are met. In OpenBrIM, the national design code with any state amendments are library components that are easy to modify. The Summary Report is dynamic and is always updated based on the model properties (see Figure 8).

Figure 8  Summary report of code check.

C. Parametric 2D Drawings and Quantities

Perhaps the most valuable use of the OpenBrIM model is the automated production of 2D Plans. Although OpenBrIM does not include CAD capability, it can dynamically link 2D drawings to reference sheets in MicroStation. This way, all the clients’ MicroStation CAD requirements can be met including using the full capabilities of MicroStation, such as “named boundaries” and intelligent sheet borders. 

OpenBrIM is also being used for dynamic quantity reporting including a full embodied carbon analysis. These quantities will be exported to various databases with additional attributes, such as design notes and specifications. 

D. Fabrication Models

The above features more than justify the use of OpenBrIM, but creating an exchangeable model for fabrication has the potential to change the industry. It has been said that the design and fabrication models must be distinct; however, the value proposition makes it worth continuing our pursuit.

None of the needed model features in the fabrication model are beyond the capabilities of OpenBrIM. However, we are not getting the full support of the fabricators that are involved in the National Steel Bridge Alliance (NSBA). There are three possibilities we see for why this is the case:

  1. There are fabrication model needs that are beyond the abilities of OpenBrIM that we have not yet learned.
  2. The industry is committed to the use of the Industry Foundation Class (IFC) format that is being developed by Building Smart.
  3. The fabricators get paid to create the fabrication models, and they don’t want to give that up. This is understandable, as they have invested years to develop these bridge modeling capabilities.

Despite the obstacles, we are committed to developing these fabrication models in OpenBrIM since we are required to deliver models that should meet LOD 400 standards and reflect the fabricators processes.

Conclusion

There are clearly challenges in developing collaborative design and fabrication models for curved steel girder bridges, including the unique processes that fabricators have developed over the years and their trade secrets. We are hoping that the continued “openness” that we experienced in recent workshops and that seems to be a characteristic of our industry will prevail and those who share the most will be rewarded the most.

Building Smart has been endeavoring for years to identify everything that is needed for the IFC bridge format. The philosophy of this project is essentially the exact opposite. We are using OpenBrIM to parametrically model a curved steel girder bridge and make as many iterations as necessary to converge on a solution that is exchangeable based on feedback from fabricators. OpenBrIM has demonstrated the ability to make the adjustments to the model geometry quickly when they know what is needed. Obviously, the amount of feedback we have received from fabricators is a small sample size compared to the large number of processes that exist. It may be naïve to believe that we can account for all the potential combinations, but through persistence, it is achievable, and that the return on investment will add value to our industry for years to come. 

The National Steel Bridge Alliance is the key organization to facilitate modeling all the different fabrication processes. OpenBrIM can develop a “suite” of parameters that can be turned on or off depending on the specific fabricators who will ultimately be involved in the project. It would help if there was a convergence of fabrication practices in the coming years to arrive at a more-or-less standardized fabrication practice. We recommend that a fabricator representative be included in the design team to facilitate integrated project delivery.

Whether our objective is achievable, difficult, or impossible seems to depend not so much on the task itself but rather on who is making the evaluation.■

About the author  ⁄ Douglas J. Dunrud P. E.

Doug Dunrud, P. E., is Vice President and Technical Principle at WSP in Sacramento. He has been active in the development of Building Information Modeling (BIM) for bridges for the last 20 years.(Douglas.Dunrud@wsp.com)

STRUCTURE magazine