TL;DR
This article provides a rubric for structural engineers to evaluate Scan to BIM deliverables. It covers critical factors from point cloud quality and control network accuracy to BIM model fidelity for structural members. The guide details how to verify member sizes, check for analytical model readiness, and ensure the model is prepared for coordination with other disciplines. It emphasizes the importance of understanding tolerance stacking, material assignments, and the documentation of assumptions for concealed structural elements. This framework helps teams confirm that the as-built model is a reliable foundation for structural analysis and design in renovation or retrofit projects.
# Key Factors in Scan to BIM for Structural Engineers
TL;DR
- Assess point cloud density and accuracy at key framing lines and connections.
- Verify the global accuracy of the scan data via the control network closure report.
- Confirm modeled member sizes against direct point cloud measurements.
- Check that Revit families and material assignments match as-built conditions for the specified LOD.
- Ensure the model's analytical lines are clean for export and the model is geolocated for coordination.
Jump to:
How is point cloud quality verified for structural analysis? · What defines a reliable control network? · How should tolerance stacking be managed in a structural BIM model? · What is the process for verifying structural member sizes? · How is BIM model fidelity evaluated for structural elements? · Is the model ready for structural analysis? · How is the model prepared for coordination with other disciplines? · How should concealed structure be documented? · Where to go next
This article provides a quality assurance framework for structural engineers and BIM coordinators who receive Scan to BIM deliverables. It is not a guide for selecting a partner, but rather an internal rubric to apply when validating that a delivered point cloud and Revit model are fit for purpose in a structural retrofit or assessment project. The goal is to confirm the as-built data is a reliable foundation for analysis and design.
Applying a consistent evaluation process ensures that the investment in Scan to BIM services yields a trustworthy digital asset. While a firm may use a companion guide for selecting from among different Scan to BIM providers for structural teams, this rubric is the next step: verifying the final product. It outlines specific checks for point cloud data, model accuracy, and overall usability for structural engineering workflows.
How is point cloud quality verified for structural analysis?
The foundation of any reliable Scan to BIM deliverable is the point cloud. For structural purposes, its quality is measured by density, completeness, and cleanliness. Density must be sufficient to clearly define the geometry of structural members, including flanges, webs, and connections. This is especially critical at gusset plates, moment connections, and column base plates, where precise geometry is required for analysis. Check these areas specifically to ensure there are no data gaps or "holidays" caused by line-of-sight obstructions during scanning.
Cleanliness refers to the absence of noise, stray points, or temporary objects that were present during the scan. A quality point cloud has had these artifacts removed. The data should represent only the permanent built condition. Verification involves visually inspecting the cloud in software like Autodesk ReCap or Navisworks. A comprehensive point cloud QA process confirms that the data is fit for modeling.
Accuracy is the final check. Review the scan registration report provided by the scanning firm. This document should show that the cloud-to-cloud error between individual scan setups is low and that the overall registered point cloud accuracy is within specification, typically ±5mm. Using a mix of hardware, such as a NavVis VLX3 for rapid capture of large areas and tripod-mounted terrestrial scanners for high-detail setups at critical connections, allows providers to balance speed with precision.
What defines a reliable control network?
While point-to-point registration ensures local accuracy between adjacent scans, a survey control network provides global accuracy across an entire site. This is non-negotiable for large facilities, such as an 80,000–120,000 sq ft industrial plant or warehouse, where registration errors can accumulate over distance, causing the point cloud to drift, twist, or bend.
A reliable control network is established by a field crew using survey-grade instruments to place and measure control targets throughout the facility. The laser scans are then tied to these known coordinate points. The defining document is the control network closure report. This report demonstrates that the survey traverse started at a known point, looped through the site, and "closed" back on the starting point with a statistically insignificant amount of error.
When reviewing deliverables, ask for and inspect this report. A well-executed control network ensures that a column on one side of the building is correctly located relative to a column on the other side. This global accuracy is essential for verifying building plumb, detecting overall structural deformation, and ensuring that any new structural elements designed in the model will fit correctly when fabricated and installed. The process of point cloud registration is foundational to a trustworthy model.
How should tolerance stacking be managed in a structural BIM model?
Tolerance stacking is the accumulation of acceptable variances that can lead to an unacceptable overall deviation. In Scan to BIM, this includes scanner measurement uncertainty, registration error, and modeling interpretation. While a registered point cloud may have a stated accuracy of ±5mm, this applies to individual points, not necessarily the modeled element derived from them. A BIM model is an idealized interpretation of imperfect, real-world conditions.
Managing this requires a practical understanding of how tolerances compound. The modeler makes best-fit judgments to place a column centerline or establish a floor elevation. To verify these judgments, the structural reviewer must spot-check critical dimensions. Do not rely solely on the model; measure directly between points in the point cloud to validate overall building dimensions, floor-to-floor heights, and primary grid spacing.
Compare these direct point cloud measurements to the corresponding dimensions in the Revit model. Any significant discrepancies should be questioned. The goal of a Scan to BIM tolerance analysis is not to achieve impossible perfection but to ensure that the model's geometric idealization remains within an acceptable envelope of deviation for structural design and coordination purposes.
What is the process for verifying structural member sizes?
Accurately identifying the sizes of existing structural members is a primary function of Scan to BIM for retrofits. Verification of the modeled elements is a straightforward but critical quality check. The most effective method is to use a 3D section box in Revit or Navisworks to isolate an individual member (beam, column, or brace).
With the element isolated, orient the view to look down its profile. The point cloud should appear as a fuzzy outline around the crisp lines of the modeled family. This visual overlay immediately reveals how well the chosen profile fits the scanned data. Check that the flange width and depth of the modeled section align with the densest areas of the point cloud. Measure the point cloud directly in several locations to confirm the dimensions and account for fabrication tolerances or coating thickness.
This check should be performed on a representative sample of members of different types and sizes. Pay special attention to any elements identified as non-standard, built-up, or showing signs of damage or deflection. A reputable provider will document their methodology for size identification and flag any members where the size could not be determined with high confidence from the scan data.
How is BIM model fidelity evaluated for structural elements?
Model fidelity refers to how well the BIM represents the actual physical asset, both geometrically and informationally. For a structural model, this means using the correct Revit families, assigning appropriate materials, and modeling to the specified Level of Development (LOD). Most structural Scan to BIM projects are delivered at a level between LOD 200 and LOD 350.
To evaluate fidelity, first check that the correct families are used. A wide-flange beam should be modeled with a "W-Shape" family, not a generic box extrusion. This ensures that the section properties and analytical parameters are correct. Second, verify that the placement and orientation of members accurately reflect the as-is condition, capturing the actual load path of the structure.
Finally, inspect the embedded information. Are material properties like "Steel, ASTM A992" assigned correctly? Are elements placed on the correct worksets and categorized properly? A high-fidelity model is not just a geometric representation; it is a structured database. Understanding the differences between LOD 200, 300, and 400 is key to specifying and verifying a deliverable that meets the project's needs without being over-modeled.
Is the model ready for structural analysis?
A primary use of a structural as-built model is as a starting point for analysis. A model's readiness for this workflow can be quickly assessed within Revit. Switch to the "Analytical Model" view to see the simplified, single-line representation of the structure used for analysis software.
In this view, check for connectivity. The analytical nodes of beams, columns, and braces should be coincident. Gaps or misalignments between these nodes will cause errors when the model is exported to programs like RISA-3D, RAM, or ETABS, requiring significant manual cleanup. The analytical lines for beams and columns should align with the member centerlines unless offsets are intentionally modeled to represent eccentricity.
A clean analytical model is a sign of a thoughtful modeling process. The provider has not just traced the point cloud; they have constructed a model with an understanding of its end use for structural engineering. If the analytical model is messy, disconnected, or nonsensical, the deliverable will require substantial rework before any analysis can begin, diminishing the value of the Scan to BIM process.
How is the model prepared for coordination with other disciplines?
A structural model rarely exists in isolation. It must coordinate with architectural, MEP, and other models. Preparation for this is a key quality factor. The first step is to confirm the model is built on the correct shared coordinate system. When the structural model is linked into a master host model, it should land in the correct position and orientation without manual adjustment.
The model should also be lean and efficient. Check that it has been purged of unused families, views, and links. The file size should be reasonable for the scope of the project. A bloated file can slow down the entire coordination process.
Finally, the model's accuracy directly impacts coordination. Because the structural, architectural, and MEP models are derived from the same point cloud, they should align perfectly. This shared origin is the "single source of truth" that makes clash detection meaningful. When evaluating the structural deliverable, it is a good practice to link it with the other trade models to confirm this alignment. Reliable as-built accuracy in design coordination prevents costly conflicts during construction.
How should concealed structure be documented?
Laser scanners can only capture what is visible. Structure concealed behind finishes, above hard-lid ceilings, or within concrete slabs will not appear in the point cloud. A quality Scan to BIM deliverable must have a clear and consistent method for documenting how these inferred elements are handled.
When reviewing the model, look for the system used to differentiate scanned reality from educated assumption. This can be achieved through several methods:
- Revit Phases: Placing inferred elements on a separate "Assumed" or "Existing-Inferred" phase.
- Design Options: Using design options to isolate assumed content.
- Custom Parameters: Applying a text or checkbox parameter to elements to flag them as "Inferred" vs. "Scanned."
- Graphic Overrides: Using specific line types (e.g., dashed lines) or colors for elements that were not directly captured.
Regardless of the method, it must be applied consistently throughout the model. Furthermore, the deliverable should include documentation—either as notes in the model or a separate PDF—explaining the basis for the assumptions. This could be information from selective demolition, GPR scans, or original building drawings. This transparency is crucial for risk management and informs the engineering team where further investigation may be needed.
Evaluation Rubric for Structural Scan to BIM
| Factor | Why It Matters Structurally | How to Verify in the Deliverable |
|---|---|---|
| Point Cloud Quality | Defines the accuracy of member geometry, connections, and placement, which is the basis for all analysis. | Check registration report for ±5mm accuracy. Visually inspect for density, completeness at connections, and lack of noise/ghosting. |
| Control Network | Ensures global accuracy across large sites, preventing drift that misrepresents building plumb, layout, and deformation. | Review the control survey closure report for minimal error. Verify alignment of distant elements in the point cloud. |
| Member Sizing | Incorrect member sizes lead to invalid analysis results and incorrect load capacity calculations. | Use section boxes to visually compare modeled profiles to point cloud data. Spot-check dimensions with direct measurements in the cloud. |
| Model Fidelity (LOD) | Ensures the model contains the necessary geometric and informational detail (e.g., materials, profiles) for analysis and documentation. | Confirm correct Revit families are used. Check for proper material assignments and that the LOD (200-350) matches the scope. |
| Analytical Readiness | A clean analytical model saves significant time by allowing direct export to analysis software without extensive rework. | Inspect the analytical model view in Revit. Check for coincident nodes and correct centerline alignment. |
| Coordination Readiness | A properly located and lean model is essential for multi-disciplinary clash detection and integrated design. | Confirm the model uses shared coordinates. Link with other trade models to verify alignment. Check for reasonable file size. |
| Assumption Documentation | Clearly identifies areas of uncertainty where structure was not visible, managing risk for the design team. | Look for a consistent system (phases, parameters) to flag inferred elements. Review accompanying notes explaining the basis for assumptions. |
Where to go next
This internal rubric provides a framework for validating Scan to BIM deliverables. By systematically checking these factors, structural teams can ensure the as-built model is a trustworthy asset for design, analysis, and coordination. For teams looking to specify these requirements upfront, it is helpful to understand how to write a 3D scan deliverable specification that clearly defines these expectations for a provider.
The insights gained from a verified model are a critical input for any building scanning for structural assessment project, enabling more accurate analysis and better-informed design decisions for retrofits and renovations.