You’ve built the model in your go-to software, the geometry is correct, the loads are on, and the analysis runs without errors. But deep down, there’s a nagging question: does this model truly represent how the building will behave? Gaining real confidence in our structural models means moving beyond the default settings and challenging the convenient assumptions we often start with. It’s about embracing the philosophy of engineered precision found in Part 4 over the prescriptive paths of Part 9.

This gap between a “correct” analysis and a “confidently correct” design is where great engineering happens. The computer will always give you an answer, but it’s up to you to decide how much to trust it. An effective model isn’t just one that accurately represents the primary load paths; it’s one that is also “relatively easily modified to investigate the structure’s sensitivity to changes in various input parameters”.

Here are a few critical areas to reconsider in your next structural model to transform it from a simple code-checker into a powerful predictive tool.

1. Foundation Rigidity

One of the biggest assumptions we make is at the lowest point of the structure: the boundary conditions. We often simplify our foundation connections as either perfectly pinned or perfectly fixed. In reality, they are almost always somewhere in between. Neglecting this can have a significant effect on your frame’s stiffness and the distribution of forces.

So, how do you test this assumption? You bracket the solution.

The Pinned vs. Fixed Check

This is the classic first step. Run your analysis twice:

1. Run #1: Assume all column bases are pinned. This will typically maximize your frame’s lateral drift and minimize your foundation moments.
2. Run #2: Assume all column bases are fixed. This will minimize your drift but maximize your foundation moments and potentially the moments at the first level of your columns.

Pro-Tip: The results from these two runs give you an envelope. Your structure’s actual behaviour will lie somewhere between these two extremes. If your members work under both scenarios, you can have high confidence. If a member is failing in one case but not the other, it tells you that the design is highly sensitive to your foundation assumption, and you need to investigate further.

Getting Closer to Reality with Soil Springs

A more refined approach is to use soil springs. Your geotechnical report will often provide a modulus of subgrade reaction (k), which has units of force per unit area per unit deflection (e.g., kPa/mm). You can use this to calculate translational and rotational spring constants for your footings.

For Canadian practice, the Canadian Foundation Engineering Manual is your go-to reference. While detailed soil-structure interaction is a complex field, modeling your footings with springs derived from the geotech report gives you a much more realistic starting point than a simple pinned or fixed assumption. This is a core concept in specialized analyses like the practical guide to seismic flexible retaining wall design, where the soil’s behaviour is just as important as the wall’s.

The key takeaway is to understand how sensitive your superstructure is to what’s happening at the foundation. Varying your boundary conditions from pinned to fixed to springs is the best way to find out.

2. Diaphragm Rigidity

Diaphragms are the unsung heroes of the lateral load resisting system (LFRS). They collect wind and seismic forces and distribute them to your frames and shear walls. But how they distribute those forces depends entirely on their in-plane stiffness.

The NBCC allows us to classify diaphragms as flexible, rigid, or semi-rigid. This isn’t a property of the material itself, but rather its stiffness relative to the stiffness of the vertical supporting elements.

• A rigid diaphragm, like a thick concrete slab in a high-rise, distributes lateral load in proportion to the relative stiffness of the vertical LFRS elements (e.g., shear walls, braced frames).
• A flexible diaphragm, like a typical plywood or steel deck roof on a low-rise building, distributes load based on tributary area, like a series of simply supported beams.

Assuming the wrong type can lead to gross errors in how forces are shared between your frames and walls.

The Rigid vs. Semi-Rigid Check

If there’s any doubt, a sensitivity check is in order. This is especially critical in buildings with mixed LFRS systems or irregular layouts.

1. Run #1: Assume a rigid diaphragm. In your software, this is often a simple checkbox or constraint. Look at the shear forces distributed to each wall and frame.
2. Run #2: Model the diaphragm with its actual in-plane stiffness (semi-rigid). This means using shell or plate elements that have been assigned properties representing the deck. For steel decks, the CSSBI’s S37 Design of Steel Deck Diaphragms provides guidance on calculating the flexibility factor, F, which you can use to find the equivalent shear area (\(A_{v}\)). For wood, the CSA O86 and the Wood Design Manual are your guides. This step is especially critical for complex structures that push the boundaries of Part 9 and require a more detailed Part 4 analysis.

Key Takeaway: You’ll often find that the stiffest vertical elements (like concrete shear walls) attract significantly more load under the rigid diaphragm assumption. If you have a long, narrow steel deck diaphragm, a flexible or semi-rigid assumption might show that the frames closest to the load take the brunt of the force, regardless of their stiffness. Knowing which assumption governs is critical for an efficient and safe design.

3. Probing the Load Path

An effective model must accurately represent the primary load paths. But how robust are those load paths? What happens if one element is compromised or just not as effective as you assumed? A great way to build confidence in your system’s redundancy is to selectively turn members “off.”

This doesn’t have to be a full progressive collapse analysis. It can be a simple series of “what-if” checks to understand how forces redistribute.

Simple “What-If” Scenarios:

• Braced Frames: In a multi-story X-braced frame, what happens if you remove one of the tension braces at a lower level? Does the load successfully transfer through the beam to the adjacent brace, or does it cause a catastrophic failure in the beam or its connections? This can help you identify if your beams need to be designed as part of a Vierendeel system in a brace-out scenario.
• Moment Frames: You’ve designed a moment frame assuming fully rigid beam-column connections as per CSA S16. What happens if you release the moments at a few key connections to simulate a more flexible, partially-restrained condition? How much does the building drift increase? This tells you how critical that connection’s rigidity is to the overall frame stability.
• Shear Walls: In a system with two coupled shear walls, what’s the effect of reducing the stiffness of the coupling beams? This simulates the impact of cracking under severe load as per CSA A23.3. You can check how much shear is redistributed to the individual wall piers.

Pro-Tip: These checks aren’t just for a final design. Doing them early helps you identify the most critical elements in your LFRS. The elements whose removal causes the biggest problems are the ones that deserve the most attention during detailing, review, and site inspection. This is how you move beyond simply following the code to developing true engineering judgment.

Confidence Through Questioning

A structural model that runs without errors is just the starting point. True confidence comes from understanding not just what the results are, but why they are what they are. By systematically questioning your assumptions and testing the sensitivity of your key parameters—foundations, diaphragms, and critical members—you move from being a software operator to an engineer who has a deep and intuitive grasp of the structure’s behaviour.

What’s the one parameter you always check in your models? Share your go-to sensitivity check in the comments below.

Disclaimer: This blog post is for informational purposes only and should not be taken as specific engineering advice. Always consult the latest edition of the National Building Code of Canada and relevant CSA standards for your projects.