CIV102 Matboard Bridge Design
CIV102 · Structures & Materials · University of Toronto
Engineered a simply-supported matboard box-girder bridge designed to withstand a 452 N dynamic point load. Developed a Python structural analysis pipeline to evaluate 8 distinct failure modes across 5 major iterations, achieving a theoretical capacity of 1125 N and sustaining a 1250 N cumulative load in physical testing.
CTMFs
Iteration and refinement · Trade-offs · CAD (Onshape)
At a glance (one-pager)
Compression governed at FOS 2.49 · Physical Testing Validates 450 N Sustained Load
The computational model predicted a theoretical failure load of 1125 N based on a 2.49 factor of safety. In the lab we stepped through increasing point loads: 133 N, 266 N, 400 N, and 450 N, with failure at 520 N and a cumulative load total of 1250 N. Ultimate structural failure occurred exclusively at the physical matboard splice joint, proving the primary cross-section design successfully resisted all calculated buckling and shear modes. Full design rationale, iterations, and calculation detail appear in our team design report [3].
Matboard Sheet
One rectangular sheet per team (32 in × 40 in); approximate mass 750 g per the course handout [2].
Matboard Thickness
Nominal 0.05 in stock; governs all section properties.
Load Case 1
Unweighted train; total weight split evenly across six axles [2].
Load Case 2
Example base-case loaded train from the handout (e.g. 182 N locomotive + 2 × 135 N freight cars) [2].
Final Web Height
Converged after five iterations; maximises I and shear buckling within deck-height rules.
Max Diaphragm Spacing
Raises critical shear buckling stress.
The objective was to computationally design and physically fabricate a high-capacity beam bridge operating under strict geometric and material constraints. The project demanded a rigorous balance between bending stiffness, shear resistance, and physical mass optimization using only mm thick matboard and contact cement, as specified in the course design brief [2].
Process, tools & reflection
The physical bridge ultimately failed at a splice joint that none of our models captured. Going into this project, I was already aware that models are inherently simplifications of reality, and I was exploring pragmatism after our praxis project. However, I didn't truly realize the danger of relying on them until I watched our bridge collapse under the limitations of imperfect construction quality.
This failure reinforced the perspective I was introduced to in Praxis I [1] because it showed me that the math did not matter if we ignored human imperfections. It highlighted that true engineering design begins where the theoretical model ends, turning the exploration of pragmatism I had going into the project into a fundamental value. It taught me that my duty as an engineering student is to apply human judgment to bridge the gap between models and actual constructability.
CTMF names and working definitions follow Praxis I and II lecture materials (slides), except where the narrative notes a different source.
Iteration and refinement
ConvergeApplication
Iteration and refinement in this case refers to the use of scripts and loops to mathematically optimize different parameters. Before fabrication, we ran five major computational iterations using a Python script that re-evaluated eight failure modes while plotting shear and bending envelopes as flange build-up and web height changed. This iterative loop became our convergence engine, letting us immediately see how each geometric adjustment shifted the governing failure mode under Load Cases 1 and 2 [2].
Assessment
This approach advanced my agnostic (impartial) lens on knowledge as it guarded our team from anchoring to widening or optimizing one sole aspect of the bridge. Because every change could be evaluated quickly, we were able to refine the bridge geometry systematically instead of relying on intuition. The process ultimately converged on a configuration that balanced bending resistance with plate buckling limits while remaining within the material constraints of the matboard sheet.
Trade-offs
ConvergeApplication
To distribute material effectively, we conducted a parameterized trade-off analysis mapping how widening the flange improved compression while sharply reducing Case 3 plate buckling resistance. This directly guided our decision to reallocate horizontal material into a 120 mm vertical web, significantly increasing bending stiffness without triggering plate buckling.
Assessment
Making these relationships visible clarified how each geometric decision affected the competing failure modes. It was highly effective for our team, providing an objective way to navigate our strict material constraints. I will definitely rely on it to justify resource allocation in future projects, provided that I ensure to consider fabrication tolerances.
CAD (Onshape)
RepresentApplication
Computer Aided Design (CAD) creates digital 2D or 3D geometric representations used to validate design aspects. Our teammate Karan used Onshape to build a 3D model of the final bridge, defining diaphragm spacing and cross-sectional geometry. Translating the parameters from our Python analysis into a spatial model made the design legible and allowed us to plan the cutting layout on the 813 × 1016 mm matboard sheet [2].
Assessment
The CAD model was incredibly useful as it allowed us to coordinate the construction of our physical structure. This tool was the perfect fit for our representation phase, providing a precise digital blueprint to guide our physical fabrication. Moving forward, catching spatial conflicts digitally will be a non-negotiable step before I make any physical cuts. However, I must remain conscious of designing geometry that exceeds human precision limits, a lesson to remain pragmatic.
- [2] Univ. Toronto, Fac. Appl. Sci. Eng., "CIV102 Matboard Bridge Design Project," Toronto, ON, Canada, CIV102 BDP handout, Fall 2025, Rev. 2, Oct. 30, 2025. PDF
- [3] D. Angelo, K. Chawla, C. Peng, and A. Xiao, "CIV102 Bridge Project Design Report," design proj. rep., Team 701, Univ. Toronto, Toronto, ON, Canada, Nov. 23, 2025. PDF
Numbers match the full portfolio reference list.