Jack Campbell

Step 1 - Generative Design Framework

  • Design Decision 1: Setting bridge specifications to maximize the number of satisfied users.
    • Design variables:
      • Number of lanes. More lanes can increase usage but require greater size, could reduce aesthetics, and may diminish user experience.
      • Type of lanes. Lanes can serve various transport modalities. Lane type modulates accessibility, user satisfaction, safety, and number of users.
      • Incline angle. Steeper bridges might be able to cross longer distances while remaining passable but could worsen pedestrian experience.
      • Height. Same principle as for incline angle: higher bridges can span longer distances but become dangerous or impassable for pedestrians and cyclists.
    • Evaluators:
      • Surface area of roadway. Greater surface area likely translates to increased usage and satisfaction but also greater cost and reduced aesthetic appeal.
      • Amount of traffic. Measuring the number of users can be done in practice by counting or in a model by estimating traffic rates for different lane types.
      • Degree of curvature. Allows a bridge to span greater distances but decreases accessibility and worsens experience for traffic without motorized vehicles.
      • Height. Same principle as for degree of curvature.
    • Most important trade-offs to consider:
      • Speed of transit vs. pedestrian experience. Faster travel by automobiles and rail is more efficient and increases usage but harms pedestrian experience.
      • Bridge size vs. cost and aesthetics. Larger bridges can cross greater distances and serve more users but are liable to cost more and obstruct views.
      • Multimodal transportation vs. efficiency. Providing for more transit options on a increases accessibility but decreases overall usage and speed of travel.
      • Accessibility vs. safety. Allowing for multimodal transit enhances accessibility but increases risk of dangerous accidents (i.e. between automobile and cyclist).
  • Design Decision 2: Selecting bridge materials to minimize environmental impact.
    • Design variables:
      • Function. The type of traffic a bridge serves determines its structural requirements and the range of serviceable materials.
      • Location. Where a bridge is sited influences the materials that can be used.
      • Length. Longer bridges likely require stronger materials while shorter bridges might have more freedom in selecting materials.
      • Materials for various components. The materials themselves could be the main determinant of the environmental impact, and can vary by component.
    • Evaluators
      • Life cycle assessment. This will measure the embodied carbon of bridge materials including extraction, manufacturing, use phase, and end-of-life.
      • Load capacity. Sustainable materials still need to support all loads the bridge will bear, so this captures any trade-off between sustainability and strength.
      • Expected lifespan. Maintenance/replacement frequency of materials has implications for cost and sustainability.
      • Cost. Sustainable materials could be more expensive and not worthwhile from the developer's perspective.
    • Most important trade-offs to consider:
      • Sustainability vs. cost. Sustainable materials are likely to be more expensive so will need their increased cost to be outweighed by their environmental benefits.
      • Sustainability vs. strength. Safety depends on load-bearing reliability, so sustainable materials must still have sufficient compressive and tensile strength.
      • Sustainability vs. availability. Sourcing sustainable materials might require extending project deadlines if those materials are less available.
      • Lifespan vs. aesthetics. Aesthetic materials might not last as long, affecting sustainability.
  • Design Decision 3: Choosing the most aesthetically-pleasing yet structurally-sound bridge form.
    • Design variables:
      • Width. A bridge's width is closely tied to its function, and different functions serve different types of traffic which require different load capacities.
      • Number of supports. Different numbers of supports are needed depending on environment and bridge function.
      • Number and type of overhead structures. Trusses, towers, cables, and cantilevered arms are parts of different bridge forms.
      • Degree of curvature. If a bridge needs to elevate (i.e. to avoid obstacles or to reach a greater distance) that likely eliminates certain candidate forms.
    • Evaluators
      • Cost. Price surely varies across bridge forms and influences choice of form.
      • Volume. Measures the amount of materials required, perhaps negatively correlated with environmental impact and positively correlated with cost.
      • View obstruction. Forms with overhead structures will obscure views of the surroundings but could also look good.
      • Structural integrity. Varying load capacities across forms will influence choice depending on bridge function.
    • Most important trade-offs to consider:
      • View obstruction vs. aesthetics. In beautiful natural settings developers might prefer bridges without overhead structures, and vice versa.
      • Aesthetics vs. strength. Preferential bridge forms may not be structurally sound.
      • Cost vs. load capacity. It may be expensive to increase a bridge's strength for certain forms, leading to selection of the strongest form that minimizes cost.
      • Cost vs. accessibility. Certain bridge forms are likely more amenable to certain types and volume of traffic as well as multimodal transportation.

Step 2 - Generative Design Study

I decided to run a Generative Design study with Design Decision #1: Setting bridge specifications to maximize the number of satisfied users. The design variables I included were length, width, midpoint elevation, and number of supports per side. These specifications could modify the bridge for different purposes, making the bridge smaller and flatter to favor pedestrian traffic or making the bridge larger and more curved to favor long-distance motorized traffic. The 4 design variables were varied within arbitrary but realistic value constraints: length from 400 ft to 800 ft at increments of 50 ft, width from 24 ft to 72 ft at increments of 12 ft (multiples of 12 ft to ensure the roadway always accommodates an integer number of lanes since 12 ft is the recommended width of highway lanes), midpoint elevation from 0 ft to 50 ft at increments of 5 ft, and number of supports per side from 2 to 10 at increments of 1. Constants were the height and thickness of the bridge as well as estimated material cost and traffic rate (determined via online statistics that are described in the additional submission document). The evaluators are construction cost to estimate cost based on volume, traffic flow to estimate daily bridge usage based on traffic rate and the number of lanes, pedestrian passability based on traversal length and incline, and structural integrity based on the roadway surface area relative to the number of supports. Equations and logic behind the evaluators are described in the additional submission document. Construction cost directly translates to cost and should be minimized, traffic flow directly measures daily usage and should be maximized since more bridge usage is a positive, and the final two evaluators are unitless and should be numerically minimized despite being named after positive attributes. These evaluators trade-off with each other: more traffic flow requires a wider bridge which hurts cost and structural integrity, more supports helps structural integrity but hurts cost, etc. Pedestrian passability is unrelated to traffic flow and better pedestrian passability is actually better for cost and structural integrity but is nonetheless an interesting measure. All 4 evaluators were used as goals (with respective minimization and maximization objectives) for the study so that each would be accounted for. The study was run as an optimization with other settings default.

Step 3 - Generative Design Study Results

  • Screenshot of results of Generative Design study:
  • image
  • Explanation of Generative Design results:
    • The results of the Generative Design study include numerical the values for the inputs and evaluators for each alternative, a scatterplot of the evaluator values for the alternatives, and a Dynamo visualization of the currently-selected alternative. The scatterplot captures structural integrity (to be minimized) as the vertical axis, traffic flow (to be maximized) as the horizontal axis, pedestrian passability (to be minimized) as the size of the data points (smaller size indicating lesser value), and construction cost (to be minimized) as the color of the data points (dark-blue is greatest cost, green is middling cost, and red is lowest cost). It is evident that better traffic flow correlates to worse structural integrity and worse cost but better pedestrian passability. It also seems like traffic flow and cost are tightly related since discrete values for traffic flow mostly correspond to discrete cost values. On the other hand, for the most part structural integrity has little impact on cost except in the case of the worst-possible traffic flow. Structural integrity and traffic flow (the latter estimating bridge usage and thus how useful its construction was) are likely the most important evaluators, so the design decision should be from the bottom-right quadrant of the plot. Cost is likely the next most important evaluator followed by pedestrian passability, so the decision should be from a green-red data point and if possible from a smaller data point. With this in mind, I perceive the selected alternative in the above screenshot as optimal.
  • Image of Dynamo study graph:
  • image