Concept Bridge for Rural Applications

The Problem:

Hello, My name is Cadel Conner and I am going into 10th grade. Behind my house on a dirt road is a bridge. Located on a Level B road, this bridge has slowly deteriorated over time. Farmers drive their expensive, heavy farm equipment over this bridge repeatedly during planting and harvest. The bridge is made of timber, and when the boards of this bridge began to rot, farmers became warry of the bridge. But to bypass this bridge they would have to drive miles on the highway. As a Level B road, the county has not invested in improving this bridge, making minimal repairs and replacing timbers as needed. The Level B road beside my house is also a popular road for cyclists and runners. The timber bridge has wide gaps between the timbers: wide enough for bike tires to get stuck in.

The Solution:

I decided to investigate the possibility of replacing the timber bridge with a prefabricated, 3D-printed bridge that would be more durable, require less maintenance, and be friendlier to cyclists and runners who want to enjoy the quiet rolling hills of this road. Manufacturing a 3D-printed bridge and transporting it to our town (or any town with the same challenges) has many advantages. The bridge building process could be made much more efficient by surveying sites and feeding the surveys into a manufacturing pipeline where bridges are manufactured in a factory and transported to the installation site. Plastic is abundant and decays slowly meaning this bridge would be low maintenance. Using generative design can reduce design cost and time while also ensuring the design is as strong as needed and as efficient as possible.

Overview:

To complete a proof of concept for this solution I broke the task up into several steps. First, I knew I would need to have an accurate survey of the bridge site so I could provide all the correct data to a design team. To do this I created a 3D scan of the site and imported it into Fusion 360. The next step was to create a bridge within the scan using generative design. Then the generative design was optimized. Finally I 3D-printed a model of the bridge to test the printability of the design.

Drone with camera (optional)

3d Printer (optional)

Fusion 360

Blender

measuring tape

I created a 3D scan of the bridge site using a drone and photo stitching software named WebODM. First, using the drone I took pictures all around the bridge at a 45 degree angle. The pictures were then imported to WebODM which stitched them together into a 3D rendering of the site. I then imported the 3D rendering into Fusion 360 and scaled it using measurements taken on the bridge.

In Fusion 360 I built a basic frame for the bridge with just a top and two sides that would abut the same riverbank walls of the original bridge. You can see this basic frame on the left side of the included image. This frame included no supports and simply gave generative design a starting point for building structural elements to provide the needed strength. Here is an explanation of the necessary data fields for generative design in Fusion 360:

• Study Settings

This brings up a slider that can determined how detailed we want our model to be. Default value is good

for most applications.

• Preserved Bodies

These are the bodies we are connecting or strengthening using generative design.

• Obstacle Bodies

These bodies help dictate where the generated structures can and cannot go.

• Starter Shape

Generative design starts its process by filling or connecting the shape. Usually this starter shape is very rudimentary and will take more carving away to optimize. Optionally, the user is able to create shapes that the computer will start with instead.

• Structural Constraints

These can tell where and how your model is supported. In my case I selected all faces touching the river bank as fixed structural constraints because they would be supporting the forces both vertically and laterally.

• Structural Loads

This tells the computer how much load the structure needs to support, and which face will support the load. In my case I selected the bridge's top face and gave it a 50,000 lb load. (I chose this after doing a little research on the weight of a typical combine harvester.)

• Objectives and Limits

We have the option between minimizing the mass of your generated object or maximizing the stiffness and setting a safety factor. The Safety Factor just multiplies all the loads by itself and should be used to create safety headroom.

• Manufacturing

Here we set our manufacturing method. For this application I selected 3D-Printing. Do not forget to set your tool direction. The tool direction specifies in what direction (bottom-up, left-right, etc) the model will be printed.

• Study Materials

Set your desired material from a library made by Fusion 360, whether it be plastic or metal.

Once you have filled out these fields you should use the checklist feature in Fusion 360 Generative Design to make sure your study is read for generation. Finally. click Generation to launch the generative design. Here Is a video tutorial that I found useful.

https://www.youtube.com/watch?v=PSSt8wswNJQ

Combining the 3D scan of the site and the generated bridge design created a finished product that I am proud to show off. To do this I imported both the new bridge and 3D scan into Blender. First I alligned the two models then removed from the 3D scan all faces that made up the old bridge. Finally I extruded the scan downward To make a solid mesh that could be edited. Then came a grueling process of cleaning loose faces, patching holes, and sculpting until the finished model was done.

Unfortunately I was unable to test the strength of my prototype because I do not have a 3D printer capable of printing nylon (the material I used for this bridge) and the physics would not scale. But I still 3D printed the bridge for a few reasons:

1. To reveal any artifacts in the model.
2. To test whether all overhangs were 3D-printable.
3. To visualize my model.
4. To create a model to demonstrate the viability of 3D printing bridges.

Throughout the project I ran into some challenges that I want to mention here in the hope that it could help someone replicate this project.

• My computer was too slow for WebODM.

I tried various ways of running WebODM on my own computer, but in the end I just don't have a powerful enough computer. To solve this problem I found and used a cloud service that runs WebODM by the name of WebODM Lightning.

• The generative design was adding material on top and well outside the bridge

Initially after running the generative the design the model was extending over and beyond where the bridge should be. I discovered that it is necessary to wrap the bridge in an obstacle body that sets boundaries for where the generative design can build the model.

• The generative design was far too bulky and couldn't be optimized

Because I selected 3D printed manufacturing the generative design was accounting for printing overhang and would not remove support material. I discovered I needed to configure the tool direction under the Manufacturing settings to have the generative design account for the overhang in the right direction.

• Generative design left the model with a high FOV(factor of safety) and did not fully optimize, which resulted in a very bulky design

After struggling to figure out why the designed bridge had so much mass which couldn't be reduced, I found that I needed to "seed" the design by removing the frame, leaving only the supports that generative design had created, then include the supports as a starter shape and generate again. This problem is described in more detail in this post: https://forums.autodesk.com/t5/fusion-360-design-validate/generative-minimize-mass-and-safety-factor/m-p/8572922#M181871

Overall I am very happy with my design and proud of how I worked towards it. My final design has just over 10,000kg mass, can hold just over 4,000,000 lbs, and can be 3D-printed without overhangs over 60 degrees. The design meets all of the goals that I set out to meet. Something I would like to try in the future is using PETG plastic because its material qualities are close to nylon (the material I used for this bridge) but PETG is commonly found in plastic bottles and other food packaging, making it abundant and cheap. While 3D-printing may never be viable for the huge super bridges of concrete I think this technology could help bring low budget solutions to small towns.