Automotive Generative Design - Fusion 360
Challenge
How might we integrate generative design in automotive manufacturing?
Design Statement
Our model created by Fusion 360’s generative design software demonstrates the capabilities and limitations of generative design on a large scale of mass production. We ensured that current manufacturing was minimally disrupted and current technologies could be repurposed for new generative designed parts. The final product is produced by additive manufacturing inspired by MX3D’s innovative metal 3D printing technology.
My Role
As lead design researcher, I conducted an analysis of the manufacturing process and uncovered processes we could rework to transition into additive manufacturing while cutting down manufacturing space and reusing current machinery. Additionally, I took on the role of designing and organizing our presentations and reports.
Timeline
4 Months: August 2019 - December 2019
Team Members
2 Industrial Designers + 3 Mechanical Engineers
Sponsor
Autodesk Fusion 360
Design Process
What is Generative Design?
Fusion 360’s new generative design tool is proven to reduce material costs and sustain structural integrity for parts. For our project, we used focused on the goals to use less material, decrease manufacturing complexity, and consolidate parts while creating a comparable structure.
Defining the Problem
Stakeholder Maps
Mapping our stakeholder relations gave us insights on how to evaluate repair and maintenance requirements for the body structure. We needed to ensure the mechanical components would be accessible and safe for repairs to be completed. Evaluating regulations further developed our understanding of the desired performance of the crash structure to keep the users safe and alive.
Interviews
We conducted interviews with a variety of mechanics and engineers to further understand how generative could be integrated into the current process. We found that the most common machinery are 5-axis robotic arms with specialized attachments for each stage. The welding process is the most space consuming process while the stamping is the most time consuming.
Academic Research
Anatomy of Parts
Manufacturing Process
Currently, the frame of the Camaro is formed with stamping dies. The parts are then welded together with over 700 five-axis robotic arms.
Mounting Points for Mechanical Components
Forces
We derived our crash structure from measurements of NCAP-MGA-2012-018, a report on the 2012 Chevrolet Camero. These measurements were proven to withstand a crash and keep the passengers safe inside. Our other inputs were to ensure the weight of our structure would not hinder the cars ability to function due to an unbalanced weight.
Manufacturing
We evaluated multiple methods of manufacturing, and we found that metal 3d printing would be the most viable. MX3D has pioneered metal 3d printing in Amsterdam using the same 5-axis robots in current car manufacturing plants, and its parts are durable enough for long lifetimes.
Prototyping
The process of prototyping and testing with generative design pushed our team to use many iterations to better understand how the program used our inputs. By varying the inputs each time, we were able to design our final test to create a desirable form.
Obstacle geometry was used to inform the program where not to generate material.
The starting shape gave the tool a form to begin adding to or subtracting material from.
Mounting points were used to keep shapes and locations for the final model.
Forces were applied to different mounting points to determine how much material would be needed to deflect during a crash.
Testing
The team ran simulations with a few of the results we acquired from the generative design tool. During the first tests, we found there to be too small of a minimum deflection; therefore, we decreased the force inputs to cause the software to generate a proper crash structure.
Final Design
By using generative design, we were able to reduce the current 20 parts down to one. The final design included all required mounting points for the subframe, drivetrain, and body panels while also having adequate clearance for the engine, suspension, and maintenance work.
Building this structure would need to be done through additive manufacturing because of its complex geometry, which would take between 68-168 hours using a singular 5-axis robot, which is comparable but higher than the current sub 6 minute process using over 700 robots, 70 hours per robot.
We did uncover that it would be possible to reuse the current robotic arms, which meant integrating this type of structure would not be too disruptive or expensive.
In conclusion, our project gave Autodesk insights on how Fusion 360 would need to be developed further for a large part to be created with generative design.