Revisiting a project from their youth enables engineers to showcase the lessons learned from experience.

Altair has submitted this post.

Written by: Rama Annamraju, Applications Engineer, TrueInsight LLC

Each year, the Indian Institute of Bombay conducts an annual science and technology festival to provide a platform for India’s student community to develop and showcase their technical prowess. I was humbled at the chance to participate in 2009, when I was in my second year/sophomore of undergrad as a Mechanical Engineering student.

The event I took part in was called “Full Throttle: Earth Bound.” The challenge was simple: buy an off-the-shelf Internal Combustion Engine (ICE) RC car and replace the stock steering, chassis, and suspension with custom designed components. My team of five students purchased a Red Cat 3.5cc Nitro Buggy with a 2.4GHz, 2-channel remote and a 2-speed automatic transmission with centrifugal clutch (Image 1).

The project seemed fun and achievable at the beginning, after all it is just 3 things to do (1) replacing the chassis, (2) changing the steering rack and (3) just swapping the suspension with some cast components. But the engineering gods weren’t going easy on us and things didn’t go as smoothly as expected. Throughout this article I want to recall all the major challenges we went through, and then delve into how I would use today’s technology to tackle these problems with more efficiently.

The 2009 Concept      

For the chassis replacement, we bought a stock aluminum plate and used a CMM (Coordinate Measuring Machine) to measure all dimensions accurately. A rudimentary drawing was generated from the data, shown below in Image 2.

The first challenge was the lack of exposure to CAD/CAE tools which is evident in our drawing above and the lack of Geometrical tolerancing (GD&T) in it. Recording the readings was the easy part but importing those into a CAD tool was not. We were craving for a proper tool with a flat learning curve to both understand and execute the idea in our minds. Eventually we were saved by the technicians who have some real-world experience and hence we were able to utilize their help and use the CMM data to directly feed into the CNC machine for the cutting and drilling operations. However, bringing it close to a finished part needed some cold working using a hammer and a bench vise. The finished output from the CNC machine is shown below in Image 3.

As for the steering mechanism, we replaced the stock adjustable control rod with a homemade spoke, which connected to the servo motor. This was a simple solution, considering the complexity of the other tasks at hand. It proved to be a poor design choice when it came to the loads that the steering rack goes through while cornering. On the race day we had to keep replacing the steering rods often, luckily, they were cheap and easy to re-produce. We considered it a good design back then and in hindsight it was the most effective and cheap component of the entire project.

Redesigning the suspension proved to be the most difficult step. The stock suspension was a double wishbone system, injection molded from nylon, and which contained complex features that would be hard to replicate. It weighed in around 270 grams (9.5 Oz) for all 8 components. Alternate plastics were ruled out, as the tooling would cost as much as the car itself and 3D printing was not an option available to us back in 2009. Therefore, we decided we would use a cheap metal that can be gravity cast. We approached a local foundry to discuss a potential material that could be cast within our budget. Because of its lower density, aluminum was the lightest and most cost-effect option available.

Since there were no major casting simulation tools readily available to us, we had some limitations on how to optimize the design. We decided to switch materials from nylon to aluminum but maintained the existing design with the hope that the selection panel would accept our retrofit. We provided the foundry with the stock suspensions for them to create a mold, and then cast replicas of the suspension in aluminum, as shown in Image 4.

The resultant cast model had many defects, but luckily they were functional. One defect that had to be accounted for was shrinakge. Shrinkage effects caused some of our dimensional tolerances to be out of acceptable range. To resolve the dimensional tolerance issue, we used a file to smooth out as much of the part as possible. The other biggest concern was the weight of the components they each weighed 145 grams (5 OZ) on a average which essentially quadrupled the total weight.

Race Day 2009

On the 2009 race day, a qualifier race determined the starting order for the finals. As soon as the qualifier race started, we noticed our mistake: our car was lagging due to the weight of the casted part.

To counter the weight of the casting, we added some NO₂ to the fuel with the hopes of increasing the power. However, since the problem was weight and not power, the car had low torque which made the car struggle to accelerate evenly. The race was a bust, and we were out in the qualifiers.

Tackling This Problem Today

Thirteen years later in 2022, and with the experience I have gained working as a Design Engineer, all I can think of is how much better we could have performed if we had access to the simulation tools available today. Specialized solutions such as Altair Inspire Cast, which is used to design, simulate, and optimize new/retrofit lightweight components for both performance and castability, would have dramatically accelerated our entire design process and improved our racing competitiveness.

I decided to revisit the old suspension design and experiment with various changes to see how understanding the casting process would change my design. Throughout this process, I used Inspire Cast to create a crude design of the old suspension arm (Image 6) and simulate the casting process.

Inspire Cast includes sketch and geometry modelling tools, which I used to model the part directly instead of using a third-party CAD tool. Inspire Cast also has the capability to import various third-party CAD files, if I wanted to import a CAD model directly.

After I finished modeling the bracket, I needed to follow the Inspire Cast setup wizard to simulate the casting. Inspire Cast has an easy-to-use workflow that can be broken down into five simple steps:

1. Define the cast part.
2. Define a gate/runner system.
3. Create a virtual mold.
4. Select the casting process.
5. Run the analysis.

For the current setup, I chose the Gravity Casting option. The gravity casting parameters were defined by the filling time set at 2.5 seconds, then I hit “Run” to start the analysis, as shown in Image 8.

I ran both solidification and filling analyses to compare the results. Using Inspire Cast, I can see multiple result types and insights into the casting process. Inspire Cast can generate the following results:

However, not all result types are needed to make changes to the design. From my experience in the competition, reducing post-processing steps would be my top priority. I started by looking at the temperature changes during the solidification process, since that is typically an indicator for shrinkage.

This plot correlates well with the total volume shrinkage plot below. Shrinkage not only requires additional post processing, but in most cases it can also affect the mass of the part.

Next, I wanted to look at the how the temperature dissipated during the filling phase. With this result, you can determine the temperature required for two fronts of material to fuse and study the risk of cold welding. It is important to identify cold welding, as this can cause the part to have structural weaknesses.

Lastly, I wanted to look at which spots would be the last areas to fill, in order to predict where bubbles may form. Knowing where you may get bubbles enables you to reposition the overflows to prevent porosity. Air bubbles will impact die casting more than sand molds, as they are less susceptible to porosity thanks to the permeability of sand.

While I could have completely redesigned and optimized this chassis component for performance, weight and castability using Inspire Cast, our biggest design challenge at the time was simply understanding and accounting for shrinkage defects. Inspire Cast would have provided this needed insight allowing our team more time to experiment with alternative materials and design changes that could have improved our chances to win.

Suspension Design 2.0

With this knowledge from the casting simulation, I wanted to see if I can make some subtle changes to the design and still be able to get the same results in terms of the strength and stiffness all the while reducing the weight if the components. I redesigned the suspension to add thinner walls as shown in Image 13.

Altair’s Inspire Cast allows me to check the weight of the components right in modeling phase, my new components were optimized to weigh just 67 grams (2.4 Oz) each which is more than a 50 percent mass reduction. I can now export the model to Altair Inspire and analyze it for stiffness and strength. See Image 14.

From an automotive engineering perspective, reduced weight is a big advantage to reduce energy consumption. In addition to that, the ability to maintain required stiffness and strength to withstand the forces during operation is the sign of a successful redesign.

Looking back at our initial race failure, we knew that our suspension design was too heavy which led to poor control of the RC car. We also spent a lot of time postprocessing our casted suspension parts to remove defects. With the Altair simulation tools, we could have significantly lightened the suspension design (more than 50%) to give us more control and understanding potential casting defects in order to reduce our postprocessing time, would allow us to focus more on design and innovation than quality assurance.

The Future of Casting Design

The experience I gained during the competition was extremely valuable in helping me realize how to understand and solve a problem as a team.  Upon revisiting this problem years later as an engineering professional, it’s amazing just how powerful and pervasive simulation has become across the entire development cycle.  If our team would have had access to a product like Inspire Cast to explore and address the Design for Manufacture (DfM) for this casted part, I feel it would have been game-changing.

Today, I would approach this competition completely differently leveraging a simulation-driven design strategy with Altair Inspire together with Altair Inspire Cast.  This approach would incorporate and streamline loads definition, topology optimization (otherwise known as generative design) for lightweighting, structural analysis to meet performance objectives, manufacturability for a variety of different casting processes, and the validation of die designs and processes all prior to producing a physical prototype and tooling.

Perhaps I will do just that and hopefully have the opportunity to share the outcome in a follow up article!

To learn more, visit Altair.

---

Acknowledgments

1. I participated in the competition with a team of five.
2. The model car was manufactured by RedCat Racing.
3. The price of the RC Car in 2009 was INR: 18,000 which converts to about $360 (USD in 2009) or about $490 (USD) today.
4. The budget for the modifications was close to INR: 7,000 which converts to about $140 (USD in 2009) or about $190 (USD) today.