Feel guilty about over-designing?
If any of you are space enthusiasts like I am, you are probably aware of the successful unmanned SpaceX Dragon capsule’s trip to the ISS and its return today. As I listened to NASA’s broadcast of the vehicle re-entry, I could not help but think of the multitude of simulations Elon Musk and his team would have done in order to ensure success.
If I were a part of his design team, I can imagine the loads that this vehicle would experience – a huge mass hurtling at shocking speeds subjected to extreme thermal conditions and excessive vibration. And this metal enclosure needs to bring back fragile scientific experiments safely back to earth. And don’t forget landing in the ocean and the shock loads upon impact. The closest analogy that each of us may have experienced to something like this would be when we had to drop an egg without breaking it!!
Now, all of us might not be designing space vehicles that sees unearthly (no pun intended) loading conditions. But we cannot undermine the complexity of the types of models built today. Each design poses a variety of challenges from the various loads and work environments to the life expectancy of the product. The first lesson coming out of engineering school and walking into the industry has been to eliminate the effects of the loads and prolong life simply by over-designing way beyond what is optimal. When we ask ourselves why, we realize that engineering has often taken a backseat compared to other economic pressures such as consumer demand, first-to-market, customer expectations, and internal team goals. While product designs have transcended into the 21st century, we continue to design products with a 20th century engineering foundation.
In the utopian world, a product would go through multiple cycles of prototyping and re-testing until it is optimal cost-wise and outlasts its intended life expectancy. In reality, this vision may be closer to you than you may think. If your products are designed inside SOLIDWORKS, then make sure you look into this special feature set called Design Optimization that is a part of Simulation Professional. All you are required to do is to set up a sample run of the loads that the product experiences, and then hand over the geometry to SOLIDWORKS and Optimization.
The software takes the model and the permitted variables, and begins a design sensitivity analysis using DOE (Design of Experiments). Based on the local trend of each variable, and its impact on the other variables and the desired end results, it arrives at an optimal combination. Once you put this process to test, you would be surprised as to how you lasted without a tool like this for your product designs. For example, let us consider an impeller spinning at a given speed. An impeller consists of a number of blades that is optimized for best pressure and flow throughput for a fluid. The typical engineer would probably work on a few hand calculations and come up with the angle of the impeller blade, and the number of blades required. Imagine if the wheel could be at least modified, if not re-invented. You can tell SOLIDWORKS the pressure of the fluid, the temperature of the fluid, the motor speed (to account for minimizing vibration and chatter) and the rpm of the impeller (to account for the centrifugal forces) and ask Optimization to give you the optimal number of blades, the thickness, the radii of curvature and the length of each blade. The output is a product that will meet the desired goals of minimal mass, least vibration, and enough rigidity to last many years in the field. This concept could easily be extended to any of your designs.
A month or so ago, I spoke to a product engineer who builds huge plastic crates. His design needs to withstand stack loads, and the weight of the products it houses. His challenges were immense – dealing with intricate features, and rising material costs and field-failure if the product is not designed correctly. He used optimization to determine how to define the corner braces on the geometry. The end result was a shape that he claims he could not have come up with at all without this tool. The solution is capable of being manufactured, and gives the model the desired rigidity, and gives him the edge over his competitors.
So the next time you are looking to decide how the thickness of your sheetmetal product, or the desired radius, or the number of holes on a pattern, or the desired length of that gusset edge, do not be left in doubt. And more importantly, do not overdesign simply because you are unsure and would rather place a safe bet even if it means money down the drain for your company. Instead, use Design Optimization to come up with a product that looks better, can be manufactured faster, and is, on the whole, cheaper!