Gear design is a challenging and technically-demanding activity, but it is critical to the performance of complex mechanical systems like helicopters. It is an arduous process that involves constant engineering tradeoffs between performance and weight. A recent conversation with a highly respected design engineer from a top aerospace OEM gave me a better appreciation for how difficult a job like this can be – and how Sentient’s DigitalClone technology can make a huge difference.
Design Optimization – 32 design options vs 1 design
There are a number of specifications that a gear engineer considers during the design optimization process – such as base material, gear tooth geometry, protective coating, heat treatment, and even the lubrication requirements. The ideal solution is to evaluate all possible combinations to arrive at the optimal mix that will maximize design life, minimize weight, and ensure manufacturability. In all but the most heavily-funded programs, there is rarely sufficient funding to support that approach.
One OEM gear expert said that after his program budget was slashed by R&D cuts, he had to dramatically shrink the scope of his program – driving him to use engineering discretion to identify the most likely candidate for the optimal solution and use that as the prototype design. “I simply didn’t have the budget to build and test multiple design options,” he says.
Contrast that with the capabilities provided by Sentient’s proprietary modeling technology DigitalClone that provides microstructure-level analysis of components and comprehensive life predictions for each application. DigitalClone allows this engineer to explore all design tradeoffs for his gear, which equates to a total of 32 design options that can be fully evaluated prior to selecting to the best prototype configuration.
Enhanced Prototype Design Data – 99% more data points
NASA typically recommends at least 6 test data points for a new design, for statistical significance, according to the OEM engineer. Due to the challenges related to refining the test stand setup, “we usually conduct eight tests to ensure we get at least six usable results,” he noted. The end product is a total of 6 to 8 data points that indicate the fatigue limit of each of the prototype gears that were tested, which is then used by the engineer during his effort to predict the life expectancy.
DigitalClone computational testing typically consists of 50 iterations of virtually simulated fatigue testing. When applied individually to the 32 design options, this yields a total of 1,600 data points for comparison purposes. Anytime you give an experienced engineer 99.5% more data before he has to finalize a design, there is no doubt a significant benefit to the users of the end product.
Reduced Cost – 37% cheaper than physical testing
According to the OEM engineer, he estimates the typical cost for each physical test iteration are currently at about $10,000 each – and he expects those costs to increase. As a result, he normally budgets at least $80,000 for physical testing during design optimization on each new gear design.
DigitalClone modeling and testing of the type of gear that this OEM engineer was evaluating would typically be estimated at less than $50,000 – a 37.5% cost reduction when compared to the standard physical testing. This is exceptional value, considering that the DigitalClone models can be utilized through the entire product lifecycle to optimize/refine designs, predict life expectancy, enhance supply chain visibility, and even assist in root cause analysis of field issues.
And this is just for a single gear – the benefits increase by an order of magnitude when applied to a system (such as a gearbox or transmission).
This article was written by Jason Rios, General Manager of Transportation & Aerospace