Dana Holding Corp.—a U.S.-based, Tier 1 global supplier of axles, driveshafts, sealing and thermal-management products, off-highway transmissions, and service parts—recently underwent a cultural change from a largely cost/manufacturing-driven company to an engineering-driven one. This meant a shift in focus, from how to control costs and manufacture efficiently, to how to innovate. Frank Popielas, Senior Manager of Global CAE in the Dana Power Technologies Group and head of CAE for Dana, spoke with AEI Assistant Editor Matthew Monaghan about this renewed emphasis on innovation, supported by the right engineering tools.
With such an expansive portfolio of products, how do you go about managing that complexity?
By standardizing, in this case utilizing SLM, which we started to deploy just recently in North America. It’s helping us to follow the same approach everywhere. It doesn’t matter where we are located; we are utilizing the same set of input data. This helps us to ensure that its relational quality is guaranteed based on what we validated and correlated with the real data. This helps us also to reduce our time itself, because it helps us to get to the right data fast. At the same time, it helps us to reduce the simulation time since we follow what is approved as a standard. So you have a much higher probability to complete a simulation, as we can more likely avoid such things as convergence issues with simulation, etc., because it’s verified that that approach is the correct one.
What kind of efficiencies are you experiencing as a result of this template-based approach?
Because you use the time you spend inside the software package more efficiently, the utilization of software, as well as compute hardware time-wise, is going down, and thus overhead cost is also going down per project. Since we have the template structure in place, it is easier to the data, and establish a history track and trace related to it. At the same time, we use this for decision making and for online collaboration with other engineers, to discuss results and possible design and process improvements since the other engineers also have to have access to the same set of simulation data. Time savings-wise, just from what I’ve mentioned, we’ve seen at least 20-25% savings.
Are you seeing a move toward more full-system simulation as opposed to simulating individual components?
Subsystems from a CAE perspective are common practice already. The next is when we really go into the complete system, meaning interaction with the customer directly to understand and consider correctly the influence our product or subassembly has when assembled into the final system, in our case typically the whole powertrain or vehicle. This is when you talk about the system engineering aspect but from a virtual perspective. This is what we call nowadays system modeling and simulation, the combination of the traditional system engineering approach with the traditional CAE. This is about the optimal synthesis between low- and high-fidelity models from a 1-, 2-, and 3-D perspective but all in the virtual space. That’s the future direction, in engineering, complete virtualization of engineering.
Is collaboration between departments around Dana benefiting from the SLM approach?
SLM is also being used as a collaboration tool, especially since we are not a centralized company, we have a lot of tech centers around the world. When you look back to how we did it before and where we don’t have SLM yet deployed, this is where we try to get information over the phone to the other engineers and explain results, or send lengthy emails and try to explain how to understand certain things and why it was done, etc. When you are within the environment of SLM, you have right away the history for this project available, so that helps to communicate the information, what is part of that model or that project. Even across different departments, it’s easy to get information across about the data that found its way into that project because it’s all combined and referenced there.
Do you see the amount of physical testing continuing to reduce?
That trend will definitely continue. I see the physical test just as a final validation of the simulation results or whatever legal requirements are in place in order to validate a product. In the future, the product is simulated and optimized from all the different aspects before you make the actual product. It depends really on the requirements from the customer side if you still have to make a prototype, or sometimes maybe we are able based on the confidence we have in the simulation data that we can go straight to production.