Skip to main content

Ignoring the bigger picture…

Let’s look at the basics of decision making and risk management for a moment. The later you decide, the more you know. The earlier you decide, the more time you have to work on implementation. The timing of the decision creates its own reality and consequences. Now, apply this logic to product development. Since implementation takes a long time, you want to armed with as much information as possible up-front when making design decisions. Better simulation answers, sooner in the design process, will help balance the trade-off between rational choices and playing the odds. Given this, one would expect that industry spends a good amount of simulation time and money upfront to aid good product design decisions.

Yet, trends, in the automotive industry among others, show that much of the CAE today is actually done much later in the design cycle – AFTER the design has been nearly firmed up and the vehicle is being tested.

There are two reasons for this. Firstly, today’s CAE tools (CFD/FEA) are data hungry – they need tons of data as input into the simulation model.  And this level of detailed data is only available once the design is finished. Secondly, there is a general propensity to get the simulation answers tightly correlated with the test data. This has driven the analysts to continue enhancing their simulation models to double-digit accuracy forgetting the importance of availability of the answers in a timely fashion.

All this might be pardonable except for the fact that the majority of the product cost is committed very early in design phase. Most of the decisions in this early phase are being made by phone calls and meetings based on “past experience” and “rudimentary” back-of-the-envelope calculations. Maybe, this is a bit of an exaggeration but not too far from the truth.

The goal of any Director of Product Development should be to close the gap between when the product costs are committed and the CAE money is spent – to affect the product costs before they are firmed up, and to enable informed decision making. This is the “Zone of Opportunity” for simulation.

In this early design phase – the land of uncertainty – is where system modeling comes in. And it is a goldmine of opportunities to save costs and make a robust product. It provides powerful insight into the system you are about to build. What will the overall performance of the product look like under different operating conditions? What key variables will have a significant effect on the overall performance? What trade-offs can I make at the macro level?

Yes, System modeling is less accurate than other rigorous methods due to simplifications and assumptions made.  In fact, any simulation (whether experimental or digital) is a reduction of reality and limits interpretation. Knowing the underlying assumptions help quantify and define the risks. Without this, one is taking the risks blindly. And a late root cause discovery may force the product development team into expensive risk-mitigating solutions.

Yet, System Modeling has been somewhat of an elusive capability for most companies. There are a few reasons for that. Firstly, it is hard. It is hard because it requires expertise on a wide variety of subjects that engineers don’t necessarily have. As an example, a good system model of the engine requires a good grasp on the combustion, heat transfer, fluid mechanics (cooling & lubrication and two-phase flow), structural dynamics, testing methodologies and standards to name a few. Secondly, it is grossly misunderstood and is sometimes called Lumped Parameter or 1D. It is not uncommon to hear, “We already do 3D, who should I go back to 1D?” In addition, it does not produce the sexy color plots of the more popular CAE tools, FEA and CFD. System Models give you numbers, or graphs at best. Very valuable information for decision making, but one that needs a lot of interpretation and cross-examination.

OEMs end up relegating the system modeling to select few who are experts but work in Research (“not related to production”) or to engineers whose bandwidth is already taken by everyday production release related activities. Given the lack of time and lack of priority, Engineers try to hack it with the minimal knowledge they have or ask Dr. Google for the path to the answer. Not an effective approach.

System Modeling should be a priority goal for every product design team with focus on two key aspects that make it successful. Firstly, the timing. The answers MUST be delivered during the “zone of opportunity” when the product design can be modified based on what we learn from the system model. Secondly, the right talent. Have a dedicated team whose role is system modeling – with the right mix of broad skill set that is required for your product.

As you review your product development strategy and review what key improvements you want to make – add System Modeling to that agenda – Upfront System Modeling during the initial design phase. Also, be willing to embrace trends and not only final answers. This will help balance the trade-off between rational choices and playing the odds and create a more robust product that meets the customer specifications.

That’s why they brought you onboard, isn’t it?


Sudhi Uppuluri is the Technical Director and Co-Founder at Computational Sciences Experts group. He has over 15 years of experience in system simulation of automotive and aerospace systems. He worked as a consulting engineer and sales manager at Flowmaster USA for 8 years. He has various technical publications on related subjects in SAE and AIAA journals. He holds a Masters in Aerospace Engineering from the University of Illinois at Urbana-Champaign and a Certificate in Strategy and Innovation from the MIT Sloan School of business.

Leave a Reply