Simulating a drive cycle can have many benefits. Several hot topics right now include improving fuel economy and emissions measures, but there's another way simulating a drive cycle can help you optimize a new vehicle design. A drive cycle simulation can help you predict transient thermal behavior and maximize the capabilities of powertrain and exhaust components. To do this, you need to simulate the thermal activity of these components over a drive cycle—which isn't easy to do.
Definition of a Drive Cycle
A widely agreed upon definition for drive cycle is a dataset representing the speed of a vehicle versus time. Depending on the required output of the analysis, many simulation methods can be used to represent a drive cycle, both steady-state and transient.
Many manufacturers use simplified, steady-state models to answer design problems. These models plan for the theoretical worst-case scenario and often lead to brute force methods for thermal management strategies. Over-engineering thermal management can lead to costly late-stage design decisions elsewhere so critical metrics like fuel economy can still be met. A transient simulation method gives you more visibility into real-world use cases so that manufacturers can not only design for the actual worst-case situations but also optimize for the typical driving scenarios the vehicles will undergo.
Simplified steady-state simulations may also fail to predict particularly high temperatures because of the rapid-changing nature of drive cycle scenarios. One example is a soak scenario with a vehicle at high speed which then rapidly comes to a stop; the convective cooling will drop off, but you still have energy that needs to dissipate from the exhaust components. Another example is a spike in exhaust temperature caused by rapid acceleration; steady-state simulations would miss the effect of lower speeds and reduced convection at the beginning of the acceleration.
Because of this need, new drive cycle simulation strategies have been developed that highlight the importance of simulating for real-world use cases. Our drive cycle methodology uses surrogate modeling techniques to rapidly simulate drive cycles that represent these real-world scenarios.
What You Need to Run a Drive Cycle Simulation
If you want to run a transient drive cycle simulation, you will need a cycle definition of speed versus time and inputs to characterize your heat sources. If you use a simulation tool like the Exhaust Extension, the heat source inputs would be:
- exhaust gas mass flow rate
- exhaust gas inlet temperature
- and engine speed
These inputs would each need to be specified as a function of time over the drive cycle(s) of interest.
Getting Data for Your Simulation
Getting the inputs for drive cycle simulations is challenging. To get this data, you can simulate a system model of the drive cycle or measure the inputs using an instrumented prototype. System models can be tough to create because they are dependent on component data you may or may not have access to. Obtaining the data from a prototype isn't easy either, mainly because prototypes are not usually available early in the design process. If you want to get started with drive cycle simulations but aren't able to get the data you need right away, one solution is to use data from a previous vehicle model with a similar powertrain. Previous vehicle data may be a good starting point as you can scale the data appropriately to fit your scenario.
Simulating Different Drive Cycle Types
You can simulate any drive cycle type with the surrogate modeling method used in the Drive Cycle Extension. It is possible to model any drive cycle type because the inputs are user-defined. With this flexibility, you can simulate any regulatory drive cycle such as WLTP, NEDC, SORDS, or FTP-75 as well as any OEM-specific cycles.
Our next software release in December of 2018 will include additional capabilities in the Drive Cycle Extension that will allow for even more flexibility. Building on the 1-dimensional speed-based surrogate modeling first released in early 2018, you can now specify multiple independent variables as the basis for the surrogate model in your drive cycle simulations. The benefit is that you will be able to create greater accuracy when something changes in the drive cycle that is not a function of speed (such as the engine load when climbing a hill when the vehicle speed remains the same). This additional capability will make the surrogate modeling we use for the drive cycle extension even more applicable to different types of drive cycles and vehicles.
Comparing Computational Power Between Methods
Using a CFD method, the computational power required to run a transient simulation of a complete drive cycle would be overwhelming. You could run a simplified method such as a single hot soak or a short portion of a drive cycle, but to get the complete picture using 3D geometry, you will want to model using a coupled simulation or specialized method that will accelerate part of the problem to enable you to get a comprehensive analysis.
The computational requirements of a 1-D system modeling tool are so low that there is not any additional complexity to simulate a drive cycle versus a more simplified scenario. Your run time will be longer but still only minutes or hours since you're not simulating a model using 3D geometry with millions of elements. However, system models typically cannot provide answers to geometry-dependent design problems (heat shield design, packaging analysis, etc.)
With a full, two-way coupling method for every time step in your drive cycle, you would simulate a new set of CFD conditions. Then you would couple those CFD results to your thermal model using the standard methods which map surface temperatures from your thermal model onto CFD and then map convection coefficients & fluid temperatures from your CFD code to your thermal model. In this method, you assume that the CFD results are a function of time—so you would have to simulate them at every time step. This method is faster than simulating using CFD alone, but it is still time-consuming. Recent results showed full two-way coupling as three times slower than surrogate modeling techniques and, in some cases, up to ten times slower depending on the modeled scenario.
With surrogate modeling, the same method used in the Drive Cycle Extension, instead of considering the convection conditions to be a function of time, we model them as a function of vehicle operating conditions (such as speed or other inputs). Since the drive cycle will oscillate between different states, we can model a small set of conditions and then re-use those conditions based on the current state of the vehicle at any time step. Instead of running hundreds or thousands of time steps in CFD, you will run between five and 20 steady-state CFD simulations (for a 1-dimensional surrogate) and then reuse those simulations via the surrogate model during the drive cycle.
When is Your Team Ready to Simulate Transient Drive Cycles?
If you and your team are already comfortable with thermal modeling using CAE software and you have access to the data needed to run a drive cycle simulation, then you're ready to investigate a new drive cycle method. We would love to talk to you about the different ways you can run a drive cycle and which method would work best for your project.
If you already have a thermal simulation process down, but you do not have access to the data needed to run a drive cycle, we can help by showing you how to use the data you do have to prove out the process. We can help you find a working process through one of our service contracts and stay with you until you can make it your own.
If you don't have a great thermal simulation process yet, don't worry. Creating a robust thermal modeling system is the first step to creating a better process overall. You will set your team up for success and the ability to perform transient drive cycle simulations in the future. We would love to continue the conversation and help you consider all the factors of learning software yourself or working with us on your next project.
Automotive is Always Looking to Optimize
This type of simulation is not easy. Developing a new process, getting the data you need, and proving out the solutions will all take time and significant effort but the alternative is not to move forward. To push the boundaries on how you can optimize your vehicles, you will have to evolve the way you design them. Customer demands will not lessen. The industry will not stop moving toward better fuel economy, increased safety, and fewer warranty claims. You will need to change your methods to create more drastic improvements. While it is a challenge to change, adopting simulation methods that offer more insight into product behavior will give you an advantage in your design process.