For many design teams, tighter emissions regulations have expanded the definition of a powertrain. Aftertreatment systems are now a critical factor that impacts engine operations. Design teams must include treatment technologies as they optimize powertrain parameters.

“The aftertreatment system, engine, and transmission need to be optimized as a complete system instead of individual elements,” said Chris Mays, Senior Technical Specialist at BorgWarner’s Advance Engineering Group.

“For example, getting the SCR aftertreatment to its operating temperature quickly means that the engine can move to more efficient operation sooner. The engine needs to interact with fuel injection, throttles, and variable valve train components as well as the transmission to enable a rapid warm-up. This is a system decision to sacrifice lower engine efficiency for a short time to achieve a net overall system benefit for the full operation cycle.”

When all these systems share information continuously, there’s a lot of data traveling over networks. Getting the right data to each node at the right time is a challenging task. If the combined systems and networking links aren’t considered as elements of a larger system, problems can arise.

“Another challenge associated with communication between the engine and transmission is obtaining accurate load information at very light loads,” said Michael Pipho, technical lead at John Deere Power Systems. “Aftertreatment technologies also increase the amount of data on the network structure due to additional control requirements and the need to communicate aftertreatment information to operator interfaces. The increased sharing of data among systems can result in high network loading and cause some data latency issues.”

Most developers are using a number of networks linked together using gateways. That provides the necessary bandwidth without adding a lot of complexity.

“As the engine sub-systems develop, particularly those related to aftertreatment, it is necessary to use multiple networks separating machine and application data from sub-system monitoring,” said Mike Cullen, Perkins’ Product Marketing Specialist. “One example is NOx sensors, which run on a different network to that of the machine. As the number of variables increase, the limitations of a network become apparent. To maintain a responsive system and avoid latency issues it is sometimes necessary to rethink the network architecture of a machine.”