Multi-domain Simulation

Intuitive and Flexible Modeling The key to capturing integrated physical behavior of a system is to have accurate and appropriate simulation models available. In order to accurately quantify the efficiency of an electric drive system, it is best to model the system in a single simulation environment. This environment needs to incorporate important loss mechanisms such as magnetic effects (saturation, thermal dependencies), electrical effects (electro-thermal coupled transistor behavior), mechanical loading, along with control algorithms. In automotive power electronics, self-heating behavior and statistical production variation can give rise to unpredictable complex system interactions. Inclusion of these physical domains allows for optimization at the component, sub-system, system and control software levels to meet overall vehicle design criteria. Modeling these behaviors is difficult - how does one go from physical devices to models with enough detail to accurately describe system behavior? Primary challenges include access to proprietary data about particular components or subsystems, inclusion of enough detail for system relevant quantities (e.g. thermal effects on electrical signals) without overly complicated formulas (e.g. co-dependent equations based on physics first principles), and methods for validating that the model is accurate. SaberRD addresses these modeling needs with an extensive library of over 30,000 physical models in all domains mentioned as well as providing graphical tools for creating or characterizing models beyond those available in the library. These tools include the ability to bring in datasheet characteristics or measurement data and use optimization algorithms to match model performance to the desired component behavior, all without requiring a user to have knowledge of modeling methods or programming languages. For those who need additional flexibility and capability, SaberRD supports open standard Hardware Description Languages (HDLs), including VHDL-AMS and OpenMAST. Another important source of system models is finite element solvers, computational fluid dynamics solvers and electromagnetic field solvers. Generation of behavioral models for power semiconductor devices (IGBTs, MOSFETs, and diodes) from device simulators (such as Synopsys TCAD tools) allows for early and accurate loss calculations for Hybrid and Electric Vehicle DC/DC converters and motor drive inverters. Extraction of S-element data for signal integrity analyses from electromagnetic field solvers such as those from CST AG permits physical layer validation of signal integrity for internal control communication busses such as CAN, LIN and FlexRay. Moreover, SaberRD helps companies to protect their investment in their existing model libraries by providing a high degree of flexibility to reuse models originally created in other software (e.g. SPICE). Figure 2 shows an example of an alternator charging system that models mechanical dynamics of the crankshaft to rotor, electro-thermally coupled active electronics (diodes), a thermo-magnetically coupled machine model, as well as system heat sinks, electrical protection and electrical network loads. This level of detail in simulation allows for a deep understanding of the electrical system characteristics of the alternator and AC rectification scheme including ripple voltage and harmonics. The simulation models used to build up the alternator charging system testbench, including the multi-domain model of the alternator, are taken directly from SaberRD's existing model library or have been created using one of several device characterization tools. For example, the alternator regulator contains a power MOSFET that controls the current that is supplied to the rotor windings. This model can be created based on semiconductor datasheet information using the Power MOSFET characterization tool in SaberRD. The tool supports both modeling of pure electrical or coupled electrothermal behavior. In addition, the environment around the transistor has been modeled to represent the dynamic thermal behavior of the circuit by using a thermal impedance network. The alternator model includes several effects critical to the design, including the electric, magnetic and mechanical dynamics to represent the translation of the rotational power into electrical power made available to the vehicle power network after rectification. The core of the alternator is also modeled to take into account the dynamic thermal effects that reflect the impact on the alternator's capability to supply electrical current. In order to do so, the core has been equipped with a thermal network model, including the air windage (e.g. cooling) effects. In order to construct the testbench with a realistic electrical consumer and supply environment that would be found in a vehicle, additional parts have been added from the standard SaberRD model library: Lithium Ion battery, cabling, and configurable power loads to model various mission profiles of the vehicle. Consumer loads are incorporated to model load dump scenarios and can be either be used directly from the model library or modeled with a load modeling tool. This feature allows users to graphically define loads and configure models with different operating characteristics (e.g. cyclic switching). From here many important system simulations can be performed to validate and optimize resonant load behavior, load dump and transient suppression protection or detailed field current regulator implementations. Important accurate fault behavior can be analyzed for conditions such as shorted or open diode connections, or field winding shorts in the alternator. The system model also accounts for dynamic charging and discharging behavior of the battery depending on the system loadings and the alternator's ability to supply sufficient electrical power. The alternator's supply ability is impacted by the rotational speed of the armature shaft as well as losses due to mechanical friction, damping and thermal behavior associated with the alternator components. All of these effects can be taken into account using the SaberRD simulation and modeling solution. Vehicle platform optimization can no longer afford to overly simplify or ignore the impact of electrical systems on size, mass, placement, performance and cost of the components and subsystems that define the final production vehicle. SaberRD's links to TCAD device simulation tools provide another important opportunity for hybrid-electric and full electric vehicle applications: co-optimization of the devices (IGBTs, MOSFETs, and diodes) and the application (inverters, DC/DC converters, etc.). Rather than having to rely on repackaging of existing classes of power devices and then utilizing circuit techniques to compensate for less than ideal performance of devices, it is now possible to perform virtual device iterations and generate accurate circuit-level models in a short enough timeframe to link the efforts of the circuit designers and device designers in real time. For different applications, different device characteristics can affect overall efficiency of the circuit. Understanding the application needs and being able to create accurate circuit simulation models from detailed device physics allows power semiconductor companies to more quickly develop differentiated solutions specifically designed for vehicle power electronics challenges. Add to this the capability to extract thermal impedance models from detailed 2-D and 3-D geometric and material data in the TCAD environment and then quickly generate an equivalent thermal impedance network for system simulation and now there is a comprehensive tool flow addressing two of the most critical aspects of hybrid-electric and electric vehicle design long before physical prototypes are even available.

 

Support for a Model Supply Chain Today's complex electrical systems are composed of components and sub-systems designed, developed and manufactured by numerous different companies. In order to understand system behavior, it is important to have a common language between different tiers of the supply chain through which to communicate requirements, performance and anomalous operation characteristics of the physical content of the system. As domain experts with respect to the components and subsystems they deliver, suppliers are typically the best equipped to create simulation models that accurately reflect performance of real hardware. Further, they have the most direct access to performance measurements and test data required for model validation. The needs of a sub-system or system integrator have to be balanced against each contributing supplier's ability to protect their intellectual property, yet still provide portable models that allow sub-system and system-level integration and test in a single simulation environment. The hybrid and electrical vehicle community has been looking to adopt methods and technology gained in other domains with similar power electronics content such as the aerospace industry. Saber�® tools have been used successfully for power electronic system design, validation and Federal Aviation Administration (FAA) certification at major aircraft OEMs for over a decade. Typical aircraft power systems involve different voltages busses (115AC, 230AC, 28VDC, 5VDC) driven by either fixed or variable frequency generators (400Hz fixed, 380Hz - 800Hz variable) and a wide variety of loads, many of which include active power factor correction (PFC) circuitry and internal DC/DC converters. System stability is critical especially due to negative incremental resistance of DC/DC converters loads. These systems, while larger and more complex than typical automotive systems, have much in common with hybrid-electric and electric drivetrain power architectures. SaberRD models and simulation technology built up over years of usage in aerospace applications can readily be leveraged for hybrid-electric and pure electrical vehicle power systems. Saber's multi-domain modeling capabilities and robust simulation algorithms typical of switch-mode power systems coupled with the ability to protect intellectual property have put Saber technology at the center of these efforts. SaberRD supports standard encryption mechanisms for industry standard HDLs (e.g. IEEE standard for VHDL-AMS encryption) that allow for portability and protection from one level of the supply chain to the next. SaberRD's analysis, post processing and report generation capabilities enable clear specification and validation definition as well as easy mechanisms to transfer pertinent data from the simulation environment to other useful forms for report generation and documentation. Another complexity arises when the different companies within a supply chain use different tools and/or languages (HDLs, programming languages, or other data standards such as IBIS or Touchstone) to model physical behavior of components. Figure 3 illustrates the range of modeling languages and formats supported by SaberRD to support modeling of physical behavior from a variety of available sources. Multi-domain simulation of physical systems is only a part of the overall tool chain necessary for system design, validation and production of power electronic systems for hybrid-electric and electric vehicles. SaberRD supports various important classes of integrations that enable extensive digital modeling and simulation of such systems, including: â�¢ Digital verification through co-simulation with popular digital logic simulators â�¢ Design and verification of embedded algorithms/software through co-simulation and model import with/from MathWorks�® Simulink�® and other tools â�¢ Validation of embedded control systems running on virtual ECUs through Synopsys Virtual Prototyping solutions â�¢ Verification of board-level analog electronics through integration with popular PCB design environments â�¢ Verification of electrical wiring and cabling through integrations into Saber Harness and other third-party design environments â�¢ Generation of power semiconductor models in connection with Synopsys TCAD device simulation products Complex power electronics, in conjunction with stringent space, weight, performance and cost considerations, require integrated optimization of vehicle electrical systems. Electrical system impact on thermal, mechanical and magnetic behavior (and vice-versa) needs to be accurately accounted for to inform pragmatic and feasible system-level tradeoffs for viable vehicle development. Software controls of power systems also require physically representative models to validate system stability and control algorithms over the full range of normal and anomalous operation. Hybrid-electric and electric vehicle systems are driving traditionally mechanical-based engineering organizations to incorporate more and more complex electronics into their platform designs. Optimization at the software, hardware, and system level is necessary - linking component, sub-system and system integrators earlier in the design cycle. SaberRD provides a unified environment, an extensive set of model libraries and modeling tools, as well as intellectual property protection allowing for early virtual integration in a multi-domain simulation environment suitable for novices and experts. www.synopsys.com/saber