Customer Story Mahindra & Mahindra

Modeling and Real-Time Simulation of an Automobile HVAC System in SimulationX

Learn how Mahindra & Mahindra uses ITI’s simulation solutions to develop reliable, powerful and energy-efficient air conditioning systems for the automotive industry.

The task at hand

Simulation with the SimulationX libraries pneumatics, thermal-fluids and industrial utilities

Mahindra & Mahindra Limited is an industry leader in the production of passenger and commercial vehicles, headquartered in Mumbai, India. To maintain their competitive edge, Mahindra conducts extensive research and development activities in the Mahindra Research Valley at Mahindra World City near Chennai, India. Part of these research activities includes the testing and calibration of a fully automatic temperature controller (FATC). To this end, Mahindra & Mahindra commissioned ITI to create a detailed automotive heating, ventilation and air-conditioning (HVAC) model using the modeling software SimulationX in order to dynamically determine the automotive cabin temperature based on the inputs from the electronic control unit (ECU) and environmental conditions. The HVAC model would need to be imported into Matlab/Simulink, where it would be connected to an existing system model before being exported to dSpace and Labcar real-time environments.

The solution

Real-time simulation with SimulationX

Since the goal of the HVAC model was the testing and calibration of an FATC, the model inputs and outputs were set as the signals to and from the ECU. Additional environmental inputs such as ambient temperature and solar radiation were also provided. In its simplest form, the refrigeration cycle of an automotive air-conditioning system can be considered as four components: compressor, condenser, expansion valve and evaporator, each separated by a volume of refrigerant of an adequate size to appropriately represent the dynamics of the system.

Additional dynamics, such as mechanical inertias, can be included or neglected based on the added value of the extra information in regards to the modeling goals as weighed against the extra model complexity and solver time expenditure. The fluid calculation model also plays a critical role in the simulation speed. In this case the table-based fluid property calculations from TLK were chosen because their quick calculation speed is ideally suited to real-time applications. Both R134a and R1234yf were considered as potential refrigerants.

The compressor was modeled as a volumetric transformer with a speed set at a fixed ratio of the engine RPM, which was an input to the model. The isentropic and volumetric efficiency were calculated dynamically from tables provided by Mahindra based on the pressure ratio and RPM. These tables were modeled using the SimulationX 2D-Map elements, which were able to directly import the table data from Microsoft Excel. The condenser and evaporator were modeled using heat exchangers from the Industrial Utilities Library. These heat exchangers use the NTU method to calculate heat transfer between two fluids.

The setting Set Efficiency was used in each of the heat exchangers, and the efficiency was calculated using SimulationX Function blocks to determine the number of transfer units and the heat capacity ratio of the fluids. Detailed heat exchanger geometry was used to determine the heat transfer area. This geometry along with the dynamic fluid properties calculated automatically in the SimulationX fluid libraries were used to calculate the convective heat transfer coefficient for the fluids. The proportion of the heat exchanger in the liquid, two-phase and gas region was also calculated dynamically and adjusted based on the current fluid conditions. [...]

The benefits

Creating a realistic simulation model of an automotive air-conditioning system

Using the SimulationX Pneumatics, Thermal-Fluids and Industrial Utilities Libraries it was possible to create a realistic simulation model of an automotive air-conditioning system. The HVAC model was able to be reduced to 23 ODE states with system time constants allowing a solver step-size of 1 millisecond. Using the Code Export Wizard the model was exported to a Simulink S-Function. In Simulink the model was connected to an existing system model and further exported to the dSpace and Labcar real-time platforms, allowing Mahindra & Mahindra to successfully test and calibrate a fully automatic temperature controller.

Simulation in the automotive industry

More projects in the field of automotive industry are available on our engineering pages. Read more.

Project examples

Read the complete paper

The complete project report is available in our customer center.

enter customer center