INAS Testimonials
Continental Automotive Romania - logo

Usage of Ansys in Thermal-electric simulation of a EMR4 HV Inverter

Written by Simina Albu

VITESCO TECHNOLOGIES Romania

Vitesco Technologies is a leading international developer and manufacturer of cutting-edge drive systems for sustainable mobility. With intelligent system solutions and components for electric, hybrid, and internal combustion drive systems, Vitesco Technologies is making mobility clean, efficient, and affordable. The product portfolio includes electric drives, electronic controls, sensors and actuators, and exhaust gas treatment solutions.

 

Task Description

The HV Inverter is composed of many subassemblies: DC and AC Areas, DC Link, Cooling Area, EMC Filter, Power Module. The focus of this analysis is the DC Connector. The geometry inside the DC Connector is part of the customer interface and was created in this simulation model based on a datasheet, it does not contain the real structure. The ambient temperature is 85°C.

HV-Inverter Structure

The finite element network has an approximative number of 1 970 000 nodes and 1 060 000 elements. The types of elements used are in majority Tet10 and Hex20.

 

The contact resistance values were adopted according to the customer requirements:



Contact resistances

The currents applied in the model are the following:

Currents

A total power loss of 10.7 W was applied on the DC Link Capacitors, it was defined as Internal Heat Generation. The power loss corresponding to one element was divided to the volume: 1.8W* (Power loss on a DCLink element) / 4.2007e-005 m^3 (volume of a DCLink element) = 40469 W/m^3



Some temperature conditions were applied to the interfaces to the battery and motor:

 

Interface temperatures: 85°C to the battery (DC), 165°C to the motor (AC)

 

The convection inside the inverter was defined initially at 5 W/m2K @ 85°C. Outside the inverter, there were 2 cases that were simulated: 5 W/m2K @ 85°C and 12 W/m2K @ 85°C. The last value is the result of the calibration simulation.

 

Solution

The temperature in the DC area decreases for Scenario 2 as the convection value is higher outside the inverter:

Scenarios overview

 

The reason of using 12 W/m2K @ 85°C as an external convection is the output of a calibration simulation that was performed on a different simulation model that includes only the DC Connector and an additional set of cables.

 

The contact resistances were defined based on the customer requirements, and adjusted considering the model and the other specifications:

 

Contact resistances for the calibration model

The ambient temperature is 19°C. The current applied is a parameter that varies between scenarios (330, 290 and 200A). Also, the application of the temperature condition at the end of the cables varies between scenarios. The convection parameters were adjusted based on testing conditions (hot air flow from power supply on the cables), including the value of 12 W/ m2K which was used in the entire inverter simulation model:

Calibration boundary conditions

The resistance measured in the test at 19°C is 116 μΩ for sensor position 1 to sensor position 2 and 380 μΩ for sensor position 5 to sensor position 6.

 

The temperature rise of the connector busbars (delta temperature to ambient temperature) in steady-state is:

         Scenario #A: 51-55 K for 330 A     

         Scenario #B: 41-43 K for 290 A

         Scenario #C: 21-22 K for 200 A

Measurement results for 330A DC Current; Position of the sensors

 

The simulation results reveal a deviation of max. 3°C from Measurement to TE (TE is the Connector supplier).

 

The average temperature increase from simulation has a low deviation of 3°C compared to test and TE (supplier requirements). The temperature values at cable ends match only for 200 A and 290 A. The electrical resistance differences (533 µΩ vs. 380 µΩ measured) are due to the temperature difference at the end of the cables and lack of information on the internal geometry (cooper circuit of the DC Connector).

The calibration analysis reveals a good calibration between simulation, measurement and TE at 200 A and 290 A.

For the entire inverter simulation, the obtained results are critical, but reveal an improvement for the case with 12 W/ m2K external convection. 

 

Customer benefits

By calibrating the measurement not only with the simulation, but also with the customer requirements (TE) ensures the product follows the specifications, can withstand certain loads and can function properly.