THE IDEA: VIRTUAL PROTOTYPING TO SUPPLEMENT TEST CHAMBER MEASUREMENTS
Airbus Defence and Space, a division of Airbus Group, develops, maintains and operates systems for civilian and military security authorities around the world. Understandably, reliable operation in hostile electromagnetic environments is a must, and so military standards specify precise methods for testing.
Anechoic chamber testing is used to characterize the behavior of devices for electromagnetic compatibility (EMC). Since Airbus Defence and Space’s products are used by military contractors, they are tested in accordance with the MIL-STD-461 standard. This specifies the precise set-up of the testing chamber, including the choice of absorbers, the arrangement of the workbenches and the grounding of the system.
However, the device under test (DUT) is often only available for a short period, which means that testing different designs or configurations in the chamber can be difficult. For this reason, Airbus Defence and Space decided to supplement their testing process with virtual prototyping using electromagnetic simulation....
THE CHALLENGE: ENSURING COMPATIBILITY WITH CHANGING EMC TESTING STANDARDS
Under the previous edition of the standard, MIL-STD-461E, the receiving antenna is placed at the same level as the DUT, which is placed on a conductive table. The receiving antenna and the conductive table are connected electrically by a copper plate. In MIL-STD-461F, the receiving antenna no longer has a direct electrical connection with the conductive table. Instead, the antenna is grounded via the chamber floor and the outer wall, which includes both pyramidal absorbers and a layer of ferrites.
Two different calibrated antennas are used to measure emissions by the DUT between 2 MHz and 200 MHz: a rod antenna for frequencies up to 30 MHz and a biconic antenna for frequencies above 30 MHz, both placed 1 m from the DUT. The setup of the measurement using a rod antenna has changed (as shown in Figure 1 and Figure 2) between the last two editions of military standards, MIL-STD-461E and MIL-STD-461F (published in 2007).
Figure 1 Test set-up for measuring of emissions from a DUT, MIL-STD-461E (in use up 2007)  and MIL-STD-461F , with simulated and measured results inset.
Figure 2 A virtual test chamber, designed to MIL-STD-461E (left) and MIL-STD-461F (right) for frequencies up to 30 MHz.
The reason is that a discontinuity occurred at the band edges at approximately 30 MHz, when comparing the emission results of a vertically polarized biconic antenna and the results measured with the rod antenna.
Because the ground connection to the wall affects the measurement results, it has to be included in the simulation. The wall is much larger than a typical device, and by design, very lossy. In order to evaluate the effect of the change in standard and verify the accuracy of virtual prototyping, Airbus Defence and Space turned to CST STUDIO SUITE®.
The frequency domain solver (F-solver) was used to model the test chamber itself and extract the receiver’s antenna factor (AF). The F-solver offered fast, accurate results for the test chamber at low frequencies, and was able to handle losses both in the metals and in the absorbers and ferrites.
Using the schematic circuit simulation tools, the 3D chamber model was then connected to analytical cable models to replicate the experimental set-up. System Assembly and Modeling (SAM) was then used to link the simulation and post-processing tasks, including S-parameter calculations for the test chamber and antenna, the excitation of the system with the AC Task, and the normalization of the results. With SAM, the calculations can be chained together and carried out automatically, and results easily compared across different projects.
THE RESULT: ACCURATE CHARACTERIZATION OF THE EFFECT OF THE TEST CHAMBER
By calculating the AF of the receiver antenna and using this to normalize the simulation results, Airbus Defence and Space were able to compare their simulation results directly with measurements taken in the test chamber. The simulation and measurement agreed very closely: the simulation accurately calculated not only the frequencies of peaks in the field strength spectrum at the antenna, but also their amplitude.
As expected, the simulation results show that at very low frequencies, the two standards gave equivalent results, but significantly different results at the higher end of the low frequency spectrum. This variation was borne out by the measurements.
References  20 August 1999, MIL-STD-461E, United States Department of Defense  10 December 2007, MIL-STD-461F, United States Department of Defense