IEC/IEEE 62704 2 Edition 1 0 201 7 06 INTERNATIONAL STANDARD NORME INTERNATIONALE Determining the peak spatial average specific absorption rate (SAR) in the human body from wireless communications dev[.]
General considerations 1 0
The three relevant elements that define the exposure conditions in vehicular environments are: the communication device(s) with antenna(s), the vehicle model, and the location of the exposed subject
Communication devices usually comprise one or more transceivers linked to a single antenna To connect multiple transceivers, multiplexers and/or power combiners may be necessary, along with an RF transmission line, such as a coaxial cable, that runs from the transceiver or combiner to the antenna connector.
The term "transceiver" encompasses both individual transceivers and more complex systems that may include multiple transceivers, combiners, and additional devices along the RF signal path Components placed before the cable leading to the antenna are considered part of the transceiver Each transceiver is equipped with an RF port, which is the connector for the cable, and it is essential to understand the RF signal characteristics, including frequency, bandwidth, and average power, at this port.
Key features of antennas include their geometrical dimensions, physical construction materials, and electrical characteristics such as frequency response, return loss, gain, and radiation pattern Additionally, any electrical or mechanical tuning mechanisms and the specific mounting locations are also important considerations.
The vehicle body’s metallic components and antenna placement are crucial factors in determining exposure scenarios The design and characteristics of the vehicle, such as windows, should accurately reflect the standard use of the communication device while avoiding unnecessary complexity in computational modeling Additionally, the pavement model must be incorporated into the simulation.
Vehicle modelling 1 0
For reliable and repeatable simulation results, a specific CAD model of the vehicle is provided in this document It is essential to use this standardized vehicle model for successful simulations, as results may not be applicable to other vehicle types or different antenna installation conditions, such as non-metallic roof installations.
The standardized vehicle model outlined in this document is applicable for compliance assessment across all vehicle models, provided specific conditions are met For both roof mount and trunk mount antennas, the bystander distance must be defined according to installation requirements, ensuring that the separation distance does not exceed the minimum compliance distance specified This same separation distance applies to the passenger, except for roof mount configurations where the passenger is partially shielded by the metal roof Consequently, the vehicle model's impact can be disregarded in simulations using the standardized vehicle model, streamlining the evaluation process.
FDTD codes can utilize either uniform or graded meshing algorithms Uniform meshing typically sets the computational model's resolution based on the anatomical details of the human model representing the exposed subject In contrast, graded meshing allows for reduced mesh resolution in certain areas outside the human body model, following specific guidelines, to enhance the execution speed of numerical simulations and expand the computational domain's geometrical dimensions However, maintaining consistent meshing quality is crucial.
IEC/IEEE 62704-2:201 7 – 1 1 – © IEC/IEEE 201 7 resolutions as defined and used for the validation and reference models are also used for all related exposure simulations
The vehicle body computational model primarily consists of metal sheets with perfect electric conductor (PEC) properties It is crucial to inspect the meshed representation of the standard CAD model to ensure the continuity of these metal sheets and to prevent any unwanted shorting between different metal parts during the meshing process The metal sheets can be represented as thin layers where the PEC condition is applied to contiguous voxel faces along a coordinate plane, or as a mix of thin and volumetric objects To maintain consistency and validate the mesh for the standard vehicle model, a maximum mesh step of 10 mm will be implemented.
In standard vehicle models, the electromagnetic fields scattered by glass surfaces and dielectric components are considered a second-order effect and are therefore not included Additionally, rear window defogger elements, which consist of high resistivity conductors, act as electromagnetic scatterers that can reduce the RF energy flow through the window For the scope of this document, the impact of defoggers is disregarded.
Communications device modelling 1 1
Before evaluating the RF energy exposure from mobile radio antennas, it is essential to confirm that the electromagnetic emissions from the transceiver equipment are minimal in comparison to the overall exposure levels This verification can be achieved by consulting the radiated emission data found in the EMC compliance report for the transceiver, which should be assessed based on measurements or other methods endorsed by internationally recognized EMC standards.
According to IEC/IEEE 62704-1, the RF source should be modeled as a resistive generator in the FDTD model Source excitation must typically be applied at the antenna feed-point, unless specific circumstances arise, which must be detailed and justified in the assessment report.
Identifying and quantifying fixed losses is essential to avoid overestimating RF exposure While cable insertion losses at the antenna feed-point can often be ignored through RF power scaling, this may lead to a conservative bias in exposure assessments If cable losses can be accurately determined based on specifications and installation length, they should be factored into the assessment by reducing the net input power by the cable loss minus 0.5 dB For example, a cable loss of 1.25 dB results in a radiated power that is 0.75 dB lower than the power at the transceiver port Losses under 0.5 dB should be disregarded to maintain a conservative assessment Additionally, return loss from antenna mismatch can also be neglected, further enhancing the conservative nature of the RF exposure evaluation.
In any case, proper justification shall be provided to quantify the cable insertion losses and return loss if they are introduced in the computational analysis
FDTD simulations can be conducted at any chosen power level due to the linearity of the simulated fields under specific electromagnetic exposure conditions, allowing for results to be scaled to the actual maximum average output power of the communication device.
To ensure valid results, the antenna must be accurately modeled to reflect its physical characteristics, which may involve using a combination of wires, patches, and dielectric or metallic objects Simplifications can be applied to reduce model complexity, such as incorporating lumped reactive elements in the FDTD model instead of electrically small reactances like loading coils, which are used to phase different sections of long antennas Components smaller than one-tenth of the wavelength in the local dielectric are considered electrically small, and any lumped elements in the antenna model must be validated according to established procedures.
Antenna models must be positioned on the vehicle in accordance with the test setup outlined in this document and the requirements specified in the antenna installation and product manual This document details the test requirements for both rooftop and trunk mount antennas, recognizing that both the antenna and the vehicle act as radiating structures Typically, the antenna feed-point is situated at the base of the antenna, where it connects to the vehicle body, which is the most common configuration However, alternative feed-point locations, such as the center of a sleeve dipole antenna, may occur, necessitating proper justification in the simulation report Additionally, matching networks may be utilized at the antenna feed-point to ensure appropriate matching to the RF source impedance.
In certain scenarios, the matching network can be excluded from the FDTD model, as the computed results can be adjusted to achieve the required antenna input power.
The antenna feed-point impedance and total power radiated under matched impedance conditions can be determined through simulations at the test frequency By integrating the steady-state magnetic field around the feed-point edge, the input impedance is calculated, allowing for the steady-state feed-point current (I) to be derived This current is then utilized to calculate the steady-state voltage (V) across the feed-point gap, leading to the computation of the net average input power.
21 in (1 ) where the asterisk indicates the complex conjugate
SAR can be normalized to power input (P) to calculate SAR per watt The whole-body average and peak spatial-average SAR for 1 g or 10 g can be derived by scaling the normalized SAR with the maximum power rating at the transceiver or antenna port Additionally, the final results may incorporate time averaging as permitted by national regulatory exposure standards, such as the 50% duty cycle for specific PTT mobile radios.
A Thevenin model of the source feeding the antenna is shown in Figure 1 The voltage source
V S and the real source impedance R S are employed in the computations
The antenna impedance is expressed as \$Z_A = R_A + jX_A\$ Under ideal impedance matching conditions, known as conjugate matching, the available power from the source represents the maximum power that can be dissipated in the load, where \$Z_A = R_S\$.
The operating condition for achieving maximum power at R S is illustrated in Figure 2 This condition corresponds to the matched load resistance, which is depicted by a power metering device that measures the maximum power transmitted from a radio.
Figure 2 – Voltage and current at the matched antenna feed-point
Therefore, it is possible to establish the Thevenin voltage V S that would be representative of a radio transmitter with maximum rated power P max and the source impedance R S as follows: max S
When a transmitter is connected to an antenna with input impedance \( Z_A \), the radiated power cannot exceed the maximum rated power \( P_{\text{max}} \) due to energy dissipation in the antenna and energy reflection at the antenna port Energy dissipation arises from dielectric and ohmic losses, while reflection occurs from impedance mismatch Ignoring these losses leads to a conservative bias in compliance assessments, as ideal (lossless) material properties are assumed, making the radiated power \( P_{\text{rad}} \) equal to the input power \( P_{\text{in}} \) Additionally, neglecting mismatch losses introduces further conservative bias, represented by the mismatch loss factor.
The reflection coefficient at the antenna port is represented by Γ, and to eliminate mismatch loss, the computed fields are scaled by a factor of 1/η post-simulation This adjustment effectively sets the radiated power, \( P_{rad} \), equal to the maximum power, \( P_{max} \), resulting in a reflection coefficient of Γ = 0 This indicates that there are no mismatch losses, allowing the radio transmitter to radiate its maximum power efficiently.
The FDTD model eliminates the impact of ohmic losses from the matching network components, resulting in a more conservative bias in the evaluation.
Exposed subject modelling 1 4
The subject will be modeled in simulations using standard bystander and passenger models derived from the "Visible Human" dataset These models, available in a generalized FDTD format, encompass 39 distinct tissue types with dielectric properties and mass densities based on Gabriel's research and sourced from publicly accessible databases The dielectric properties are calculated using the Cole–Cole formulas, taking into account frequency dispersion characteristics.
The Visible Human model represents a standing man, which has been rearranged to create sitting (passenger) and properly positioned bystander models using specialized software These models, along with their tissue structures, are available in voxel-based format for integration into custom or commercial FDTD codes Detailed information about the file format and standardized human models, including tissue parameters, can be found in Annex A Additionally, Annex A includes reference data on tissue dielectric constants and conductivities at selected frequencies to ensure accurate implementation of tissue parameters based on the Cole–Cole model.
For compliance assessment, human models must not be modified or used at a different voxel resolution than the specified 3 mm, applicable up to 1 GHz, which corresponds to no worse than one-twelfth of the wavelength inside the body A study by Gosselin et al [5] found that within the 30 MHz to 1 GHz frequency range, a voxel resolution of 3 mm yields a convergence of approximately 0.5 dB for 1 g peak spatial-average SAR and about 0.2 dB for whole-body average SAR Another study [6] reported even better convergence, achieving better than 0.3 dB for peak spatial-average SAR and better than 0.15 dB for whole-body average SAR when using a 3 mm meshed bystander human body model.
1 ,5 mm and 1 mm resolution These studies indicate that the results are not likely to change substantially with better than 3 mm resolution models
The actual resolution in terms of wavelength is generally significantly better than the worst-case scenario of one-twelfth of a wavelength, which is derived from a maximum frequency of 1 GHz and the tissues with the highest dielectric constant.
Accurate modeling of pavement is essential for bystander exposure simulations Common pavements like asphalt and concrete exhibit stable electric properties across a broad frequency range Typically, these pavements, even when wet, along with various soil conditions, can be effectively modeled as a dielectric slab with a relative dielectric constant of 8 and a conductivity of 0.01 S/m The pavement should be represented by a dielectric slab that meets the minimum thickness, relative dielectric constant, and conductivity specifications outlined in Table 1.
Relative dielectric constant Conductivity Slab thickness
Exposure simulations using thicker dielectric slabs reveal minimal impact on the peak spatial-average SAR for 1 g, 10 g, whole-body average SAR, and SAR distributions in the exposed bystander model It is important to note that the bystander model should not be in direct contact.
Remcom Inc offers Version 1.2 of VariPose®, based in State College, PA, USA This information is provided for user convenience and does not imply IEC's endorsement of the product.
The IEC/IEEE 62704-2:2017 standard specifies the use of a dielectric slab to accurately simulate the insulation gap that shoes create between the feet and the ground, with a required distance of 10 mm between the feet and the ground.
Figure 3 depicts the bystander and passenger (driver) models provided by this document for SAR simulations
Figure 3 – Bystander model (left) and passenger/driver model (right) for the SAR simulations
Exposure conditions 1 5
Vehicle-mounted antennas can expose both passengers, including the driver, and bystanders to radiofrequency (RF) radiation.
The placement of antennas on the vehicle must adhere to the specified test setup for rooftop or trunk mount installations, following the manufacturer's product manual for optimal positioning Different antenna locations may necessitate various exposure configurations for SAR evaluation, utilizing bystander and/or passenger models It is essential that all exposure configurations maintain a conservative separation distance from the antenna to the models, aligning with the actual installation, product safety guidelines, and operational requirements.
All relevant test configurations for standard passenger and bystander human bodies must be simulated to assess Specific Absorption Rate (SAR), unless compliance is established through alternative methods and exposure limits, such as the Maximum Permissible Exposure (MPE) limits outlined in IEEE Std 95.1-2005 or the reference levels set by the ICNIRP guidelines from 1998.
Passenger exposure can happen in both the front and rear seats of a vehicle For front seat exposures, it is essential to evaluate antennas located on the roof of the standardized vehicle In contrast, rear seat exposures should be assessed for antennas mounted on both the roof and trunk Additionally, front seat evaluations must consider both the driver and passenger sides.
The assessment of exposure to electromagnetic fields (EMF) from antennas in vehicles depends on their proximity to passengers If an antenna is mounted near the roof's side, exposure should be evaluated with a passenger model seated on the same side For centrally mounted antennas, the assessment should focus on the driver side In vehicles with rear seats, evaluations should include the center and one lateral side of the rear seats When antennas are not centrally located, the closest lateral side must be assessed for both passenger and bystander models In vehicles with two rows of seats, three passenger locations should be evaluated: one in the front and two in the rear If both rows are equidistant from a roof-mounted antenna, the assessment will still cover three locations, focusing on the side of the antenna For instance, with an antenna on the right side of the roof's centerline, the assessment would include the front right seat, center rear seat, and right rear seat.
The bystander separation distance from the antenna must adhere to the specifications outlined in the manufacturer’s user manual, ensuring it meets or exceeds the compliance evaluation distance Consistency with the manufacturer’s antenna installation manual is crucial for all test configurations Different mobile radio products necessitate varying compliance distances due to their RF output power, which can range from a few watts to over 100 W This document details the procedures for positioning the bystander model relative to the antenna on a standard vehicle for SAR simulation, accounting for variations in vehicle dimensions The compliance evaluation distance is defined as the measurement from the antenna surface to the nearest point on the bystander model In cases where vehicle dimensions prevent achieving this distance, adjustments to the antenna location on the standard vehicle may be necessary to meet the required separation distance If no minimum separation distance is specified in the user manual, the bystander to vehicle test distance should comply with local regulations Figure 5 provides a top view of potential bystander positions adjacent to the vehicle for both roof mount and trunk mount antenna configurations.
NOTE Some national regulatory bodies require manufacturers to include the minimal separation distance within the product manual
Figure 4 – Passenger and driver positions in the vehicle for the SAR simulations
Figure 5 – Bystander positions relative to the vehicle for the SAR simulations
Accounting for variations in population relative to the standard human body model 1 8
Whole-body average SAR adjustment factors 1 8
The standardized bystander and passenger models outlined in this document are essential for simulating exposure from vehicle-mounted antennas Individual exposure levels can vary due to physical and anatomical differences among people Notably, the whole-body average Specific Absorption Rate (SAR) can fluctuate significantly at certain frequencies, particularly influenced by body height and the whole-body resonance effect Research indicates that under plane wave exposure conditions, there is a marked enhancement in whole-body SAR around 75 MHz for the average adult, especially when the electric field of the incident wave is co-polarized with the body In contrast, shorter individuals, such as children, experience resonance at higher frequencies based on their height Consequently, for a given exposure setup and frequency, the whole-body average SAR can differ considerably depending on the height of the exposed subject.
To accurately assess the whole-body average Specific Absorption Rate (SAR) for bystanders near a vehicle, it is essential to use various body models that account for different heights and body compositions under identical exposure conditions Although the highest whole-body average SAR from these models should be reported, this evaluation process can be time-consuming and impractical.
The whole-body average Specific Absorption Rate (SAR) varies with different body sizes, as estimated for specific test configurations involving trunk or roof mount antennas and exposure conditions for bystanders and passengers To derive frequency-dependent SAR adjustment factors, these configurations were analyzed These adjustment factors must be applied to the whole-body average SAR evaluated for the defined bystander and passenger models, ensuring a conservative estimate of the whole-body average SAR under the corresponding test conditions.
The whole-body average Specific Absorption Rate (SAR) adjustment factors, derived from the IEC/IEEE 62704-2 computational study, are detailed in Tables 2 to 5, with further information available in Annex A The study highlights that the variation in whole-body average SAR across different human body models is significantly influenced by exposure conditions, such as whether the individual is a passenger or bystander, and the type of antenna mount (roof vs trunk) Consequently, separate whole-body adjustment factors are established for these varying exposure conditions Additionally, the study distinguishes between antenna locations, considering the center roof and side roof placements, with side roof defined as any position with a lateral offset from the roof center exceeding a quarter of the roof's width, while smaller offsets are categorized as center roof locations.
Linear interpolation or extrapolation of adjustment factors is applicable for frequencies ranging from 30 MHz to 1 GHz and for bystander distances that are less than the maximum distances specified in the tables However, these factors do not apply to configurations involving antenna installations that are not addressed in this document.
Table 2 – Whole-body average SAR adjustment factors for the bystander and trunk mount antennas
Bystander distance from the trunk mount antenna
NOTE 1 The factors in this table apply to all bystander positions with trunk mount antenna in Figure 5
NOTE 2 The values at 30 MHz, 800 MHz, and 1 000 MHz in this table were derived by linearly extrapolating the available results and taking the maximum between the extrapolated value and unity (see Annex B)
Table 3 – Whole-body average SAR adjustment factors for the bystander and roof mount antennas
Centre roof mount antenna Side roof mount antenna Centre and side roof mount antenna with bystander
20 cm from the vehicle 40 cm from the vehicle 20 cm from the vehicle 40 cm from the vehicle
Table 4 – Whole-body average SAR adjustment factors for the passenger and trunk mount antennas
Passenger location in the vehicle
Back seat, centre Back seat, side
Table 5 – Whole-body average SAR adjustment factors for the passenger and roof mount antennas
Centre roof mount antenna Side roof mount antenna
Passenger location in the vehicle Back seat, centre Back seat, side Front seat Back seat, centre Back seat, side Front seat
Peak spatial-average SAR adjustment factors
The evaluation of peak spatial-average SAR for 1 g and 10 g across various human body models has led to the determination of conservative adjustment factors (Annex A) As anticipated, different human body models under similar exposure conditions yield varying peak spatial-average SAR values, primarily due to differences in local tissue composition The same exposure conditions and body models utilized in the whole-body SAR variation study were also applied to assess peak spatial-average SAR Notably, the relationship between peak spatial-average SAR values and the height, mass, or age of the exposed subjects is inconsistent Consequently, the derivation of adjustment factors for all exposure conditions is based on a conservative evaluation, selecting the highest factor from all models.
Tables 6 to 9 present the peak 1 g and 10 g Specific Absorption Rate (SAR) adjustment factors for bystander, trunk, and roof mount antenna configurations, derived from the IEC/IEEE 62704-2 computational study These factors must be applied to the peak spatial-average SAR for 1 g and 10 g, as evaluated for the defined bystander and passenger models, to ensure a conservative estimate of peak spatial-average SAR under the specified test conditions Additionally, linear interpolation or extrapolation of these adjustment factors is required for all other frequencies ranging from 30 MHz to 1 GHz, as well as for bystander distances that fall below the maximum distances listed in Tables 6 to 9.
Table 6 – Peak spatial-average SAR adjustment factors for the bystander model and trunk mount antennas
MHz 1 g SAR factor 1 0 g SAR factor
NOTE 1 The factors in this table apply to all bystander positions and distances with trunk mount antenna in Figure 5
NOTE 2 The values at 30 MHz, 800 MHz, and 1 000 MHz in this table were derived by linearly extrapolating the available results and taking the maximum between the extrapolated value and unity (see Annex B)
Table 7 – Peak spatial-average SAR adjustment factors for the bystander model and roof mount antennas
Centre roof mount antenna Side roof mount antenna Centre and side roof mount antenna with bystander > 50 cm from the vehicle
20 cm from the vehicle 40 cm from the vehicle 20 cm from the vehicle 40 cm from the vehicle
Table 8 – Peak spatial-average SAR adjustment factors for the passenger model and trunk mount antennas
Back seat, centre Back seat, side
Table 9 – Peak spatial-average SAR adjustment factors for the passenger model and roof mount antennas
Centre roof mount antenna Side roof mount antenna
Back seat, centre Back seat, side Front seat Back seat, centre Back seat, side Front seat
6 Validation of the numerical models
Validation of antenna model
To validate the antenna model used in FDTD simulations, a series of test setups have been established These setups can be implemented both experimentally and numerically, offering reference solutions for the antenna's near-field regions.
In the experimental validation setup, the antenna is mounted on a square or circular ground plane to simulate vehicle operating conditions Near-field measurements are conducted to establish reference results for comparison with simulation predictions Specifically, gain monopole antennas with inductive loading coils undergo validation testing to verify the accuracy of simulation models All simulations for validating the antenna on the finite ground plane utilize the same FDTD grid resolution that will be applied in future exposure simulations involving vehicle and human body models, ensuring consistency and relevance to actual exposure scenarios.
The antenna ground plane for this experimental setup must have a minimum diameter of 1,000 mm if circular, or dimensions of at least 1,000 mm × 1,000 mm if square, with the antenna positioned at the center To ensure optimal performance, a separation of at least one wavelength from surrounding metallic objects is required; if this is not feasible, RF absorbers should cover these objects It is recommended to place absorbing materials laterally at a distance of at least half a wavelength from the ground plane to reduce residual reflections, while absorbers below the ground plane should be positioned at a minimum distance of a quarter wavelength The support structure for the ground plane should be made from non-conductive materials, such as wood or low dielectric constant plastic, to limit interference with the antenna In cases where maintaining these separation distances is challenging, particularly at low frequencies (below 300 MHz), measurements should be conducted in an anechoic or quasi-anechoic environment, such as an anechoic chamber, to meet the necessary conditions.
The electric and magnetic field magnitudes will be measured vertically above the ground plane, with the antenna positioned on it The measurement will be conducted at a distance of 200 mm from the antenna, as illustrated in Figure 6.
The IEC/IEEE 62704-2:2017 standard specifies that measurement points for electric and magnetic field values must be evenly spaced along the measurement line, with a distance of 100 mm or 0.1 wavelength, whichever is smaller.
Figure 6 – Experimental setup for antenna model validation
The antenna's feeding cable should be routed vertically downward from the center of the ground plane at the feed-point, continuing towards the RF source beneath the absorbers positioned below the ground plane.
Prior to conducting field measurements, it is essential to measure the power delivered to the antenna using a directional coupler with power meters or a network analyzer This allows for the estimation of forward and reflected power at the antenna port, enabling the calculation of the power accepted by the antenna after accounting for losses in the feeding cable Field measurements should be carried out with calibrated electric and magnetic field probes that have known measurement uncertainty, adhering to best practices for electromagnetic field measurement To reduce measurement uncertainty, it is advised to avoid using RF survey probes with sensors larger than 5 cm, as they may not accurately resolve the field distribution near the antenna and could interfere with it Automating the probe positioning during measurements is also recommended to enhance accuracy.
The numerical results must deviate from the measurements by no more than one standard uncertainty, as outlined in section 7.2.4 Due to potential conservative biases in the numerical antenna model—such as ignoring mismatch loss by enforcing perfect matching or neglecting metal losses by using Perfect Electric Conductor (PEC)—it is permissible for the numerical results to overestimate the measured outcomes, provided this overestimation falls within the established bias limits beyond the measurement uncertainty.
Reference results can be generated through precise high-resolution numerical modeling of the antenna under test, rather than relying solely on the measurements outlined in section 6.1.2 The numerical simulation model used for this validation is defined to be consistent with the specified parameters.
IEC experimental setup described in 6.1 2 The computed magnitudes of electric and magnetic field values can be used as reference data instead of measured data
For the numerical reference model of the antenna under test, a high-accuracy full-wave model must be simulated using a numerical code that preferably employs a methodology other than FDTD, such as the method of moments (MoM) for metallic structures The MoM is favored for its superior accuracy in modeling metallic structures Similar to FDTD, the MoM can model a source signal amplitude that generates a predefined radiated power, providing a valuable comparison for both near-field and far-field regions Additionally, FDTD or finite element method (FEM) codes can be utilized to achieve very high-resolution modeling of the antenna, which is not possible in vehicle exposure simulation configurations.
Numerical validation is ideal for antennas like straight wire monopoles of varying lengths and wire antennas with inductive coils, which are typically placed on the radiating element to effectively phase the induced currents along different sections These antenna structures can be accurately simulated using various numerical methods, including the Method of Moments (MoM), high-resolution Finite Element Method (FEM), or Finite-Difference Time-Domain (FDTD) techniques It is essential to report the convergence and accuracy of these simulations.
Validation of the human body model
The numerical validation of the bystander and passenger models aims to ensure the accuracy of the modeling within a specific FDTD code by simulating exposure to a defined RF source, such as a plane wave This validation confirms that the models are correctly integrated into the simulation project and adheres to the SAR simulation procedures outlined in this document and IEC/IEEE 62704-1 It evaluates the deviation of computed SAR results from reference values, using standard voxel body models without modifications Simulations are conducted in free space with a vertically polarized incident plane wave, assessing front and back exposure conditions separately The frequency for validation simulations should closely match the actual exposure configuration, with all SAR results normalized to an average plane wave power density of 1.0 W/m² The passenger and bystander models are simulated at least 200 mm away from the perfectly matched layers at the computational domain boundaries, with alternative boundary conditions as specified in IEC/IEEE 62704-1.
The reference Specific Absorption Rate (SAR) results for the bystander model are presented in Table 1, while those for the passenger model are shown in Table 1 These results were calculated for plane wave exposure, utilizing 3-mm resolution models that are uniformly meshed across the entire computational domain.
To ensure valid and applicable test results for simulation accuracy assessment, all compliance simulations must utilize the 3-mm resolution standard body models of bystanders and passengers.
Table 1 0 – Peak spatial-average SAR for 1 g and 1 0 g and whole-body average SAR for the front and back plane wave exposure of the 3-mm resolution bystander model
1 g SAR 1 0 g SAR Whole-body average SAR
Front Back Front Back Front Back
Figure 7 depicts the described benchmark validation configuration featuring the bystander exposed to a front and back plane wave
Figure 7 – Benchmark configuration for bystander model exposed to a front or back plane wave
Table 1 1 – Peak spatial-average SAR for 1 g and 1 0 g and whole-body average SAR for the front and back plane wave exposure of the 3-mm resolution passenger model
1 g SAR 1 0 g SAR Whole-body average SAR
Front Back Front Back Front Back
The corresponding validation configuration with the front and back plane wave exposure is depicted in Figure 8
Figure 8 – Benchmark configuration for passenger model exposed to a front or back plane wave
The validation shall be considered successful if the computed peak spatial-average SAR for
The measured values of 1 g and 10 g are within ±10%, and the calculated whole-body average Specific Absorption Rate (SAR) is within ±5% of the reference values outlined in Tables 10 and 11 Additionally, the locations of the peak spatial-average SAR must align with the specifications detailed in Annex C for the respective configurations.
Validation of the vehicle numerical model
General
The numerical model and simulation configurations outlined in section 6.3 confirm the accurate representation of the vehicle model and its meshing within the FDTD grid This validation test effectively simulates the electromagnetic field generated by a precisely characterized quarter wave monopole antenna mounted on the vehicle trunk, allowing for a thorough assessment of the vehicle model's accuracy The chosen frequency for this specific validation simulation is also defined.
6.3.2 and 6.3.3 shall be the closest to the actual frequency of the exposure configuration to be simulated for compliance evaluation.
Vehicle model validation for bystander exposure simulations
The configuration, illustrated in Figure 9, incorporates the pavement model outlined in section 5.4 The standard vehicle model will be meshed using the same FDTD grid resolution designated for exposure simulations during compliance evaluations Additionally, a quarter wave monopole antenna will be positioned at the center of the trunk.
NOTE Relative to the coordinate system of the standard vehicle CAD model, the antenna base is located at
Figure 9 – Configuration for vehicle numerical model validation
The lengths of the quarter wave antenna at selected frequencies are listed in Table 1 2
Table 1 2 – Antenna length for the vehicle model validation configurations
The electric and magnetic field values will be calculated along a vertical line positioned 20 cm behind the vehicle body, with measurements taken at ten locations starting from 20 cm above the ground and spaced at 20 cm intervals, as illustrated in Figure 9 This line is situated within the symmetry plane of the car.
The reference results for the validation model were derived as an average from various simulation tools that comply with IEC/IEEE 62704-1 standards, as shown in Table 1 All results are normalized to an average power output of 1 W from the antenna.
Table 1 3 – The reference electric field (top) and magnetic field (bottom) values for the numerical validation of the vehicle model for bystander exposure
Position above the ground cm
33 MHz 80 MHz 1 50 MHz 450 MHz 800 MHz 1 000 MHz
Position above the ground cm
33 MHz 80 MHz 1 50 MHz 450 MHz 800 MHz 1 000 MHz
Successful validation occurs when the deviation calculated using Formula (5) in section 7.2.3, based on the reference results in Table 1 3, is less than 30% This threshold for maximum allowed deviation reflects the variability of values generated by different codes at a specific location compared to the mean reference value provided in Table 1 3 for that location.
The achievable deviation of Specific Absorption Rate (SAR) from the target values set for the bystander benchmark simulation models in section 8.2 is limited to 1.7%, significantly lower than the deviations observed in the electric and magnetic field values calculated in section 6.3.2.
Vehicle model validation for passenger exposure simulations
Vehicle model validation for passenger and driver exposure simulations will follow the same procedures as the validation for bystander exposure, as outlined in section 6.3.2 The key difference lies in the specific locations where electric and magnetic fields are measured for comparison with reference values The coordinates for these test points, which are detailed in Table 1, are strategically placed in both the back and front seat areas of the vehicle.
Table 1 4 – Coordinates of the test points for the standard vehicle validation simulations for the passenger
Test point Location X mm Y mm Z mm
The validation test results, obtained using various simulation tools that comply with the requirements outlined in this document and validated according to IEC/IEEE 62704-1, are presented in Table 1 All results are normalized to an average power emission of 1 W from the antenna.
Table 1 5 – The reference electric field (top) and magnetic field (bottom) values for the numerical validation of the vehicle model for passenger exposure
33 MHz 80 MHz 1 50 MHz 450 MHz 800 MHz 1 000 MHz
33 MHz 80 MHz 1 50 MHz 450 MHz 800 MHz 1 000 MHz
Successful validation occurs when the deviation calculated using Formula (5) in section 7.2.3, based on the reference results in Table 1 5, is below 45% This threshold for maximum allowed deviation reflects the variation observed in values generated by different codes at a specific location, compared to the reference value provided in Table 1 5 for that location.
The achievable deviation of Specific Absorption Rate (SAR) from the target values for the passenger benchmark simulation model is limited to 20%, significantly lower than the deviations observed in the electric and magnetic field values calculated within the unloaded vehicle model.
General considerations
The SAR computational procedure, like any experimental method, carries a level of uncertainty in numerical exposure evaluations under specific conditions and in the representation of antennas To reduce this uncertainty, all SAR evaluations must utilize computational code validated in accordance with Clause 8 of IEC/IEEE 62704-1 This validation is typically conducted by the software vendor However, for SAR evaluations involving vehicle-mounted antennas, not all validation tests outlined in IEC/IEEE 62704-1 are necessary, specifically sections 8.3.3 and 8.3.4.
The IEC/IEEE 62704-2:2017 standard specifically addresses exposure simulations for portable transmitters designed to operate within 200 mm of the human body, and therefore, certain elements may be excluded from consideration.
Validation tests for bystander and passenger exposure evaluations are outlined in Clause 6 These tests must be conducted at frequencies that align with the operating bands of the antenna for which numerical Specific Absorption Rate (SAR) evaluations are necessary, and the findings must be documented.
Contributors to overall numerical uncertainty in standard test configurations
General
The validation test procedures outlined in Clause 6 assess the overall uncertainty of SAR simulations in standard test configurations by evaluating contributions from various components of the numerical model Specifically, computational and modeling uncertainties are categorized into four areas: a) uncertainty of the numerical algorithm; b) uncertainty of the numerical representation of the antenna model; c) uncertainty of the numerical representation of standard vehicle and ground models; and d) uncertainty of SAR evaluation in standard bystander and passenger models.
The physical model of the antenna and vehicle is initially represented through a CAD model, which is then rendered in the simulation grid Two distinct uncertainty contributions are identified: the first pertains to the numerical representation of the vehicle CAD model and pavement modeling using a quarter wave antenna, as specified in Table 1 2 The second uncertainty relates to the antenna modeling utilized in the actual exposure simulation.
Uncertainty of the numerical algorithm
The evaluation of the numerical algorithm's uncertainty will follow the guidelines outlined in Table 3 of IEC/IEEE 62704-1 It is important to note that the contribution from mesh resolution (device rendering) is excluded from this assessment, as it has already been addressed in sections 7.2.3 and 7.2.4 The combined standard uncertainty, as specified in Table 3, should be recorded directly in column g of Table 16, excluding the mesh resolution contribution.
Uncertainty of the numerical representation of the vehicle and pavement
The standardized vehicle model allows for a priori evaluation of uncertainty in its numerical representation, significantly influencing the electromagnetic (EM) field distribution both inside and outside the vehicle The impact of the vehicle's metal body and the specific ground model on exposure assessments is crucial Uncertainty in Specific Absorption Rate (SAR) simulations can be analyzed using a quarter wave antenna positioned at a predefined location on the vehicle, as outlined in section 6.3 This analysis involves comparing reference results from the standard vehicle model with those obtained from the numerical tool used It is essential to assess this uncertainty separately for bystander and passenger exposure conditions, requiring the computation of electric and magnetic field amplitudes at designated reference test points, as detailed in Tables 1-3 for bystander exposure and Tables 1-4 and 1-5 for passenger exposure.
The uncertainty for the 1 g and 10 g peak spatial-average Specific Absorption Rate (SAR) is determined by the maximum deviation of the calculated electric and magnetic field amplitudes from their respective reference values This maximum deviation for each validation is computed accordingly.
SAR 2 max ref num 2 ref 2
The uncertainty contribution for the whole-body average Specific Absorption Rate (SAR) is determined by the combined deviations of all test points from their respective reference values Each validation's combined deviation is calculated accordingly.
WB ref cert vehicle_un [%] 1 00 , (6) where
E ref (P i ) is the reference magnitude of the electric field evaluated at the test point P i ;
H ref (P i ) is the reference magnitude of the magnetic field evaluated at the test point P i ;
E ref max (P n ) is the maximum electric field reference magnitude across all the evaluated test points P i , which occurs for some i = n;
H ref max (P n ) is the maximum magnetic field reference magnitude across all the evaluated test points P i , which occurs for some i = n;
E num (P i ) is the magnitude of the electric field numerically determined at the test point P i using the numerical code;
H num (P i ) is the magnitude of the magnetic field numerically determined at the test point
MAX is the operator that selects the maximum value of the data set in parentheses
The highest deviations calculated using Formulas (5) and (6) are recorded in column c of Table 1 6 The standard uncertainty contributions derived from these deviations are computed as specified in column g of Table 1 6 and are also included in that column.
Uncertainty of the antenna model
Vehicle-mounted antennas vary significantly, ranging from simple wire monopole radiators to more complex designs like monopoles with reactive coils Each antenna type must be individually evaluated for compliance, even if supplied with the same mobile radio product Typically, these antennas are mounted on the vehicle's metal body, relying on the ground plane near the mounting location, such as the trunk or roof To accurately assess antenna modeling uncertainty, a similar operating condition must be established to evaluate the near-field produced by the transmitting antenna reliably For antennas mounted on a metal ground plane, the reference condition for uncertainty assessment is outlined in section 6.1, with electric and magnetic field values determined at all test points as specified in sections 6.1.2 and 6.1.3 Uncertainty components are calculated based on the maximum deviation of the simulated numerical values from the corresponding reference values.
2 max ref num 2 ref 2 ainty 1
IEC/IEEE 62704-2:201 7 – 33 – © IEC/IEEE 201 7 where
E ref (P i ) is the reference magnitude of the electric field evaluated at the test pointP i ;
H ref (P i ) is the reference magnitude of the magnetic field evaluated at the test pointP i ;
E num (P i ) is the numerically determined magnitude of the electric field evaluated at the test point P i;
H num (P i ) is the numerically determined magnitude of the magnetic field evaluated at the test point P i ;
E ref max (P n ) is the maximum electric field reference magnitude across all the evaluated test points P i , which occurs for some i = n;
H ref max (P n ) is the maximum magnetic field reference magnitude across all the evaluated test points P i , which occurs for some i = n;
MAX is the operator that selects the maximum value of the data set in parentheses
The maximum deviation, calculated using Formula (7), is recorded in column c of Table 1 6 Additionally, the standard uncertainty contribution derived from this deviation is specified in column g of Table 1 6.
If the simulated antennas are straight wire monopoles, as outlined in section 6.3 for vehicle numerical model validation, no additional uncertainty contribution is necessary since it is already accounted for in the results evaluated in section 7.2.3.
Uncertainty of SAR evaluation in the standard bystander and passenger models
The uncertainty contribution related to the standard bystander and passenger models in FDTD grid and SAR calculations is assessed under the plane wave exposure conditions outlined in section 6.2 The human body models will be represented in the FDTD grid with a resolution of 3 mm, as required for compliance exposure simulations from vehicle-mounted antennas Separate evaluations will be conducted for both the bystander and passenger models The specifics of the exposure conditions are detailed in section 6.2, and the SAR uncertainty contribution is calculated based on the variations in the computed peak spatial-average SAR values for 1 g.
The whole-body average Specific Absorption Rate (SAR) values for bystander and passenger models are derived from the reference SAR values listed in Tables 10 and 11 Relative deviations for all exposure conditions outlined in section 6.2 will be calculated using the formula: \( \text{ref num body\_model ref y uncertaint} \, [\%] \).
SAR ref is the reference peak spatial-average SAR for 1 g, peak spatial-average SAR for 1 0 g or whole-body average SAR values;
SAR num is a peak spatial-average SAR for 1 g, peak spatial-average SAR for 1 0 g or whole- body average SAR values determined using the numerical code
The maximum deviation for the peak spatial-average SAR for 1 g and/or 10 g, as well as the whole-body average SAR, is calculated using Formula (8) and presented in column c of Table 16 The standard uncertainty contributions derived from these deviations are detailed in column g of Table 16.
Uncertainty budget
The overall uncertainty budget shall be reported and described according to Table 1 6
Table 1 6 – Numerical uncertainty budget for exposure simulations with vehicle mounted antennas and bystander and/or passenger models a b c d e = f(d,h) f g = c × f / e h
Deviation/ uncertainty Prob dist Div c i
Numerical model of the vehicle 7.2.3 R √ 3 1 ∞
SAR evaluation in the standard human body model 7.2.5 R √ 3 1 ∞
NOTE 1 Column headings a to h are given for reference
The table includes several abbreviations: a) Div refers to the divisor used to calculate standard uncertainty, which depends on the probability distribution in column d and the effective degrees of freedom ν eff in column h; b) 1 g, 1 0 g, and WB denote the uncertainty components of the peak spatial-average SAR for 1 g and 1 0 g, as well as the whole-body average SAR; c) R indicates rectangular probability distributions; d) k represents the coverage factor; and e) c i signifies the sensitivity coefficient.
The sensitivity coefficient ci is applied to convert each uncertainty component into the corresponding standard uncertainty for the SAR
The validation results for the standard vehicle model and human body model were independently produced using up to five distinct numerical codes, as detailed in sections 6.2 and 6.3 These findings indicate that the maximum combined uncertainty contribution is less than a specified threshold.
The combined standard uncertainty of numerical Specific Absorption Rate (SAR) results is influenced by the antenna model's uncertainty contribution, particularly when evaluated separately Achieving lower uncertainty is possible through specific implementations of recommended procedures For instance, benchmark model simulations, computed independently with three different numerical codes in accordance with IEC/IEEE 62704-1 standards, showed SAR deviations of less than 17% for bystander exposure and 20% for passenger exposure, relative to the target values defined in sections 8.2 and 8.3.
General
The benchmark models outlined in Clause 8 are essential for evaluating the accuracy of Specific Absorption Rate (SAR) calculations related to bystander and passenger exposure under standard simulation conditions These assessments utilize a standard vehicle model equipped with a monopole antenna positioned on the trunk, alongside the standard passenger or bystander model during benchmark simulations It is crucial that these simulations are conducted using codes that have been validated in accordance with the procedures specified in IEC/IEEE 62704-1.
Benchmark for bystander exposure simulations
The bystander exposure configuration features a standard vehicle model equipped with a trunk-mounted monopole antenna The standard bystander model is located 1.0 meters behind the vehicle, as illustrated in Figure 10 For benchmark simulations, the antenna lengths corresponding to specific operating frequencies listed in Table 12 will be utilized.
Figure 1 0 – Side view (top) and rear view (bottom) benchmark validation configuration for bystander and trunk mount antenna
The distance \( d \) between the antenna and the bystander is defined as the horizontal distance projected onto the ground, measured between the nearest vertical lines touching the antenna and the closest point on the bystander model Simulations will be conducted with the bystander model oriented towards the vehicle, while all other conditions and parameters will adhere to the specifications outlined in this document.
The bystander and vehicle models are defined in separate coordinate systems, with the bystander's position determined by the coordinates of the bounding box corner closest to the front of the right foot This corner is located at (−291, 424, 10) mm in the vehicle coordinate system, which corresponds to the bystander bounding box center at (3, 596, 931) mm, as detailed in Annex A.
Benchmark test configurations will be simulated at frequencies specified in Table 1 2, which are nearest to the actual compliance exposure simulation frequencies The antenna for the benchmark test frequency must be positioned at the center of the trunk along the vehicle's symmetry line For the standard vehicle model, this center point is located in the trunk area.
IEC d is 37 cm from the trunk edge at the rear window, which corresponds to (0, −575, 1 080) mm relative to the vehicle coordinate system
The benchmark simulation will utilize the 3-mm resolution standard bystander model to calculate the 1 g and 10 g peak spatial-average Specific Absorption Rate (SAR) as well as the whole-body average SAR These SAR results will be normalized to an average power output of 1 W from the antenna and will be compared against the reference data listed in Table 17 To successfully pass the benchmark test, the computed SAR values must fall within the expanded uncertainty specified in Table 16 when compared to the corresponding reference values in Table 17 Additionally, Annex C provides the peak spatial-average SAR locations for each frequency for further reference.
Table 1 7 – Reference SAR values for the bystander benchmark validation model
Frequency 1 g peak spatial-average SAR 1 0 g peak spatial-average SAR Whole-body average SAR
Benchmark for passenger exposure simulations
The passenger exposure configuration features a standard vehicle model equipped with a trunk-mounted monopole antenna, with the standard passenger model placed centrally in the back seat, as illustrated in Figure 1.
The passenger and vehicle models are defined in distinct coordinate systems, with the passenger's correct position determined by the coordinates of the bounding box corner closest to the model's right foot This corner is positioned at (−291, −2,726, 284) mm in relation to the vehicle's coordinate system, which aligns with the bystander's bounding box center coordinates of (3, −2,145.5, 872) mm, as detailed in Annex A.
Figure 1 1 – Benchmark validation configuration for passenger and trunk mount antenna
For benchmark simulations, the antenna lengths corresponding to the specific operating frequencies listed in Table 1 2 must be utilized Simulations of the benchmark test configurations should only be conducted at the frequencies specified in Table 1 2 that are nearest to the actual compliance exposure simulation frequencies.
Table 1 8 – Reference SAR values for the passenger benchmark validation model
Frequency 1 g peak spatial-average SAR 1 0 g peak spatial-average SAR Whole-body average SAR
The antenna should be positioned at the center of the trunk, aligned with the vehicle's symmetry line In the standard vehicle model, this central point is located 37 cm from the trunk edge at the rear window, corresponding to the coordinates (0, −575, 1 080) mm in the vehicle's coordinate system.
The 3-mm resolution standard passenger model shall be used for the benchmark simulation to compute 1 g and 1 0 g peak spatial-average SAR and the whole-body average SAR The SAR
IEC results must be normalized to an average power of 1 W transmitted by the antenna and compared with the reference data in Table 1 8 To meet the benchmark test, the computed Specific Absorption Rate (SAR) values should fall within the expanded uncertainty outlined in Table 1 6 when compared to the reference values in Table 1 8 Additionally, Annex C provides the peak spatial-average SAR locations for each frequency for further reference.
General
The SAR results must be reported for all simulated configurations, even those that do not yield the highest peak spatial-average or whole-body average SAR values Additionally, the report should include pertinent information regarding the simulated configurations and antenna modeling Clause 9 outlines the recommended procedures for documenting numerical SAR evaluations specifically for vehicle-mounted antennas.
Test device
The test device, a mobile radio, must be described with key details including its operating frequency band, maximum transmit power, applicable duty cycle, and the geometry and location of the antenna on the vehicle.
Simulated configurations
The article will detail the simulation configurations, highlighting key aspects such as the transmit frequency, antenna model name or kit number, and comprehensive specifications including antenna type and length Additionally, it will specify the antenna's location on the vehicle, the exposure condition (whether bystander or passenger), and provide precise details regarding the exposure location, including the distance of bystanders from the antenna and the positioning of passengers inside the vehicle.
Software and standard model validation
The report must include a detailed description of the simulation software, specifying its name and version It is essential to document the software's compliance with the requirements outlined in this document and IEC/IEEE 62704-1, along with all relevant validation test results Additionally, evidence of the validation of standard vehicle and human models, following the specified procedures, should also be documented.
Antenna numerical model validation
The validation results of the antenna modeling will be presented, highlighting the reference results and the discrepancies between the simulated outcomes and the reference data outlined in sections 6.1 and 7.2.4 Additionally, if necessary, further results may be obtained from simulations, including a comparison of measured and simulated return loss, which is frequently requested by regulatory authorities to substantiate the antenna model validation.
Results of the benchmark simulation models
The benchmark simulations for passenger and bystander exposure configurations at relevant frequencies will be reported, highlighting any deviations from the reference data outlined in Clause 8.
Simulation uncertainty
The report must include an evaluation of computational and modeling uncertainty as outlined in Clause 7 and Table 1 6 Additionally, it should detail the minimum and maximum mesh resolution along with the absorbing boundary conditions of the computational domain Regulatory bodies may impose further requirements based on the simulation uncertainty for the acceptance of results.
SAR results
All simulated configurations must report the applicable 1 g and/or 10 g peak spatial-average SAR and whole-body average SAR results, including detailed descriptions of the simulation configurations as outlined in section 9.3, to ensure a clear association between individual SAR results and their respective configurations Furthermore, SAR results should be adjusted according to the factors specified in section 5.6, with both the scaled and original (non-scaled) SAR values presented together.
The report must clearly identify the maximum scaled peak spatial-average SAR and the maximum scaled whole-body average SAR from all simulated conditions It should include SAR distribution plots for the maximum scaled peak spatial-average SAR in a 2D color map format, corresponding to the cross section of the human model (sagittal or coronal) that contains the peak spatial-average SAR location If a single plot does not adequately illustrate the peak SAR location, multiple cross sections should be utilized Additionally, the report must provide electric and magnetic field distributions in the cross section aligned with the antenna.
Annex A (normative) File format and description of the standard human body models