IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY. VOL 37, NO 2. MAY I995

Measurement of Equivalent Power

183

Density and RF Energy Deposition in the Immediate Vicinity

of a 24-GHz Traffic Radar Antenna Quirino Balzano, Senior Member, IEEE, John A. Bergeron, Jules Cohen, Life Fellow, IEEE,

John M . Osepchuk, Life Fellow, IEEE, Ronald C. Petersen, Member, IEEE, and Leon

Abstract- In response to reports of alleged health effects associated with the use of hand-held traffic radars, e.&, testicular cancer, a study was undertaken to quantify: 1) The distribution of the electric field ( E ) in the immediate vicinity of the antenna aperture of a typical 24-GHz hand-held traffic radar; and 2) the relative match and the corresponding depth of penetration of the absorbed energy when the antenna aperture is positioned in contact with material having properties simiiar to human tissue. The former .measurements are important for assessing a large number of measurements reported by others [11-[31 using techniques that are not expected to be reliable indicators ofpower density at distances small compared with the size of the radiator; the latter are important for verifying predictions that energy from the antenna at 24-GHz will be predominantly absorbed in superficial tissue, Le., the skin. The results of tbe first part of the study indicate that the equivalent-plane-wave power density in the plane of the aperture of the radar antenna is approGmately one- half of the corresponding values reported in the literature; the results of the second part of the study indicate that the depth of penetration into material simulating human skin is approximately 0.5 mm.

I. INTRODUCTION

ECENTLY, several reports have appeared in the media R claiming that the use of hand-held traffic radars may cause injury to the user. In spite of a lack of a scientific basis for such claims, these reports have led to a U.S. Senate hearing, a number of lawsuits and a ban on the use of hand-held radars in the State of Connecticut. The questions of maximal exposure level and distributions of the absorbed energy are rarely considered in media reports of injury. While several studies in the scientific literature report measured near-field data for a large number of devices [1]-[3], there are questions relating to the measurement protocol, Le., the use of large (with respect to wavelength) probes placed in close proximity to an antenna aperture of approximately the same size, and the interpretation of the results, i.e., reporting the results as “power density.” A specific measurement protocol has been followed rigorously and a wealth of data has been recorded. The purpose of this study was to estimate the relative accuracy and interpretation of the 24-GHz results reported in the literature and, hence, their utility for addressing the question of exposure. In addition,

Manuscript received March 2, 1994; revised December 15, 1994. The authors are members of the Science and Technology Committee of the

Electromagnetic Energy Association, Washington, DC 20037 USA. IEEE Log Number 9410262.

M. Roszyk

while the depth of penetration for plane-waves incident on a planar tissue surface can be readily calculated, there is no reason to believe that the penetration depth associated with the complex fields near a 7.5 cm diameter antenna aperture will necessarily be the same as for the planar case. Therefore, as a second part of the study, the penetration depth with the antenna aperture in contact with a tissue (skin) equivalent medium, was measured.

11. SYSTEMATIC ERRORS ASSOCIATED WITH THE USE OF “LARGE” PROBES

There are many studies in which microwave power density directly in front of typical traffic radars is reported [I]-[3]. In each study, relatively large measurement probes (compared with the wavelength and size of the radar’s antenna aperture) were used and the measurements were made at distances sufficiently close to the antenna that questions relating to the interpretation of these measurements arise. Sipce literally thousands of traffic radars have been measured following such procedures, it is important to assess the magnitude of the systematic error in the reported results. This portion of the study was designed to provide a reference to the existing data.

There are two pitfalls that may seriously influence measure- ment accuracy when electrically large electric field (E-field) sensors are used to evaluate equivalent power density near the aperture of small antennas. First, the electromagnetic fields close to such antennas are extremely complex spatially and in phase and orientation. The measurement probes used in the studies reported in the literature [ 11-[3] are substantially larger than the wavelength (at 24-GHz) and, hence, the non- uniform fields have extreme gradients over the length of the probe sensors. Since the probes used in these studies have been calibrated in uniform, plane-wave fields, there is no reason to believe that the same calibration factors will give accurate readings for the very different conditions near antenna apertures with dimensions comparable with those of the probe.

The second pitfall is associated with the energy stored in the E - and H-fields near the antenna aperture, especially at distances small with respect to the size of the aperture. An antenna stores a substantial amount of energy in both the E- and H-fields in the space immediately adjacent to the aperture. To illustrate this phenomenon for a horn antenna, note that the entire aperture contributes to the electric field at each point

0018-9375/95$04.00 0 1995 IEEE

184 IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 31, NO. 2, MAY 1995

(a)

A e l e m e n t a r y radiators

(bi

Fig. 1. antenna.

(a) Near-field region of a horn antenna. (b) Far-field region of a horn

in free space. (See Fig. l(a).) In other words, the aperture should be considered an aggregate of a large number (infinite in the limit) of small subapertures (elementary radiators), each with its own illumination and each contributing to the total field at any point of measurement. In the space adjacent to the aperture, the contributions from each elemental radiator are not necessarily in phase with the others and their phases change very rapidly from point to point even over a short distance. This is because of the substantial variation in phase, path length and direction between each elemental radiator and the measurement point. The energy associated with rapid spatial changes of the electric or magnetic fields is called evanescent because it disappears at sufficiently large distances from the aperture, i.e., where small changes of the coordinates of the measurement point do not significantly change the path length from the elemental radiators. (See Fig. l(b)). Thus, the direction of the Poynting vector remains constant only at distances where the direct paths from every elemental radiator on the antenna aperture to the observation point are essentially parallel.

It is obvious that close to the elemental radiators the total electric field (the measurable physical entity) embodies both the evanescent and the transported energy. Making a precise measurement of the E-field in these conditions and assigning the entire value to energy transport, Le., equivalent power density, overestimates the radiation fields and can lead to a substantial overstatement of the total radiated power. It is essential in making near-field measurements to somehow separate the total power into the component radiated and the power stored in the fields near the aperture.

Attempts to relate the measured field values to the RF power available from the antenna aperture in order to assess the accuracy of the measurement methodology (and verify that the results are consistent everywhere in the field) have not been reported in the earlier studies [ 11-[3]. Further, measurement probes that were calibrated in a uniform, plane-wave field [4], with probe elements substantially larger than the wavelength of the fields being measured, were used in these studies. Thus,

there is no reason to assume that the use of a probe with 1 ” diameter E-field sensors placed at a distance of 5 cm from a 7.5-cm diameter antenna aperture will produce accurate results. In order to assess the magnitude of the measurement error associated with the results reported, in the literature, measurements, using miniature probes, of the electric field distribution near the aperture of a 24-GHz FALCON (Kustom Signals) hand-held traffic radar are compared with reported measured results for the same radar using conventional RF survey meters.

111. MATERIALS AND METHODS

The E-field probes used for this study were an EIT model 979 [5 ] and a NarddBRH Model 14 Implantable Probe. The sensors of the EIT probe are 2.5 mm long, or about 116 of a wavelength at 24.15 GHz (A = 1.25 cm). The sensors of the Narda probe are 1 mm long or less than 1/12 of A. While the length of the latter is more satisfactory for examining the details of the E-field structure in the near- field of a small aperture at 24-GHz, because of sensitivity limitations associated with the short electrical length of the antennas, the EIT probe was used for most measurements. The EIT probe is shown in Fig. 2(a); the Nard-RH probe is shown in Fig. 2(b). Each probe has been calibrated using a horn antenndmicrowave assembly, identical to that of the FALCON model radar, with the aperture short circuited in one plane with a metal septum to obtain linear polarization. The modified horn is shown in Fig. 3. The magnitude of the square of the electric field strength [El2 was measured with an EIT digital radiation monitor model 679A, which has an optical link between the sensor and the measuring equipment, i.e., a millivoltmeter.

Each probe has three sensors; the EIT 679A has three independent ports each of which can be zeroed and calibrated independently. Channel “3” of the EIT probe (see Fig. 2(a)) was used to determine the polarization of the field along the antenna axis at a distance of 10 cm from the aperture by first rnaximizing the reading of the instrument, then rotating the probe by 90 degrees and repeating the field strength measurement. The ratio of the two values was found to be larger than 20 dB, showing that the field was linearly polarized.

Sensors “1” and “2” were independently calibrated to pro- vide 1/2 the reading of sensor “3” in the identical position with respect to the aperture. The probe was then rotated about its axis at various points in the E plane 10 cm from the aperture. The readings from the probe at each point varied less than 1.5 dB, which is acceptably isotropic for this type of measurement.

The extent to which the transmission lines feeding the probe may interact with the fields, thereby resulting in errors, was also examined. To evaluate the decoupling between the sensors and their feed lines, the probe handle was aligned in the direction orthogonal to the E-field 10 cm from the aperture. The probe handle in this condition is also orthogonal to the axis of the horn assembly (position 1 in Fig. 4). The values of E-field recorded on channel “3” were identical in this position to the corresponding values previously measured. The probe handle was then rotated to position 2 (Fig. 4) which aligned the

BALZANO et NI.: MEASUREMENT OF EQUIVALENT POWER DENSITY AND RF ENERGY DEPOSITION I85

(b)

Fig. 2. forward. (b) NardaEXRH Model 14 implantable probe.

(a) EIT Model 979 Isotropic RF probe. Channel “3” is shown facing

sensor feed lines with the E-field, and the sensor orthogonal to the direction of E. The field strength was again measured and the readings found to be 20-25 dB below those of position 1. Similar procedures were used for sensors “1” and “2” with analogous results. Thus, any errors associated with coupling of the transmission lines were sufficiently small so as not to effect the overall accuracy of the measurements. In addition, the linearity of the sensors was determined by observing the reading of each channel to ensure that it varied as cos2, with respect to the angle between the sensor and the field. Similar tests were conducted on the NardalBRH implantable probe to ensure its applicability. The NardaBRH probe was also found acceptable in terms of isotropic response, linearity and decoupling of the sensors from the feed lines. Thus, the accuracy of the probes was found acceptable for measuring the strength of the E-field near the aperture in question.

The probes were positioned by means of a robot (Intelledex Model Microsmooth 660) during the actual measurement of the distributions of the square of the electric field strength ( / E l 2 ) . The robot was controlled by an IBM PS2 computer which also recorded the output of the EIT millivoltmeter (by means of a Hewlett Packard multimeter). The positioning accuracy of the robot arm was f l mm. The measuring apparatus is shown in Fig. 5. The optical links between the

Fig. 3 . calibration.

A horn antenna from a traffic radar as modified for purposes of

Fig. 4. E-field calibration geometry.

probe assembly and the EIT millivoltmeter can also be seen. A 2 ft. square section of microwave absorber was placed behind the probe to eliminate reflections from the robot arm. Earlier scans along the axis of the horn very clearly indicated the presence of a reflected wave in the region close to the antenna. Results of a typical scan without the absorber are shown in Fig. 6. Note the periodicity of the pattern, which indicates the presence of a scatterer at a constant distance from the E-field sensor. The scatterer was found to be the robot arm grip holding the box which converts the dc signals from the E-field sensors to optical signals.

Final calibration of the EIT and the Narda/BRH implantable probes, in terms of the square of the magnitude of the electric field strength, (V2/m2) was carried out using a Narda Broadband Isotropic Radiation Monitor, Model 8606 (with a Model 8621 probe) as a secondary standard. The broadband probe (8621) was sent to the manufacturer and calibrated one

I86 IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 37, NO. 2, MAY 1995

c.. 0 5 E u s 3 0 4

8 5 03 P e! E P 02

e

a

- U

H 0 1 m

00

Z -centimeters

Fig. 7. Vertical scan along the antenna axis (with absorber in place). The ordinate is proportional to \ElZ. The distance beyond which E decreases as l / r (approximately 15 cm) can clearly be seen.

distances. The readings of the Narda 8621 were recorded at 50 cm (maximum extension of the robot arm) while the probe was rotated around its axis. The readings were within 1 dB. The average measured value was considered the calibration field which was used to calibrate both the EIT and the NARDABRH Implantable probes. Also, from an on-axis measurement of the power density at 50 cm, it is possible to determine (within 1 dB) the total power radiated by the aperture if the gain of the hom assembly and the radiated Fig. 5. Measurement apparatus lncludtng the robot, probe, absorber, and

FALCON hand-held traffic radar.

I 0 5 10 15 20 25 M 35

0.0 I

Z - centimeters

Fig. 6. Vertical scan along the antenna axis (without absorber) showing the effect of reflections from the robot arm. The ordinate is proportional to \El'.

week before the measurement program began. The EIT and Narda/BRH probes were calibrated as follows.

From the axial scans it is possible to determine the distance ( T ) from the aperture beyond which the E-field decreases as 1/r. Beyond this point, /El = I C / T and it is meaningful to define power density S as lEI2/377. It is appropriate to use a conventional Narda Probe, e.g., Model 8621 at these

power are known from ather far field measurements.

Iv. MEASURED VALUES OF IEI2 NEAR THE APERTURE OF A TYPICAL 24-GHz RADAR

Fig. 7 shows a typical scan obtained by moving the EIT probe away from the aperture. At T equal to zero, the sensors are resting against the plastic lens cover of the aperture. Note how /El2 changes rapidly within the first 10 cm from the aperture and then begins to fall off. The ordinate of Fig. 7 is in volts squared per centimeter squared (V2/cm2). It can be seen in Fig. 7 that the near-field (Fresnel zone) of the antenna extends out to about 2 times the aperture diameter (-14 cm).

Fig. 8 shows a typical distribution of IEI2 over the aperture of the radar at essentially T = 0 (a distance where the sensors almost graze the plastic cover). Note the mountain-like shape of the pattem. The horn feeds the aperture with an illumination that is not quite axisymmetric, i.e., the peak of the illumination is slightly off axis.

Fig. 9 shows a typical distribution of IEI2 over an area of 13 cm x 13 cm at a distance of 5 cm from the aperture. Note the deep crater in the center of the pattem. At this close distance (5 cm), the contribution to the on-axis E-field from one half of the antenna is almost in phase opposition with the contribution from the other half. It is now clear why it is not possible to rigorously define power flow, or power density, close to an RF source: The energy is being transported in many directions. In addition, the energy is still being stored in the space immediately surrounding the aperture and at some

BALZANO et ai.: MEASUREMENT OF EQUIVALENT POWER DENSITY AND RF ENERGY DEPOSITION

/--- I /

15

187

8.5

Fig. 8. not axisymmetric.

/E(' distribution ( r = 0 cm). Note that the aperture illumination is Fig. 10. IEl' distribution (r = 9.5 cm). Note that the distribution is becoming smoother than at closer distances indicating that most of the energy is being propagated away from the antenna.

0 4

%.

E 03 2

8 0.2

P e 6i 2 0.1

0 5 .c

U 0

f iz X -centimeters

5 5

Fig. 9. indicating energy is being returned to the source.

/El2 distribution ( r = 5 cm). Note the depression along the axis

locations, e.g., 5 cm from the aperture, it is clear that the energy is actually travelling back toward the aperture.

Fig. 10 shows the dstribution of ]El2 at a distance of 9.5 cm from the aperture. At this distance, there is less structure to the distribution indicating that the phenomenon of energy storage is no longer as prominent as for much closer distances. In principle, energy storage continues to some extent to at least a distance d F = 2D2/A D = 80 cm, (D = antenna diameter) . The distance d F marks the end of the Fresnel transition zone and the beginning of the Fraunhofer zone, where the radiative characteristics of the antenna fully materialize.

Power density can be evaluated, at least in an opera- tional fashion (since the theoretically rigorous definition i.e., ReExH*, does not lead to measurable results), as follows. If one postulates that the square of the electric field strength at the aperture is proportional to the power radiated, then the average value of /El2 over the aperture must be proportional to the average power density Pa over the aperture. Since,

Po = PIA

where P is the total power radiated and A is the area of the aperture in cm2, it follows that the maximum power density

is given by the ratio of the maximum value of IEJ2 to the average value of [El2. From the distribution shown in Fig. 8, this ratio and the power density in the plane of the aperture are, respectively

and

P(0 )dmax = 1.5 P/A.

For a typical 12 mW Falcon radar

Pamax = 1.5 x 12/38 mW/cm2 = 0.45 mW/cm2

At a distance of 5 cm from the plane of the aperture it is difficult to rationalize a formula for power density because of the rapid variations of (El2 , especially in the central portion of the distribution (Fig. 9). Fig. 11 shows the distribution of JEI2 in a plane which includes the antenna axis (corresponding to the distribution shown in Fig. 9). Note the depth of the crater in the center of the distribution. If IEI2 is averaged over the central portion of the aperture, e.g., a circle with T < 2.5 cm, which includes the highest measured value of /El2 shown in Fig. 1 1 , this average value will be proportional to the maximum power density at a distance of 5 cm. Averaging around the peak values of [El2 makes much more sense in a possible relation to power than just the peak value of JEI2 at a point around which E is changing very rapidly in both amplitude and phase. With the above relation between ]El2 and power density, averaging IEI2 as indicated yields the following relation for the maximum power density at 5 cm distance from the aperture:

P(5)amax = 0.5 P(O)d,,.

For the typical 12 mW radar examined, the maximum power density at 5 cm is

P ( 5 ) d m a x = 0.23 mW/cm2.

IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 37, NO. 2. MAY 1995

~~

-6 4 -2 0 2 4 6

X -centimeters

Fig. 1 1 . axis.

[El2 distribution ( r = 5 cm) in a vertical plane through the antenna

10 20 25 30 35 00

Z -centimeters

Fig. 12. plantable probe. The ordinate is IEI2.

Vertical scan along the antenna axis using the NardaBRH im-

Generally, the output power of 24-GHz hand-held traffic radars made less than 100 mW. Thus, based on the above measurements and analyses, the muximum power density that might be detected at 5-cm distance should be less than 2.3 mW/cm2, approximately one-half the 4.6 mW/cm2 reported in [l] for the same radar.

The above measurements were repeated using the NardalBRH implantable probe with similar results. Fig. 12 shows the fall-off of /El2 versus distance from the aperture along the axis of the antenna. The features of this distribution are compIeteIy analogous to the results shown in Fig. 7 obtained with the EIT probe.

The complete set of /El2 distributions obtained with the NardaBRH probe agree with those obtained with the EIT probe. Fig. 13 shows that the distribution of (El2 at 5 cm is practically the same as the corresponding distribution shown in Fig. 9. Some details may be different because of the different sensor size, but there is no doubt that the deep crater in the field distribution is a characteristic of the antenna near field, not an artifact of the measurement method or the probe used.

- w

-8 5

Y - centimeters X - centimeters

-8 5

Fig. 13. probe. Note the depression along the axis similar to that shown in Fig. 9.

(E)' distribution (T = 5 cm) using the NardaBRH implantable

\ / Gunn Diode V Recei -- Assembly A ^ ^ ^ _

ver

,A *s X

port1 Port2 . - Directional ' Coupler port

Fig. 14. reflection and absorption measurements.

Block diagram of a typical traffic radar and the setup used for

v. POWER REFLECTION AND ABSORPTION MEASUREMENTS

The traffic radar microwave assembly (see Fig. 3) is a three-port RF device, whose reflection and transmission char- acteristics can be determined experimentally by the mews of a directional coupler and power meter. The: numbering of the ports and experimental set up is shown graphically in Fig. 14 and a picture of the test assembly is shown in Fig. 15. A Hewlett Packard (HP) 20 dF3 directional coupler is used in the experiments (HP K752D); the RF power sensor is an HP 8485A; the power meter an HP 437B. The initial set of measurements verified that insertion of the directional coupler between the Gum diode generator and port 1 did not introduce any changes in the distribution of 1EI2 in the plane of the aperture (T = 0.2 cm). The distribution of [El2 (obtained using the assembly shown in Fig. 13) is shown in Fig. 16 and is totally analogous to the corresponding distribution shown in Fig. 8. The NardmRH implantable probe was used for these measurements.

The incident and reflected RF power was measured at port 1 by connecting the output arm of the directional coupler to the power meter. The reflected power at port 1 was measured with the directional coupler oriented as in Fig. 14; the incident power was measured by inserting the directional coupler with its output arm connected to port 1. The power coupled to port 2 was measured by replacing the diode receiver assembly with a matched load connected to the power meter. Fig. 3 shows the microwave assembly as the power coupled to port 2 is being measured while the horn is radiating in free space.

.

BALZANO el a/.: MEASUREMENT OF EQUIVALENT POWER DENSITY AND RF ENERGY DEPOSITION I89

Fig. 15. surements.

Photograph of the setup used for reflection and absorption mea-

TABLE I

Open Circuited Short Circuited Horn Hom

Incident Power at 17.5 mW 15.5 mW Port 1

Reflected Power at Port 1 0.4 mW 2.0 mW Power Coupled to Port 2 0.2 mW 11.6 mW Radiated Power 16.9 mW 1.9 mW

The results of the RF power measurements, which are accurate to 4~0.1 mW, are shown in Table I. In the “open horn” case the horn is radiating in free space; the “short circuited horn” measurements were made while pressing a copper sheet much larger than the aperture against the dielectric radome of the horn. The “short circuited’ horn measurements were made to establish the RF pathways within the microwave assembly, specifically the pathway at the RF energy reflected at the horn aperture.

Ohmic and leakage losses (leakage at the dielectric gap between the rim of the horn and the shorting metal sheet) account for the 1.9 mW loss shown in Table I. Practically all the RF power transmitted through port 1 and reflected from port 3 is found at port 2.

If the aperture of the horn assembly is held near or pressed against various parts of the human body (chest, abdomen, thighs and head) one finds that the reflected power at port 2 is small (0.2-0.4 mW). Maximum power is reflected when the aperture is approximately 2.5-4 cm from a relatively flat surface of the human anatomy. In this case, the power level at port 2 is only 1.5 mW.

The results point clearly to the fact that RF energy at 24- GHz is readily absorbed by the human body. Human skin is a tissue rich in water with a thickness varying approximately from 0.3-2.0 mm over the body. The Debye relaxation fre- quency for water is approximately 16 GHz and the dielectric constant is relatively low (E’ = 3 1) and the dissipation constant

-3 5

Fig. 16. IE12 distribution in the plane of the antenna aperture (T = 0.2 cm). Note that the distribution is identical to the corresponding distribution shown in Fig. 8.

relatively high (E” = 38) [6], i.e., water is a good absorber at 24-GHz. In addition to the absorption characteristics, it is also important to know the penetration depth of the absorbed energy.

VI. SKIN DEPTH MEASUREMENTS .

Classical theory and existing values for E’ and E” for skin would predict that the depth of penetration for a 24-GHz plane electromagnetic-wave incident on a planar layer of skin tissue would be about 1.5 mm. That is, approximately 86 percent of the energy would be deposited in the first 1.5 mm of tissue. Since the fields in the aperture of a traffic radar are highly nonuniform, and a portion of the total energy is‘evanescent (not radiated), it is not intuitively clear that the depth of penetration into tissue-simulating material placed against the antenna aperture would be the same as that expected for a uniform plane wave.

Ideally, the penetration depth would be measured by means of a small E-field probe implanted within the tissue equivalent material. Since the wavelength of an electromagnetic wave in tissue (approximately 2-2.5 mm at 24-GHz) is considerably smaller than in air, even the 1 mm (Narda probe) is too large for such measurements. Instead, measurements in air, of the transmission of energy through various thickness of skin equivalent medium (moist chamois skin, approximately 0.4 mm thick), were related to the penetration depth.

This model is not totally correct, because skin in humans is normally followed by a layer of fatty tissue. If, however, there is substantial attenuation in the skin, reflections at the skidadipose interface will have minimal effect on the result and the procedure followed here will be essentially correct. To this end, a chamois skin 0.3 m x 0.3 m in size was wetted with 20 g of water and stretched over the horn assembly. The chamois skin thickness varied from about 0.3-0.4 mm. The section of the chamois skin that was 0.4 mm thick was used for both single and double layer measurements.

The distribution of (E(’ was measured at a distance of 2 mm from the plane of the antenna aperture with one and then two thicknesses of chamois in place. The resulting distributions

190 IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 37. NO. 2. MAY 1995

Fig. 17. moistened chamois skin stretched across the antenna aperture.

/El2 distribution ( T = 0.2 cm) with a 0.4-mm thick layer of

(measured on a 0.25 cm grid) were then compared with the results found earlier (Fig. 4), Le., in absence of the chamois. Results show that the regular pattern of Fig. 8 is substantially attenuated by the presence of a 0.4 mm thick layer of moist skin and further attenuated by the double layer (0.8 mm). Examples of the attenuated distributions are shown in Figs. 17 and 18. The attenuation is far from uniform because the skin is not a homogeneous isotropic material. Given the irregular shape of the skin-attenuated /El2 distributions at 2 mm from the aperture, the attenuation constant of skin at this frequency was evaluated as the ratio of the total electric energy available in presence of the skin to that and in absence of the skin. The following definition is used:

L I E 2 dA

/)El2 d A ’ Attenuation =

where is the square of the field strength with the skin in place, IE21 is the corresponding value in the absence of the skin, and dA is the elemental area on the aperture of the hom antenna.

The results of these measurements indicated that the at- tenuation in the chamois is 17 dB/mm with an uncertainty of f 2 dB/mm. This corresponds to a penetration depth of approximately 0.5 mm, about one-third of the value expected for a plane wave in skin. Although the above model would be more accurate if the skin layer were followed with a layer of simulated fat (as in the human body), in view of the high gttenuation measured the error expected from not using the addtional adipose layer is small since most of the energy is absorbed in the skin. The difference between the classical prediction and the values measured above are probably asso- ciated with the predominantly nonparallel nature of the fields at the antenna aperture and an inadequate equivalence of skin with the chamois simulant. Reduced depth of penetration has previously been reported in the literature for similar near-field exposure conditions [7].

.07,

Fig. 18. skin stretched across the antenna aperture.

/El2 distribution (T = 0.2 cm) with a 0.8-mm thick layer of chamois

VII. CONCLUSION

The above measurements indicate that: 1) The results of the studies of the equivalent power density in the immediate vicinity of the aperture of traffic radars that have been reported in the literature are reasonably correct. A comparkon of these results with the corresponding results reported in [l], [3] indicate that the latter are on the high side, by about a factor of two times. 2) When placed close to the surface of the body, most of the radiated power is absorbed with very little reflected back into the antenna. For the unit measured (12 mW), the maximum power density incident on the skin when the antenna is in close proximity is less than 0.5 mW/cm2. 3) More than 95 percent of the energy is absorbed in the first millimeter of depth when the antenna is placed in contact with tissue. Since the depth of penetration at infrared wavelengths (IR) is comparable to the values found here, it is noteworthy to compare the above value with the maximum permissible exposure of 100 mW/cm2 for IR lasers [8].

REFERENCES

[ l ] P. D. Fisher, “Microwave exposure levels encountered by police traffic radar operators,” Michigan State University, MSU-ENGR-9 1-007, Aug. 30, 1991.

[ 2 ] R. C. Baird, R. L. Lewis, D. P. Kremer, and S. B. Kilgore, “Fieldtrength measurements of speed measuring radar units,” Nat. Bur. of Standards Rep. No. NSMIR 81-2225, 1981.

[3] P. D. Fisher, “Microwave exposure levels encountered by police traffic radar operators,” IEEE Trans. Elecfromugn. Compat.. vol. 35, no. I , pp. 3 6 4 5 , 1993.

[4] NARDA Microwave Corp., “Operation and maintenance manual for broadband isotropic radiation monitors,” Model 8608/8606/8607, 198, pp. 4/29-4/30.

[SI H. Bassen, “Internal dosimetry and external microwave field measure- ments using miniature electric field probes,” in Proc. Symp. Biolog. Effects and Meusurement of Radio FrequencyhWcrowave.7, Rockville, MD, Feb. 16-18, 1977, pp. 136-151.

[6] R. Pethig, Dielectric und Electric Properties of Biological Materials. New York, NY: John Wiley, 1979, p. 139.

[7] I. Chatterjee, 0. P. Gandhi, M. J . Hagmann, and A. Riazi, “Plane-wave spectrum approach for the calculation of electromagnetic absorption under near-field exposure conditions,” Bioelectromagn.. vol. 1, pp. 363-377, 1980.

[8] ANSI 2136.1-1993. American National Standard,for the Safe Use of Lasers. New York, NY: American National Standards Institute, 1986.

BALZANO et 01.: MEASUREMENT OF EQUIVALENT POWER DENSITY AND RF ENERGY DEPOSITION 191

Quirino Balzano (S963-M’72-SM’83) was born in Rome, Italy in December of 1940. In 1965, he received a Doctorate of Engineenng degree in electronics from the University of Rome. Italy

Dunng 1966 he was at FIAT, SPA, in Tunn, Italy From 1967-1974, he was employed by Raytheon Co , in the Missile Systems Division. He was in- volved in the research and development of planar and conformal phased arrays. Since 1974, he has been with the Plantation, Florida operations of Mo- torola Inc., where he is the Corporate Vice President

and Director of the Flonda Electromagnetics Research Laboratory His main interest is in the biological effects of the human exposure to RF electromagnetic energy

Dr Balzano has wntten over 50 papers on RF dosimetry near electro- magnetic sources and the biological effects of RF energy Dr Balzano is a charter member and past director of the Bioelectromagnetics Society, a scientific society wholly dedicated in the research of the biological effects of electromagnetic fields He received the IEEE Vehicular Technology Society Paper Pnze Award in 1978 and 1982 and a certificate of merit from the Radiological Society of North America in 181 for the treatment of tumors with RF energy

John A. Bergeron received the B.S. degree in biology from Brown University i n 1951 and the doctorate degree from Comell University in 1955.

After nine years in the Biology Department at Brookhaven National Laboratory, he joined a solar energy conservation group in the General Physics Department of GE’s Research Laboratory, 1964, now GE Corporate Research and Development, During his tenure, he has had varied staff and managerial positions, currently he is a Physiologist in the Environmental Laboratory. His ongoing research concerns human endogenous ELF fields and the intensity- time (dBldt-time) relationship for exciting human magnetophosphenes. Over the years his research has ranged over the entire nonionizing electromagnetic spectrum; visible light (spectral effects of pigment interactions, photosynthe- sis), microwaves (energy disposition in human phantoms), RF (dosimetry in the near field of an antenna, 450 MHz; magnetic resonance imaging, 47 MHz, 1.5 T; and ELF (25 kHz magnetic fields and human immune cells: electric shock, dc to 200 kHz; Calcium efflux from chick brain).

Dr. Bergeron belongs to numerous scientific societies, including the Bio- electromagnetic Society and the American Society for Photobiology, has numerous publications and several patents and participated in the development of the national standard for RF-microwave exposures (ANSI C95.1- 1982) and the recent revision (ANSVIEEE C95.1-1992). He is co-chair of the subcommittee (SC3) given the charge by IEEE SCC28 of developing ELF (0-3 kHz) exposure guidelines.

Jules Cohen (M’39-SM’55-LS’83-LF90) received the B.S. degree in elec- trical engineering from the University of Washington in 1938. His early engineering employment was as an engineer in the Substation Design Section at the Bonneville Power Administration. As a commissioned officer in the US. Navy during WW 11, he started his career in electronics and has continued in that field. From 1952-1988, he was managing partner, sole owner or president of the consulting engineering firm bearing his name. Although retired from firm management, he continues to practice engineering consulting at the firm Jules Cohen & Associates, P.C., Consulting Engineers. He is a Life Fellow of the SMITE, a member of the EMC Society, a member of Tau Beta Pi, and a charter member of the Bioelectromagnetics Society.

John M. Osepchuk (A’51-M’5&SM’71-F’78- LF93) received the A.B. degree and ;Ph.D. degrees from Harvard University in 1948 and 1957, respectively. He has worked on high-power microwave devices and applications including microwave ovens. Since 1968, he has concentrated on radiation hazards and safety and through various committees (e.g., IEEE-COMAR) has worked toward restoring a rational public view of microwaveiRF technology. He is a Consulting Scientist with the Raytheon Research Division.

Dr. Osepchuk is a Fellow of IMP1 and a member of Phi Beta Kappa, Sigma Xi, and BEMS. He has served as Division IV appointee to the ADCOM of the IEEE Society for Social Implications of Technology (SSIT) and has recently played a key role in the forming and operation of the Electromagnetic Energy Policy Alliance (EEPA). He presently is also past Chairman of COMAR and Executive Secretary of the IEEE Standards Coordinating Committee 28 which is responsible for the C95 standards on safety from radiofrequency radiation hazards.

Ronald C. Petersen (S’59-A’61-M’89) received the B.S.E.E. (1968) and M.S.E.P (electro- physics-1970) degrees from the Polytechnic Institute of Brooklyn.

He served four years as an Aviation Electronics Technician in the U S . Manne Corps. He has been with AT&T Bell Laboratones since 1960, the last 24 years with the Radiation Protection Department where his responsibilities are nonionizing radiation protection for AT&T. In this capacity, he serves as the Corporate resource for all nonionizing radiation matters.

Mr. Petersen is a member of the Institute of Electncal and Electronics Engineers (IEEE) Standards Coordinating Committee SCC-28 (Radiofre- quency Radiation Hazards) and the IEEE Committee on Man and Radiation (COMAR), and a member of the National Council on Radiation Protection and Measurements (NCRP). He is also a member of the Board of Directors and Chairman of the Science and Technology Committee of the Electromagnetic Energy Association (EEA). He is a Visiting Lecturer at Rutges University, Department of Radiation Science and has published several papers and chapters of books in related subjects

Leon M. Roszyk received the B.S.E.E. degree in electrical engineering from the University of Illi- nois and the MBA degree from Illinois Benedictine College. He is a consulting engineer specializing in electrical product design. He was employed by Sunbeam Corporation for 36 years. His experience covers a wide variety of consumer as well as indus- trial and commercial products and markets. Prior to retirement, Mr. Roszyk served as vice-president of research and engineering at Northern Electric Co. (now an operating div. of Sunbeam-Oster House-

hold Products) and was directly responsible for research, development and quality control of all consumer products. He directed the engineering efforts which led to the successful introduction of Sunbeam’s Low Magnetic Field electric blanket heating system in 1989.

Mr. Roszyk is a licensed professional engineer in the states of Illinois, Pennsylvania, Mississippi, and Florida and a member of the Science and Tech- nology Committee of the Electromagnetic Energy Association, headquartered in Washington, DC.