PROM 1 anti-personnel landmines Possibility of activation by physical contact with a metal detector

Transcription

1 EUROPEAN COMMISSION DIRECTORATE GENERAL JRC JOINT RESEARCH CENTRE Technical note N S.P.I PROM 1 anti-personnel landmines Possibility of activation by physical contact with a metal detector Authors: Fernandez Manuel Lewis Adam Littmann François Ispra March 2001

2 Page 2 / 40 JRC David Wilkinson Alois Sieber, John Dean Fernand Sorel, Fivos Andritsos Martyn Dowell, Iain Shepherd Distribution List ISIS ISIS/TDP ISIS/RIT ISIS/MIA External to the JRC Daniela Dicorrado-Andreoni Wolfgang Boch, Pascal Collotte, Jean Jacques Lauture Peter Billing Hans Schiebel Ollie Allerhead Gerhard Vallon Martin Foerster Theodor Steinbuchel Jernej Cimpersek, Eva Veble Pat & Eddie Banks Filip Filipovic Damir Gorseta John Flanagan Arben Braha Janko Petrovic Robert Doheny, Jan Koster, Curt Larsson, David Lewis, Robert Suart, Steve Vermeer Kaj Horberg George Zahaczewsky Harold E. Bertrand Denis Reidy Christine Lee, Karin Breiter Jacques Roosenboom Arnold Schoolderman Pieter Jan de Veer Yogadhish Das, Jack D. Toews Richard Beech, David Lewis Alastair McAslan, Paddy Blagden Hemi Morete Christian Desmoulins, Denis Duret Claudio Bruschini Pasquale Nardone DG RELEX DG INFSO ECHO Schiebel Guartel Vallon Foerster Ebinger ITF BiH Commission on Demining BHMAC CROMAC UN MACC, Kosovo AMAE, Albania CCPDR ITEP ExCom ITEP Secretariat U.S DoD IDA JUXOCO NVESD MOD Netherlands TNO Test Department Hemburg DRES DERA Chertsey GICHD UNMAS CEA EPFL ULB

3 Page 3 / 40 Head of Unit Name: Alois Sieber Date: 19 March 2001 Institute Director Name: David Wilkinson Date: 23 March 2001 Signature: Original signed by Alois Sieber Signature: Original signed by David Wilkinson Disclaimer This report is an investigation of a specific problem that may occur with metal mine detectors and one class of mine. Its purpose is to help to define safe operating procedures. It is not intended as a general statement of the relative merits of different types or models of detectors. Legal Notice The information in this document may not be disseminated, copied or utilized without the written authorization of the Commission. The Commission reserves specifically its rights to apply for patents or to obtain other protection for the matters open to intellectual or industrial protection. The distribution of this document is limited to the persons given in the distribution list. Neither the Commission of the European Communities nor any person acting on behalf of the Commission is responsible for the use that might be made of the following information.

4 Page 4 / 40 Abstract PROM 1 anti-personnel landmines Possibility of activation by physical contact with a metal detector M. Fernandez, A. M. Lewis & F. Littmann European Commission Joint Research Centre TP 272 Via E. Fermi 1, Ispra (VA) Italy Abstract Measurements of the horizontal detection distance of a VPROM1 inert mine have been made with five different models of commercial metal detectors, two having differential coils and three having non-differential coils, in air and in highly magnetic soil, with the heads horizontal and tilted. The VPROM1 is the training version of the PROM1 antipersonnel bounding fragmentation mine. The measurements were made in response to a request by a demining agency, which suspected that some fatal accidents involving activation of PROM1 s had been due to physical contact of a metal detector with the protruding pronged fuze. It was thought that the presence of regions of reduced sensitivity straight in front and behind the head of a differential coil detector had prevented the detector giving its alarm sound early enough to provide adequate warning. In the work reported here, horizontal detection distances at prong height of the order of 10cm were recorded. Significant reductions in detection distances were found in sectors 30º wide. No similar effect was found for the non-differential coil detectors. The results are therefore consistent with the above hypothesis. The waveform of the detector current appeared to play no part. Highly magnetic soil had only a small effect. All the detectors, including those with differential coils, were sufficiently sensitive to make it possible to detect the PROM 1 in a preliminary sweep, made well above the prong height.

5 Page 5 / 40 Contents 1 Introduction Experimental Measurements The aim The metal detectors The target The method The tests In-air & In-soil Main Measurements Ebex 420 PB Foerster Minex 2FD Guartel MD Schiebel AN Vallon ML 1620C Tilt measurements on the differential detectors The aim The method The results Verification of minimum sweep height The aim The method The results Recommendation Additional Observations Conclusions Annex References Acronyms Measurement data... 35

6 Page 6 / 40 1 Introduction The PROM1 is a bounding fragmentation anti-personnel mine (fig. 1.1) [1] with a very high metal content (approximately 2500g) and a very powerful charge. It may be triggered by tripwires (fig. 1.2) and/or by contact with the four-pronged fuze. The activation mechanism is described in [2]. After ignition, the body of the mine, including the main charge, is launched into the air. At around 70cm the anchor cable (or tether wire) joining the bottom of the mine, which still remains in the ground, and the base of the main part of the mine body will be pulled taut. The resulting jar pulls a percussion cap onto a fixed firing pin, initiating the detonator and then the main charge. The mine is then shattered in all directions with an effective lethal range of more than 20m. The PROM1 is normally olive green, so that visual detection in grass is difficult. The total height of the mine is 26cm. In the standard Yugoslav National Army mode, it is buried with the prongs 11cm above ground level. A variant called the PROM PK was manufactured from 1977, the most important modification being the reduction of the tether wire length to 20cm. After several severe causalities caused by PROM1 mines on deminers, the European Commission Joint Research Centre was approached by Mr Eddie Banks, Technical Advisor to the Bosnia and Herzegovina Commission for Demining [3]. In an accident investigation report for the US State Department, Mrs. Pat Banks had proposed a mechanism, which might have caused such accidents. She suggested that there was a region of low sensitivity ("weak spot") at the front of the halo of some metal detectors, where the detection range was significantly reduced, even for high metal-content mines. If such a detector were swept in a forward motion, a signal would not be audible until its head was extremely close to the prongs of a PROM1. The operator would be unable to assess the danger before contact with the mine fuze trigger. Mr and Mrs. Banks asked the JRC to see if this mechanism could be confirmed or refuted scientifically. A multi-national technical evaluation of performance of commercial off the shelf metal detectors in the context of humanitarian demining has recently been completed the International Pilot Project for Technology Co-operation Consumer Report (IPPTC) [4]. However, all targets used in IPPTC had a very low metal content compared to the PROM1, so that the results are not directly pertinent to this case and additional work was needed.

7 Page 7 / 40 Figure 1.1: PROM1 Figure 1.2: Fuze and tripwire

8 Page 8 / 40 2 Experimental Measurements 2.1 The aim To assess the level of risk that the mine might be triggered in the search process, the key parameter to know is how close the head can get to the mine, before the operator receives an alert from the detector. This technical note documents the experiments that were performed at the JRC to measure this distance, for various detectors under different conditions. 2.2 The metal detectors Three metal detectors were loaned by Mr Banks [5] which had been used by a commercial demining contractor in Bosnia and Herzegovina (BiH). In order to have a larger spectrum of sensors, the JRC also used three of its own detectors. The following table provides the overview of all metal detectors examined (fig ). Detector BiH JRC Serial Number Ebinger Ebex 420 PB X 2494 Foerster MINEX 2FD X Guartel MD8 X X & Schiebel AN 19/2 X Vallon ML 1620C X 104 Figure 2.2.1: The metal detectors in front of the Gauss Laboratory

9 Page 9 / The target A VPROM1 (fig ) was used to realize the tests. The VPROM1 is an inert training version of the PROM1. Externally, it is a geometrical replica of the PROM1, has a similar mass of metal and is free from explosive. The internal structure (fig ) differs somewhat from that of the working mine. This will have no effect on the measurements because the electromagnetic penetration depth in steel for metal detectors is only a fraction of a millimetre, much less than the wall thickness. Essentially, the detectors see the external surface of the mine. The VPROM1 with its model fuze weighs 2 360g. Figure 2.3.1: VPROM1 Figure 2.3.2: VPROM1 and inner simulant part

10 Page 10 / The method The experiment was conducted at the Carl-Friedrich Gauss Laboratory of the Unit for Technologies for Detection and Positioning at the Joint Research Centre in Ispra (Italy). The laboratory is a non-metallic building designed specifically for tests on mine detectors sensitive to metal [6]. The measurement method is described below and was generally the same for all metal detectors. Specific details are given for each detector. Step 1: Set up A VPROM1 was placed upright on the ground, far from any metallic object. Above the mine, a square plastic board (fig ), transparent to metal detectors, was placed horizontally at a known height. A millimetric graph paper sheet was taped on the board, with the centre of the sheet located exactly above the prongs of the VPROM1. Starting from the centre, radial lines were drawn every 15. The altitude referred to below for each test is taken between the upper part of the plastic board and the upper part of the prongs (fig ) of the mine. Step 2: Calibration The detectors were first calibrated, in order to allow the sensitivity setting always to be restored to the same level, even if the detector drifted, was switched off or was adjusted in error. The sensitivity of the detector was set to the almost highest possible level, the detector being far away from any metal object. The sensitivity was then reduced until no more sound was audible. The detector was brought near to the calibration object: a non-magnetic, stainless steel sphere of diameter 19mm (fig ). The sphere is mounted in the cream-coloured silicone-rubber block. The maximum height at which this sphere could be detected with each detector was then recorded (fig ). Note: For the differential detectors, for which no continuous adjustment was available, the calibration sphere detection height varied by around ± 10% from day to day. Step 3: The horizontal detection distances For each selected altitude ( 2cm, 8cm and +18cm), the entire board was scanned in the following manner. Starting from the edge of the paper, the search head was slowly moved forward towards the mine, so that the centre of the coil followed one of the 15º radial lines (fig ). The detector was stopped at the point where its alarm first sounded and its position (intersection of the radial line and the edge of the coil closest to the mine) was marked. The plotted lines therefore show the separation between coil and mine not the locus of the centre of the coil. The process was repeated for each radial line, all around the mine (fig ). The orientation of the detector with respect to the paper was kept the same for all lines; that is to say, the handle always pointed in the same direction. The altitude was then changed and the operation repeated. Note: Between the point where the first alarm sounded and a clear continuous sound, it was frequently necessary to move the head a few centimetres further towards the mine target. The plots correspond here to the first initial sound even if hesitant.

11 Page 11 / 40 Figure 2.4.1: The VPROM1 below the plastic board Figure 2.4.2: Altitude above the prongs Figure 2.4.3: Reference steel sphere in silicone-rubber block Figure 2.4.4: Calibration

12 Page 12 / 40 Method of measurement of the horizontal detection distance with the orientation of the handle constant (0 ) Figure 2.4.5: Measuring method (the centre of the search head tracking the radial lines in the direction of the target) Figure 2.4.6: Example with the Guartel MD8

13 Page 13 / The tests Two main series of measurements were made: First in air inside the Gauss laboratory at various altitudes above the prongs (-2, 8 & 18cm). Then outside, with the VPROM1 target partially buried in a highly magnetic soil from Naples (fig & 2.5.2), to a depth of 17.5cm (the fuze trigger protruded 8.5cm above the soil). The in-soil measurements were made only at one altitude of 18cm. Additional measurements were made to check the effect of tilting the search head of the differential detectors, at the altitude of 8cm: Roll (10, 20 ). Pitch (-10, 10 ). Finally, measurements of the maximum vertical height detection in the highly magnetic soil were made to confirm that it was possible to detect the mine safely above the prongs. Precision: The error in the measurements of the detection distances is estimated at ± 20mm, including the error in the placement of the mine, the error due to the subjective nature of the limit (sound) and reading of the calibration sphere detection height. The error in maintaining the constant yaw angle of the search head is estimated at ± 5. For the differential detectors for which no continuous adjustment was available, the day to day repeatability of the horizontal detection distance is around ± 10%, proportional to the day to day calibration drift. Figure 2.5.1: VPROM1 buried in the highly magnetic soil Figure 2.5.2: In-soil measurement

14 Page 14 / 40 3 In-air & In-soil Main Measurements 3.1 Ebex 420 PB Figure 3.1.1: Ebex 420 PB Figure 3.1.2: Search head Figure 3.1.3: Control box Preliminary remark: The Ebex 420 PB loaned from Bosnia-Herzegovina did not always maintain constant sensitivity and it was necessary to recheck the calibration frequently. Detector principle: The Ebex 420 PB is a pulsed induction single receiving coil detector, without auto-zero [7]. The shape of the search head is oval, length 26cm & width 15cm. Calibration sphere detection height: 19cm Sensitivity control: Continuous adjustment, without any pre-selectable positions marked. Special details of method: The heads of all the detectors tested are designed to partially rotate about a horizontal axis (pitch) so that the handles can be held at a convenient slope with the head horizontal. Since the Ebex 420 has a metal shaft, the calibration is affected by changing the head angle in this way. Therefore, care was taken not to alter the head angle during the test. All the other detectors used in this study have non-metallic shafts attached to the head. Results: The horizontal detection distance graphs are all almost circular (fig ). At the altitude of -2cm, the horizontal detection distance is around 35cm. The detection range measured in highly magnetic soil was greater (fig.3.1.5), but not significantly so. Summary: The Ebex 420 PB tested detects a VPROM1 with a sufficient safety margin at all angles.

15 Page 15 / 40 Ebex 420 PB Horizontal detection distance in cm for different altitudes cm 8cm 18cm Figure 3.1.4: Horizontal detection distance with sketch showing orientation and scale of head in-air in-soil Ebex 420 PB Horizontal detection distance in cm In-air & In-soil Altitude 18 cm Figure 3.1.5: Comparison between in-air and in-soil measurement

16 Page 16 / Foerster Minex 2FD Figure 3.2.1: Foerster Minex 2FD Figure 3.2.2: Search head Figure 3.2.3: Control box Detector principle: The Foerster Minex 2FD is a continuous induction, differential receiving coils detector, with auto-zero [8]. The shape of the search head is elliptic, length 29cm & width 21cm. Note: The differential arrangement of the search head coil is designed to help in an accurate pinpointing of a target, by producing a null when the middle axis of the search head is in front, above or behind the target. Calibration sphere detection height: 19 to 23cm Sensitivity controls: Three selectable positions (maximum sensitivity, position H, was used). Special details of method: This detector displayed drift of the zero: i.e. it continued to beep after withdrawal from the mine, so it was necessary to manually zero between two sweeps. The signal begins with a very low volume, so the detection decision threshold was more subjective than for the other detectors. Results: The horizontal detection distance graphs (fig ) showed a loss of sensitivity in the front (and behind) the search head. Between -30 and +30, there is a noticeable reduction of the detection range at all heights. The maximum loss of sensitivity is around 5 and +5 taking into account the precision on the constant yaw angle of the search head. For the altitude of 18cm, no signal at all was heard when scanning in a forward direction (0 ). For the altitude of -2cm, the horizontal detection distance at 0 and 180 is around 14cm. The detection range measured in highly magnetic soil is almost unchanged (fig ). Summary: The Foerster Minex 2FD tested detected the VPROM1 but showed regions of reduced sensitivity in front of and behind the search head.

17 Page 17 / 40 Foerster Minex 2FD Horizontal detection distance in cm for different altitudes cm cm 40 18cm Figure 3.2.4: Horizontal detection distance with sketch showing orientation and scale of head Foerster Minex 2FD Horizontal detection distance in cm In-air & In-soil Altitude 18 cm in-air in-soil Figure 3.2.5: Comparison between in-air and in-soil measurement

18 Page 18 / Guartel MD8 Figure 3.3.1: Guartel MD8 Figure 3.3.2: Search head with a white line on the sighting bar Figure 3.3.3: Control functions Preliminary remark: The MD8 loaned from Bosnia-Herzegovina (s/n ) was not used because it tended to sound continuously, even in the total absence of metal, and also had an excessive tendency to auto-zero. Instead, the JRC s MD8 (s/n ) was used in the test. Note: The differential arrangement of the search head coil is designed to help in an accurate pinpointing of a target, by producing a null when the middle axis of the search head is in front, above or behind the target. Detector principle: The Guartel MD8 is a pulsed induction, differential receiving coils detector, with auto-zero [9]. The shape of the search head is circular, diameter 32cm. Calibration sphere detection height: 19 to 22cm Sensitivity control: Three selectable positions (maximum sensitivity, position III, was used). Special details of method: The sound from the MD8 is not sustained when it is held immobile above a metal piece, but vanishes as the detector automatically re-zeroes. The head was therefore swept fast enough to avoid the detector zeroing-out while it approached the target. The detection limit was marked directly after the beginning of the beep. Since it was more difficult to measure the detection limit when the head was moving, the measurements were repeated several times for accuracy.

19 Page 19 / 40 Results: The horizontal detection distance graphs (fig ) showed a loss of sensitivity in the front (and behind) the search head. Between -30 and +30, there is a noticeable reduction of the detection range at all heights. The maximum loss of sensitivity is around 5 and +5 taking into account the precision on the constant yaw angle of the search head. For the altitude of 18cm, no signal at all was heard when scanning in a forward direction (0 ). For the altitude of -2cm, the horizontal detection distance at 0 is 8.5cm and 5.6cm at 180. The detection range measured in highly magnetic soil is almost unchanged (fig ). Remark: In the detector manual [9], the manufacturers note the presence of the region of reduced sensitivity and draw attention to the possibility of making use of it as a feature: The halo search head is of differential detection design. That is, it tends to reject targets outside the halo loop. The maximum rejection of outside targets is in line with the white line on the sighting bar and minimum at right angles. In this way a small target may be detected under the search halo with a large target close to (150 mm) the halo in line with the sighting bar. Summary: The Guartel MD8 tested detected the VPROM1 but showed regions of reduced sensitivity in front of and behind the search head.

20 Page 20 / Guartel MD8 Horizontal detection distance in cm for different altitudes cm 8cm 18cm Figure 3.3.4: Horizontal detection distance with sketch showing orientation and scale of head in-air in-soil Guartel MD8 Horizontal detection distance in cm In-air & In-soil Altitude 18 cm Figure 3.3.5: Comparison between in-air and in-soil measurement

21 Page 21 / Schiebel AN-19 Figure 3.4.1: Schiebel AN- 19/2 Figure 3.4.2: Search head Figure 3.4.3: Control box Detector principle: The Schiebel AN-19/2 is a pulsed induction single receiving coil detector, without auto-zero [10]. The shape of the search head is circular, diameter 26cm. Calibration sphere detection height: 23cm Sensitivity control: Continuous adjustment, no pre-selectable positions marked. Special details of method: None. Results: The horizontal detection distance graphs are all almost circular (fig ). At the altitude of -2cm, the horizontal detection distance is more than 35cm. The detection range measured in highly magnetic soil was smaller (fig ), but not significantly so. Summary: The Schiebel AN 19/2 tested detects a VPROM1 with a sufficient safety margin at all angles.

22 Page 22 / Schiebel AN 19/2 Horizontal detection distance in cm for different altitudes cm 8cm 18cm Figure 3.4.4: Horizontal detection distance with sketch showing orientation and scale of head in-air in-soil Schiebel AN 19/2 Horizontal detection distance in cm In-air & In-soil Altitude 18 cm Figure 3.4.5: Comparison between in-air and in-soil measurement

23 Page 23 / Vallon ML 1620C Figure 3.5.1: Vallon 1620C Figure 3.5.2: Search head Figure 3.5.3: Control box Detector principle: The Vallon ML 1620C is a pulsed induction single receiving coil detector, with auto-zero [11]. The shape of the search head is elliptic but slightly flattened, length 30cm & width 17cm. Calibration sphere detection height: 33cm Sensitivity controls: Continuous adjustment with graduated scale set to 6 out of 7. Two current levels: set to mode P (upper level). Internal switches: set to 1 (normal soil) and 5 (50Hz filter). Special details of method: The above settings were selected because they give maximum sensitivity, in accordance with the protocol. The in-soil test was also repeated at the mineralised soil setting (2). Note that in practice it would not be known in advance that a PROM 1 is present, so the current would not necessarily have been reduced to M mode. With the above settings, the detector is very sensitive, as indicated by the calibration height. The signal vanishes when the detector is immobile over the mine due to auto zeroing. The Vallon ML 1620C does not give a signal when it is removed from the mine after it has auto-zeroed. Results: The horizontal detection distance graphs are all almost circular (fig ). At the altitude of -2cm, the horizontal detection distance is more than 45cm. The detection range measured in highly magnetic soil (position 1 & 2) was smaller (fig ), but not significantly so. Summary: The Vallon ML 1620C tested detects a VPROM1 with a sufficient safety margin at all angles.

24 Page 24 / 40 Vallon 1620C Horizontal detection distance in cm for different altitudes cm cm cm Figure 3.5.4: Horizontal detection distance with sketch showing orientation and scale of head in-air in-soil in-soil Vallon 1620C Horizontal detection distance in cm In-air & In-soil (position 1 & 2) Altitude 18 cm Figure 3.5.5: Comparison between in-air and in-soil measurement (with (2) and without (1) magnetic switch on)

25 Page 25 / 40 4 Tilt measurements on the differential detectors 4.1 The aim In a field situation, the deminer would not necessarily have the detector exactly in a horizontal plane with respect to the mine body, especially since the angle at which the mine is buried may vary. Therefore, tests were conducted on the differential coil detectors to determine the effect of tilting of the head (roll & pitch) on the sensitivity patterns, to find out if the regions of reduced sensitivity were enlarged, reduced or moved substantially. Note: it was unnecessary to twist about the third axis, the yaw axis, since a rotation about this axis is implicit in all the plots already available. 4.2 The method A small plastic block was taped to the underside of the head of each detector, so that the head rested at an angle when placed on a flat surface. The angle was calculated from the position of the block and its height to an estimated precision of ± 10%. All these tests were conducted in air, at an altitude of 8cm above the prongs, using the same method to measure the horizontal detection distance. The following angles were used: Axis Angle roll +10 roll +20 pitch +10 pitch -10 Pitch Roll The results Tilting the head in the roll axis (fig & 4.3.2) by 10 and 20 tended to move position of the reduced sensitivity regions in the direction of the tilt but the minimum horizontal detection distances remain almost at the same level. Tilting the head by 10 in the pitch axis (fig & 4.3.4) had no clearly discernible effect, within the precision of the measurement, for either detector. The argument that there is a risk associated with the reduced sensitivity regions is therefore not affected by the possibility that the head or mine might be tilted.

26 Page 26 / 40 Guartel MD8 Horizontal detection distance in cm for different roll angles (right side up) Altitude 8 cm Figure 4.3.1: Tilt test for the Guartel MD8 roll axis Foerster Minex 2FD Horizontal detection distance in cm for different roll angles (right side up) Altitude 8 cm Figure 4.3.2: Tilt test for the Foerster Minex 2FD roll axis

27 Page 27 / 40 Guartel MD8 Horizontal detection distance in cm for different pitch angles (+ front up, - front up)) Altitude 8 cm Fig 4.3.3: Tilt test for Guartel MD8 pitch axis Foerster Minex 2FD Horizontal detection distance in cm for different pitch angles (+ front up, - front up)) Altitude 8 cm Fig : Tilt test for Foerster Minex 2FD pitch axis

28 Page 28 / 40 5 Verification of minimum sweep height 5.1 The aim The existence of a risk associated with the low sensitivity regions of the differential coil detectors is confirmed by the experiments above. It is therefore important to consider how a safe operating procedure might be defined, where mines of this pattern are known or suspected to be present. Since the PROM 1 contains a large amount of metal, it is detectable from relatively large heights. Therefore, it is possible in principle to search for PROM 1 mines in a preliminary sweep, conducted at a height great enough to avoid the risk of touching the prongs. Having made the preliminary sweep, the deminers could then go on to sweep close to the ground to detect minimum metal mines. A test was conducted to verify that detection at a safe height in this manner was possible for all detectors (fig ). Figure 5.1.1: Sweep above the VPROM1 prongs in highly magnetic soil 5.2 The method The Guartel MD8 and Foerster Minex 2FD have both three positions of sensitivity, giving approximately the same calibration heights for the two detectors. In order to get equivalent settings, the other detectors were adjusted to similar sensitivity, using the calibration sphere detection height as described above. The VPROM1 was buried in the magnetic soil with the prongs protruding 11 cm above the surface as in the Yugoslav Army Mode. The detection heights were measured to a precision of about ±20mm, which is sufficient for the purpose intended.

29 Page 29 / 40 The Vallon ML 1620C was set for the lower current level (setting M) and for lightly mineralised soil (internal switch position 2), consistent with what might reasonably be used in conducting a general preliminary sweep. It was not necessary to use the ground learning button on the Foerster Minex 2FD Note: This test is not intended to compare generally the performance of one detector against another, for which purpose a higher level of precision and a larger number of targets and conditions are required. Again, we refer the interested reader to the IPPTC report [4]. 5.3 The results All the detectors detected the mine at least at (fig ): 40cm above the soil (29cm above the prongs) for the highest sensitivity setting. 30cm above the soil (19cm above the prongs) for the medium sensitivity setting. 20cm above the soil (9cm above the prongs) for the minimum sensitivity setting. Height above the soil (the prongs are protuding 11 cm above the soil) cm Schiebel Guartel Vallon Foerster Ebinger 0 High Medium Low Figure 5.3.1: Detection heights for VPROM1 in highly magnetic soil 5.4 Recommendation It would be therefore feasible to make a preliminary sweep at a height of about 20cm above the soil on a medium sensitivity setting. In this way, any PROM1 present would be safely detected but without picking up the small metal object. Then another sweep, at a lower height, will be needed to detect mines with lower metal content.

30 Page 30 / 40 6 Additional Observations In addition to the question of the sensitivity pattern, which was the main object of this investigation, we wish also to mention here some other possible aggravating factors: Two of the detectors brought from BiH and loaned to us (MD8 and Ebex 420 PB) had poor electronic performance. In particular, the field-used MD8 performed much worse than the JRC s example, which has been used mainly in laboratories and, on average, for only a few hours a month. We do not know if the examples loaned to us are representative of detectors currently in field use, or are unusually bad. The results do confirm the need to test on a regular basis the performance of detectors used in the field in order to avoid possible hazards arising from circuits or coils deteriorating after prolonged use in harsh conditions. The design and placement of the detector controls should not be overlooked. It might be possible for a detector to be manually re-zeroed accidentally for example. If the manual zero-reset is positioned under the user s thumb, where it is very easy to push by accident, a possible hazard arises (fig. 6.1), especially if the detector gives no indication that it has been re-zeroed. Automatic re-zeroing could also be hazardous, but only if the detector was moved extremely slowly, so that the target was zeroed-out, rather than detected. Another very significant hazard arises with mines of this type as a result of standard search protocols. When the deminer finds an indication of metal, it is common practice to attempt to pinpoint the source by making passes with the detector in two perpendicular directions. A deminer following such a procedure would be likely to touch the prongs of a PROM1 even if the detector had given the indication at a safe distance from the mine, especially if it was concealed in vegetation. Figure 6.1: Manual zeroing button of MD8

31 Page 31 / 40 7 Conclusions Differential receive coil metal detectors: The measurements show that there are regions of reduced sensitivity in front and behind the search head of the two differential coils metal detectors. The sensitivity pattern is clearly described by the manufacturers in the relevant operator manuals and, in fact, highlighted as a potentially useful feature. In the work reported here, horizontal detection distances at prong height of the order of 10cm were recorded. Significant reductions in detection distances were found in sectors 30º wide. The detectors use different current forms (pulsed and continuous wave) and have search head with different shapes (circular and elliptic) but show similar loss of sensitivity. Tilting of the search head of the differential coil detectors had only a small effect on the sensitivity patterns. The regions of reduced sensitivity were moved but not significantly so. Highly magnetic soil had only a minor effect on the shape and size of the sensitivity graph. Single receive coil metal detectors: All the single receive coil metal detectors tested, detected the VPROM1 with a sufficient safety margin at all angles, independently of the current forms and the shapes of the search head. Highly magnetic soil had only a minor effect on the shape and size of the sensitivity graph. Observations: The tests reported here were deliberately conducted with a sweep pattern different from that recommended by the manufacturers, but which might conceivably occur in practice. The state of maintenance of detectors after rigorous field use should not be overlooked as a contributing factor in accidents. Two of the detectors loaned from BiH showed poor electronic stability. There may be some risk of missing a mine due to accidental manual or automatic re-zeroing, depending on the design of the detector and the manner in which it is held and swept. The importance of matching Operating Procedures and training to the particular detector in use is especially apparent from this observation. Despite the large metal content of the VPROM1, the weak spots in front and behind the search heads of the two differential coil metal detectors tested were large enough that it would be possible to get very close to the prongs before the detector sounds. If a detector of this pattern were moved forward or backward towards a PROM1 at any realistic sweeping speed and the search height matched the height of the mine prongs, a deminer would have only a fraction of a second to react before the mine was activated. We consider that these results should be reflected in future operating and training demining procedures.

32 Page 32 / 40 Recommendations: Differential receive coil detectors, with a coil format similar to those tested here, must be swept laterally and not in a forward / reverse motion. Personnel should be trained to follow the operating procedure recommended by the manufacturer for the specific model of detector in use. When the presence of PROM1 s is considered possible, the deminer should make a preliminary sweep at a height of about 20cm above the soil on a medium sensitivity setting. In this way, any PROM1 present would be safely detected, without false alarms from small metal objects. Another sweep, at a lower height, should then be conducted to detect mines with lower metal content.

33 Page 33 / 40 8 Annex 8.1 References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] Jane s Mines and Mine Clearance Colin King (ed.) Third edition Jane s Information Group, Coulsdon, Surrey, UK Brassey s Essential Guide to Antipersonnel Landmines Eddie Banks Brassey s (London) 1997 Mrs. Pat and Mr Eddie Banks Private correspondence with JRC July11, 2000 International Pilot Project for Technology Co-operation Consumer Report A multi-national technical evaluation of performance of commercial off the shelf metal detectors in the context of humanitarian demining Dean. J. T., Zahaczewsky G., Das Y., Lewis D., Roosenboom J. (Editors), Joint Research Centre, Ispra, Italy, 2001 Loan of Metal Detectors from BiH Mr Eddie Banks Private Correspondence with JRC October 5, (Test and Evaluation Facilities) Operators Manual for the Metal Detector EBEX 420 PB Ebinger Prüf- und Ortungstechnik GmbH, Köln, Germany (undated) Minex 2FD Compact Metal Detector Operating Instructions Institute Dr Foerster, Reutlingen, Germany, 2000 MD8 Mine Detector Handbook June 1999 Guartel Ltd., London, UK, 1999 AN 19/2 Mine Detecting Set Operating Manual Schiebel, Vienna, Austria (undated) Metal Mine Detector ML 1620 C Operation Manual Issue 1/98 Vallon GmbH, Eningen, Germany, 1998

34 Page 34 / Acronyms JRC IPPTC PROM1 PROM PK VPROM1 Joint Research Centre of the European Commission International Pilot Project for Technology Co-operation Consumer Report Anti-personnel fragmentation bounding mine from the former Yugoslavia Modified PROM 1 with reduced length tether wire Training version of the PROM1 mine, has no explosive, almost the same metal content and a broad yellow band around the top of the mine body

35 Page 35 / Measurement data Ebinger In-air & In-soil in-air in-soil Ebinger -2cm 8cm 18cm 18cm

36 Page 36 / Foerster In-air & In-soil in-air in-soil roll roll pitch pitch Foerster -2cm 8cm 18cm 18cm

37 Page 37 / Guartel In-air & In-soil in-air in-soil roll Roll pitch pitch Guartel -2cm 8cm 18cm 18cm

38 Page 38 / Schiebel In-air & In-soil in-air in-soil Schiebel -2cm 8cm 18cm 18cm

39 Page 39 / Vallon In-air & In-soil in-air in-soil 1 in-soil 2 Vallon -2cm 8cm 18cm 18cm 18cm

40 Page 40 / Sweep heights Position Calibration (cm) Height (cm) Altitude (cm) Ebinger Foerster H M L Guartel III II I Schiebel Vallon