Patent Application
20180245994
The present description relates to a device and a method for hysteresis compensated force sensing, and in particular to a device and a method for automatic compensation of a hysteresis related deviation of sensing signals at magnetic field based force sensing.
Hysteresis is a known phenomenon in ferromagnetic materials like iron (Fe), nickel (Ni), cobalt (Co) and their alloys. It is the inability of the material to return back to its original state when an external force or magnetic field previously applied to it is removed. Hysteresis can occur due to the mechanical pressure applied on the material (elastic hysteresis) and also due to the applied magnetic field (magnetic hysteresis).
In any sensor design and development industry, it is critical for the sensing unit to be capable of measuring the best signal without noise or other distortions. Noise can be compensated by the implementation of analogue filters and also by software algorithms. Apart from noise, the other major distortions include temperature, humidity, vibration, as well as hysteresis. Depending on the quality of the test object's alloy, the measured hysteresis may be larger than ±2.5% which may not be acceptable in some applications.
There may be a need for hysteresis compensation at a magnetic field based force sensing, a corresponding programme element and computer readable medium.
According to an exemplary embodiment there is provided a sensor device for measuring forces applied to an object to be sensed with the sensor, the sensor comprising a magnetic field generating element being adapted for generating a magnetic field in order to generate a magnetic flux in an object to be sensed, a first magnetic field sensing element being adapted to sense a magnetic field depending on a variation of the generated magnetic flux in said object to be sensed, a driving unit being adapted to drive the magnetic field generating element with a first driving signal and with a second driving signal, and an evaluating unit being adapted to evaluate a first sensing signal based on the magnetic field sensed by the first magnetic sensing element resulting from driving the magnetic field generating element with the first generating signal, and a second sensing signal based on the magnetic field sensed by the first magnetic sensing element resulting from driving the magnetic field generating element with the second signal, wherein the first driving signal is an alternating current with a first predetermined frequency and the second driving signal is an alternating current with a second predetermined frequency, which is different from the first predetermined frequency.
Thus, it is possible to eliminate the hysteresis effect from the sensing results. As the mechanical stress of an applied force to an object is higher close to the surface, also the hysteresis will be larger in the outer regions beneath the surface. As different frequencies of externally applied magnetic fields result in a different penetration depth owing to the skin effect, at least two different frequencies provide at least two sensing signals having a different absolute hysteresis effect depending on the penetration depth. This can be used to eliminate deviations from the sensing signals resulting from the hysteresis effect.
According to an exemplary embodiment the first signal has a first predetermined amplitude and the second signal has a second predetermined amplitude, wherein the first predetermined frequency is higher than the second predetermined frequency and the first predetermined amplitude is lower than the second predetermined amplitude.
Thus, the penetration effect can be supported and further the expected signal strengths can be compensated beforehand. As the higher frequency will represent the areas beneath the surface having a stronger hysteresis effect, the lower amplitude makes a later normalizing easier, or even superfluous.
According to an exemplary embodiment the first generator signal and the second generator signal are applied alternately, so that either the first generator signal or the second generator signal is applied to the magnetic field generating element.
Thus, the signals can be separated with respect to the generating frequency. No disturbances are expected, as the frequencies are clearly separated at the driving unit while generating a magnetic field, as well as the evaluation unit while evaluation the signals. A common clock for synchronized switching can be used by the driving unit and the evaluation unit.
According to an exemplary embodiment the first generator signal and the second generator signal are applied at least temporally overlapping, so that at least temporarily both the first generator signal and the second generator signal is applied to the magnetic field generating element.
Thus, the sensing procedure can be shorter. If a switching can be avoided, also transits can be avoided. The separation of the signals can be realized with e.g. band pass filters. The overlapping frequencies are an alternative to the clearly temporary separated application of different frequencies. It should be noted that also a combination of a frequency separation and a frequency overlapping can be applied, In particular when using more than two different frequencies, e.g. four frequencies. In this case two frequencies can be applied alternately and other two alternately applied frequencies can be applied parallel to the first two frequencies.
According to an exemplary embodiment the evaluating unit is adapted to determine the difference between the first sensing signal and the second sensing signal at an applied force, and to determine a correction value for at least one of the first and second sensing signals in order to compensate a hysteresis related deviation between the respective sensing signal sensed when applying said force to an hysteresis biased object to be sensed and a sensing signal sensed when applying a force to an hysteresis unbiased object to be sensed.
Thus, the hysteresis effect onto the sensing signals can be eliminated. It should be noted that at hysteresis biased objects a signal of an unbiased object is not available and only a theoretical value. However only this theoretical value represents the true force applied, wherein at a biased object it is to be distinguished between the signal portion resulting from the applied force and the signal portion resulting from the hysteresis. Consequently, if it is possible to determine the amount of contribution of the hysteresis, the signal portion relating to the force can be determined and thus the true force.
According to an exemplary embodiment the correction value is determined based on an algorithm being representative for the object to be sensed.
Thus, a fast elimination of the hysteresis effect onto the sensing signals can be achieved. The algorithm may be provided by a manufacturer of the object or may be determined when installing the sensor by e.g. applying a spectrum of forces before and after a maximum force, i.e. traveling along the entire hysteresis characteristic.
According to an exemplary embodiment the sensor device comprises an initializing unit being adapted to determine the algorithm being representative for the object based on a test cycle including at least two recorded sensing signal sets, each sensing signal set comprising a first sensing signal and a second sensing signal at an applied force.
Thus, the sensor is capable to analyze the respective object to be sensed itself. The sensor detects the characteristic of the object and determines the algorithm which serves as a base for further sensing procedures. This can be supported by a particular test cycle. This algorithm can successively be adapted and improved so that the sensor, in particular the evaluation unit learns the algorithm. This is of relevance as the hysteresis of an object may change depending on the applied maximum force ever, as soon as this maximum force is lower than the force where the hysteresis arrives at its saturation.
According to an exemplary embodiment the sensing signal sets are recorded at at least first and second applied forces, wherein at least one sensing signal set is recorded before applying a force being higher than the first and second applied forces, and wherein at least one sensing signal set is recorded after applying a force being higher than the first and second applied forces.
Thus, sensing values can be provided which include values of both branches of the hysteresis characteristic. This makes the calculation of a correction value and thus the compensation more exact. In this case the first force may be applied before having reached a higher force corresponding to the turning point of the hysteresis characteristic, whereas the second force is applied after having reached said higher force corresponding to the turning point of the hysteresis characteristic. In particular if the first and second force is the same, the compensation may be very exact. In particular if the hysteresis is symmetrical, the true value is the average of the both sets recorded at said force. It should be noted that in this case two sensing values before and after having applied the higher force at a single frequency may be sufficient to determine the median and thus the true value for force determination.
According to an exemplary embodiment the sensor comprises a normalizing unit being adapted to bring the first sensing signal and the second sensing signal of an actually applied force into a relation corresponding to a relation of previously determined first and second sensing signals of a previously applied maximum force.
Thus, the evaluation is facilitated. If a signal of one frequency is multiplied/divided by a factor so that the maximum signal values of both frequencies are the same, a direct comparison of the both signals is possible. The calculation of the compensation is easier when having normalized the sensing values beforehand.
According to an exemplary embodiment the sensor device comprises as the first magnetic field sensing element a first inductance arrangement having a first coil with a corresponding main sensing direction, and as a second magnetic field sensing element a second inductance arrangement having a second coil with a corresponding main sensing direction, wherein first inductance arrangement and the second inductance arrangement are connected, wherein the first coil and the second coil are oriented so as to have the corresponding main magnetizing directions oriented toward an object to be sensed, wherein the first coil and the second coil with respect to the corresponding main sensing directions are connected anti-parallel, in particular such that in a homogenous magnetic field applied to the first coil and the second coil the signal of the first coil and the signal of the second coil results in an at least partially compensation of signals at the first coil and the second coil, leading to a resulting signal over the first inductance arrangement and the second inductance arrangement.
Thus, the signal strength can be improved, so that together with the hysteresis compensation a more exact force sensing can be provided. As the compensated signals of the first and second coil result in zero when having applied no force or a fixed offset force, the sensor electronics may have a higher resolution for sensing the resulting signal of the first and second coil when having applied a force to be detected.
According to an exemplary embodiment the sensor further comprising a flux concentrator with a first leg and a second leg, wherein the first coil is wound around the first leg and the second coil is wound around the second leg.
Thus, the sensing of the magnetic field can be done more effectively to the object to be sensed. Stray effects can be minimized.
According to an exemplary embodiment the flux concentrator comprises a further leg, wherein the magnetic field generating element is a coil wound around the further leg, wherein the further leg is located at a joint of a yoke branch of the first leg and a yoke branch of the second leg.
Thus, the generated magnetic field can be more efficiently applied the object to be sensed.
According to an exemplary embodiment the yoke branch of the first leg and the yoke branch of the second leg form a “V”, in particular are orthogonal with respect to each other.
Thus, a clear separation of the different types of applied forces can be achieved. In particular parasitic effects can be eliminated. If for example a sensor with such a flux concentrator is applied to an object to sense torque, an offset bending will be automatically compensated as bending will have the same effect to the first and second magnetic field sensing element so that no bending related signal will result. The torque impact however will result in a anti-parallel effect, so that torque related signals will add but not compensate.
According to an embodiment the yoke and/or flux concentrator is formed in one piece, in particular made of a sinter material. According to an embodiment the branches of the “V” are orthogonal, i.e. include an angel of substantially 90°. According to an embodiment the legs extend orthogonal from the yoke/branches and receive the sensor coils. According to an embodiment the generator coil is arranged at the leg extending from a junction point of at least two branches of the yoke. According to an embodiment the transit from the yoke branches to the respective legs is rounded, in particular with a radius being larger than the diameter of the legs.
According to an exemplary embodiment the first inductance arrangement further has a third coil with a corresponding main sensing direction, wherein the second inductance arrangement further has a fourth coil with a corresponding main sensing direction, wherein the third coil and the fourth coil are oriented so as to have the respective main magnetizing directions oriented toward an object to be sensed, wherein the first coil and the third coil are connected in series and the second coil and the fourth coil are connected in series; wherein the series connection of the first coil and the third coil with respect to their main sensing directions are connected anti-parallel to the series connection of the second coil and the fourth coil with respect to their main sensing directions, in particular such that in a homogenous magnetic field applied to the first coil, the second coil, the third coil and the fourth coil results in an at least partially compensation of signals at the first coil, the second coil, the third coil and the fourth coil, leading to a resulting signal over the first inductance arrangement and the second inductance arrangement. According to an embodiment the generator coil is arranged at the leg extending from a crossing of a yoke having four sensing coils. According to an embodiment the transit from the yoke branches to the respective legs is rounded, in particular with a radius being larger than the diameter of the legs.
Thus, additional settings can be achieved when dimensioning the coil parameters, in particular the relation of the first and third coil to each other and the relation of the second and fourth coil with respect to each other. A difference in the number of windings may shift the zero point so that directions of the applied torque can be determined.
According to an exemplary embodiment the first leg, the second leg, the third leg and the fourth leg are arranged at corners of a rectangle, in particular a square, with respective end faces of the legs oriented toward the object to be sensed.
Thus, a more or less symmetrical sensor design can be achieved. The sensing failures can be reduced.
According to an exemplary embodiment the further leg is arranged in the centre of the rectangle, in particular the square, with a respective end face of the further leg oriented toward the object to be sensed.
Thus, a better coupling can be achieved, so that the stray losses can be reduced. The magnetic field generator in the centre also avoids un-symmetries and provides an efficient build-up of the sensor unit.
According to an exemplary embodiment the sensor device comprises a flux concentrator, wherein the flux concentrator has wound there around a coil, which coil serves as a magnetic field generator and as magnetic field sensing element at the same time, wherein the sensor signals are gained by either a current detection between the driving unit and the coil, or a voltage drop over the coil. The coil then is driven by the both frequencies.
Thus, only one coil for generating and sensing is required. This design is an alternative to the design with a generator coil and two sensing coils or a generator coil and four sensing coils.
It may be an aspect of the description to compensate a hysteresis effect of an object to be sensed, so that a gained signal representing the applied torque can be freed of a hysteresis related signal portion to arrive at a “true” applied force related signal.
It should be noted that the above features may also be combined. The combination of the above features may also lead to synergetic effects, even if not explicitly described in detail.
In other words, the approach described herein may be summed up as follows:
The Active Sensing Technology approach described herein enables real-time compensation for unwanted signal hysteresis. This can be done with one of the following three activities:
a) Taking measurements with different IOF (Internal Operating Frequency)
b) Taking measurements when using different (changing) generator magnetic field strength
c) Taking measurements by using a combination of both: different IOF's and different generator magnetic field strength
To compensate for unwanted signal hysteresis, at least two different measurements may be taken and the results of these measurements may be processed further to achieve the desired results: reducing the effects of signal hysteresis.
Taking the first torque measurement using the IOF of 2.000 Hz (2 kHz) and storing the measurement result (A), followed by the second torque measurement using a different IOF, like 10.000 Hz (10 kHz) and storing the measurement result as (B). The measurement results A and B have to be normalised first before the difference will be built between them.
The torque measurement taken at the higher IOF (like 10.000 Hz) (B), will show a much larger effect of the signal hysteresis than the measurement taken at a lower IOF (like 2.000 Hz) (A).
Depending on the application specific circumstances the result of the difference (A-Normalised)−(B-Normalised) may have to be multiplied by a final correction factor to meet calibration expectations.
Taking mechanical force measurements (like torque forces) with a higher IOF will result in mechanical stress measurements taken from the surface of the test object. The unwanted effects of the signal hysteresis are strongest at the outer layers (surface) of the test object. Therefore, the torque measurement B will show a much larger signal hysteresis than the torque measurement taken at a much lower IOF. A high IOF (like 10.000 Hz) and above will have a very limited penetration into the surface of the test object. The to be expected measurement depth is in the area of 50 μm to 200 μm.
Taking mechanical force measurements with a low IOF (in the area of 500 Hz for example) will result in mechanical stress measurements taken at the outer and the inner layers of the test object. With a low IOF, the Active Torque Sensor can penetrate deeper and deeper below the surface of the test object. Depending on several physical design factors, the measurement depth into the test object (below the surface of the test object) can be several millimetres (example 1.5 mm to 2.0 mm).
Identical to Example 1 with the exception that instead of using different IOFs, the IOF is constant for both measurements to be taken A and B. The difference between the measurement A and B will be that the magnetic flux density (emanated by the sensing module) will change. The magnetic flux density of the generator (LG) can be changed by changing (increasing or decreasing) the electric current (IG) with which the magnetic field generator LG is powered. In this example, the first measurement (A) will be taken while IG is high and the second measurement (B) will be taken while IG is much lower. The rest of the signal processing is then identical to Example 1.
When using a large current (AC or DC) to power the magnetic field generator LG (like 100 mA), mechanical stresses will be measured from the surface of the test object and from layers that are deeper into the test object. When using a small electric current (like 25 mA) then only the outer layers of the test object will be penetrated by the emanating magnetic flux lines.
For example, the device described herein may be used for measuring mechanical forces in drilling pipes that have been tooled from unhardened materials. Mechanical force measurements may be executed directly at pipes. The metal pipes may be ferromagnetic, unhardened, and manufactured from low cost material.
When an external magnetic field is applied to a ferromagnetic material such as iron, the atomic dipoles align themselves with the magnetic field. Even when the field is removed, part of the alignment will be retained: the material has become magnetized. Once magnetized, the magnet will stay magnetized indefinitely. To demagnetize, it requires heat or a magnetic field in the opposite direction. This is the effect that provides the element of memory in a hard disk drive.
The relationship between field strength H and magnetization M is not linear in such materials. If a magnet is demagnetized (H=M=0) and the relationship between H and M is plotted for increasing levels of field strength, M follows the initial magnetization curve. This curve increases rapidly at first and then approaches an asymptote called magnetic saturation. If the magnetic field is now reduced monotonically, M follows a different curve. At zero field strength, the magnetization is offset from the origin by an amount called the remanence. If the H-M relationship is plotted for all strengths of applied magnetic field the result is a hysteresis loop called the main loop. The width of the middle section is twice the coercivity of the material.
A closer look at a magnetization curve generally reveals a series of small, random jumps in magnetization called Barkhausen jumps. This effect is due to crystallographic defects such as dislocations. The larger the mechanical stress measurement range is, the larger the measurement hysteresis will be.
These and other aspects will become apparent from and elucidated with reference to the embodiments described hereinafter.
Exemplary embodiments will be described in the following with reference to the following drawings.
This patent application describes a magnetic principle based non-contact mechanical force sensing technology that no longer requires any type of treatment of a ferromagnetic test object 2 from where the measurements have to be taken and is very low complexity in design. The following terminology will be used in this specification: sensor, sensor unit, sensing module, sensor system, and test object.
The sensing module has to be placed nearest to the test object. The test object is the item from where mechanical forces have to be measured. Together, the sensor and the test object will build a sensor system. The test object is a ferromagnetic material like a power transmitting shaft where the mechanical forces have to be measured from. The sensing module is the item that needs to be placed nearest the test object 2 where the mechanical forces have to be measured from. The SCSP electronics is the Signal Conditioning and Signal Processing electronics.
As can be seen in
The test object can be of any shape like a symmetrically shaped shaft or beam, flat non-symmetrically shaped beam, tube, etc. and has to be tooled from ferromagnetic material.
The physical principle described here is to measure the magnetic energy dissipation caused by the applied mechanical stresses/forces at the test object where the measurement will be taken from. A raw signal output of the sensor may have a better quality and a better performance when the test object 2 has been hardened. Better performance means that the measurement signal hysteresis may be larger when using un-hardened test object. It is not necessary that the entire test object has to be hardened. It is sufficient when the hardening process has been applied at least to the sensing location.
The sensing module 10, 50 comprises a magnetic field generating element 50, which can be a inductor (coil), mounted in a holder or a fixture that allows to place this magnetic field generator nearest to the actual sensing location. The inductor (coil) is electrically driven by a signal generator that operates at certain frequencies and is capable to drive the inductor load. The coil may be wound around a flux concentrator. The alternating magnetic field which essentially is a form of energy that travels through the flux concentrator into the surface of the test object 2 and from there again back into the flux concentrator. Depending on the amount and the orientation of the mechanical stress that is applied to the test object 2, the magnetic energy flow will be modulated, meaning that with a high mechanical stress applied a higher amount of energy will be dissipated by the test object 2. In reverse, when the test object is in a relaxed condition, i.e. no mechanical stress applied, the energy dissipation in the test object 2 is at its lowest level assuming that all the other conditions are kept constant or identical.
The sensor electronics has two main functions: generating the signal to drive the magnetic field generator, i.e. inductor or coil to be done by the driver 200, and to recover and process the signal modulation, coming back from the sensing module to be done by an evaluation unit 300, which can be considered as the signal conditioning and signal processing module of the sensor electronics: SCSP module. The electronic signal generator 200 that is driving the magnetic field generator coil 50 is operating with e.g. a constant output voltage. A higher or lower magnetic energy dissipation caused by the mechanical stresses that are applied to the test object results in a modulation of the electric current flow through the magnetic field generating inductor. The SCSP module 300 of the sensor electronics has the task to identify and to measure the amount of the modulation of the supply current that is driving the magnetic field generator coil. After the sensing module has been placed in the correct position and orientation in relation to the “to-be-measured” mechanical forces, a sensor initialization process may be executed by an initializing unit 310. During the initialization process the sensor electronics will identify the optimal signal generator and signal processing settings and will store these settings in an internal non-volatile memory. When using a non-volatile memory for the sensor settings, then the initialization process has to be executed only the very first time when the sensor will be used together with a specific test object 2 and at a specific sensing location.
The orientation with which the magnetic flux lines will travel through the surface of the test object will decide about which mechanical forces will be detected and measured by the sensor. Therefore great care has to be taken when placing the sensing module at the test object. Depending on the angular orientation of the main axis of the flux concentrator, a specific mechanical force will be detected and measured more easily, while the signals from other mechanical forces will be suppressed or cancelled. In an orientation where the magnetically effective axis of the flux concentrator is placed onto the test object 2 with a zero degree angle in relation to the main axis of the test object is most suitable to measure torque forces that will be applied to the test object and will almost completely suppress the effects caused by bending forces. The sensing module orientation where the magnetically effective axis of the flux concentrator is placed onto the test object with a 45 degree angle in relation to the main axis of the test object is most suitable to measure bending forces that will be applied to the test object and will almost completely suppress the effects of applied torque forces. Any angle between these two positions >zero degree to <45 degree will result that the sensor signal output represents a mixture of both torque forces and bending forces
The sensing technology allows the use of a very wide range of ferromagnetic materials for the test object. Basically anything can be used as the test object material as long as “a permanent magnet will be strongly attracted” to the test object. However, when using lower cost alloys, the risk will increase that the measurement output signal of a sensor system will be effected by a noticeable larger measurement or sensing hysteresis. Depending on the quality of the test object alloy, the measurement hysteresis may be larger than +/−2.5% which may not be acceptable in some applications. By sending out e.g. successively one-after-another two different types of electric driver signals that are responsible for generating the magnetic field in the inductor the magnetic field generated by the sensing module will differently penetrate the surface of the test object. The driver signals can vary in frequency and in the signal amplitude, i.e. amount of magnetic energy. Depending on the combination of these two parameters the magnetic flux lines that travel beneath the surface of the test object will travel at different depth. For example, a high generator frequency will cause that the magnetic field will travel nearest the surface of the test object. A larger driver signal amplitude will cause the magnetic flux lines to be pushed deeper beneath the test object surface.
When using alternating two different driver signals like a first signal with a high frequency and a low driver signal amplitude, and a second signal with lower frequency at high driver signal amplitude then the first signal will travel very near the surface of the test object, for example between two pole shoes of a flux concentrator and the second signal will travel deeper beneath the test object surface.
When a specific value of mechanical stress is applied to the test object like a specific bending force or a specific torque, then there will be a difference between these two measurements, here called dU, with dU=U1f1−U1f2 or dU=U2f1−U2f2. Depending on what the mechanical stress history was before this specific sensing has been taken, the hysteresis differential signal is either positive or negative. In other words, when applying a particular force, the signal for the first and second frequency driving signal depends on the “history” of the applied force. When starting from a low force and increasing the force, to the sensing force, the signal will be U2f1 and U2f2, whereas when lowering the force from Fmax back to the same sensing force, the signal will be U1f1 and U1f2. The hysteresis differential signal will have a negative sign when the current mechanical stress sensing is executed after a much higher mechanical forces have been applied to the test object. The actual amount of dU and its polarity will indicate correctly by how much the sensor output signal has to be corrected to substantially reduce or even to eliminate the effect of the signal hysteresis. Based on the dU gained from the signals U1f1, U1f2 or U2f1, U2f2, the “real” signal value can be determined, which is on the center line of the hysteresis. It should be noted that this “real” signal value which is the force related portion of the signal cannot be sensed at hysteresis biased materials of test objects, as the hysteresis masks this signal value. The “real” signal value is a much more exact measure for the actually applied force. In other words, the described hysteresis correction allows a more exact determination of the applied force. The described method does not render necessary to have applied the Fmax, as only one set of sensing values is sufficient to compensate the hysteresis, in particular that set gained when increasing the force starting from zero, without having reached Fmax.
This principle explains hysteresis compensation by the use of two alternating different driver signals such that, a first signal has high frequency and low signal amplitude whereas the second signal will have lower frequency and high signal amplitude.
When applying a generator signal in an interleaving manner, alternating the first frequency is applied for a particular time and then the second frequency is applied for a particular time to the generator coil 50. The evaluation unit 300 may apply the same interleaving rhythm when evaluation the sensing signals. In case the driving unit 200 and the evaluation unit 300 are implemented as a single electronic entity, the driving unit 200 and the evaluation unit may use the same clock so as to separate the signals for the respective frequency. As an alternative the generator frequencies may be applied at the same time so as to have a generator signal with two or even more frequency components. In this case the evaluation unit may separate the sensing signals by applying e.g. a band pass filters. This may render a clock obsolete. It should be noted that also more than two frequencies may be applied, wherein some of them may be applied in an interleaving mode and some overlapping. In this case both a clock and respective filters are required.
Sensing coils 20 and 40 are connected in series and form sensing arrangement 2040. The sensing arrangement 1030 is connected anti-parallel to sensing arrangement 2040.
It should be noted that the principles described here may be applied to other sensor geometries, as far as the sensor principle is based on a magnetic sensing principle.
It should be noted that the term ‘comprising’ does not exclude other elements or steps and the ‘a’ or ‘an’ does not exclude a plurality. Also elements described in association with the different embodiments may be combined.
It should be noted that the reference signs in the claims shall not be construed as limiting the scope of the claims.
| Number | Date | Country | Kind |
|---|---|---|---|
| 15188528.2 | Oct 2015 | EP | regional |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/EP2016/073870 | 10/6/2016 | WO | 00 |
