Measurement of the homogeneous temperature of a coil by increasing the resistance of a wire

Abstract

The invention relates to a method of measuring the temperature of a coiled component comprising the injection of a known DC current into a gauge wire (1) made of resistive material, the resistance of the gauge wire varying with temperature according to a known law, the measurement of potential difference between the terminals (7a, 7b) of said gauge wire, and a step of calculation transforming the potential difference into a mean temperature of the gauge wire, said gauge wire (1) being wound inside the coil, and arranged as a series of “outbound” turns (5) and a series of “inbound” turns (6) associated pairwise with a geometry and a position that are substantially equal. It also relates to a component made in order to be able to implement this method and the measurement device as a whole.

Claims

1. Method for measuring the temperature of a coiled component, a method comprising the injection of a known direct current into a gauge wire made from resistive material, the resistance of the gauge wire varying with the temperature, the measurement of a potential difference between the terminals of said gauge wire, and a calculation step converting the potential difference into the mean temperature of the gauge wire by determining its resistance, said gauge wire being wound inside the coil, and arranged in a series of “outbound” turns and a series of “inbound” turns associated in pairs with a geometry and position that are substantially equal, a method characterised in that: said gauge wire has a diameter in a range from 0.05 mm to 0.25 mm, and in that winding the gauge wire produces at least twenty turns inside the coil, the variations in resistance of the gauge wire being between 2 and 8 ohms for a temperature varying between −60° C. and 200° C., and the length being adjusted so as to obtain said variations in resistance in said temperature range, in order to apply said method to a high-powered coiled component for aeronautical applications.

2. Method for measuring the temperature of a coiled component according to claim 1, in which the relationship between the temperature and the resistance is linear in the temperature range measured.

3. Method for measuring the temperature of a coiled component according to claim 1, in which the material of the wire is copper.

4. Method for measuring the temperature of a coiled component according to claim 1, in which the measurement of the potential difference uses two complementary wires attached to the terminals of the gauge wire.

Description

DESCRIPTION OF AN EMBODIMENT OF THE INVENTION

(1) A non-limitative embodiment of the invention is now described in more detail with reference to the accompanying drawings, on which:

(2) FIG. 1 is an axial section of a coiled component;

(3) FIG. 2 is the outline diagram of two turns circulating the measuring current in opposite directions;

(4) FIG. 3 shows the principle of measuring with four wires on the gauge wire.

(5) A typical coiled component, for example as transformer as shown in FIG. 1, comprises two active windings 2 and 3. They are configured so that there is an outside winding 3 surrounding the inside winding 2, the whole enveloping a column 4 with a central core.

(6) The heating of the component is due essentially to the Joule losses in the active windings because of the high currents used. It is a case of estimating the heating of the high-power coiled components, preferably for aeronautical applications. A small-diameter copper gauge wire 1 is therefore wound on a cylinder between the two active windings. Measuring the variation in resistance of the gauge wire 1 related to the variation in resistivity of the material in the component as a function of temperature makes it possible to obtain a temperature measurement representing that of the inside of the coil, and therefore exceeding the temperature that is observed on the skin of the component.

(7) Copper is chosen because it makes it possible to obtain correct measurements with small wire diameters. In addition, it is a common material in electronics, compared for example with platinum used in certain temperature measurement equipment.

(8) Moreover, the device is easy to integrate in the manufacture of the component described since it suffices to wind the gauge wire 1 at the same time as the inside active winding 2, on its external surface, before assembling it with the rest of the component, which does not require any additional operation. In general terms, the components the temperature of which it is wished to monitor have a diameter of between 1 and 30 cm. The diameter of the gauge wire used is generally between 0.25 and 0.05 mm, which, at a nominal resistance of 6 ohms at ambient temperature (20° C.), gives rise to a length of gauge wire of between 17 meters and 1.5 meters, that is to say at least twenty turns. This length may have an impact on the final diameter of the component, the wire being able to represent between 0.1% and 10% of the total volume of the conductors. It can therefore be seen that this device disturbs the geometry of the component in a proportion of the same order of magnitude as the percentage of gauge wire, which is small compared with conventional means.

(9) FIG. 1 presents an embodiment with two active windings. In a variant concerning a coiled component with more than two active windings, the gauge wire is wound inside the coil placed between the two innermost windings. In another variant, concerning a single active winding, the gauge wire is wound against the inside face of this winding.

(10) The temperature of the coiled components is monitored in a range of around −60° C. to +200° C. In this temperature range, the resistivity of the copper as a function of temperature is linear and is expressed in the form:
ρ=ρ.sub.0.Math.(1+α.Math.θ)  (1) α=0.00427 (coefficient of variation of the resistivity of copper as a function of temperature) θ=temperature expressed in ° C. ρ.sub.0=resistivity of copper at 0° C. in ohms.Math.meters (1.6 10.sup.−8 ohms.Math.m)

(11) For a gauge wire of given length and cross section, a resistance is therefore obtained that is expressed in a similar manner:
R=R.sub.0.Math.(1+α.Math.θ)  (2) R.sub.0=resistance of the gauge wire at 0° C. in ohms.

(12) In addition, the potential difference being given by Joule's law, in this case the temperature is obtained easily by applying a known current I, expressed in amperes, passing through the gauge wire, by measuring the potential difference U, expressed in volts, at the terminals of said gauge wire by the formula:
θ=(U/I.Math.R.sub.0−1).Math.(1/α)  (3)

(13) For the coiled components envisaged, the value of R.sub.0 sought during this calibration is between 2 and 8 ohms. This makes it possible to have variations in the value of the resistance of several ohms, between 2 and 8 ohms, over the range of temperature variations expected for the component in operation. The amplitude of this variation in resistance for the temperature range sought (from −60° C. to +200° C.) makes it possible to have a measurement with a precision greater than 1%, appreciably improved compared with that of conventional means, as detailed below in an example.

(14) It can be envisaged to use materials other than copper. If the resistivity of the material is not a linear function of temperature, the relationship between the temperature and the measurements of the potential difference U will simply be a little more complex to program. Advantageously, this material will have a resistivity of between 1 and 10 10.sup.−8 ohms.Math.m, preferably between 1 and 7 10.sup.−8 ohms.Math.m, and will make it possible to manufacture gauge wires the resistance of which will vary substantially in the ranges mentioned above for the range of operating temperatures of the coiled component.

(15) To be able to use this formula, it is however necessary to eliminate the sources of parasitic potential differences. In the case of the coiled component, the winding of the gauge wire having a magnetic flux pass through it, an electromotive force, equal to the derivative of the magnetic flux passing through the winding, will appear at the terminals (Faraday's law).

(16) In order to compensate for this electromotive force and to simplify the associated measuring circuit, the coiling of the gauge wire is effected by folding the wire on itself at its middle, and then winding this double wire. Thus two series of turns are created, associated in pairs in an “outbound” turn 5 and an “inbound” turn 6 as illustrated in FIG. 2. These two turns have substantially the same position in space and the same form. It is therefore the same magnetic flux Φ that passes through them and thus the electromotive forces created at their terminals are equal and of opposite signs. The resultant of the electromotive forces observed at the terminals of the gauge wire therefore remains substantially zero.

(17) Other geometrical arrangements of the winding of the gauge wire can be envisaged for thus matching the “outbound” turns 5 and “inbound” turns 6 in pairs. In all cases, it is important to keep the “outbound” and “inbound” parts of the conductor as close as possible to each other, including at the start and end of the winding, in order to thus guarantee the equality of surface area of turns through which the magnetic flux passes.

(18) By injecting a direct current, it is then possible to determine the temperature at the core of the component by means of formula (3). A filtering of the “low-pass” type is used to dispense with any residual voltage related to the difference in equivalent surface area through which magnetic flux passes for the “outbound” and “inbound” turns.

(19) Preferably, the voltage is measured using the so-called 4-wire method, or Kelvin method. In this method, the terminals 7a and 7b of the length of gauge wire that corresponds to the known value R.sub.0 of the resistance at the reference temperature and which is used in the calculations are considered. Two complementary wires 8a and 8b are attached to these terminals by brazing, soldering or any other connection means. Next the gauge wire 1 is connected by its two ends to a current-generating means 10 and the two complementary wires 8a and 8b are connected to a means for measuring potential, a voltmeter 9. The impedance of the voltmeter 9 being very high, the current that passes through the connection wires is negligible and the potential difference is measured with great precision over the exact length of wire corresponding to the resistance R.sub.0. The intensity of the current passing through the gauge wire is moreover indicated with good precision by the current-generating means.

(20) On the other hand, the copper wires are provided with a certain tolerance on their variation in radius. Typically, the mean radius may vary by +/−2.5% for wires with a diameter of 0.1 mm. The uncertainties on the temperature measured will therefore be around 5% if the nominal data are relied on.

(21) Preferentially, the precision of measurement is improved further by calibrating the gauge wire before integrating it in the component. Because of the order of magnitude of a few ohms of the resistance of the gauge wire (see the example provided in table (1)), the calibration can be carried out with a micro-ohmmeter in order to achieve precisions of around 0.2% on the resistance R.sub.0. When the operator has identified the precise length corresponding to the theoretical value of the resistance over 6 m (see table (1) for an embodiment with the accepted tolerances), he connects, to the corresponding terminal 7a and 7b, the complementary wires 8a and 8b used for the potential measurement and then winds the gauge wire in the coiled component. This measurement means, with the calibration carried out, makes it possible to have a temperature probe having a precision of +/−0.3%, to be compared with the current typical mean value of 1% with temperature probes placed against the component. In addition, the manufacturing cost of a temperature probe according to the invention is less.

(22) In a variant embodiment, errors introduced by the uncertainties as to the current value I supplied by the means 10 are dispensed with in an additional manner by directly measuring the resistance of the gauge wire between the terminals 8a and 8b. For this purpose, a resistor of known value is placed at a point on the circuit of the current I not subjected to the variations in temperature of the coiled component. The variation in potential is measured at the terminals of this resistor and the resistance of the gauge wire is obtained directly by a ratio between the two potential differences measured.

(23) Compared with a component equipped with a temperature gauge, the assembly installed in the aircraft therefore consists of this modified component with a current generator, a voltmeter and a computing module able to supply the temperature from the measurements made, the latter three components being similar in complexity to ohmmeters available on the market.

(24) TABLE-US-00001 TABLE (1) Example concerning an autotransformer, evaluation of the measurement area on the temperature between −55° C. and +175° C. Mean length of a turn: 15 cm Total number of turns: (“outbound” and “inbound”): 40 Length of wire used: 6 m Mean radius of wire: 0.07081035 mm; manufacturing tolerance: +/−2.5% Temperature Nominal Maximum ° C. Minimal values values values Resistance (in −55 4.456 4.466 4.476 ohms) of the 25 6.700 6.710 6.720 gauge wire 175 10.908 10.918 10.928 adjusted around 6 m after calibration at +/−0.01 ohms Tolerance of the −55 −0.224 0 0.224 resistance of the 25 −0.149 0 0.149 invention 175 −0.092 0 0.092