Setup for Calibrating a Tool-Workpiece Thermocouple by the Method of Electric Contact Heating
Abstract Generated abstract
The paper describes an apparatus for calibrating natural tool, workpiece thermocouples used to measure temperatures in metal cutting. It critiques conventional bath and coil heating methods as slow, labor-intensive, and insufficiently representative of the localized heating in cutting, then presents a contact-heating setup in which high current passes through electrodes and a hard-alloy tool element while thermoelectric signals are recorded and filtered. The article gives the device’s main specifications, measurement sequence, circuit corrections based on Kirchhoff’s laws, and practical simplifications when contact resistances are made equal. The proposed method permits calibration to high temperatures in about 5 to 10 minutes, with checks for equal junction temperatures and reduced parasitic thermoelectric effects, including calibration on actual tool inserts or similar cutting tools.
Full Text
SETUP FOR CALIBRATING A TOOL–WORKPIECE THERMOCOUPLE BY THE METHOD OF ELECTROCONTACT HEATING
V. A. Naumov
(Presented by the Department of Machine Tools and Metal Cutting)
Purpose and technical data of the setup
At present, the generally accepted method for measuring temperatures in the cutting zone of metals has become the method of the natural thermocouple: tool–workpiece, most fully described in [1]. However, for its successful application, careful calibration of the pair under study—machined metal–hard alloy—is necessary. The most widespread of the existing calibration methods consists of the following (Fig. 1). A rod made of the machined material and another rod made of hard alloy are placed in a bath with a molten heat carrier (lead, antimony, etc.) 3; their cold ends are connected to a galvanometer (millivoltmeter) 2. Between them is placed the junction of a reference thermocouple connected to a pyrometric millivoltmeter 1. Heating is carried out by various methods, in particular by means of a heating coil 4. During heating and cooling, the readings of millivoltmeters 1 and 2 are recorded simultaneously, and calibration curves are constructed from the data obtained.
Fig. 1.
P. I. Bobrik [2] proposed a calibration device, schematically shown in Fig. 2. Two specimens—one of the material being machined, the other of hard alloy (the tool material)—are brought together end to end in a special fixture and heated by an electric heating coil. The readings of the millivoltmeters of the working and control thermocouples are recorded in the same way as in the preceding method. The junction of the control thermocouple is placed in a groove between the end faces of the specimens under investigation.
Fig. 2
The disadvantages of these methods are:
-
Both processes are not good models of the cutting process, since during calibration substantial regions of the specimens under investigation are subjected to heating, whereas in cutting the source of heat generation is a very small zone of deformation and friction.
-
The calibration processes are very lengthy (especially the first) and labor-intensive (manufacture of special specimens, rods of hard alloy, etc.).
-
It is difficult to monitor and take into account the influence of parasitic thermal EMFs.
In the metal-cutting laboratory of TPI, a design has been developed for an installation that makes it possible to carry out the calibration of the natural tool–workpiece thermocouple very quickly and with very good quality; moreover, calibration can be performed on ordinary hard-alloy inserts with maximum approximation to the conditions occurring during cutting.
The setup operates on the principle of heating the junction of two dissimilar metals when a high-power current is passed through them.
Technical data of the setup
- Power consumption — 15 kW
- Maximum current at the calibrated junction — 350–500 A
- Calibration time — 5–10 min
- Temperature range up to which calibration can be performed — 1200–1500°C
Considerably greater heating is also possible
Design and operating principle of the setup
The schematic diagram of the setup is shown in Fig. 3. The setup consists of a three-phase regulating transformer of the RNT 220-12 type with voltage regulation limits of 0–220 V, a welding transformer 4 with an output power of 300–500 A, a system simulating the contact of the cutter with the workpiece, and a system for recording electrical signals. From the material being machined two rods with conical ends are made (in Fig. 3, rods 5 and 11). With their conical ends these rod electrodes are pressed against a hard-alloy rod 7. Instead of electrode 7, an ordinary tool holder may be installed, with which the cutting process is performed. The ends of the power winding of the welding transformer are connected to electrodes 5 and 11. Between one of the electrodes and the hard-alloy rod there is installed a thermal receiver 6 of a platinum–platinum-rhodium thermocouple 10.
Fig. 3
The system for recording electrical signals consists of switches 8, 9, 12, 16, 19; a smoothing filter made up of inductances \(L_1\)—13 and \(L_2\)—15, capacitance \(C\)—14; and recording millivoltmeters and a compensating potentiometer PP. The power expended in the process is recorded by ammeter 3 and voltmeter 2.
The setup operates as follows. By means of transformer 1, the voltage is smoothly raised to the required limit. A high current, passing through the contacts, heats them. After 2–3 seconds the thermal process at the contact pads stabilizes, as is seen from millivoltmeters 10 and 18 (the pointers remain steady). In this stable state, measurements are made, conventionally denoted by the following indices:
a) The potential difference \(U_I\) and the equivalent emf \(E_{\text{eqv }I}\). This measurement is carried out by alternately connecting first the millivoltmeter and then the compensating potentiometer into the circuit.
b) The potential difference \(U_{II}\) and the compensating emf \(E_{\text{eqv }II}\).
c) The potential difference \(U_{III}\) and the corresponding emf \(E_{\text{eqv }III}\).
d) The total potential difference \(U_p\) with the current leads from specimens 5 and 11 connected in parallel, and the corresponding emf \(E_{\text{eqv }p}\).
These measurements are carried out with the switch positions given in Table 1.
Table 1 (to Fig. 3)
| Toggle-switch index in the diagram of Fig. 3 | \(U_I\) | \(E_{\text{eqv }I}\) | \(U_{II}\) | \(E_{\text{eqv }II}\) | \(E_{\text{eqv }III}\) | \(U_{III}\) | \(U_p\) | \(E_{\text{eqv }p}\) |
|---|---|---|---|---|---|---|---|---|
| 8 | a | a | a | a | a | a | a | a |
| 9 | b | b | a | a | b | b | b | b |
| 12 | — | — | — | — | — | a | a | |
| 19 | a | b | a | b | b | a | a | b |
The letters denote the positions of the switches for a given measurement. Toggle switch 16, not included in Table 1, serves to measure the polarity of the instrument when the direction of the current in the circuit is changed.
Together with the thermoelectric emf, the alternating current heating the contact pads also enters the measuring circuit. To smooth it, a filter is used, consisting of a bank of capacitors \(C\)-5000 mf (in Fig. 3—14) and chokes \(L_1\) and \(L_2\), 13 and 15, having the following data: number of turns \(W = 300\), active resistance \(R = 6.0\), diameter of the winding wire 1.2 mm. A check on an electron-beam oscillograph showed that a filter with the above-listed characteristics passes an alternating component of less than 0.01 \(\mu a\) in current and 0.1 \(\mu v\) in voltage, which is quite negligible.
After completion of the measurements at a given temperature regime, the voltage is again raised to the next temperature level, and the set of measurements is repeated. In this way data are obtained for the entire temperature interval of interest to us.
Processing of the Results Obtained
The results obtained during calibration must be processed accordingly, because the electrical circuits during
cutting and during calibration differ from one another. In Fig. 4 the calculation circuits for determining the electromotive forces are shown. The calculation is carried out using the superposition method and Kirchhoff’s laws [3]. Fig. 4a presents the general circuit diagram (measurement of \(U_{\mathrm{p}};\ E_{\mathrm{eqv}}\)). Fig. 4b presents the circuit for measuring \(U_1;\ E_{\mathrm{eqv}}\).
Fig. 4.
Figs. 4c and 4d respectively present the circuit for measuring \(U_{\mathrm{II}};\ E_{\mathrm{eqvII}}\) and \(U_{\mathrm{III}};\ E_{\mathrm{eqvIII}}\).
In these circuits
\[
R_{\mathrm{t}}
\]
— resistance of the transformer winding;
\[
E_1,\ R_1
\]
— the first unknown e.m.f. and the corresponding contact resistance, which is, as it were, its internal resistance;
\[
E_2 \text{ and } R_2
\]
— the second unknown e.m.f. and its resistance;
\[
R_{\phi_1} \text{ and } R_{\phi_2}
\]
— ohmic resistances of the filters;
\(R_{mv}\)—the internal resistance of the millivoltmeter;
\(E_{\mathrm{eqv}}, R_{\mathrm{eqv}}\)—the compensating e.m.f. and the resistance of the compensating device.
In all measurements the potentiometer and the millivoltmeter are connected into the circuit one after the other; therefore the circuits \(4a, б, в, г\) must be considered in the same sequence: first with the millivoltmeter, and then with the potentiometer, or vice versa.
From circuit \(4б\) we obtain (for the millivoltmeter)
\[ U_I = \frac{E_1(R_t+R_2)R_p+E_2R_1R_p} {R_2R_t+R_1R_2+R_pR_t+R_pR_1+R_pR_2}, \tag{1} \]
where: \(R_p=R_{mv}+R_\phi+R_y\),
\(R_y\)—the resistance of the circuit conductors.
When measuring by the potentiometric method, at the moment of measurement the current of the measuring circuit is equal to zero, and for circuit \(4б\)
\[ E_{\mathrm{eqv}I}(R_2+R_1+R_t)=E_2R_1+E_1(R_2+R_t). \tag{2} \]
For circuit \(4в\)
\[ U_{II} = \frac{E_2(R_t+R_1)R_p+E_1R_2R_p} {R_1R_2+R_2R_t+R_pR_t+R_pR_1+R_pR_2}, \tag{3} \]
\[ E_{\mathrm{eqv}II}(R_1+R_2+R_t)=E_1R_2+E_2(R_1+R_t). \tag{4} \]
For circuit \(4г\)
\[ U_{III} = \frac{(E_1-E_2)R_tR_p} {R_1R_2+R_2R_t+R_pR_t+R_pR_1+R_2R_p}, \tag{5} \]
\[ E_{\mathrm{eqv}III}(R_1+R_2+R_t)=(E_1-E_2)R_t. \tag{6} \]
In formulas (1, 3, 5) the denominators are equal; therefore the following proportion can be obtained:
\[ \frac{E_1(R_t+R_2)+E_2R_1}{U_I} = \frac{E_2(R_t+R_1)+E_1R_2}{U_{II}} = \frac{(E_1-E_2)R_t}{U_{III}}. \tag{7} \]
From proportion (7) one can compose 3 equations; however, it contains 4 unknowns \(E_1;\ E_2;\ R_1;\ R_2\). Taking \(R_1=R_2\), which can be achieved by making the ends of the electrodes of the same diameter, we obtain
\[ E_1=U_I\left(1+\frac{R_tR_1+R_1^2}{2R_1R_{mv}+R_tR_p}\right), \tag{8} \]
\[ E_2=U_{II}\left(1+\frac{R_tR_2+R_2^2}{2R_2R_p+R_tR_p}\right). \tag{9} \]
It is simpler to use the potentiometric method. From equations (2, 4, 6)
\[ E_1=E_{\mathrm{eqv}I}+E_{\mathrm{eqv}}\frac{R_1}{R_t}, \tag{10} \]
\[ E_2=E_{\mathrm{eqv}II}-E_{\mathrm{eqv}III}\frac{R_2}{R_t}. \tag{11} \]
In practice, everything turns out to be considerably simpler. Having made electrodes with equal diameters of the conical ends, we practically obtain that, in measurements, \(E_{\mathrm{eqv}I}=E_{\mathrm{eqv}II}\) and \(U_I=U_{II}\), while \(U_{III}=E_{\mathrm{eqv}III}=0\).
Therefore
\[ E=E_{\mathrm{eqv}I}=E_{\mathrm{eqv}II}, \tag{12} \]
\[ E=U_I K=U_{II}K, \tag{13} \]
\(\kappa\) is the correction that must be introduced for the presence in the circuit of the filters \(R_{\phi_1}\) and \(R_{\phi_2}\).
From the results obtained, a calibration curve is constructed.
In concluding the article, let us emphasize the necessity of precisely three, at minimum, measurements at each temperature step.
- \(U_{\mathrm{I}};\ E_{\mathrm{equiv\,I}}\) give the true value of the tool--workpiece thermoelectric power. 2. \(U_{\mathrm{II}};\ E_{\mathrm{equiv\,II}}\) give the value of the tool--workpiece thermoelectric power controlled by the reading of the platinum--platinum-rhodium thermocouple. 3. \(U_{\mathrm{III}} = E_{\mathrm{equiv\,III}} = 0\) indicates that the temperature at the heating points is identical.
If the temperature between junction 2 and junction 1 differs, then \(U_{\mathrm{III}} \ne 0\) and \(E_{\mathrm{equiv\,III}} \ne 0\). Such double control makes it possible to obtain highly reliable data. The entire calibration process takes 5--10 minutes. It is recommended to repeat the calibration 2--3 times in order to verify good reproducibility and agreement.
The advantages of the described calibration method are:
-
The speed of calibration.
-
The absence of parasitic emfs, which is due to the fact that only a small region of the contacting bodies is heated.
-
The possibility of carrying out calibration on a tool actually used in the cutting process: a hard-alloy insert, a drill, a reamer, etc.
References
-
M. I. Danielyan. Tool Wear and Thermal Phenomena in Metal Cutting. Mashgiz, 1954.
-
P. I. Bobrik. Surface-layer quality, temperature and force relationships in cutting certain grades of heat-resistant alloys. In: Machinability of Heat-Resistant and Titanium Alloys. Kuibyshev, 1962.
-
L. A. Bessonov. Theoretical Foundations of Electrical Engineering. Gosizdat, Higher School, 1942.