Frigichips testing for the heat flow measurements

Size: px

Start display at page:

Download "Frigichips testing for the heat flow measurements"

Pearl Wiggins
5 years ago
Views:

1 Frigichips testing for the heat flow measurements V. Bahjil, S. DubniCka Department of Physics and Applied Mechanics Technical University in Zvolen, Slovakia Abstract The present paper is devoted to the critical analysis of the individual heat flow measurement methods. These methods are mutually compared fiom the point of their practical use. The special and new method based on the fi-igichips' performances is presented and described. 1 Introduction The energy crisis impact us stronger and stronger from year to year. The prices of energies climb higher and higher. This is the reason to figure the dissipation energy reduce investments. One of them which directly touches not only people but all the alive on this planet is the thermal energy. The legislation of many countries of the world has dealt with this topic. The standards dictate the minimal thermal resistance (R[~~K/w]), which should be met in the building industry. In the manufacturing sphere there are being sought the new technologies, which will provide not only the higher quality and utility of the product or production, but the economic production too. The heat losses are the crucial ones. To determine the heat flow density value (q[~m-2]) there are used the measuring devices based on very different principles. But all of them have one common property, i.e. they are not able to measure the q directly but the quantities which relate it more or less directly. Simply the heat flow is transformed into electric current which we are able to measure very easily. In our research we have obtained some very considerable results. See for example Bahyl and MarEok 1995, Kotrik 1997 and DubniCka Of course we are not the first who intend to use the fi-igichip as the heat flow measuring device. See for example the paper by Dittmann and Schneider 1992.

2 1.1 Thermoelectric voltage Let us deal with the problem from its physical essence and not from the side of its technical usage. There are two basic thermoelectric events here - the Peltier and the Seebeck effect. It is possible to rise up the thermoelectric voltage 5 in the closed electric circuit which is realized with two different metal wires l and 2, if for the contact temperatures of the metals is true T, > Tb. This thennoelectric voltage is given as: 4 = a.(t,, -T,), where, k I? = -.ln' (1) e I202 and a is the value which defines the metals contact performances, k is the Boltzmann constant e is the electron charge no,, noz -the amount of electrons in the volume unit of the metals. This effect relates to the potential well difference and to the different value of the Fermi energy for the conduction electrons in the different metals [2]. The usage of this event in the measuring accuracy technique and in the adjusting technique is very large. Figure 1 : The thermoelectric voltage [l] 1.2 The Seebeck phenomenon If we change the arrangement in Fig. 1 in such a way that we include the voltmeter in it, we will measure zero voltage. This is caused by the fact that in the connecting points of the metals is rising the voltage but it is of the same value and of opposite direction. Of course this is true only if the metal contacts "a" and "b" are of the same temperature. In the case of different contact temperatures, the contact electromotive voltage will still have the opposite sign but a different value. The result measured voltage is named thermoelectric voltage 5 and it is given as the difference between the contact voltages. This phenomenon is named after the German physicist Seebeck - the Seebeck phenomenon [2].

3 \ 1 Figure 2: The Seebeck phenomenon [2] 1.3 The Peltier effect The opposite of the Seebeck phenomenon is the Peltier event. If we replace the voltmeter in Fig. 2 by the direct current voltage source, we will observe the changes of temperature on the contacts "a" and "b". The one contact will be act as the heat source for its surroundings and the other will act as the heat sink for its surroundings [3]. This phenomenon is widely used in the freon less cooling systems in the cosmic technologies and in the electrical engineering. 1.4 The thermoelectric converters [4] The thermoelectric converters are such kinds of devices which are used in the direct change of the thermal energy in the electric one and back. As the base of them there are two semiconductor posts (See Fig. 3.). One of them is of the N type and the second one is of the P type and they are connected with the well conductive (electric and thermal) metal shim. The heat flow which is going through the semiconductors will cause the change in the charge carriers concentration. The holes concentrate on the warmer side in the P type semiconductor. This caused the rise of the so-called space charges and of course the potential difference V,. Similar is true for the N type semiconductor but in this case the free electrons are taking place here instead of the holes. The different concentration of them will cause the potential V,. The resultant voltage of such pair is given as the potential addition U = V, + V,. The efficiency of one element is defined as the ratio of the electric power generated with the element and of the heat flow Q, which is flowing through the element: 'Ic = (0,-@,)l@, 0,,=(0,+0,)12 ["K] z = a' I(p,y) [K-']

4 650 Compututior~ul Methods and E.~per.imel~tal Measur.es a,b - Seebeck factor [vk'] 0, - the hot side temperature -the cool side temperature e2 qc - the Carnot efficiency 0," - the mean temperature ["C] p - specific electric resistance [Q.m] X - the heat conductivity coefficient [~m-'k'] According to the operating temperature we divide the Peltier elements into three groups: the low temperature ones K the middle temperature ones K the high temperature ones K. The assembly schedule for the individual elements is dependent upon this division. The low temperature of thermoelectric converters have very simple scheme of assembly. (See Fig. 4a.) The mean and high temperature converters have the cascade type of the assembly scheme (See Fig. 4b.) or they can have the segmented branching (See the figure 4c) too. heot flow density W J J cooling Figure 3. The basic scheme for the Peltier element.

5 Figure 4a,b,c. The variety of the Peltier elements assembly schemes. The production of the thermoelectric converters includes of course very many problems which should be overcome. One of them, of the crucial importance is the excellent electrical contact between metals and semiconductors or the contact to the end electrical outlet. Next it is the anti corrosion protection of the semiconductors' surface under the atmospheric or high temperature conditions. For more information on this question we recommend for example the Handbook of the high voltage electronics (Hehnan 1984). 2 Our experiments We have realized the set of experiments in the thermotechnical laboratory of the Technical university in Zvolen. In our experiments we have been concentrated before all on the problem of the calibration of our converters i.e. we should determine the hnction = f (P), where P is the power of the generated electric current. We have devoted great amount of interest to determination of the efficiency of the fiigichip as a hot engine. I.e. we paid our attention to the quite opposite event of the frigichip as it is in common use. We decided to R.I' experimentally determine the value of the efficiency?l = -, where R is the Q electric resistance (see Fig. 6) and I is the electric circuit generated with the heat Q. All our experiments we have realised in the steady state conditions and we have used quite classical experimental arrangement. The heat flow line havs been

6 represented with the set of aluminium joists between which the frigichips and the control thermocouples were placed. The calorimeter loaded with melting ice has been used as a heat sink. The temperatures have been measured in 8 different places. This enables us to determine the heat flow losses in the different AI joists on the way of heat from the room conditions to the heat sink in the calorimeter. For the reference temperature point we have used the calorimeter with zero Celsius degree. Our measuring device is connected with the personal computer and it is hlly automated. The measured values have been gathered every 30 seconds. The scheme of our computer aided the working place as in the Fig. 5. The scheme of the Peltier element circuit is in the Fig. 6. The determination of the value of the loading resistor has been the very important task of our work. The results of our measurements are given in the figure 7. It is easy to see that we get the maximal performance of our "converters" we got for the loading resistor RLR = 3.9R. The resistance of ampere-meter is RA = 10.5 R. In common we get R = 14.4 R. As we can see in the Fig. 7, the dependence of the Peltier element performance on the loading resistor has well-defined maximum. Figure 5. The scheme of the data gathering computer aided system. L Figure 6. The Peltier element circuit connection.

7 For Fig. 5: For Fig. 6: T 1 -T2 - thermoelements TEM(PC) - thermoelectric U 1 -U4 - terminal on measurement of the voltage changer - Peltier element P.M.M - switch of the measuring spoil V - voltmeter VM - voltmeter A - ampere-meter PP - personal computer Rz - loading resistance 3 The results We have realized 12 independent experiments, in which we were looking for the dependence of the performance of the studied Peltier element (PE) upon the heat flowing through it. As it is in the common use, first we determined the function U = f ( ~ t in ) which U is the PE voltage and At is the temperature difference between the PE plates. We have obtained the equation U = (0.022 f 0.006).At + ( f )[V]. (7) As the q value in the regression equation (1) equals and from this we can conclude that this value is statistically equal to zero we can take our results as the precise ones. (The value of q should be zero from the theoretical point of view.) The next proof of our preciseness we can get from the thermal conductivity coefficient h. According to our computations and measurements we have obtained the value A,, = 1.99 f 0.65[~m-'.~-']. (8) As it is impossible to find this value in the scientific papers, we ask for it the Melcor Peltier Corporation directly. Via we have obtained the next form k = (ko+kl*t,, +k2*~~:,).10~[~crn-'.k'~], (9) where : ko = ;kl = ;k2 = ];TA, = 0.5 * (T, - T,) After substitution of our measured values into the form (3) we obtained /2'~F=1.712k0.003 [w~-'.k-']. (10) We take this as a very good agreement with our measurements and as the direct proof of our correctness. The efficiency of the PE is in comparison with their efficiency in the cooling systems very low. In the conditions of the loading resistance 14.4 R we have obtained the next value vpe = S f ~. (1 1) This value we do not find as a definite one, because of the fact that from our theoretical computations in accordance with Heimann at all we obtained the value 770pT = ' ~. (12) Our experimental device did not allow us to measure under the conditions of rather great amount of the heat flow. So we decided to install our measuring

8 654 Conip~ttatio~zal Methods and Esperimental Measures device into the wood drying kiln. Our results are depicted in the Fig. 7. We can conclude that the electric power performance of the PE is the linear function of the heat flow through the ~eltier 'lernent. 4 The graphic representation of some results The dependence of the electric power on the size of the load resistance. Figure 7. The dependence of the electric power as the function of the loading resistor. A dependence of the electric power on the temperature difference. 1 Q l Tx-Ty PC] l Figure 8. The Peltier element graduation function (The resultant voltage as the function of the temperature difference)

5 Conclusions As it has been mentioned at the beginning of the present paper we have been dealing with the possibility to use the PE as the direct heat flow measuring device for rather long time.

9 5 Conclusions As it has been mentioned at the beginning of the present paper we have been dealing with the possibility to use the PE as the direct heat flow measuring device for rather long time. It is quite sure that PE is possible to use as the basic element in the heat flow measuring device. We think that it will not take a long time till we are able to meet it in the technical practice. Moreover we regard them much more precise than the classical ones, fore example the Schmidt carpets. So if we wish fully utilise their sensitivity, we need more precise experimental device and we should realise more experiments. But what we have done till now serves as the clear proof that we are on the true way. From the point of view of the practice it is important to solve the question of the so called self calibrating system for the PE connected as the heat flow measuring device i. e. in the sense of the Seebeck phenomenon. Normally the PE is used as the heat source or sink part of many modem systems i. e. in the sense of the Peltier phenomenon. The calibrating of the heat flow measuring PE is possible in the obvious way, by usage of the well heat conducting externally heated thin plates. We suppose that such an arrangement will allow us to determine the straight line from Fig. 7 in much more uniformly distributed data. The thermal contact of the PE with the measured object plays very important role in the heat flow measurements. Other way we can introduce rather large systematic errors in the results. So it is inevitable to use the thin layer of the thermo-paste to be sure that the thermal contact is good. Of course such a kind of past have been used in our experiments too. 6 Acknowledgement This paper has been done under the support of the Scientific and technical project No. VTP 2903 which is highly appreciated by the authors. 7 References [l] Bahjd V, MarEok M, 1995 The thermal conductivity of the frigichips for using in the heat flow measurements in the vacuum drier. In: Vacuum Drying of Wood '95, Slovak Republik, High Tatras, October 8-12, 1995, TU Zvolen, 1995, pp [2] Kotrik E, 1997 The Utilization of Frigichips for Measurement of Heat Flows and Thermophysical Characteristics of Materials and Constructions. In Zbornik 7. International conference, 5 Section, Budova a energia 2., KoSice , SF TU KoSice, 1997, [3] DubniEka 5; 1999 The Heat Flow Measurement by the Usage of the Peltier Elements, in Proc. "NovC smery vo erobnych technologi8ch 2000n, jun 2000, PreSov [4] Dittmann H, Schneider V. B, 1992 Ein Mefigerat fir den Warmestorm. Physik und Didaktik, 3, [S] Heiinan J at all., 1984 PfiruEka silnoproude elektroniky, Praha SNTL 780 p.

FRIGICHIPS TESTING FOR THE HEAT FLOW MEASUREMENTS

FRIGICHIPS TESTING FOR THE HEAT FLOW MEASUREMENTS Vladimír Bahýl, Štefan Dubnička Department of Physics and Applied Mechanics, Technical University in Zvolen, T. G. Masaryka 4, SK 96 53 Zvolen, Slovakia