Measurement of Operating Conditions of Rolling Piston Type Rotary Compressors

EISUKE SAKURAI MAJOR APPLIANCE PRODUCTS ENGINEERING LABORATORY TOSHIBA CORPORATION 4-1, UKISHIMA-CHO KAWASAKI-KU, KAWASAKI 210 JAPAN A rolling piston type rotary compressor was modified for measurement of dynamic pressures in the compression and suction chambers, and instantaneous ~ngular position of the shaft. The compressor connected with a compressor load stand was tested at three different loads by changing discharge gas pressure and different running speeds by controlling line frequency between 35 to 75 herts. The analog data on pressures and angular position of the shaft was converted into digital data at the local site and processed at the remote site for plotting P-V diagrams and angular position shift curves. Indicaced gas work, frictional loss, and mechanical and indicated efficiencies were obtained with high accuracy, based on the P-V diagram. The results show that mechanical and indicated efficiencies at higher running speeds are less than at lower speeds, over operation speed between 35 to 75 hertz. The shaft does not rotate at constant speed during one cycle, particularly under operation at low frequency. The fluctuation of the speed of the shaft increases as the average shaft speed decreases. IHTHODUCTION Controlling cooling or heating capacity of an air conditioner with a variable-speed compressor becomes popular recently as one of the most effective ways to improve SEER of the air conditioner. As inexpensive, highly reliable power semiconductors have been developed, even household air conditioners begin to be equipped with frequency convertors using the semiconductors to control running speeds of induction motors in the refrigerant compressors. A rolling piston type rotary compressor combined with the frequency convertor may be suitable for a small-sized air conditoner requiring a wide control range in capacity because high volumetric efficiency may be kept constant for the rotary compressors, unlike reciprocating compressors, as the running 60 JAMES F. HA!vtiLTON, PROFESSOR THE R. W. HERRICK LABORATORIES PURDUE UNIVERSITY . \'ll~ST LAJ?AYETTE INDIANA 47907 USA speed becomes higher. J, rolling piston type rotary compressor designed for constant-speed operation was picked up and then modified for investigating operating conditions over wide speed range under various load. The compressor with a two-pole, single-phase induction motor was operated at-various. rotational' shaft speeds between 1880 to 4390 rpm by changing the input frequency between 35 to 75 herts. Three different loads were applied on the compressor by changing discharge gas pressure. _ Two kinds of measurements of the operating conditions will be presented mainly in this paper; measurement-of P-V (pressure-volume) diagrams, which is fundamental to analyze thermodynamic losses of the compressor_r, and measurement of shaft speed fluctuation in one cycle, which is related to the accuracy of the measureMent of P-V diagrams. EXPERIMENTAL PRINCIPLE3 P-V diagrams The measurement of P-V diagrams was done as precisely as possible. The following are basic concepts of the methods for measuring the diagrams. (1) measurement of instantaneous angular position of the shaft in rotation, not to be affected by the revolving flutter due to the change of the load in one cycle (2) accurate conversion of the angular position measured to the cylinder volume, depending on basic equations which express the geometric relationship between volume and angular position (3) accurate calculation of the area of the P-V diagram In order to make the methods above possible digital computers for sampling and processing data was used in the measurement system. The conversion from angular position to cylinder volume, and the calculation of the area of the P-V diagram were carried out by _using geometric equations [ 1 ] and trapezoidal rule respectively programmed in the computer.Instantaneous shaft speed · Instantaneous angular position'of the rotating shaft was measured by using two sensors for detecting reference angular position and countipg the number of reg~lar ang~lar intervals. There were two different ways to utilize the signal from the second sensor: for measuring P-V diagrams and instantaneous shaft speeds. While the cylinder pressures were sampled with an A-D convertor at every constant angular interval detected, for the P~V diagram, the signal of constant angular interval was sampled at every constant time interval for the measurement of instantaneous shaft speed. Instantaneous shaft speed was obtained by differentiating the data of angular interval with respect to time numerically with the computer. EXPEHIEENTAL JtPP AnATUS Test compressor An existing rolling piston type rotary compressor with a displacement volume of 10.3 cc per revolution was modified for measurement of the dynamic pressures in the compression and suction chambers, and the instantaneous angular position of the shaft. Cro?ssections of the compressor are shown in Figure 1 and 2. The existing shell of the compressor was replaced with the bolted shell des.igned for the installation of: transducers, a timing gear, cable connectors, etc. Operation of compressor The compressor for the test was connected to a hot gas load stand, similar to the one shown in reference f2l. The stand was equipped with an orifice system for measuring mass flow rate. Three different loads, as shown in Table 1, were applied on the compressor by changing the discharge gas pressure. The frequency of power source was changed with a frequency convertor v:hich consisted of an AC-DC convertor, a DC motor, and an AC ~enerator. The frequency range was limited to 35 to 75 herts due to the limitation of the frequency generated by the convertor while the compressor might run properly out of the frequency range. To make the induction motor of the compressor drive most efficiently at various frequencies, the voltage supplied was changed so as to have aconst;;mt value which is ratio of voltage to frequency. From some preliminary experiments, the ratio was determined to be maintained at 2.0 volt per hert. 61 Dynamic pressure measurement Three piezoelectric pressure transducers were used for measuri-ng--the· aynami"c pres~ sures in the cylinder as shown in Figure 4. Since the compression and suction processes occur simultaneously in different chambers in rolling piston type compressors, at least two pressure transdu~ers were required in the suction and compression chambers respectively to obtain P-Y diagrams. These two pressure transducers were mounted near the suction port and the discharg-e p'ort respectively. A third pressure transducer was mounted between these two s·o that it may measure the pressure of either the suction or compression process. The pressure measured with the third transducer was used as a reference pressure to deterr::2.ne relative pressure level between the two dynamic pressures. All of the three pressure transducers were recessed as shown in Figure 3. ~ith the recessed mount, the pressure at the point desired can be measured without interrupting other functions of the compressor. The recessed mount may cause gas dynamic due to the contained volume of the adapter system (3,4 ], contamination of the measurement by these dynamic effects is avoidable by proper adjustment of the contained volume. Calibration of the pre3sure transducers was carried out prior to installing them in the compressor. The calibration includes both the pressure sensitivity and the static Temperature sensitivity as discussed in reference [ ·5, 6 J • Angular position measurement ~ measurement system composed of two proximitor probes and a timing gear, ':Jas utilized to provide accurate data concerning the angular position of the shaft in the presence of rotational speed fluctuations. The arrangement of the probes and the gear on the upper side of the motor is sho~~ in Figure 1 and 5. A gear with 178 teeth was selected to obtain high resolution of the angular position. One of the two probes was set horizontally as close to the gear as possible so as to recognize each tooth clearly. The other one was set vertically as to detect a notch on the upper side of the gear. The notch location corresponded to the top dead center as a reference angular position. SIGNAL PROCESSING Signal processing for P-V diagrams The signal processing for P-V diagrams may be separated into two major ones: processing analog data to digital at alocal site, and processing the digital data to draw and integrate P-V diagrams at a remote.site. Ihe signal conditioning process at the local site is shown in Figure 6. The sampling interval of the pressure signal in the A-D convertor was controlled by an external clock terminal using the signal from the proximitor probe facing at the gear teeth. Since the number of the gear teeth was 178, each sampling for pressure should occur at every 2.02 deg •. The signal from the proximitor probe detecting the reference angular position of the top dead center was used as an external trigger source to start sampling. The wave form of the trigger source was also sampled after triggering occurred. The signal of the trigger source sampled was used for obtaining not only the initial angular position when the triggering took place, but also the average rotational speed of the shaft so that an internal clock worked to generate constant sampling time-interval for this sampling. Each signal was independently sampled, averaged over fifty times, and then punched out on paper tape. The paper tape then processed with a largescale digital computer. The flow chart for the processing is shown in Figure 7. The rotational speed of the shaft, n , is s 1 l T (i-j)llt ( 1 ) mllt + nl'lt From an interpolation of the assumed parabolic curve for the local voltage-time history around a peak shown in Figure 8, the two peak locations are found E(i-1)-E(i+l) m 5 2[E(~+lt-2E(1)+E(1-l)] (2) n 5 E(j-1)-E(j+l) 2 [E (j+l) -2E (j) +E (j-1)) (3) The initial angular position, Gi• at which the triggering occurs in advance of the first peak of the top dead center is then G. = 360llt (l-i-m) l · T (4) The angular position of each pressure sample specified with the number of sampling order, k (k.?:1), is 8 = 360 (k-ll + e. (5) NG 1 F·inally the volumes of the suction and compresslon chambers are determined as a function of the shaft angular position using the geometric equations in reference [1) • Determination of absolute pressure levels for the two major dynamic pressures in suction and compression chambers was carried out. as shown in Table 2, basing on the third dyn~~i~_pressure, PM, and static suction gas pressure,P. The ttlree areas in P-V diagram which corr~spond to effective gas work, we, 62 oversuction loss, ls, and overcompression loss, lc, are respectively "'rn ;co (~2(Pc-Pe)dVc + .(Pd-Pe)f83dvc (6) JtJ1 82 ·. 1 = J (P -P )dV (7) s e s s 81 1 = [ 83 (P -P )dV (8) c 8 c d c 2 where 8 1 : angular :position at which the compressin process starts (=32°) 82 : angular position at which Pc starts to exceed Pd e3 : an~;ular position at vlhich the discharge process finishes(=339°) During the transitional range from 83 to 81 no gas v!ork by the rotating motion of the shaft is necessary since the suction chamber is connected to the compression chamber. neexpansion process of the trapped gas in the additional volume including the volume of discharge port occurs during this :['ange, however, thermodynamic loss of the reexpansion is not defined an this paper because reexpansion loss may be obtained with not only the P-'ll diagram but also the data on thermodynamic mixing process in the single chamber and inverse flow through the snction port that were not measured. Therefore, indicated gas work, w., and indicated · efficiency, ?i' are d§fined as w. l '?i (9)