A Study on the Performance Improvement of a Rotary Compressor

Of many elements influencing the performance of a rotary compressor, the valve system plays an important role. In order to evaluate the performance, some basic experimental studies were made on the discharge coefficients and the proper dimensions of the valve system. Also, since the valve systems have previously been swdied by using rather simplified models, some effort' were made in this study to describe the valve system wilh more refined model. A valve stopper, two-plated valve and a valve port are modeled as contacting elements in finite clement analysis. Also initial configuration of the upper valve plate is considered. Results of the analysis were the deflection of a valve plate as a function of a rotation angle, mass flow history through the valve port, capacity o£ a compressor, the v"locity losses, compressing work and the quantity of back flow. 1. INl'RODUCTION In order to evaluate the performance of a rotary compressor, the valve system and compressing mechanism shO!Ild be considered.[! ,2 J The movement of a reed valve is in11 uenced by rna ny conditions: the valve port type, the shape of the valve plate, the height of the valve stopper and etc. In other words, the dynamic behavior of the reed vnlve is determined by its mass, stiffness, initial deflection and forces acting on it according to the behavior of a reed valve. [3] Some characteristics of valve ports of various shape.< for performance evaluation shcni!d be determined by experiments. Mass flow rate and forces on valve are determined by the port type. Also. two-plated valve has an advantage in scaling tightly the valve port before opening the valve because of the flexibility of a lower valve plate and the pushing force of the valve and port system should he determined in advance. The proper din1ensions of the valve port and stopper can be studied by the experiments on discharge coefficients under various conditions. In case of considering pressure loss in mmrressor valves, steady one-dimensional flow as in ductcu flow is usnally assumed.[5] Tn this experiment. the maximum velocity through a model valve port is 87m/s and so equations of incompressihle flow can be used. Because in this range there arc no difference between valves based on incomprcssihle r!ow theory and those based on compressible flow theory, which means that one may assume the densitv of the fluid be constant. Discharge coefficient Cd is defined by the ratio of the effective mass flow ·n1te ,;, and the theoretical mass flow rare 1i11h. 221 ,;,th = AJ2p(!1 Pz) A :port arm P 1 P 2 : pressure difference p : density 2.1. Experimental set-up and performance The real size of a valve port is so small that maoy difficulties are expected in e;<periments. Thus large models are made for easy manipulation. And air is used instead of refrigerant. TI1e prohloms arising due to the difference between a real model and a te.,ting model, can be handled bv sinlilitude. Reynold number Re is generally used. The kinematic viscosity of air at 20° C and ~t 1 bar is L5X10m!s and that of refrigerant at 125°C and at 12 bar is 0.3498Xl06m!s. However. in actu" a! compressors, the refrigerant is mixed with tbe lubricant. This make.' the refrigerant more v iscous. Transition from laminar and turbulent boWldary layer occurs at Rccrir =0.333X10 5 depending on general turbul7nt ~eve! in ~ncoming f~ow. Surface r~ughness of the testing model should not affe ct Recrlr. A schematlC dtagram ol the cxpenmental set-up JS shown m Ftg.l. E1 is made to consider the effect of a cylinder. ln real compressor, the valve port is p arr! y covered with the cylinder. E2 handles the distance between a valve port and a valve plate. The hun· dling equipment is a micrometer. E3 measures the mass flow rale. From the differance between the total pressure and the static pressure measured at one poim in the duct, one can calculate th e velocity at !hat point. In the same way, one can calculate at other points over the cross section a nd by in· tegrating velocity over the cross section, one can calculate the the mass flow rate. To analyz e the experimental data, a simple computer program was developed. The pressure is measured by the dig ital manometer which is connected with the AfD convertor to transfer the signal to the microcom puter. The signal is sampled 300 times at 0.5sec rale. These data are used after averaging. The pressure at each side of a valve parr is obtained in the same way. This di;charge coefficient, the averag e velocily, Reynold's number and things like !hat can be obtained. Table 1 shows dimensions of testing valve port models. ' 2.2 Discharge coefficient and proper shape 'The discharge coefficients correspooding to valve port types and distance between !he valve port and the valve plate arc shown in Fig.2, Fig.3, and Fig.4. The calculation of the theoretical mass flow rate was based on the port areas. Fig.2 shows that there exists optimal value of !he diamete r of contacting part (DCP). TI1e comparisons according to !he ratio of port area and port plate thickness are shown i,n Fig.3. Except the case that the thickness is so thin that the port plays like another orif ice, the thicker the valve plate is, rhe more the pressure loss is. Fig.4 shows the discharge coeffic ients as a fWlction of the radius of protrusion (ROP). The discharge coefficient of port 6 compared with th at of port 1 is lower when the distance is below 1.7mm and over 2.7mm. This is very disadvantageou s con· ditioo. During most of valve opening time, a valve plate contacts wim the valve stopper. Tllis indicates the discharge coefficient ncar the valve stopper height is important. And Bernoulli effect pla ys greater role in port 6 than in port 1 because port 6 has greater radius of protrusion. In the real compressor, a valve port is slightly covered with the guttered cylinder to &,charg e the Table.1 Dimensions of te.<ring valve port models