 |
||||
 |
 |
 |
 |
 |
 |
|
 |
The Kinetic Water Pump DIAPHRAGM MODELA way to greatly reduce steam condensation is illustrated in Figure 3. With this configuration, working fluids other than water can be used. It requires a flexible, stretchable material such as rubber to separate the working fluid from the water to be pumped. Most such materials are good thermal insulators. The system would be limited by the working temperature of the material. This design functions in a similar manner to the one shown in Figure 2. High-pressure steam or other vapor enters the Chamber in which there is a diaphragm sealed around the edges. The diaphragm is forced upward and causes water to flow into the Acceleration Tube. With a partial vacuum ahead of it, the water column accelerates and accumulates kinetic energy until it hits the Check Valve. CALCULATIONS OF PERFORMANCE Computer programs have been written to simulate the performance of the Kinetic Pump, taking into consideration tube diameter and length, water turbulence and drag on the walls of the tube, steam pressure and volume, and other parameters of different geometries of the system. The program keeps track of the enthalpy changes by interpolating in steam tables and uses this to calculate the overall efficiency. The Table 1 gives the theoretical performance values from the computer simulations. Each row is calculated for a set of input parameters for one kilogram of steam per cycle. (In an actual device, the quantity of steam per cycle would be considerably smaller). The pump pressure in each case in this table is 70 bars (1,015 psi), although the program is designed to run with any pump pressure. Note that the pump works well when the steam pressure is only 20 bars even though the system is pumping water to 70 bars. Efficiencies are higher for higher temperatures.
Figure 3. Diaphragm separation of working fluid and pumped liquid. Many of the details shown in Figure 2 are left out of this drawing.
Steam pressures are given in bars (example, 70 bars = 1,015 psi). The reason for the odd steam temperatures is that the computer calculations were made on the Kelvin scale (427 ° C = 700 K). Energy and Enthalpy are given in kilojoules (kJ). The final column gives the theoretical efficiency. In all of these cases, the pump pressure is 70 bars. The numbers are for one kilogram of steam for each cycle. The initial pressures and temperatures are those with which the steam enters Chamber A. The final pressures and temperatures are those of the steam after expansion when the water stops flowing. The “volume of water pumped” is the quantity of water in cubic meters pumped per cycle at 70 bars.
A number of parameters may be varied to get the desired results. Having a larger diameter Acceleration Tube allows the Tube to be shorter, provides lower velocity water flow, and offers less friction of the water against the wall. It also provides less heat loss to the wall. Making the Tube longer and having a greater length of water also gives lower water velocity. It should be recognized that the efficiencies listed in the last column are theoretical efficiencies. If we assume that the system will attain only 80% of that, the operational efficiencies would still be excellent. This number represents the energy of the pumped water compared to the heat input. If this were a turbine, the energy would yet have to go through a gearbox and a mechanical pump before the high-pressure water would be produced. EXPERIMENTAL RESULTS Since steam was not initially available, an experimental Kinetic Pump (called “Model 1”) was constructed that ran with compressed air. A small tank (called a “driver tank”) was mounted above the Acceleration Tube on the right side (above Chamber A of Figure 2). The driver tank was filled with compressed air, and the valve from the compressor was closed. A valve at the bottom of the tank was quickly opened to allow the compressed air to accelerate the water in the tube. The pressure in the Surge Tank was typically considerably higher than the pressure in the driver tank. With this simple system, the constant pressure portion of the water acceleration was missing, and only the adiabatic expansion portion was involved. About twice as much water would be pumped if both the constant pressure and adiabatic expansion were involved. The experimental results were surprisingly close to the computer calculations. Table II gives typical experimental results and a comparison with the computer results. Since this was a small system with only a 3/4-inch ID Acceleration Tube, about a third of the energy was lost to friction between the water and the tube wall (in both experiment and calculation). A larger diameter tube would reduce the loss considerably. Note that with only 20.6 psi in the driver tank, water was still pumped into the 100 psi Surge Tank. This table gives some experimental results of a small Kinetic Pump with a 3/4 inch inside diameter Acceleration Tube that was 137 inches long. Some units are English, and some are metric. The last column compares experiment with calculation. “Pumped Water” means the amount of water that was pumped into the high-pressure Surge Tank for each pulse. The pressure in the Surge Tank is listed as “Pump Pressure.” The “Free Run” gives the distance through which the water can accelerate before it strikes the Check Valve. The volume of the driver tank was 240 cubic centimeters.
We later constructed Model 2, which was larger and had an Accelerator Tube with an ID of 1.87 inches. Rather than having a flow of about 40 ml per pulse as in Model 1, Model 2 gave about 250 ml per pulse. Runs were made with 300-psi pressures in the Surge Tank, and the unit performed near calculational values. Our compressor would not go to higher pressures. Introduction | Part 1 | Part 2 | Part 3 | Part 4 |
|
|
 |