1、NASA TECHNICAL NOTE NASA TN D-3953 cr) - OI m d z c 4 rn 4 z (ACCESSION NUMBER) ITHRU) I f: 2 2 -t Lg e l ). (PAGES) (CATEGORY) (NASA CR OR TMX OR AD NUMBER) . 4 LOCAL HEAT-TRANSFER COEFFICIENTS AND STATIC PRESSURES FOR CONDENSATION OF HIGH-VELOCITY STEAM WITHIN A TUBE by Jack H. Goodykoontz und Rob
2、ert G. Dorsch Lewis Research Center Cleveland, Ohio NATIONAL AERONAUTICS AND SPACE ADMINISTRATION WASHINGTON, D. C. MAY 1967 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,- NASA TN D-3953 LOCAL HEAT-TRANSFER COEFFICIENTS AND STATIC PRESSURES FOR CON
3、DENSATION OF HIGH-VELOCITY STEAM WITHIN A TUBE I I I By Jack H. Goodykoontz and Robert G. Dorsch . Lewis Research Center Cleveland, Ohio . NATIONAL AERONAUTICS AND SPACE ADMINISTRATION For sale by the Clearinghouse for Federal Scientific and Technical Information Springfield, Virginia 22151 - CFSTI
4、price $3.00 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-LOCAL HEAT-TRANSFER COEFFICIENTS AND STATIC PRESSURES FOR CONDENSATION OF HIGH-VELOCITY STEAM WITHIN A TUBE by Jack H. Goodykoontz and Robert G. Dorsch Lewis Research Center SUMMARY Local he
5、at -transfer and static-pressure data were obtained for steam condensing in vertical downflow inside a tube. A 0.293-inch-inside-diameter by 8-foot-long copper tube was used as the test condenser. Inlet vapor velocities ranged from approximately 300 to 1000 feet per second with complete condensing o
6、ccurring in the test section. The condenser was cooled with water flowing countercurrently in an annulus around the con- denser tube. rate, and correlates with the product of local quality and the square of the test-fluid total mass velocity. In general, high values of the coefficient occurred at th
7、e inlet and decreased with length. The mean condensing heat-transfer coefficient varied from 3860 to 11 850 Btu per hour per square foot per OF over a test-fluid total mass velocity range from 64 900 to 336 000 pounds per hour per square foot. Static-pressure changes in the two-phase region varied f
8、rom a net increase of 1.32 pounds per square inch to a net decrease of 35.34 pounds per square inch. Axial static- pressure profiles were a function of the vapor flow rate and heat flux. Overall friction- pressure losses for the two-phase region were computed from measured static-pressure changes an
9、d correlated in terms of common pipe -friction parameters that included flow rate, total condensing length, and specific volume of the vapor at the condenser inlet. The local condensing heat-transfer coefficient is a function of the local vapor flow I NTRO DUCT ION As part of an overall program at t
10、he Lewis Research Center concerned with Rankine-cycle space -power systems, an experimental study of condensing inside tubes was initiated. Water was selected as the working fluid for simplicity of apparatus and instrumentation. A wide range of inlet vapor velocities (50 to 1000 ft/sec) was desired
11、in the studv to movide a sufficient range of variables for data analvsis and for comnari- Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-1 1 I son with theoretical results. (ratio of inertia to gravity forces) becomes large, and thus the flow patter
12、ns within the condenser should be similar to those in a zero-gravity space environment. lower end of the desired inlet vapor velocity range (65 to 232 ft/sec). Local heat- transfer coefficients were determined for steam flowing in vertical downflow within a 5/8-inch-inside-diameter stainless-steel t
13、ube. Because of the low inlet velocity, the pressure drop was negligible during these tests. In the investigation reported herein condensing data were obtained with high inlet vapor velocities (300 to 1000 ft/sec). Both local heat-transfer and local static-pressure data were obtained for steam conde
14、nsing inside a tube. The data were taken with the downstream vapor-liquid interface located within the condenser tube so that complete condensing occurred. With complete condensing, the overall friction-pressure drop for the condenser could be computed with reasonable accuracy from the measured pres
15、sure change. mounted vertically with vapor flowing downward. The condenser was cooled by water flowing upward (countercurrently) in an annulus around the tube. The range of variables employed was as follows (symbols are defined in appendix A): Further, at high vapor velocities the Froude number -I I
16、n the first phase of the study reported in reference 1, data were obtained at the The test condenser was a 0.293-inch-inside-diameter by 8-foot-long copper tube I Variable I Test-fluid flow rate, wt, lb/hr Inlet vapor pressure, Pi, psia Inlet vapor velocity, vvi, ft/sec Condensing length, Lc, ft Coo
17、lant flow rate, wk, Ib/hr Coolant temperature, tk, OF Inlet Exit Inlet vapor qualities Range 30.5 to 158 15.03 to 39. 34 313 to 1018 1.1 to 6.7 405 to 2180 61 to 99 94 to 206 APPARATUS AND PROCEDURE Description of Rig A schematic drawing of the test facility is shown in figure 1. The test-fluid side
18、 of the facility was a once-through system using demineralized and deaerated water. The coolant loop used demineralized water that was continuously recirculated. Steam at 100 pounds per square inch gage was used as the heat source, and cooling-tower water 2 Provided by IHSNot for ResaleNo reproducti
19、on or networking permitted without license from IHS-,-,-Remotely operated pressure regulator - vapor entered at the top and coolant flowed countercur- rently in the annulus between the inner and outer tubes. The inner tube was a thick-wall copper pipe with a measured outside diameter of 0.541 inch,
20、a measured inside diameter of 0.293 inch, and a total condensing length of 8 feet. The outer jacket was a copper tube with a 0.7Sinch outside diameter and a 0.040-inch wall. The annular gap between the inner and outer tubes was 0.064 inch. The small gap induced high coolant velocities at low coolant
21、 mass flow rates. This combination provided a large coolant temperature change per unit length of annulus so that accurate local heat flux determinations could be made (appendix B). Thick-wall tubing was used for the condenser so that the junctions of the wall thermocouples could be deeply embedded.
22、 A bellows at the downstream end of the condenser allowed relative motion between the inner and outer tubes. Spacer pins in the annulus maintained concentricity. The inner diameter of the inlet vapor line changed from 1.049 to 0.293 inch at a distance of 18. 5 inches upstream of the condenser. A bel
23、l-shaped transition section accommodated the change in diameter. A stainless-steel ring with an inner diameter of 0.293 inch was placed at the inlet of the condenser (fig. 2) to reduce axial heat conduc- tion in the thick-wall copper tube. The test section, as well as all vapor lines, was lagged wit
24、h blanket insulation to minimize heat losses. location of the vapor-liquid interface could be varied by throttling the condensate. A 0.5-inch-outside-diameter glass tube (fig. 1) was installed between the end of the con- denser tube and the condensate throttle valve to allow visual observance of the
25、 interface when it was moving into or out of the condenser. - A condensate flow control valve was installed downstream of the condenser so the Instrumentation The condenser was provided with instrumentation to measure vapor inlet tempera- ture, condensate exit temperature, condenser tube-wall temper
26、atures, coolant temper- atures, vapor inlet pressure, static pressures inside the condenser at different axial positions, and condensate and coolant flow rates. Figure 3 and tables I and I1 show the 6 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-l
27、ocation of the temperature and pressure measuring stations with respect to the inlet of Temperatures were measured with iron-constantan thermocouples with the thermo- couple wires insulated with magnesium oxide and swaged inside a 1/16-inch-outside- diameter stainless-steel tube. The vapor inlet tem
28、perature was measured upstream of the transition section in the 1.049-inch-inside -diameter tube, where dynamic effects of flowing vapor would be negligible. Condensate temperature was measured 6 inches downstream of the end of the condenser. longitudinal plane after an initial spacing of the first
29、two thermocouples of 0.14 and 0.98 foot downstream of the condenser inlet (table 11). The bare physical junction of the thermocouple was placed 0.034 inch from the inner surface of the condenser tube. A copper disk, with a hole through the center, was soldered to the end of the thermocouple sheath t
30、o aid in installation. The physical junction of the thermocouple was flush with the surface of the disk (fig. 3). The location of the measured temperature was arbitrar- ily taken as the location of the physical junction for reasons discussed in appendix B. The 1/16 -inch-outside-diameter sheath of t
31、he thermocouple projected radially outward. A bellows between the probe and the outer jacket was used to accommodate relative motion between the inner and outer tubes. selected for the cooling passage in order to obtain accurate local heat flux data. Because of this small gap and the desirability of
32、 deeply embedding the thermocouple junction in the tube wall, it was not possible to run the sheathed leads axially or to wrap them around the tube. The radial installation employed had the disadvantage of providing a possible fin effect. The magnitude of the fin effect of the cylindrical sheath on
33、the measured wall tem- peratures was investigated analytically. (The lead wires were thermally insulated from the sheath.) The rate of withdrawal of heat from the tube wall by conduction through the stainless-steel sheath immersed in the coolant stream (ref. 2, pp. 25-29) was compared with the conve
34、ctive-heat-transfer rate for an equal surface area of the unmodified copper tube wall. The calculated ratio of the two heat removal rates was near unity for the range of test conditions employed. In view of this result and the fact that the thermo- couple junction was embedded deeply in the copper w
35、all, it was concluded that the fin effect did not have an important influence on the measured wall temperatures. As an additional check on the magnitude of the fin effect, a wall temperature error analysis was made by using the method of reference 3. The method of reference 3 con- siders the thermoc
36、ouple as a heat sink in a plate that is exposed to fluids on both sides at different temperatures. Equations are developed to give the temperature error as a function of the convective heat-transfer coefficients, thermal conductances of the plate , the condenser. Condenser tube-wall temperatures wer
37、e measured at 4-inch intervals in one This radial wall thermocouple installation was dictated by the 1/16-inch annular gap 7 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-and thermocouple wires, and temperature differences between the plate and the
38、 fluid. Application of the technique of reference 3 indicated that the error in measured wall temperature would be less than 1.0 F. It was concluded, therefore, that any temper- ature error caused by fin effects on the thermocouples would be small in comparison with other sources of error. 1 foot wi
39、th the exception of the inlet and exit region. The exact axial location is shown in figure 3 and table I. The plane of the coolant thermocouples was displaced 90 from the plane of the condenser tube-wall thermocouples, and the junctions were located at the center of the annular gap. Pressures were m
40、easured with mercury-filled U-tube manometers with a hydro- static water leg on the condenser side and atmospheric pressure on the reference side. The connecting lines from the manometer to the pressure tap on the condenser consisted of transparent tubing so that visual observation would give assura
41、nce of an all-liquid hydrostatic head. Static pressures were measured at 16 axial positions from upstream of the transition section to the end of the condenser. The locations of the positions are shown in figure 3 and table I. The static pressure of the inlet vapor was measured at four positions ups
42、tream of the condenser. One static pressure tap was located in the 1.049-inch-inside-diameter tube upstream of the transition section. The other three were axially spaced in the small-bore connecting line between the transition section and the condenser. Condensate flow rate was measured by using a
43、modified weigh tank technique, which consisted of measuring the time required to fill a known volume. The temperature of the condensate was measured so that the flow rate could be evaluated in mass units. The measuring station consisted of a quick-shutoff valve positioned downstream of a 2 -inch out
44、side-diameter by 3-foot-long metal tube. The tube was equipped with a sight glass that allowed visual observation and timing of the motion of the liquid interface when the quick-shutoff valve was closed. It was assumed that the flow rate did not change when the quick-shutoff valve was closed, since
45、neither the pressure in the condenser nor the pressure drop across the condensate flow control valve changed. The coolant flow rate was measured by a commercial turbine-type flowmeter. Coolant temperatures (fig. 3) were measured in one axial plane at a spacing of Test P roc ed u re Initial startup o
46、f the facility consisted of elimination of noncondensibles from the system. The boiler was filled with demineralized water and isolated from the remaining part of the system, which was evacuated to a vacuum of 28 or 29 inches of mercury. Ten to fifteen percent (6 to 9 gal) of the original boiler inv
47、entory was then boiled off to the 8 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-. atmosphere to rid the water of any dissolved or entrained gas. The air content of the water in the evaporator was reduced to an order of 2 parts per million, on a m
48、ass basis, as determined by a gas analyzer. The boiler was then opened to the system, and with the condensate flow control valve fully open, high-velocity vapor flowed through the system to the receiver tank and out the vacuum line to purge any residual air pockets. The vacuum line was then closed.
49、The pressure level in the receiver varied between 26 and 29 inches of mercury vacuum, depending on the temperature of the condensate. For each run, however, the receiver-tank pressure was constant. Establishment of test conditions consisted of setting a coolant flow rate and building steam pressure inside the coils of the boiler and throttling the outflow of the c
