Subject: Space-tech Digest #155 Contents: Pressurization pumps (2 msgs) Isp calculation program available (1 msg) Fiberglass (5 msgs) ------------------------------------------------------------ Date: Mon, 28 Jun 93 10:53 PDT To: space-tech@cs.cmu.edu Subject: Ejector pumps for liquid helium From: Bruce_Dunn@mindlink.bc.ca (Bruce Dunn) > Paul Dietz writes: > If you are going to put a gas generator into the helium system, > how about putting it downstream of the helium tank, where it could > feed an ejector pump? I don't know how well such pumps would work > on a cryogen like helium (likely not too well), but if it did work > you could get by with a low pressure He tank. The ejector pump could > be combined with the gas generator that you already imagine will > heat the helium. In the posted design, using a liquid bipropellant gas generator, the generator/helium vaporizer is already downstream of the liquid helium tank. Liquid helium from the tank is mixed with the gases from the generator. If this were arranged correctly, there would probably be some pumping effect (ie. a net gain in pressure from the liquid helium entering the generator to the gaseous helium/carbon dioxide/water stream exiting the generator). I have not relied however on this potential pumping effect, and have conservatively assumed that the liquid helium entering the vaporizer must be at a higher pressure than the output of the generator. This requires the provision of a separate system to generate hot helium gas to pressurize the liquid helium tank (a gaseous helium bottle, and a heat exchanger in the gas generator). If some ejector pumping effect could be achieved by the gas generator, then this could lead to a slightly simplified design. The liquid helium would enter the gas generator and be vaporized and pumped to a higher pressure by the bipropellant hot gas flow. A small proportion of the generated gaseous helium would be diverted back to pressurize the liquid helium tank. Starting with a modest tank pressure, the system could bootstrap itself up to whatever pressure is desired. The ejector effect would only have to supply a small differential pressure (perhaps 0.1 or 0.2 MPa), not the total output pressure (4 MPa). This would eliminate the gaseous helium bottle and the heat exchanger. The mass saving is small, but the greater benefit would be in simplification of design. Such a scheme would be worth looking into further, but probably should not be assumed in a proof-of-concept design. -- Bruce Dunn Vancouver, Canada Bruce_Dunn@mindlink.bc.ca ------------------------------ Date: Mon, 28 Jun 93 20:30:23 -0400 From: dietz@cs.rochester.edu To: space-tech@cs.cmu.edu Subject: jet pumps I was reading a bit about jet pumps for rockets: D. G. Elliot, Investigation of a Gas-Driven Jet Pump for Rocket Engines (in Progress in Astronautics and Rocketry Vol 2, pages 497-541. Am. Roc. Soc., 1960) Investigation of jet pumps for rockets go back to at least 1936. Elliot claimed to have a concept that would require a few times the gas of gas-generator cycle turbopumps, at a similar pump mass. All the pumps used in rockets have the same very general form: fluid is accelerated to high speed, then the dynamic pressure of the fluid flow (1/2 rho v^2) is converted into static pressure in a diffuser. In centrugal pumps, this is done by spinning the fluid out from the axis by a rapidly moving impeller. Jet pumps exploit the entrainment of one fluid by another. Aside from valves, the only moving parts are the fluids passing through the pump. No high speed rotating machinery is required. Eliiot's scheme was tested with water and air, with so-so results; components were tested that improved on some of the problems. I don't know if the idea was pursued further. His idea is to accelerate a liquid, the driver liquid, with a compressed gas. The acceleration is done in a nozzle with an intimately mixed gas/liquid mixture. This works best if the liquid has a low vapor pressure and the gas a low molecular weight and low condensation point. The liquid is then directed onto a surface on which a low pressure liquid is flowing. The impingement mixes the fluids, entraining the low pressure fluid, and causing the gas to separate. Centrifugal effects complete the separation, and the fluid is directed into a diffuser where it is slowed and pressurized. The gas, now at low pressure, is either directed out an exhaust line, or first directed back into the low pressure propellant tank so that any entrained propellant droplets can be recycled (incidently pressurizing the tank). In Elliot's scheme some of the output of the pump is recycled to act as the driver liquid. Typically about 40-50% of the liquid gets recycled. One could also, I suppose, use a separate pressurized liquid tank for the driver liquid; the remaining 50-60% of the propellant would come from a low pressure tank. This would reduce the tank mass by on the rough order of 50% over a purely pressure fed design. As mentioned, the scheme is best for propellants with low vapor pressure, to reduce losses of propellant in the gas. Fortunately, hydrogen peroxide has very low vapor pressure; he estimates a vapor loss of < 0.2% at a discharge pressure of 600 psia using decomposed hydrazine as the driver gas. JP-5 and hydrazine are similar low vapor pressure fuels. He also outlines a scheme in which the gas is directed into the thrust chamber rather than being exhausted. This increases both the gas flow and driver fluid requirements, and would only work with low vapor pressure propellants like peroxide. Paul F. Dietz dietz@cs.rochester.edu ------------------------------ Date: Mon, 28 Jun 93 19:35 PDT To: space-tech@cs.cmu.edu Subject: Isp program From: Bruce_Dunn@mindlink.bc.ca (Bruce Dunn) I have prepared a help file for the Isp program which I obtained from Mitchell Burnside Clapp. The program does not come with its own help file, and is somewhat confusing to use as a result. I would appreciate comments on the clarity of the file and on its technical content. This posting will also act to give people an idea of what the program can do. I can arrange to send the program by surface mail to anyone E-mailing me a postal address. George Herbert has stated that he may be able to place the file on an FTP site as well. Help file follows: xxxxxxxxxxxxxxxxxxxxx The files in this package comprise a program for calculating the specific impulse of rocket engines. The program assumes that users have basic textbook knowledge of the meanings of various terms involved in the characterization of the performance of rockets. This program requires an MS-DOS computer which has a math co-processor. For those without access to a machine with a math co-processor, it may be noted that the program has been found to run with at least one shareware software co-processor emulator, namely Q387, Version 3.3 by QuickWare. The files required for the program to run are as follows: allprop.dat data on rocket propellants inp.exe a utility program which is used to specify the propellants, chamber pressure etc. isp.exe the program which actually calculates the Isp thergg.dat a file containing needed data therii.dat a file containing needed data therss.dat a file containing needed data not needed for execution are: hlp.doc contains a copy of help screens appearing in inp.exe and isp.exe readme.txt this file, written by a user of the program to help others in tackling it the first time isp.exe is run, it will create two additional files: output.dat the output from the program, readable with a text editor tabout.dat a binary file containing (presumably) information on the arrangement of data in the output file each time the isp.exe is run, output.dat and tabout.dat are overwritten by new versions To calculate the Isp of a rocket engine, proceed as follows: 1) Run inp.exe. You will get a page of instructions, followed by a series of menus which allow you to enter data on the propellants to be burned in your engine, their proportions (either pre-determined by you, or via instructions for the "optimizer" which systematically tries different propellant proportions), the chamber pressure, and the expansion ratio(s) or exhaust pressure(s) which you wish to have the Isp calculated for. When you specify propellants, their characteristics (density, MW, heat of formation etc.) are looked up from the data stored in the propellant library (allprop.dat). Other menu items allow you to change how the program does the calculations, and what the output looks like [it is suggested that initially these be left at their default values]. Once all the information is determined, you select the menu item which allows you to exit inp.exe. The program will then save the relevant information in a file, using a file name provided by you. 2) Run isp.exe. The program will prompt you for the name of the file which you created in step 1. Once this is given, isp.exe will carry out the calculations. The results will flash on the screen, usually too quickly to been seen. Don't panic, the program also writes the information to the file output.dat. 3) Read the file output.dat with your favorite word processor or text editor. The output tables are created with spaces, not tabs. Therefore, to make sense of the tabulated data, it is essential to use a fixed width font. The output table can also sometimes use long line lengths, and when using a word processor to view the file, individual data lines will wrap onto new lines, destroying the formatting of the tables. In this case, either select a smaller font or a larger paper size, such as legal paper in landscape mode, to prevent line wrapping. 4) If you are not happy with the results in the tables, you don't need to create a new input file from scratch. Run inp.exe again and select the menu item indicating that you wish to create a new input file. The program will then give you the option of basing the file on a previous one. Once the input file has been modified, it can be saved under the old name or under a new name. The following points should be noted. - The program is not written for SI units, and uses a grab-bag of units. For example, energy is in calories rather than joules, temperature is Kelvin, and pressures are in psi. - The program calculates the conditions of the chamber and throat, then calculates the conditions and Isp for any specified expansion conditions. Expansion conditions may be given either by specifying the exit pressure, or by specifying the nozzle area expansion. You can specify more than one expansion condition; for example in a single run you can calculate the Isp for exit pressures of 30 and 15 psi, and area expansion ratios of 4, 5 and 6. - The program calculates two Isp values: "Isp (optimum)" [which is listed as just "Isp" in the output tables] and "Isp (vacuum)". Isp (optimum) is the Isp for an engine in which the gas pressure at the plane of the exhaust is exactly equal to the ambient atmospheric pressure. If you ask to have Isp calculated for an exit pressure of 14.7 psi (sea level atmospheric pressure), then the calculation will determine what nozzle size is necessary to expand the gas to 14.7 psi. The output table will give "Isp" which is the Isp for that engine expanding the gas to 14.7 psi, and exhausting the gas against an outside atmospheric pressure of 14.7 psi. The output table will also give "Isp (vacuum)" which is the Isp that the same engine would have if operating in a vacuum. - The program does not directly calculate the performance of under or over expanded engines (where the calculated pressure at the exhaust plane is lower or higher than ambient atmospheric pressure). There is sufficient information however in the output tables to be able to estimate Isp for such conditions (see below). - The output of the program is the traditional rocket engineers "Isp", or specific impulse. The units of Isp are seconds, and due to the way that a pound of force and a pound of mass are defined in the U.S. measurement system, the unit includes the reciprocal of the acceleration of gravity at the earth's surface. To convert Isp into SI units, multiply Isp by the acceleration of gravity, 9.8 m/s^2, to get exhaust velocity in m/s. An Isp of 300 s thus becomes 300*9.8 = 2940 m/s. Through unit conversion, the same 2940 m/s is also equal to 2940 Nm/kg. In other words, 1 kilogram of propellant will give a thrust of 2940 newtons for 1 second, or 1 newton for 2940 seconds etc. - The program calculates a theoretical Isp by the "shifting equilibrium" method, which assumes that the exhaust species are constantly coming to chemical equilibrium as they expand. This is overly optimistic, in that in practical engines the exhaust process is so quick that some energy releasing processes in the exhaust aren't fast enough, and the actual exhaust products are not in chemical equilibrium. The calculation process also ignores a variety of losses occurring in real world engines such as incomplete combustion, nozzle friction and the use of propellants for purposes other than thrust (such as film cooling, powering turbopumps and providing tank ullage gas). In practice real world Isp of most engines ranges from about 90 to 95% of the calculated theoretical shifting equilibrium Isp. The following table compares the theoretical results of calculations for a number of Rocketdyne engines, with performance data taken from some older Rocketdyne specification sheets (NB. engines are modified from time to time - do not compare the specifications of these engines with those of current versions of the engines). Engine Condition MR Chamber E Real Isp Calc. Isp Fraction F1 sea level 2.27 982 16 265 290.7 0.911 RS-27 sea level 2.24 702 8 262.5 287.3 0.912 H1 sea level 2.23 700 8 263 287.0 0.916 MA5-boost sea level 2.25 639 8 259.1 284.1 0.911 MA5-sust. sea level 2.27 735 25 220.4 251.1 0.877 J2 vacuum 5.5 763 27.5 425 446.2 0.952 SSME vacuum 6 3260 77.5 453.5 465.2 0.974 MR = oxidizer/fuel mixture ratio Chamber = chamber pressure in psi E = nozzle area expansion ratio Real Isp = Isp from Rocketdyne data sheet Calc. Isp = Isp calculated with this program, with corrections for non-optimum expansion, as described below The F1, RS-27 and H1 engines are LOX/kerosene engines designed as first stage engines for boosters. All are slightly overexpanded at sea level (the F1 has an exit pressure of about 6.8 psi, and the RS-27 and H1 about 12.3 psi). The MA5 engine uses the same propellants, but is a three chamber engine, with two boost chambers and nozzles designed for sea level operation, and 1 sustainer chamber and nozzle designed for high altitude efficiency. At sea level, the booster engines are slightly overexpanded (exit pressure 11.2 psi) while the sustainer is very badly overexpanded (exit pressure 2.8 psi). The J2 and SSME burn hydrogen and oxygen, and are optimized for vacuum operation (although the SSME also operates in the overexpanded condition at takeoff). The SSME engine has an extremely effective staged combustion cycle, which probably accounts for its attainment of a very high fraction of its theoretical Isp. All engines in this table are pump fed, stealing energy in one way or another from the propellants in order to power the pumps. Pressure fed liquid fueled engines would be expected to yield a somewhat higher fraction of real to theoretical Isp because of the lack of this parasitic loss. Calculation of Isp for under or over expanded engine. Principle: For a given nozzle, the program calculates both Isp (optimum), and Isp (vacuum). The extra Isp in a vacuum comes from the additional thrust generated by the difference between the exhaust pressure and ambient pressure (in this case zero), times the area of the exit plane of the nozzle. For a given engine (when the area of the exit plane is a constant), variations from Isp (optimum) are proportional to the difference between the nozzle exit pressure and the ambient pressure. If for example a nozzle is calculated to have an exit pressure of 10 psi, Isp (vacuum) minus Isp (optimum) for the engine is proportional to 10-0=10. If this same engine were operated in flight at an ambient pressure of 5 psi, then the gain in Isp over Isp (optimum) would be half the gain realized in operating in a vacuum. If however the engine were operated at sea level (14.7 psi) then there would be a reduction in thrust due to the difference in atmospheric pressure (the exit pressure is less than the ambient pressure). The difference between the calculated Isp (optimum) and the new effective Isp would be proportional again to the pressure differential (14.7-10=4.7 psi), but in the negative direction. Let: Pa = ambient atmospheric pressure [specified by the investigator] Pe = exhaust pressure at the exit plane [listed as "pressure" in the output tables] Io = Isp (optimum) [listed as "Isp" in output tables] Iv = Isp (vacuum) [listed as "Isp (vacuum)" in output tables] In = Isp (non-optimum) [the Isp for under or overexpanded engines, which is to be calculated] then: In = [Io] + [(Iv - Io) x (Pe - Pa)/(Pe - 0)] in words: the effective Isp for an under or over expanded engine [In] equals the Isp for optimal expandion [Io], plus how much Isp for that engine increases in a vacuum [lv-Io] times the actual pressure differential in question [Pe-Pa] divided by the pressure differential associated with vacuum operation [Pe-0]. dropping un-necessary brackets and the un-needed zero, which was inserted to make the derivation of the equation clear, we get: In = Io + (Iv - Io) x (Pe - Pa)/Pe This equation can be used to estimate the Isp of an engine which is not optimally expanded. Note that when the engine is under expanded, Pe is greater than Pa, and there is a gain in Isp. When Pe is less than Pa, there is a loss in Isp. -- Bruce Dunn Vancouver, Canada Bruce_Dunn@mindlink.bc.ca ------------------------------ Date: Fri, 2 Jul 93 13:18:20 -0400 From: dietz@cs.rochester.edu To: space-tech@cs.cmu.edu Subject: fiberglass I was wondering if someone could tell me about fiberglass. I noticed that E-glass fibers (which cost $1.65/kg in bulk) have a tensile strength/density ratio some 8 times higher than the high strength steel Bruce is proposing for P2. How much of this strength actually ends up in fiberglass/epoxy composite? Paul F. Dietz dietz@cs.rochester.edu ------------------------------ Date: Fri, 2 Jul 93 11:06:05 PDT From: korn@cadre.com (Roger Korn) To: space-tech@cs.cmu.edu Subject: re:fiberglass Fabrication density ratios of 65% glass/resin are achievable with vacuum bagging and up to 75-80% with pressure autoclaving. You also need some axial strength, so, assuming a 30 degree filament inclination, the efficiencies would be: 8 * .65 * cos 30 = 4.5 for vacuum bag fab and 5.54 for autoclave fab. Elongations at max fiber stress approach 9% which may be undesireable. Roger Korn korn@cadre.com ------------------------------ To: uunet!cs.rochester.edu!dietz@uunet.UU.NET Cc: space-tech@cs.cmu.edu, gwh@lurnix.COM Subject: Re: fiberglass Date: Fri, 02 Jul 93 15:52:51 -0700 From: gwh@lurnix.COM I don't have the exact numbers at work, do at home (this source in fact...). However, recall the loss of strength that accompanies other composite fibers from raw fiber strength to strength of a resin/fiber composite, per my earlier posting... For Kevlar 49, you go from 2650 MPa to 1380 MPa in a unidirectional composite, and down to 517 MPa for bidirectional layups. So you're looking at about a factor of six reduction for that fiber. I'll try and post tonight the whole table I took the below figures from... -george william herbert Retro Aerospace -----begin included file (previous mail to space-tech) Kevlar 49, 90 degree woven bidirectional layup (has max strength along x, y axies, much lower along intermediate angles) in Epoxy matrix: Ultimate Tensile Strength: 517 MPa Density: 1.33 +-45 degree woven XAS Carbon fiber, (closer to even strength in all axies) in Epoxy: UTS: 240 MPa Density: 1.53 +-90 degree woven XAS, Epoxy (again, strong only along x, y axies) UTS: 625 MPa Density: 1.53 Unidirectional XAS, Epoxy (strength only along one axis) UTS: 2040 MPa Density: 1.57 Unidirectional Kevlar UTS: 1379 MPa Density: 1.38 The fibers involved, without a matrix, have UTS of roughly 2650 MPa (Kevlar) and 3850 MPa (Carbon XAS). There are slightly stronger fibers out there, but their performance in an actual composite layup is similar in bi- and multi-directional usage. Their primary usages are in uni-directional layups. As you can see, there is a factor of 7 reduction in strength from unidirectional fiber to even bidirectional composite layup; for a real spherical tank, I would guess that a limiting stress would be on the order of 165 MPa. [For the metric-impaired (Including Me, most of the time): 1 MPa = 145 PSI in other words, Carbon XAS has a fiber strength of about 560,000 PSI but in a non-directional composite it's closer to 35,000 PSI (and you have to have a safety margin on top of that...)]. _Design with Advanced Composite Materials_, Ed. Leslie N Phillips, Springer-Verlag, New York (1989) pp. 5, 18,21, esp. 73 (table) ------------------------------ Date: Fri, 2 Jul 93 20:33:13 -0400 From: dietz@cs.rochester.edu To: gwh@lurnix.lurnix.com Subject: Re: fiberglass Cc: space-tech@cs.cmu.edu I notice Sutton has the following information (5th ed., page 327): Material Tensile Strength Density Strength/Density Modulus (kpsi) (lb/in^3) (mpsi) D6aC steel 230 .283 813 29 Maraging steel 200-300 .289 865 27.5 Glass composite (Owens-Corning No. 2-901, 600 kpsi fiber, in 20% epoxy) 170 .072 2360 4.6 Organic composite (Dupont PRD-49 in 30% epoxy) 250 .050 5000 11 That glass in the composite is, I think, S-glass, which I think (in the commercial form, S-2) cost $6/kg in 1985. On the face of it, this would suggest that glass composites are stronger than D6aC steel, even if one has to divide by two for spherical tanks, or downrate the glass by 25% to account for use of the weaker E-glass. Note the lower modulus, however. Paul ------------------------------ Date: Fri, 2 Jul 1993 17:46:01 MST From: "Richard Schroeppel" To: space-tech@cs.cmu.edu Subject: fiberglass, composites I must be missing something in the discussion of fiberglass, etc. for tanks. Is there no need for compressive strength? What part of the structure is transmitting the engine's thrust to the payload, and for that matter, to the tanks and fuel within? Rich Schroeppel rcs@cs.arizona.edu ------------------------------ End of Space-tech Digest #155 *******************