ALBERT

Description: Viewgraphs from a presentation on multiple robot systems in space are included. Topics covered include categories of robots in space; scenarios of robot applications in space; some characteristics of robots in space; and some interesting problems and issues.

Description: Redundant search often makes Prolog programs inefficient. Prolog execution corresponds to a depth-first traversal of an AND/OR graph; often a large subgraph will be searched successfully, with failure resulting from a subsequent small one. It is shown how reordering of Prolog clauses and especially goals can prevent unnecessary search. The restrictions on reordering are characterized and how they may be detected is shown. A new system of calling modes for Prolog, geared to reordering, and a system for inferring them automatically are designed. An improved method for determining a good goal order for Prolog clauses is presented, and it is used as the basis for a reordering system, showing how it can be guided by information about modes and restrictions to generate reordered Prolog that behaves correctly.

Description: Growing interest in a future manned mission to Mars illuminated a critical need for more information on the Martian environment, surface conditions, weather patterns, topography, etc. While the Viking landers provided valuable information of this type, the information came from fixed locations. There is a real need for Viking type of information from a number of locations on the Martian surface in order to adequately survey the planet for future landing and exploration sites. Current site survey mission discussions range from Mars orbiters to sample return missions. The limited data return from the former and the extreme expense of the latter suggest consideration of a 'middle ground' mission which provides needed survey information for an acceptable investment. Utah State University (USU) designed a Mars Lander/Rover (MLR) for use in gathering needed environmental and surface information from Mars. Philosophically, the MLR resembles a mobile Viking; that is, it moves from location to location on the Martian surface, measuring environmental conditions, analyzing soil samples, charting topographical features etc. Measured data is then telemetered to earth for further analysis. Conceptually, it was envisioned that MLR survey locations would be rather widely separated. In that sense the MLR was not a terrestrial vehicle limited to local movement about a fixed location. Rather, it would have the capability for movement over long distances to reach widely separated locations. The design focus, then, was upon a Mars Lander/Rover that leaves an orbit around Mars, reenters and soft lands on the Martian surface and moves sequentially to widely scattered locations to sample, measure, and analyze Martian environmental and surface conditions. Primary goals were payload mass and size definition, characterization of the Martian atmosphere, selection of sampling locations, identification of alternative design concepts, selection of a preferred concept, team organization, and preparation for the detailed design phase.

Description: An essential ingredient in scientific work is data. In disciplines such as Oceanography, data sources are many and volumes are formidable. The full value of large stores of data cannot be realized unless careful thought is given to data access. JPL has developed the Pilot Ocean Data System to investigate techniques for archiving and accessing ocean data obtained from space. These include efficient storage and rapid retrieval of satellite data, an easy-to-use user interface, and a variety of output products which, taken together, permit researchers to extract and use data rapidly and conveniently.

Description: New methods for generating spur, helical, and spiral-bevel gears are proposed. These methods provide the gears with conjugate gear tooth surfaces, localized bearing contact, and reduced sensitivity to gear misalignment. Computer programs have been developed for simulating gear meshing and bearing contact.

Description: Systematic methods for the design of adaptive force and position controllers for robot manipulators within the hybrid control architecture are presented. The force controller tracks the desired force setpoints while the position controller tracks the desired position trajectories. The present control schemes are computationally efficient and suitable for implementation in online control with high sampling rates.

Description: High-pressure, high-temperature seal flow (leakage) data for nonrotating and rotating Raleigh-step and convergent-tapered-bore seals were characterized in terms of a normalized flow coefficient. The data for normalized Rayleigh-step and nonrotating tapered-bore seals were in reasonable agreement with theory, but data for the rotating tapered-bore seals were not. The tapered-bore-seal operational clearances estimated from the flow data were significantly larger than calculated. Although clearances are influenced by wear from conical to cylindrical geometry and errors in clearance corrections, the problem was isolated to the shaft temperature - rotational speed clearance correction. The geometric changes support the use of some conical convergence in any seal. Under these conditions rotation reduced the normalized flow coefficiently by nearly 10 percent.

Description: Difficulties encountered in the modeling of fault-tolerant systems are discussed. The Hybrid Automated Reliability Predictor (HARP) approach to modeling fault-tolerant systems is described. The HARP is written in FORTRAN, consists of nearly 30,000 lines of codes and comments, and is based on behavioral decomposition. Using the behavioral decomposition, the dependability model is divided into fault-occurrence/repair and fault/error-handling models; the characteristics and combining of these two models are examined. Examples in which the HARP is applied to the modeling of some typical fault-tolerant systems, including a local-area network, two fault-tolerant computer systems, and a flight control system, are presented.

Description: The Common Data Format (CDF), a structure which provides true data independence for applications software and has been developed at the National Space Science Data Center, is discussed. The background to the CDF is reviewed, and the CDF is described. The conceptual organization of the CDF is discussed, and a sample CDF structure is shown and described. The implementation of CDF, its status, and its applications are examined.

Description: The integration of CAD/CAM/mapping with image processing using geographic information systems (GISs) as the interface is examined. Particular emphasis is given to the development of software interfaces between JPL's Video Image Communication and Retrieval (VICAR)/Imaged Based Information System (IBIS) raster-based GIS and the CAD/CAM/mapping system. The design and functions of the VICAR and IBIS are described. Vector data capture and editing are studied. Various software programs for interfacing between the VICAR/IBIS and CAD/CAM/mapping are presented and analyzed.

Description: An analysis was performed to determine the effects of inner-ring speed and press fits on roller bearing fatigue life. The effects of the resultant hoop and radial stresses on the principal stresses were considered. The maximum shear stresses below the Hertzian contact were determined for different conditions of inner-ring speed and load, and were applied to a conventional roller bearing life analysis. The effect of mean stress was determined using Goodman diagram approach. Hoop stresses caused by press fits and centrifugal force can reduce bearing life by as much as 90 percent. Use of a Goodman diagram predicts life reduction of 20 to 30 percent. The depth of the maximum shear stress remains virtually unchanged.

Description: Thin, adherent sputtered MoS2 and ion plated metallic (Au, Ag, Pb) lubricating films are primarily used in precision contacting triboelement surfaces where wear debris formation is critical and high reliability requirements have to be satisfied. Detailed structural and compositional characterization of solid film lubricants is of prime importance. It is this information from the nano-micro-macro level which is needed to interpret and improve the frictional behavior and assure long endurance lives. The purpose of this paper is to summarize in a concise review the solid lubricant film structure and morphology and their effects on the tribological properties of the lubricant systems. The tribological performance of thin lubricating films has significantly advanced through progressive understanding of the film parameters such as adhesion, cohesion, interface formation, nucleation and microstructural growth, critical film thickness and substrate finish, and temperature. Sputtered MoS2 and ion plated Au, Ag, and Pb films are separately discussed and evaluated in terms of the above film parameters to establish the most desirable film structures and thicknesses in order to achieve effective lubrication.

Description: The FDDI Token Ring Protocol controls communication over fiber optic rings with transmission rates in the range of 100 megabits per second. It is intended to give guaranteed response to time-critical messages by using a 'timed token' protocol, in which noncritical messages may be transmitted only if recent movement of the token among stations has been sufficiently fast relative to a 'target' token rotation line (TTRT). In this paper, two important properties of the protocol are proved. The first is that the average token cycle time is bounded above by the TTRT, and the second is that the maximum token cycle time is at most twice the TTRT. Each property is treated first under the assumption that all overheads are negligible, and second with certain sources of overhead taken into account explicitly. It is found that the proposed standard protocol can be improved for situations of practical interest by a slight modification.

Description: A technique has been developed for modeling solid films in rolling element bearings. For a given bearing geometry an effective viscosity can be calculated for a solid film as a function of the film shear modulus and ball-race friction coefficient. The calculated effective viscosity can subsequently be used as an input to a numerical model of cage motion and stability. Results from a sample calculation of effective viscosity and prediction of cage stability for a turbopump bearing are presented for films of molybdenum disulfide (MoS2) and polytetrafluoroethylene ()PTFE).

Description: Spur gear surface fatigue endurance tests were conducted to investigate hot forged powder metal AISI 4620 and 4640 steel for use as a gear material, to determine endurance characteristics and to compare the results with machined AISI 4340 and 9310 steel gear materials. The as-forged and unground AISI 4620 gear exhibited a 10 percent fatigue life that was approximately one-fourth of that for AISI 9310 and less than one-half that for the AISI 4340 gears. The forged and finish ground AISI 4620 gears exhibited a 10 percent life, approximately 70 percent that of AISI 9310 and slightly better than that of AISI 4340. The AISI 4640 hot forged gears had less fracture toughness and slightly less fatigue life than the AISI 4620 test gears.

Description: The prototype demonstrates the feasibility of using Ada for expert systems and the implementation of an expert-friendly interface which supports knowledge entry. In the Ford LISP-Ada Connection (FLAC) system LISP and Ada are used in ways which complement their respective capabilities. Future investigation will concentrate on the enhancement of the expert knowledge entry/debugging interface and on the issues associated with multitasking and real-time expert systems implementation in Ada.

Description: Two expert systems were written in OPS5 by the expert, a Ph.D. astronomer with no prior experience in artificial intelligence or expert systems, without the use of a knowledge engineer. The first system was built from scratch and uses 146 rules to check for duplication of scientific information within a pool of prospective observations. The second system was grafted onto another expert system and uses 149 additional rules to estimate the spacecraft and ground resources consumed by a set of prospective observations. The small vocabulary, the IF this occurs THEN do that logical structure of OPS5, and the ability to follow program execution allowed the expert to design and implement these systems with only the data structures and rules of another OPS5 system as an example. The modularity of the rules in OPS5 allowed the second system to modify the rulebase of the system onto which it was grafted without changing the code or the operation of that system. These experiences show that experts are able to develop their own expert systems due to the ease of programming and code reusability in OPS5.

Description: The proposal selection process for the Hubble Space Telescope is assisted by a robust and easy to use query program (TACOS). The system parses an English subset language sentence regardless of the order of the keyword phases, allowing the user a greater flexibility than a standard command query language. Capabilities for macro and procedure definition are also integrated. The system was designed for flexibility in both use and maintenance. In addition, TACOS can be applied to any knowledge domain that can be expressed in terms of a single reaction. The system was implemented mostly in Common LISP. The TACOS design is described in detail, with particular attention given to the implementation methods of sentence processing.

Description: The principle focus of one of the RICIS (Research Institute for Computing and Information Systems) components is computer systems and software engineering in-the-large of the lifecycle of large, complex, distributed systems which: (1) evolve incrementally over a long time; (2) contain non-stop components; and (3) must simultaneously satisfy a prioritized balance of mission and safety critical requirements at run time. This focus is extremely important because of the contribution of the scaling direction problem to the current software crisis. The Computer Systems and Software Engineering (CSSE) component addresses the lifestyle issues of three environments: host, integration, and target.

Description: A cooperative program was initiated by JSC with the University of Houston - Clear Lake to support research in computing and information systems. The objective of this program is to provide continuing long-term research in support of the numerous mission and mission-related endeavors of NASA/JSC. The Research Institute for Computing and Information Systems (RICIS) concept is presented in form of viewgraphs. History of RICIS, its organization, and research status are discussed.

Description: Quality of software not only is vital to the successful operation of the space station, it is also an important factor in establishing testing requirements, time needed for software verification and integration as well as launching schedules for the space station. Defense of management decisions can be greatly strengthened by combining engineering judgments with statistical analysis. Unlike hardware, software has the characteristics of no wearout and costly redundancies, thus making traditional statistical analysis not suitable in evaluating reliability of software. A statistical model was developed to provide a representation of the number as well as types of failures occur during software testing and verification. From this model, quantitative measure of software reliability based on failure history during testing are derived. Criteria to terminate testing based on reliability objectives and methods to estimate the expected number of fixings required are also presented.

Description: The ability to determine the contents of a scene by a machine has obvious applications. There are reasons for seeking horizontal, vertical and diagonal line segments of varying lengths in order to interpret the content of a scene. Such line segments aid in determining edges that determine the boundaries of objects within the scene. Algorithms for each of the above as well as for data compacting are presented.

Description: The instability and unbalance response of dissymmetric rotor-bearing systems containing periodic coefficients when modeling produces matrices with a large number of degrees of freedom are discussed. It is important to solve the equations and then predict the dynamic behavior of the system. This can be done knowing the instability areas and the unbalance response in the stable areas. One deals here with a large number of equations and a reduction of the number of degrees of freedom of the system is achieved through a pseudo modal method. This method is shown to give satisfactory results.

Description: Means were developed to measure the effect of vibration on human beings in a tank vehicle so as to determine the limits to which they might be taken in terms of operating effectiveness and also in terms of well being. Improved hardware for making these measurements, a method for relating tank hull input forces to ride quality, and installation and calibration of instrumentation to measure the hull forces are discussed.

Description: Except in the near Earth phase of GIOTTO's mission to Comet Halley, the HGA (high gain antenna) on board GIOTTO was the only designed means of up/down communications. The spacecraft spin stabilization required that the HGA be despun at the same rotational rate of nominally 15 rpm in order to keep the HGA pointing accurately to a Earth. A dual servomotor despin mechanism was designed and built by SEP of France for this purpose. The expected thermal environment suggested that dry lubrication was preferable to wet for the ball bearings but there existed no relevant data on the torque noise spectrum of candidate solid lubricants. Therefore ad hoc torque noise tests were run with two solid lubricants: ion plated lead film plus lead bronze cage (retainer) and a PTFE composite cage only. The lead lubrication showed the better spectrum up to the mission lifetime point so it was selected for continued test over some 20 times the Halley mission life, with periodic torque spectrum monitoring. The spectrum remained well within the pointing error budget over the 100 million revolutions covered.

Description: In the early 1970s, an antenna despin drive was developed for MBB solar science satellite HELIOS. A problem with high bearing drag torque that was encountered on the two flight models of this drive, after successful tests were completed on twelve bearings, an engineering model, and the qualification unit is discussed.

Description: The workings and function of the Preloader Vector Sensitive Latch are described. A discussion of docking systems used in the U.S. manned space flight programs is included to show how docking systems have evolved, and to highlight the potential advantages of a preloadable vector sensitive latch in such systems.

Description: Governing equations of motion are derived for a bulk-flow model of the leakage path between an impeller shroud and a pump housing. The governing equations consist of a path-momentum, a circumferential - momentum, and a continuity equation. The fluid annulus between the impeller shroud and pump housing is assumed to be circumferentially symmetric when the impeller is centered; i.e., the clearance can vary along the pump axis but does not vary in the circumferential direction. A perturbation expansion of the governing equations in the eccentricity ratio yields a set of zeroth and first-order governing equations. The zeroth-order equations define the leaking rate and the circumferential and path velocity distributions and pressure distributions for a centered impeller position. The first-order equations define the perturbations in the velocity and pressure distributions due to either a radial-displacement perturbation or a tilt perturbation of the impeller. Integration of the perturbed pressure and shear-stress distribution acting on the rotor yields the reaction forces and moments acting on the impeller face.

Description: Discussed is a magnetic bearing which was designed and tested in a flexible rotor both as support bearings and as a vibration controller. The design of the bearing is described and the effect of control circuit bandwidth determined. Both stiffness and damping coefficients were measured and calculated for the bearing with good agreement. The bearings were then placed in a single mass rotor as support bearings and the machine run through two critical speeds. Measurements were made of the vibration response in plain bushings and magnetic bearings. Comparisons were also made of the theoretical calculations with the measured peak unbalance response speeds. Finally, runs were made with the magnetic bearing used as a vibration controller.

Description: Anti-swirl vanes are used by some manufacturers to delay the full development of half speed circulation in annular clearance spaces. The objective is to reduce the aerodynamic cross-coupling in the forward direction. The novel feature of a jet damper is a number of tangential nozzles discharging against the rotor surface speed. Some preliminary results on a 33.9 Kg rotor demonstrate that significant reductions in amplitude are obtained at the synchronous critical speeds.

Description: This paper is concerned with the non-linear vibration performance of a rigid rotor supported on three bearings, one being surrounded by a squeeze-film damper. This damper relies on the pressure built up in the squeeze film to help counter-act external forces arising from unbalance and other effects. As a result a vibration orbit of a certain magnetude results. Such vibration orbits illustrate features found in other non-linear systems, in particular sub-harmonic resonances and jump phenomena. Comparisons between theoretical prediction and experimental observations of these phenomena are made.

Description: Described is an experimental study to identify the damping laws associated with a squeeze-film vibration damper. This is achieved by using a non-linear filtering algorithm to process displacement responses of the damper ring to synchronous excitation and thus to estimate the parameters in an nth-power velocity model. The experimental facility is described in detail and a representative selection of results is included. The identified models are validated through the prediction of damper-ring orbits and comparison with observed responses.

Description: A great deal of research has recently been conducted to solve the subsynchronous rotor vibration problems in high-performance turbomachinery. Particularly, the destabilizing effect of the labyrinth seal on compressors or turbines has been investigated for many years. In spite of many efforts the dynamic effect of the labyrinth seal had not been fully determined from qualitative and quantitative points of view. But from theoretical and experimental work, the dynamic characteristics of the labyrinth seal have been established. The results of recent theoretical and experimental works are presented.

Description: Based on Childs finite length solution for annular plain seals an extension of the bulk flow theory is derived to calculate the rotordynamic coefficients and the leakage flow of seals with parallel grooves in the stator. Hirs turbulent lubricant equations are modified to account for the different friction factors in circumferential and axial direction. Furthermore an average groove depth is introduced to consider the additional circumferential flow in the grooves. Theoretical and experimental results are compared for the smooth constant clearance seal and the corresponding seal with parallel grooves. Compared to the smooth seal the direct and cross-coupled stiffness coefficients as well as the direct damping coefficients are lower in the grooved seal configuration. Leakage is reduced by the grooving pattern.

Description: The K1501 compressor, driven by a steam turbine, is used to transport synthetic gas in fertilizer plants of 1000 tons daily production. The turbo-compressor set, which had been in operation since 1982, vibrated rather intensely, and its maximum load was only about 95 percent of the normal value. Damaging vibration to pads and gas-sealing labyrinths occurred three times from 1982 to 1983 and resulted in considerable economic loss. From the characteristics of the vibration, we suspected its cause to be rotor instability due to labyrinth-seal excitation. But, for lack of experience, the problem was not addressed for two years. Finally, we determined that the instability was indeed produced by labyrinth-seal excitation and corrected this problem by injecting gas into the middle-diaphragm labyrinths. This paper primarily discusses the failure and the remedy described above.

Description: A review of previous experimental measurements of squeeze film damper (SFD) forces is given. Measurements by the authors of SFD pressure fields and force coefficients, for circular centered orbits with epsilon = 0.5, are described and compared with computer predictions. For Reynolds numbers over the range 2-6, the effect of fluid inertia on the pressure fields and forces is found to be significant.

Description: Earlier work has shown that oil-film forces can be modelled by linear coefficients. Identification techniques were used to generate numerical values for these coefficients. This paper has shown the invalidity of applying the perturbation techniques normally used in bearing studies to derive expressions for linearized coefficients to represent a cavitated oil-film. An alternative approach was developed based upon energy techniques to obtain estimates for linearized coefficients. Some current work being undertaken suggests that an alternative analytical approach is possible. These results will be reported in due course.

Description: An instability associated with an intershaft squeeze film damper is described. The squeeze film is located between the intershaft bearing outer race and the low-speed shaft of a five-bearing, two-spool test rig. The instability is dominated by response of the third system mode to destabilizing excitation of the type described by Hibner, et al. Installing a spring cage in place of the intershaft damper removes the instability and produces satisfactory performance throughout the operating range.

Description: The prediction of rotordynamic coefficients for gas seals is achieved with the aid of a two-volume bulk flow model based on turbulent rotationally symmetric 3D flow calculations including swirl flow. Comparison of cross-coupling and damping coefficients with measurements confirm this approach. In particular the theoretically predicted phenomenon that labyrinth damping is retained without inlet swirl is confirmed. This is important for the design of high pressure compressors, where labyrinth damping is a major contribution improving rotor stability. Discrepancies are found when comparing theory with measured direct stiffness and the cross-coupling damping coefficients. First measurements of labyrinth seals on a recently installed test rig operated with water are presented. Since forces are larger than on test stands operated with air and since individual chamber forces are obtained phenomena like inlet effects may be studied.

Description: The results of numerous investigators have shown the importance of labyrinth seal inlet swirl on the calculated dynamic stiffness of labyrinth seals. These results have not included any calculation of inlet leakage swirl as a function of geometry and sealing conditions of the given seal. This paper outlines a method of calculating the inlet swirl at a given seal by introducing a radial chamber to predict the gas swirl as it goes from the stage tip down to the seal location. For a centrifugal compressor, this amounts to including the flow path from the impeller discharge, down the back of the disk or front of the cover, then into the shaft seal or eye packing, respectively. The solution includes the friction factors of both the disk and stationary wall with account for mass flow rate and calculation of radial pressure gradients by a free vortex solution. The results of various configurations are discussed and comparisons made to other published results of disk swirl.

Description: The rotordynamic coefficients of an incompressible-flow annular pressure seal were determined using a bulk-flow model in conjunction with two different friction factor relationships. The first, Hirs' equation, assumes the friction factor is a function of Reynolds number only. The second, Moody's equation, approximates Moody's diagram and assumes the friction factor is a function of both Reynolds number and relative roughness. For each value of relative roughness, Hirs' constants were determined so that both equations gave the same magnitude and slope of the friction factor. For smooth seals, both relationships give the same results. For rough seals (e/2 H sub 0 = 0.05) Moody's equation predicts 44% greater direct stiffness, 35% greater cross-coupled stiffness, 19% smaller cross-coupled damping, 59% smaller cross-coupled inertia, and nominally the same direct damping and direct inertia.

Description: The dynamic characteristics of spiral-grooved seals are theoretically obtained by using the Navier-Stokes equation. First, with the inertia term of the fluid considered, the flow and pressure in the steady state are obtained for the directions parallel to and perpendicular to the groove. Next, the dynamic character is obtained by analyzing the steady state and by analyzing the labyrinth seal. As a result, the following conclusions were drawn: (1) As the land width becomes shorter or the helix angle decreases, the cross-coupling stiffness, direct and cross-coupling damping, and add mass coefficients decrease; (2) As the axial Reynolds number increases, the stiffness and damping coefficients increase. But the add mass coefficient is not influenced by the axial Reynolds number; (3) The rotational Reynolds number influences greatly the direct and cross-coupling stiffness and direct damping coefficients; and (4) As the journal rotating frequency increases, the leakage flow decreases. Therefore zero net leakage flow is possible at a particular rotating frequency.

Description: For modelling the turbulent flow in a seal the Navier-Stokes equations in connection with a turbulence (kappa-epsilon) model are solved by a finite-difference method. A motion of the shaft round the centered position is assumed. After calculating the corresponding flow field and the pressure distribution, the rotor-dynamic coefficients of the seal can be determined. These coefficients are compared with results obtained by using the bulk flow theory of Childs and with experimental results.

Description: The principal sources of vibration related reliability problems in high pressure centrifugal gas compressors are the re-excitation of the first critical speed or Resonant Subsynchronous Vibration (RSSV), and the forced vibration due to rotating stall in the vaneless diffusers downstream of the impellers. An example of such field problems is given elsewhere. This paper describes the results of a test program at the author's company, initiated in 1983 and completed during 1985, which studied the RSSV threshold and the rotating stall phenomenon in a high pressure gas compressor.

Description: Various computer programs based upon the SIMPLE or SIMPLER algorithm were studied and compared for numerical accuracy, efficiency, and grid dependency. Four two-dimensional and one three-dimensional code originally developed by a number of research groups were considered. In general, the accuracy and computational efficieny of these TEACH type programs were improved by modifying the differencing schemes and their solvers. A brief description of each program is given. Error reduction, spline flux and second upwind differencing programs are covered.

Description: A large scale, general purpose finite element computer program, ANSYS, developed and marketed by Swanson Analysis Systems, Inc. is discussed. ANSYS was perhaps the first commercially available program to offer truly interactive finite element model generation. ANSYS's purpose is for solid modeling. This application is briefly discussed and illustrated.

Description: Basic hand grips and sensor interactions that a dextrous robot hand will need as part of the operation of an EVA Retriever are analyzed. What is to be done with a dextrous robot hand is examined along with how such a complex machine might be controlled. It was assumed throughout that an anthropomorphic robot hand should perform tasks just as a human would; i.e., the most efficient approach to developing control strategies for the hand would be to model actual hand actions and do the same tasks in the same ways. Therefore, basic hand grips that human hands perform, as well as hand grip action were analyzed. It was also important to examine what is termed sensor fusion. This is the integration of various disparate sensor feedback paths. These feedback paths can be spatially and temporally separated, as well as, of different sensor types. Neural networks are seen as a means of integrating these varied sensor inputs and types. Basic heuristics of hand actions and grips were developed. These heuristics offer promise of control dextrous robot hands in a more natural and efficient way.

Description: Perhaps the most critical structural system aboard the Space Station is the Solar Alpha Rotary Joint which helps align the power generation system with the sun. The joint must provide structural support and controlled rotation to the outboard transverse booms as well as power and data transfer across the joint. The Solar Alpha Rotary Joint is composed of two transition sections and an integral, large diameter bearing. Alpha joint bearing design presents a particularly interesting problem because of its large size and need for high reliability, stiffness, and on orbit maintability. The discrete roller bearing developed is a novel refinement to cam follower technology. It offers thermal compensation and ease of on-orbit maintenance that are not found in conventional rolling element bearings. How the bearing design evolved is summarized. Driving requirements are reviewed, alternative concepts assessed, and the selected design is described.

Description: A framework is presented for developing computationally unified numerical algorithms for solving nonlinear equations that arise in modeling various problems in mathematical physics. The concept of computational unification is an attempt to encompass efficient solution procedures for computing various nonlinear phenomena that may occur in a given problem. For example, in Computational Fluid Dynamics (CFD), a unified algorithm will be one that allows for solutions to subsonic (elliptic), transonic (mixed elliptic-hyperbolic), and supersonic (hyperbolic) flows for both steady and unsteady problems. The objectives are: development of superior unified algorithms emphasizing accuracy and efficiency aspects; development of codes based on selected algorithms leading to validation; application of mature codes to realistic problems; and extension/application of CFD-based algorithms to problems in other areas of mathematical physics. The ultimate objective is to achieve integration of multidisciplinary technologies to enhance synergism in the design process through computational simulation. Specific unified algorithms for a hierarchy of gas dynamics equations and their applications to two other areas: electromagnetic scattering, and laser-materials interaction accounting for melting.

Description: Turbo Prolog is a recently available, compiled version of the programming language Prolog. Turbo Prolog is designed to provide not only a Prolog compiler, but also a program development environment for the IBM Personal Computer family. An evaluation of Turbo Prolog was made, comparing its features to other versions of Prolog and to the community of languages commonly used in artificial intelligence (AI) research and development. Three programs were employed to determine the execution speed of Turbo Prolog applied to various problems. The results of this evaluation demonstrated that Turbo Prolog can perform much better than many commonly employed AI languages for numerically intensive problems and can equal the speed of development languages such as OPS5+ and CLIPS, running on the IBM PC. Applications for which Turbo Prolog is best suited include those which (1) lend themselves naturally to backward-chaining approaches, (2) require extensive use of mathematics, (3) contain few rules, (4) seek to make use of the window/color graphics capabilities of the IBM PC, and (5) require linkage to programs in other languages to form a complete executable image.

Description: Dynamic model verification is the process whereby an analytical model of a dynamic system is compared with experimental data, and then qualified for future use in predicting system response in a different dynamic environment. There are various ways to conduct model verification. The approach adopted in MOVER II employs Bayesian statistical parameter estimation. Unlike curve fitting whose objective is to minimize the difference between some analytical function and a given quantity of test data (or curve), Bayesian estimation attempts also to minimize the difference between the parameter values of that function (the model) and their initial estimates, in a least squares sense. The objectives of dynamic model verification, therefore, are to produce a model which: (1) is in agreement with test data, (2) will assist in the interpretation of test data, (3) can be used to help verify a design, (4) will reliably predict performance, and (5) in the case of space structures, facilitate dynamic control.

Description: This paper describes a new type of electromagnetic bearing, called the Eddy-Current Bearing, which works by repulsion between fixed AC-electromagnets and a conducting rotor. The following advantages are expected: inherent stability, higher load carrying capacity than DC-electromagnetic bearings, simultaneous radial, angular and thrust support, motoring and generating capability, and backup mode of operation in case of primary power failure. A prototype is under construction.

Description: The flexural vibrations of a rotating shaft, running through one or more critical speeds, can be reduced to an acceptably low level by applying suitable control forces at an intermediate span position. If electromagnets are used to produce the control forces then it is possible to implement a wide variety of control strategies. A test rig is described which includes a microprocessor-based controller, in which such strategies can be realised in terms of software-based algorithms. The electromagnet configuration and the method of stabilising the electromagnet force-gap characteristic are discussed. The bounds on the performance of the system are defined. A simple control algorithm is outlined, where the control forces are proportional to the measured displacement and velocity at a single point on the shaft span; in this case the electromagnet behaves in a similar manner to that of a parallel combination of a linear spring and damper. Experimental and predicted performance of the system are compared, for this type of control, where various programmable rates of damping are applied.

Description: The results are given for an experimental program concerned with a parametric identification of the damping and inertial coefficients of a cylindrical squeeze-film bearing, through an analysis of transient response data. The results enable the operating range for which a linear model of the squeeze-film is appropriate to be determined. Comparisons are made between the estimated coefficients and theoretical predictions. Presentation is by courtesy of the Council of the Institution of Mechanical Engineers, London.

Description: Presented is an analytical method to determine static and dynamic characteristics of annular parallel-grooved seals. The governing equations were derived by using the turbulent lubrication theory based on the law of fluid friction. Linear zero- and first-order perturbation equations of the governing equations were developed, and these equations were analytically investigated to obtain the reaction force of the seals. An analysis is presented that calculates the leakage flow rate, the torque loss, and the rotordynamic coefficients for parallel-grooved seals. To demonstrate this analysis, we show the effect of changing number of stages, land and groove width, and inlet swirl on stability of the boiler feed water pump seals. Generally, as the number of stages increased or the grooves became wider, the leakage flow rate and rotor-dynamic coefficients decreased and the torque loss increased.

Description: A multistage centrifugal compressor, in natural gas re-injection service on an offshore petroleum production platform, experienced subsynchronous vibrations which caused excessive bearing wear. Field performance testing correlated the subsynchronous amplitude with the discharge flow coefficient, demonstrating the excitation to be aerodynamic. Adding two impellers allowed an increase in the diffuser flow angle (with respect to tangential) to meet the diffuser stability criteria based on factory and field tests correlated using the theory of Senoo (for rotating stall in a vaneless diffuser). This modification eliminated all significant subsynchronous vibrations in field service, thus confirming the correctness of the solution. Other possible sources of aerodynamically induced vibrations were considered, but the judgment that those are unlikely has been confirmed by subsequent experience with other similar compressors.

Description: A multiflexible body dynamics code intended for fast turnaround control design trades is described. Nonlinear rigid body dynamics and linearized flexible dynamics combine to provide efficient solution of the equations of motion. Comparison with results from the DISCOS code provide verification of accuracy.

Description: Some important aspects of design modeling for shape optimization are discussed for both stamped sheet metal components and cast solid components. For stamped components the basis for the modeling approach is a boundary design function. Design parameters control the shape of 2-D regions. For more complex, folded plate components, the 2-D regions can be assembled using translation and rotation operations. The analysis model is automatically created using a mesh generation procedure requiring only boundary data. For less complex solid components, it was found that this approach is not suitable. For these structures, the finite element models are typically created using very sophisticated graphical modeling systems. A new approach which overlays a parameterized surface design model on an existing analysis model is described. To summarize, the future needs for solid shape design is described in terms of an extension of the previously described 2-D capability.

Description: The octree technique is developed into the finite octree, and an overview is given. Modeler requirements are given. The octree discretization is discussed along with geometric communication operators. Geometric communication operators returning topological associativity and geometric communication operators returning spatial data are also discussed and illustrated. The advantages are given of the boundary representation and of geometric communication operators. The implementation plays an important role in the integration with a variety of geometric modelers. The capabilities of closed loop processes within a complete finite element system are presented.

Description: The structure to a geometry based finite element preprocessing system is presented. The key features of the system are the use of geometric operators to support all geometric calculations required for analysis model generation, and the use of a hierarchic boundary based data structure for the major data sets within the system. The approach presented can support the finite element modeling procedures used today as well as the fully automated procedures under development.

Description: An intelligent tracker capable of robotic applications requiring guidance and control of platforms, robotic arms, and end effectors has been developed. This packaged system capable of supervised autonomous robotic functions is partitioned into a multiple processor/parallel processing configuration. The system currently interfaces to cameras but has the capability to also use three-dimensional inputs from scanning laser rangers. The inputs are fed into an image processing and tracking section where the camera inputs are conditioned for the multiple tracker algorithms. An executive section monitors the image processing and tracker outputs and performs all the control and decision processes. The present architecture of the system is presented with discussion of its evolutionary growth for space applications. An autonomous rendezvous demonstration of this system was performed last year. More realistic demonstrations in planning are discussed.

Description: This paper describes the use of computer graphic simulation techniques to resolve critical design and operational issues for robotic systems. Use of this technology will result in greatly improved systems and reduced development costs. The major design issues in developing effective robotic systems are discussed and the use of ROBOSIM, a NASA developed simulation tool, to address these issues is presented. Three representative simulation case studies are reviewed: off-line programming of the robotic welding development cell for the Space Shuttle Main Engine (SSME); the integration of a sensor to control the robot used for removing the Thermal Protection System (TPS) from the Solid Rocket Booster (SRB); and the development of a teleoperator/robot mechanism for the Orbital Maneuvering Vehicle (OMV).

Description: The Flight Telerobotic Servicer (FTS), a multiple arm dexterous manipulation system, will aid in the assembly, maintenance, and servicing of the space station. Fundamental ideas and basic conceptual designs for a shuttle-based telerobot system have been produced. Recent space station studies provide additional concepts that should aid in the accomplishment of mission requirements. Currently, the FTS is in contractual source selection for a Phase B preliminary design. At the same time, design requirements are being developed through a series of robotic assessment tasks being performed at NASA and commercial installations. A number of the requirements for remote operation on the space station, necessary to supplement extravehicular activity (EVA), will be met by the FTS. Finally, technology developed for telerobotics will advance the state of the art of remote operating systems, enhance operator productivity, and prove instrumental in the evolution of an adaptive, intelligent autonomous robot.

Description: The ACCESS truss was telerobotically assembled in order to gain experience with robotic assembly of hardware designed for astronaut extravehicular (EVA) assembly. Tight alignment constraints of the ACCESS hardware made telerobotic assembly difficult. A wider alignment envelope and a compliant end effector would have reduced the problem. The manipulator had no linear motion capability, but many of the assembly operations required straight line motion. The manipulator was attached to a motion table in order to provide the X, Y, and Z translations needed. A programmable robot with linear translation capability would have eliminated the need for the motion table and streamlined the assembly. Poor depth perception was a major problem. Shaded paint schemes and alignment lines were helpful in reducing this problem. The four cameras used worked well for only some operations. It was not possible to identify camera locations that worked well for all assembly steps. More cameras or movable cameras would have simplified some operations. The audio feedback system was useful.

Description: An overview of Automation Technology Branch facilities and research is presented. Manipulator research includes dual-arm coordination studies, space manipulator dynamics, end-effector controller development, automatic space structure assembly, and the development of a dual-arm master-slave telerobotic manipulator system. Sensor research includes gravity-compensated force control, real-time monovision techniques, and laser ranging. Artificial intelligence techniques are being explored for supervisory task control, collision avoidance, and connectionist system architectures. A high-fidelity dynamic simulation of robotic systems, ROBSIM, is being supported and extended. Cooperative efforts with Oak Ridge National Laboratory have verified the ability of teleoperators to perform complex structural assembly tasks, and have resulted in the definition of a new dual-arm master-slave telerobotic manipulator. A bibliography of research results and a list of technical contacts are included.

Description: The resolved rate law for a manipulator provides the instantaneous joint rates required to satisfy a given instantaneous hand motion. When the joint space has more degrees of freedom than the task space, the manipulator is kinematically redundant and the kinematic rate equations are underdetermined. These equations can be locally optimized, but the resulting pseudo-inverse solution has been found to cause large joint rates in some cases. A weighting matrix in the locally optimized (pseudo-inverse) solution is dynamically adjusted to control the joint motion as desired. Joint reach limit avoidance is demonstrated in a kinematically redundant planar arm model. The treatment is applicable to redundant manipulators with any number of revolute joints and to non-planar manipulators.

Description: As part of the Intelligent Task Automation program, a team developed enabling technologies for programmable, sensory controlled manipulation in unstructured environments. These technologies include 2-D/3-D vision sensing and understanding, force sensing and high speed force control, 2.5-D vision alignment and control, and multiple processor architectures. The subsequent design of a flexible, programmable, sensor controlled robotic assembly system for small electromechanical devices is described using these technologies and ongoing implementation and integration efforts. Using vision, the system picks parts dumped randomly in a tray. Using vision and force control, it performs high speed part mating, in-process monitoring/verification of expected results and autonomous recovery from some errors. It is programmed off line with semiautomatic action planning.

Description: The thrust of automation and robotics for space applications has been proposed for increased productivity, improved reliability, increased flexibility, higher safety, and for the performance of automating time-consuming tasks, increasing productivity/performance of crew-accomplished tasks, and performing tasks beyond the capability of the crew. This paper provides a review of efforts currently in progress in the area of robotic vision. Both systems and algorithms are discussed. The evolution of future vision/sensing is projected to include the fusion of multisensors ranging from microwave to optical with multimode capability to include position, attitude, recognition, and motion parameters. The key feature of the overall system design will be small size and weight, fast signal processing, robust algorithms, and accurate parameter determination. These aspects of vision/sensing are also discussed.

Description: Part 1 presented an overview of the unified approach to manufacturing software. The specific characteristics of the approach that allow it to realize the goals of reduced cost, increased reliability and increased flexibility are considered. Why the blending of a components view, distributed languages, generics and formal models is important, why each individual part of this approach is essential, and why each component will typically have each of these parts are examined. An example of a specification for a real material handling system is presented using the approach and compared with the standard interface specification given by the manufacturer. Use of the component in a distributed manufacturing system is then compared with use of the traditional specification with a more traditional approach to designing the system. An overview is also provided of the underlying mechanisms used for implementing distributed manufacturing systems using the unified software/hardware component approach.

Description: For several years, a broad, unified approach to programming manufacturing cells, factory floors, and other manufacturing systems has been developed. It is based on a blending of distributed Ada, software components, generics and formal models. Among other things the machines and devices which make up the components, and the entire manufacturing cell is viewed as an assembly of software components. The purpose of this project is to reduce cost, increase the reliability and increase the flexibility of manufacturing software. An overview is given of the approach and an experimental generic factory floor controller that was developed using the approach is described. The controller is generic in the sense that it can control any one of a large class of factory floors making an arbitrary mix of parts.

Description: The conventional control and monitor software currently used by the Space Center for Space Shuttle processing has many limitations such as high maintenance costs, limited diagnostic capabilities and simulation support. These limitations have caused the development of a knowledge based (or model based) shell to generically control and monitor electro-mechanical systems. The knowledge base describes the system's structure and function and is used by a software shell to do real time constraints checking, low level control of components, diagnosis of detected faults, sensor validation, automatic generation of schematic diagrams and automatic recovery from failures. This approach is more versatile and more powerful than the conventional hard coded approach and offers many advantages over it, although, for systems which require high speed reaction times or aren't well understood, knowledge based control and monitor systems may not be appropriate.

Description: SWAN is an expert system and natural language interface for assessing the war fighting capability of Air Force units in Europe. The expert system is an object oriented knowledge based simulation with an alternate worlds facility for performing what-if excursions. Responses from the system take the form of generated text, tables, or graphs. The natural language interface is an expert system in its own right, with a knowledge base and rules which understand how to access external databases, models, or expert systems. The distinguishing feature of the Air Force expert system is its use of meta-knowledge to generate explanations in the frame and procedure based environment.

Description: Constraints are used in the CAMPS knowledge based planning system to represent those propositions that must be true for a plan to be acceptable. CAMPS introduces the make-mode for interpreting a constraint. Given an unsatisfied constraint, make evaluation mode suggests planning actions which, if taken, would result in a modified plan in which the constraint in question may be satisfied. These suggested planning actions, termed delta-tuples, are the raw material of intelligent plan repair. They are used both in debugging an almost-right plan and in replanning due to changing situations. Given a defective plan in which some set of constraints are violated, a problem solving strategy selects one or more constraints as a focus of attention. These selected constraints are evaluated in the make-mode to produce delta-tuples. The problem solving strategy then reviews the delta-tuples according to its application and problem-specific criteria to find the most acceptable change in terms of success likelihood and plan disruption. Finally, the problem solving strategy makes the suggested alteration to the plan and then rechecks constraints to find any unexpected consequences.

Description: The Space Shuttle community is currently developing a number of knowledge-based tools, primarily expert systems, to support Space Shuttle operations. It is proposed that anticipating and responding to the requirements of the operations environment will contribute to a rapid and smooth transition of expert systems from development to operations, and that the requirements for verification are critical to this transition. The paper identifies the requirements of expert systems to be used for flight planning and support and compares them to those of existing procedural software used for flight planning and support. It then explores software engineering concepts and methodology that can be used to satisfy these requirements, to aid the transition from development to operations and to support the operations environment during the lifetime of expert systems. Many of these are similar to those used for procedural hardware.

Description: The IDEAL (Integrated Design and Engineering Analysis Languages) modeling methodology has been formulated and applied over a five-year period. It has proven to be a unique, integrated approach utilizing a top-down, structured technique to define and document the system of interest; a knowledge engineering technique to collect and organize system descriptive information; a rapid prototyping technique to perform preliminary system performance analysis; and a sophisticated simulation technique to perform in-depth system performance analysis.

Description: Where the Knowledge Based Software Assistant (KBSA) is now, four years after the initial report, is discussed. Also described is what the Rome Air Development Center expects at the end of the first contract iteration. What the second and third contract iterations will look like are characterized.

Description: The generation of a set of tools for software lifecycle is a recurring theme in the software engineering literature. The development of such tools and their integration into a software development environment is a difficult task because of the magnitude (number of variables) and the complexity (combinatorics) of the software lifecycle process. An initial development of a global approach was initiated in 1982 as the Software Development Workbench (SDW). Continuing efforts focus on tool development, tool integration, human interfacing, data dictionaries, and testing algorithms. Current efforts are emphasizing natural language interfaces, expert system software development associates and distributed environments with Ada as the target language. The current implementation of the SDW is on a VAX-11/780. Other software development tools are being networked through engineering workstations.

Description: An overview is presented of the Automated Software Development Workstation Project, an effort to explore knowledge-based approaches to increasing software productivity. The project focuses on applying the concept of domain specific automatic programming systems (D-SAPSs) to application domains at NASA's Johnson Space Center. A version of a D-SAPS developed in Phase 1 of the project for the domain of space station momentum management is described. How problems encountered during its implementation led researchers to concentrate on simplifying the process of building and extending such systems is discussed. Researchers propose to do this by attacking three observed bottlenecks in the D-SAPS development process through the increased automation of the acquisition of programming knowledge and the use of an object oriented development methodology at all stages of the program design. How these ideas are being implemented in the Bauhaus, a prototype workstation for D-SAPS development is discussed.

Description: Control of on-orbit operation of a spacecraft requires retention and application of special purpose, often unique, knowledge of equipment and procedures. Real-time distributed expert systems (RTDES) permit a modular approach to a complex application such as on-orbit spacecraft support. One aspect of a human-machine system that lends itself to the application of RTDES is the function of satellite/mission controllers - the next logical step toward the creation of truly autonomous spacecraft systems. This system application is described.

Description: For the evolving discipline of programming, acquisition of programming knowledge is a difficult issue. Common knowledge results from the acceptance of proven techniques based on results of formal inquiries into the nature of the programming process. This is a rather slow process. In addition, the vast body of common knowledge needs to be explicated to a low enough level of details for it to be represented in the machine processable form. It is felt that this is an impediment to the progress of automatic programming. The importance of formal approaches cannot be overstated since their contributions lead to quantum leaps in the state of the art.

Description: An overview of the knowledge acquisition component of the Bauhaus, a prototype computer aided software engineering (CASE) workstation for the development of domain-specific automatic programming systems (D-SAPS) is given. D-SAPS use domain knowledge in the refinement of a description of an application program into a compilable implementation. The approach to the construction of D-SAPS was to automate the process of refining a description of a program, expressed in an object-oriented domain language, into a configuration of software parts that implement the behavior of the domain objects.

Description: Results of preliminary research on the design of a knowledge based fault diagnosis system for use with on-orbit spacecraft such as the Hubble Space Telescope are presented. A candidate data structure and associated search algorithm from which the knowledge based system can evolve is discussed. This algorithmic approach will then be examined in view of its inability to diagnose certain common faults. From that critique, a design for the corresponding knowledge based system will be given.

Description: Applications of linguistic principles to potential problems of human and machine communication in space settings are discussed. Variations in language among speakers of different backgrounds and change in language forms resulting from new experiences or reduced contact with other groups need to be considered in the design of intelligent machine systems.

Description: Recently the artificial intelligence community has turned its attention to the process of discovery and found that the history of science is a fertile source for what Darden has called compiled hindsight. Such hindsight generates weak heuristics for discovery that do not guarantee that discoveries will be made but do have proven worth in leading to discoveries. Triangulation is one such heuristic that is grounded in historical hindsight. This heuristic is explored within the general framework of the BACON, GLAUBER, STAHL, DALTON, and SUTTON programs. In triangulation different bases of information are compared in an effort to identify gaps between the bases. Thus, assuming that the bases of information are relevantly related, the gaps that are identified should be good locations for discovery and robust analysis.

Description: It is demonstrated that when iterative-deepening A asterisk (IDA asterisk) is applied to one type of resource allocation problem, it uses far less storage than A asterisk, but opens far more nodes and thus has unacceptable time complexity. This is shown to be due, at least in part, to the low-valued effective branching factor that is a characteristic of problems with real-valued cost functions. The semi-optimal, epsilon-admissible IDA asterisk sub epsilon search algorithm that the authors described was shown to open fewer nodes than both A asterisk and IDA asterisk with storage complexity proportional to the depth of the search tree.

Description: An architecture for a knowledge-based simulator is described. The task of scheduling represents an area in which such a tool might be applied. More specifically, scheduling for crew and ground support activities for the shuttle and space station would benefit from the application of knowledge-based simulation. The knowledge-based simulator would allow the crew and support personnel to schedule and reschedule activities in a timely and flexible manner in order to examine and test possible plans.

Description: The development of a simulation assistant for modeling discrete event processes is presented. Included are an overview of the system, a description of the simulation generators, and a sample process generated using the simulation assistant.

Description: A methodology for automatic mathematical modeling is described. The major objective is to create a very friendly environment for engineers to design, maintain and verify their model and also automatically convert the mathematical model into FORTRAN code for conventional computation. A demonstration program was designed for modeling the Space Shuttle Main Engine simulation mathematical model called Propulsion System Automatic Modeling (PSAM). PSAM provides a very friendly and well organized environment for engineers to build a knowledge base for base equations and general information. PSAM contains an initial set of component process elements for the Space Shuttle Main Engine simulation and a questionnaire that allows the engineer to answer a set of questions to specify a particular model. PSAM is then able to automatically generate the model and the FORTRAN code. A future goal is to download the FORTRAN code to the VAX/VMS system for conventional computation.

Description: A method is proposed whereby the underlying domain knowledge is represented such that illustrative examples may be generated on demand. This method has the advantage that the generated example can follow changes in the domain in addition to allowing automatic customization of the example to the individual.

Description: Progress is reported on the development of SCOTTY, an expert knowledge-based system to automate the analysis procedure following test firings of the Space Shuttle Main Engine (SSME). The integration of a large-scale relational data base system, a computer graphics interface for experts and end-user engineers, potential extension of the system to flight engines, application of the system for training of newly-hired engineers, technology transfer to other engines, and the essential qualities of good software engineering practices for building expert knowledge-based systems are among the topics discussed.

Description: A Failure Modes and Effects Analysis workstation (FMESassist) was designed for use during the development of the space station. It assists engineers in the complex task of tracking failures and their effects on the system. Engineers experience increased productivity through reduced clerical loads, reduced data inconsistency, and significantly reduced analysis time. System developments benefit from a more thorough analysis than was available using previous methods.

Description: The C Language Integrated Production System (CLIPS) is a forward chaining rule-based language. The requirements necessary for an expert system tool which is used for development, delivery, and training are examined. Because of its high portability, low cost, and ease of integration with external systems, CLIPS has great potential as an expert system tool for delivery and training. In addition, its representation flexibility, debugging aids, and performance, along with its other strengths, make it a viable alternative for expert system development.

Description: CLIPS is a language for writing expert systems applications on a personal or small computer. Here, the CLIPS programming language is described and compared to three other artificial intelligence (AI) languages (LISP, Prolog, and OPS5) with regard to the processing they provide for the implementation of a knowledge based system (KBS). A discussion is given on how CLIPS would be used in a control system.

Description: The implementation of an inference engine for embedded diagnostic systems is described. The system consists of two distinct parts. The first is an off-line compiler which accepts a propositional logical statement of the relationship between facts and conclusions and produces data structures required by the on-line inference engine. The second part consists of the inference engine and interface routines which accept assertions of fact and return the conclusions which necessarily follow. Given a set of assertions, it will generate exactly the conclusions which logically follow. At the same time, it will detect any inconsistencies which may propagate from an inconsistent set of assertions or a poorly formulated set of rules. The memory requirements are fixed and the worst case execution times are bounded at compile time. The data structures and inference algorithms are very simple and well understood. The data structures and algorithms are described in detail. The system has been implemented on Lisp, Pascal, and Modula-2.

Description: The Diagnostic System Development Shell (DSDS), an expert system development shell for diagnostic systems, is described. The major objective of building the DSDS is to create a very easy to use and friendly environment for knowledge engineers and end-users. The DSDS is written in OPS5 and CommonLisp. It runs on a VAX/VMS system. A set of domain independent, generalized rules is built in the DSDS, so the users need not be concerned about building the rules. The facts are explicitly represented in a unified format. A powerful check facility which helps the user to check the errors in the created knowledge bases is provided. A judgement facility and other useful facilities are also available. A diagnostic system based on the DSDS system is question driven and can call or be called by other knowledge based systems written in OPS5 and CommonLisp. A prototype diagnostic system for diagnosing a Philips constant potential X-ray system has been built using the DSDS.

Description: A basic paradigm that allows representation of physical systems with a focus on context and time is presented. Paragon provides the capability to quickly capture an expert's knowledge in a cognitively resonant manner. From that description, Paragon creates a simulation model in LISP, which when executed, verifies that the domain expert did not make any mistakes. The Achille's heel of rule-based systems has been the lack of a systematic methodology for testing, and Paragon's developers are certain that the model-based approach overcomes that problem. The reason this testing is now possible is that software, which is very difficult to test, has in essence been transformed into hardware.

Description: The techniques described provide an effective tool which knowledge engineers and domain experts can utilize to help in evaluating and refining rules. These techniques have been used successfully as learning mechanisms in a prototype adaptive diagnostic expert system and are applicable to other types of expert systems. The degree to which they constitute complete evaluation/refinement of an expert system depends on the thoroughness of their use.

Description: Verification and validation of expert systems is very important for the future success of this technology. Software will never be used in non-trivial applications unless the program developers can assure both users and managers that the software is reliable and generally free from error. Therefore, verification and validation of expert systems must be done. The primary hindrance to effective verification and validation is the use of methodologies which do not produce testable requirements. An extension of the flight technique panels used in previous NASA programs should provide both documented requirements and very high levels of verification for expert systems.

Description: The NASA Lewis Research Center and the U.S. Army Aviation Systems Command share an interest in advancing the technology for helicopter propulsion systems. In particular, this presentation outlines that portion of the program that applies to the drive train and its various mechanical components. The major goals of the program are to increase the life, reliability, and maintainability; reduce the weight, noise, and vibration; and maintain the relatively high mechanical efficiency of the gear train. The current activity emphasizes noise reduction technology and analytical code development followed by experimental verification. Selected significant advances in technology for transmissions are reviewed, including advanced configurations and new analytical tools. Finally, the plan for transmission research in the future is presented.

Description: The morphology and properties of the Variable Polarity Plasma Arc (VPPA) weld composite zone are intimately related to the physical processes associated with the keyhole. This study examined the effects of oxide, halide, and sulfate additions to the weld plate on the keyhole and the weld pool. Changes in both the arc plasma character and the bead morphology were correlated to the chemical environment of the weld. Pool behavior was observed by adding flow markers to actual VPPA welds. A low temperature analog to the welding process was developed. The results of the study indicate that oxygen, even at low partial pressures, can disrupt the stable keyhole and weld pool. The results also indicate that the Marangoni surface tension driven flows dominate the weld pool over the range of welding currents studied.