Comparison of Subsonic LD for 993/94 SSTO
and 1994/95 RLVThe NASA ASDP has added a significant space component to the overall AEM program. Parts of the NASA ASDP project have been used in courses, in addition to the capstone design courses, to provide meaningful design problems. For example, the primary structure of the Mars habitation module was used as a class design project in one of the Department's structures classes. Student displays of the design class models are a visible expression of the impact the program has had in the AEM Department. A three meter long model of the single stage Mars transfer vehicle hangs in the stairwell just outside the Department's main office. Faculty, students and visitors cannot help but be aware of this senior class project. The current AEM design curriculum is now a vital and dynamic part of the students' undergraduate experience. This year the class was divided into two components, spacecraft and aircraft design and their projects are presented below.
The activities and projects this year centered around the Aircraft Team's participation in AGATE (Advanced General Aviation Transport Experiments). This design competition, jointly sponsored by the National Aeronautics and Space Administration (NASA) and the Federal Aviation Administration (FAA), is part of a major national effort to rebuild and revitalize the U.S. general aviation sector. Considering a number of revitalization goals, the competition provided design challenges in four technical areas: integrated cockpit systems, propulsion, noise and emissions, integrated design and manufacturing, and aerodynamics. The students' design will be critiqued by a panel of industry and government experts. The competition awards will be made at the Experimental Aircraft Association (EAA) annual convention at Oshkosh, Wisconsin, in July 1995. In this competition, general aviation aircraft are defined as: fixed-wing, single- engine, single-pilot, and propeller-driven aircraft for 2-6 passengers. They have a speed specification of 150-300 knots (177-345 m.p.h.) and a range specification of 800-1,000 nautical miles. Stressing innovation, national needs, and commercial potential, the competition asked that the preceding design challenges be addressed with an eye towards making general aviation flight easier and more convenient while improving safety, comfort, reliability, dependability, and performance.
With these challenges in mind during fall quarter, aircraft teams performed the initial conceptual design of two aircraft. Both designs showed innovation in such areas as overall configuration, aerodynamic design, powerplant choice, and advanced cockpit instrumentation. During the first half of the following quarter, the team consolidated its designs using the best areas of both preceding aircraft, and completed the conceptual design of their aircraft, the Wyvern.
Beyond the basic definition of a general aviation aircraft, the Wyvern had the following characteristics: four places including pilot, 250 knot cruise speed (287m.p.h.), a 1000 nautical mile range with a 45 minute reserve fuel supply, minimized pilot workload, and improved safety and comfort. Its conventional configurations utilize a canard, pusher propeller, and a diesel engine. Its glass cockpit and electronic engine and flight controls increase safety and performance while decreasing pilot demands and required experience. In addition to analyses of the plane's structure, weight, center of gravity, performance characteristics, stability, systems configuration, propulsion systems and avionics, extensive modeling of the aircraft was performed using the 3-D solids modeling program, Pro/Engineer. During winter and spring quarters, the aircraft team was broken into several project groups that focused on specific areas of the Wyvern.
Fly-By-Wire Systems Group: Jason Bergerson, Shawn Erieau, and Rick Kerner
This group worked on a simulation to show the feasibility of using a fly-by-wire controls system in the Wyvern. After investigating and comparing mechanical and fly-by-wire systems, they determined not only that the plane would require a fly-by-wire system due to its control methods, but also that such a system would be feasible. They then developed a short-period, vertical plane simulation of the Wyvern using the computer packages MATLAB and SIMULINK. By allowing perturbations such as canard movement and wind gusts, they determined that the Wyvern would be controllable using a fly-by-wire system.
Investigation of Computational Fluid Dynamics (CFD): Eric Engebretsen
During part of spring quarter, this group looked at the feasibility of performing CFD analyses on the Wyvern. CFD was judged important because it is a modern design tool that can yield detailed aerodynamic data around and upon an aircraft. This team investigated several methods of converting the fuselage of the plane, as created in ProEngineer, into a three-dimensional mesh suitable for CFD.
Wind Tunnel Model Group: John Alabach, David Graf, Glenn Lehrke, and David Swanson
Starting in the middle of winter quarter and continuing though spring quarter, this group constructed a 1/16th scale model of the Wyvern and tested it in the Department's subsonic recirculating wind tunnel. The model was constructed using a variety of materials and methods. Using the ProEngineer description of the airplane, two fuselage molds were manufactured on a CNC milling machine in the Department Shop. These molds were used to create the fuselage from two vacuum-formed plastic halves. The wing and canard were manufactured from PVC and aluminum, respectively, also using the CNC mill. The canards were designed to be adjustable to allow testing to determine the effects of varying canard angle.
The wind tunnel model was tested through a range of airspeeds, angles-of-attack, and canard angles. This data was needed to determine the plane's lift, drag, and longitudinal (pitch) stability characteristics. Using the departmental computerized data acquisition system, the data was collected, analyzed, and compared to the predictions made when the Wyvern was designed. The data revealed certain difficulties with the lift from the wings and the longitudinal stability at certain angles-of-attack. These problems were attributed to difficulties in the construction of the NLF (Natural Laminar Flow) airfoil section on the model and not specifically to the airplane design.
Integrated Cockpit Systems Group: Jason Bergerson, Steve Ford, Kristine Geske, Paul Tran, and Terry Wehrle
This group began in the second half of winter quarter by learning the program Alias Studio and applying it to the airplane models that were created in ProEngineer. During winter quarter, the group worked on defining and analyzing the cockpit area of the Wyvern.
Part of the group used the computer animation program Alias to analyze aspects of the physical environment of the cockpit. This program allowed the group to examine a computer model in ways similar to a full scale physical mock- up. Using Alias, the group analyzed the visibility from the cockpit, identified problems, and suggested corrections to the design. Also, the group examined the size and ergonomics of the cockpit by placing an animated figure within the aircraft. In addition, Alias proved useful as a tool for presenting the overall aircraft design.
The rest of the group reanalyzed and redesigned the cockpit systems. Thought was given to the development of a simple interface between the pilot and the aircraft systems.
BD-5 Control System Redesign Group: Randy Bierwerth, Kevin Joda, Ali Naughton, Jerome Socha, and PJ Vitoff
During winter quarter, this group left the AGATE design team and turned their attention to the department's BD-5 aircraft. This kit plane, which is still under construction, was the focus of the aircraft design team during the 1993-1994 academic year. The group undertook the task of designing, fabricating and testing a new control linkage system for the BD-5. They designed a system that moved elevator and aileron control from the original side-stick controller to a center-stick controller. Considering calculated loads and the Federal Aviation Regulations, parts were designed in Pro/Engineer, analyzed using finite element analysis of stresses in ANSYS, and fabricated in the department shop. The group then built a testbed for their controls system, installed it, and tested it to determine whether it met their design criteria.
Alan Estenson was the aircraft design teaching assistant.
Dramatic changes in the politics and economics of space utilization have motivated the need for the design of a new vehicle for launching satellites and supplying future space stations. Although the Space Shuttle has achieved great success over the past fifteen years as a versatile provider of the means for space science research, as well as satellite deployment, retrieval, and repair, it is time to utilize the engineering knowledge gained from the Space Shuttle to design a launch system that will make the U.S. competitive in the international space market. The large support infrastructure of the Space Shuttle, which facilitated its early success, now results in a per launch cost greater than $400 million. Because other satellite launch systems like the French Ariane rocket are less expensive (about $60 million per launch), the percentage of satellites launched by the U.S. has dropped from 100% to 30%. To regain this market loss, knowledge gained from the Space Shuttle program must be cost effectively applied.
NASA examined the future of its space program in the Access to Space Study. This study consists of three options: an upgrade of the current Space Shuttle program, a multiple- launch vehicle fleet consisting of both expendable rockets and small reusable multi-staged vehicles, and fully reusable launch vehicles. NASA concluded that single stage-to-orbit, reusable launch vehicles would be the most economical decision. This vehicle should also be a pure rocket incorporating near-term and advanced technologies. By developing a new launch vehicle with these specifications, annual costs can be lowered resulting in a smaller, overall investment.
With these thoughts in mind, NASA requested that students participating in the NASA- sponsored aerospace design class investigate a vehicle configuration capable of achieving these requirements. As part of this program, students are given the opportunity to communicate and consult directly with NASA engineers. NASA also provides many state- of-the-art computer programs to help carry out the analyses. The NASA/Marshall Space Flight Center (MSFC) is the University of Minnesota sponsor and has been extremely helpful. This year Marshall sent Systems Engineer Bill Pannell to advise the students early in the design process, and later sent Steve Cook (former U of M graduate and current Deputy Program Manager for X-33 Engineering Technology) to critique the design at the end of the year. Students concluded their design experience by presenting the design to engineers at NASA/MSFC. An account of the design process follows.
This year's conceptual design is a second iteration of the 1993-1994 SSTO (Single Stage To Orbit) vehicle design and is termed the Reusable Launch Vehicle (RLV). The SSTO is a single-stage-to-orbit, lifting body design. This type of design was continued in the RLV. The design team began the quarter by analyzing the SSTO and found that it contained two areas which needed major development. First, the SSTO was designed solely for hypersonic flight. Wind tunnel testing proved unacceptable subsonic performance for the vehicle. The data showed that the vehicle could only trim at negative angles-of-attack points with very little lift. This would make the vehicle incapable of a safe landing. Drawing on this data and advice from MSFC, it was theorized that a more delta-shaped vehicle might resolve this problem. The planform shape of both the SSTO and RLV are shown on the next page. The second problem of the SSTO is its extremely heavy modular tank and aeroshell structure. It was determined that this is unrealistic and that an integral tank structure is the only choice for reducing the weight of the vehicle.
The first step in the actual design process was to establish a list of requirements for the vehicle. The most obvious requirement is that the vehicle must have a primary mission of launching and retrieving commercial satellites, which stems from the above statements about the economics of the vehicle. For this reason, the vehicle needs to be designed with a $50-60 million launch cost as a goal. This will ensure market competitiveness. The vehicle must launch satellites to a 200 n.m. altitude at 28.5û inclination. This is the orbit that provides the best coverage of the US geography and allows the most leniency for achieving other higher inclination orbits from the US. By having a 9072 kg (20,000 lb.) payload bay with 4.6m x 9.1m (15ft x 30ft) dimensions, the vehicle will be able to carry two average-sized satellites each mission. Drawing on NASA's experience, as well as experience with the SSTO, it was determined that a more integral tank design should be used, not a modular design like the SSTO. The integral tank design uses the tanks of the vehicle as the supporting structure with very little truss work to carry loads. This should result in a large reduction in the total weight of the vehicle. Another requirement is that the vehicle have a delta wing shape, providing better performance for subsonic flight than the rectangular shaped planform of the SSTO. The last requirement isthat it must have a vertical take-off and horizontal landing or VOTHOL. The current state of technology does not allow for a horizontal take-off (this requires very high thrust air-breathing engines) and since the vehicle is a lifting body, the lift allows it to land horizontally.
To maintain focus on important tasks, the design team determined five goals to be accomplished by the end of the school year. These are outlined below:
Moving from this general list of requirements to an initial vehicle configuration was the next action. The design team proceeded by selecting the main engines, and estimating the aerodynamic characteristics for a delta-shaped vehicle. This data was used as input into a NASA/MSFC program called OPGUID used for trajectory optimization. From the trajectory, the amount of fuel necessary to achieve the required orbit was determined. The design process then became an iterative procedure. The fuel was positioned into the vehicle to see if it fit. If the fuel didn't fit, a slightly modified shape was designed, new aerodynamic coefficients were determined, and a new trajectory with different fuel requirements was calculated. This process eventually narrowed in on a RLV configuration (from HABP, a NASA/MSFC program used to calculate the aerodynamics).
The next step was to use ProEngineer (a solid modeling program) to realistically represent the vehicle and its components. ProEngineer was one of the most valuable design tools for the team. It was used as a layout tool to show how components of the vehicle interacted when assembled. By inputting material properties, data such as mass distribution and center of gravity location could be determined. These numbers were used by the design team in an iterative process of designing a component, modeling that component, checking the mass and its effect on the center of gravity, and then redesigning according to these effects.
With an initial baseline configuration designed, it was now necessary to go back and perform more in-depth analyses to determine if the design met its payload requirement. The class was split into project groups during the winter and spring quarters to carry out these investigations. The project groups included subsonic aerodynamics, hypersonic aerodynamics, radio controlled (R/C) model, and structures. The team also formed an animation group in order to realistically present their ideas to the outside world. The passages that follow contain more detailed accounts from each of these project groups.
The subsonic aerodynamics group found in early fall quarter that there is very little information available on the subsonic characteristics of lifting body vehicles. They concluded that the best way to obtain information was to complete a SSTO model and test it in the University of Minnesota subsonic wind tunnels. As was previously stated, they found that the SSTO had poor subsonic characteristics and attributed this to the rectangular platform. Thus, a delta shape design was established.
The subsonic analysis proceeded using two methods -- computational fluid dynamics (CFD) and wind tunnel testing. A CFD program from NASA Ames called Overflow along with a grid generating program from Mississippi State University called Genie++, was used to perform the CFD analysis. Since this was the first year that the class had used CFD, the main goals were to generate CFD models for the RLV and validate them through experimentation in the wind tunnel. Future classes could then build on this work by taking these validated models and incorporating them into the actual design process. The CFD grid was modeled by generating the surface of the RLV and then constructing a grid ten body lengths in front and behind the RLV. Appropriate boundary conditions were applied and the geometry was processed on a Silicon Graphics Indy computer until the solution converged.
The next step was to validate the results from the CFD code by testing the vehicle in the U of M subsonic wind tunnel. The wind tunnel model was created by taking the ProEngineer geometry for the body and fins of the RLV and sending it to the computer is numerically controlled (CNC) milling machines in the AEM Machine Shop. The model was milled and the fins and flaps attached. It was then placed in the wind tunnel and tested for various body flaps and fin flap deflections.
Comparison of Subsonic LD for 993/94 SSTO
and 1994/95 RLV
Comparison of Wind Tunnel Results to
Overflow
This data also compared well to the CFD predictions. A typical graph is shown above right. This is a comparison of the coefficients of lift for CFD to the wind tunnel.
Further investigations are being performed into the RLV's controllability and dynamic stability, through the use of a radio controlled model. This large 1:27 scale model was created by a three axis CNC milling machine provided graciously by Milltronics of Chanhassen, MN. A picture of the milling process is shown below.
Construction of Radio Controlled Model
R/C Model in Flight
The model was created approximately six feet in length to provide kinematic similarity to the actual vehicle. The vehicle is equipped with sensors for determining velocity, angle-of- attack, control surface deflections, and an on board computer to record this information. Currently, the model is undergoing preliminary flight tests without the electronics. The model is being launched with a Hi-Start and is achieving about five feet in altitude. A picture of one launch is shown on the previous page. As the flight testing matures, a rocket engine will be used to launch to higher altitudes which will enable more in-depth experimentation to be undertaken.
Hypersonic CFD was carried out to give re-entry aerodynamic coefficients, as well as a temperature distribution, for the vehicle at a critical heating condition. The grid used to perform this analysis, along with a typical solution is shown below.
RLV Hypersonic CFD Grid
Hypersonic CFD Solution
This CFD analysis will be compared to data obtained from a hypersonic wind tunnel model that is being built. This model is being constructed by the students in the AEM Machine Shop with the help and expertise of the Shop Manager Research Engineer, Dave Hultman. To create this model, the ProEngineer geometry of the RLV was exported to the Machine Shop and sent to a CNC milling machine, much like the subsonic model. Pressure taps will be added later to give direct measurements to correlate with the CFD solution. The model will be tested at the NASA/MSFC Trisonic Wind Tunnel later this summer. The data obtained from this experimentation will be used to validate the CFD code and provide insight into areas that would be too difficult to implement in the CFD model, such as flap deflections.
The design goals of the structures' team were to provide stress and failure analysis for the major components of the RLV, namely, the fuel tanks and skeletal truss. The tanks form to the outer skin of the RLV, occupying most of the volume. Therefore, it was necessary to place the lightweight skeletal truss between the various tanks to support the inertial loading of each tank and connect to the RD-704 engines. A preliminary conceptual design of the integrated tank-truss structure is shown schematically below.
Integrated Tank-Truss Structure
The total weight of the tanks and the truss had to be lower than 70% of the dry weight at landing for the RLV, so the primary design goal was to minimize the thickness of the tanks as well as the cross-sectional area and number of truss spars, without sacrificing safety and reliability. The dry weight was calculated from the trajectory analysis performed by the orbital mechanics discipline. The tanks were characterized as semi-integral tanks, designed to withstand thrust loads at take off, as well as the aerodynamic loads. Structural analyses was performed using the solid modeling program ProEngineer and the finite element mesh program Pro-MESH that interfaced to ANSYS. Loads and boundary conditions were also applied in Pro-MESH.
Stress Distribution for RLV LOX Tank
The tanks were designed to withstand the stresses encountered during maximum inertial loading during ascent. Loads applied on the tanks include internal and dynamic pressure, and inertial loading. The stress distribution on the liquid oxygen tank during this critical loading condition is shown qualitatively below.
RLV Truss Structure
The truss structure for the RLV (above) was also analyzed at this critical loading condition. Particular finite elements had to be placed on the truss structure to model the inertial loading and stiffness of the fuel tanks. The truss was refined over several iterations, and in the end, the minimum weight calculated for the truss and tanks exceeded the allowable dry weight. Future design considerations will investigate eliminating the truss all together and designing a fully integral tank structure. In order to communicate these ideas to the outside world, the design team developed an animation for a typical mission scenario of the RLV, and including the team's concepts for ground operations, launch, satellite deployment, and landing. This was accomplished with an animation program called "Alias Studio."
The detailed analyses performed through the use of the three physical models (subsonic, hypersonic, and R/C), the accurate ProE models, the CFD analysis of the aerodynamics, and the computer-aided structural analysis has lead to a very realistic vehicle design. Aerodynamic studies have shown that the design is both controllable and stable in all flight regimes, and has the lift to drag ratios necessary to fulfill both cross range and safe landing requirements. This was a major objective of the class.
There are still many areas that need refinement. Currently, the structure of the vehicle is much too heavy. Future work should investigate using a fully integral tank structure (no truss) to lighten the weight of the vehicle. Preliminary CFD results show high temperatures on the sharp leading edge of the RLV. The radius on this edge should be significantly increased. Finally, a more detailed look into the operations of the vehicle is necessary if it is to demonstrate economic viability including more involvement of a commercial enterprise such as Northwest Airlines. While these areas still need future refinement, the design team has established a comprehensive process that includes accurate physical modeling and integration, aerodynamic analysis using wind tunnel experiments and CFD, and a computer-aided structural analysis.
Fred Moeller was the Spacecraft teaching assistant.
The design class students spoke with a number of local elementary and high school students about Aerospace Engineering Careers, NASA, Space Shuttle, the University of Minnesota, AEM classes, Design, and CAD tools. Some students also participated in the Mid-Continent Space Development Conference, February l7-l8, 1995, in Iowa. The conference was sponsored by the Iowa Space Grant Consortium and brings many recognized leaders in the aerospace industry to the Midwest to talk about a variety of aerospace topics. Students participating in the conference were Martin Annett, Jeffrey Baker, James Browing, Shawn Dockter, Kelly Elblein, Jeffrey Jackson, Brandon Powell, Laura Sachi and Daniel Scharf.
NASA decided to change its annual Advanced Design Program (ADP) Summer Conference format from a single conference to multiple mini-conferences at each of its participating centers. Our students, Martin Annett, Brandon Powell, Kortney Rivard, Michael Stram and Brian Tutt headed off to the conference at the NASA Marhall Space Flight Center (MSFC) in Huntsville, Alabama, June l0-l3, 1995. A special presentation was given by our students to the NASA/MSFC engineers that they were working with during the year on their Lifting Body Reusable Launch Vehicle design. Other schools participating in the conference were: Central Florida University, Georgia Tech University, Kansas State University, Vanderbilt University and West Virginia University, and Florida A&M/FSU. Students also went on several tours: NASA MSFC Trisonic Wind Tunnel, NASA MSFC Productivity Center, Huntsville Space and Rocket Center, IMAS Movie: "Destiny in Space and the Marshall Space Flight Center."
Instructors for the course were Professors Andew Vano, Yiyuan Zhao and Mike Meixel.