Software
This is a very wide subject matter and I am not sure how well I'll be able to cover all the available material. My company of course uses many forms of software in our own aircraft design/development work and that will of course be listed and reviewed however, there are many other packages available both, commercial and freeware, much of which is unknown to me. Therefore, as with the references above, I will from time to time, be depending on the HBA's membership to supplement the information herein in order to provide the readers with a more complete listing. As before, any supplementary information should be sent my way through private messaging within this board.
As software has advanced and matured, a fairly substantial number of folks today appear to be of the opinion that we should by now have the ability to develop a computer based system that essentially becomes our own private engineering team, allowing even the amateur to design a safe and functional aircraft with only a few basic inputs. While this is an interesting goal for someone to try, all can be assured that this type of "expert system" has as of yet not been developed. Although the analysis of airplanes has advanced, the software discussed in the following sections can really be thought of as nothing more than an advanced tool set - a tool set that still must be understood, along with the sciences of flight, structures and overall configurational design. Even an advanced piece of software will still not enable one to understand the design of an airplane any more than a hammer and tape measure will enable an apprentice carpenter to understand the construction of a house. The designer still needs to understand the technology of the design process in order to use these tools in an effective and realistic manner.
No, this does not require a university level education in engineering (although that certainly doesn't hurt) - there are many here who are essentially self taught, doing the job of research and asking a lot of questions in order to teach themselves the proper techniques and forms of these disciplines in order to enable themselves to design that plane they've maybe dreamed of for many years. The software may help, but in of itself, is no substitute for that hunt for information.
The other consideration one must ponder herein is simply that the true value of the software is, for the most part, directly related to the cost. With only one or two exceptions, the more expensive the package the better results it will deliver - there really are no cheap shortcuts. However, by the same measure, the more expensive the software, often times the more complex it is. But then in this realm, the better the results, the more complex the program must be in order to do its job. Here is where every designer must balance between making that purchase and then going through the learning curve, or simply hiring a consultant to cover the bases. Given the prices of some of this stuff, the latter often seems to be the best solution.
Software - CFD
CFD simply stands for computational fluid dynamics. The flow analysis software is usually fairly complex, requiring specific experience and knowledge to not only properly model and setup the analysis but to also get meaningful results. Without said experience, CFD can also stand for confusion, frustration and despair.
There are several versions of this type of software ranging in cost from web based freeware to professional packages costing upwards of $50,000. The more expensive, the better the software and the more dependable the results, assuming of course the operator has specific knowledge and experience to use it. CFD works in a similar manner to a finite element modeler in structural analysis - in essence the shape to be analyzed is divided up into a multitude of small elements over which a flow is superimposed. The software then calculates a series of physical variables that act on said element as a result of that flow, the results of which can then be manipulated to get various forms of data ranging from pressures and forces, to design coefficients. But it must be understood that CFD is really not a design tool - it is more a tool for optimization.
A few years back there was an aero analysis presentation that was sponsored by United Technologies. The speaker was one of the principal designers for the turbine arm of the company, although he also had a background in airframe applications of the solver technologies. One of the more eye opening parts of the presentation was a point where he indicated that for the most part, unless the analysis is conducted properly and by specifically trained individuals, they consider the vast majority of CFD results no better than clown puke. In other words, the results are filled with pretty colors that are often arranged into an impressive presentation, but outside of that, because of the way the CFD software was used, the results have no dependable use. In his experience he felt that it is critical to understand that at no point is CFD a design tool, nor is it really an analysis tool. It is more than anything else an optimization tool, a job it cannot do without some form of accurate environment basis and hardware based verification.
In the design of a new engine, the vast majority of the work is done by experienced engineers, the results of which are tested along the way as the engine program matures. The engine is then built (usually based on an advancement of an existing engine core, or similar) and tested in available facilities. This is then followed by a CFD model, the environment of which is massaged and tweaked until the CFD results match those of the hardware tests. Only then is the mathematical geometry model modified, in small increments, to generate a new configuration or geometry that advances the propulsion product.
Over the years I come across several very experienced aero types who specialize in CFD modeling and analysis. And interestingly enough, virtually all agree with that statement and approach. True, I have not heard the others express it as colorfully as at the UT presentation, but in general the identical sentiment comes out over and over again - CFD is not a design tool and unless you are very well versed in the technology and in its application, depending on the results may simply lead you down the wrong path. Yes, it may be fun to play with, but then you're just playing, not designing.
In short, there is no software that will do the airplane design for you, aero nor structural. There is no shortcut to knowledge and experience. We already have more than enough design information publicly available to develop a safe airplane, even a very high end one. The trick then is learning that and not trying to find shortcuts. However, for those still interested in what's available, the following is a short list of the more mainstream packages.
VSAERO - Desktop CFD - VSAERO couples integral methods for potential and boundary layer flows for low runtimes (a complete Boeing 727 in 300 seconds). Flowfield properties are computed for off-body velocity surveys and on/off-body streamlines. The ability to calculate internal and external flows, non-uniform inflow and body rotation, makes VSAERO applicable to fluid flow problems in aerospace, automotive and marine engineering. Special purpose modules like FSWAVE and ROTOR expand VSAERO's simulation capabilities to include nonlinear hydrodynamic wave effects on ships and helicopter rotor/ fuselage interactions. Zonal coupling to Navier-Stokes codes is also available. Running on a wide variety of computers, from Cray Supercomputers to desktop PCs, VSAERO is used worldwide. VSAERO has been used in the development of Rutan Voyager and Beech Starship aircraft, the Stars and Stripes racing yachts and the Sunraycer solar automobile.
VLAERO+ - VLAERO+ is a planar vortex lattice method for the aerodynamic analysis of subsonic and supersonic aircraft configurations. With it's own GUI, VLAERO+ is ideally suited for the preliminary design environment where it can be used to quickly produce loads, stability and control data. Thousands of calculations have demonstrated that VLAERO+ is not only extremely simple to use but is also highly accurate within the limitations of the governing equations. Geometry is represented by a series of trapezoidal patches. Camber and twist (including airfoil sections, wing twist, and control surface deflections) are easily specified using common aerodynamic design parameters. Output data are clearly tabulated, including interpolated geometry, surface pressures, force and moment coefficients, and distributed loads.
VLAERO+ includes a graphical user interface for geometry creation, program execution, and solution visualization. Intuitive toolbars provide geometry creation and editing functions. Component ordering and input file creation are performed transparently without the need to edit fixed format text files. Airfoil data management and specification are greatly simplified. Execution of the flow solver is controlled by the interface, and the results are processed for visualization.
Features available from the graphical interface include:
Toolbar for quick intuitive access to file and visualization functions and asymmetric models.
Display window for quick visualization of geometry, airfoils, and flaps.
Geometry editor allows addition, deletion, and sorting of model components.
Component operations for translation, scaling, and stretching of components.
Control surface editor allows graphical specification of control surfaces.
Wing editor for quick generation/modification of model geometry using standard wing design data.
Advanced functions include viscous correction calculator, ground effect model generator, and batch processing.
MGAERO - Although this software is often advertised for application to aerodynamic preliminary design, it must be understood that it is still a fairly complex (and expensive) piece of software that requires a fairly detailed level of knowledge in order to obtain useful data. However, in the terms of mainstream aerospace applications, it can be said that this is one of the more useful pieces of software for running quick and relatively accurate trades. The Cartesian Euler code allows you to rapidly model and analyze the most complex configuration. Cartesian embedded grids simplify grid generation and automatic component intersections simplify geometry definition. Multi-stage Runge-Kutta integration with multi-grid acceleration yield an efficient solution on Unix and PC platforms (not for Windows machines). The user can create geometries as sectional data, generate wireframe components, or provide an IGES file from your CAD system. All these can be used to develop MGAERO input.
MSES/MISES - Whether you need a single airfoil section, multiple sections for a highlift system, or an airfoil for a cascade, this suite of codes allows you to quickly translate your design requirements into an optimum configuration. Two-dimensional codes driven with X-Window GUIs allow rapid geometry changes, parametric flow studies, and configuration development; MSES for multi-element airfoils in subsonic and transonic flow; MISES for airfoils in a transonic cascade. Both codes offer inverse design, forced and natural transition, and direct and inverse interactive boundary layer methods. Grid generation is automatic using a streamline coordinate solution, allowing the grid to adapt to the evolving flow field.
NSAERO - The answer to all your fluid flow questions in a single code. Focus on problem solution and not on multiple flow code inputs. Get one software product - NSAERO - for all flow speeds and Reynolds numbers, including gaseous combustion, Rapid problem set up; flexible and general boundary conditions; structured, unstructured and hybrid grids; and many turbulence and chemistry model options give you access to the most complex applications.
OMNI3D, our interactive 3D data visualization tool and TECPLOTTM, the industry standard visualization software, are recommended with NSAERO to give you immediate feedback from results and flexible display of flow properties. NSAERO, TECPLOT, and our GridgenTM preprocessing code provide you the ideal software system for fluid dynamics analysis.
TECPLOT is a trademark of Amtec Engineering, Inc., Bellevue, Washington. Gridgen is a trademark of Pointwise, Inc. , Dallas Texas.
USAERO - Challenged by relative motion aerodynamic or hydrodynamic simulations? Multi-store release from a complex aircraft? Trains passing in a tunnel? Ships with rotating propellers operating near the free-surface? USAERO is the engineer's choice for these and other transient calculations. USAERO's unique coupling of potential flow and boundary layer methods with a time-stepping procedure for arbitrary motions assures timely and cost effective assessments of unsteady surface pressures and loads.
USAERO calculates the transient aerodynamic characteristics of complex configurations in arbitrary motion. It is based on a time-stepping procedure, which allows relative motions of configuration components. As integral potential flow and boundary layer methods are the basis of this CFD software, the aerodynamics solution is only required on the boundary surfaces for each time step and requires only surface meshing.
Furthermore, while components or bodies may involve relative motions, no regridding is required by the solution scheme. Because of these features, USAERO supports practical engineering solutions for such problems as maneuvering aircraft, formation flying, aircraft stores carriage, gust response, rotor/body interactions, and train passing and tunnel entry. Special application modules, FPI and FSP, couple with USAERO to provide, respectively, flight-path integration calculations with six degrees of freedom and ship nonlinear free-surface simulations.
XFOIL - Although not really a CFD tool, it is used widely enough and given the headings used, seems best to fit here. XFOIL is an interactive program for the design and analysis of subsonic isolated single-segment airfoils. Written by MIT professor Mark Drela, it is viewed as one of the better section design/analysis packages available.
XFLR5 - XFLR5 uses XFOIL as its computation kernel and adds a graphical user interface for Windows operating systems. You still need the XFOIL manual to find your way around. XFLR5 also offers a 3D wing design capability, using two different calculation schems. The one similar to MIAReX uses the built in XFOIL kernel to determine local wing section properties. MIAReX is a calculation method for Xfoil and multi-airfoil wings, based on formulae developed by James C. Sivells & Robert H. Neely in NACA TN-1269 (1947). Basically it takes 2D wing section data and integrates the various 2D section properties across the span to arrive at a 'semi-3D' solution.
AVL: Athena Vortice Lattice Method - Yet another design program by professor Mark Drela and Harald Youngren. AVL is an extended vortice lattice method (VLM) software that supports aircraft configuration development by offering aerodynamic analysis, trim calculation and dynamic stability analysis, among other things.
Personal Simulation Works - PSW is a streamline-body design and analysis package for the PC, based on what's generally referred to as a panel code. It includes programs for surface definition, CFD flow analysis, and visualization. Its three principal elements are Loftsman/P, Cmarc, and Postmarc. An optional fourth component, Digital Wind Tunnel, performs stability and control analysis for aircraft. The program was written up in the December 1995 issue of Sport Aviation and has seen some moderate level of success. Current status is unknown as the web site does not seem to have been updated in more than seven years.
Software - Airplane Design
This is sort of an interesting subcategory in that it represents software that's being presented or marketed as an automated design system. The interesting thing is that in several cases the software may come close however, the advertising tends to skim over the limitations or shortcuts that were incorporated in order for the tool to be more user friendly at the amateur level. While these simplifications do not invalidate the program, they do impose certain loose tolerances that cause less than precise results. That does not necessarily make it a bad tool, just one that has to be used with a certain amount of care and restraint.
Airplane PDQ - AirplanePDQ (by DaVinci Technologies) is a conceptual/preliminary design tool for light homebuilt and general aviation aircraft. It is specifically designed to be intuitive and easy to use by amateur airplane designers. AirplanePDQ includes tools for doing a first cut analysis of aircraft performance, handling, and stability and control as well as a CAD component for developing your aircraft drawings. Initial airplane sizing is done rapidly using a wizard-based approach. The designer enters a few basic performance requirements, chooses from among a wide variety of aircraft configuration options (high wing or low wing, tricycle gear or tail-dragger, canard, three-surface, pusher, twin, etc...) and then the wizard performs sizing calculations and generates an initial three-view drawing of the aircraft based on the designer's inputs. Once the initial drawing has been generated, the designer can modify the design using the CAD engine to meet his or her needs. The analysis tools are then used in an iterative fashion to verify that the design is meeting the performance goals and determine what further changes to the design are needed for safety, handling qualities or performance reasons. The analysis tools not only generate detailed performance estimates, they also guide the designer in sizing and adjusting the configuration of the aircraft to help ensure that the design is safe and practical. Although it does have some level of limitations, this is about as close to an expert system as we have int he light plane industry. For the price, it is a good beginning tool for those wishing to experiment with their ideas. However, based on input from several members of this board, it is unclear whether this software is still being developed and/or supported in any way. Web site has not been updated since 2007.
X-Plane - I am generally against using this program for any real world design or analysis work since this is one of the best examples of the computer theory of garbage in-garbage out. To date the program tends to be used by individuals with little or no actual design background, who seem to be of opinion that this is an expert system that will provide them with everything they need to develop a safe plane. And of course nothing could be further from the truth.
In order to provide a more compete answer I made a number of phone calls to those who are familiar with this program and the science behind it. Furthermore, I’ve also contacted several individuals who are intimately familiar with CFD modeling and those who are or have been involved in developing simulator software, be it for the civilian/amateur market or the higher end, mainstream aerospace level. In inquiring whether X-Plane is specifically useful for analysis or dependable flight prediction, the answers were pretty interesting, if for no other reason than their consistency.
The following is a paraphrased version of the responses I’ve been able to gather. I’ll try to do this in two parts: The first part will concentrate on what a flight simulator actually is and some of the functionality behind it; the second part will discuss the math basis involved and the resulting limitations.
As the name implies, the purpose of the modern computer flight simulator is to give the operator a fairly seamless representation of an airplane in flight. Furthermore, many flight simulators like this also incorporate various subroutines that enhance said experience with systems functionality so that the would be pilot not only flies the plane but also operates the equipment within. As such, the focus of the program is on the environment inside and outside the airplane and that is of course why we see the emphasis of many of these pieces of software at simulating not only the cockpit but also a seemingly realistic set of scenes depicted outside the window.
But in order to be truly functional and realistic, the software must be able to do all this in real-time at a rate of many cycles per second, in order for the flight experience to be smooth (I believe X-Plane updates at 15 times per second). To allow this to happen, the math behind the program must be relatively simple since anything complex would bog even the most powerful desktop workstation nearly to a stop. For this reason the machine’s ability to make analysis decisions is extremely limited – in real terms it’s more like a program that has a particular and almost predictable set of reactions to any particular set of inputs. One can think of this as “Action A” causes “Reaction B”.
One gentleman I spoke with equated this action/reaction to sort of a clockwork mechanism with a multi-speed transmission – you move one particular lever in this manner and the mechanism will react (maneuver the plane) in a prescribed way. There is no thinking or analysis in this act/react scenario, just a physical reaction to a set of preprogrammed variables (gears designed during the input phase). And this is where the software falls apart as a design tool. The variables and constants that one would be normally looking for in a typical analysis package (like CFD software) must actually be predefined by the builder of the airplane model. If one inputs the incorrect variables, the results will be meaningless since the program has no way of knowing or telling you whether those values are accurate or not, or even correct. Without having the ability to judge the correctness of input values, the program has demonstrated that it will allow fly nonsensical configurations that normally would not even get off the ground, let alone stay in the air. As such, if one does not have a background in aircraft design, the likelihood that the configuration the designer has input will behave in the same way in the real world, is pretty low.
The other limitation the program has for the purpose of design is the math that makes it work. The math behind the functionality is called blade-element-theory. This is sort of a macro element methodology that was developed to analyze propeller blades and propeller performance. Given the very finite expanse of a typical prop the methodology works since the information covers a very finite structure operating in a finite space and volume of fluid. However, when considering multiple blades the theory failed to produce accurate results so in order to get a better performance model, the blade element theory was combined with momentum theory, which was able to better account for and model air flow behavior in front and behind the blade. All this works quite well for the finite expanse of a typical prop however it was never intended for analyzing a full sized plane. Furthermore, to keep the mathematics relatively simple, the X-Plane software does not incorporate the momentum theory thus making any information that one might actually get from a typical run a bit suspect.
The blade-element theory has another limitation and that is that it cannot very well deal with non-wing bodies and bodies within wings (like flying wings or blended wing-bodies), as well as numerous other geometry shapes. For this reason some folks who are playing with configurational variations have encountered some fairly inconsistent results, despite making only very slight changes in their model geometries. Making a minor geometric tweak and slightly moving a fin and getting a 50% drag rise is one of the examples seen recently, even by folks who use this board.
And of course we can further nitpick by bringing up things like boundary layer and/or viscous effects, wakes, propulsion effects, mass and momentum issues, compressibility, induced velocities, etc., most of which the program has no way of addressing or dealing with. Taking all the above into account, we can only conclude that X-Plane is a very good flight simulator for the pilot but is far short of what one could call an engineering design tool.
So then the question comes up regarding the FAA’s certification of the program: “If this is such a basic tool and so inaccurate, how can it get certified by the FAA for pilot proficiency training?” That one I had to do a bit of digging on but the answer is simpler than you think. First off, for basic level flight and simple maneuvers the software is satisfactory in presenting the pilot with the experience of flying, navigating and operating the aircraft. For that you don’t need the sophistication that you’d need for any form of dependable analysis. Furthermore, to get the certification, the program variables are tweaked so that the aircraft’s behavior as the pilot sees it, conforms to the Pilot Operating Handbook. In other words, they input the airplane model but then go back in and start tweaking and prodding the variables until the airplane behaves just like the manual says it should. In this way it meets the FAA’s requirements and the pilot’s, so certification is feasible.
So given all this, how dependable can it be? Well, it depends. It depends on what information you’re trying to get from it, how good you are at aircraft design and programming in the correct variables, and what part of the flight envelope you’re looking at. When I asked those I contacted this question I got a variety of answers so the following is sort of an average. The numbers assume the designer is
very experienced however is hampered by the finesse or accuracy of the program’s math. The percentages are a best guess at how close the model and its results could be to the real word hardware. They can easily be worse but are unlikely any better.
For conventional configurations (Cessna, Piper, etc.):
Level flight - 90%
Maneuvering - 75% to 85%
Abrupt maneuvers or near stall – 60% to 70%
Stall, unusual attitude, corners of operating envelope - less than 60%
Unconventional configurations (highly dependent on layout)
Level flight - 80% or lower
Maneuvering - 70%
Abrupt maneuvers or near stall - 60% or less
Stall, unusual attitude, corners of operating envelope - low (no one gave me any guesses here)
In short, the consensus among those I spoke with is that X-Plane is a very good flight simulator and obviously has a loyal and enthusiastic following. It is however not a design tool nor should it be thought of as one unless you are using it for your own education or entertainment. If you wish to model your own design, do so with the understanding that this is only a potential representation but none of the information you gain from said exercise should be depended upon, used or applied in any way towards developing an actual flight vehicle.
ADS (Aircraft Design Software) - Developed by OAD. ADS is particularly suitable for aircraft designers, homebuilders, university staff and students, as well as for pilots and future aircraft owners. If you are a homebuilder, ADS can help you design or modify a light aircraft. You do not wish to spend time learning to use complex software and you are looking for something user-friendly. ADS makes it possible for you to size the dimensions of your future aircraft, without any fuss and at a very low cost, while ensuring the greatest chance of success.
If you are a professional aircraft manufacturer and you want to design an aircraft, you need an efficient, fast and accurate tool to analyse the market and design the best product in the least possible time; the product which best meets your specifications. ADS enables you to achieve optimisation, i.e. to find the best configuration of the aircraft at the planning stage so that it meets the requirements of the specifications with maximum efficiency and in as short a time as possible.
If you are student or teacher, ADS is a tool perfectly tailored to your needs. You want to understand and explain. ADS is of considerable assistance in providing a better understanding of the aircraft design process, viewing the effects of a parameter variation on aircraft geometry and performances, apprehending aircraft design in a comprehensive way, teaching both the interest of the analytical approach where everything is broken down to the smallest detail and the synthetic approach where all the details form a whole.
I've been following the development of this tool for some time and so far have been pretty encouraged with what I've seen. No, I do not have a copy as of yet but may invest in one to use as a reference and check to the software we've developed in-house. The software is available in several versions, a couple on the bottom end being quite affordable, yet with a nearly full functionality.
Advanced Aircraft Analysis - Developed by the DAR Corporation: Advanced Aircraft Analysis (AAA) has been making inroads in the aerospace industry as a tool for aircraft design, stability, and control analysis software. The program provides a framework to support the iterative and non-unique process of aircraft preliminary design. The AAA program allows students and preliminary design engineers to take an aircraft configuration from early weight sizing through open loop and closed loop dynamic stability and sensitivity analysis, while working within regulatory and cost constraints.
AAA is used for preliminary and Class II design and stability and control analysis of new and existing airplanes. Class II design incorporates detailed weight & balance, aerodynamics, stability & control calculations including trim analysis and flying qualities used in conjunction with the preliminary design sequence. Class II design accounts for power plant installation, landing gear disposition and component locations on the airplane. Furthermore, it uses more sophisticated methods than Class I and requires more detailed information of the airplane to be known. The accuracy of Class II methods is therefore greater than Class I methods.
Advanced Aircraft Analysis can be used for small (civil), military and transport airplanes. The program is designed to assist in the design learning process while reserving for the user the individual creative judgment which is essential to the process of airplane design.
The design methodology used in Advanced Aircraft Analysis is based on Airplane Design I-VIII, Airplane Flight Dynamics and Automatic Flight Controls, Parts I and II, by Dr. Jan Roskam, and Airplane Aerodynamics and Performance, by Dr. C.T. Lan and Dr. Jan Roskam. AAA incorporates the methods, statistical databases, formulas and relevant illustrations and drawings from these references.
The last sentence is critical to understanding this tool in that it virtually requires that one takes the university courses that use Dr. Roskam's books in order to gain a functional understanding of the methodologies used. Some years ago we purchased this software for our company and found it extremely cumbersome for the design process. The software seemed quite useful for analysis of existing configurations but the working requirements were such that the program required input that one would normally assume an analysis program would generate, not need supplied.
I have known a few individuals that give this software good marks but all admit that it is much more an analysis tool and that the learning curve is relatively steep if one is not familiar with Dr. Roskam's curriculum. Personally, I probably would not recommend this to anyone, especially for homebuilding applications.
Software - Structures
Software - CAD