 |
| Part 1: History of Flight Simulation |
| |
"Our experience is composed rather of illusions than of wisdom acquired."
|
Joseph Roux, Meditations of a Parish Priest, 1886
|
| |
Synthetic flight training has been around for quite a while. Flight Simulators were first introduced during World War I, but it was not until World War II that their use became widespread. Obviously, they were rudimentary devices and were employed to teach simple procedural techniques.
It was however, the advent of the computer that, in the 1950s, gave rise to simulators that were to become regarded a more serious means of synthetic training. Unfortunately, these devices displayed all of the inherent problems encountered by computers of that generation; they were large, slow, unreliable analogue machines that suffered greatly from data drift, and they did not have motion. As a result, good simulation of the aviation environment was still not available.
During the 1960s, the inception of the digital computer began to allow greater fidelity in the simulation of the aviation environment. These computers offered greater computational speed, higher degrees of accuracy as well as vast improvement in reliability. The leap in technology at this time allowed attempts to rectify what was until this point, seen as the major deficiency of flight simulation – the lack of a visual scene.
One of the more interesting solutions to this problem involved a closed circuit television system moving over a fixed 3D-terrain board. The computers in this style of system would move the camera over the terrain board in response to the selected flight path and expand the image to a more or less appropriate size for aircrew viewing. Obviously there are many limitations to this design, not the least of which is the finite dimension of the terrain. Also, consider the impact on the training outcomes when you could be flying low level, ingressing towards the target to drop your load of 1000 pound bombs, only to be confronted by the carcass of a dead fly that on first inspection appeared as if straight out of "Land of the Giants"! Interestingly, the F1-11 crews at Amberley were still training on this style of device as late as the 1990s.
It was not until the 1970s that aircraft manufacturers and their customers generated the need for realistic, high fidelity synthetic training devices. The advent of large, wide bodied passenger aircraft and the associated expensive training bills demanded cheaper forms of training. Full Motion flight simulators answered the need. Faster, more powerful computers now meant more accurate and faithful reproductions of the flight models that the aircraft manufacturers were now providing, six degrees of motion platforms, computer generated visual images and appropriate aural and proprioceptive cuing were all possible. This all added up to an environment in which nearly all aspects of Fixed Wing commercial aviation could be trained and tested with the appropriate degree of fidelity.
Since this time the basic configuration of flight simulators has not changed, as a typical device it still has all of the assets mentioned above. However, what has changed is the power and expense of the computers that are now used and the methods of delivering the required elements that make up the training environments. In fact, the environments are now becoming so realistic and at such relatively cheap costs that they are able to support the exhaustive training demands of military customers, in particular those demanded by the helicopter environment. This was an area of flight training that until the mid 1990s was not possible to the degree of fidelity that was being provided within the Fixed Wing community.
The Illusion
Flight simulation is all about illusion, that is, convincing the aircrew they are using a device that it is not tethered to the ground and that they are actually flying a real aircraft. But how can a synthetic aviation environment ever be realistic enough? The answer lies in understanding the type of training that you wish to provide and in considering whether the "real" environment is actually needed to achieve the desired training outcomes. For example, is the aircraft really needed to train cockpit procedures? Probably not, therefore, once a set of skills has been identified, it is simply a case of creating an appropriate environment in which to train, such that the skills may be transferred without hindrance to the aircraft at a later date.
|
| |
Consider also that total fidelity of a real world environment may not necessarily be required to train and perform a given sequence or manoeuvre. In fact, if pursued this "total" solution would be unreasonably expensive, thereby negating one of the prime motivations of simulation - cost. |
| |
Thus, the illusion is achieved by the creation of an appropriate environment for a particular skill set. To achieve this illusion for flight the cockpit, controls, switches, lights and interfaces must function as per the real aircraft. Sounds heard in the aircraft have to be faithfully reproduced and the handling characteristics have to be accurate. The current generation of flight simulators also supports the illusion by providing; wrap around day/night, colour, photo-realistic visual scenes that can be viewed from any seat in the cockpit and motion bases capable of reproducing dynamic in-flight and on-ground accelerations.
For example, imagine that the simulated aircraft is overloaded and pulled to the hover. The aircrew should not only hear the rotor droop and the engines spool up, but also see the appropriate indications on the instrument panel, feel the correct aerodynamic responses and the extra amounts of control inputs required to complete the transition to flight.
If the conduct of the sequence has been performed without any anomaly or distraction then the creation of the illusion has been successful and it is reasonable to assume that the skill would be successfully transferred to the real environment.
Once operational, the FF&MS will have been accredited the most realistic level of illusion, or to Level 5. This means that for a certain set of sequences the device may be used for training without the need to use the aircraft. This is called a Zero Flight Time (ZFT) Simulator. Obviously, not every sequence that is possible in the aircraft will be possible in the FF&MS, but its design and extreme level of fidelity will enables the conduct of the vast majority of flight and mission sequences, hence its level 5 status.
FF&MS Sub Systems. In support of the illusion there are several major sub systems that make up the S-70A-9 FF&MS. These are described below.
The computer system. The computer system is the heart of any flight simulator and is often referred to as the "Host". It receives and controls inputs from other computer sub-systems and, as can be seen in Figure 1, distributes these outputs via microprocessors to the various control systems - the motion jacks, speakers etc. A simulator can operate to some extent with any one of its other systems off-line, but if the computer fails it will be unusable.
|
| |
 |
| |
The Host computer also contains the mathematical models of the aircraft and environments being simulated and constantly calculates a wide variety of equations that determine what the aircraft should be doing under the prevailing conditions.
There is a finite amount of time allowed for the conduct of these calculations and for their results to be seen by the aircrew. This time or delay is an acceptable by-product of the system and is known as Transfer or Transport Delay (TD). Basically, the smaller, or faster, the TD, the more realistic the device as it is closer to emulating real world reactions and resultants. Interestingly, slow TDs have also been associated with simulator sickness, a condition which needs to be avoided at all costs. The time allowed varies dependant upon the level of fidelity of the device. The FF&MS is a Level 5 flight simulator, therefore the most stringent limitations are imposed, so the calculations must be complete and the results manifest in, at the most, 100ms.
Like all computer systems there are many factors that go into determining what the system’s final TD will be. Processor speeds, amount of memory and bus speeds are just a few obvious examples of what may impact the TD. One that may not be as obvious, but which had significant bearing on the originally designed TD of the FF&MS is the configuration and performance of the visual system and its Iteration Rate.
Iteration Rate (IR), or the speed at which the computer conducts the same calculation or activity, is a vital component of TD. For example, in the Visual System this means that only a certain number of objects, or polygons, could be drawn in each sweep of the screen by the projector. (This is also known as the screen redraw rate) The slower the IR, the more surfaces can be drawn per projector sweep, therefore the more realistic the image. But the slower the TD the less realistic the "feel" of the simulator and the harder it is to conduct the required flight sequences.
To cater for the unique size and configuration of the FF&MS Visual System the original design TD figures had to be slowed down to the upper allowable limits in order to improve the quality of the visual scene to an acceptable level. This was achieved such that no deterioration in the feel of the device was apparent
The visual system. As was mentioned earlier, the configuration of the visual system design of the FF&MS is quite unique. The design is relatively conventional and consists of seven digital channels providing day/night/NVG colour images of realistic Australian and regional terrains, environments and equipment. Five of these channels provide signals to the five digital projectors mounted above the flight deck. These projectors force the image through a Back Projection Screen (BSP) which refracts the image onto a series of optically perfect glass mirrors that form a continuos Field of View display. In turn, the image is then reflected towards the aircrew, converging at the design eye-points.
The configuration of a continuous field of view visual system is shown in Figures 2 and 3. This configuration is similar to that of the S-70A-9 FF&MS. |
| |
 |
 |
| |
Embedded within this design but not shown, are another two separate digital visual channels which feed signals to either the Chin Window displays for day/night viewing, or to a dedicated simulated NVG system for aided operations. It is this design, along with the size of the mirror system that makes the FF&MS’s system unique.
Traditional visual systems have been hampered by the lack of natural optical flow cuing at the aircrew’s horizontal and vertical periphery In all cases these have been structural limitations. This restriction has been significant, as appropriate, natural cuing is a vital factor in enabling aircrew to determine speed, rates of closure, height etc. and as such are intrinsic to all flight operations.
To some extent the advent of continuous field of view devices have gone a long way to overcoming these deficiencies. Until recently, however, these systems have only been able to provide, at most, 1800 of viewable image, and typically closer to 1400 in the horizontal and 400 in the vertical. Many attempts to overcome this limitation have been trialed in the past. Most interestingly one configuration has images that are projected onto the inside of a large sphere that is mounted around the cockpit structure. Although this configuration provided the natural cuing mentioned above, image quality is dependant upon a large increase in the number of projectors, hence requiring a larger overall structure, and the image cannot be simultaneously collimated as required for the dual pilot environment of S-70A-9 operations.
The FF&MS overcomes this visual range deficiency by providing 2200 of horizontal image, equal to the natural viewed area of the human eye, and 600 in the vertical. This is the largest flight simulation mirror system currently in use in the world and the result is that natural optical flow cuing is obtained. The cues are immediately obvious the first time that you "fly" the FF&MS and the effect greatly enhances the overall illusion.
The obvious limitation of this configuration is the vertical size of the image. Helicopter operations typically use everything from the floor adjacent to the chin windows to the roof structure behind the overhead transparencies.The lack of overhead windows in the s-70A-9 FF&MS also creates some concern as they are used, generally, at AOB greater than 300. Although it is not impossible to incorporate a solution for this deficiency within the configuration of this visual system, what possible solutions that are available are prohibitively expensive and would not greatly enhance the overall training environment provided by the FF&MS. Hence, this area of the simulator is used for enhancing the ambient light within the cockpit.
To aid in overcoming this physical limitation, the FF&MS visual design incorporates two flat panel, projected digital images for the Chin Window displays that are separate from the viewing area of the main image. These are uncollimated displays that are mounted within the bounds of the chin window transparencies such that the cockpit structure naturally breaks the images from that which is projected onto the main mirror system. They provide similar image brightness and resolution as that of the main image from digital projectors that are mounted alongside the cockpit doors and provide a realistic albeit limited solution to the lack of vertical field of view inherent within the design.
Chin window displays are a very recent innovation within the helicopter flight simulation community. Recent advances in technology have enabled the use of this visual enhancement, but there are still many limitations, coupled mainly to the performance of the class of projectors currently used and the physical limitations of where they have to be mounted. The S-70A-9 FF&MS is one of the first flight simulators in the world to incorporate this advance.
For NVG operations the FF&MS is fitted with an NVG system that simulates the natural environment via two sets of NVG goggles that emulates recently operationally fielded equipment. This is new technology that is not without its limitations, but it offers the first real attempt within the industry to provide a solution to a relatively risky training environment. |
| |
| |
| |
|
Part Two: Enhancing the Illusion |
|
|
|
|
| |
| |
| |
| |
|
 |
 |
The U.S. Military saw the need for the "pilot maker" but lacked the money to buy them. In February 1934, the U.S. Army Air Corps was ordered to fly the airmail across the United States. Army pilots lacked experience in flying at night or in bad weather and five pilots were killed in the first few days of the mail flights. In an effort to overcome the problem the Army arranged for Link to visit the Newark Airport in New Jersey to demonstrate his trainer. On the day of the demonstration the weather turned stormy, but Link was able to fly in safely and convinced the Army that instrument flight was practical and could be taught in his trainer. In the photo above (courtesy of the USAF Museum of Flight) a basic Link trainer can be seen mounted on the bellows mechanism. To the right, connected by cable, is the "crab" output devic. This is a three wheeled electric motor driven mechanism, which moves across a map and traces the aircraft's track. Inputs to the pilot were from the operator via the console on the table. |
|
 |
|
Link's trainer was used for instrument flight training for almost all US and Allied pilots during World War II. Upon entering the war, the armed forces contracted for Link's entire production. The company built 6,271 trainers for the Army and 1,045 for the Navy. In 1945, an AT-6 Texan training airplane cost $10 per hour to operate while a Link Trainer cost 40 cents an hour. More important, no pilot was ever injured in the "crash" of a Link Trainer - although folk-legend has it that one trainee pilot had become so absorbed in his sequence, that when told over the "radio" that his fuel had ran out, he broke his ankle in his haste to escape before the Link hit the ground. |
|
|
| |
 |
| |
|
| |
 |
| |
| |
| |
| |
| |
|
