Personal tools

Chapter 8: Expert Essays

From X-Plane Wiki

Jump to: navigation, search


Tuning the Handling in X-Plane

If X-Plane is set up and flying, but aircraft seem to be too sensitive in pitch or pull to one side, the simulator may need to be tuned.

Before performing the following, make sure the joystick and/or other control devices are set up and calibrated. See Chapter 3 for instructions on this.

To easily see whether the controls are properly calibrated, go to the Settings menu and click Data Input & Output. There, select the rightmost checkbox next to 'joystick ail/elv/rud, as seen in the following image.


This will cause X-Plane to display on the flight screen the aileron, elevator, and rudder inputs from the flight controls (such as a joystick, rudders, yoke, etc.). With this done, close the Data Input & Output screen. The inputs for the various stick deflections should now be visible the top left corner of the screen, as seen in the image below.


With properly configured controls, when the stick/yoke/pedals are centered, the aileron, elevator, and rudder joystick inputs all read around 0.0. When the controls are pushed full left and forward, they should read around -1.0. When the controls are pulled full aft and right, they should read around 1.0. If these are the results obtained, then the joystick is calibrated. If not, the joystick is not calibrated—no wonder the plane is not flying correctly! See Chapter 3 for information on calibration.

If the controls are indeed properly calibrated as per the above test, but the plane still is not flying correctly, it’s time to look at the first level of control response tuning. Go to the Settings menu and click Joystick & Equipment. In that window, click the Center tab. Move the stick or yoke around. Doing so should move little rectangles around in a box on the lower half of the screen, and when the stick is centered, the rectangles should (ideally) go to zero size. Since no hardware is perfect, though, simply center the controls and hit the button labeled CENTER THE YOKE AND PEDALS AND HIT THIS BUTTON, as highlighted in the following screenshot.


This will tell X-Plane that the hardware is indeed centered. When using PFC hardware, there will be little buttons across the bottom of the window that will set the center position of each axis.

With that done, close the Joystick & Equipment window and move the flight controls to the centered position. Check to see if the data output (which should still be on the screen from the pre-test in the above paragraphs) is around 0.000 when the controls are centered. If it is, then the hardware works fine and the center point was set successfully. If the data output does not read near 0, the hardware is either of poor quality (or failing) or the center point was not properly set.

With the center point set correctly, try flying the plane once again. If it still does not handle correctly, read on to tune the next level of control response.

Open the Joystick & Equipment window and select the Center tab once again. Look at the three sliders labeled control-response (one labeled pitch, one roll, and one yaw) at the top left of the screen, as seen in the following screenshot.


If these three sliders are fully left, then the control response is linear; that is, a 50% stick deflection in the hardware will give 50% control deflection in the aircraft. Likewise, 100% stick deflection in the hardware will give 100% control deflection in the aircraft.

If the problem being experienced is that the plane feels too responsive in the simulator, try dragging the sliders all the way to the right. This will give a non-linear response. Set this way, 0% hardware deflection will still give 0% control deflection in the sim, and 100% hardware deflection will still give 100% control deflection. The difference lies in between—50% stick deflection in the hardware might only give 15% control deflection in the sim. In other words, while the hard-over roll rate in the sim will remain unchanged no matter what these sliders are set to, fine control will be increased for smaller, partial deflections, since the flight controls will move less for a small-to-moderate stick deflection in the hardware joystick or yoke. This will give a nice fine-pitch control and slow, detailed roll control.

If, after changing the control response, the aircraft still does not fly as it should, read on.

The next level of control tuning is stability augmentation. If the plane still feels squirrelly or overly sensitive, go back to the Center tab of the Joystick & Equipment window and try dragging the three sliders on the right side of the screen (labeled stability-augmentation—shown in the following image) all the way to the right.


This will cause X-Plane to automatically counteract any stick input to some degree, resisting rapid or large deflections in pitch, heading, and roll. Basically, it is like always having an autopilot on that smooths things out. This is obviously very fake, but in the absence of a perfect flight control system and g-load and peripheral-vision feedback, this can help smooth out the airplane’s flight characteristics. Try flying with those sliders at various places, bearing in mind that full left should be most realistic (with no artificial stability added).

If, after doing all of the above, the aircraft still does not fly as it should, nothing more can be done within the simulator. It is now time to tweak the airplane. In the real world, if a plane is pulling to one side or the other, a pilot will bend the little trim tab on the aileron one way or another. This bending of the aileron trim tab counteracts any imperfections in the shape of the airplane or dynamics of the propwash or mass distribution inside the plane. The same thing can be done in X-Plane—one can bend a trim tab a bit one way or the other to make the plane fly true.

To do this, first exit X-Plane and open Plane-Maker (found in the X-Plane 9 installation folder). Go to the File menu and open the plane that is pulling left or right, as shown in the following screenshot.


Go to the Standard menu and click Control Geometry, shown in the following image.


In this window, click the Trim & Speed tab, as shown in the following screenshot.


Look at the far right-hand column of controls in the top half of the screen, labeled trim tab adjust, highlighted below.


This is a measure of how much the trim tabs are bent on each axis. The top control is the elevator, the middle is the aileron, and the bottom is the rudder (as per on the far left side of the screen). A value of 0.000 means that the trim tab is not bent at all. A value of 1.000 means the tab is bent so far that the control is fully deflected by the trim tab—this is way too far. Try bending the trim tab just a little bit—maybe set the value at 0.05 or at most 0.10. This would correspond to being enough force to deflect the controls 5% or 10% due to the trim tab. A positive value corresponds to bending the trim tab up or right, depending on whether it is pitch, yaw, or roll. Thus, if the plane needs to roll right a bit more (or needs to stop rolling left), then enter a positive number for the aileron control. The same goes for the rudder if the plane needs to pull right a bit more, or for the elevator if the plane needs to pull up a bit more. Tweak the trim tabs as needed, save the plane, and exit Plane-Maker. Then, open up X-Plane and try flying the plane again. The plane should noticeably pull one way or another based on how the trim tabs were bent. The trim tab controls may need to be tuned again to get the plane to fly as straight as is desired.

Factors Affecting X-Plane's Performance

X-Plane users tend to notice either that the simulator runs extremely fast, giving them 100 frames per second (fps), or that it is dismally slow, topping out at 20 fps. At identical rendering settings, this is due almost entirely to the hardware in the computer.

Some people today have computers with a 500 MHz Pentium III processor, 128 MB of RAM, and 8 MB of VRAM (perhaps an 8 year old system), while others have quad-core 3000 MHz processors, 4096 MB of RAM, and 512 MB of VRAM (perhaps only one year old). There is more than a 6x difference in speed between those two setups, since the RAM speed, bus speed, video card speed, and many other things also influence the computer’s performance.

Many people do not understand what determines a computer’s performance. The three biggest factors are the amount of RAM in the system, the speed of the CPU, and the speed of the graphics card. A fourth factor, which determines a system’s ability to display high quality video textures, is the amount of RAM on the video card (called VRAM).

Coming up short in any of the above categories will create a “bottleneck” in system performance, limiting the ability of the rest of the components.

For instance, using higher quality textures than can be stored in VRAM will slow X-Plane significantly (see Texture Resolution in Chapter 3 for more information on VRAM and textures), regardless of any other factors.

Conversely, even if the system’s video card has 2 GB (that is, 2048 MB) of VRAM and X-Plane is running at a low screen and texture resolution (eliminating any RAM problems), if the computer’s CPU or video card are too slow, then X-Plane’s performance will be poor.

For information on optimizing X-Plane’s frame rate, see Setting Up X-Plane for Best Performance.

Tuning the Autopilot

Occasionally, the autopilot in X-Plane might act up. It may sort of wander down the localizer or wander around in pitch when it should be holding altitude. It might wander around in heading, or perhaps flicker its wings madly left and right as it tries too hard to hold a heading. Whatever the problem, the autopilot constants can be adjusted by the user in order to make the plane hold its desired path more tightly.

These autopilot constants can be adjusted in Plane-Maker by doing the following:

1. Open Plane-Maker by opening the X-Plane 9 folder and double clicking on Plane-Maker.exe.
2. Load the airplane that needs adjusting by clicking the File menu and selecting Open, as shown below.
3. Find the .acf file just as you would when selecting a craft within the X-Plane sim (see Opening an Aircraft in Chapter 4).
4. Go to the Expert menu and click on Artificial Stability, as below.
5. Go to the Autopilot tab and check the use custom autopilot constants box, as seen in the following image.
6. A number of controls will appear that specify the autopilot constants for your airplane.

Let’s examine what each of these autopilot constant controls do. First, let's talk about correcting heading.

Tuning Autopilot Roll

Roll Error for Full Aileron

This control is found in the middle box of the Artificial Stability window, in the right column, highlighted in the following image.


When flying a real plane, a pilot decides on a roll angle to make a turn. He or she then decides to deflect the ailerons a certain amount to achieve the desired bank angle.

Imagine that you want 45 degrees of bank, and the plane is currently at 0 degrees of bank. You wouldn’t apply just a touch of aileron to get there, but rather a strong dose of it. After all, you are a whole 45 degrees away from the desired roll angle. Conversely, imagine you are at 29 degrees of roll, and you want 30 degrees of roll. You only need one more degree of roll, so you wouldn’t put in full aileron to get to 30 degrees—that would overshoot it for sure. Instead, you would look at the controls and notice that you are only a little off of from the desired bank angle, so the plane would need only a little bit of aileron.

Now, how many degrees off of a desired bank angle would a pilot have to be to put in full aileron? One degree? Ten? One hundred? The roll error for full aileron control specifies to the autopilot how many degrees off the aircraft must be from the desired roll angle before it puts in full aileron. If this is set to a very small number, the autopilot will put in full aileron for even the tiniest of roll errors—not good! This will cause the plane to over-control and flutter madly left and right like an over-caffeinated pilot! On the other hand, if this control is set to a very large number, like 100 degrees, then the autopilot will hardly put in any aileron input at all. In that case, the plane will always wander off course a bit, because it will never move quickly enough to get back on course.

Now, a smart pilot might say, "I would never input full aileron, ever." Fair enough. But realize that the autopilot will be limited to about 50% travel or so, and it will automatically back off of the controls as the airplane speeds up, just as a good pilot would. Thus, what this control really determines is how aggressively the ailerons are applied. If the plane tends to steer too unaggressively to the command bars, a smaller number is probably required here. This will tell the autopilot to require a smaller deflection to really crank in the ailerons. Conversely, if the plane flutters left and right like a plastic bag in a 50-knot wind, then the autopilot needs to be told not crank in so much aileron. To do that, enter a larger number here, so that the autopilot waits for a larger error to develop before responding with so much force.

A good starting point for this control is 30 degrees. This means that if the roll angle is off by 10 degrees, the plane will apply one-third aileron to correct when at low speed—not a bad idea.

Roll Prediction

This control is found in the middle box of the Artificial Stability screen, at the top of the left column, highlighted in the following image.


When a pilot flies, he or she tends to look into the future to decide when to add to or back off from the flight controls. This is simple anticipation. The roll prediction control tells the autopilot how far into the future it should look. If the plane tends to wander slowly left and right, always behind its mark, or it overshoots and then wanders slowly off in the wrong direction like a tired drunk driver, then it clearly is not anticipating enough. In that case, an increase is required in the roll prediction to make the autopilot anticipate more. If, however, the airplane starts flopping back and forth hysterically every frame, the autopilot is clearly anticipating too much; a smaller roll prediction is needed.

One second is a good starting point for this control—after all, when real pilots fly, it’s a good bet that they enter controls based on where the plane will be in one second, rather than where it is at the moment.

Roll Tune Time

This control is found in the middle box of the Artificial Stability window, in the middle of the left-hand column, seen in the following screenshot.


In the real plane, a pilot will trim out any loads with trim if it is available. The roll tune time determines how long the autopilot takes to run the trim. A real pilot probably takes more than just a few seconds to do this. However, if the autopilot waits too long to trim out the loads, it may be slow and late in getting to the correct angle.

A good starting point for this control is 5 seconds.

Localizer CDI Gain

This control is found in the middle box of the Artificial Stability window, at the bottom of the right-hand column, highlighted in the following image.


If a pilot is one degree off the localizer when flying an ILS, he or she needs to decide how many degrees of heading correction are called for to correct that. If s/he corrects only one degree, the craft will be flying right towards the airport, never intercepting the localizer until it gets to the transmitter on the ground. Usually, if a pilot sees a one degree error in the localizer (one dot on the CDI), s/he would enter about 10 degrees of heading correction, thus forcing the plane to nail that HSI now. The localizer CDI gain control sets the number of degrees of heading change that the autopilot will pull for each degree of error on the localizer (which is the same as saying for each dot of CDI deflection).

A good starting point for this control is 10 degrees.

Localizer CDI Prediction

This control is found in the middle box of the Artificial Stability window, at the bottom of the left column, as seen in the screenshot below.


A good pilot does not fly an ILS based on where the CDI is at the moment. A pilot that flies like that wanders around in S-turns all the way down the localizer! A good pilot flies the plane based on where the HSI CDI will be in the near future. The localizer CDI prediction control tells the autopilot how far into the future it should be looking when following the CDI. It should be looking at least a few seconds into the future. The higher this number is, the more the autopilot will anticipate. If the plane is wandering back and forth slowly across the localizer, always S-turning, it probably needs a bigger number here. More anticipation will prevent those endless S-turns. However, if too big a number is entered here, then the plane might never join the localizer. This is caused by the autopilot anticipating so far ahead that it turns away from the localizer as soon as the needle comes alive, shying away to avoid an over-shoot. Obviously, that is too much anticipation!

A reasonable number for this control is between 2 and 4 seconds.

Summary of Roll Settings

In summary, enter the number of degrees of bank error that should give a very strong aileron response in the roll error for full aileron control. Enter the number of seconds the system should anticipate in the roll prediction control, the number of seconds required to trim out the load in the roll tune time, the number of degrees of heading change per degree localizer error in the localizer CDI gain, and the number of seconds of anticipation in HSI CDI deflection in the localizer CDI prediction.

To tune these controls, it is a good idea to first forget about the ILS and simply try to get the aircraft to perfectly hold a heading. Tweak the roll error for full aileron to give as strong a response as desired, and tweak the roll prediction to give the desired anticipation. Fly the plane around in heading mode, snapping the heading bug left and right and tweaking those constants until the plane follows the heading bug perfectly. Then, after the heading mode is perfect, adjust the localizer values while flying ILSs to tune the localizer. If the plane flies S-turns across the localizer, the localizer prediction needs to be greater. If the autopilot never even latches on to the localizer, continually turning away from it, then the localizer prediction needs to be decreased—the craft is clearly over-anticipating.

Next we will discuss correcting pitch; the discussion will be almost exactly the same as roll, really.

Tuning Autopilot Pitch

Pitch Error for Full Elevator

This control is found in the bottom box of the Artificial Stability window, at the top of the right column, as seen in the following screenshot.


This control is to pitch as the roll error for full aileron (from above) control is to roll. It determines how much error between desired and actual pitch is required for full elevator deflection. Remember that the autopilot will automatically reduce the control deflections as the plane speeds up, limiting to maybe 50% control deflection, so it isn’t necessary to worry about the system really going to full deflection.

To configure this control, forget about the ILS for a minute and just fly vertical speed or pitch sync mode. If the plane is sloppy about getting the nose up to track a new vertical speed and just takes too long to get there, then a smaller pitch error for full elevator value is needed. This will cause the plane to be more aggressive with the elevator. Of course, if the plane starts flapping about madly, a larger value is needed, telling the plane to stop deflecting the elevator so much unless it has a larger error between the actual and desired pitch. Put the autopilot in pitch sync mode, then hold the CWS button down and quickly pitch the nose, letting go of the CWS button. If the autopilot is slow and sloppy in holding that new pitch, then a smaller number needs to be entered here to make the thing more aggressive.

Pitch Prediction

This control is found in the bottom box of the Artificial Stability window, at the top of the left column, highlighted in the following screenshot.


A good pilot will input flight controls by predicting where the plane will soon be. The pitch prediction control determines how far into the future the autopilot will look. If the plane is always wandering up and down when trying to hold a given vertical speed, always a few steps behind where it needs to be, then more anticipation is clearly called for—the pitch prediction control needs to be set to a larger number. Conversely, if the plane is always afraid to get where it needs to be, resisting motion towards the desired pitch, then it is probably anticipating too much, and a smaller number is called for. Once again, these numbers need to be tuned in pitch and roll modes, or maybe heading and vertical speed modes, to get them set perfectly, with nice, snappy, precise autopilot response, before the autopilot is tested on an ILS.

A good starting point for this control is one second.

Pitch Tune Time

This control is found in the bottom box of the Artificial Stability window, in the left-hand column, as seen in the following image.


It sets the time require to trim, similar to the 'roll tune time control described above. If this is set to too small a number, the plane will constantly be wandering up and down as it plays with the trim, as it will always be too quick to modify the trim. A real pilot would wait until s/he is sure that the trim needs modifying.

This control should probably be set between 5 and 10 seconds.

Glideslope CDI Gain

This control is found in the bottom box of the Artificial Stability window, at the bottom of the right column, as seen in the following screenshot.


It tells the autopilot how much it should change the pitch for each degree of glideslope error. For example, if it is set to 5 degrees (a reasonable value), the autopilot will pitch up 5 degrees for each degree it is below the glideslope. The greater the number entered here, the more the command bars will move to meet the glideslope.

Glideslope CDI Prediction

This control is found in the bottom box of the Artificial Stability window, in the left-hand column, highlighted in the following image.


A good pilot will anticipate where the glideslope will be in the near future as he or she controls the pitch. If the pitch is not anticipated enough, the aircraft will be correcting up and down all the way down the glideslope. If the pitch is anticipated too much, the craft will never get to the glideslope, as it will always be shying away from it as soon as the needle starts to close in.

A good starting point for this control is 8 seconds.

Pitch Degrees per Knot

This control is found in the bottom box of the Artificial Stability window, at the bottom of the left column, seen in the following image.


It determines how many degrees the autopilot will pitch the craft up or down in order to correct for a one-knot difference between the actual speed and the one set in flight level change mode. A good starting point is 0.2 degrees.

Summary of Both Pitch and Roll Controls

To summarize, remember that there are two things happening with these controls: the amount the autopilot moves the command bars, and the amount it moves the controls to capture those command bars (see the table below). Therefore, if the command bars are not behaving as they should, one of the command bar variables needs to be set. On the other hand, if the command bars are fine, but the airplane isn’t tracking those bars correctly, one of the flight control tracking variables needs to be set so that the autopilot will “grab” the bars.

Remember there are two steps to tuning these autopilot controls:

1. Decide how to move the bars (CDI gain and CDI prediction), then
2. Decide how to move the controls (pitch and roll error, pitch and roll prediction).

Finally, remember that there is one number that controls how hard we try to get to our target (CDI gain, roll and pitch error—think of this as a “spring constant”) and one number that controls our anticipation (CDI prediction, roll and pitch prediction—think of this as a damping constant).

Amount to Move the Command Bars on the ILS Amount to Move the Controls to Track the Bars
localizer CDI gain roll error for full aileron
glideslope CDI gain pitch error for full aileron
Amount to Anticipate the Command Bars on the ILS Amount to Anticipate the Attitude to Track the Bars
localizer CDI prediction roll prediction
glideslop CDI prediction pitch prediction
Time to Trim the Forces
pitch tune time
roll tune time

Setting Autopilot Constants Quickly

Now that we’ve discussed what each control does, let's look at how to set these things up quickly. First, launch X-Plane and open the aircraft that needs modifying. Go to the Special menu and click the Set Autopilot Constants menu item. A window (shown in the image below) will appear that looks identical to the Artificial Stability window discussed above.


The settings here can be changed while flying in order to determine what the autopilot constants need to be for the plane. Be aware, though, that these settings will be lost the second X-Plane is closed or another aircraft is opened. This in-simulation version of the Artificial Stability window from Plane-Maker is for experimentation only. Once desired settings have been determined, be sure to write them down on a piece of paper or enter them into a text document so that they can later be entered into Plane-Maker, where the settings can actually be saved.

One final note: Some users try to configure a really aggressive autopilot system that has huge anticipation, huge gains, and tiny maximum pitch and roll errors for full deflections. That would be a very strong, very aggressive autopilot that may seem to work perfectly. A problem arises, though, as soon as the craft is flown at a low frame rate. When this happens, the plane will start shaking violently on autopilot because that autopilot is not being run fast enough to see the very rapid results of its overly strong inputs. If such settings are to be used, be sure to configure the scenery or weather so that they will really slow the simulator down (ideally to its minimum frame rate) while tuning the autopilot. Only then can it be certain that the constants entered that will always work, because a higher frame rate will never hurt. The easiest way to slow the sim down for this is to set three broken layers of clouds and plenty of buildings to be rendered (see Chapter 3 for more on setting rendering options).

Designing an Artificial Stability System

For users creating a VTOL (a vertical take-off and landing aircraft) or a fighter, it may be necessary to design an artificial stability system in order to make the craft feel stable, even though, in reality, it isn’t. This is especially common in fighter jets and helicopters—fighters are most maneuverable if unstable, and helicopters simply have nothing to naturally make them stable. Control systems are designed to make these craft seem stable. These typically work by adding some input in addition to the joystick/yoke input in order to make the craft do what the pilot wants.

A common example of this kind of stability system in the civilian world is the yaw damper. A pilot’s feet still move the rudders, but the yaw damper system adds some additional rudder deflection for the pilot to damp out the rotation rates of the plane. The amount of rudder deflection added depends on what the control system engineer decides is necessary—in the case of designing custom aircraft for X-Plane, that “engineer” is the user.

To create a system to add stability in Plane-Maker, first load the aircraft to be modified. Open the Expert menu and select the Artificial Stability menu option (as shown in the image below).


In the window that opens, select the Art Stab tab.


Designing a Yaw Damper

By way of example, consider a yaw damper again. Its purpose is to add some rudder deflection to whatever the pilot inputs with his or her feet, stopping the aircraft’s rotation. This is seen in high-end Mooneys and most jets. The yaw damper’s designer must consider how much rudder is desired to stop the rotation—full rudder? Half? Perhaps just 1/10 of the max rudder deflection is needed. Obviously, if the plane is only wagging its tail a little bit, only a little rudder is needed to stop it. However, if the plane swings around quickly, then the damper system needs to put in a lot of rudder to stop the rotation quickly.

To decide how much extra rudder input is necessary, the designer first needs to know how much “wag” the system needs to compensate for. In X-Plane, designers enter a fraction of the rudder input per degree per second of rotation rate.

For instance, imagine the plane’s tail is swinging (from turbulence, varying crosswind, the pilot stepping on the rudder, etc.) at 90 degrees per second—that is, the tail moves in one second from being straight in line to pointing full left or full right. In a real airplane, 90 degrees per second of tail-wagging will feel like a lot. Kicking the rudders a bit in a Cessna 172, for example, will shake its tail at about 35 degrees per second. So, let’s imagine that 90 degrees per second is such a high rotation rate that the control system needs to put in full rudder to oppose it. That means that if the plane is rotating at 90 degrees per second, the yaw damper will put in full rudder to oppose that motion, and at 45 degrees per second it will put in half rudder to oppose that motion. At a measly 9 degrees per second, the yaw damper will put in only 1/10 rudder to oppose that motion. At the 35 degree per second tail-wag of a Cessna 172, the control system would put in as much as about 35% rudder deflection to stabilize the plane’s yaw motion. This does not sound like an unreasonable constant.

To enter those settings in X-Plane, once again open Plane-Maker’s Artificial Stability window from the Expert menu. Select the Art Stab (that is, artificial stability) tab. The above example deals first with the heading: target deg sideslip control, found in the second column from the left in the bottom box of this window, highlighted in the following image. A value of 0 would be entered here, meaning the plane always tries to stabilize at 0 sideslip.


For the fraction deflection per degree difference control immediately to the right of that, simply enter 0, meaning the system is not trying to achieve a desired sideslip, only to damp out the tail wagging by opposing rotation.

For the fraction deflection per degree per second control to the right of that (highlighted in the following image), enter a value 0.0111.


This number comes from dividing 1.000 (that is, full rudder) by 90 (the rotation rate in degrees per second that full rudder should be applied at). 1 / 90 = 0.01¯ , which is rounded to 0.0111 in Plane-Maker. Put another way, that equation is 1 / rate for max yaw. A value of 0.01111 is pretty reasonable. Try entering this for the 172, saving the plane in Plane-Maker, and loading it again in X-Plane. Pop the rudders left and right and notice how the plane damps out faster, as would a real plane if such a yaw damper were installed.

Now, if even more stabilization is needed, try entering 0.1 in the fraction deflection per degree per second control. This means that if the plane is rotating at 10 degrees per second, the rudder will deflect fully to oppose it. (10 degrees per second times 0.1 control per degree per second = 1.00, or full deflection.) A rotation rate of 10 degrees per second means that it will take 9 seconds for the plane’s tail to move 90 degrees—a very slow rotation rate. With a constant of 0.1, even this rate will be opposed by full rudder. Yikes! If such a plane were taken into turbulence, the air would certainly be kicking the plane around at over 10 degrees per second, so the craft would give full rudder deflection first one way, then the other. The plane would over react to every angular rotation induced by the turbulence by kicking the rudder to full in order to oppose that rotation.

Obviously, this constant of 0.1 is high. Customer support had a call, though, with someone who had entered a constant of 3.0—thirty times higher than the hypothetical case above. This means that for a rotation rate of 1/3 degree per second (at which rate it would take a whole four and a half minutes to move thru 90 degrees of heading), the system would put in full opposing rudder. Even the tiniest hint of rotation in a given direction would make the rudder slam hard over to counter it. Needless to say, any time this plane met even a touch of wind, the rudder would slam from one stop to the other in a wildly exaggerated effort to counter the turbulence. If one must kill a fly buzzing around in a china shop, don't do it with a sledgehammer—the results won’t be pretty. This particular plane handled alright if there was no turbulence—since nothing was rotating the plane, the flight controls didn’t have to move to oppose that rotation. As soon as the slightest imperfection came along to move the plane, though, such as turbulence, movement from the pilot hitting a flight control, a bird-strike, an engine-failure, a bumpy landing, or flying into changing winds, the controls went crazy.

One thing that can only be learned by actually getting a pilot’s license and getting up in the sky is that it is a very imperfect world up there. The plane is constantly barraged by all manner of imperfections, perturbations, and external winds and forces, and, much like with a boat, these imperfections must be anticipated in the design.

Stabilizing Pitch

Now that we’ve discussed heading stability, let’s move on to pitch. If a plane is not very stable in pitch, users may want to “lock it down” a bit. Plane-Maker’s Artificial Stability window is used as above if users want to avoid artificial stability controls found in X-Plane (these are applied to all aircraft in the sim and thus aren’t applicable any particular craft—see Chapter 3), but rather want to design their own to mimic one that might be installed in a real plane.

For our example airplane, we will again open Plane-Maker, click the Expert menu, and open the Artificial Stability Window. Once more, select the Art Stab tab. This time, enter maybe 20 degrees in the pitch: target deg angle of attack (highlighted in the following image)—this should be enough to stall the plane.


Enter 0.1 for the fraction deflection per degree difference (highlighted in the image following), so that if the angle of attack is 10 degrees off, the plane applies full elevator to capture the desired angle of attack.


Finally, enter 0.05 for the fraction deflection per degree per second (highlighted in the following image) so that if the nose is coming up at a rate of 20 degrees per second, the system will apply full elevator to stop it.


These are all pretty aggressive constants (meaning a lot of elevator is brought in to counteract a small amount of motion), but they aren’t extreme.

There are two reasons for using these aggressive constants. First, the plane needs to have lower rates in pitch than in yaw. This is because if the plane is moved left and right a bit, not that much will change in terms of flight control—the vertical stabilizer, which is being broadcast to the air, is small. But, if the plane is tilted up or down a bit, then the entirety of both the wing and horizontal stabilizer is exposed to the air. The effect will be much greater than in yaw, where only the vertical stabilizer is offset, simply because the wing is so much bigger. A plane sees a much greater effect for each degree of change in the angle of attack than in sideslip, so it needs lower rates of pitch than yaw to keep within comfortable (safe) G-loads. For this reason, we enter higher constants in pitch than in yaw to really work hard to counter those pitch rates.

The second reason for entering higher constants in pitch than might seem advisable is that, quite simply, X-Plane cheats. The simulator will automatically reduce these settings as the plane speeds up, because it knows that at high speeds it is better to enter smaller control deflections to keep from breaking things! This means that the constants entered here are only fully applied near stall speeds where control authority is mushy. The artificial stability controls relax and phase out as the indicated airspeed (air pressure on the controls) builds up.

To see this scheme in practice, open up “Austin’s Personal VTOL” in Plane-Maker. Go to the File menu, click Open Aircraft, open the Austin's Designs folder, and select Austin's Personal VTOL. Now open the Artificial Stability screen again from the Expert menu. Notice that only low-speed constants (in the top box of the screen rather than the bottom, highlighted in the following image) are set here, designed to phase out rotation rates to make the craft easy to fly.


Look at the rotations that are targeted with full-scale stick deflections in hover—a max of 30 degrees pitch, 45 degrees roll, and 45 degrees per second rotation rate in yaw (now you know what the 0.02 and 0.01 do as well).

Next, open up X-Plane and load up this aircraft (noting that it starts off with its thrust vector at 90 degrees, straight up). Add power to rise up off the ground and work on hovering. Slide left and right, then fore and aft, up and down, all using small control deflections. Then, click on the little switch on the panel labeled ART STAB. This will turn the stability augmentation off in order to fly without it. Viva la difference! For an even more extreme case, try turning off the artificial stability in Austin’s Death Trap at 300 knots.

Setting Up Advanced Networks

Suppose a user wants to have two computers running X-Plane, one with the instrument panel on the pilot's side, and one with a panel on the copilot’s side. This is called having a master machine and a copilot's machine. Or, suppose a user wants two panels, and maybe a center radio panel as well. Maybe she or he wants an Instructor Operator's Station (called an IOS) to control weather, time of day, and aircraft failures. Or maybe s/he wants a separate computer for an out-the-window view (called an external visual), or to set up a multiplayer session to fly formation with friends.

Maybe the user wants all of the above, all at once!

What all of these setups have in common is the use of multiple computers. Each of these computers needs X-Plane installed (scenery and all). They also need to be on the same network, with IP addresses that are the same for the first three numbers (e.g., 10.2.2.*** or 192.168.1.***) and subnet masks of

Once each of the computers has an IP address that meets this requirement, they may be set up something like this, for example:

Setup of Six Networked Computers
IP Address Description Master machine—joysticks plugged in here Copilot's machine (.acf file with copilots instrument panel used here) IOS (instructor’s station for initiating failures, setting weather, moving the plane, etc...) External visual, left view External visual, center view External visual, right view

Of course, the LAN must be set up so that the computers can talk to each other, and the network must be ready for an X-Plane multi-computer setup.

With the network configured, open the Net Connections window from X-Plane’s Settings menu on each of the computers. Select the middle tab, Inet 2, as seen below.


This tab allows the user to configure a multi-computer X-Plane system. On each computer, simply check the box describing the job of each computer and enter the IP address of whatever other computers are called for by the text description. With that done, the sim should be ready to fly!

Let’s go through the specific example of setting up a copilot’s instrument panel using a second computer.

Setting Up a Copilot’s Station

First, we will need two computers, each running their own copy of X-Plane. These need to be joined together with either a single crossover Ethernet cable or a pair of Ethernet cables hooked to an Ethernet hub. The computers should form a simple LAN, configured as normal within the Mac OS or Windows, whatever the case may be.

The airplane file to be used will need two copies, both either created or modified using Plane-Maker. The first copy (for instance, named “Boeing 747.acf”) should have the pilot-side instrument panel. If the user is content with the default panel layout, any of the stock planes could be used.

With the first version of the plane ready (the pilot-side version), simply make a copy of the airplane file and add "_copilot" to the end of the name—for instance, if the file “Boeing 747.acf” was used for the pilot-side version, the copilot-side version would be named “Boeing 747_copilot.acf”. The copilots-side airplane should be saved in the same folder as the pilot-side plane.

Next, open the copilot-side copy in Plane-Maker and tweak the instrument panel as desired for the co-pilot’s side of the craft. Save it when finished and close Plane-Maker.

There should now be two copies of the same plane, each with its own instrument panel, with names in the format of "Insert plane name.acf" and "Insert plane name_copilot.acf". Both files should be in the same folder.

Simply copy that whole aircraft folder from one of the computers over to the other, putting the aircraft folder in the same directory on the second computer. For example, if, on the first computer, the folder was located in:

C:\Documents and Settings\Pilot\Desktop\X-Plane 9\Aircraft\Boeing 747\

then, on the second computer, it would be located in:

C:\Documents and Settings\Copilot\Desktop\X-Plane 9\Aircraft\Boeing 747\

With that done, open X-Plane on each computer, move the mouse to the top of the screen, click on the Settings menu, then select Net Connections, as shown in the following image.


On the pilot’s machine, go to the Inet 1 tab of the Net Connections window. Check the first check box labeled IP of extra visual/cockpit (this is master machine) and enter the IP address of the copilot's machine. For instance, in the image below, the copilot’s machine has an IP address of


On the copilot’s computer, go to the Inet 2 tab of the Net Connections window. Check the box labeled IP of master machine (this is extra cockpit) and enter the IP address of the first machine (the pilot’s, configured in the paragraph above). For instance, in the following image, the pilot’s machine has an IP address of


Now, on the lower left, click on the 'aircraft name reading suffix' and enter "_copilot". This means that no matter what plane is opened on the pilot’s machine, this computer will add "_copilot" to the name of the plane that it needs to open.

Now, on the pilot’s machine, open the "Insert plane name.acf" file. If everything is set up correctly, the pilot’s machine will send all the appropriate data to the copilot's machine (because the IP address of extra visual/cockpit box is checked), the copilot’s machine will get the message (because the IP address of master machine box is checked), and the copilot's machine will apply the name "_copilot" to the aircraft name (because of the name suffix that was entered), and it will open the copilot’s plane on the copilot’s machine.

Setting Up Multiple Monitors

A very commonly asked question deals with how to set up a multiple-monitor simulator. Often, this is in reference to using three monitors in particular. There are two ways to do this. The first is to use one computer with multiple monitors (either hooked directly to the video card or to a video splitter, like the Matrox TripleHead2Go), and the second is to use multiple computers with one monitor attached to each.

Obviously, with all other things being equal (e.g., hardware, rendering options), having one monitor per computer with multiple computers will give the highest frame rate, simply because there is more computing power behind each bit of display. However, using a powerful video card with a high fill rate, it should be possible to use one video card or computer to drive many monitors.

We will first examine the better of the two options (in terms of performance rather than cost efficiency)—using one computer per monitor, with the computers networked as specified in the preceding section, Setting Up Advanced Networks.

Multiple Computers, Multiple Monitors

Let’s assume we are to use four computers and four monitors: one cockpit and three external visuals (a common setup). Go to the Rendering Options screen on each of the three external-view computers in X-Plane. Enter a field of view of 45 degrees for each of them. Enter a lateral offset of -45 degrees for the left screen, 0 for the center screen, and 45 degrees for the offset, with 0 vertical offset on all screens. This will simply yield a 45+45+45 degree field of view. If this is drawn out on a piece of paper, it becomes apparent that the 45-degree offsets on the left and right screens will cause them to perfectly sync up with the center screen.

From there, the monitors need to physically be moved around the “cockpit” (that is, where a user will sit when flying the sim) in a semi-circle describing a 135-degree field of view. If this is not done, then the horizon will seem to not be straight as the craft pitches and rolls, caused by the “fisheye lens” effect. If a 135 degree field of view is described in a flat plane or in an arc of monitors that describe less than 135 degrees of arc, fisheye distortion will result, apparent as a horizon that seems to bend and distort between monitors.

In some cases it is not desirable for the monitors to wrap around the “cockpit,” but instead to simply be lined up beside one another in a flat plane (as when the monitors are stacked against a flat wall). In that case, an offset in degrees should not be used, but rather an offset ratio. In the case of using offset ratio, a ratio of 1.0 will cause the lateral offset for that copy of X-Plane (in linear distance) to be an amount equal to the distance between the user and the monitor. So, if the user is six feet from the monitor, and an offset ratio of 1.0 is used, then the center of that monitor should be 6 feet off to the right to line up.

Now, sometimes people sit on the ground and see the horizon does not line up, so they enter vertical offsets on some of the display machines only in order to get the horizons to line up. They quickly become confused when everything breaks down as they pitch and especially roll. Vertical offsets should not be set on some machines but not others. As soon as this is done, things start getting messed up. What often happens is that a user will fly with a cockpit in the center screen, where the center of the screen as far as scenery is concerned is probably about 3/4 of the way up the monitor (in order to leave room for the instruments), while using external visuals on the lateral displays, whose screen centers as far as scenery is concerned is right in the center of the monitor. In that case, the viewpoint center needs to be set in Plane-Maker for whatever airplane is being flown. This should be set to the center of the monitor—384 pixels as of this writing, or halfway up the 768 pixel height.

One Computer, Multiple Monitors

If the cost of a multi-computer setup is prohibitive, a single computer can be used to drive multiple monitors. Since the virtual demise of the Matrox Parhelia video card, a video splitter like the Matrox TripleHead2Go is most often used. These video splitters trick the operating system (Windows or Mac OS) into seeing the three monitors together as a single super-wide display. To configure this in X-Plane, simply specify the “single” display’s resolution in the Rendering Options screen, being sure to also set a wide field of view. The video splitter will distribute X-Plane’s output automatically across all three monitors. Of course, to do this with a decent frame rate, the computer will need a very powerful video card.

There are plenty of other ways to have multiple monitors on both Mac and Windows computers, but the rule of using them in X-Plane is simple: If the monitors appear to be one big desktop in the operating system, then they can form one big window in X-Plane (using the wide resolution and wide field of view that results). Just set the pixel resolution and field of view in the Rendering Options screen in X-Plane to match whatever monitor real-estate is available. This will allow wrap-around visuals from one computer.

Flying Helicopters

The following is a description of how helicopters are flown in the real world, along with the application of this in X-Plane.

All manner of different helicopter layouts can be found in reality, but we will discuss the standard configuration here—a single overhead rotor with a tail rotor in the back. Here's how this works: First, the main rotor provides the force needed to lift the craft by continually maintaining the same rotor RPM for the entire flight. The amount of lift generated by the main rotor is only varied by adjusting the blade pitch of the main rotor blades.

So, imagine the one-and-only operational RPM of a helicopter is 400 RPM. When the craft is sitting on the ground, the rotor is turning 400 RPM, and the pitch of the rotor’s blades is about zero. This means that the rotor is giving about zero lift! Because the blades have zero pitch, they have very little drag, so it is very easy to move them through the air. In other words, the power required to turn the rotor at its operational RPM is pretty minimal. Now, when the pilot is ready to go flying, he or she begins by pulling up on a handle in the cockpit called the "collective." When this happens, the blades on the rotor go up to a positive pitch. All the blades on the main rotor do this together at one time—"collectively." Of course, they are then putting out a lot of lift, since they have a positive pitch. Equally apparent is the fact that they are harder to drag through the air now, since they are doing a lot more work. Of course, since it is a lot harder to turn the blades, they start to slow down—if this were allowed to happen, it would be catastrophic, since the craft can’t fly when its rotor isn’t turning! To compensate, at that point any modern helicopter will automatically increase the throttle as much it needs to in order to maintain the desired 400 RPM in the rotor.

To summarize, this is the sequence for getting a helicopter in the air in X-Plane:

1. While on the ground, the collective handle is flat on the ground. This means the rotor pitch is flat, with minimum drag and zero lift. In X-Plane, a flat collective corresponds to the throttle being full forward, or farthest from the user. The automatic throttle in the helicopter is obsessively watching the rotor’s RPM, adjusting the throttle as needed to hold exactly 400 RPM in the example above. On the ground, with the collective pitch flat, there is little drag on the blades, so the power required to hold this speed is pretty low.
2. When the user decides to take off, s/he does so by raising the collective up by pulling it up from the floor of the helicopter. In X-Plane, this is done by easing the throttle on a joystick back down toward you. This increases the blade pitch on the main rotor and therefore increases its lift, but it also increases the drag on the rotor a lot. The rotor RPM begins to fall below 400 RPM, but the auto-throttle senses this and loads in however much engine power it has to in order to keep the rotor moving at exactly 400 RPM.

3. More collective is pulled in until the blades are creating enough lift to raise the craft from the ground. The auto-throttle continues adding power to keep the rotor turning at 400 RPM no matter how much the collective is raised or lowered.

Once the craft is in the air, the first-time helicopter pilot’s first crash is no doubt beginning. This inevitability can be delayed for a few moments using the anti-torque pedals.

The main rotor is of course putting a lot of torque on the craft, causing it to spin in the opposite direction (because of course for every action there is an equal and opposite reaction—the rotor is twisted one way, the helicopter twists the other way). This is where the anti-torque pedals come in. The rotational torque on the helicopter is countered with thrust from the tail rotor. Just push the left or right rudder pedal (such as the CH Products Pro Pedals) to get more or less thrust from the tail rotor. If rudder pedals aren’t available, the twist on a joystick can be used for anti-torque control. If the joystick used does not twist for yaw control, then X-Plane will do its best to adjust the tail rotor’s lift to counter the main rotor’s torque in flight.

Incidentally, the tail rotor is geared to the main rotor so that they always turn in unison. If the main rotor loses 10% RPM, the tail rotor loses 10% RPM. The tail rotor, like the main rotor, cannot change its speed to adjust its thrust. Like the main rotor, it must adjust its pitch, and it is the tail rotor’s pitch that is being controlled with rudder pedals or a twisting joystick.

Once the craft is in the air and the collective pitch of the main rotor is being adjusted (in X-Plane, using the joystick throttle), try holding the craft 10 feet in the air and adjusting the tail-rotor pitch with the anti-torque pedals (e.g., rudder pedals or a twisting stick) to keep the nose pointed right down the runway. From here, the joystick should be wiggled left, right, fore, and aft to steer the helicopter around.

Here is how this works: If the stick is moved to the right, then the rotor blade will increase its pitch when it is in the front of the craft, and decrease its pitch when it is behind the craft. In other words, the rotor blade will change its pitch through a full cycle every time it runs around the helicopter once. This means that it changes its pitch from one extreme to the other 400 times per minute (7 times per second) if the rotor is turning at 400 RPM. Pretty impressive, especially considering that the craft manages to stay together under those conditions! Now, while it seems that the right name for this might be the "helicopter destroyer," the fact that moving the stick sends the blade pitch through one cycle every rotation of the rotor blades means we call the control stick the cyclic stick. So, we have the collective, cyclic, and anti-torque controls.

Let's talk more about the cyclic. When the stick is moved to the right, the rotor increases pitch when it is in the part of its travel that is in front of the helicopter. This will increase the lift on the front of the rotor disc, causing it to tilt to the right—remember that the gyroscopic forces are applied 90 degrees along the direction of rotation of the gyroscope. Now that the rotor is tilted to the right, it will of course drag the craft off to the right as long as it is producing lift.

The fascinating thing is that the rotor on many helicopters is totally free-teetering; it has a completely "loose and floppy" connection to the craft. It can conduct zero torque (left, right, fore, and aft) to the body of the helicopter. Maneuvering is only achieved by the rotor tilting left, right, fore, and aft, dragging the top of the craft underneath it in that direction. The helicopter body is dragged along under the rotor like livestock by a nose-ring, blindly following wherever the rotor leads.

Use the above information to hover perfectly. Once that is mastered, push the nose down to tilt the rotor forwards. The lift from the rotor acting above the center of gravity of the aircraft will lower the nose of the helicopter, and the forward component of lift from the rotor will drag the craft forward as it flies along.

Flying the Space Shuttle

Read this chapter before attempting Space Shuttle landings in X-Plane if you want your virtual pilot to live!

The first rule of flying a glider—quite unlike flying a powered plane—is this: Never come up short. When bringing a powered plane in for landing, if the pilot thinks the craft will not quite make it to the runway, it is no big deal. She or he just adds a bit more power to cover the extra distance. If a little more speed is needed, it is again no problem—just add power.

Gliders play by a different set of rules, though. There is no engine to provide power, so when setting up a landing, a pilot must be sure to have enough altitude and speed to be able to coast to the airport, because if s/he guesses low by even one foot, the craft will hit the ground short of the runway, crashing. Gliders must never be low on speed or altitude, because if they ever are, there is no way of getting it back—a crash is assured. (Thermals, or rising currents of air, provide the exception to this rule. These can give efficient gliders enough boost to get the job done, but thermals will typically provide less than 500 feet per minute of vertical speed—not enough to keep even a lightweight Cessna in the air!)

Now, with the Space Shuttle, it is certainly true that the aircraft has engines—three liquid-fuel rockets putting out 375,000 pounds of thrust each, to be exact. (To put this in perspective, a fully-loaded Boeing 737 tips that scales around 130,000 pounds, so each engine of the orbiter could punch the Boeing straight up at 3 Gs indefinitely. That is not even considering the solid rocket boosters attached to the Shuttle's fuel tank that provide millions of pounds of thrust!)

So, the Space Shuttle has engines; the problem is fuel. The orbiter exhausts everything it’s carrying getting up into orbit, so there is nothing left for the trip down. Thus, the ship is a glider all the way from orbit to its touch-down on Earth. With the final bit of fuel that is left after the mission, the orbiter fires its smaller de-orbit engines to slow it down to a bit over 15,000 miles per hour (that’s right—it slows down to a bit over 15,000 miles per hour!) and begins its descent into the atmosphere.

So, if a user wants to fly the Space Shuttle, and the Space Shuttle is a glider from the time it leaves orbit to the time it touches down on Earth, that user must bear in mind the cardinal rule of gliding: Always aim long (past the landing point), not short, because if ever you aim short, you are dead, because you cannot make up lost speed or altitude without engines. Aim long since the extra speed and altitude can always be dissipated with turns or speedbrakes if the craft winds up being too high, but nothing can be done if it comes up short.

In observance of this rule, the Orbiter intentionally flies its glide from orbit extra high to be on the safe side.

But there is one problem. It would appear that if the Orbiter flies its entire approach too high, it will glide right past Edwards. In reality, this doesn’t happen for the following reason.

For most of the re-entry, the shuttle flies with the nose way up for extra drag, and it makes steep turns to intentionally dissipate the extra energy. The nose-up attitude and steep turns are very inefficient, causing the shuttle to slow down and come down to Earth at a steeper glide angle. If it ever looks like the orbiter might not quite be able to make it to the landing zone, the crew simply lowers the nose to be more efficient and level it out in roll to quit flying the steep turns. This makes the orbiter then glide more efficiently, so the crew can stretch the glide to Edwards for sure. The extra speed and altitude is the ace up their sleeve, but the drawback is they have to constantly bleed the energy off through steep turns (up to 70 degrees bank angle!) and drag the nose up (up to 40 degrees!) to keep from overshooting the field.

We will now walk through the re-entry process from the beginning as it is done both in the real Shuttle and in X-Plane.

After de-orbit burn, the shuttle heads for the atmosphere at 400,000 feet high with a speed of 17,000 miles per hour and a distance of 5,300 miles from Edwards (equivalent to landing in the Mojave Desert after starting a landing approach west of Hawaii—not a bad pattern entry!). In reality, the autopilot flies the entire 30-minute re-entry, and the astronauts do not take over the controls of the shuttle until the final 2 minutes of the glide. The astronauts could fly the entire re-entry by hand, but it is officially discouraged by NASA, for obvious reasons. These speeds and altitudes are way outside of normal human conception, so our ability to "hand-fly" these approaches is next to nil.

During the first one hundred NASA Shuttle missions, the craft was hand-flown for the entire re-entry only once, by a former Marine pilot who was ready for the ultimate risk and challenge.

In contrast, users flying the Shuttle in X-Plane will have to complete the entire mission by flying by hand. There is not yet an autopilot for the Space Shuttle in X-Plane yet.


Go to the File menu and select Load Situation, then click the Space Shuttle: Full Re-entry button. X-Plane will load the craft at around 450,000 feet, in space, coming down at a speed of Mach 20. Control will be limited in space (the craft is operating off of small reaction jets on the Orbiter, set up as "Puffers" in Plane-Maker), but once the shuttle hits atmosphere, there will be some air for the flight controls to get a grip on and the craft will actually be able to be controlled. The ship will first hit air at about 400,000 feet, but it will be so thin that it will have almost no effect.

The airspeed indicator at this point will read around zero—interesting, since the craft is actually moving at over 17,000 mph. The reason for this is that the airspeed indicator works based on how much air is hitting it, just like the wings of the orbiter do. In space, of course, that’s very little. The indicated airspeed will build gradually as the craft descends. Under these conditions, even though the Shuttle is actually slowing down, the airspeed indicator will rise as it descends into thicker air that puts more pressure on the airspeed indicator. This oddity of the airspeed indicator, though, is useful, since the air is also putting more pressure on the wings. This means the airspeed indicator is really measuring how much force the wings can put out, which is really what a pilot is interested in here.

Restated, the airspeed indicator indicates the craft’s true airspeed times the square root of the air density. It indicates lower speeds in thin air, but the wings put out less lift in thin air as well, so the airspeed indicator works very well to tell the pilot how much lift can be put out by the wings.

Note: If the airspeed indicator reads more than about 250 knots, the wings have enough air to generate the lift to carry the aircraft. If the airspeed indicator is showing less than about 250 knots, then the wings do not have enough air hitting them to lift the Shuttle, so it is still more or less coasting in the thin upper atmosphere, where the air is too thin to do much for controlling flight.

As the airspeed indicator on the HUD gradually starts to indicate a value (as the aircraft descends into thicker air), it means the craft is starting to ease down into the atmosphere at 15,000 mph like a sunburned baby trying to ease into a boiling-hot Jacuzzi—very carefully and very slowly. Remember, if the craft was going 15,000 mph in the thick air of sea level, it would break up into a million pieces in a microsecond. The only reason it survives at 15,000 mph up here is the air is so thin that it has almost no impact on the ship. Again, the airspeed indicator tells how much the air is really impacting the craft; 250 knots is a "comfortable" amount. The trick is to get the craft moving much slower than 15,000 mph by the time it gets down to the thick air of sea level—and to have it doing so at Edwards Air Force Base. This is what the re-entry is for, to dissipate speed while descending so that the Orbiter is never going too fast for the thickness of the air that it is in. It should only descend into the thicker air once it has lost some speed in the thinner air up higher. The whole thing should be a smooth process wherein the ship doesn’t get rammed into thick, heavy air at too high a speed.

Now, as the Orbiter begins to touch the outer molecules of the Earth's atmosphere, users will notice a slight ability to fly the ship as some air begins to pass over the wings. At the same time, the HUD should begin showing speed. Notice the picture of the Orbiter on the right-hand EFIS display. The Atlantis already has this display retrofitted over its old steam gauges (the EFISs from the Atlantis are modeled very accurately in X-Plane—astronauts could use it for familiarization for sure). Both the Orbiter and the path down to Edwards should be visible. The goal is to stay on the center path. If the craft gets above it, it is either too fast or too high and might overshoot the landing. If it gets below it, it is either too slow or too low and might not make it.

Remember that the line is drawn with a large margin for error, so if a pilot stays on the line, he or she will have plenty of extra energy. Getting below the line a little will only tap into the speed/altitude reserve. Getting below the line a lot will keep the craft from reaching Edwards.

The Orbiter must stay near the center green line. This green line represents the desired speed for the early part of the re-entry, the desired total energy for the middle part of the re-entry, and the desired altitude for the final phase of the re-entry. This is the way NASA set up the EFIS. If the craft is too fast or too high (meaning it is above the center line) then it is time to dissipate some energy. Put the Shuttle in a steep bank, pull the nose up, and hang on!

The real Orbiter will have it nose up about 40 degrees and be in a 70 degree bank to try to lose energy while moving at 14,000 mph, glowing red hot, hurtling through the upper atmosphere on autopilot, and leaving a ten mile-long trail of ionized gas behind it while the astronauts just watch.

Go into some steep turns to dissipate energy as needed to keep the ship from going above the center green line. Look at the little blue pointer on the far left-hand side of the far right display. That indicates how high the nose is supposed to be. The green pointer is where the nose is now—they need to match. The pointers just to the right indicate the desired and current deceleration. These indicators, though, will not be used to fly by. Look at the little pointer up top on the horizontal scale. That is the computer's estimation of how much bank angle the craft probably needs to stay on the center green line. Pilots should follow the computer's recommendation or their own intuition for how much bank to fly, but they must certainly keep the nose up (in order to stay in the upper atmosphere) and fly steep banks to dissipate the extra speed and altitude. It might be tempting to just push the nose down if the craft is high, but don't. The aircraft would drop down into the thick air and come to an abrupt stop from the tremendous drag, keeping it from ever making it to Edwards. It would wind up swimming in the Pacific somewhere around Hawaii.

Now, as the pilot makes those steep turns, the aircraft will gradually be pulled off course. For this reason, the turn direction should be switched from time to time to stay on course. Turn left awhile, then right, then back to the left again. This is what the real Orbiter does—it slalom-skis through the upper atmosphere at Mach 20. Watch Edwards on the center EFIS display. This is the destination. Hit the ‘@’ key to see the Orbiter on a flyby. Watch carefully—it’s going fast. Hit the ‘w’ key to get back in the cockpit (being sure that the caps lock is off).

As the ship approaches Edwards, right on the center green line on the right-hand display, there should be a sort of a circle out past Edwards. This is the Heading Alignment Cylinder, or H.A.C. The aircraft will fly past Edwards at about 80,000 feet, then fly around the outside of the H.A.C. like it’s running around a dining room table. After coming around, it will be pointed right at Edwards. If the craft is still on the green line, its altitude will be just right for landing as well. In the real Shuttle, this is usually where the pilot will turn off the autopilot and hand-fly in.

The craft should now be doing about 250 or 300 knots, coming down at about 15,000 feet per minute or so (about 125 miles per hour of descent rate). Needless to say, pilots do not want to hit the ground with that 125 miles per hour descent rate. Do not aim for the runway without expecting to become a smear on it. Instead, aim for the flashing glideslope lights 2 miles short of the runway that NASA has thoughtfully provided. If they are all red, the craft is too low. If they are all white, it is too high, so the speed brakes need to be hit using the ‘6’ key or the mouse. If the lights are half red and half white, the Orbiter is right on its glideslope (about 20 degrees). Airliners fly their approach at 125 knots with a 3 degree angle of descent, while the Space Shuttle uses 250 knots and a 20 degree descent angle—not too unusual considering pattern-entry started west of Hawaii, actually.

To recap: the craft should be at 250 knots, on the green line, lined up with the runway. It should be facing half red, half white glideslope lights with the flashing strobes by them. This approach configuration should be held until the craft is pretty close to the ground (3 degree glideslope to the runway), then the descent should be leveled and the gear put down (using the ‘g’ key or the mouse). Pull the nose up for a flare as the runway approaches, causing the Orbiter to touch down smoothly. Lower the nose then and hit the parachute and even the brakes if the craft will be allowed to roll out.

Now, if a user can just repeat that process another hundred times in a row without a single hitch, s/he will be as good as NASA.

Special thanks to Sandy Padilla for most of the Shuttle re-entry information!

Flying on Mars

NASA has very exact data on the atmospheric pressure, density, and temperature on Mars. They also have very exact data on the gravity of Mars, as well as rough topographic maps for the entire planet and very detailed maps for some areas. Furthermore, the laws of physics, which are programmed into X-Plane, are exactly the same on Earth as on Mars. X-Plane needs atmospheric pressure, density, temperature, gravity, and topographic maps to deliver an engineering-accurate flight simulation.

X-Plane can simulate flight on Mars.

Use the "Set planet to Mars" option (found in the Location menu) to go to Mars; use the corresponding "Set planet to Earth" option to go back.

Introductory Letter

The following is an email sent by Austin Meyer, author of X-Plane, to the X-Plane community, at 4:35 AM on February 24, 2000. It is reprinted here in its original, coffee-fueled form.

I DID POSSIBLY THE MOST EXCITING THING I HAVE EVER DONE TONIGHT. (OK, technically I finished it THIS MORNING). As some of you may know, I have been gathering data on Martian atmosphere, gravity, surface "texture", and topography for X-Plane from various NASA sites ( , for example) I do NOT yet have the TOPOGRAPHY for Mars, but I DO have everything else, and I have gotten it all entered into X-Plane and designed two planes to fly on Mars as well, and have been experimenting with deign and flight on Mars for the last 6 hours or so. (Could I be the first human to fly a real-time flight simulation of Mars? I have seen many "movies" of "flying" over Mars terrain, but NONE have been hooked to an actual realistic FLIGHT MODEL... has NASA done a REAL-TIME simulation of Mars flight in a PILOTED aircraft? Has ANYONE?) Well, I have for the last 6 hours, AND IT IS FRIGGIN FASCINATING. First of all, the atmosphere is ONE PERCENT as thick on Mars as it is on earth... INDICATED airspeed is proportional the square root of the air density, so the INDICATED airspeed is ONE TENTH the true airspeed. The result? If you take off with 60 knots on the airspeed indicator, your REAL speed is SIX HUNDRED KNOTS! (about Mach 1) Take it from me, Mach-1 takeoffs are quite a thing to behold, when the plane will barely leave the runway at that speed. While there is almost no AIR for you, you do have the (sort of) advantage of only about ONE THIRD the GRAVITY, so it is three time easier to get airborne! Result? A take-off in a well-designed airplane can occur at a "mere" 400 knots or so, indicating all of 40 knots on the airspeed indicator! Sound easy? IT ISN'T, BECAUSE WHILE YOUR GRAVITY (WEIGHT) IS ONLY ONE-THIRD OF EARTH'S, YOUR ==>INERTIA<== IS STILL THERE IN FULL FORCE! So you are flying with only 1/3 the total lift of what you are used to having to stay in the air, which seems fine UNTIL IT COMES TIME TO TRY TO TURN OR FLARE!!!!! THEN you see that while the lift for STAYING airborne is only 1/3 of Earth's, the INERTIA, and thus the lift needed to CHANGE DIRECTION (this includes the landing flare!) IS STILL THERE IN FULL FORCE! The problem is, you DON'T HAVE THAT KIND OF LIFT, SINCE THE AIR IS SO THIN! Bottom line: All airplanes on Mars are AIRBORNE TITANICS: Ripping blissfully along, unaware of their impending doom due to their inability to TURN against their tremendous inertia. Landings are impossible without arresting gear. If you can work the flare out right (it IS possible with advance planning) then you will touch down doing about 400 mph. Now how do you stop? ->PARACHUTE? NOPE!!!! 400 mph is only 40 mph worth of drag due to the thin air. You will run off the end of the runway going 100 mph with the chute only "seeing" 10 mph: USELESS for slowing down ->BRAKES? NOPE!!! You only have one-third gravity, so only 1/3 of your weight on the wheels. NO TRACTION! ->Reverse thrust? NOPE!!!! With only 1% atmosphere, jet or prop engines can put out basically no thrust... just barely enough to keep the airplane in flight at mach-0.85.. the jet plane needs a JATO to take off! So how do you stop? I finally went with ARRESTING GEAR. I know of no other way to avoid blasting off the end of the runway at 200 knots with the chute uselessly deployed and brakes uselessly locked. Speaking of which, CRASHES are interesting. No air drag to slow the tumbling planes down, and little gravity to drag them to a stop against the ground! Crashes look like "the Agony of Defeat" from the Olympics where the guy on the downhill ski-jump bites it near the top of the ramp and tumbles on and on and on, powerless to stop an accident that started hundreds of yards earlier! (though on mars, at 400 mph, your plane will tumble across the plains for MILES!) CRUISING ALONG OVER MARS is SPECTACULAR, with the scary red-orange Martian sky, new Martian rocky-red terrain textures, VISIBLY thinner air(!) (due to modified lighting in OpenGL, modified fog in OpenGL, and visibility of stars).. you really can tell you are halfway between air and space! Returning to Earth, you feel like you are flying in soupy water! Yuk! So what sort of planes can fly on Mars? Not anything from Earth, that's for sure. Not enough lift or thrust. A Cessna or Boeing will just sit there on the ground without even moving. Put them in the air and they drop like beveled bricks with no wings. Both of my Mars-plane concepts are much like the U-2 Spy plane (designed to operate at around 100,000 ft, in similar density air) one with a HUGE high-bypass jet engine built AROUND THE FUSELAGE, and another with a smaller rocket engine in the tail, like the X-15. The rocket plane has a lower-thrust engine, with plenty of fuel, for about 30 minutes of flight or so... the JET plane can fly for hours! My designs are realistic (again, based on the U-2, with reduced weight for the lower structural needs (lower gravity) and modern (composite) materials). The rocket-plane is pretty much guaranteed feasible (known technology across the board) but the jet-powered one I am not sure about since Mars has so little OXYGEN in the atmosphere it may be impossible to keep a turbofan engine running. (My Mars jet-plane has twice the average fuel-consumption, though, to simulate injection of liquid oxygen or nitrous oxide). Bottom line, I now know it IS possible to build and fly a piloted plane on Mars and I now know what it would be like. (though I used a 10,000 ft runway with arresting wires... none of those on Mars now I admit).


With the latest versions of X-Plane, the Martian terrain is finally available. We’ll now discuss touring that terrain.

X-Plane has planet-wide Mars elevation data thanks to NASA's Mars Orbiting Laser Altimeter (or MOLA, a satellite orbiting Mars that gathered terrain elevation data on the entire planet). All of this scenery has been gridded for X-Plane. Contrary to common assumption, Mars is not flat, with a few meteor impacts here and there—not even close!

For example, even though Mars is one half the radius of Earth (with one quarter the surface area), it has canyons that make our Grand Canyon look like a fish pond (30,000 feet deep!). It has a volcano 65,000 feet tall. The atmosphere is gone to essentially pure vacuum at 155,000 feet, so if one climbs the volcano, he or she is about halfway to space! Mars has far more topographic variety than Earth, on a planet with only a quarter of the surface area, so the sight-seeing by air is intense. The max allowable visibility in X-Plane has been raised to 60 miles when on Mars so that the grand vistas can be taken in.

Aircraft designed for Earth will simply not fly on Mars. It is recommended that one of the aircraft found in the Mars Planes folder (within the Aircraft folder) be used.

Taking off and landing is interesting. A pilot is tempted to pick a nice, high spot to make an airport so that there are no obstructions to a landing approach. However, the air is so thin up there that aircraft can hardly fly! In that case, maybe a pilot wants an airport at the bottom of a canyon or meteor crater where the air is thicker. That could work, but watch out for the canyon or crater wall when approaching and departing, or face impacting the crater wall at the speed of sound! Of course, the speed of sound is around the minimum speed needed to fly on Mars.