Heatblur Simulations posted a lengthy description about the Inertial Navigation System of their upcoming F-14 Tomcat. They detail how they tried to reproduce accurately the system performance, including its imperfection. Inertial Navigation Systems are tracking the aircraft movement in space to understand its position. These systems do not rely on radio or GPS navigation, and aren’t 100% accurate. Some conditions and actions may lead the system to accumulate some deviation –  Heatblur is working hard reproducing the system and its flaws to the finest degree of realism.

This is a great introduction about inertial navigation system – it is an interresting description that fit other aircrafts as well!

 

“Dear All,

As mentioned previously; one of our main goals is to release the F-14 as complete as possible. One big item that we’re beginning to fully square away is the navigational systems, and in particular a new simulation of the inertial navigation system. The navigational systems tie in heavily with elements such as the TID and CAP, and we’ll go into more detail on these aspects in the future.

Enjoy a technical, physics based update on the navigational systems below, written by F-14 developer Krzysztof Sobczak (Ph.D Physics)!

Introduction

A good combat jet should provide the crew with means to navigate without external navigational aids or guidance. The way to achieve it is to equip the aircraft with an inertial navigation system (INS). An INS system measures and integrates sensed inertia forces (acceleration) and rotational velocities to calculate aircraft position and linear velocity. A good navigation system can precisely guide an aircraft on a route to a mission objective hundred or thousand miles-long, and then back to the home base, safely and reliably. Such a system is even more important when an aircraft is designed to operate over the ocean, far away from any ground-based TACAN or visual references.

The INS used on the F-14 is a multi-unit Carrier Aircraft Inertial Navigation System (CAINS) designated as AN/ASN-92. As you have already discovered, this system is the centre of this development update.
AN/ASN-92 features
The AN/ASN-92 INS is the primary navigation system on the F-14 and provides the crew and the other aircraft systems with:

  • Current latitude and longitude;
  • Attitude;
  • Heading true and magnetic;
  • Own ground speed and ground track;
  • Ability to store and display three waypoints, a fixed point (FP), an initial point (IP), a surface target (ST), a home base (HB), a defended point and a hostile area;
  • Range, bearing, command course, command heading and time-to-go to a selected destination point;
  • Calculated wind speed and direction;
  • Calculated magnetic variation;
  • Continuous monitoring of the status of the unit, and in case of failure inform the crew with advisory lights and appropriate acronyms displayed on the TID;
  • Backup navigation modes in case of partial system failure.

Although from the crew member’s point of view, the INS is used mostly for navigation, it is also essential for proper operations of other aircraft equipment. For example, the attitude is necessary for the radar. The attitude and the own position are required for some weapon delivery modes, particularly for long shots. Even more distressing to the crew, a complete failure of the INS renders weapons such as the AIM-7 inoperable.

The same information is used for data-link operations – when using erroneous INS data, own tracks and targets received from cooperating aircraft will not match and result in false contacts being displayed on the TID. These are only a few examples, and the INS data is used whenever aircraft position or attitude is required.
Construction and principles of operation

AN/ASN-92 is built from multiple components, but there are two particular components which constitute the core of the system: the inertial measurement unit (IMU) and the navigation computer.

The IMU is a three-axis, four-gimbal, all-attitude unit containing two gyros and three accelerometers. The gyros and the accelerometers are mounted to a platform that is free to rotate respect to the base (aircraft). The four-gimbal system provides gimbal-lock free rotation and uses torquer motors to correct platform attitude errors. The gyros sense angular rotation about their sensitive axes and are the source of information about the aircraft attitude. They also stabilise the whole platform and keep the constant orientation of the accelerometers respect to the ground. Two accelerometers are used to measure acceleration in the horizontal plane; the third accelerometer measures vertical acceleration. The sensitive axes of the accelerometers are orthogonal. The sensed acceleration signal is integrated in the computer and used to calculate aircraft velocity and displacement from the initial position. The attitude of the platform is also corrected continuously to account for the effects associated with the Earth’s rotation and device inaccuracies.

This design is widespread for gimballed inertial navigation systems. It was used for the F-14, but also for the Space Shuttle and many other aircraft of the era. Below, you can find a sketch of an IMU from the JA37 flight manual – this model is almost the same as the model used for the F-14.

An INS device like the AN/ASN-92 requires a high precision of measurements of the acceleration and the attitude, because even the smallest inaccuracy can result in a significant error when accumulated over extended time.

Consider an example: the inertial platform is slightly tilted from the nominal position, let’s say by 0.002 degrees. Then, the horizontal accelerometers are no longer parallel to the ground, and this means that they start to be sensitive to gravity. If not corrected, this gravitational component is interpreted by the navigation computer as a horizontal acceleration. If the wrong attitude is kept constant for one hour, it will result in an error of the measured position of over one nautical mile. It is a significant inaccuracy, and it comes as a result of such a minimal alignment error.

The accuracy of the INS degrades with time – usually the longer they operate in the navigation mode, the higher the error they accumulate.

INS alignment procedure

An INS device must be prepared before it is ready for navigation. This process is called alignment. Before the alignment begins, the RIO has to input aircraft coordinates and altitude.

Upon selection of the alignment mode, alignment routines are read into the computer and the first stage – the coarse alignment – is initiated. The platform is levelled using the accelerometer output, and the initial rough estimation of the aircraft heading is performed.

The second stage – fine alignment – uses the precise measurement of gyroscope drift to calculate aircraft’s true heading. This is possible because of the Earth’s rotation and utilises the mentioned before Shuler tuning. At no point of alignment, is the magnetic heading used, and the whole process relies only on the sensing of the non-inertial movement of the platform within the 3d space.

For shore-based operations, the whole alignment process should be finished within 8 minutes. It is possible to pre-align the aircraft on the ground, which allows for a quick-reaction response. This reduces the alignment time to 2 minutes but requires aircraft to be tied down in the alert position.

Carrier-based alignment is slightly more complicated than ground alignment because the ship is constantly moving. Thus, to support the process, ship’s INS data is transfer to the aircraft using data-link or deck-edge cable. The carrier-based alignment process should complete in 10-12 minutes. In case of the ship’s data being unavailable, ship’s true heading and speed have to be manually entered by the RIO.

Performance

A fully aligned AN/ASN-92 INS, in accordance with the requirements of the navy specification, for the latitude of 45 degrees North, should provide the following performance:

  • 3 arc minutes for heading,
  • 2.5 arc minutes for pitch and roll,
  • Position error rate of 0.75nm per hour (CEP),
  • Velocity error of 3 feet per second.

All values stand for standard deviation and assume a normal distribution of the error.
The RIO can decide to finish the alignment and switch to the INS navigation mode at any point after coarse alignment criteria have been met. However, a premature selection of the navigation mode will significantly degrade the navigation quality.

In-flight alignment of the F-14 INS is impossible. In case of an in-flight INS failure or a takeoff without proper INS alignment, two additional backup navigation modes are available. They provide dead-reckoning navigation using attitude information from the IMU or the AHRS (Attitude and Heading Reference Set), airspeed from the CADC (Central Air Data Computer), stored wind data and magnetic variation.

The RIO can improve (restore) the precision of the INS in-flight by updating the aircraft position:

  • With the radar by locking on the known reference point (waypoint);
  • Using TACAN signal and known coordinates of the TACAN station stored as a waypoint;
  • By overflying a visual reference point;
  • Using data-link, either when flying in close formation, or by hooking a radar track of the cooperating aircraft.

Updating the aircraft’s INS position in flight may introduce a greater error than before the update, and the accuracy is limited by the precision of the method used to update. Thus, updating has a greater usefulness when utilized as a backup navigation method when navigation stability is significantly reduced.

Simulation

Designing an INS (IMU) is an engineering challenge, which requires consideration of such problems as calibration, alignment, Earth’s rotational motion, inertia forces, thermal stability, analogue-digital converters precision, all different types of correction which have to be applied to keep the device precise over extended time, and many more. Simulating an INS platform is very similar – it is a complex undertaking.

At Heatblur, we decided to develop an entirely new mathematical model to simulate the AN/ASN-92 for our F-14. We included all the potential sources of errors contributing to the final precision of the device, and recreated the characteristic behaviour of a gimballed INS platform. The result is a set of algorithms providing an authentic representation of the AN/ASN-92 in DCS, yet optimised to have almost no impact on CPU performance.

Because a picture is worth a thousand words, below, you can find a plot which is the result of a test run of our simulation of the AN/ASN-92.

In this example, the aircraft was parked, and the initial misalignment of the IMU was equal to 0.0005 degrees – i.e. it was relatively low. The figure represents the magnitude of the INS calculated position error as a function of time. As you can observe, the rate of change is not constant, and there are periods when the magnitude of the error decreases. This oscillatory behaviour is a known effect, described by German engineer Maximilian Schuler in 1923 ( https://en.wikipedia.org/wiki/Schuler_tuning). The theoretically predicted period of this oscillation is equal to 84 minutes. In our model, those oscillations come as a natural product of the simulated physical processes. Finally, as you can see, the aircraft does not have to move for the IMU to accumulate errors.

From the functional point of view, our simulation of the AN/ASN-92 is an authentic virtual representation of the real unit and contains all features described above. Most of them are already implemented, and the final missing bits should be finished within the next weeks.
You can expect that your INS will:

  • Let you navigate to any destination point;
  • Drift and will not be 100% accurate;
  • Communicate with other aircraft systems and simulated INS inaccuracies will affect their performance;
  • Require proper alignment;
  • Use stored heading alignment method reducing the alignment time to less than 2 minutes if the appropriate checkbox is selected in the mission editor;
  • Sometimes fail and force you to use the backup modes.

Test flight results

To conclude this update, we would like to present to you a record of two test flights which we performed using our F-14 in DCS. We took off from Nellis AFB, climbed to 12000 ft, flew to Creech, then hit the dirt and turned to Groom Lake. After a zoom climb over Groom Lake, we descended back to ground level and flew straight to Lake Meade, crossed it to the Hoover Dam, passed Boulder City, and then back to Nellis. The whole route was almost 40 minutes long.

Before the first flight, we let the INS run the alignment until the “fine align” status was reached – it took a bit less than 8 minutes. The second flight was preceded by a partial alignment, stopped after 4 minutes.

For both flights, we recorded the true aircraft position and the INS calculated position.
We loaded the exported data to Google Earth and prepared graphics comparing the true flight path and the INS-sensed flight path.

The error of the calculated INS position at the end of the flight was equal to 0.4 nm for the fully aligned case, and over 4 nm for the 4 minutes long alignment.
If you want to take a closer look at the results, you can download the recorded flight paths and open them in Google Earth:

Fine alignment: https://drive.google.com/open?id=1yI…qvBaA6hnUHwDQx
4 minutes alignment: https://drive.google.com/open?id=1Bx…M3ZHA6QqRaxZKE

Many thanks for reading!
We will return to the AN/ASN-92 in one of the next development updates when we take a closer look at the practical side of using the navigation systems of the F-14.

As always, thank you for the support!

Sincerely,
Heatblur Simulations

LEAVE A REPLY

Please enter your comment!
Please enter your name here