AN ASSESSMENT OF MODERN METHODS FOR

ROTOR TRACK AND BALANCE

THESIS

Michael J. Renzi, Ensign, USNR

AFIT/GAE/ENY/04-J11

DEPARTMENT OF THE AIR FORCE

AIR UNIVERSITY

AIR FORCE INSTITUTE OF TECHNOLOGY

Wright-Patterson Air Force Base, Ohio

APPROVED FOR PUBLIC RELEASE; DISTRIBUTION UNLIMITED.

The views expressed in this thesis are those of the author and do not reflect the official
policy or position of the United States Navy, United States Air Force, Department of
Defense, or the United States Government.
AFIT/GAE/ENY/04-J11

AN ASSESSMENT OF MODERN METHODS FOR ROTOR TRACK AND

BALANCE

THESIS

Presented to the Faculty

Department of Aeronautics and Astronautics

Graduate School of Engineering and Management

Air Force Institute of Technology

Air University

Air Education and Training Command

In Partial Fulfillment of the Requirements for the

Degree of Master of Science in Aeronautical Engineering

Michael J. Renzi, BS

Ensign, USNR

June 2004

APPROVED FOR PUBLIC RELEASE; DISTRIBUTION UNLIMITED.

AFIT/GAE/ENY/04-J11

AN ASSESSMENT OF MODERN METHODS FOR ROTOR TRACK AND

BALANCE

Michael J. Renzi, BS
Ensign, USNR

Approved:

Donald L. Kunz (Chairman) date

LtCol Robert A. Canfield (Member) date

Maj Richard G. Cobb (Member) date

AFIT/GAE/ENY/04-J11
Abstract

One routine maintenance item facing the helicopter industry today is the issue of

rotor track and balance (RT&B). While the task of reducing vibrations is often

overlooked as simply an unimportant “maintenance” concern, what should not be

overlooked is the extensive amount of time and money committed by maintenance to

smoothing an aircraft.

If there were a way to make the process of rotor track and balance more efficient

it would be a huge boost to the helicopter industry in both time and money. The first

steps towards research into new and improved methods is to evaluate what is currently

used in the field, determine if there is room for improvement and if so what can be

improved.

While each company may use a slightly different approach to correct the problem,

each method has essentially the same objective─ to reduce vibrations in the helicopter

structure due to main and tail rotor rotation.

This document reflects the findings of a study done to gather information and evaluate

the different RT&B methods that currently exist, pinpointing the existing weaknesses in

the process. In most all cases, a qualitative approach was used in determining problems

and comparing current systems as the actual proprietary algorithms used by RT&B

companies were unavailable.

iv
 Acknowledgements

I would like to acknowledge my advisor, Dr. Kunz for his help in the thesis

process and the U.S. Army Aviation & Missile Command, Aviation Engineering

Directorate, Aeromechanics Division (AMSAM-RD-AE-A), Redstone Arsenal, AL for

their contributions to this document, especially Dr. Jon Keller and Mr. Eric Bale. I

would also like to thank Dr. Mark Hollins of Naval Air Warfare Center Aviation Division

(NAWCAD), Patuxent River NAS, MD for his contributions to this document.

Michael J. Renzi

v
 Table of Contents

Page

Abstract ............................................................................................................................. iv

Acknowledgments................................................................................................................v

List of Figures .................................................................................................................. viii

List of Tables .......................................................................................................................x

List of Symbols .................................................................................................................. xi

1. Introduction ............................................................................................................. 1-1

1.1 General ............................................................................................................ 1-1

1.2 Problem Statement .......................................................................................... 1-1
1.3 Objectives........................................................................................................ 1-1
1.4 Research Methods ........................................................................................... 1-2
1.5 Chapter Summary............................................................................................ 1-2

2. Background .............................................................................................................. 2-1

2.1 Introduction ..................................................................................................... 2-1

2.2 Defining Track and Balance............................................................................ 2-1
2.3 Aircraft Harmonics.......................................................................................... 2-1
2.4 Brief History.................................................................................................... 2-2
2.4.1 Rotor Tracking History .................................................................................. 2-3
2.4.2 Rotor Balance History ..................................................................................... 2-4
2.5 Typical Track and Balance Procedure............................................................. 2-5
2.5.1 RT&B Equipment............................................................................................ 2-5
2.5.2 Aircraft Adjustments .................................................................................... 2-11
2.5.3 Typical Smoothing Operation ...................................................................... 2-17
2.6 Why Track and Balance? .............................................................................. 2-18
2.6.1 Vibration Affecting Humans ........................................................................ 2-18
2.6.2 Vibration Affecting Component Life ........................................................... 2-26
2.7 Chapter Summary.......................................................................................... 2-28

3. Algorithm Types....................................................................................................... 3-1

3.1 Introduction ..................................................................................................... 3-1

3.2 Algorithm Overview........................................................................................ 3-2
3.2.1 Non-learning Algorithm .................................................................................. 3-2
Page

vi
 3.2.2 Learning Algorithms ....................................................................................... 3-5
3.3 Chapter Summary............................................................................................ 3-6

4. Discussion, Conclusions and Recommendations ..................................................... 4-1

4.1 Introduction ..................................................................................................... 4-1

4.2 Blade Uniformity............................................................................................. 4-1
4.3 Rotor Adjustments........................................................................................... 4-4
4.4 Equipment Reliability...................................................................................... 4-8
4.5 Mechanic Education ........................................................................................ 4-8
4.6 User Error ........................................................................................................ 4-9
4.7 Learning Algorithms ..................................................................................... 4-12
4.8 HUMS ........................................................................................................... 4-13
4.9 Conclusions and Summary ............................................................................ 4-14

Appendix A. Fort Eustis Experimental Test Results ..................................................... A-1

Bibliography ...............................................................................................................REF-1

vii
 List of Figures

Figure Page

1. Accelerometer ................................................................................................ 2-5

2. Accelerometer mounted on tail of OH-58 Kiowa.......................................... 2-6

3. Closeup of Figure 2........................................................................................ 2-6

4. Tachometer .................................................................................................... 2-7

5. Tracker ........................................................................................................... 2-8

6. Tracker mounted to CH-47 Chinook ............................................................. 2-8

7. Data acquisition unit and user interface......................................................... 2-9

8. Typical equipment setup .............................................................................. 2-10

9. Weight pins on AH-64 Apache.................................................................... 2-12

10. Weight added to blade on AH-64 Apache ................................................... 2-12

11. Pitch link on AH-64 Apache........................................................................ 2-13

12. Typical blade cross section .......................................................................... 2-14

13. Airfoil with trim tab added .......................................................................... 2-14

14. Trim tab on blade of AH-64 Apache ........................................................... 2-15

15. Trim tab on OH-58 Kiowa Blade................................................................. 2-16

16. Tab bending tool .......................................................................................... 2-16

17. Subjective human vibration response curves............................................... 2-20

18. Mean errors (5 counters) by level ................................................................ 2-21

19. Vibration and no vibration errors for each counter...................................... 2-21

20. Mean error (by frequency) for each counter................................................ 2-22

21. Mean heading tracking error scores............................................................. 2-24

viii
22. Workload versus vibration........................................................................... 2-25

23. Frequency spectrum ....................................................................................... 3-1

24. Coefficient development using a polar plot ................................................... 3-3

25. CH-47 blade construction .............................................................................. 4-1

26. Patched blade ................................................................................................. 4-2

27. Pitch link adjustment on CH-47 Chinook...................................................... 4-5

28. Pitch link lock on CH-47 Chinook ................................................................ 4-6

29. Pitch link lock on CH-47 Chinook ................................................................ 4-6

30. U.S. Army AH-64 Apache adjustable pitch link design.............................. 4-10

31. U.S. Army UH-60 Black Hawk adjustable pitch link design ...................... 4-11

ix
 List of Tables

Table
Page

1. Total aircraft system comparison, reliability and corrective maintenance .. 2-27

2. Life-cycle savings per aircraft resulting from vibration reduction .............. 2-30

x
 List of Symbols

Acronyms
AMRDEC Aviation Missile Research Development and Engineering Center
ATABS Automated Track and Balance Sets
AVA Aviation Vibration Analyzer
Cpm Cycles per minute
DSS Dynamic Solutions Systems
HUMS Health Usage and Monitoring System
IMD Integrated Mechanical Diagnostics
ips inches/sec
NAWCAD Naval Air Warfare Center Aircraft Division
rpm revolutions per minute
RT&B Rotor Track and Balance
THUMS Total Usage and Monitoring System
VATS Vibration Analysis Test Sets
VMEP Vibration Management Enhancement Program

xi
 AN ASSESSMENT OF MODERN METHODS FOR ROTOR TRACK

AND BALANCE

I. Introduction

1.1 General

Helicopter rotor track and balance (RT&B) is currently a major high-cost

maintenance item in today’s fleet of helicopters. Smoothing vibrations in today’s

helicopters involves an extensive amount of maintenance man-hours and aircraft flight

hours. [Keller, 2004] High maintenance time eats away at the life-cycle usefulness of the

aircraft, burns high cost fuel, and detracts from the operational readiness of the aircraft.

For these reasons, the most efficient method possible of smoothing main and tail rotor

vibrations should be used to reduce all the costs of this necessary task.

1.2 Problem Statement

What are the current methods used to smooth helicopter vibrations induced from

the main and tail rotors? Which method is most efficient? What are the biggest areas of

improvement that can be made in today’s popular techniques for smoothing helicopter

vibrations?

1.3 Objectives

The objective of this research is to determine what methods currently exist for

helicopter rotor track and balance, and pinpoint weaknesses and suggest areas that need

improvement.

1-1
1.4 Research Methods

There is very little written material on the subject of helicopter rotor track and

balance. Therefore, research required TDY trips to Aviation Missile Research

Development and Engineering Center (AMRDEC) at Redstone Arsenal, AL , Naval Air

Warfare Center Aircraft Division (NAWCAD) at Patuxent River NAS, MD, and a trip to

the U.S. Army National Guard Unit in Akron, OH. Additionally, phone interviews were

conducted with representatives in the field of RT&B. Incidentally, all photos in this

document, unless otherwise noted, were taken on one of the TDY trips mentioned above.

1.5 Chapter Summary

Helicopter rotor track and balance is a major maintenance cost of today’s

helicopter fleets. For this reason, the most efficient method of smoothing should be

employed to save on the high cost of man-hours, fuel, aircraft flight hours, and aircraft

unavailability. This document will report on the major players in the RT&B market and

point out areas that could use improvement.

1-2
 II. Background

2.1 Introduction

Before specific discussion of how rotor track and balance is performed, a more

general background is necessary on the basics of rotor track and balance and why it is

necessary

2.2 Defining Track and Balance

What exactly do we mean by rotor track and balance? Historically, the term

“track” refers to the actual vertical location of each blade tip while the rotor is spinning.

When the tips of each blade are all passing through the same plane, the helicopter is said

to have a perfect track. The term “balance” refers to both the mass balance and the

aerodynamic balance of the rotor. The problem is more complex than say balancing a

wheel/tire on an automobile because the rotor and blade assembly has inherent

aerodynamics that must be considered as well. The terms “track” and “balance” aren’t

necessarily the best terms to describe this maintenance process. The real objective of

RT&B is to smooth the aircraft by reducing vibrations created by the rotating main and

tail rotors. Whether or not the aircraft rotor blades are in perfect “track” is really not

important (as pioneers once thought) if the aircraft structure is vibration free. Helicopter

rotor smoothing is a more generic term which better suits the process but the term “rotor

track and balance” is the industry term for the process. [Johnson, History].

2.3 Aircraft Harmonics

A helicopter is a complex machine of systems and subsystems, many of which are

rotating or moving in some way. Among the rotating components are the main and tail

2-1
rotors. The two rotors are responsible for the majority of the violent vibrations felt in a

typical helicopter. This discussion will be limited to the main rotor.

Inherent to the spinning rotors are different types of vibration and different

vibration frequencies. The main rotor produces both vertical and lateral type vibrations.

Vertical vibrations are most commonly the result of aerodynamic differences between the

blades resulting in asymmetrical lift characteristics. The most common reason for a

lateral vibration is a mass imbalance somewhere in the rotor assembly. Again,

asymmetries are the typical cause for this. Another reason for lateral vibration is two or

more blades out of vertical balance causing a rolling motion in the helicopter.

The main rotor head has several inherent vibration frequencies associated with its

rotation. The largest amplitude of vibration due to the main rotor occurs at a frequency

equal to the rotating speed of the rotor. RT&B is designed to smooth this “1-per-rev”

frequency of vibration from the spinning rotor. Vibrations at higher frequencies, namely

n-per-rev frequencies, are inherent in the spinning rotor. While some companies claim to

also be able to smooth the higher harmonics with their equipment, the reality is that only

the 1-per-rev can be directly adjusted. Any other improvements are an indirect result of

the 1-per-rev adjustment. [Robinson, 1]

2.4 Brief History

The process of RT&B, like anything else, has evolved over the years, as

technology has improved and new ideas have been tested.

2-2
2.4.1 Rotor Tracking History

Over the past few decades, several methods of rotor tracking have been

employed. The first method used was flag tracking. In this case, the tip of each blade is

marked with a different colored chalk or grease pencil and allowed to pass by a white rag

or “flag” mounted to a long, lightweight, vertical pole. With the helicopter on the

ground, the flag was moved slowly toward the rotor disk until contact was made with the

blade tips. The relative path of each blade could be seen on the rag and then appropriate

adjustments could be made to correct for these “track splits.” This method was

dangerous and did not allow for track measurements off the ground. [Robinson, 2]

The next track method employed was called electro-optical tracking. Developed

in the 1960s by Chicago Aerial, this method relied on opto-electronics to determine rotor

track. A photographic image of each blade tip is captured by the tracker during a

revolution and compared to determine relative track of each blade. [Johnson, History]

Then, in the late 60's/early 70's, Chadwick-Helmuth invented a method called

strobe light tracking. While a competent method, its downside is that it requires

significant operator skill level. Making use of a strobe light and reflective blade tips, this

method requires the operator to manually adjust a dial “and visually locate a group of

targets in space and remember their relative locations.”[Johnson, History] This method

was abandoned by most mainly because of the skill required to perform it properly.

Then, in the 1980's, the electro-optical method became popular again, primarily

due to new developments by Stuart Hughes. Up-to-date electronics and technology made

this the method of choice and it is still the primary method used today. [Johnson, History]

2-3
2.4.2 Rotor Balance History

The first methods used to balance helicopter rotor blades were little more than

simply a static balance of blades, trying to match weights in order to create as

symmetrical a rotor weight as possible. Soon enough there was a demand for improved

smoothing methods, so technicians found themselves mounting vibration measurement

devices on the rotor in conjunction with a strobe light to determine the amplitude of

imbalance and phase angle. This procedure would show how much weight needed to be

added and where it needed to be added, much like a tire-balancing machine. Soon, there

was a recognized need to be able to account for not only a mass imbalance but also an

aerodynamic imbalance of the rotor blades. Techniques were developed using pitch

links, trim tabs, and weights to methodically adjust blade angle of attack, camber, and

weight step by step to balance the rotor system. The technician would plot amplitude and

phase of vibration for each iteration and then make adjustments based on a table of values

created for that particular aircraft. These “manual” algorithms required a lot of skill on

the part of the technician and therefore made it hard to get good results. [Robinson, 2]

In the 1980's, as computers began to become more powerful and available,

computer based algorithms were developed that made use of vibration measuring devices

to develop recommended adjustments to the rotor head/blades. New digital technology

has overtaken older analog methods. Digital equipment is much faster, accurate, more

economical, and offering many more options. Vibration measurement devices in

conjunction with computer algorithms is the approach used today in helicopter smoothing

[Robinson, 2].

2-4
2.5 Typical Track and Balance Procedure

Generally speaking, RT&B procedures are very standard throughout the industry.

Type of equipment and helicopter adjustments are virtually the same throughout the

industry.

2.5.1 RT&B Equipment

While there are quite a few different companies that produce RT&B equipment,

each having a method with their name on it, the premise of each method is very much the

same. Each system uses some type of vibration sensor. A piezo-electric accelerometer,

similar to the one depicted in Figure 1, is commonly used to sense vibration in a chosen

direction.

Figure 1. Accelerometer

2-5
Figures 2 and 3 shows an accelerometer mounted on the tail of a U.S. Army OH-58

Kiowa located at Redstone Arsenal, AL. Figure 2 depicts a typical mount location for an

accelerometer measuring vibration in the vertical plane.

Figure 2. Accelerometer mounted on tail of OH-58 Kiowa

Figure 3. Closeup of Figure 2

2-6
 In addition to the accelerometers (or other type of vibration sensor), a tachometer is

also needed for input of rotor speed. The typical mount location for the tachometer is on

the non-rotating ring of the swashplate. Figure 4 shows an example tachometer used for

this application. The tachometer shown in Figure 4 is a magnetic pickup which produces

an electrical impulse when a ferrous metal interrupter mounted to the rotating ring of the

swashplate, passes close by.

Figure 4. Tachometer

The third input commonly used is from the tracker. The example tracker shown in

Figure 5 is currently used by the U.S. Army for RT&B. Figure 6 shows the same tracker

in Figure 5 mounted to the structure of a U.S. Army CH-47 Chinook (owned by the

Akron, OH National Guard Unit). The tracker collects data on the relative location of the

2-7
 Figure 5. Tracker

Figure 6. Tracker mounted to CH-47 Chinook

2-8
blade tips in the rotating disk. The above equipment is then wired to some type of data

storage unit, ready to receive data. Usually, a user interface allows the user to interact

with the system and visualize the vibrations. Figure 7 shows an example of a data

acquisition unit and a user interface. Incidentally, the example equipment seen in Figures

1-7, is part of the U.S. Army’s Aviation Vibration Analyzer (AVA) system. A typical

RT&B equipment configuration on the aircraft is shown in Figure 8.

Figure 7. Data aquisition unit and user interface [AVATM, 2-1]

2-9
 Figure 8. Typical equipment setup [AVATM, 1-2]

The problem of rotor track and balance is relatively specialized, so therefore,

helicopter manufacturers rely on outside companies to develop solutions for their aircraft.

Traditionally, RT&B systems have been stand alone items bought and sold exclusively

for smoothing. These systems were usually separate from the aircraft, requiring

equipment hookup and wiring to be run each time a RT&B was on the maintenance

schedule. Recently, however, there has been an industry-wide push towards integration

of RT&B vibration measurements into an aircraft Health Usage and Monitoring System

or HUMS. The HUMS system not only collects data required for rotor smoothing, but

monitors vibrations from different onboard equipment as well. The HUMS systems are

collecting data continuously for monitoring of overall aircraft health. The idea is to

detect trends in vibration data, indicating component failure or preferably a warning

which gives the time left to component failure. While stand-alone RT&B systems

certainly still exist, the preferred data collection is through some sort of HUMS.

2-10
2.5.2 Aircraft Adjustments

Typically, the data received from the above hardware is fed through an algorithm

which determines the best adjustments to make to smooth the aircraft vibrations.

Possible adjustments that can be made to the helicopter rotor/blade assembly include

weight addition or subtraction, blade sweep adjustment, pitch link adjustment, or trim tab

adjustment. While, obviously, each company designs their aircraft slightly differently,

the RT&B adjustments are all very similar from aircraft to aircraft. Most helicopters

have similar designs for adding weight. On a typical four-bladed helicopter, the root of

two adjacent blades has weight that is added or taken away. Each blade has the

capability of having weight added to it but the two adjacent blades are the only ones

touched because adding weight to a blade is the same as subtracting from the opposite

one. Weight adjustments correct for lateral vibrations in the aircraft. Figure 9 shows an

example of the weight pins found on the U.S. Army’s AH-64 Apache. Figure10 shows

another blade on the same Apache with weight added to it.

The U.S. Army’s UH-1 Huey is a two-bladed helicopter that is designed to hold

lead pellets in the rotor head. If an imbalance is detected and weight must be added,

ounce-sized lead pellets are added to either side to correct the imbalance. Some

helicopters have the ability to adjust blade “sweep,” moving it forward or aft. This

adjustment is also made for lateral vibration smoothing.

Pitch link adjustments are another common adjustment made to smooth a

rotorcraft. The swashplate moves in response to the cyclic and collective inputs of the

2-11
 Figure 9. Weight pins on AH-64 Apache

Figure 10. Weight added to blade on AH-64 Apache

2-12
pilot, and pitch links make the connection between the rotating ring of the swashplate and

the pitch varying housing on the rotor blade. The pitch link on each blade can be

adjusted

separately and therefore, the angle of attack of each blade can be adjusted individually.

Typically, a helicopter will have a locking turnbuckle that can be extended or

compressed, therefore changing blade angle of attack.

Figure 11 shows an example of an adjustable pitch link on a U.S. Army AH-64

Apache. In this case the turnbuckle is extended to increase angle of attack and

Figure 11. Pitch link on AH-64 Apache

compressed to decrease angle of attack. Angle of attack of the blade affects the drag and

lift properties of the individual blade. Usually correcting for vertical vibrations will call

2-13
for a pitch link adjustment, however, as an example, increasing the angle of attack on a

particular blade will produce the desired increase in lift, but it also increases drag on the

blade causing the blade to want to “lag.” Any change in lead/lag can affect the lateral

balance. In this way, the adjustments are closely related and affect one another.

The other adjustment that can be made to correct for vertical vibration is a trim

tab adjustment. Figure 12 shows a typical airfoil cross-section without a trim tab.

Figure 12. Typical blade cross section [Chinook]

Trim tabs are attached to the trailing edge of each blade and usually they are designed to

be adjusted by manually bending them up or down by the prescribed angle amount. They

are essentially the same as a “flap” on a fixed wing aircraft. The purpose of the trim tab

is to effectively increase or decrease camber of the airfoil (blade) which, according to

basic aerodynamic theory, increases or decreases blade lift. Figure 13 shows the same

LIFT INCREASES
DRAG INCREASES

TRIM TAB OR “FLAP”

Figure 13. Airfoil with trim tab added [Chinook]
2-14
blade cross section as Figure 12 except a trim tab has been added to the trailing edge.

Again, lift and drag always go hand and hand, meaning an increase in lift also gives an

increase in drag. Figure 14 shows the trim tab on a blade of a U.S. Army AH-64 Apache.

The Apache is unique in that the trim tab spans the entire length of the blade.

Figure 14. Trim tab on blade of AH-64 Apache

Most manufacturers have only short trim tabs on their helicopter blades. Figure 15

shows the trim tab attached to the trailing edge of a blade found on the U.S. Army’s OH-

58 Kiowa. Trim tabs are adjusted using a specially designed tool. Figure 16 shows an

example of a trim tab bending tool made by Chadwick-Helmuth. The handle of the tool

is pushed or pulled vertically for bending while the dial indicator measures the bend

2-15
amount. The Kaman H-2 is unique in that it was designed with a dynamic trim tab that

can be adjusted by the pilot in flight until the aircraft was smooth. [Whitten, 2004]

Figure 15. Trim Tab on OH-48 Kiowa Blade

Figure 16. Tab bending tool [www.chadwick-helmuth.com]

.

2-16
2.5.3 Typical Smoothing Operation

If an aircraft rolls off the assembly line or a blade is replaced, a smoothing

operation must be performed from the beginning. The first step in a smoothing operation

is to statically balance the rotor blades. Static balance is accomplished by matching the

chord moment and span moment of all the blades. A test jig with load cells is usually

used to accomplish this. Weights are added to the blade in various locations or

"pockets," set up by the manufacturer, to correct for non-uniformities.

The second step is to achieve perfect ground track. The aircraft rotor assembly is

brought up to speed and the tracker is used to measure track differences in the individual

blade track. Then, based on the measurements, usually pitch link adjustments are made

to correct for these track differences. Perfect track is the starting point for the vibration

smoothing process.

After ground track is accomplished, a ground vibration check is performed. The

rotor assembly is, again, brought up to speed on the ground while vibration data is

collected. This data is used to determine what adjustments need to be made to correct

vibrations. The goal for the ground vibration check is to bring the aircraft within safe

acceptable flying limits.

As soon as vibration levels are low enough to permit safe flying, test flights can

be performed. Vibration levels in the aircraft structure are very dependent upon flight

conditions. Vibrations are known to vary primarily with airspeed. Ideally, vibration data

would be taken at every speed to create a complete database of the vibration state.

However, this would be very impractical, as the aircraft can fly at an infinite number of

speeds. The convention in industry is to use a finite number of acquisition speeds. Each

2-17
company may have a different set of flight conditions to establish their test conditions but

a common example would be to collect data 1) at hover 2) 80 knots forward flight 3) 120

knots forward flight 4) 150 knots forward flight. A test flight is done and data is

collected at each flight condition. After flight data is collected, the data will be used as

input to an algorithm, which determines the appropriate adjustments necessary to smooth

the aircraft. Each of these test flight/data acquisition/aircraft adjustments together are

known as one iteration in the RT&B process. Obviously, the goal is to smooth the

aircraft to acceptable levels in as few iterations a possible.

Usually, RT&B is performed during routine intervals as part of regular

maintenance. Any time changes are made to the rotor/blade assembly (e.g. a blade is

replaced), a rotor track and balance must be performed.

[Studer, 2004]

2.6 Why Track and Balance?

Why is rotor smoothing important? Excessive vibration levels over time can lead

to crew fatigue, avionics damage, and structural fatigue of the aircraft. Obviously, any

one of the above can be detrimental to the mission and therefore, should be avoided.

2.6.1 Vibration Affecting Humans

Research has shown the detrimental effects vibration can have on the human body

functions. A study was published jointly by the Office of Naval Research and Boeing

[Human Factors, 1969] in which the subjective human response to different vibration

levels is defined. Figure 17 shows the result of the subjective test illustrating the

different levels of reaction to vibration. In this same study, effects of vibration on human

sensory-motor processes was noted. The study investigated the effect human vibration

2-18
has on speech and hearing intelligibility, visual performance, horizontal and vertical

compensatory tracking, response to workload, and general effects on crew activities.

Results showed that while not every sensory-motor process was affected (at least from

the results of this particular study) there were significant negative effects from the

vibration levels. The speech and hearing intelligibility test was inconclusive, as "[it was]

reported that binaural threshold increased with vibration, but non-systematically. The

changes which were noted were judged too small to have operational significance.

Vibration likewise had little effect on speech intelligibility, except that subjects tended to

speak in short bursts under the lower frequency vibration conditions." [Human Factors,

28] The results of the visual performance study concluded that "vibration severity

and frequency affected readability." [Human Factors, 28] Figures 18, 19, and 20 show

the results of the visual performance study. The explanation for the dramatic decrease in

readability at 10Hz is explained by the coinciding critical flicker frequency of the human

eye. Another study tested different lighting conditions in combination with different

vibration levels. Again, the results consistently show a trend in reduced reading

performance with increased vibration levels. An interesting finding resulting from this

study was that red light seemed to prevent reading decrement in comparison with white

light.

Human horizontal and vertical compensatory tracking was also studied. Results

showed that vertical tracking was negatively affected by vibration, while horizontal was

not affected severely. Figure 21 shows the results of the test. "The relationship between

2-19
2-20
Figure 17. [Beaupeurt, 20]
 Figure 18. [Beaupeurt, 30]

Figure 19. [Beaupeurt, 30]

2-21
2-22
Figure 20. [Beaupeurt, 31]
vibration severity and compensatory tracking was explored by Chaney and Parks, whose

seven subjects performed wheel, column, and foot tracking tasks under vertical whole

body vibration . . . of special interest . . . is their finding that tracking proficiency

decreases systematically as vibration severity increases." [Human Factors, 38] In fact the

study "found that the immediate effects of vibration were indicated on the terrain

following task in which tracking under vibration was 39 percent poorer than tracking

without vibration." [Human Factors, 41]

Vibration and workload was also studied, by evaluating the test subjects' ability to

adjust a three-inch dial indicator using a large and small knob, vertical and horizontal

levers, and vertical and horizontal thumbwheels. Time to complete the task was the

primary evaluation means. Figure 22 shows the results of this testing. Increased

vibration seemed to consistently increase errors and time required to make proper

adjustments.

In summary, the Boeing/ONR report showed that there is a direct relationship

between vibration level and degradation of human sensory-motor processes.

2-23
2-24
Figure 21. [Beaupeurt, 37]
2-25
Figure 22. [Beaupeurt, 40]
2.6.2 Vibration Affecting Component Life

Excessive vibration also shortens the life of components and subsystems found on

an aircraft. A study performed in Fort Eustis, Virginia by the Sikorsky Aircraft Division

explored the effects of vibration on helicopter reliability and maintainability. [Veca,

1973] The study compared two equivalent groups of USAF H-3 helicopters, one group

was equipped with a rotor mounted bifilar vibration absorber and the other without this

absorber. The vibration absorber found on the H-3 helicopter absorbs the n-per-rev

higher harmonics of vibration from the main rotor. RT&B works to correct the 1-per-rev

vibration directly while indirectly correcting the n-per-rev. The absorber does not

replace the need for RT&B.

The Fort Eustis testing showed a significant increase in failure rate of the

helicopter group without the vibration absorber. The vibration absorber decreased

vibration by 54.3% and as a result the overall aircraft failure rate dropped by 54% and

corrective maintenance was reduced by 38.5%. In addition, life cycle costs were shown

to be reduced 10% by the reduced vibration levels. In the end, the study showed that

improved reliability and reduced maintenance/life-cycle costs can be achieved by

reducing vibration levels. The findings also suggest that the aircraft's useful life can be

extended purely by a reduction in vibration levels.

Table 1 shows a comparison of data pulled from USAF records on normal

maintenance and subsystem reliability of the two groups of H-3s. Failure rates were

calculated by taking the ratio of total number of recorded failures and total accumulated

flight hours, averaged for the two groups. The second comparison in Table 1 was

calculated by taking the ratio of total required maintenance man-hours to total flight

2-26
hours of the aircraft. Table 1 illustrates the dramatic impact of excessive vibration on

health of the aircraft.

Table 1. [Veca, 27]

2-27
 Aircraft component reliability is also affected by vibration. The Fort Eustis study

compared component reliability on board both groups of H-3s. The results of the

comparison show a dramatic difference in reliability for nearly every component.

Appendix A includes tables comparing reliability which clearly show the negative effect

vibration can have on component reliability.

Vibration reduction also decreases life-cycle cost of the aircraft. Table 2

illustrates the life-cycle savings per aircraft resulting from vibration reduction. The study

found that savings over the life-cycle of the aircraft from vibration reduction totaled over

$350,000, which is approximately a 10% savings.

The results of this study suggest that component failure can be blamed on fatigue

over time, and that reducing vibration levels can therefore increase the usable life of the

aircraft components. An increased component life and reliability also saves maintenance

man-hours and, therefore, money as proven by the cited study.

2.7 Chapter Summary

Track and balance are two terms describing the dynamics of the rotating rotor

blade assembly. Track refers to the relative path of the blade tips to one another, while

balance includes a mass balance as well as an aerodynamic balance. The primary method

of track used today is electro-optical tracking and the method of balancing has evolved

into computer aided algorithms. Necessary equipment for a RT&B process includes

accelerometers, a tachometer, a tracker, some type of data acquisition unit, and a user

interface. The equipment is used to gather vibration data on the aircraft, which leads to

rotor adjustments to smooth the aircraft. Typical adjustments include: weight changes,

pitch link lengthening or compressing, and trim tab adjustments. RT&B is performed on

2-28
an aircraft to decrease vibration levels. Decreased vibration levels are desired because

excessive vibration has a negative effect on both human performance and component life.

2-29
 Table 2. [Veca, 94]

2-30
 III. Algorithm Types

3.1 Introduction

The best method of visualizing helicopter vibrations for analysis is through use of

a polar chart. Amplitude and phase angle of vibration are plotted on the polar chart.

Figure 23 shows an example of a polar plot used to plot vibration level. Industry

standard for vibration amplitude is to report it in ips (inches per second). The

accelerometers measure in units of g’s. This data undergoes a fast Fourier transform

(FFT) from the time domain into the frequency domain, and is then integrated to give

magnitudes in mean vibration velocity rather than g's. Figure 23 shows an example of a

frequency spectrum which is vibration amplitude (ips) plotted vs. frequency (usually

cpm). The dominant 1-per-rev frequency can be seen followed by the higher harmonics

of decreasing amplitude.

Figure 23. Frequency spectrum [www.dssmicro.com]

3-1
3.2 Algorithm Overview

All current RT&B systems work very similarly. Each method relies on

predictable aircraft response to adjustments. A database of response to different

adjustments is created empirically for each aircraft or "type" of aircraft. Test flights are

performed where single adjustments are made beforehand and the response of the aircraft

from the vibration data is recorded. After a database is "complete," the information can

be used by an algorithm in future smoothing problems to determine the proper

adjustments to make to achieve a smooth aircraft. The "trick," or proprietary part of the

algorithm is developing the proper adjustment solution from an arbitrary vibration

magnitude and phase angle. In each adjustment case, only discrete adjustments can be

made, therefore, the algorithm must account for this and develop a compromise in the

case where an initial solution calls for a fractional adjustment to be made. [Studer, 2004]

Current algorithms can be categorized into two categories: non-learning and learning.

3.2.1 Non-Learning Algorithm

The non-learning type algorithms rely on aircraft response to adjustment to be

consistent with that type of aircraft, relying on the initial database developed, as

discussed above. Figure 24 shows an example of a polar plot used to visualize aircraft

vibration. In this example, a preliminary test flight is done without any adjustment to the

aircraft. This initial test establishes a baseline, and in this example, the baseline is a

vibration of 0.5 ips at a phase angle of 45º. After establishing a baseline, an arbitrary

weight adjustment is made on this aircraft, adding 5 oz. of weight to the black blade.

Then, a second test flight is performed, with data recorded just as before in the first test

flight. The results of the second test flight show a new vibration amplitude of 0.5 ips at

3-2
225º phase angle. These two test flights show that for this particular aircraft, adding

weight to the black blade results in 0.2 ips of magnitude change per one oz. of weight

added, in the direction of 225º on the polar chart. The 0.2 ips/oz. is merely one of the

many "coefficients" developed and entered into the database for that particular aircraft.

The algorithm then calls on this information from the database when developing an

adjustment solution for the particular vibration magnitude and phase angle on the aircraft.

[Keller, 2004]

Figure 24. Coefficient development using a polar plot

The companies using a non-learning type algorithm are Chadwick-Helmuth,

Dynamic Instruments, Goodrich, Helitune, Intelligent Automation Corporation, RSL

Electronics, and Smith Aerospace.

3-3
 Chadwick-Helmuth is a company based out of El Monte, CA. Recently, the

company was purchased by Honeywell, and now calls itself Honeywell Chadwick.

Chadwick-Helmuth was one of the original RT&B companies, started in 1954 by Jim

Chadwick. The company carries a broad range of products designed to smooth vibrations

in both fixed and rotary wing aircraft. The product range starts with the entry level

“Vibrex 2000” and ends with a “Vibralog” HUMS. [Welborne, 2004]

Dynamic Instruments is a company based in San Diego, CA. Dynamic

Instruments is the creator of the ATABS/VATS (Automated Track and Balance

Sets/Vibration Analysis Test Sets) RT&B system used by the U.S. Navy/Marine Corps.

It has been around since the early 1980s and is just recently starting to be replaced by a

new Goodrich HUMS. [Whitten, 2004]

Goodrich (formerly known as BFGoodrich) is a large company based out of

Charlotte, NC. Recently, (past ten years) they won the contract with the U.S. Navy and

U.S. Marine Corps for its new IMD HUMS (Integrated Mechanical Diagnostics Health

and Usage Monitoring System). Still in the testing phase at Patuxent River NAS, this

system is a big upgrade from the portable ATABS/VATS system currently being used by

the Navy/Marine Corps. [Hollins, 2004]

Helitune is a small company located in Devon, England. Helitune supplies

RT&B systems primarily to the U.S. Coast Guard. They offer several different systems

for rotor smoothing in their product line known as “Rotortuner.” [www.helitune.com]

Intelligent Automation Corporation (IAC) is the developer of the new system

undergoing testing by the U.S. Army. Their system is a HUMS known as VMEP

(Vibration Management Enhancement Program) or AVA II (Aviation Vibration

3-4
Analyzer). The VMEP is a culmination of lessons learned over the last fifteen years or

so, integrated into a permanent, on-board management system. [Keller, 2004]

RSL Electronics is a company based out of Israel. RSL recently released a new

T-HUMS (Total Health Usage and Monitoring System). [www.rsl.co.il/]

Smith Aerospace is a large company with offices located both in the U.S. and

England. Originally, Stuart Hughes (of the UK) and Scientific Atlanta teamed up

together to create a RT&B system (which is known today as AVA). Then the two

merged to create a company called Global Associates. Global Associates was then

renamed later to Smith Aerospace, which is the company’s name today. AVA is the

primary system used today by the U.S. Army. [Studer, 2004]

3.2.2 Learning Algorithms

There are two companies which currently use learning algorithms to develop

recommended rotor adjustments. Learning type systems use neural networks. The

results of each adjustment are used to continually update the coefficients. The idea is to

develop a database that is specific to the one particular aircraft, with hopes that future

adjustment recommendations will be more effective than the last.

Dynamic Solutions Systems (DSS) is a small company located in San Marcos, CA.

DSS produces a system they call "MicroVibe". The MicroVibe is a learning-type

algorithm similar to what is used by ACES. MicroVibe only recommends a single

adjustment per iteration. The downside to this is that this system typically requires more

iterations than systems that call for multiple moves to be made. [Johnson, 2004]

ACES is a company located in Knoxville, TN specializing in aviation

maintenance solutions. Primarily they deal with balancing different rotating assemblies

3-5
on aircraft. They carry four different products which perform helicopter rotor track and

balance. The ACES systems use an algorithm based on only one adjustment per iteration.

Each iteration tweaks the algorithm slightly, in an attempt to get more accurate results in

future tests. The ACES system is very similar to the system used by Dynamic Solutions

Systems. [Lucas, 2004]

3.3 Chapter Summary

There are two primary types of algorithms in use today for RT&B, learning and

non-learning. Non-learning relies on a database developed empirically for each aircraft

type used to predict an aircraft’s response to adjustments. A learning type algorithm is

similar, however, the database is continually updated with every iteration made to

develop a more accurate aircraft database. Both types are seen among the available

systems found today in the RT&B industry.

3-6
 IV. Discussion, Conclusions and Recommendations

4.1 Introduction

Helicopter rotor track and balance is a very costly part of a helicopter’s

maintenance. Costs include: wear and tear on the aircraft, fuel, crew maintenance time,

and unavailability of aircraft for mission. For this reason, every helicopter owner wishes

to perform the very important smoothing of the aircraft as efficiently as possible. There

are certain “problem” areas which currently plague the smoothing process.

4.2 Blade Uniformity

Blade uniformity is one of these problem areas. The reason for needing a rotor

smoothing process is the result of non-uniformities. Each rotor blade, in fact, is not made

equally. Today’s rotor blades are relatively complex in construction. Figure 25 shows

Figure 25. CH-47 Blade Construction (Chinook)

4-1
how the blades on the CH-47 Chinook are constructed. Manufacturing techniques are not

perfect. In fact, the consensus in the helicopter industry is that the blades are far from

perfect and practitioners assume or simply count on the fact that no two blades are

identical. So even beginning with brand new blades, there will almost always be an

imbalance in the rotating assembly. But manufacturing irregularities aside, there are

other reasons for non-uniform blades. The military performs many of its missions with

the aid of helicopters. Some of these missions include combat scenarios in which there is

a chance of blade damage from enemy fire. It is common for a military helicopter blade

which has been damaged in combat or otherwise to be patched and put back in service.

Figure 26 shows an example of a damaged blade which has been patched.

Figure 26. Patched Blade

4-2
Obviously, with a hole blown through a blade and then patch material used for a repair,

the static and dynamic properties of the blade will change.

Another reason for non-uniformity in the rotor blades is the construction materials

used. The honeycomb core of the blade is porous and capable of absorbing moisture. In

wet, tropical climates the blades will absorb water, which, depending on the severity of

the conditions, can become a big problem in terms of uniformity.

Another problem facing helicopters, currently serving, especially in the Middle

East, is blade erosion. Blade erosion is a function of quantity and type of debris in the

air. When the debris is sand, as in the Middle East case, blade erosion is accelerated

dramatically. The erosion along the leading edge is not necessarily perfectly uniform, so

it can create a permanent imbalance in the rotor assembly that must be corrected.

Simply from a blade uniformity standpoint, an initial rotor smoothing process is

always necessary to correct for non-uniformities from the manufacturing. It is true that

quality control can be increased and the manufacturing process could be tweaked so that

all blades leave the factory nearly perfect, if not perfect, but when the cost is considered

for the number of helicopters in service, this would be impractical. Plus, eventually, the

work environment will change the blade uniformity. So, the bottom line is that blade

non-uniformities cannot be eliminated unless drastic changes are made in design and

manufacture. Therefore, these non-uniformities must be accepted and dealt with.

[Keller, Bale, 2004]

4-3
4.3 Rotor Adjustments

As stated previously, the RT&B process is very time consuming. U.S. Navy’s

Petty Officer Scott Beckman stationed at Patuxent River NAS explained how the

maintenance crews really push to get a helicopter smooth within their eight hour shift,

with a typical smoothing operation taking between five and six hours. AFIT’s very own

Capt. Justin Eggstaff, USMC, who is a AH-1 Cobra pilot, talked similarly about the time

required to smooth a typical helicopter. He also spoke of a time when he remembers it

taking over 15 hours to smooth an aircraft! [Eggstaff, 2004]

So what about the operation is so time consuming? Well, the test flight itself

takes time. However, it is necessary to gather vibration sensor data while the aircraft is

in flight. The test flight will always be necessary. Part of the big push towards the

integrated HUMS is that test data can be collected while the aircraft is already in the air

for a mission related flight. The argument can be made, however, that if vibration levels

are too high, the aircraft will be grounded for mission related flights. In this case,

specific RT&B flights are necessary. Other than the test flights themselves, which

usually last no longer than ½ hour, the mechanics must make the required rotor

adjustments called for by the program. Making these rotor adjustments is by far the most

time consuming part of the RT&B process. On a trip to the Akron, OH, U.S. Army

National Guard Unit, smoothing was being performed on a CH-47 Chinook. After a test

flight, a pitch-link adjustment had to be made on a blade on the rear rotor. This one pitch

link adjustment took roughly 1 ½ hours! Figure 27 shows this pitch link adjustment

being performed. Figures 28 and 29 show the wire lock being installed on this pitch link.

Weight addition

4-4
 Figure 27. Pitch link adjustment on CH-47 Chinook

or subtraction usually isn’t too difficult or time consuming. Most mechanics prefer to

avoid trim tab adjustments because they can be tricky. Common adjustments are in 0.001

in. of tab movement, which often can be difficult to achieve accurately. Some tabs are

physically difficult to bend as well, leading to problems with the adjustment. The fatigue

life of the tab is a factor as well. Some manufacturers put a limit on the number of tab

adjustments due to fatigue concerns. [Studer, 2004]

Rotor track and balance adjustments are not easy to make. No matter what

algorithm is used to find the adjustment solution, the fact remains that these time

4-5
Figure 29. Pitch link lock on CH-47 Chinook

Figure 28. Pitch link lock on CH-47 Chinook

4-6
consuming adjustments must still be made. If manufacturers designed easier adjustment

methods for their aircraft, required smoothing time would decrease dramatically. The

best design would be one which allowed adjustments to be performed midflight. Weights

that could remotely move radially as well as circumferentially around the rotor head,

pitch links that could be remotely lengthened or shortened, and trim tabs which could be

remotely adjusted, similar to the pilot adjustable trim tab found on the H-2, would be an

ideal design. These adjustments could be made by the pilot from the cockpit, midflight,

to essentially “tune” the rotor assembly until it was perfectly smooth (much like the old

method of balancing a tire). As discussed previously, environmental conditions have a

big impact on blade-uniformity, so this type of design would be very advantageous in

climates that require frequent smoothing (e.g. sandy environments that cause heavy blade

erosion).

Disadvantages to this type of design would include the likely weight addition to

the existing designs, adding to the life-cycle costs of the aircraft, and the initial cost of

the aircraft would likely increase. Advantages would be the huge savings in maintenance

and required flight time of the aircraft, which can really add up, considering the

frequency of smoothing operations and the time and cost of each iteration. Often initial

cost and weight are the primary considerations to purchasing an aircraft with

maintenance requirements being secondary, explaining why this design is not

widespread. For example, in the military, RT&B is not considered a big priority to the

contract bidders. Main emphasis is on mission effectiveness vs. initial cost/weight.

However, considering the life cycle maintenance costs, RT&B should be a priority, and

ease of adjustment should be demanded of the manufacturers and should be pushed for in

4-7
future designs. In the meantime, adjustment designs on existing aircraft must be

accepted, as they cannot be redesigned and retrofitted (easily).

4.4 Equipment Reliability

Equipment reliability is another key factor in the timeliness of smoothing

operations. If the equipment is unreliable, with one essential component broken, then

obviously, the smoothing operation will fail. Often equipment failures can be subtle,

making troubleshooting difficult. A fatigued wire which has broken inside its insulation

can be tough to spot. If an accelerometer fails, reporting erroneous data, the smoothing

operation will fail. Equipment with built-in diagnosis can dramatically decrease

troubleshooting time. [Keller, 2004]

4.5 Mechanic Education

As a follow-on to the above discussion, in order to be an effective troubleshooter,

a good complete understanding of the system is necessary. One area of the RT&B

process that needs serious improvement is maintenance technician education. In general,

the more educated a person is about the task, the more effective he or she is at performing

the task. This couldn’t be more relevant than in the process of RT&B. The more familiar

the user is with the system, usually the more effective he or she is with succeeding in

smoothing the aircraft. Because of the large number of variables, troubleshooting is often

an important job of the technician. As stated earlier, sometimes it is impossible to

smooth an aircraft with the current blade configuration. For example, if there is a

problem blade that makes smoothing impossible, the maintenance technician must

recognize that because any effort put into smoothing will be futile. Perhaps one of the

4-8
blades is irregular, or perhaps there is a bad component on the aircraft preventing the

smoothing operation from succeeding (for example, a bad damper). If the maintenance

technician does not realize this, he or she could attempt to find balancing solutions

forever without results. [Bale, 2004]

Eric Bale of Avion, Inc. is contracted by the U.S. Army to educate maintenance

crews about proper RT&B procedures. He calls himself the “myth buster.” With every

class he teaches, he corrects false notions of the mechanics, which often interfere with

successful smoothing. One of the most common errors of a mechanic is to not trust the

equipment. Many mechanics will modify adjustment recommendations to whatever he or

she thinks is the proper adjustment, thinking he or she knows better than the equipment.

The RT&B equipment is designed to work properly. Assuming all components are

working properly, and setup is correct, any given adjustment recommendations are the

proper adjustments to make.

4.6 User Error

Sometimes user error can occur. Most systems include printers to print out

required adjustments, or at least have the option of printer hookup. Printers are highly

recommended for use with RT&B equipment. The sequence from when the equipment

determines the proper adjustments to make, relays this information to the mechanic, and

then the mechanic actually makes the required adjustments may seem trivial, but it is

very important that this sequence is performed error-free. If the system does not have a

printer to print out required adjustments, it is common for the mechanic to write the

required adjustments on his or her hand for example. Addition or subtraction of weight,

extension or compression of pitch links, or movement of a trim tab are all indicated by a
4-9
+ or -. If these crucial signs are smudged on a hand, and a + is interpreted as a -, then

obviously the process will fail. Also, even if the correct adjustments are acknowledged,

mistakes can still be made. It is not always immediately obvious which way the pitch

link turnbuckle should be rotated to achieve the desired effect. A very easy mistake to

make is to

Figure 30. U.S. Army AH-64 Apache adjustable pitch link design
accidentally rotate the pitch link turnbuckle the wrong way, leading to smoothing failure.

[Bale, 2004] Proper training can help reduce the chance of mechanic error. Figure 30

4-10
and 31 show two different pitch link designs. Figure 30 shows a AH-64 Apache pitch

link, designed by McDonnell Douglas. The adjustment increment used is a “flat.” The

Figure 31. U.S. Army UH-60 Black Hawk Adjustable Pitch Link Design

pitch link has a hexagon perimeter in the center, with the algorithm telling the mechanic

how many of these “flats” to rotate and which direction. Using flats as a reference can be

vague, especially when dealing with fractions of flats, where the rotation amount is really

left to the interpretation of the mechanic.

A much better design is found on the Sikorsky H-60. Figure 31 shows the design

of an adjustable pitch link found on the H-60. Instead of using “flats” as a rotation

reference, Sikorsky uses notches located around the perimeter of the pitch link, secured
4-11
from rotation with a cotter pin. This style design removes all the guess work from the

adjustment. All recommendations are given with a definite number of notches to rotate

the turnbuckle. This type of “foolproof” design is what manufacturers should be aiming

for in their designs.

4.7 Learning Algorithms

Aircraft are designed to be smoothable in every flight regime. Over time,

however, some aircraft develop problems where they can no longer be smoothed in every

regime. A compromise must be made when choosing which adjustments to make. The

consensus is to concentrate on smoothing the flight regime used most by the aircraft.

Unexpected aircraft response is often attributed to fatigue and wear on the

structure/components which changes the properties of the structure so that the structure

no longer matches the original design specs. In these cases, smoothing may also be

difficult because many system algorithms rely on data that was collected from the aircraft

when it was new. So, gradually the actual aircraft will no longer match the model used

by the RT&B algorithm and will not respond appropriately to any adjustments made.

[Bale, 2004]

This is where the argument for a learning algorithm becomes strong. If the

algorithm is tweaked with every iteration, any property changes to the airframe are

compensated for. Initially, having a system that “learns” sounds like a great idea. The

big disadvantage to the learning system comes when there is user error. If the smoothing

equipment develops an adjustment or set of adjustments to be made, and the adjustments

are made incorrectly (for example, moving a pitch link turnbuckle the wrong way by

4-12
accident), this can create some serious errors in the future performance of the algorithm.

After an incorrect adjustment is made, assuming the user believes that he or she made the

correct adjustment, when, for example, he or she made the opposite of the recommended

adjustment, the algorithm will update its coefficients based on the response of the aircraft

to the opposite of the indicated adjustments. Any future adjustment recommendations

made by the algorithm will be made using the incorrect data input, rendering the

algorithm corrupt. Indeed, the idea of a learning algorithm is promising. However,

before this type of algorithm is implemented in the hands of the average maintenance

technician, it must be deemed fool-proof first. An adjustment system must be designed

in which it is impossible to make an incorrect move by mistake.

The ACES and DSS systems are learning systems geared towards the “seasoned

pro” who is conscious about making correct moves every time. Plus, these systems only

call for one adjustment per iteration, so if an incorrect move were made, it would be

immediately obvious after the next test flight. However, while it is likely the user will

recognize if he or she made an adjustment error, the algorithm will not, and the erroneous

coefficient change cannot and will not be fixed. These companies should include an

option to return the coefficients back to where they were before the previous iteration to

allow for cases when the user recognizes an error. [Johnson, 2004]

In the case of a non-learning algorithm, the appropriate resolution for an aircraft

which no longer responds to adjustments as it “should” or used to, is for a new script file

to be written for that particular aircraft. The new script file will be specific to the aircraft

compensating for any changes that have occurred in aircraft reponse behavior. [Bale,

2004]

4-13
4.8 HUMS

One improvement that is being instituted already is using permanently installed

equipment on board the aircraft for RT&B, usually part of a HUMS. Having

permanently installed equipment eliminates the need for the time consuming equipment

hookups required with traditional systems. The rough, exposed wiring can also become a

hazard, as the U.S. Navy has had at least one fatality from a crew member getting tangled

up in the wiring during a smoothing session. [Brown, 2004] Permanently installed

systems are also more reliable. The equipment components endure less abuse, not having

to be installed and uninstalled at each smoothing. The connection wires will eventually

become fatigued from repeated movement. So, clearly there is an advantage to

permanently installed equipment, in time savings, safety, and reliability. Also, because

the system is permanently installed, data can be collected at any time, so the system can

give an early indication if a rotor smoothing is required. The disadvantage to the

permanently installed system is that each helicopter requires its own system, adding to

the cost of operating a fleet of helicopters.

4.9 Conclusions and Summary

Even with the best algorithm in the world (which many companies would claim

of their own) the adjustments and test flight still must be done. The existing equipment

and algorithms currently in use, do work. When used properly, they will all develop

accurate solutions and allow smoothing to be achieved. The best possible case would be

a complete solution with only one iteration. There are existing algorithms which are

capable of only one iteration for a solution. It seems the majority of problems lie

elsewhere. The major time consuming factors could be eliminated if improvements were

4-14
made in blade quality, rotor adjustment design, maintenance technician education,

eliminating the possibility of user error, equipment reliability, moving towards

permanently integrated systems, and developing “fool-proof” methods so that learning

algorithms could be trusted. If money and R&D were to be put into the RT&B process, it

should be spent trying to fix the problems outlined above rather than developing new

algorithms.

4-15
APPENDIX A
The following Tables and Figures are the result of the

1973 Fort Eustis study.

A-1
A-2
A-3
A-4
A-5
A-6
A-7
A-8
A-9
A-10
A-11
A-12
A-13
A-14
A-15
A-16
A-17
A-18
A-19
A-20
A-21
A-22
A-23
A-24
A-25
A-26
A-27
A-28
A-29
A-30
A-31
A-32
A-33
A-34
A-35
 Bibliography

AVATM. "Operator's, Aviation Unit, and Intermediate Maintenance Manual Including

Repair Parts and Special Tools List for Test Set, Aviation Vibration Analyzer
(AVA). PN 29313107. NSN 6625-01-282-3746. 7 March 2002.

Bale, Eric. Avion, Inc. Instructor. Conversation with. 22 Jan 2004.

Beaupeurt, J.E., F.W. Snyder, S.H. Brumaghim, R.K. Knapp. Ten Years of Human
Vibration Research. "Research on Low Frequency Vibration Effects On Human
Performance." Human Factors Technical Report D3-7888. Office of Naval
Research Contract Nonr-2994(00); The Boeing Company, Wichita, KS. August
1969.

Beckman, Petty Officer Scott. U.S. Navy. AD1. Conversation with. 8 March 2004.

Brown, Nikki. NAWCAD. Technician. Conversation with. 13 April 2004.

Chinook. "The Rotor Blade." http://www.chinook-

helicopter.com/standards/areas/blade.html

Eggstaff, Capt. Justin. U.S. Marine Corps. AH-1 Cobra pilot. Conversation with. 5
March 2004.

Hollins, Mark. NAWCAD. Head Engineer. Conversation with. 13 April 2004.

Johnson, Lloyd. DSS. Director of Engineering. Conversation with. 15 April 2004.

Johnson, Lloyd. "HISTORY: Helicopter Rotor Smoothing." Article. n. pag.

http://www.dssmicro.com/theory/dsrothst.htm. 25 August 2003.

Keller, Jon. AMRDEC. Engineer. Conversation with. 22 Jan 2004.

Lucas, Mike. ACES. Aviation Application Support Specialist. Conversation with. 7

April 2004.

Robinson, Mike. "Helicopter Track and Balance Theory."

http://www.amtonline.com/searchscripts/search.asp?SearchSiteURL=\articles\199
9\amt_02-99\amt_02-99-4.html. February 1999.

Sharp, Don. Goodrich. Program Manager. Conversation with. 15 April 2004.

Studer, John. Smith Aerospace. Head Engineer. Conversation with. 13 Feb 2004.

REF-1
Veca, Angele C. Vibration Effects on Helicopter Reliability and Maintainability. U.S.
Army Air Mobility Research and Development Laboratory Technical Report 73-
11. Fort Eustis, Virginia. April 1973.

Welborne, Murdock. Honeywell-Chadwick. Staff Engineer. Conversation with. 5 April

2004.

Whitten, Paul. Dynamic Instruments. Vice President. Conversation with. 9 April 2004.

REF-2
 Form Approved
REPORT DOCUMENTATION PAGE OMB No. 074-0188
The public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and
maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of the collection of information, including
suggestions for reducing this burden to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (0704-0188), 1215 Jefferson Davis Highway,
Suite 1204, Arlington, VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to an penalty for failing to comply with a collection of
information if it does not display a currently valid OMB control number.
PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.
1. REPORT DATE (DD-MM-YYYY) 2. REPORT TYPE 3. DATES COVERED (From – To)
04-06-2004 Master’s Thesis Jun 03 – Jun 04

4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER

AN ASSESSMENT OF MODERN METHODS FOR ROTOR TRACK AND 5b. GRANT NUMBER
BALANCE
5c. PROGRAM ELEMENT NUMBER

6. AUTHOR(S) 5d. PROJECT NUMBER

Renzi, Michael J., Ensign, USNR 5e. TASK NUMBER

5f. WORK UNIT NUMBER

7. PERFORMING ORGANIZATION NAMES(S) AND ADDRESS(S) 8. PERFORMING ORGANIZATION

REPORT NUMBER
Air Force Institute of Technology
Graduate School of Engineering and Management (AFIT/ENY) AFIT/GAE/ENY/04-J11
2950 Hobson Way
WPAFB OH 45433-7765
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)
Dr. Jonathan A. Keller
US Army RDECOM, AMRDEC,
11. SPONSOR/MONITOR’S REPORT
Director, Aeromechanics Division NUMBER(S)
AMSRD-AMR-AE-A, Bldg. 4488, Rm. B-346
Redstone Arsenal AL 35898-9000
12. DISTRIBUTION/AVAILABILITY STATEMENT

APPROVED FOR PUBLIC RELEASE; DISTRIBUTION UNLIMITED.

13. SUPPLEMENTARY NOTES

14. ABSTRACT

One routine maintenance item facing the helicopter industry today is the issue of rotor track and balance (RT&B). While the task of
reducing vibrations is often overlooked as simply an unimportant “maintenance” concern, what should not be overlooked is the
extensive amount of time and money committed by maintenance to smoothing an aircraft. If there were a way to make the process of
rotor track and balance more efficient it would be a huge boost to the helicopter industry in both time and money. The first steps
towards research into new and improved methods is to evaluate what is currently used in the field, determine if there is room for
improvement and if so what can be improved. While each company may use a slightly different approach to correct the problem,
each method has essentially the same objective─ to reduce vibrations in the helicopter structure due to main and tail rotor rotation.
This document reflects the findings of a study done to gather information and evaluate the different RT&B methods that currently
exist, pinpointing the existing weaknesses in the process. In most all cases, a qualitative approach was used in determining problems
and comparing current systems as the actual proprietary algorithms used by RT&B companies were unavailable.
15. SUBJECT TERMS

Helicopter Rotor Track and Balance, Helicopter Vibration

16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 18. NUMBER 19a. NAME OF RESPONSIBLE PERSON
ABSTRACT OF Dr. Donald L. Kunz
a. REPOR b. ABSTRAC c. THIS PAGE PAGES 19b. TELEPHONE NUMBER (Include area code)
T T
U UU 104 (937)785-3636 e-mail: [email protected]
U U
Standard Form 298 (Rev. 8-98)
Prescribed by ANSI Std. Z39-18