A study of carburetor/induction system icing in general aviation accidents

5-1920

NASA CR- 143835

A STUDY OF CARBURETOR/INDUCTION SYSTEM ICING IN
GENERAL AVIATION ACCIDENTS

Richard W . Obermayer and William T . Roe

MANNED SYSTEMS SCIENCES
Northridge, Calif. 91324

March 1975

I^ATioNAL TECKNICAL
INFORMATION SERVICE

U S Depstmenl of Commorce
Springfi.W, VA. 22151

Prepared for

NATIONAL AERONAUTICS AND SPACE ADMINISTRATION
Flight Research Center
Edwards, Calif. 93523

1. Report No.

NASA CR-143835

2. Government Accession No.

3. Recipient's Cataiog No.

4. Title and Subtitle

A STUDY OF CARBURETOR/ INDUCTION SYSTEM ICING IN
GENERAL AVIATION ACCIDENTS

5. Report Date
March 1975

6. Performing Organization Code

7. Author(s(

Richard W , Obermayer and William T .Roe

8. Pertofmins Organization Report No.

10. Wwk Unit No.

9. Performing Organization Name and Address

Manned Systems Sciences
8949 Reseda Boulevard
Northridge, California 91324

IT, Contract or Grant No.
NAS4-2191

12. Sponsoring Agency Name and Address

National Aeronautics and Space, Administration
Washington, D. C. 20546

13. Type of Report and Period Covered
Contractor Report - Final

14. Sponsoring Agency Code

IS. Suppiementary Notes

16. Abstract

The purpose of this study was to perform a current assessment of the frequency
and severity of carburetor/induction icing in general -aviation accidents. The
available literature and accident data from the National Transportation Safety Board
were collected. A computer analysis of the accident data was performed. Between
65 and 90 accidents each year involve carburetor /induction syBtem icing as a prob-
able cause/factor. Under conditions conducive to carburetor /induction icing,
between 50 and 70 percent of engine malfunction /failure accidents (exclusive of
those due to fuel exhaustion) are due to carburetor/induction system icing'. Since
the evidence of such icing may not remain long after an accident, it is probable
that the frequency of occurrence of such accidents is underestimated; therefore,
some extrapolation of the data was conducted in this study . The problem of
carburetor /induction system icing is particularly acute for pilots with less than
1000 hours of total flying time. The severity of such accidents is about the same
as any accident resulting from a forced landing or precautionary landing . About
144 persons, on the average, are exposed to death and injury each year in acci-
dents involving carburetor /induction icing as a probable cause/factor.

PRICES souEo 10 mm^

17. Key Words (Suggested by Author(s))

Carburetor ice
Induction system icing
Icing accidents

19. Security Qassif. (of this reporti
Unclassified

18. Distribution Statement

Unclassified - Unlimited

20. Security Classif, lof this page)
Unclassified

21. -

Category: Q3

22. Price*

*For sale by the National Technical Information Service, Springfield, Vir^ginia 22151

1

TABLE OF CONTENTS

Page

SUMMARY 1

INTRODUCTION 2

SURVEY ANALYSES 3

Search for Accident Data 3

Literature Search , , ...... 4

Trends in the Future Use of Carburetor-type

Aircraft 8

THE COMPUTER DATA BASE 9

DATA ANALYSIS 20

Examination of Aircraft and Engine Makes Under
Conditions Conducive to Carburetor/Induction

System Icing ....,.....,..,.,.... 20

Probable Cause/Factor by Year 20

Composite Carburetor/Induction Ice Cause/Factor , . 22

Pilot Experience as a Factor in Carburetor/

Induction System Icing Accidents ......... 24

Composite Indicator to C/I Ice Conditions 25

Analysis of Accidents Occurring Under

Conditions Conducive to C/I Ice ..-....,. 29

Severity of Carburetor/Induction System

Icing Accidents 29

DISCUSSION 32

The Difficulty of Diagnosing Carburetor/

Induction System Icing .............. 32

A Context for Data Interpretation ......... 32

Severity of Carburetor/Induction System Icing ... 33

CONCLUSIONS . 34

REFERENCES ................. 35

SELECTED BIBLIOGRAPHY 35

APPENDIX A: INFORMATION SOURCES 39

HI

Preceding page blank

LIST OF FIGURES

Figure No. Page

1 Aircraft Analysis Sheet . 10

2 DOT Icing Probability Curves 26

LIST OF TABLES
Table No. Page

I Accident File Contents . - , . , 13

II Accident File Codes 15

III Comparison of Aircraft Make Sensitivity
to Carburetor/Induction System Icing
Conditions 21

IV Comparison of Engine Make Sensitivity
to Carburetor/Induction System
Icing Conditions , 21

V Accidents with Icing as Probable

Cause/Factor by Year .,.,.. 23

VI Composite C/I Ice Cause by Year 23

VII Composite C/I Ice Cause by Investigator's
Judgment that Conditions are Conducive
to C/I Ice 24

VIII Effect of Pilot Experience on C/I Ice as

Accident Cause/Factor . . , 25

IX Number and Percentage of Accidents that
Occurred Under Conditions Appropriate
for C/I Ice by Year . 27

X Number and Percentage of Accidents that
Occurred Under Conditions Appropriate
for C/I Icing by Cause Determination ... 28

XI Accident Cause/Factor by Composite

Indicator of C/I Conditions 30

XII Aircraft Damage Due to C/I Icing 30

XIII Number of Accidents Involving Injury

Due to C/I Icing 31

XIV Number of Occupants Involved in C/I Icing

Accidents 31

IV

A STUDY OF CARBURETOR/INDUCTION SYSTEM
ICING IN GENERAL AVIATION ACCIDENTS

By Richard W. Obermayer and William T. Roe

Manned Systems Sciences, Inc.
Northridge, California

SUMMARY

The purpose of this study, was to perforin, a current
assessment of the frequency and severity of carburetor/induction
icing in^general-aviation accidents. The available literature
and accident data from the National Transportation Safety Board
were collected. A computer analysis of the accident data was
performed. Between 6 5 and 90 accidents each year involve
carburetor/induction system icing as a probable cause/factor.
Under conditions conducive to carburetor/induction icing,
between 50 and 70 percent of engine malfunction/failure
accidents (exclusive of those due to fuel exhaustion) are due to
carburetor/induction system icing. Since the evidence of such
icmg may not remain long after an accident, it is probable
that the frequency of occurrence of such accidents is under-
estimated; therefore, some extrapolation of the data was
conducted in this study. The problem of carburetor/induction
system icmg is particularly acute for pilots with less than 1000
hours of total flying time. The severity of such accidents is
about the same as any accident resulting from a forced landing or
precautionary landing- About 14 4 persons, on the average, are
exposed to death and injury each year in accidents involving
carburetor/induction icing as a probable cause/factor.

INTRODUCTION

The conditions in the carburetor and induction system of
aircraft engines are such that ice can be formed, even on sunny,
warm days. There are ordinarily some indications to the pilot
of the onset of such icing such as a decrease in RPM or manifold
pressure. If the pilot does not take appropriate timely action
engine malfunction/failure may occur, and at times airstart of
the engine may be difficul-t or impossible- The pilot must be
aware of the possibilities of carburetor/induction icing and
routinely make proper use of carburetor heat controls. The
pilot must also be aware that use of carburetor heat controls at
the wrong times can also lead to difficulty.

It is generally known that carburetor/induction system
icing causes a number of accidents each year. The purpose of
this study is to make a current assessment of the frequency and
severity of the problem, to provide a basis for determining if
corrective research should be conducted.

A broad definition of "carburetor ice" will be used in this
report. Generally, the term carburetor/induction system ice
will be used in this report to indicate icing throughout the
fuel, carburetor, or induction systems while in operation.
Carburetor/induction system icing consists of three categories:

(1) Impact ice, which occurs when super-cooled moisture- laden
air strikes elements of the induction system. Build-up of such
ice occurs when operating in snow, sleet, rain, or clouds.

(2) Throttle icing, which occurs when moisture in the induction
air condenses and freezes due to the expansion cooling and
lower pressure as the air passes the restriction imposed by the
throttle. (3) Fuel-evaporation icing, which occurs when cooling
of the fuel mixture by fuel vaporization causes entrained
moisture to freeze.

The current study collected available literature and
accident data, and subsequently analyzed the accident data to
determine the frequency of occurrence and severity of accidents
for which carburetor/induction system icing was a probable cause
or factor. The study included the following specific tasks:
(1) establishment of available data sources, (2) search of the
available literature, (3) collection of data from specific
individuals and organizations, and organized data bases of
accident records, (4) formation of a computer data base of
relevant accident data, and (5) performance of an analysis of
these data and a statistical summary. The results of these
efforts, along with discussion and conclusions, are presented in
this report.

SURVEY ANALYSES

Three types of searches were conducted during the course of
this study: (1) a search for accident data bases relevant to
the study of carburetor/induction icing in general aviation,
(2) a search for published literature pertinent to aviation
accident statistics, accident investigation methods, carburetor/
induction system icing, engineering for icing protection, and
analyses of pilot error-related accidents, and (3) a follow-up
survey to determine the extent of the future use of carburetors
in aircraft design.

Search for Accident Data

Based on a survey of information available in the NASA
Flight Research Center Library, an initial list was made of
federal aviation agencies, safety institutions, flight schools,
recognized authorities, foreign agencies, and manufacturers of
airframes, engines and carburetors. These sources were queried
by mail or telephone: Did they have relevant information?
Were they aware of data existing elsewhere? Who else would they
suggest we contact? In this way, data were collected and
additional sources were discovered. . When no new sources were
uncovered, the process was stopped, and it was assumed that
complete coverage had been achieved. A list of the persons and
organizations contacted is presented in Appendix A.

The search for accident data extended throughout the United
States, Canada, and Mexico. While specific documents were
elicited from many sources, it became clear there is but one
organized general aviation accident data base which is maintained
by the National Transportation Safety Board. No organized data
base IS maintained in Canada and Mexico. In this country, both
the National Transportation Safety Board (NTSB) and the Federal
Aviation Agency (FAA) conduct accident investigations, but the
maintenance of a data base is the responsibility of NTSB.

As a result, the efforts of this study subsequently
concentrated on familiarization with the NTSB data base and with
acquisition of a subset of the data base suitable for this study.
Speific training information for accident investigators was
provided by the Transportation Safety Institute, Oklahoma City,
Oklahoma (ref. 1) and personal contact established with NTSB
personnel. A description of the data base' contents is provided
in references 2, 3 and 4.

Literature Search

The Southern California area is host to many fine libraries
which were available for the conduct of this study. In
particular, the University of California, Los Angeles; the
University of Southern California, California State University,
Northridge; and Los Angeles Pierce College have all been
extensively searched. The National Aeronautics and Space
Administration Flight Research Center Library provided many
aviation-specific documents these other sources could not. As
might be expected, the literature pertinent to the topic of this
study is not large. The References and Selected Bibliography of
this report presents the results of the search of these
libraries.

Many of these documents influenced the conduct of this
study: however, three are selected for specific coimnent in this

report;

Special Study; Carburetor Ice in General Aviation ( Ref. 5) .-

The authors cite that: ^'During the latest 5-year period for

which complete data are available, there was a total of 360
general aviation accidents involving carburetor ice as a probable
cause or factor. There were 40 fatalities and 160 persons
injured, 40 of them seriously, in these accidents. The number of
persons exposed to death or injury in these accidents was 6 36;
47 aircraft were destroyed and 313 others substantially damaged."

The blame for these accidents is laid squarely on the pilot,
stating that these can be attributed to the pilot in virtually
all cases. Further, carburetor icing is considered by NTSB as
one of the "unnecessary" causal factors in general aviation
accidents, since "carburetor icing can be avoided because the
means to preclude it are readily available." While it may be
true that the pilot has the means to prevent carburetor icing,
consider the following complex procedures which are recommended
m this document to avoid such icing:

"1. Periodically check carburetor heat systems and controls
for proper condition and operation.

"2. Start engine with carburetor heat control in the "cold"
position, to avoid possible damage to the carburetor heat system.

"3. As pref light item, check carburetor heat availability
by noting heat "on" power drop.

"4. When the relative hiimidity is above 50 percent and the
ambient temperature is below 299.8*^K (80°F), use carburetor
heat immediately before takeoff. In general, carburetor heat
should not be used during taxi because of possible foreign

matter entry when intake air is unfiltered in the "alternate" or
carburetor heat "on" position.

"5. Conduct takeoff without carburetor heat unless extreme
carburetor icing conditions are present, when carburetor heat may
be used if approved by aircraft manufacturer, and when conditions
are such that there will still be ample power for takeoff without
incurring engine overheat damage.

"6. Remain alert for takeoff for indications of carburetor
icmg, especially when the relative humidity is above 50 percent,
or when visible moisture is present.

"7. With supplemental instrumentation, such as a carburetor
air temperature gage, partial carburetor heat should be used as
necessary to maintain gage temperatures to forestall icing.
Without such instrumentation, use full heat but only intermit-
tently if considered necessary.

"8. If carburetor ice is suspected of causing a power loss,
immediately apply full heat. Do not disturb throttle initially,
since throttle movement may kill engine if heavy icing is
present- Watch for further power loss to indicate effect of
carburetor heat, then rise in power as ice melts.

"9. In case carburetor ice persists after a period of full
heat, gradually move throttle to full open position and climb
aircraft at maximum rate available in order to obtain greatest
amount of carburetor heat. if equipped with mixture control,
adjust for leanest practicable mixture, (approach this remedy
with caution - although carburetor ice generally serves to
enrich mixture, the reverse can be true; if the engine is lost
through excessive leaning, an airs tart might be impossible with
an iced induction system) .

"10. Avoid clouds as much as possible.

"11. In severely iced conditions, and when equipped with
mixture control, backfiring the engine can sometimes be effective
in dislodging induction system ice. With carburetor heat
control "off", lean engine while at full throttle (observe
caution note in No. 9, above).

"12. Consider that carburetor icing can occur with ambient
temperature as high as 310. 9°K (lOO^F) and humidity as low as
50 percent. Remain especially alert to carburetor icing
possibilities with a combination of ambient temperature below
294.3 K (70 F) and relative humidity above 80 percent.
However, the possibility of carburetor ice decreases in the
range below 273. 1°K (32^F) , This is because of (a) lessened
humidity as the temperature decreases, and (b) at around 263- 7°K
(15 f) any entrained moisture becomes ice crystals which pass

through the induction system harmlessly. It should be remembered
that if the intake air does contain these ice crystals,
carburetor heat might actually cause carburetor icing by melting
the crystals and raising the moisture-laden air to the carburetor
icing temperature range.

"13. Prior to closed- throttle operation, such as for a
descent, apply full heat and leave on throughout throttled
sequence. Periodically, open throttle during extended closed
throttle operation so that enough engine heat will be produced
to prevent icing. Be prepared to remove carburetor heat if
go-around is initiated.

"14. Return control to "cold" position immediately after
landing. If carburetor heat should be further required, observe
ground operation precaution in (4), above."

The above routing use of carburetor heat, and the awareness
of possible icing conditions in the induction system, are
probably not appreciated by any but the more experienced pilots.
Consequently, this special study by NTSB was produced to improve
pilot awareness, attention, and/or carefulness to reduce the
number of accidents of this type,

Aiycraft Carburetor Icing Studies (Ref. 6) .- The National
Research Council of Canada conducted an investigation into the
prevention of carburetor icing in aircraft by the use of
gasoline-soluble inhibitors. In addition to, or instead of,
using gasoline additives, the possibility of treating interior
components of the carburetor on which ice is formed was
investigated.

An engine test procedure was developed that (1) produced
sufficient carburetor ice to affect engine performance, and
(2) allowed engine performance to be monitored. It was possible
to select temperature, humidity, and throttle plate settings to
give the most severe icing conditions.

The importance of this study is that it established a basis
for hoping that the best possible solution to the problem of
carburetor icing can be achieved — that of eliminating the
problem. In this study the use of a teflon-coated throttle
plate was studied and found to virtually eliminate any ice
formation on the plate. The use of ethylene glycol monome thy 1
ether at 0.10 - 0.15 percent by volume in the gasoline and the
teflon-coated plate was shown to prevent both carburetor and
fuel system icing. Coating the shaft at the point where it
enters the barrel wall may be required to eliminate all ice
deposits.

)

Icing-Protection Requirements for Reciprocating-Engine
Induction Systems (Ref. 7) .- This study provides an extensive
and comprehensive discussion of the various types of icing
associated with the generic term of "carburetoi: icing", and,
from these investigations, establishes criteria for safe
operation and for design of new induction systems.

The opinion is reinforced that "the symptoms associated
with induction-system icing are not always discernible and
recognition of icing conditions generally requires considerable
operational experience." And, "The icing problem remains a
serious hazard, despite the fact, that the induction systems on
most licensed aircraft are capable of providing sufficient
protection in all weather conditions if th^ protection is
correctly applied and at the proper time" .,'' Further , a study by
Weick (ref . 8) of light-airplane power plant failures in 1947 is
cited, in which it is indicated that "34 percent of the forced
landings were caused by ineffective carburetor heat or by non-
application or improper application of carburetor heat" . This
analysis was based on one-seventh of the total 9 2-53 privately
operated airplane mishaps recorded by the Civil Aeronautics
Board for the year.

The methods and techniques for this research are presented,
including both airborne and ground-based test installations.
The temperature and the pressure of the air supplied to either
the airborne or ground-based induction system were carefully
controlled and measured at the carburetor inlet. Moisture
content of the air was regulated by steam jets to control
humidity and by water sprays to provide water in excess of
saturation. Instrumentation was provided to measure rates of
air flow, fuel flow, and simulated rain injection, as well as
carburetor and supercharger pressures, carburetor metering
pressures, and other variables affected by icing.

This study also reflects on a possible approach to the
carburetor icing problem: displaying information to the pilot
which will allow him to more readily detect and diagnose
carburetor/induction system icing. It is pointed out that
numerous instruments have been designed for this purpose,
including (1) interruption of a light beam focussed on a
photocell, (2) the blocking of special air passages to upset a
pressure-differential system, (3) a comparison of pressure drop
through 'an ice-free air passage and the induction system, and
(4) the change in capacitance between charged plates due to
icing.

The authors conclude that none of these ice detectors is .
entirely satisfactory because of the difficulty in locating the
sensing elements in positions in which they will register any
type of icing. Further, many detectors have been erratic and

indicate falsely. The authors feel that "a suitable ice detector
must be an extremely reliable instrument to retain pilot con-
fidence." In general, the use of ice-warning devices had been
unsatisfactory and none had been accepted for general use at
that time (circa 1950).

Trends in the Future Use of Carburetor- type Aircraft

While fuel-injected engines retain the danger of induction
system icing, the danger of. icing in a carburetor is removed.
High-performance engines using fuel injection are available in
general aviation. If fuel-injected engines were to eventually
exclude carburetor designs from future aircraft, then the
nature of the carburetor/induction system icing problem could
change. To test this hypothesis, and to collect data on future
trends in aircraft manufacture, a questionnaire was sent to the
major airframe manufacturers to determine their future plans.
Unfortunately, the attempt was fruitless, and no meaningful data
were collected in regard to this question. Informal communica-
tion with the four major engine manufacturers (Continental,
Franklin, Avco-Lycoming, and Pratt & Whitney) indicate a desire
to increase production of fuel-injection- type engines, but the
situation is not expected to change in the near future.
Examination of existing aircraft designs and lists of available
engine types for future design indicates that the carburetor -type
aircraft will predominate in the near future.

THE COMPUTER DATA BASE

The computer data base was formed using data provided on
ffiagnetic tapes by the Safety Analysis Division, Bureau of
Aviation Safety, National Transportation Safety Board, Department
ot Transportation. Data were requested on all accidents involv-
ing engine malfunction/failure in general aviation from 1968
JvIf^Tnil III '^^^ magnetic tapes provided were compatible with

zne IBM 360 computer system on which they were produced, and
therefore required conversion to run on the CDC Cyber 70 computer
system at the NASA Flight Research Center.

4-K 7^^ information potentially contained for each accident in
the data base is shown in Figure 1 in the form of the analysis
Sheet which is completed by the accident investigator and
subsequently key-punched for computer entry. While only 55 of
over 350 data items are mandatory for each accident,, the total
amount of data is large and much is unnecessary for this study,
consequently, to increase the efficiency of analysis, the data
base was reduced in content and number of accidents.

4-u ^ff^t, the content of each accident record was reduced to
those Items of information relevant to the current study. The
41 Items selected are presented in Table I. The codes used for
those items which are not entered directly as a numerical
quantity are given in Table 11.

Secondly, only accident records for the years 1969 through
19 73 were retained (five years). Accident records for the few
accidents which occurred outside the continental United States
were also discarded, as most of these contained very little
information. Also, only accident records for fixed-wing
aircraft with reciprocating engines were retained. It was
believed that the problems of rotary-wing aircraft were
inconsistent with the problems of fixed-wing aircraft. Only
reciprocating- engine- type aircraft accidents were retained, since
the term carburetor ice" does not apply to turbine engines.

Lastly, accident records which involved fuel exhaustion were
removed from the data base since (1) a large number of engine-
malfunction/f allure accidents are due simply to running out of
gasoline (approximately 170 accidents per year) , and (2) such
accidents are irrelevant to a study of carburetor icing. The
final data base contained 3555 records of the original 5930 '
records.

S ummary - The final data base of 3555 records is character-
ized by the following conditions:

^ARD I^O. 00 IDEHIIFriNG MfOnMATION

^ OtPARlUBE ^0(NT 1 1>^T l-LAHNED ENfiOUTE lANDSNG POINT

INTlNDtD MSIIHtnOH

li

M

]? IE

II

—I

iL
— 1

2

a

M

li

161

21

_l

^

B

30 SI

31

.J

It

MJijl

.ica

11

ii

JH!

^

A4«

■« atUb

M Sri

51

5! 5S

M

55ur«iuii,iL,

11 J!mE

'■i

!*|J,'«>i5,

%l"7li»s!*3f».l5slS

r-

79

1

1

X Crlt-

-

-

CA

ID

NO.

«4

AC

CII

lEf

n

SI1

re/

Pit

OT

D

AM

CAM WO. *7 MIDICAl fACTDRS

isU

17 SI

22 -a

I*

n^

11

rn

^

T'

'1'

CARD

NO. Ha INJUIII

5

CMD NO

t3

INiURKS

)4 1<i

OTKER GOVtBWt^HT

r^

F

rt-'
]9

r

M

M
* 11 |»

N

z

r 3

_,

^

-^

±_ X

Figure 1. (page 1 of 3)

10

CAtD NO. 17 BMim^PSOreUER FAILUH DATA

Figure 1. (page 2 of 3)

11

SPEClAt DATA

fAKD NO. ZJ AEIIAL tifWlKAltOt*

/i/

~r~T

i/i/imi/i/ ■& / i / s

II E li h - ?} y V » i' Jg >i i u I *1 M <5 I « ; j; } '« j ^2

mp MO. 17 eOiHimw MTWUM AIIIC««rT(THI 5 t.\KXt.n\

iizt

Figure 1- (page 3 of 3)

Aircraft Analysis Sheet (from Manual of Code Classifications,
Aircraft Accidents and Incidents, National Transportation
Safety Board, Department of Transportation, Washington, D.C.,
June 1970, Third Edition).

12

u>

TABLE I

■s

ACCIDENT FILE CONTENTS

NO,

LABEL

CHAR

FORMAT

NAME

001

FILENO,

5

F5.0

FILE NUMBER

002

MONTH

2

F2.0

MONTH OF ACCIDENT

^

003

YEAR

2

F2.0

YEAR " "

004

HOUR

2

F2.0

HOUR

005

LOC

2

F2.0

LOCATION OF ACCIDENT

^

006

ACMAKE

3

F3,0

AIRCRAFT MAKE

007

ACMOD.

2

F2.0

AIRCRAFT MODEL

008

ENGMAKE

2

F2.0

ENGINE MAKE

009

ENGMOD.

2

F2.0

ENGINE MODEL

010

NO.ENG.

1

Fl.O

NUMBER OP ENGINES

Oil

DAMAGE

1

Al

AIRCRAFT DAMAGE

012

KIND

l'

A1,1K

KIND OP PLYING

013

PHASEl

2

A1,1X

PHASE OF 1ST ACCIDENT {TYPE =

= ENG.MALF./FAIL.)

014

TYPE2

1

Al

TYPE OF SECOND ACCIDENT

015

PHASE2

2

A1,1X

PHASE OF SECOND ACCIDENT

016

ALT.

5

F4.0,1X

ALTITUDE OF OCCURRENCE

017

PLAWD

1

Al

FORCED LANDING

018/019

CIRCUM1/C1RCUM2

2

2A1 .

EMERGENCY CIRCUMSTANCES

020

TERRAIN

1

Al

TERRAIN

021

TOTHRS

5

F4.0,1X

TOTAL HOURS - PILOT

022

PLT.CNTRL

1

Al

PILOT AT CONTROL

023

INJURY

1

Al.

INJURY INDEX

TABLE I

(Page 2 of 2)

NO.

LABEL C

024

TOTABRD

025

PARTPWR

026

TOTPWR

027

SKYCOND

028/029

PRECIP1/PRECIP2

030

DE^^^TSPRD

031

TEMP.

032

INDAIR

033

UNDET

034

FUELHTR

035

CARBICE

036

FUELDEICE

037

CONDCI

038

AIEQUIP

039

ICEFUEL

040

ICECARB

041

ICE.IND

AR

FORMAT

3

F3.0

I

Al

1

Al.

1

Al

2

2A1

2

F2.0

3

F3.0

1

IX

1

Fl.O

1

Fl.O

1

Fl.O

1

Fl.O

1

Fl.O

1

Fl.O

NAME

1 Fl.O

1 Fl.O

1 Fl.O

1 Fl.O

TOTAL ABOARD
PARTIAL POWER LOSS
COMPLETE POWER LOSS
SKY CONDITIONS
PRECIPITATION (2 CODES)
DEW POINT SPREAD
TEMPERATURE
- NOT APPLICABLE -
PROBABLE CAUSE/FACTOR:

INDUCTION AIR/PREHEAT
CONTROLS, MISUSE

UNDETERMINED CAUSE

FUEL HEATER

CARB. DE-ICING SYSTEM

FUEL DE-ICING SYSTEM

CONDITIONS CONDUCIVE
TO CARB. ICE

IMPROPER USE: ANTI-
ICE, DE-ICE

ICE IN FUEL

ICE IN CARBURETOR

ICE IN INDUCTION
SYSTEM

TABLE II
ACCIDENT FILE CODES

AIRCRAFT DAMAGE

D - DESTROYED

S - SUBSTANTIAL

M - MINOR

N - NONE

a - UNKNOWN/NOT REPORTED

KIND OF FLYING

A - INSTRUCTIONAL
B - NONCOMMERCIAL
C - COMMERCIAL
D - MISCELLANEOUS

TYPE OF ACCIDENT

A GRND/WATER LOOP/SWERVE

B DRAGGED WING, POD, FLOAT

C WHEELS -UP

D WHEELS DWN LNDG - WATER

E GEAR COLLAPSED

F GEAR RETRACTED

G HARD LANDING

H NOSE OVER/DOPWf

I ROLL OVER

J OVERSHOT

K UNDERSHOOT

L COLLISION WITH A/C

M COLLISION - GRND/WATER

PHASE

N

P

Q
R
S
T

COLLISION -

BIRD STRIKE
STALL

FIRE OR EXPLOS
AIRFRAME - FAIL
ENGINE - TEARAWAY
U ENG MALF/FAIL

V PROP/ROTOR FAIL

W PROP/ROTOR ACC - PERS
X JET INTAKE EXH - PERS

Y PRpP/JET/ROTOR BLAST
Z TURBULENCE

A - STATIC

B - TAXI-

C - TAKEOFF

D - INFLIGHT

E - LANDING

B - UNKNOWN/NOT REPORTED

15

TABLE II

(Page 2 of 4)

TERRAIN

A - MOUNTAINOUS

B - HILLY

C - ROLLING

D - LEVEL, FLAT

E - FROZEN

F - ROCKY

G - SANDY

H

-

DENSE WITH TREES

I

-

CITY AREA

J

-

PLOWED

K

-

WATER

Y

-

OTHER

8

—

UNKNOWN/NOT

REPORTED

PILOT AT CONTROLS

A - PILOT IN COMMAND
B - COPILOT
C - DUAL STUDENT
D - CHECK PILOT

E - NONE

F - BOTH PILOTS

Y - OTHER

g - UNKNOWN/NOT REPORTED

INJURY INDEX

F - FATAL

G - SERIOUS

M - MINOR

N - NONE

3 - UNKNOWN/NOT REPORTED

PARTIAL/COMPLETE POWER LOSS

A - ONE ENGINE

B - TWO ENGINES

C - THREE ENGINES

D - FOUR ENGINES

a - UNKNOWN/NOT REPORTED

16

TABLE II - . (Page 3 of 4)

FORCED LANDING - PRECAUTIONARY LANDING
CODE

A FORCED LANDING ON AIRPORT/SEAPLANE BASE HELIPORT

B FORCED LANDING OFF AIRPORT ON LAND

C FORCED LANDING OFF AIRPORT ON WATER (DITCHING)

D PRECAUTIONARY LANDING ON AIRPORT

E PRECAUTIONARY LANDING OFF AIRPORT

a UNKNOWN/NOT REPORTED

EMERGENCY CIRCUMSTANCES

CODE

A LOW ON FUEL

B SMOKE IN COCKPIT

X PASSENGER DISTURBANCE

D FALSE FIRE WARNING

E LATERAL CONTROL PROBLEM

F PITCH CONTROL PROBLEM

G DIRECTIONAL CONTROL PROBLEM

H ADVERSE/UNFAVORABLE WEATHER

I APPROACHING DARKNESS

J SUSPECTED OR KNOWN AIRCRAFT DAMAGE

K SUSPECTED MECHANICAL DISCREPANCY

L DOOR/PANEL OPEN

M AIRFRAME BUFFET

N UNUSUAL NOISE

P PHYSICAL CONDITION OF PASSENGER

Q FUMES IN CABIN

' R PROP/ENGINE VIBRATION '

% UNKNOWN/NOT REPORTED

17

TABLE II

(Page 4 of 4)

SKY CONDITION (Local weather at time of 'accident)
CODE

A

CLEAR

B

SCATTERED (Above 1,000 feet)

C

SCATTERED (Below 1,000 feet)

D

BROKEN

E

BROKEN/LOWER SCATTERED

F

OVERCAST

G

OVERCAST/LOWER SCATTERED

H

PARTIAL OBSCURATION

I

OBSCURATION

%

UNKNOWN/NOT REPORTED

PRECIPITATION OF ACCIDENT SITE

CODE

A

HAIL (A) H

SNOW SHOWERS (SW)

B

SLEET (E) I

THUNDERSTORM (T)

C

DRIZZLE (L) J

FREEZING DRIZZLE (ZL)

D

RAIN (R) K

FREEZING RAIN (ZR)

E

RAIN SHOWERS (RW) L

NONE

F

SNOW (S) %

UNKNOWN/NOT REPORTED

G

SNOW GRAINS (SG)/SNOW

PELLETS (SP)

18

(1) Included accidents for the years 1969, 1970, 1971,
1972, and 1973.

(2) Included only engine malfunction/failure accidents.

(3) Included only fixed-wing, reciprocating engine accidents
which occurred within the continental U.S.

(4) Excluded accidents due to fuel exhaustion.

19

DATA ANALYSIS

The analyses performed on the computer data base consisted
primarily of crosstabulations of variables selected from the 41
variables included in the data base, and subsequent statistical
analysis of the resulting comparisons. The primary tool used for
this purpose was the Statistical Package for the Social Sciences
(SPSS) (ref. 9). SPSS is an integrated package of programs for
statistical analysis, data base management and file manipulation.
Using these programs, selected variables may be transformed or
receded, new variables created through computations based on
existing variables, and statistical procedures executed. This
package of programs provides efficient analysis of a given data
base with minimal programming effort. In addition, special-
purpose statistical programs were used, and hand calculations
performed in selected cases. The principal results of all
analyses are summarized in the following sections.

Examination of Aircraft and Engine Makes Under Conditions
Conducive to Carburetor/Induction System Icing

As part of an initial examination of the data base to
determine data base variables important to the study of
carburetor/induction system icing, aircraft makes and engine
makes were compared under conditions conducive and not conducive
to carburetor/induction system icing. One of the cause/factor
variables included in the data base was an indication of whether
the accident investigator considered the conditions conducive to
carburetor icing. For the present it is assumed that the
conditions are not conducive to icing if the accident investiga-
tor did not indicate such conditions existed; the validity of
this assumption will be discussed later.

The results comparing aircraft makes are presented in
Table III, while the results for engine makes are given in
Table IV. As no practical or statistical significance exists
between the major manufacturers of either airframes or engines,
no distinction between makes was carried through the subsequent
analyses. These results contradict informal reports that
specific manufacturers produced designs more or less prone to
carburetor icing difficulties.

Probable Cause/Factor by Year

Selected causes/factors, indicated by the accident
investigator, are tabulated in Table V by year. The causes/
factors selected are: (1) induction air, preheat cockpit
controls, (2) power plant failure for undetermined reasons,
(3) anti- icing, de-icing, and carburetor de-icing systems.

20

TABLE III
COMPARISON OF AIRCRAFT MAKE SENSITIVITY TO
CARBURETOR/INDUCTION SYSTKM ICING CONDITIONS

NO. OF ACCIDENTS THAT OCCURRED UNDER -

PERCENT OF TOTAL
ACCIDENTS THAT
OCCURRED UNDER
CONDITIONS CON-
DUCIVE TO ICING

MRCRAPT
MAKE

CONDITIONS
CONDUCIVE
TO ICING*

OTHER
CONDITIONS

A

B
C
D

E
F
G-
H
Other

13

17

126

7

11

97

5

8

32

88

298

989

43

84

935

39

73

690

12.9

5.4

11.3

14.0

11.6

9.4

11.4

9.9

4.4

*As defined by the accident investigator.

TABLE IV

COMPARISON OF ENGINE MAKE SENSITIVITY TO

CARBURETOR/ INDUCTION SYSTEM ICING CONDITIONS

NO.

OF ACCIDENTS THAT OCCURRED UNDER -

PERCENT OP TOTAL
ACCIDENTS THAT

CONDITIONS

OCCURRED UNDER

MRCRAFT

CONDUCIVE '

OTHER

CONDITIONS CON-

MAKE

TO ICING*

CONDITIONS

DUCIVE TO ICING

A

6

86

6.5

B

144

1376

9.5

C

153

1437

9.6

D

7

246

2.8

E

3

37

7.5

Other

3

57

5.0

Total

316

3239

8.9

*As def

ined

by the accident

investiqator.

21

(4) weather conditions conducive to carburetor/induction system
icing, (5) improper use, or failure to use, anti-icing de-icing
equipment, (6) ice in fuel, (7) ice in carburetor, and (8) ice in
the induction system. More than one (usually five or more)
cause/factor may be indicated by the investigator for a given
accident.

A number of results accrue from examination of Table V:

(1) There is a statistically significant trend for
accidents within the selected data base to decrease over the five
year period examined (P<.05).

(2) Relatively few accidents involve identifiable failures
of the controls and systems associated with carburetor/induction
system icing.

(3) There are a large number of powerplant failures for
undetermined reasons .

(4) There are approximately 65 accidents per year, on the
average, for which evidence of ice is found in the fuel or
induction system.

(5) The number of cases for which ice in the carburetor is
cited as a probable cause/factor often exceeds the number of
cases for which it is cited that conditions are conducive to
carburetor/induction system icing. Apparently the investigators
may omit indicating that the conditions are conducive to
carburetor icing when other, more specific, evidence exists.

Composite Carburetor/Induction Ice Cause/Factor

It was noted in Table V that a number of probable causes/
factors may be indicated by the accident investigator either
singly or in combination. A composite cause/factor was formed to
avoid this confusion. The composite indicator was defined as a
probable Carburetor/Induction Ice Cause/Factor _if any of the
following cause/factors were noted, by the accident investigator:
(1) induction air/preheat controls, or (2) the carburetor
de-icing system, or (3) improper use of anti-ice/de-ice
equipment, or (4)~Tce in the carburetor, or (5) ice in the
induction system.

The composite carburetor/induction ice cause/factor (C/I Ice
Cause/Factor) is tabulated in Table VI over the five year period
included in this study. While the number of carburetor/induction
system ice accidents may appear to decrease over the five year
period (as did the total number of accidents in the data base)
the number of accidents in years 1969 and 1970 are not statis-^
tically significant from the number in years 1972 and 1973
(P<.33). Based on the composite index, the number of such
accidents averages 7 4 per year.

22

TABLE V
ACCIDENTS WITH ICING AS PROBABLE CAUSE/FACTOR BY YEAR

PROBABLE CAUSE OF ACCIDENT

Misuse of induction air,
preheat cockpit controls

Powerplant failure -
reason undetermined

Carburetor de-icing system
malfunction

Conditions conducive to
carburetor/induction
system icing

Improper operation -
failed to use anti-
icing/de-icing equipment

Ice in fuel

Ice in carburetor

Ice in induction system

Total

YEAR

1969 1970 1971 1972 1973

Number of Accidents

5'
206
1

79

85
4

82
2

228

7

47

58

3

59

1.

1
203

1

70

64

7

70

3

1

158

2

65

69

8

63

1

1

173
3

55

50
4

52
4

Total

760

730

730

661 674

8

968

14

316

326
26

326
11

3555

TABLE VI
COMPOSITE C/I ICE CAUSE BY YEAR

YEAR

1969

1970

1971

1972

1973

Number

of Accidents

Total

WAS C/I ICE

CAUSE OF
ACCIDENT?*

No
Yes

666

94

659
71

658
72

587
.74

615
59

3185

370

Total

760

730

730 .

661

674

3555

♦Composite C/I Cause: Misuse of induction air/preheat controls,
carburetor de-icing system malfunction, improper use of '
de-icing system, ice in carburetor, or ice in induction
system.

23

Also, as previously noted in Table V, Table VII indicates
that a large nvunber of carburetor/induction system icing acci-
dents are not noted by many investigators as occurring when the
conditions are conducive to icing; consequently/ a better
indicator of the conditions appropriate to such accidents must be
constructed.

Pilot Experience as a Factor in Carburetor/Induction

System Icing Accidents

Pilot experience, in terms of total number of flying hours,
is cross-tabulated with the composite C/I Ice Cause/Factor in
Table VIII. The percent of carburetor/induction ice accidents
decreases with the total number of pilot hours; however, the
decrease is statistically significant (P<.05) only for pilot
hours greater than 1,000 hours. Stated in other terms, only
pilots with over 1,000 hours experience seem to possess the skill
and knowledge to perform significantly better in avoiding
carburetor/induction system icing than pilots with less than 1,000
hours total flying time. About 68% of the carburetor/induction
icing accidents involve pilots with less than 1,000 total hours.

-TABLE VII
COMPOSITE C/I ICE CAUSE BY INVESTIGATOR'S JUDGMENT
THAT CONDITIONS ARE CONDUCIVE TO C/I ICE

Were Conditions Conducive to C/I
Ice? (Investigator's Judgment)

No

Yes

Number of Accidents

Total

WAS C/I ICE No
CAUSE OF
ACCIDENT?* Yes

3173
66

12

304

3185
370

Total

3239

316

3555

♦Composite C/I Ice Cause; Misuse of induction air/preheat
controls, carburetor de-icing system malfunction, improper
use of de-icing system, ice in carburetor, or ice in induction
system.

24

TABLE VIII
EFFECT OF PILOT EXPERIENCE ON C/I ICE AS
ACCIDENT CAUSE/FACTOR

Pilot
Experience

Hours

Total
Accidents

for
Experience
Range

Cause of

Accident

C/I Ice

Other

Niomber of
Accidents

Percent of

Total
Accidents

Number of
Accidents

Percent oi

Total
Accidents

10 to 100

477

70

14.7

407

^ 85.3

100 to 500

1000

125

12.5

875

87.5

500 to lOOC

519

55

10.6

464

89.4

>1000

. 1559

120

7.7

1439

92.3

>10

3555

370

10.4

1559

89.6

Composite Indicator to C/I Ice Conditions

As previously noted, there is reason to believe that condi-
tions are appropriate for carburetor/induction system icing more
often than indicated by the accident investigators. A composite
indicator (Conditions Appropriate) was constructed based on the
following: (1) conditions were considered appropriate by the
accident investigator, or (2) there was an indication of
precipitation at the accident site, or (3) the temperature and
dew point were within the profile presented in Figure 2. The
temperature/dew point profile of Figure 2 is generated by the
Department of Transportation (taken from Gardner and Moon, 1971);
the solid line profile was that used in the computer implementa-
tion and is to be considered an approximate indicator of the
existence of conditions appropriate for carburetor icing.
Unfortunately, the existence of temperature/dew point data was
rare in the NTSB data base.

The number of accidents occurring when conditions are
appropriate for icing according to the composite indicator are
presented in Table IX. Approximately 18% of the accidents in the
reduced data base used in this study occurred when the conditions
were conducive to carburetor/induction system icing. Based on
these data, one may estimate the maximum number of, accidents due
to this cause at about 128 per year.

It is interesting to note in Table X that about 13% of the
accidents for which the cause is undetermined occur when the
conditions are appropriate for carburetor/induction icing.

25

ICIN6 PROBABILITY CUWV^S

i 1 1 L

Figure 2. DOT Icing Probability Curves. (The heavy solid
line is the profile implemented in the computer
analysis . }

26

TABLE IX
NUMBER AND PERCENTAGE OP ACCIDENTS THAT OCCURRED UNDER
CONDITIONS APPROPRIATE FOR C/I ICE* BY YEAR

(YGar(s)

Total

Accidents

for

Year(s)

Were

! Conditions Appropriate for
C/I Ice?

Yes

No

Number of
Accidents

Percent of

Total
Accidents

Number of
Accidents

Percent of

Total
Accidents

1969

760

126

16.6

634

83.4

1970

730

134

18,4

596

81,6

1971

730

121

16.6

609

83.4

1972

661

144

21.8

517

78.2

1973

674

113

16.8

561

83.2

1969 to

1973

3555

638

18.0

2917

82.0

*Conditions appropriate for C/I ice: Indicated by accident
investigator, temperature/dew point within profile, or
precipitation at accident site.

27

TABLE X
NUMBER AND PERCENTAGE OF ACCIDENTS THAT OCCURRED UNDER
CONDITIONS APPROPRIATE FOR C/I ICING* BY
CAUSE DETERMINATION

Number of

Accidents

Percent

Number of accidents for which cause was

determined

2587

100

Number of these accidents for which

conditions were appropriate for

C/X icing

511

14.4

Number of these accidents for which

conditions were not appropriate for

C/I icing

2076

85.6

Number of accidents for which cause

was not determined

968

100

Number of these accidents for which

condition was appropriate for

C/I icing

127

13.1

Number of these accidents for which

conditions were not appropriate for

C/I icing

841

86.9

*Conditions appropriate for C/I ice: Indicated by accident
investigator, temperature/dew point within profile, or
precipitation at accident site.

28

Analysis of Accidents Occurring Under Conditions
Conducive to C/I Ice

Table XI presents an analysis of accidents occurring during
conditions conducive to carburetor/induction icing. At least
319 out of a total of 638 accidents are attributable to
carburetor/induction icing, or about 50%, Another 127 accidents
occurring under appropriate conditions are due to undetermined -
reasons. Thus a total of 446 accidents might be due to such
icmg causes, or an estimated maximum of 70%; it might be
expected that about 50% of these are due to carburetor/induction
'■^^"?:. Consequent ly, it may be estimated that, when the
conditions are app ropriate, between 50% and 70% arf^ dn^ tn
carburetor/ induction system icincr" ~

Severity of Carburetor/Induction System Icing Accidents

It should be clear that, while a number of accidents may
have^involved carburetor/induction system icing as a cause/factor,
the flights eventually terminated with a forced landing and
possible collision with a ground object. As is the case with
other engine/malfunction failure accidents, approximately 80%
resulted m substantial damage to the aircraft, and the
remainder resulted in destroyed aircraft (see Table XII)
However, as indicated in Table XIII, 230 of 370 accidents (62%)
resulted m no personal injury and only 14% resulted in serious or
±atal injury. The distribution of occupants involved in C/I icina
accidents is presented in Table XIV; the net result is that
between 720 (for C/I ice conditions) and 977 (including possible
accidents with undetermined reasons) persons have been involved
""I? such accidents over the five year period considered in this
study {between 144 and 195 per year) .

29

TABLE XI
ACCIDENT CAUSE/FACTOR BY COMPOSITE INDICATOR
OP C/I CONDITIONS

■

NUMBER OF ACCIDENTS

TOTAL

CONDITIONS APPROPRIATE

ACCIDENT CAUSE/FACTOR

FOR C/I ICE?

C/I ICE

NO C/I ICE

UNDETERMINED

Yes

319

192

127

638

No

51

2025

841

2917

Total

370

2217

968

3555

TABLE XII
AIRCRAFT DAMAGE DUE TO C/l ICING

NUMBER OF ACCIDENTS

TOTAI

AIRCRAFT DAMAGE

MINOR

SUBSTANTIAL

DESTROYED

NOT REPORTED

C/I ice

1

309

60

370

Dndetermined,

with C/I ice

conditions

80

47

127

Other

2425

632

1

3058

Total

1

2814

739

1

3555

k

30

TABLE XIII
NUMBER OF ACCIDENTS INVOLVING INJURY DUE TO

C/I ICING

NUMBER OF ACCIDENTS

TOTAL

Accident Cause

INJURY INDEX

None

Minor Serious

Fatal

UnJcn.

C/I Ice

230

S7 34

19

370

Undetermined,

with C/I Ice

Conditions

51

38 22

16

12 7

Other

1803

652 347

255

1

3058

Total

2084

777 403

290

1

3555

TABLE XIV
NUMBER OF OCCUPANTS INVOLVED IN C/I ICING ACCIDENTS

1 . 1

NUMBER OF ACCIDENTS

TOTAL

Accident Cause

NUMBER ABOARD

SIR* 1 2 3 4 5 6 >6

C/I Ice

Undetermined,
with C/I Ice
Conditions

Other

Total

143 154 38 27 4 2 2

51 51 13 7 2 3

1 1350 1030 289 281 59 23 25
1 1544 1235 340 315 65 25 30

370

127
3058
3555

Accident Cause

TOTAL PERSONS INVOLVED
IN ACCIDENTS

C/I Ice

720

Undetermined,
with C/I Ice
Conditions

257

Other

6105

Total

7082

*Not reported

31

DISCUSSION

The Difficulty of Diagnosing Carburetor/Induction System Icing

A fundamental problem affecting the study of the severity
of carburetor/induction system icing in general aviation acci-
dents is the inherent difficulty of diagnosing carburetor/
induction icing. If ice has caused an accident, the ice will
melt and the resulting water may evaporate; so, unless the
investigator is astute and has an initial strong suspicion of
carburetor icing, the cause/factor associated with the accident
may not be properly identified.

Since this type of accident may be so easily mis-diagnosed,
it is believed that the results of this study underestimate the
magnitude of the problem. A broad definition of "carburetor
icing" has been taken, and some reasonable extrapolations to
include the probable number of "undetermined" accidents which
may be carburetor icing have been made in this study. Neverthe-
less, it is believed that the estimates made in this study are
quite conservative and that a number of additional accidents
could have included carburetor/induction icing as a probable
cause/factor. Further, the findings of this study concentrate on
the accident data available; a large number of forced landings
occur each year which are not included in the accident data base.

A Context for Data Interpretation

Over a five-year period analyzed in this study, an average
of 65 carburetor ice accidents per year are identified by the
accident investigators, and one may extend this estimate to
approximately 90 probable carburetor/induction icing accidents
per year. That is, there are about that many general aviation
accidents per year which involve carburetor/induction icing;
however, a niimber of other probable cause/factors are generally
also involved.

Compared to the approximately five thousand total general
aviation accidents per year, and the approximately one thousand
engine malfunction/failure accidents per year, the less than one
hundred carburetor/induction accidents per year may seem small.
On the other hand, it must be recognized that carburetor/
induction icing can only occur when the weather conditions are
appropriate. Viewed in this context, the picture changes, for
when the weather conditions are conducive to icing, 50 percent
to 70 percent of all engine malfunction/failure type accidents
(which are not due to fuel exhaustion) include carburetor/
induction icing as a probable cause/factor. That is,
if the weather conditions are conducive, if the engine

32

malfunctions or fails with fuel still available to the engine,
then carburetor/induction icing is involved more than half of
the time .

Further, carburetor/induction system icing is a widespread
problem. Such icing can occur anywhere, even at temperatures
over BlO.g'^K (lOO^F), and is equally likely throughout the
year. Pilots with long experience become involved in such
accidents, although more than two-thirds of carburetor/induction
system icing accidents involve pilots with less than 1000 hours
of total flying time. Apparently, any pilot, wherever he/she
may be flying, at any time of the year, may encounter carburetor/
induction system icing.

Severity of Carburetor/Induction System Icing

Of course, as with many other accident causes/factors, the
injurious and damaging aspects of an accident for which
carburetor/induction icing is a probable cause/factor are
associated with the terminating forced or precautionary landing.
One hundred, and forty-four persons per year, on the average, are
involved in carburetor icing accidents, about 4 of these
accidents involve fatalities, about 7 per year involve serious
injury, and almost all of these accidents involve substantial
damage or destruction of the aircraft. Consequently, carburetor/
induction system icing accidents are just about as serious as
any other accident involving a forced landing or precautionary
landing. Other types of accidents occur more often: for
example, accidents involving fuel exhaustion occur about twice
as often as those involving carburetor/induction icing; but
when weather conditions are conducive, carburetor/induction
system icing is the most common cause for engine malfunction/
failure.

33

CONCLUSIONS

Carburetor/induction system icing is a serious
threat existing in general aviation operations. Given that
fuel is available, under the proper weather conditions, which
occur frequently and regularly throughout the year, carburetor/
induction icing is the most common cause of engine malfunction/
failure in general aviation. Overall, this cause amounts for
about 10 percent of engine malfunction/failures, making it a
major cause for concern. Carburetor/induction icing is a
particular threat to the pilot with less than 1000 hours of
total flying time.

While engineering analysis was not included in the current
study, the review of the literature indicates a strong
possibility that this problem may be solvable. The problem
may yield to (1) efforts to eliminate icing through
carburetor/induction system design and fuel additives, and/or
(2) efforts to either increase pilot awareness through improved
display of icing information or automation of the pilots
function for the control of icing countermeasures . The payoff
would be high: 65-90 accidents per year could be eliminated
along with the attendant death, injury and damage.

The analysis of accident data is severely limited for
investigation of a problem as insidious and difficult to detect
as carburetor/induction icing. Further investigation of the
problem is recommended using flight or ground-based facilities
permitting the control of induction icing. Under controlled
conditions, the ability of the pilot to detect, recognize and
take proper action can be accurately assessed. It is suspected
that given such data it will be found that the findings of this
study, which label carburetor/induction icing as a major problem,
have underestimated the magnitude and frequency of carburetor/
induction system icing occurrences.

34

REFERENCES

1. Transportation Safety Institute, Basic aircraft accident

investigation procedures and techniques . Oklahoma City,
Oklahoma: Department of Transportation, October 19 71.

2. National Transportation Safety Board, Manual of code

classifications: Aircraf-t accidents and incidents .
(3rd Ed.) Washington, D.C.; Department of Transportation,
June 1970.

3. National Transportation Safety Board, Manual of aircraft

and engine code classifications . Washington, D.C. :
Department of Transportation, June 1973.

4. Bureau of Aviation Safety, User instructions: ADP programs .

Automated aircraft accident and incident information
system . Washington, D.C: National Transportation Safety
Board, November 1971.

5. Bureau of Aviation Safety, Special Study: Carburetor ice in

general aviation . Washington, D.C: National Transporta-
tion Safety Board, NTSB-AAS-72-1, January 1972.

6. Gardner, L. and Moon, G. Aircraft carburetor icing studies .

Ottawa, Canada: National Research Council of Canada,
Fuels and Lubricants Division, 1971.

7. Coles, W.D-, Rollin, V.G. and Mulholland, D.R. Icing -

protection requirements for reciprocating-engine induction
systems . Washington, D.C: U.S. Government Printing
Office, 36th Report of the National Advisory Committee for
Aeronautics, Technical Report 9 82, 1950.

8. Weick, F.E. Powerplant failures--causes and cures. Aviation

Week , 7 March 1949, 50(10), 21-25.

9. Nie, N.H., Bent, D.H., and Hull, C.H. Statistical package

for the social sciences (SPSS) . New York: McGraw-Hill,
1970.

35

SELECTED BIBLIOGRAPHY

ACCIDENT INVESTIGATION METHODS

A method of determining the danger of airplane icing .

Charlottesville, Va.: Army Foreign Science and Technology
Center. FSTC-HT-23-1818-73, 16 July 1974 (AD 781 221/7GA")

Association for the Aid of Crippled Children, Behavioral
Approaches to Accident Research . New York, 1966.

Dresser, Robert F. New Approach to Air Safety Statistics .
Federal Aviation Agency, Washington, D.C.

Haddon, W. , Jr., Sauchman, E.A., and Klein, D. Accident Research :
Methods and Approaches . Harper & Row Publishers, New York,
1964.

Johnson, W.M. "Aircraft accident investigation in Canada",
Canadian Aeronautics and Space Journal , 15 , Pp 113-116.

AVIATION ACCIDENT INVESTIGATION

Bureau of Aviation Safety. A Preliminary Analysis of Aircraft
Accident Data; U.S. Civil Aviation , 1970. National
Transportation Safety Board, Washington, D.C, NTSB-APA-71-1,
1970.

Gough, Melvin N. Study of Current Information Pertinent to

Aviation Accidents (or Incidents) , Final Report - Contract
NAS4-866, October, 1966, PR-1909-16 - 1/251 PPC 61, NASA
Flight Research Center, Edwards, California.

Sparrow, S.W. Airplane Crashes: Engine Troubles - A Possible
Explanation . NACA TN55, 1921.

AVIATION SAFETY

Hoekstra, H.C., Huange, S.D. Safety in General Aviation . Flight
Safety Foundation, Inc., Arlington, VA, 1971.

CARBURETOR ICING REPORTS

Aircraft Ice Protection , FAA Advisory Circular No. 20-73, April
21, 1971.

Barnard, D.P. and Scott, E.H. Weather or Stall-Carburetor Icing
Control . The Standard Oil'To^^ (Ohio), SAE Annual Meeting,
January 1959.

36

Carburetor Icing , Ethyl Technical Note, Ethyl Corp.

Carburetor Icing Weather Data for the United States , En jay
Chemical Co.

Gardner, L. and Moon, G. Aircraft Carburetor Icing Studies .

National Research Council of Canada; Fuels and Lubricants
Division, Ottawa (Ontario), 19 71.

"Look out for carburetor ice". FAA Aviation News , August 1967.

Perkins, P.J., Lewis, W. , and Mulholland, D.R. Statistical Study
of Aircraft Icing Probabilities at the 700- an d 500-
Millibar Levels over Ocean Areas in the Northern Hemisphere .
National Advisory Committee for Aeronautics Technical Note
3984, Lewis Flight Propulsion Laboratory, Cleveland,. Ohio,
May 1957.

Richter, K. "Carburetor heat: how to use it". Carburetor

Icing - A Problem ? The Standard Oil Co. (Ohio), SAE Paper
448b, January 1962.

Use of Carburetor Heat Control . Lycoming Division Service
Instruction No. 1148A, December 8, 1967.

GENERAL AVIATION REFERENCES

Gillespie, A.C. (ed.) 19 74 Aviation Almanac , National Aviation
Consumer Association, New Jersey, 1973.

World Aviation Directory . Public Transportation and Travel

Division, Ziff-Davis Publishing Co., Washington, D.C., 1974.

DeMuth, T.P., Jackson, H.R. and Test, L.J. Carburetor Icing

Tests in the Laboratory and in Service . Atlantic Refining
Co., SAE Paper 448c, January 1962.

Investigation of Icing Characteristics of Typical Airplane

Engine Induction Systems . NACA TN No. 1970, February 1949.

Schuel, H.J. and Burt, J.G. "New tests on carburetor icing".
E.I, duPont Co., Petroleum Refiner , November i960.

The Aircraft Engine and its Operation . Installation Engineering-
Pratt & Whitney Aircraft, May 1947.

NTSB SPECIAL STUDIES

Special Study; Carburetor Ice in General Aviation . National

Transportation. Safety Board, NTSB-AAS-72-1, Washington, D.C.
1972.

37

Special Study; U.S. General Aviation Accidents Involving Fuel
Starvation; 1970-1972 . National Transportation Safety
Board, NTSB-AAS-74-1, Washington, D.C, , 1974.

PILOT ERROR

Aircraft Design-Induces Pilot Error .
Washington, D.C.

Civil Aeronautics Board,

Kowalsky, N,B. , Masters, R.L., Stone, R.B. , Babcock, G.L. , and
Rypka, E.W. An Analysis of Pilot Error-Related Aircraft
Accidents . Department of Aerospace and Environmental
Medicine, Lovelace Foundation, Albuquerque, New Mexico,
June 1964, NASA CR-2 444.

38

APPENDIX A; INFORMATION SOURCES

GENERAL SOURCES

AEROSPACE SAFETY RESEARCH AND DATA INSTITUTE {ASRDI )

AIRCRAFT OWNERS ASSOCIATION

AIR FORCE FLIGHT DYNAMICS LABORATORY (AFFDL)

AIRWORK SERVICE DIVISION

ALLEN AIRCRAFT (DIVISION OF AAR CORP.)

AMERICAN NATIONAL . STANDARDS INSTITUTE.

AMERICAN SOCIETY FOR NONDESTRUCTIVE TESTING

AMERICAN SOCIETY OF SAFETY ENGINEERS

ANDREWS & ASSOCIATES

ATLANTIC AVIATION CORPORATION

AUTOMOTIVE ENGINEERING (SAE JOURNAL)

AVCO (LYCOMING DIVISION) CORP.

AVQUIPO INC.

BORG-WARNER CORP. (MARVEL- SCHEBLER DIV, )

BUREAU OF AVIATION SAFETY (NTSB: WASH., D.C.)

CALIFORNIA AVIATION COUNCIL

CANADIAN CURTIS S-WRIGHT LTD.

CATERPILLAR TRACTOR CO. ^

CHANDLER EVANS CONTROL SYSTEMS (DIV. OF COLT INDUSTRIES)

CHEMICAL PROPULSION INFORMATION AGENCY

CHRYSLER CORP . , DEFENSE SPACE GROUP

CIVIL AEROMEDICAL INSTITUTE

COLT INDUSTRIES

COMBUSTION INSTITUTE

COOPER- BESSEMER

E.I. DU PONT CO.

ETHYL CORPORATION ,

PAA SAFETY DATA BRANCH FLIGHT STANDARDS TECHNICAL DIVISION

FEDERAL AVIATION ADMINISTRATION (LOS ANGELES)

FIELD AVIATION CO., LTD.

FLIGHT SAFETY FOUNDATION, INC.

39

FRANKLIN GNO CORP.

GUGGENHEIM CENTER FOR AEROSPACE HEALTH AND SAFETY
JET AVIATION CORP.

JOHNS HOPKINS APPLIED PHYSICS LAB
LARSON AERODEVELOPMENT
LEAVENS BROS . , LTD ,

MINISTRY OF (CANADIAN) TRANSPORT ACCIDENT INVESTIGATION
NATIONAL AIRCRAFT ACCIDENT INVESTIGATION SCHOOL (NAAIS)
NATIONAL INSTITUTE FOR OCCUPATIONAL SAFETY AND HEALTH
NATIONAL SAFETY COUNCIL

NATIONAL TRANSPORTATION SAFETY BOARD (NTSB)
LOS ANGELES, CALIFORNIA
WASHINGTON, D,C.

ELSON AIRCRAFT CORPORATION

NONDESTRUCTIVE TESTING INFORMATION CENTER

ONAN DIVISION, ONAN CORPORATION

REGIONAL INFORMATION AND COMMUNICATION EXCHANGE

SOUTHWEST RESEARCH INSTITUTE

STANDARD OIL COMPANY (OHIO)

THE OHIO STATE UNIVERSITY:

DEPARTMENT OF AERONAUTICAL AND ASTRONAUTICAL ENGINEERING
DEPARTMENT OF AVIATION

TRANSPORTATION SAFETY INSTITUTE

TRW EQUIPMENT, INC.

TELEDYNE CONTINENTAL MOTORS, AIRCRAFT PRODUCTS DIVISION

TRANSAERO INC.

UNITED AIRCRAFT OF CANADA, LTD.

WHITE SUPERIOR, DIVISION OF WHITE MOTOR CORPORATION

UNIVERSITY OF ILLINOIS:
STAN ROSCOE
JESSE STONECIPHER

40

ACTUARIAL SOURCES

ABBOTT, WHITE, AND CO., INC.

AERO COVERAGES, INC.

AERO INSURANCE AGENCY

AIR ASSOCIATES, INC.

AIRBANC OF AMERICA, INC.

AIRWAY CASUALTY CO.

ALEXANDER & ALEXANDER, INC.

ASSOCIATED AVIATION UNDERWRITERS

AVEMCO AIRCRAFT INVESTMENT CORPORATION

AVEMCO CORPORATION

AVIATION INSURANCE SERVICE

AVIATION OFFICE OF AMERICA, INC.

RICHARD J. BERLOW & CO. , INC.

J.H. BLADES & CO., INC,

BOWES & CO. , INC.

C&A FLIGHT CREW INSURANCE, INC.

CESSNA FINANCE CORP.

C.l.T. LEASING CORP.

COMMERCIAL CREDIT EQUIPMENT CORP.

CONNECTICUT GENERAL LIFE INSURANCE CO.

CONTINENTAL ASSOCIATES, INC.

DELTA AGENCIES, INC.

ELECTRA FINANCIAL GROUP

FAIRFAX UNDERWRITERS SERVICES, INC.

G. SHANNON GROVER

HARLAN INC, OF PENNSYLVANIA

HOLTON INSURANCE AGENCY, INC.

INA REINSURANCE CO.

INSURANCE CO. OF NORTH AMERICA

INTERNATIONAL AIRCRAFT SALES

INTERNATIONAL AVIATION UNDERWRITERS, INC,

KUDLICH GENERAL INSURANCE AGENCY, LTD.

MANN-KLINE, INC.

NATIONAL AVIATION UNDERWRITERS

41

PACIFIC INTERNATIONAL UNDERWRITERS
PARKER & CO. INTERNATIONAL, INC.
PETER J. MC BREEN & ASSOCIATES, INC.
THE SOCIETY OF ACTUARIES
TEX-WIDE INSURANCE
TRANSWORLD INSURANCE BROKERS
WRIGHT & CO.

AIRFRAME MANUFACTURERS

AERONCA, INC.

AERO RESOURCES

AERO SPORT

AMERICAN AVIATION CORPORATION

ATLANTIC AVIATION

AVCO-LYCOMING

BEECH AIRCRAFT CORPORATION

BELLANCA AIRCRAFT CORPORATION

CESSNA AIRCRAFT CO.

CONTINENTAL

ENSTROM CORPORATION

EVANGEL AIRCRAFT CORPORATION

FAIRCHILD INDUSTRIES, INC.

FRANKLIN

MAULE AIRCRAFT CORPORATION

NORTH AMERICAN ROCKWELL CORPORATION

PIPER AIRCRAFT CORPORATION

PITTS AVIATION ENTERPRISES

PRATT & WHITNEY

PERSONNEL CONTACTED

WILLIAM ALLEN: FAA ACCIDENT PREVENTION & ANALYSIS
VINCENT BROPHY: FAA

JAMES CONSTANTINE: NTSB ACCIDENT RECORDS SECTION
JOHN CRAWFORD: BUREAU OF AVIATION SAFETY

42

ROBERT CRESTON: FAA: TRANSPORTATION SAPETY INSTITUTE

GENE KING: FAA, DOT

DAVE PARKER: NTSB

JOHN REED: NTSB

BOB SHAW: NTSB

JIM WEBSTER: FAA FLIGHT STATISTICS BRANCH

GREGG VAN BRUNT: FLIGHT SAFETY COUNSELLOR

43