Pressure cabin investigations. Phase I (Aircraft Branch report)

              1. l \ IN DUS. & set.
(;J I( 1, /1{2, i 8-/7 I 0
File 629. 13 / Un 3 as AIR CORPS TECHNICAL REPORT No. 4220
AIR CORPS INFORMATION CIRCULAR
PUBLISHED BY THE CHIEF OF THE AIR CORPS, WASHINGTON, D. C.
=c)> ~~w==-=::r~:r, eon-_c:,
0 <' N~ro ------------------------------~ !·
Vol. VIII October 1, 1937 No. 710 ui-r
______________________________ c~ n=wfr
~ ffi
PRESSURE CABIN INVESTIGATIO
PHASE I
(AIRCRAFT BRANCH REPORT)
UNITED ST A TES
GOVERNMENT PRINTING OFFICE
WASHINGTON : 1938
Ralph Brown Draughon
LIBRARY
JUN 19 2013
Non•Depoitory
Auburn University
-- "'
..
' '
PRESSURE CABIN INVESTIGATIONS
PHASE I
COLLABORATORS
Maj. J . G. Taylor, Chief, Aircraft Branch.
Maj. D. G. Lingle, Chief, Engineering Shops Branch.
Capt. H. G. Armstrong, Director of Physiological Research Laboratory.
Mr. A. L. Berger, Head of Supercharger Unit, Power Plant Branch.
Mr. L. D. Bonham, Materials Branch.
Mr. G. P. Young, Materials Branch.
Mr. George D. Bogert, Aircraft Branch.
Mr. R. R. Curtis, Aircraft Branch.
Dr. John E. Younger, Project Engineer, Aircraft Branch.
(III)
;
I
TABLE OF CONTENTS
Page
Intr,oduction _____ ______ ___ ______________________ ___ ______ _____ _____________________ __ _____ __ _ 1
Chapter I- Preface ______ ____ _____________ _______________________ ____ ___ .. ___ ____ ______________ _ 3
Chapter 2- Resume of Experimental Investigations ____ __ ___ ___ _ .. __ _____________________________ __ 5
Object ______ _____ ____________ ____ _____________________ ____ ____ _________ _______ ________ __ _ 5
Procedure- Experimental investigations ______________________________________ ___ ________ _____ _ 5
Table I- Outline of investigations ___ _________________________________ ___ __ _____________ _ 5
Pressure cabin windows ____ _____ ___ ___ ___ __ ______________ _________________________ __ __ _ 5
Airtight juin ts __ __ __________ ________ _______ __ __________ ____ _________________________ __ _ 5
All-welded cabin ___ ________________________ ___ _____________________________________ ___ _ 5
Fleetster fuselage experiments ____________________________ _____________ ___ __________ ____ _ 6
Safety flap for windows ________________________________________________ ________________ _ 6
Safety valve __ __ ____ ________ ____________ __ ____________________________________ __ _____ _ 6
Pressure drop __ ____ ___________ ___ _________________________ _________ _____ ________ _____ _ 9
Sudden pressure release ____ ____ _____________ ________ _____ ___ ________________________ ___ _ 9
Cockpit windows __ ________________________________________________________ _______ ___ _ 9
Air duct ______________________________________________ _______ ______ ______ ___ ___ ______ _ 11
Cabin door _ ___ ____________ ____ __________ __ ______ ____ __________ __________ _______ _____ _ 11
Pressure-proof test ______________________________________ ___ ________________ . ________ _ 14
Discharge valve __ _________________________ ______ ______ ___ ____________________________ _ 14
Check valve ____ ________ _________ ____ ____ ____ __ __ ______ __________ __ _____ __________ ___ _ 15
Inward pressure safety valve _______ ___________________________ _______ _____ ___________ __ 15
Airflow s.ilencing ___ __________________ _____ __ _____________________________ ________ _____ _ 16
Refrigerated room tests __ ___________________ _______________________ _________ _______ ____ _ 16
Table II- Supercharged cabin in refrigerated room ____ ___________ ____ . _ _ __ _______ _____ _ 21
Discussion of investigations :
Structural problems:
Design features ____ ______ __ ___ _________ ___ ______________________ __ . ------ __ - · 23
T ype of structure _________ __ ________ ______________________________ ___________ ____ _ 24
Airtightness of joints ________ ___________ ____ ______ _______ __________ _____ _____ ______ _ 25
Skin expansion __ _______________ ___________________________ ______ _________ ________ _ 26
Bulkheads ___________ _________________________ __________ __ ____ ___ __________ ___ ___ _ 28
Pilot's visibility __ ___________________ ____________________ _________ _______ ______ ___ _ 29
Mechanical problems :
Doors, exits, window frames, airtightness ___________________ __ ______ _____ ___________ _ 29
Glands for controls __ .. _______________ ___________ ___ ________ _______________________ _ 29
Control of window frosting and icing __ ______ ____________________________________ ____ _ 30
Heat insulation and frosting of exposed metals ____ ______ __ ____ _______ __________ ______ _ 30
Utilization of dynamic air pressure and energy of discharged air_ _______ ~ __________ __ ___ _ 31
Effect of supercharging on instruments ___ __ ___ ___ _____________ _____ ____ __ ___ _____ ___ _ 31
Air flow and regulation problem :
Automatic and sensitive regulation of quantity and pressure, free from noise and danger of
freezing ____________ ______ ___ __ . ____ _____ _________ ___ ___________ ___ __________ __ _ 31
Automatic cabin sealing and emergency oxygen spray ____ _____________________________ _ 31
Safety valves ___________ _____________________ __ __________ _______ __ ________ ____ ____ _ 31
Physiological requirements:
Air supply _______________ ____________ _______ ______ __ ______ ______ ______ _________ __ 31
Pressure _________ ____ _______ ________ _________ ___ _____________________ _______ _____ _ 32
Permissible rate of discharge ________ ____ _________________ __ ___ _ • ______ __ - - - - - - - - ___ _ 32 •
Oxygen supply __________ ___ ___________________________________________________ ___ _ 32
Air contamination _______________ __ __ ________ _____ ______ ______ _______ - - - - - - - - - - - _ - - 32
Air conditioning ___ ___ ___ ____ _________________ " ___ __ _________ ____ _ - - - - - - - - - - - - - - - _ - 32
Supercharging requirements ___________________________________________ _ :. ___ _____ - _ - ____ _ 32
Figures 1- 34-- -------------------------- -- - ------------ ----------- ---- --·--··-------- ---- - 7- 29
(VJ
VI
Chapter 3- Development and Processes for Construction and Sealing of All-Metal Fuselages and Compart-ments
for Supercharged, High-Altitude Airplanes __ _____ ____ ____ ______ ____ ____ ____ ____ _
Object __ ___ _____ _____ ___________________ ____ ______________ _____ ___ ______ ____ ____ _____ ___ _
Discussion __ ______ _________ ____ _____ ___ _______ ____ ___ ______ ____ ___ ____ _____ ______ ____ ___ _ Project 1- -- ----------- - - ------- --- --------------- --- ------ - - -- --- -- -- - ---------------
Project II __ _____ _______________ __ _______ ____ ____ ______ ____ ____ _____ ______ __ ___ ______ _
Project Ill ____________ ___ _______________________ ___ ___ ____ ___ ______ ___ __ _______ _____ _
Project IV __ ___ ___ ____ _____ ___ ____ ______ ____ _____ __ ________________ ________ ______ ___ _
Conclusions and recommendations __________________________________________________________ _
Figures 35-38-- --- - ---------- ------- - - - ~--------------------------------------------------
Chapter 4- Stress Analysis of Pressure Cabin __________________________ _____ _______ _______ _______ _
Object _____ ___ ____ __ _________ ____ __ ____ ______________ ____ ______ __ ______ _________________ _
Conclusions _______ ____________ ____ _____ _____ ___ _______ _____ _____ __ ______ ___ ______________ _
Discussion ____ __________ _________________________________________________ ________________ _
Temperature effect on a metal fuselage ______________ ___ ____ ____ ________ __________ _______ _
Change in fength _____ ______ ____ ______ ________ ___ ________ _______ __ ________________ _
Diametral change ___________ ____ _____ ___ _______ ___________________________________ _
Equivalent stresses ___ ______ ____ ________________________ ___________ _______ ________ _
Pressure effects on a metal fuselage ___ _____ __ ________ __ _________________ ___ _____________ _
Dimensional chamges of a cylinder ___ ___ ____________________________________________ _
Stresses in noncylindrical portions ___ __ ___ _______ ___ _______________ ____ _____________ _
Spherical ends __ __ _________ ______ __ ______ ___ _____ __ __ _______ ______ ___ ________ _
Conical ends _____ _______ _______ ____ _______ _____ ____ ___ ______ __ _______ ________ _
Bulkhead stresses at end connections __ ______________________________________________ _
Plates subject to pressure __ ________________________________________________________ _
Circular plate, simply supported __________________________________ ____ _____ _____ _
Circular plate, clamped edges ___ __ ___ _____ ___ __________________________________ _
Rectangular plate, simply supported __ __________________________________________ _
Rectangular plate, clamped edges _____ __ ____ _____ _____ ____ _____ ____ _____ ___ _____ _
Elliptfo plate, clamped edges _____________________ ____ ______ __ _____________ __ ___ _
Numerical exam pies ____________ _______________________________ ____ __ __ ____ ___ _ Rectangular cross section ___________ __ __________ _____ ___ __ _______________________ __ _
Elliptical cross section __ ____________________ __ __ _____ ____ _______ _______________ ___ _ Relative weights of elliptical and round fuselages __ ____ ______________ ____________ _____ _
Discontinuity stresses __ ______ _________ __ _______ ___ ______ __ _____ ____ ________ ___ _____ _
Rigid bulkhead assumption __ _ _____ ___ ______ ___ _____ ___ ______ ______ ___ ____ __ ___ _
Flexible bulkhead assumption _____ ________________ _____________________________ _
Appendix:
Membrance stresses in elliptical heads ____ __ _____ ____ _______ ____ _________ _____ __ _____ _____ _
Chapter 5- Some Strength Characteristics of Laminated Glass for Windows of Pressure Cabins in Aircraft_
Object ___ _______ ___ ______________________ __ __________ __________________ _____________ _____ _
Conclusions and recommendations ____ ____________ _____ _____ _____ ___ ____ ___ _________________ _
General ___ __ ___.,-_______________________________________ ___________________________ ___ ____ _ Material _______ ___________ _____ ____ ___ ______ ___ ________________ _____ ____ _____ ____ __ __ Methods of test_ __ ____ ____ ____ __ ________ ________________ __ ___ ______ ____ ______ ________ _
Results ___ ____ _____ ______________________ ______ __ ____ _____ ___ _______ _ c ____ __ ___ _____ _ _
Discussion of results ___ __ ________ ______ ____ _______ ____ _____ ____ ____ _____ ______ __ ______ _
Figures 55- 60 __________________________________________________ __________________________ _
Chapter 6- Automatic Control of Air Pressure and Flow in Pressure Cabin ___ ______ __ _______________ _
Object __ ___ ________ ________________ ___ ______ __ _______ ________ ___ _______ _________________ _
Procedure ____ ___ ____ ________________________ __ ___ ___ ____________________________ ________ _
Discussion ________________ __ _________________ ___ _____ ___ _____ ____ _______ ___ _______ _______ _
Design factors ____________________________ ___ __ _______ ______ ____ ____ _____ _____ ________ _
Compartment pressure ___ ___ _____ _____________ _____ ___ _______ ___ ______ _____ ___ __ __ _
Ventilation ___ 0 _ - - - - - _ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Noise ____ ____________ _____ ____ ____ _____ ____ ______ ____ ___________________________ _
Method of air supply ____ _____ ___ _____ ___ __ ___ _____________________________________ _
General conditions _____ _____ ___ ___ _______ _____ ______ __ ________________________ __ __ _
Existing automatic systems ______ ____ ____ ____ ____ _________________________ __ ______ _____ Experimental systems _____ ________ __________________ __ • _______________________________ _
Tentative requirements for an automatic pressure and flow control system ___ _________ _____ __ _
P age
33
33
33
33
35
35
35
35
33- 36
37
37
37
37
37
37
37
37
37
37
38
39
39
39
40
40
40
40
40
41
41
41
43
45
45
47
47
49
50
50
50
50
50
51
51
52
51- 54
55
55
55
55
55
55
56
58
58
58
59
59
61
VII
Chapter 6-Continued.
Calculat ions ____ _______________________________ ____________ _______ __ __________________ ___ _
References ______________________ __ _____ _______ ______ _____________________ __ ____________ __ _
Figures 61- 65 ______________________________________________________ _______ ______ _______ __ Chapter 7- Physiological Requirements of Sealed High Altitude Compartments __ __________________ __ _
Summary __ _______________________________________________________ ______ __ _____ _______ __ _
Object _____ ___________________________________ ______ ______ __ ____________________________ _
M:ethod and procedure ________ ____ _____________ ______ _____________________________________ _
Results and discussion ___________________ ______________ _____________ ______________ _________ _
Introduction _____ __________________ ______________ ___ ____ ______ __ _____________________ _
General physiological considerations __________ ___________________________________________ _
Temperature controL ___________________________________________ ______ ________________ _
Humidity control_ _________ ___ _____ ___ _____ ______ ___ ___ __ _____ ________________________ _
Ventilation _____ _____ _______ _________________________________________________________ _
The carbon monoxide problem __ ____________ __ ___________ _______ _______ __ _____ _________ _
The carbon dioxide problem ________________________________________________________ ___ _ Other noxious gases _________________ _______ _______ _____________ ______ __ ____________ ___ _
Barometric pressure considerations ___________ ______________ _____________________________ _
Oxygen requirements ______________________________ ____________________ _____________ __ _ Conclusions and recommendations __________________________________________________________ _
References ____ ___________________________________________________________ ______ ________ __ Tables ______________ ___________________________________ ____ ______ ______ _________________ _
Figure 66---------------------------------------------------------------- - ----- -- ---------
Chapter 8- Air Conditioning a Pressure Compartment for Sub-Stratosphere Airplane _________________ _
Object ___ __ ___________________________________________________________________________ __ _
Conclusions ______ _________ ___ ____ ______ __ ____________ ____ ____ _____ ____ ___ __ ____ __ _____ ___ Requirements ____________________________________________________________________________ _
Oxygen pressure compartment method ______ ___ _ _____ ________ _____________ _____ . _______ ___ _
Pressure compartment method ____ ____________________________ ______ ______________ _________ _
Results _______________ __________ _____ ___ _____ ______ _____________________ ______ __________ _
Temperature contra] ____ ___________ _____ ___________________________________________ ___ _
Pressure control __________________________ __ ______ _____ __ _____________ ________________ _
Humidity control_ ____ __ ___________ ___________________________________ ______ __ ________ _
Carbon dioxide controL ___________________________ ________________ __________________ __ Chapter 9.-Air Compressors and Sources of Power Suitable for Pressure Cabins ________ _______ ______ __ _ Object ___ _____________________________ ______ ______ ____________________ _________________ _ Conclusions _________ ____________ ____ ________________________ ____ _____ ____ ______ __________ _
Compressor requirements ______ __ __ _____ _______________________________ ____ ____ __________ __ _
Sources of power for dri ving compressor ______________________________ ______ ___ ____ __ ______ __ Chapter 10.-Requirements of a Pressure Cabin Experimental Airplane __ ______________ ____ ______ ___ _
Object _________ ___ __ _________________________ __ ____ ___ ___________ _______ ______________ __ Conclusions ________ _____ _______________________________ ___ ____ ____ ______________________ _
Safety in flight experiments __ _______________________________________ ____ ___ ________________ _
General specifications ____ ________ ___ ____ _______ ________________________________ ·:· __________ _
Experience of personneL ___ ______________________ ___ ___________ _______ ___ ___ _____ ___ __ _
Load factors __ ________ __________________________ __ _____ __ ____________________________ _
Airtightness of cabin _________________________________________________ __ _____ __ ____ ___ _ _
Experimental airplane __ ____ ___ ____ ______________ ____ ____ ___________________________ __ _
V~ibilUY-- - --------- -- ------ - ----- - - ---- --- ---- ----------- --- -------- ------ -- ---- ----
Ground tests ____________________________________________ _______ ______________________ _
Flight experiments __________________ __ ____ ___ _____ ___________________________ _________ _
Centralization of control equipment _____________________________________________________ _
Separate pilot's compartment ____ ___ ________ ______ ___ ____ ______________________________ _
Detail specifications ____ __________ ______________ _____________ ___ _________________________ __ _
Structural ____ __________ ____ __ _____ ______ ___ ____ _____________________________________ _
M:echanica] _______________________________________ _____ ___ _________ ____ ____ ___ ____ __ _ Airflow and regulation __ ______ _______ __ ____ __ ______ ___________________________ __ _____ __ _
Physiological ___ __ _____ _____ __________ ___ __ ___ ___________ _____ _____ ___________ ________ _
P age
62
63
56- 60
64
64
64
64
64
64
66
66
66
67
68
68
69
69
70
72
72
73- 74
72
75
75
75
75
75
75
76
76
76
76
76
77
77
77
77
77
78
78
78
78
78
78
79
79
80
80
80
80
80
80
80
80
81
82
82
VIII
LIST OF FIGURES A JD PHOTOGRAPHS
CHAPTER II
~ D~~
1. Fleetster pressure cabin, side view, showing closure of window and door space ______________ _____ ____ _
2. Fleetster pressure cabin, showing rear cone and pressure regulating device _____ _______________ ____ _
3. Fleetster pressure cabin close-up, showing rear cone and pressure regulating device ___ _____________ _
4. Fleetster pressure cabin close-up, front cone and quick release door_ _____ ___ _____________________ _
5. Automatic experimental window flap for pressure cabin _ ____ __________ ____ _____________________ _
6. Safety valve for pressure cabin ____ ______ _____ _____ __ ____________________ • ____________________ _
6A. Pressure drop-time curves for various round orifices __________________________________________ _
7. Cut-out for pilot's windows _________________________________________________________________ _
8. Plan view showing cone cut-out ____________________________________________________________ _
9. Cross section of windshield installation ____ ___________________________________________________ _
10. Pressure cabin door, interior vi ew, showing locked position of door with valve closed _________________ _
11. Pressure cabin door, interior view, showing locked position of door with valve open ________________ _
12. Pressure cabin door, showing in side construction, door in lock ed position, valve open _ _____________ _
13. Pressure cabin door, showing inside construction, door in locked position, valve closed ________ _____ _
14. Pressure cabin door, showing inside construction, door in unlocked position __ _____________________ _
15. Door release valve ____ _________________ ________________________ ___________________________ _
16. Airtight seal of door_ _______ ___________ ____ ______________ __ _____ _______ ____________________ _
17. Vertical cross section of discharge-valve muffler_ _____________________________ __ _____________ __ _
18. Plan view of cabin in refrigerated room __ ____ ______________ ________ __________ __ __ ______ ____ __ _
19. End view of cabin and air supply system ___ _____________ ___________ __________ __ _____
0
_______ _ _
20. Legend for cabin in refrigerated room- - -------------------------~------ - -- - ------------------ -
21. Pressure cabin in refrigerated room, side view, showing door and supercharger lines _______ _________ _
22. Pressure cabin in refrigerated room, front three-fourths view, showing heater and supercharger lines __
23. Pressure cabin in refrigerated room, front view, showing pilot's windows, quick release door in place ____ _
24. Pressure cabin in refrigerated room, front view, showing pilot's windows, quick release door removed ___ _
25. Pressure cabin in refrigerated room, interior front view without insulation __ _____ _____ ____ ___ _____ _
26. Pressure cabin in refrigerated room, front view showing windows iced and frosted ____ _________ ____ _
27. Pressure cabin in refrigerated room, interior front view, showing frosted windows ____ _____________ _
28. Pressure cabin in refrigerated room, front view, showing windows cleared of ice and frost_ __ ________ _
29. Pressure cabin in refrigerated room, rear interior view, showing discharge valve and felt partition ______ _
30. Pressure cabin in refrigerated room, interior front view, showing insulation ____ ____ _____________ __ _
31. Pressure cabin in refrigerated room, front view, showing windows cleared of coat of ice ____ ________ _
32. Bulkheads designed to allow for expansion of skin ______ _____________ ___ ______________ ____ _____ _
33. Air-tight torsion gland and air-tight sliding gland _ _ _________________ __________________________ _
34. Bellows type gland ________________________ _________ _________ __ _______ _____________________ _
CHAPTER III
35. Tank X at left ; tank Y at right of figure. Tank heads YY at right and Z at left, A at bottom and B
Page
7
7
8
8
9
10
10
11
11
11
12
12
13
13
14
15
15
16
17
17
17
18
19
20
21
22
22
23
24
25
26
27
28
28
29
at top of tank x ______ ____________ _____ _________________________ ____ ___________________ -- 33
36. Tank z _____________________________________________ ________ ____ __________________________ 34
37. 30-inch aluminum alloy pressure tank __ ___ _____ ___________________ __ _________ ________________ _ 35
38. Test joints for 30-inch pressure tank____ ______________________ ___ __ _____ ______ _______________ _ 36
CHAPTER IV
39. Stresses on element of pressure head_______ ______________________ ____ _________________________ 39
40. Stresses on element of pressure head____ _____________________________ ___________ _____________ __ 39
41. Conical pressure head_______________________________________________________________________ 39
42. Skin movement at bulkheads_ _ _ _ _ _ _ _ __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 40
43. Bulkhead loads from conical head _________________________________ ___ _______________________ _ 40
44. Loa,ds on rectangular plate __ ____________________________________ ____ _____________________ .: __ 40
45. Rectangular pressure cabin-------- - --------- ----------------------- ---~-------------------- - 42
46. Equilibrium of side of r ectangle______________________________________ ______________________ __ 42
47. Corner continuity of a r ectangle ___ __________________________ __ ___ _________________________ __ 43
48. Equilibrium of a quarter ellipse___________________________________ ___ _______________________ _ 43
49. Element of an ellipse ______________________________________________________________________ _ 43
50. Bending of an element___________________ _______ ______________ ___ __________________________ _ 44
51. Points of inflection for an ellipse _____________________________________ ______________________ __ 44
52. Discontinuity loads for a cylinder__________ __________________ _____ _________________________ __ 46
53. Discontinuity at a bulkhead____________________________________ ___ __________________________ 47
54. Elliptical pressure head_ ______________________________________ ____ _______________________ __ _ 49
IX
Figure
No. CHAPTER V P age
55. Cell used for burst measurements ___ ____________________________________ _________ ____________ 51
56. Data for circular plates ___________________________ ___ ____ ____ -·- _ _ _ _ _ 52
57. Data for square plates _______ ___ ______ ____ ____ ________________________ __ _________ ___________ 52
58. Characteristic figure on 8-inch diameter specimen __ ________________________ ____ __ _______ __ ___ __ 53
59. Characteristic figure on 14-inch diameter specimen__ ________________________________ ____ _____ __ 53
60. Characteristic figure on 10-inch square heat-treated glass_ ____ ___ ___ ___________ __ ______ __________ 54
CHAPTER VI
61. Required ventilation rate to maintain compartment relative humidities of 50 percent and 75 percent
under various compartment and atmospheric temperatures____ ________________________________ 56
62. Atmospheric temperatures and corresponding altitudes at which frost formation on single-pane windows
may be expected under various conditions of compartment temperature and relative humidity____ 57
63. Pressure compartment inlet temperatures at various altitudes resulting from temperature rise in
supercharger__ _____ ____________ _______ _________________________________ _______________ __ 58
64. Self-operated discharge valve ___ _________ ___________ ___ _______ __________ _____ ____ ___ _____ ____ 59
65. Automatic electrically-operated discharge valve ____ _______ ______________ ____ _____ ___________ ___ 60
CHAPTER VII
66. Oxygen dissociation curve for human blood ___ __ _______ _______ ______ __________ ______ ________ ___ 72
.\
PRESSURE CABIN INVESTIGATIONS
PHASE I
Introduction
This report covers phase I of the pressure-cabin in­vestigations
made by the various branches of the Mate­riel
Division between July 1, 1935, and June l, 1936.
The experiments conduct.ed to date on the various
components of a pressure cabin and on the appurte­nances
necessary for the proper operation of such a
cabin are described in detail by the organizations
assigned to the particular tests.
The report is comprised of several chapters for the
purpose of clarity. Although interrelated, the chapters
each cover a particular field of research, and to combine
them under one heading would only confuse the reader
with a myriad of titles and subheadings. Each author
is a competent investigator in his field, and it is only
fair to his ability to allow his personal presentation of
the problem to remain unchanged.
Due to the interrelation of the chapters, it is only
natural that repetition should occur. No attempt was
made to restrict the subject matter of any author, so
each chapter is a perfectly free expression of opinion
regarding the work undertaken. Each chapter contains
its own summary and conclusions where necessary, and
no general summary can be drawn without verbatim
repetition of the individual summaries.
(1)
..
PRESSURE CABIN INVESTIGATIONS
PHASE I
Chapter !.-Preface
(Prepared by Maj. J. G. Taylor)
HISTORY
The problem of supercharging an airplane cabin in
order to maintain low-altitude temperature and pressure
conditions during high-altitude flights was first under­taken
by the United States Army Air Corps at McCook
Field in 1920 under the supervision of Lieutenant
Foulk. An oval tank was constructed to fit within
the fuselage of the U. S. D. 9--A, P-80, and a wind­driven
gear supercharger was utilized to provide pressure
for the tank. Several flights were made during the
peTiod 1920-22; the fact was established that a super­chaTged
cabin was entirely feasible if the valve opera­tion,
temperature control, and visibi lity could be
improved.
FurtheT recommended tests with the U. S. D. 9--A
weTe abandoned. Plans were laid in 1929 and in 1934
fOT high-altitude airplanes, but it was not until 1935
that actual experimental work relevant to supercharged
cabins was again undertaken. On April 29, 1935, an
expenditure order was initiated by Capt. C. F. Greene
to "study and report on the present literature or exist­ing
development of supercharged cockpits or cabins for
high-altitude aiTplanes. Thereafter to recommend a
program of such further study and expeTiment as
may be Tequired to obtain an airplane especially
designed or rebuilt to incorporate the desired features."
As a result of the foregoing, a definite project was set
up and Dr. J.E. Younger, professor of mechanical engi­neering
at the University of California, was engaged for
1 year as project engineer to conduct the necessary
preliminary investigations and to prepare the way for
the procurement of a flying article. Dr. Younger
commenced this work on July 1, 1935.
On September 13, 1935, the project engineer initiated
an expenditure mder for the second phase of the project,
"Further Development of the Supercharged Cabin
Project." This second phase is as yet unfinished, and
the findings to date constitute the body of this report.
CONCLUSIONS
It is apparent at this time that the actual accomplish­ment
_of supercharging at least the cabin portion of
present-day airplanes is approaching the practical
stage. vVhile there are many small details yet remain­ing
to be more fully investigated or means found to
accomplish them in a simpler, more reliable way, never­theless
all signs point to the conclusion that the major
difficulties have been sufficiently analyzed to permit the
ready reduction to practice.
So far nothing has arisen in the present development
phase to greatly modify the conclusions reached from
the projects of 1922, 1929, and 1934 which, in the main,
were that the problem was relatively simple and con­sisted
chiefly of adding together existing knowledge by a
process of rationali zation, together with the physical
development of many small allied parts through a
process of cut and try. As will be noted in this report,
it was deemed expedient to first prove all this structure
and equipment to the fullest extent possible by labora­tory
means. The point has now been reached where
all these things may be readily accomplished on a fly­ing
airplane without imposing too great a burden of
weight upon it. The structural problems have shown
themselves to be of such a nature that they will not
require too great a departure from fabrication methods
normally employed in modern semimonocoq ue metal
aircraft.
The end of the laboratory phase has been reached and
one of reality can now be entered. A flying laboratory
airplane will now be fabricated in a manner analogous
to that of 1922, but with the advantage of better equip­ment
the physiological research, which has always pre­sented
the greatest unknown factor, can be carried on.
It is likely that considerable time will be required to find
ways and means of solving the physiological problems
which are bound to arise in the use of such equipment.
(3)
In order to obtain the most efficient operating alti ­tude,
from an aiTplane economy point of view, a seTies
of very detailed flight analyses under all conditions at
increasing altitudes must be made until a sufficient
pressure altitude is reached which will disclose the
obvious advantages of supercharged airplane cabins
over other means now available for sustaining human
life in the substratosphere. The purpose will be to
make possible easy and safe flight at these altitudes by
continuing along the present path step by step. It is
believed that this report, therefore, sustains and justi­fies
the outline of the situation prepared by the Aircraft
Branch in May 1934, which initiated the phase of the
work herein r eported.
PRESENT ST ATOS
The phase of the pressure-cabin project covered by
this report includes the entire theoretical investigation,
exclusive of aerodynamics, and the preliminary portions
of the experimental investigations. Further experi-
4
mental work is now in progress, and will be recorded in a
future report; these experiments include physiological ,
dynamic, static, and vibratory aspects of design, and
additional information on strength of glass and opera­tion
of control mechanism. An actual airplane is being
equipped with a cabin especially constructed for super­charging,
and flight tests will be conducted to obtain
first-hand data on all phases of high-altitude flight.
Chapter I !.- Resume of Experimental Investigations
(Prepared by Dr. John E. Younger)
OBJECT
To present a summary of the experimental investiga­tions
undertaken in connection with the development
of pressure cabins for high-altitude airplanes.
PROCEDURE
The experimental investigations were outlined to
include studies of the items contained in table I.
The plan of the experimental investigations was as
follows:
(a) To simulate as nearly as possible the pressure
cabin in sub-stratosphere flight under labora­tory
conditions on the ground.
(b) With the data obtained in these preliminary ex­perimental
investigations to design a substrato­sphere
pressure-cabin laboratory airplane.
A chronological enumeration of the investigations
performed will first be presented and then the problems
of table I and their solution as they appear from the
experimental and theoretical investigations will be
discussed in the order enumerated in the table.
Pressure-cabin window experiments.-In order to ob­tain
data on the strength and type of construction of
windows suitable for pressure cabins, an investigation
was initiated under the direction of Mr. G. P. Young,
of the Materials Branch. Attention is called to chapter
V of this report.
Construction of airtight joints.- A number of experi ­mental
investigations were carried out on the construc­tion
of airtight joints. Spot-welded, seam-welded, and
metal-sprayed, as well as various types of riveted joints,
were tested. Attention is called to chapter III of this
report.
Class of problem
Structural.. ___ __ __ __ _
Mechanicai. ___ _____ _
TABLE I
Specific problem
a. Type of structure best suited.
b. Airtightness of joints.
c. Expansion of skin under pressure.
d. 'l'emperature stresses combined witb pressure
and flying stresses.
e. Preventing of explosion in case of failure of
seam or the development of a crack due to
flying loads, vibration, bullet punctures,
etc.
J. Development of restraining bulkheads.
g. Strength of glass windows.
h. Pilot's cockpit with sufficient visibility.
a. Doors, emergency exits. " 'indow frames, air­tightness.
b. Glands for control wires, tubes, etc., to control
surfaces and engines.
c. Control of fogging and frosting of windows.
d. Automatic window flaps to bold air pressure
in case of breakage of window.
e. Utilization of dynamic air pressure and energy
of air discharged from the cabin.
/ . Effect of supercharging on instruments.
Class of problem
Air flow and regula­tion
.
Physiological. ___ ___ _ _
TABLE I-Continued
Specific problem
a. Automatic and sensitive regulation of quan­tity
and pressure, free from noise and danger
of freezing at altitudes.
b. Automatic sealing of cabin and release of oxy­gen
spray in case of failure of superchargers.
(Check valves, etc.).
c. Safety valve.
d. Inward pressure safety valve.
a. Air required per passenger.
b. Pressure required.
c. Rate of discharge of air permissible in emer-gency.
d. Quantity, etc., of oxygen in emergency spray.
e. Temperature control.
f . Humidity control.
g. Ventilation (air circulation).
All-welded experimental pressure cabin.-Quoting
from an Aircraft Branch Routing and Record Sheet 9-n
Supercharged Cabins of July 20, 1935 :
"General purpose at present development of project :
To begin work by building a simple experimental super­charged
cabin to try out structural, mechanical, and
physical features which require experimentation.
"To do this a simple structure as follows is planned:
"l. Circular fuselage section, 60 inches in diameter,
with a flat bulkhead for one end and a truncated right
cone bulkhead for the other end . The length to be
about 100 inches.
"2. The structure to be designed to resist t ypical
loadings such as would exist on the central portion of an
airplane fuselage, and a cabin pressure of 7.5 pounds
per square inch above atmospheric.
"3. A free edge on the flat end is to be left for attach­ing
other experimental bulkheads, or applying loads in
vibration and static tests.
"4. The flat bulkhead is to be designed with a 20- by
40-inch door for entering the cabin.
(5)
"5. The truncated cone end of the cabin is to taper
45° to a diameter of 30 inches. A flat circular plate 30
inches in diameter is to be designed for the 30-inch
section of this cone. The plate is to be removable, as
a door, so that plates with experimental windows,
glands, wiring t erminals, etc., may be clamped in place.
"6. Air leakage to be reduced to a minimum (welded
joints are being used).
" 7. The structural shapes are to be simple forms that
can be constructed in the repair shops."
This cabin was designed as per assembly drawing
No. X36N l66. It was planned to set this pressure
cabin up in the refrigerated room of the Power Plant
Bra nch to simulate, as neariy as possible, the conditions
of actual flight in the substratosphere. It was then
proposed to subject this article to static and dynamic
tests when under a differential pressure comparable to
that which it would be subj ected in actual flight condi­tions.
Fleetster fuselage experiments.-While these above
projects were being developed, it was noted that an
unused Fleetster fuselage had desirable characteristics
for a pressure cabin. It appeared that much progress
could ba immediately made on certain of the many
items to be investigated by the use of this fuselage.
Accordingly, the fuselage was revised for this use.
The fuselage was cleaned and stripped of its acces­sories.
The usable portion, from the rear wing spar
bulkhead to the tail skid, was cut out from the re­mainder
of the structure. Airtight cone bulkheads
were bolted in the ends. The riveted joints were rein­forced
by doubling the number of rivets. The interior
was painted with a synthetic rubber compound (see
ch. III) to make the joints airtight (see figs. 1, 2, 3,
and 4).
The joints in the reclntly constructed portion of the
fuselage, namely, in the cones at the ends, were made
airtight by the use of fabric soaked in marine glue
pressed between the metal sheets at the riveted joints.
A manhole was constructed for the larger cone. (See
figs. l and 4, and r efer to drawing No. X36G458.)
This quick release door was constructed, as may be
noted on the photographs and the drawings, so that 11,
jerk on the handle would permit the door to fly off
entirely away from the cabin. Air was supplied to
the cabin from a high-pressure air line. A pressure
gage was used to indicate the pressure in the cabin.
It was the immediate purpose in preparing this fuselage
as a pressure cabin to study the following problems:
(a) Sealing joints against air leakage.
(b) Construction and strength of cone bulkheads.
(c) Effect of sudden release of pressure on the struc­ture
and on the passengers.
(d) To check the equation for time required to re­duce
the differential pressure in the cabin to zero by
opening various sized holes.
The physiological phases of these experiments were
under the direction of Capt. ·H. G. Armstrong and are
presented by him in chapter VII of this report.
Preliminary to these experiments the cabin was
proof tested to a differential pressure of 10 pounds
per square inch. The theoretical minimum strength
of the cabin (riveted joints) was about 15 pounds per
square inch. The strength of the skin, which was 0.032
inch in thickness, was approximately double this figure.
The quick-release door was tried for operation at a
differential pressure of 5 pounds per square inch. The
door operated quite satisfactorily, and the resulting
explosion was quite startling. The· heavy door was
blown as from a cannon for a distance of 30 to 40 yards
with quite a loud muffled report.
A window, size 6 by 14 inches, was inserted in the
side of the fuselage in a wooden frame, the approxi­mate
cross section of which is shown in figure 5. The
window glass rested on a soft sponge-rubber pad one­fourth
inch thick by one-half inch wide. This was
found to be a very satisfactory seal and foundation for
6
the glass window. It was noted, however, that there
was some seepage of air through the pores of the
rubber. However, sponge rubber which is nonporous
may be obtained. The trade name of one such product
is Chicago Foam, which is manufactured in England
by the Anazote Process, and another is a composition
developed by the Liquid Carbonic Corporation, 52
Vanderbilt Avenue, New York City. There are un­doubtedly
other airtight sponge-rubber products on
the market which may be obtained. Also, ordinary
sponge rubber may be made airtight by the applica­tion
of, for example, marine glue to the inner edges.
Safety flap for windows.-Inasmuch as it appeared
that safety in the substratosphere depended prim1J,rily
upon retaining the cabin differential pressure; it
seemed advisable to attempt to develop a safety flap
for the windows such that in case of the breakage of
the windows the flap would close automatically and
hermetically seal the opening. Accordingly, the flap
shown in figure 5 was constructed. This flap consisted
of a one-eighth inch aluminum plate hinged to the top
of the window, being spring-hinged so that the impetus
given by the spring and the weight of the flap would
cause it to close when released. The flap was sup­ported
by a curved rod, the upper end fitting into a
socket in the flap and the lower end resting on the
center of the window. It was anticipated that in the
case of breakage of the window the wire support would
be released, allowing the flap to close. It was found
after several trials that a soft sponge rubber seat one­fourth
inch thick glued around the window formed an
ideal gasket on which the flap could rest when closed.
Experiments with the operation of this flap showed
that it would operate satisfactorily to retain 75 per­cent
of the cabin pressure for pressures of the order of
5 pounds per square inch. It was quite obvious that
with the use of a stronger spring to close the flap more
readily this loss could be reduced, if desirable. The
con,clusion in this case was that a window flap could
be made to operate quite satisfactorily, if it seemed
desirable. The attention of the reader is called to the
discussion of the desirability of window flaps later on
in this report.
Safet11 va./ve.-It was realized that adequate safet.y
valves would be necessary for the safe operation of the
pressure cabin. Bearing in mind the fluttering tend­ency
of ordinary spring-operated safety valves, it
appeared that a special ingenious device would have
to be developed to operate satisfactorily as a safety
valve. However, it seemed reasonable to try first the
simplest possible type of spring-loaded valve. Accord­ingly,
thevalveshownin drawing X36D529 and indicated
diagrammatically in figure 6 was constructed and put
in place. This valve was, much to the surprise of those
interested, very satisfactory and was used throughout
these series of experiments. When the supercharger
was later connected to the pressure cabin a larger valve,
constructed in the same manner, was installed. Both
these valves operated quite satisfactorily during all the
experimental work which followed . These valves were
satisfactory to relieve the pressure at about one-quarter
pound per square inch above that for which the regu­lating
discharge valve was set. In testing the safety
7
FIGURE 1.- Fleetster Pressure Cabin side view showing closure of window and door space.
FIGURE 3.- Fleet ster Pressure Cabin showing rear cone and pressure regulating device.
42060- 38-- 2
8
FIGURE 3.- FleeLsLer Pressure Cabin close-up showing rear cone and pressure regulating device.
FIGURE 4.-Fleetster Pressure Cabin close-up, front cone anrl quick release door.
of the refrigerating chamber set-up for routine opera­tion
the regulating valve was repeatedly completely
closed to check the operation of the safety valve. In
all cases they gave perfect service. It was concluded
that no further investigation was needed in this respect,
as these valves were extremely light, simple, and con­tained
no mechanism which was likely to cause trouble.
Pressure drop.- The time required for a pressure
drop from 5 pounds per square inch above atmospheric
pressure to atmospheric pressure, with various size
orifices in Fleetster fuselage (volume, 280 cubic feet),
was as fo llows:
Orifice No.
]_ _______________ --
2 ___ ___ ___________ _ 3_ - - - -- --- ------ ----
54_ _ _-_- -__--__-_-_-_-_-_-_-_-_-_-_· _- -_
5 __ __ _ -- -- ---- ------
Window __________ _ _
Diameter I
In ches
2. 31
1. 68
1. ~9
. 78
. 61
. 51
13 by 5
See figure 6.\ for plotted data.
Area
Square inches
4. 19
2 22
1. 52
. 48
65
. 29
. 20
rrime to drop
Seconds
8. 3 (compu ted).
14. 5 (experimcn .al).
21. 5 (experimental ) .
66. 0 (experimental) .
112 (experimental) .
160 (experimental) .
. 54 (computc<l ).
The following formula which neglects the effect of
t . mperature change and humidity represents the data
with fair accuracy.
In which:
V o lw,, I
to= Apn'V2gV Po
p 0 = differential pressure before orifice is opened,
pounds per square foot.
V0 = Volume of cabin, cubic feet (280 in this case).
A = Area of cross section of orifice, square feet.
W0= Weight of air per cubic foot at altitude of
cabin.
p0 = Barometric pressure, pounds per square foot,
at altitude of cabin.
g= 32.2.
Experiments on the sudden release of pressurc.- For a
detailed account of these experiments the reader is
referred to chapter VII of this report by Capt. H. G.
Armstrong.
To determine the physical effect of the release of
pressure in the cabin approximately 50 tests were made,
in which the pressure in the cabin was released. The
magnitude of the pressure and the rate of release of
pressure were increased by small increments until a
pressure of 5 pounds was reached and in which the
rate of discharge was a small fraction of a second,
practically equivalent to an explosion . All of these
tests were carried out with Captain Armstrong and the
author in the pressure cabin. No apparent physical
discomforts were found to result. Other personnel
were later brought into the experiment to further check
the findings.
During later experiments the author was subjected
to rapid raising and lowering of pressure probably
hundreds of times, yet at the present time can attri­bute
no physical discomfort to these experiments.
Also, during the later experiments other personnel were
subjected to the same treatment in the routine work,
yet complained of no physical discomfort.
It may be of interest to note here that in a later
9
set-up in the refrigerated room, on many occasions,
the pressure was increased from zero to 4 pounds per
square inch in 30 seconds with only very little physical
discomfort; th is discomfort being in connection with
the ear drums. In th is case the ears must be kept
cleared by frequent swallowing. This is approximately
equivalent to a dive of 20,000 feet per minute.
Pilot's cockpit windows.-It was originally the in­tention
to use the all-welded job previously mentioned
in this discussion to carry out t ests on windows. How­ever,
since the Fleetster fuselage proved so satisfactory
as an experimental cabin, it seemed advisable, for the
sake of saving time, to build the pilot's cockpit windows
into this job. It should be made clear that the type
of construction used in this case is not necessarily rec­ommended
for the flight article. It was used here
(
\ \
'~=:1,:T~:;Mf,N/>Q'INDAoW~
/,fol()c lj" CABIN
.SPrT .Rl/81/Ek ~MtET
f'l)R r/AP
cXPt/?ll'fliNlitlt- W,NOOW
,"'JtME, WOOl)t:I{
O/'T lftl86'ER (i'IISKET
WINDOW <i"IA~
F IGURE 5.- Automatic experimental window fl.ap for pressure cabin.
only because it was the quickest and simplest way of
constructing a job convenient for the experimental
work required. While the structμre in this case was
quite irregular and fairly light, as may be noted from
the photographs and the discussion, yet it is heavier
than necessary for practical use on an airplane. The
cockpit windows were installed as follows: A portion
of the cone at the larger end of the fus~lag~ _'IVas- cut
out as noted in figure 7. An 0.064-inch steel plate· was
then fitted around the cut portion of the cone .as noted
in figure 8. The frame of the windows was then· con­structed
and welded together into one piece and fitted
into place. The material of this frame was 0.064-inch
steel sheet. The top cover of this structure was then
formed from 0.050-inch duralumin sheet bumped to fit.
10
Valve he«d
Soft rubber
.se.a.t for valve
V<iltle spri.nq,pressure
regulated from in.side
of ca.bin.
Sa.fetlf v.altJe for pressure c.abin
F IGURE 6.
FIGURE 6A.
One-fourth inch steel bolts were used to assist in tying
the upper cover to the base plate of the windows, thus
cut out
/
C.3.bi.n
FIGURE 7.-Cut-out for pilot's windows.
Air-ti9~t joint
Secl:!on AA
Pla..n vi'ew showt'ng cone cut-out
FIGURE 8.
Steel pla.fe
Ct:lb in
aiding in resisting the upper pressure of the air on the
top of the cockpit. Seats for the window glass were
made of sponge rubber one-quarter-inch thick. The
11
glasses were held lightly in place by strips of thin sheet
screwed to the frame around the glass.
Air d1tcl. - An air duct was made by adding a sheet
aluminum structure to the under side of the lower
window plate as noted in figure 9. One-eighth inch
holes were drilled through the plate on the inside of the
cabin, so that the warm air from a supercharger or
heater could pass upward and thence be deflected over
the window to prevent it from frosting on the inside and
to prevent ice from forming on the outside. It was
hoped that the windows could be kept 'warm enough
for this purpose. Provisions were made as noted in
figure 9 for the inclusion of a thin glass window on the
inside of the cabin to form a narrow passage for the air
over the window. It was arranged for this air to be
11; Sa!ely 9/a.ss
I!§§§ Sponge rubber
~ S~ee( = Alum,ntim
Cro//s :section of windshield in sta.llaf,on
.Sr:a.le 1 "•Z"
FIGURE 9.
discharged through holes in the support of the inner
window at the top. Twice the n umber of holes were
drilled at the top as were drilled at the bottom, in order
that no appreciable pressure would be built up between
the two glasses which might prove dangerous by blow­ing
the thin inner glass inward into t he pilot 's face.
The lower support of the inner glass served as a de­flector
of the air when the inner glass window was
removed.
Cabin door.-See figures 10, 11, 12, 13, and 14.
In developing a door for the cabin the following fea­tures
were borne in mind:
(a) T hat the door must be light and easily closed.
(b) That it must not be readily possible to open
12
FIGURE 10.-Pressure cabin door, interior view showing locked position of door with valve closed.
FIGURE 11.-Pressure cabin door, interior view showing locked position of door with valve open.
13
FIGURE 12.-Pressure cabin door showing inside construction: door in locked positiou, valve open.
FIGURE 13.-Pressure cabin door sbowing inside construction: Door in locked position, valve closed.
the door when the cabin is subjected to a differential
pressure.
(c) That the operating mechanism be such as would
be normally expected in the operation of a door yet
provide for a quick release of the cabin pressure and
thence a normal opening of the door.
14
The door shown in the photographs mentioned above
was constructed with these principles in mind. It will
be noted that the door is designed to be held in place
by 14 bolts all actuated at the same time by a handle
operating either from the inside or the outside. The
principle of operation of this feature may be readily
noted from the photographs. A study of the photo­graphs
will show that in opening the door the handle
may be turned through approximately !l0° before a
cam, operating in a slot, makes the proper contact for
disc. The door was sealed around the edges as shown
in figure 16.
Press11re proof-tesl.- After completion of the door in
the pilot's cockpit the cabin was proof-tested with an
internal pressure of 8 pounds per square inch. The
operation of the door was found to be satisfactory.
The door was used hundreds of times throughout
succeeding experiments in all temperatures ranging
from +80° Fahrenheit to -45° Fahrenheit, yet at all
times operated with perfect satisfaction. To illustrate
the need of such a rlevice on the door it will be inter­esting
to relate the following incident.
During the routine testing in the refrigerated room
the author naturally became quite familiar with
features which once appeared to have elements of
danger. It is under circumstances like these that
FIGURE 14.-Pressure cabin door showing inside construction: Door in unlocked position .
withdrawing the bolts. In the process of moving
through this 90°, however, the disc on the inside of the
door, actuated directly by the handle, rotates, thus
bringing the air holes in the disc and in the seat of the
valve, on which the disc rotates, in line, thereby re­leasing
the air from the cabin. The speed of release of
the air in the cabin could be regulated by the size of the
holes and the size of the disc and its seat. After a
number of trials it was found that leakage of air could
be readily prevented between the disc and its seat by the
use of glazed leather three-sixteenths inch thick, soaked
in oil and held in place by a pad of sponge rubber as
noted in figure 15. It will be noted from t.he figure that
the leather, not being attached to the disc, would be
held firmly in place by the pressure of the air through a
number of small holes drilled in the top of the metal
automatic devices and warnings are necessary. On this
particular occasion something occurred on the outside
of the cabin while the author was working on the inside
with the cabin subjected to a differential pressure of
about 5 pounds per square inch, which would make
about 4,000 pounds on the door. Overlooking the
fact that the cabin was under pressure the author non­chalantly
started to open the door and walk out.
Naturally the loud hissing noise when the valve in the
disc became in operation, brought the absent-minded
operator back to his senses immediately. Thus a very
valuable bit of test data, though not planned, was
obtained.
Discharge valve.- For a detailed discussion of this
feature the reader is referred to chapter VI of this
report by Mr. R. R. Curtis.
15
Two valves were designed . The first was an air­pressure-
operated valve (see drawing No. X36G507),
and the second was an electrically operated valve (see
drawing No. X36G632). The first of these valves was
constructed, installed, and tested. At the writing of
this report the second valve is in the process of being
installed.
The principle prerequisites of a sa t isfactory discharge
valve are:
(a) It should hold a constant pressu re on the inside
of the cabin regardless of the barometric pressure on
the outside; that is, its operation should be independent
of the barometric pressure.
(b) The valve should be free from any possibility of
freezing.
In operation, the air-pressure-operated valve ful­filled
these two prerequisites satisfactorily for differen­tial
pressures as high as 4 pounds per square inch and
above. At low differential pressures the spring effect
of the sylphon bellows was sufficient to cause a variation
of about one-half pound per square inch when the
differential pressure was changed from about 1 pound
Pd._/ h of escaping
a,r
Sedled wall o f door
Valve .seat-durdlumin
Door h,;,.nd/e
Disc: of spont1e rubber
Disc of oil sodked leather
Door release tia./tie
FIGURE 15.
per squa re inch to about 5 pounds per square inch.
This change can be further lessened by lengthening the
sylphon bellows so that the spring effect will be lessened .
This valve has the di sadvantage that it will not close
readily when the pressure is reduced in the cabin ; for
example, if the pressure is, say, 5 pounds per square inch
in the cabin and the supercharged air supply is cut off
suddenly the discharge valve will allow the pressure
to drop to 3 or 4 pounds per square inch before it will
close. On the other hand, the valve has the decided
advantage of being very simple in operation and con­struction.
The author believes that for an experimental
laboratory ship this valve would be preferable to an
indirectly operated valve, such as one operated elec­trically.
It would be necessary, however, to provide
an arrangement whereby the valve could be closed
immediately by manual operation.
Check valve in supercharger line.- A check valve was
provided in the supercharger line so that in case of the
cut-out of the supercharger the cabin would not lose
immediately its pressure by the flow of air back through
the supercharger. This valve was simply a flap of
rubber supported by a perforated disc, thus forming a
simple rubber flap valve. This valve proved to be
very satisfactory in operation but not very rugged.
The fluttering of t he valve, due to t he high air stream,
caused it to tear away. This fact must be taken into
consideration in the design of a check valve. It
appears probable that such a valve, free from any
chance of failure, would be of a ball-and-cage type,
Outside wa.lt of door
Locking bolt
Inside wa.lt of door
Rubber f lap
-"--.__Air pressure holds flap
- air-tight
Door frame
Wooden block around door
frame , glued in place.
Air -t i9ht seal of door
FIGURE 16.
although an ordinary metal flap valve may be perfectly
satisfactory for longevity of operation.
Inward preswre safety valve.- Since it was contem­plated
in t he experiments to simulate high-a ltitude con­dit
ions by drawing a vacuum in the discharge line from
t he cabin by means of a large vacuum pump, it was
necessary to provide a means of preventing a partial
vacuum being formed inside the cabin in case of the
failure of the intake line. It should be noted in this
case, as in the cases of any flimsy shell such as an air­plane
fuselage, that a very low inward pressure on _the
walls of the cabin, for example, of the order of one-half
pound per square inch, would cause the walls to collapse
inward. This condition will also exist in the flight
article and must be provided for; for example, suppose
that at a high altitude the differential pressure in the
cabin is reduced . When the altitude is decreased and
the barometric pressure is consequently increased at
some altitude, let us say, for example, 18,000 feet, the
air pressure on the outside becomes greater than the
pressure on the inside of the cabin, and since the cabin
is hermetically sealed the walls will be crushed inward.
To provide for this emergency a very simple flap
valve was made as follows: A number of }'2-inch holes
were drilled in the side of the cabin. A flap of sponge
rubber was then_ suspended loosely over these holes. A
slight increase of pressure on the inside of the cabin
would immediately press the sponge rubber t ightly over
the patch of holes, making them completely airtight.
On the other hand, an increase of pressure on the out­side
would cause the flap to hang loosely, so that the
16
conditions of substratosphere flying could be simulated
in the "set-up." However, all the conditions could
not be simulated at the same time. It did not appear
necessary, however, that simultaneous simulation was
necessary . Figures 18, 19. and 20 show schematic
drawings and explanations of the arrangement of the
cabin and its auxiliary apparatus. As noted in the
sketch of figure 18 the air was supplied to the super­charger
(2) by meanR of pipe ( 4). The air could be
supplied with varying degrees of humidity and tem­perature.
The air from the supercharger was carried
by pipe (5) to a Y joint (7). At this Y joint one
branch of the pipe (6) led out of the refrigerated room
into the outer air and the other branch led into the
•:, .. :.·
_:- _:;: .040 Aluminum
·'._:·. structure
. ~ ...
: ·...~ . ~5;
Inside of cabin
Verticc1 { cross-section of di'scharqe valve muffler
FIGURE li.
holes were perfectly free to equalize the pressure on
both sides of the wall.
Silencing airflow noises.-It was found that lining the
insides of the air ducts which were located in the cabin
with felt about one-half inch thick almost completely
silenced the noise of the airflow. The disc harge valve
was found to be quite noisy. A simple muffier, con­structed
as shown in figure 17, would so completely
si lence the valve that it was necessary to hold the ear
close to it to determine when its operation would
start.
Experiments in the refrigerated room of the power plant
branch.- The supercharged Fleetster cabin, after tests
on the operation of the various accessories, was installed
in the refrigerated room of the Power Plant Branch
and connections made to simulate, as nearly as possible,
the conditions of substratosphere fly ing. Most of the
cabin by way of an oil strainer (8) and orifice chamber
(9), through an electric heater, and thence into the air
duct underneath the windows of the pilot's cockpit.
The air could be distributed into the cabin over the
ins ide of the windows, between the double glasses of
the windows, directly into the cabin as noted in figure
27, or under the benches, thence through a hole into
the cabin as noted in figure 25, or distributed partially
over the windows and partially through other routes
into the cabin.
Referring again to figure 18, two butterfly valves on
the same shaft operating in the two branches of the pipe
at (7) could be actuated by means of a lever on the
inside of the cabin. In one extreme position of the
lever, the pipe leading into the cabin would be entirely
closed by one of the butterfly valves, while the pipe
leading outside the refrigerated room would be entirely
(i) Cabin.
@ Supetcharger.
@ Electric motor.
© Condit ioned air supply, 6-inch.
® Compressed air line, 3-inch.
© Bypass line, 3-inch.
0 Control valves.
© Air cleaner.
17
FIGURE 18.- P lan view of cabin in refrigerated room.
tnd view of cabin and az'r s uppl1f system
Scale :- l "=1z" ·
L--~-. ;..1------ , r
1 1
,u,
I 1
11
' '
F IGURE 19.
Ol
® Flow·measuring orifice. @Seat.
@ Air heater.
@ Vacuum line.
@Vacuum tank.
® Dispersing t ube (air duct).
@ Quick-release door.
@) P ressure ind icator.
19
I
I
I
I
I I
~---j- I
1 r---- - L- ._,.....,
11
L ,
------11 Manometer.
@ Discharge line, 3-inch .
@ Pressure regulator.
@ Automatic safety va lve.
® Manual sa fety val ve.
® Entrance to chamber.
@ Obser vers' window. &z'.£ ££,?6 Lagged or insulated line.
FIGURE:.,20.-Legend for cabin in refrigerated room. (See Fig. 18.)
F!GUrtE 2l.-Pressure cabin in refrigerated room: Side view showing door and supercharger lines.
open. In the other extreme position the supercharged
air would be directed into the cabin.
The air leaves the cabin through the valve (14) and
out pipe (13) into drum (12) which may be evacuated
by a large vacuum pump. A flap valve was provided
on the side of this drum so that in case of the non­operation
of the vacuum pump the air would be dis­charged
through holes under the flap directly into the
refrigerated room. This prevented the pressure from
backing up into the cabin. The arrangement of the
various items previously discussed is shown in figures
21 to 31, inclusive.
In figure 21 we note the following items: The door;
the pilot's windows at the extreme right with the
19
orifice determined the volume of air flowing through
the cabin.
Figure 22 shows the Y joint more clearly. At the
extreme right of this picture is the electric air heater.
Just to the left of this air heater may be noted a number
of holes in the fuselage arranged in circles. These
holes permit the discharge of air from the cabin through
a large safety valve located in a cup held airtight
against the surface of the cabin by the circle of screws
visible in the photograph.
Figure 23 shows the general arrangement of the pilot's
windows. At the right of the picture just below the
windows is a pressure gage for the benefit of operating
personnel on the outside of the refrigerated room.
FIGURE 22.-Pressure cabin in refrigerated room : Front % view showing beater and supercharger lines.
supercharged air line entering just below the windows;
The Y joint in the supercharged air line in the center of
the picture with one liue leading outside the refrigerated
room along the floor and another line entering the drum
at the lower right-hand side of the picture. A calibrated
orifice is located where the flange is noted at the center of
this drum. The two rubber tubes leading up and into
the cabin are connected to the two ends of a mecury
manometer. This manometer measured the differential
pressure across the orifice. The black wire leading
from the drum up to the ceiling of the room is one lead
of an electric thermometer. The reading of this
thermometer and the differential pressure across the
Figure 24. shows the same view as figure 23, except
that the quick release door which was used in previous
experiments was removed. This view was taken before
the cabin was heavily insulated.
Figure 25 is a forward view of the viterior of the
cabin. In the center may be noted the quick release
door, used in previous experiments. It was not used
during the experiments in the refrigerated room. At
the top center of the photograph may be noted the
four sections of the pilot's windows. Underneath these
windows may be noted the air duct for rlistributing
air under the windows and into the cabin. It will be
noted that the second window from the left has small
holes which were drilled to play air over the window
plugged with small white plugs of wood. This window,
therefore, during the process of the experiments, was
not heated. Later p ictures will show it covered with
frost on the inside and with ice on the outside. The
window at the left is of single-thickness glass, the inner
th in glass being removed. The air was, therefore,
directed over this window by means of the strip of
sheet metal noted at the bottom of the glass. The two
windows at the right were of double glass as noted in
figure 9. The large safety valve may be noted to the
right of the quick release door. To the r ight of this
safety valve is the pipe conducting the intake air under-
20
on the bench between the two cushions, is a telephone
connected with the outside of the refrigerated room.
Above this telephone may be noted a second, small
safety valve.
Figure 26 shows the cabin windows with an ice forma­tion
on the outside. This ice formation was obtained
by spraying a mist of water over the surfaces. The
temperature at this time was about -45° Fahrenheit.
Figure 27 shows the inside of the windows. This
picture was taken immediately after that of figure 26 .
Note that the windows are heaYily frosted. The super­charger
had been running only a few minutes and the
temperature inside the cabin was still below zero.
FIGURE 23.- Pressure cabin in refrigerated room: Front view sbowing pilot's windows,. quick release door in place.
neath the bench at the side. At the extreme right of
the picture is the mercury manometer from which may
be read the differential pressure across the orifice.
To the left of the quick release door may be noted a
triangular strip. This is the inward pressure safety
valve and consists of sponge rubber one-fourth inch
thick attach d at the top to the fu selage side and
provided at the bottom with a handle. In testing out
the cabin for safe automatic operation to be sure that
none of t he personnel would be injured, one operator
was stationed with his hand on this valve so that he
could immediately release the pressure in the cabin by
pulling up the rubber flap . At the left of this flap
is the manometer for reading the differential pressure
in the cabin. At the extreme left of the pi cture, resting
Note at the lower parts of the glass the froRt has begun
to melt.
Figure 28 is the same as figure 26, except t hat the
photograph was taken about an hour later after the ice
had been melted from the outside and frost had been
melted on the inside by the normal action of the super­charged
air conducted between the double windows.
The window at the right was also clear but this does not
show in the photograph.
Figure 29 shows a rear view in the cabin after it had
been insulated with 1 inch of soft felt and a perforated
partition had been installed immediately aft of the
discharge valve. This insulation and partition were
necessary because the heat dissipated from the cabin
prevented the development of a low temperature in the
refrigerated room. It was also found impossible to
heat the cabin on the inside to a comfortable tem­perature.
Figure 30 shows the front end of the cabin after the
insulation had been installed. Note that the safety
valve at the right is covered with frost. This valve
ceased to operate because of this condition. However,
21
the safety valve at the left continued to operate per­fectly,
since it extended into the cabin far enough so
that it retained a fairly warm temperature. It is readily
obvious, of course, that safety valves should be installed
on the inside of the cabin with a tube connecting them
with the outside. This tube may be 3 or 4 inches in
diameter and about 3 or 4 inches long.
FIGURE 24.-Pressure cabin in refrigerated room: Front view showing pilot's windows, quick release door removed .
Time
TABLE II.-Supercharged cabin in refrigerated room Mar. 10, 1936
[Temperat ure: Degrees fahrenheit. Pressure: Inches mercury]
Cabin
Refriger- ,----------,
ated
room Dry
Relative
Wet humid-ity
Cabin
intake
Intake Cabin
Orifice tempera- differen-ture
tial
Intake Dis­charge
Super­Orifice
1 charge '
(across) (pres-sure/
pound)
---------,---- ---------- ---- ------------ ---- ---- ---- ----
Percerit Iriches
9·30_ -- ---- --- _ -- --- - ---- -44 +22 +64 49 80 8.4 0. 5 -2. 1 2 6
JO'--------------- -- ---- -44 +64 54 51 + 108 95 92 8. 5 +.5 -2. 1 1. 6 6
10:30 ____________________ -44 +73 58 38. 5 + 122 113 96 8. 5 +.5 -2. l 2.0 +6
ll ' - - - - ------ --- - 11 ·30 _____________ -_-__-_--__--_ - 43 +67 57 53. 5 + 128 117 98 8. 4 +.5 -2. l 1.8 6 12_ ___ _____ ______________ - 44 +66 55 48 + 124 I 18 100 8. 0 +.s -2. l 2. 2 +6 -43 +74 59 38. 5 + 132 122 102 8. 3 + . 5 -2.l 2. 2 +6
!1.2 _:3__0 _'_-_-_-_-_-_--__--_-__--_-__-_-_-_-_--_ -43 +68 55 42 + 132 122 102 8. 3 +.5 -2. l 2. 2 +6 1·30 _____________________ -44 +72 58 +131 121 104 8. 3 . 5 -2. l 2. 2 +6 2 ____ ______ ___ __________ - 45 +75 60 +131 122 103 8. 0 . 5 -2. l 2. 3 +6 -46 +75 -------- - --------- +131 122 104 8. 0 . 5 -2.1 2. 2 +6
The conditions obtained were considered very satisfactory.
1 Diameter of orifice = 1 inch. The windows could be kept clear when sprayed with a mist when the entire volume of air into the cabin was
forced over them.
' Approximate volume = 100 cubic foot/minutes. A double glass window arranged so that inner glass may be put in or taken out at will is recom-mended.
This is for cleaning purposes and better vision.
' Started to clear windows at 10:10.
• Out of cabin 11 to 11:05.
' 12:45 opened door.
•
22
•
FmuRE 26.-Pressure cabin in refrigerated room: Front dew showing windows iced and frosted.
Figure 31 shows the ice formation left on the pilot's
windows after a 5-hour run at -45° Fahrenheit in the
refrigerated room. Note that the windows are free
from ice formations but are quite stained. This stain
is between the two glasses. It may be concluded from
this that the inner glass should be hinged so that it
may be lifted from place when not in use, so that the
glasses may be wiped clean.
A number of runs were made on this apparatus before
satisfactory operation could be obtained, and a.number
of runs were made after satisfactory operation was
obtained for the purpose of obtaining useful data on
various conditions. A sample data sheet is shown in
table II.
23
consider the factor from two standpoint~-first, from
the standpoint of elastic expansion, and second, from
the standpoint of ultimate strength.
In theoretical discussions of the structural problems
involved in the design of pressure cabins it appears to
be customary to indicate a fixed-load factor or design
factor. This is generally placed at a figure of two and
one-half times the differential cabin pressure antici­pated.
It appears absurd to use a fixed design factor
such as this for both elastic and strength calculations
in conjunction with the usual stress analysis of the
flying structure. In the first place, there should be no
possibility of the design pressure in the pressure cabin
becoming greater than the basic design load. Safety
FIGURE 27.- Pressure cabin in refrigerated room: Interior front view showing frosted windows.
Experimental results .- On account of the great num­ber
of phases of this investigation the results and con­clusions
may be made clearer by discussing them in the
order as outlined in table I.
DISCUSSION OF INVESTIGATIONS
STR UCTURAL PROBLEMS
Design features. - In considering the structural
analysis and design of a pressure cabin it is gratifying
to know that these features may be more closely
analyzed than in the case of a general analysis and
design of the airplane structure for fl ying conditions.
In arriving at a design factor it appears desirable to
42060- 38-- 3
valves should make this absolutely impossible, even if
it were not for the fact that a slight increase of pressure
would be immediately noticed by the personnel and
could be immediately r elieved by appropriate hand­operated
valves were the safety valves to fail. How­ever,
the safety-valve feature of the design should be
absolutely perfect, even if it were necessary to install
a dozen or more of such valves.
It seems unreasonable, therefore, to assume that a
pressure in the cabin two and one-half times the normal ·
design pressure would ever occur at the same time the
maximum design aerodynamic conditions are being
realized.
It is recommended that the basic design pressure only
be used in connection with the usual design load factors
req uired in the structural design of the airplane.
24
A design pressure factor of three times the basic
differential pressure required in the cabin for the
specified flight conditions is recommended. For
example, if the airplane is to fly at an altitude of 25,0dO
feet with sea-level pressure maintained inside the
cabin, the basic-design differential pressure would be
the difference between the barometric pressure at sea
level and the barometric pressure at 25,000 feet. For
the design of the pressure cabin this difference in pres­sure
should then be multiplied by 3. As mentioned
previously, however, t his factor would not be used in
connection with the general structural design of the
In general for airplane pressure cabins with diameters
not in excess of 6 to 8 feet these load factors would not
require an increase of weight in the structure over the
normal structural weight.
The internal pressure will have a tendency to
strengthen the walls of the fuselage against flying
stresses.
Type of structure.-The simplest and lightest type of
structure is a pressure vessel of the round cylindrical
type with spherical heads. The stresses in the struc­ture
in such a vessel are all tensile stresses, except for
certain discontinuous bending stresses at the junction
of the cylindrical elements and the spherical heads,
generally referred to as discontinuous stresses. If the
FIGURE 28.- Pressure cabin in refrigerated room : Front view showing windows cleared of ice and frost.
airplane. Neither should it be used in calculating the
elastic expansions of th-e structure. A factor of 3, it
appears, should be used solely in the design of the
pressure cabin irrespective of any other stresses to
which it may be subjected.
The load factor of 3 is selected in this case for the
following reason. It is desirable to static-test the
pressure cabin with a proof pressure of twice the basic­design
differential pressure. This proof test should not
produce stresses above the proportional limit of the
material, since for aircraft materials the proportional
limit is generally above two-thirds of the ultimate
strength. A load factor of 3 would produce the mini­mum
desirable conditions.
cylinder has fairly rigid annular bulkheads, these dis­continuous
stresses will also be introduced at the
junction ·of the skin and the bulkhead. These stresses
for pressures contemplated in the pressure-cabin
design, however, will be small as will be pointed out
later in this discussion.
The problem of pressure cabin design requires the
determination, first of the simplest type of structure
not only to carry the differential pressure but to carry
this differential pressure in conjunction with the design .
flying stresses imposed upon the fuselage. Four
possibilities are immediately apparent:
(a) A circular pressure vessel with hemispherical
ends suspended inside of the normal fuselage. The
disadvantages of this type of structure, however, a re
great. Aside from the problem of disadvantage of
great weight, difficulty would be experienced in window
and door design.
(b) The skin on the outside of the fuselage frame, as
in the present standard design, to be reinforced to form
the pressure vessel. This appears to be the most
logical type of construction. The fuselage in this case
should be circular for a minimum weight, and as in a,
t he ends should be as nearly hemispherical as practical.
Due allowance must be made for the expansion of the
shell when subj ected to the differential pressure in
attaching the skin to the fuselage frame. This prob-
25
Airtightness of joints.- For a complete discussion of
this phase of the subject the reader is referred to chapter
III of this report by Maj. D. G. Lingle.
Experiments performed in connection with this in­vestigation
show that the problem of sealing the joints
of the airplane structu re is in general quite simple.
At the present stage of development it appears that
the type of joint now used in flying-boat hulls is quite
adequate, namely, a joint, in which a strip of fabric _
soaked in marine glue is placed between the two edges
of the metal sheets to form the joint. The riveting
pitch, as in t he case of boat hulls, must of necessity be
small. If it appears necessary, the riveting may be
FIGURE 20.-Pressure cabin in refrigerated room: Rear interior view showing discharge valve and felt partition.
!em of expansion is given furth er consideration in this
discussion.
(c) The skin on the outside of the fuselage to be cor­rugated
parallel to the longitudinal axis of the fuselage,
permitting the fuselage bulkheads to carry the pressure
loads. The corrugated sheets would then carry the
pressure loads from bulkhead to bulkhead in bending.
If a flat-sided structure is used, which, of course, is
impractical, it would be necessary to use some scheme
such as this. The idea does not seem practical.
(d) A structure involving a great number of small
stiffeners to operate as an integral unit in resisting the
pressure and flying stresses. This may be considered
as an extreme case of b.
It should be borne in mind that both the elasticity
and the strength of the structure must be taken into
consideration in the design.
d ivided into two classes : (1) The larger strength rivets,
and (2) t he smaller sealing rivets; that is, the larger
rivets to give the structure ample strength and the
small r ivets, closely spaced, to give the proper air­tightness.
In case that the finished joint leaks air
slightly, a coat or two of marine glue, or other bitu­mastic
nondrying paints, applied along the riveted
seam will make the joint more satisfactory. Other
types of riveted joints as well as welded joints are still
under investigation by the division and more accurate
information will be in time made available. It is not
a question, however, of finding a satisfactory joint
because the above mentioned joint is satisfactory.
The problem is one of finding th~ most satisfactory
joint.
It should be noted in this connection that such a
bitumastic impregnated joint will naturally be quite
resistant to vibration stresses which may have a ten­dency
to cause a more rigid joint, such as a welded
one, to crack.
26
The problem of making doors, windows, etc., air­tight
is discussed under the heading of "Mechanical
Problems."
Expansion of skin under pressure.-The problem of
expansion of the fuselage skin under pressure is not a
serious one and in most cases may be neglected. It
should be noted, however, that all such expansion
problems are sμb ject with great accuracy to determina­tion
by the ordinary theory of elasticity, since the
stresses for such calculations will not be above the
proportional limit of the material. There appears to
spherical structures are two of the most serious elastic
problems with which the designer must cope. The
stresses involved in these connections are called "dis­continuous
stresses" and the reader may find solutions
of such problems in standard books on elasticity under
this heading and in chapter IV.
While an exact determination of these stresses may
be obtained, the mathematics involved is not simple.
An approximate consideration of this problem, how­ever,
will serve to point out that the stresses involved
are not serious. Let us assume, for example, a differ­ential
pressure of 10 pounds per square inch and let us
assume the extreme case of perfect rigidity of the
bulkheads. If the bulkheads are spaced at intervals
FIGURE 30.- Pressure cabin in refrigerated room: Interior front view showing insulation .
be no reason, therefore, for not carrying out a complete
and accurate stress analysis of all elasticity problems
connected with the expansion of the structure. As
previously pointed out in this discussion, the calcula­tions
for the elastic distortion of the structure should
be based upon a pressure load not greater than the
basic-design pressure load. This is true because there
should be no possibility of the pressure in the cabin
rising higher than this value, and even if by a remote
possibility the pressure should rise higher than the
basic normal pre~sure, the operators would certainly
not permit this condition to exist longer than a few
seconds. The problem of the semirigid bulkheads and
the problem of the change from cylindrical to hemi-of
20 inches, which may be considered as an extreme
case, the total circumferential pressure which may
effect any one bulkhead would be 200 pounds per inch.
Under the assumption of absolute rigidity of the bulk­head
the skin immediately adjacent to the bulkhead is
under no stress. The stress in the skin on either side
of the bulkhead is tensile and increases to a maximum
half way between the bulkheads. Theory shows that
the stresses approach this maximum at a short distance
on either side of the bulkhead, less than 3 or 4 inches in
general. However, in an extreme case the radial load
carried by the bulkhead would not be over one-half the
total load of 200 pounds. However, since bulkheads
are not rigid they would expand further, reducing the
load on the bulkheads which is carried from the skin to
the bulkhead by the rivets. The load on the bulkhead
then would certainly be less than 100 pounds per inch,
and since the most extreme cases have been assumed,
it would be safe to say that this load will not amount
to over 50 pounds per inch. If two rows of rivets are
used and spaced at a pitch of 1 inch, this would mean
a tensile load on the head of each rivet of 25 pounds.
Now bear in mind that the actual stress will certainly
be less than this, which at even twice the stress is not a
serious matter.
It is advisable, however, to always take this discon­tinuity
stress into consideration, and if the design of
27
vessel obviously will not be greater than the discontin­uity
stresses at the bulkhead under consideration.
The effect of the expansion of the skin on the main
fuselage-wing connections may obviously be neglected,
since, in comparison to the heavy members involved in
this connection, the thin sheet metal used in the con­struction
of the pressure cabin would have no appre­ciable
effect. The effect of the expansion of the
fuselage on the operation of the controls should be
taken into consideration in the analysis, but in general
it will be found that if any stresses at all are involved
they will be negligible. If these stresses do exist it can
be shown that they may be offset by the temperature
FIGURE 31.- Pressure cabin in refrigerated room: Front view showing windows cleared of coat of ice.
bulkheads permits, allowance should be made for this
expansion. The weakening of the structure by the use
of expansion joints in the bulkheads is certainly in­advisable.
It is better to design the bulkhead in such
a way that it has a certain degree of flexibility to allow
for this expansion; for example, figure 32 shows a simple
design for this purpose. In figure 32 (a) the bulkhead
is an extruded I-beam. The skin is attached to this
bulkhead by means of the relatively thin sheet-metal
adapter, shaped to allow for the expansion of the skin
without exceeding the proportional limit of the material.
Figure 32 (b) shows the ordinary hat section with the
brim of the hat widened to allow the proper flexibility.
The discontinuity stress involved at the connection of
the cylindrical and semispherical sections of the pressure
stresses. This is obvious when it is borne in mind that
as altitude is attained and the differential pressure is
being built up the temperature is dropping, hence con­tracting
the airplane structure. The reader is referred
to chapter IV of this report by Mr. G. D. Bogert for
sample calculations on these conditions.
Expansion of skin under pressure.-An unsymmetrical
structure, such as a pilot's cockpit of the conventional
form or a flat-sided fuselage, deserves special considera­tion
in connection with the problem of distortion.
Such problems may in these cases be quite serious. In
the experiments carried out on the Fleetster cabin with
the irregular pilot's ,vindow structure considerable
distortional effects were noted; for example, when this
part of the structure was covered with ice an introduction
of air pressure caused excessive structural noises, in­dicating
the breakage of ice formation due to the dis­tortion.
It would probably be simpler to build a more
symmetrical structure than it would be to analyze the
distortion for the irregular structure.
Probability of explosion.-An important question which
naturally presents itself to one for consideration is the
possibility of an explosion of the supercharged cabin;
for example, when a fatigue crack in the skin develops
or when the structure is punctured. It appears that
this problem should give no great concern. This is
especially true if the cabin is designed for a load factor as
high as three, which means that the pressure stresses are
only one-third of the ultimate strength of t he structure.
An explosion results when the stresses are near the
ultimate strength, so that when the structure is slightly
weakened the stresses in the adjacent structure imme-
Bulkhea.d
(d)
Bulkhea.d
(b)
28
Bu:lkhead.s designed. to a.Clow for expans/.on of skin.
FIGURE 32.
diately become higher than the ultimate strength of the
structure. In general, this condition would be re­motely
improbable when the pressure stresses are as low
as one-third of the ultimate strength. Any failure
would simply result in a quick loss of pressure which
may further rupture the opening but which certainly
would not develop into an explosion which would
damage the flying structure of the airplane. This is
especially true, as an airplane pressure cabin would be
designed, because the structural members such as
bulkheads, stringers, etc., would limit the disintegration
of the skin.
Bulkheads.-The hemispherical shell is, of course, the
ideal type of bulkhead. This type of structure, how-ever,
offers the objection of being difficult to form.
The hemispherical shell also would occupy valuable
space in certain cases.
The ideal type of restraining bulkhead, where space
is not an important question from a structural stand­point,
is the cone.
Any flat surface used as a pressure restraining wall is,
of course, quite inefficient. In cases where it is neces­sary
to use a flat restraining bulkhead the bulkhead
may be designed by the usual conventional methods.
The designer may find, however, that a truncated cone
bulkhead may be designed for the same space which
would be occupied by the flat bulkhead yet would be
much lighter. In the design of such a bulkhead the
Crank
Ca.bin. W<ill
A,rt,ght p,ckin9
/jeo.r~ng
Air-proof .sponge rubber
,. (Held ti9ht1y in place)
Crank
Air-tight forsi'on gl.,nd
Air-tight sliding gland
FIGURE 33.
Brass tube
Control fube
an9Jysis of the outer supporting ring of the cone would
have to be given careful consideration. The methods
involved, however, are conventional and may be found
in chapter IV of this report.
Strength of glass windows.-The reader is referred to
chapter V of this report for the results of the research
on glass windows suitable for pressure cabins.
There seems to be at least a popular conception that
the windows for this type of structure should be round;
at least most artists' sketches for proposed stratosphere
airplanes have round windows. In the last analysis
there does not appear to be any obvious reason why
round windows shonld be used and they certainly
would not be as easily installed and as convenient as
rectangular windows. It appears that narrow rectan­gular
windows, arranged longitudinally with respect to
the axis of the pressure cabin in long narrow strips, are
the most desirable fron a structural and a practical
standpoint. For example, the indi vidual windows may
be 6 inches high and 12 inches long, spaced with an
interval for structural strength of about 3 or 4 inches.
Visibi lity for the pilot.- The visibility for the pilot
should not be sacrificed in the design of a p ressure cabin.
There is no need for such a sacrifice because the struc­tural
problems involved are fairly simple and the weight
of the structure is not very greatly increased by afford­ing
sufficient visibility.
MECHANICAL PROBLEMS
Doors, emergency exits, window frames, and airtight­ness.-
The reader is referred to the disc ussion of the
quick-release door and t he general-use door as used on
the Fleetster fuselage experiments. Doors and ex its
built on the principles involved in the construction of
these openings, properly refined for
practical use, will be quite satisfac-tory.
There a re, of course, many
other ways of solving the same prob­lem,
any of them being equally as
simple.
There is one question concerning
these openings on which many are
in disagreement; that is, should the
doors be made to open inward or out­ward?
It appears that the logical
answer for this, especially in connec­tion
with the emergency exits, is to
make two exits, one opening inward
and one opening outward. The
reason for this is as follows: The exit ~
opening outward requires certain ~
fairly intricate operating mechanisms. c5
In the case of a crash these mecha-nisms
may be readily di sarrangerl
so that it would be impossible to
open the door. However, in case
that it is necessary to take to the
parachute at a high altitude, then this type of emergency
door would be ideal because a simple twist would permit
the door to fly off and entirely away from the airplane.
On the other hand, the emergency door which opens
inward would require no mechanism at all for its
operation and need only be held in place, loosely, against
a gasket of soft rubber by small spring cl ips. In case of
a crash, in which the door area is distorted, the door
would not only be easily removed but in all probability it
would fly from its seat from the impact. Of course it
is obvious that immediately upon impact the air in the
cabin would be released through a dozen sources; for
example, through broken windows, seams, etc., the time
required certainly not being greater than one-half
second. The door which opens inward, of course,
could not be opened at all, in the a ir, until the pressure
in the cabin is reduced to a very low value. A valve,
however, could be designed in the door with an area of
4 or 5 square inches which would reduce the pressure to
a proper limit so that the door could be opened in, for
example, 2 or 3 seconds.
These arguments also apply to doors which open in­ward
or outward. These in vestigations, however, have
29
shown that whether it is decided to open the doors in­ward
or outward, the structural and mechanical problems
are quite simple and the resulting structure should not
be much heavier than such structures as are now used.
If the doors or emergency exits open inward, the air­sealing
problem is extremely simple since a soft airtight
rubber gasket around the door of about ),~-inch thick­ness
is all that is necessary for sealing. The pressure of
the air holds the door so firmly against this gasket that
all air leakage can readily be prevented.
The glass windows, of course, may be sealed in the
same manner. It is advisable in the case of the windo,, s
to separate the glass entirely away from the metal
structure. The structure which is designed to hold the
glass in place should not bear too heavily on the soft­rubber
gaskets. This air sealing should be left to the
pressure of the air on the glass itself. The purpose of
Ca.bi.n wa.lt
Inside
.IJellowS' tqpe gla.n.d
FIGURE 34.
this is to prevent distortional stresses from being trans­rnitted
to the glass.
Soft sponge rubber only should be used as this allows
a more even distribution of the stresses in the glass.
For example, if the metal window-frame structure is
distorted, extremely high concentrated stresses may be
introduced in the glass due to t his distortion. Properly
designed windows and doors should be perfectly air­tight.
This problem has proven to be quite simple.
Glands for control wires, t1,bes, etc., to control si,rfaces
and engines .- The principal problem is obviously that
of hermetically sealing the exits of the control wires,
tubes, etc., from the interior of the cabin to the exterior.
Two types of motion are involved, rotation and trans­lation
(or sliding). The sealing of a gland for a torsion
shaft, of course, is extremely simple; for example, figure
33 (a) shows the extremely simple gland used in the
Fleetster body in connection with the supercharger
butterfly valve control. Figure 33 (b) shows a gland
developed in connection with the Fleetster fuselage
experiments which proved satisfactory for either rota­tion
or sliding even at very low temperatures when
sprayed by a water mist so as to completely cover the
gland with ice. No appreciable resistance to its opera­tion
was noted due to the low t emperature, even though
the entire gland was covered with frost which may be
easily prevented if the gland is properly designed. For
example, the felt layer on the inside of the gland should
havE) been deeper so that the temperature of the gland
would be maintained more nearly at the temperature of
the cabin.
30
There are, of course, many other ways of constructing
such glands ; for example, a bellows arrangement as
shown in figure 34 may be used. Such a gland was
designed for use in these investigations but was not con­structed,
the reason being that the above-mentioned
design is so simple, light, and safe that it would not
seem desirable to investigate the problem further. The
bellows type of gland has at least one very serious objec­tion.
This objection may be explained as follows. In
order that the gland operate properly, ample provision
must be made for an easy access of air into and out of
t he bellows. Hence the hole through the cabin wall
must be much larger than the rod or tube passing through
the hole or additional holes must be drilled as noted in
the figure. Therefore, if the sylphon bellows should
develop a fatigue crack or become ruptured in some other
manner, a serious leakage of air would develop. This
type of gland also has the objection that considerable
space and weight is necessary for its inclusion in the
design.
Control of icing and f rosting of windows.- Tbe basic
principle involved in this feature of the design is to keep
the window glass warm enough to melt the ice or frost .
This may he accomplished in a number of ways. The
warm air from the supercharger may be directed over
the inside of the glass or it may be passed between the
two glasses of a double window, as in the case of the
investigations previously discussed. It is obvious that
other sou rces of heat may be used other than from the
intake air from the supercharger. For example, the air
in the cabin may be picked up by a small blower,
heated, and directed against the window glass.
In the experiments performed in the refrigerated
room of the Power Plant Branch the glass windows
could be kept clear under the severe conditions of - 45°
Fahrenheit temperature, while wa ter was being sprayed
on the outside of the window. One could, of course,
not find water in the liquid state at - 45° Fahrenheit.
Except for keeping windows clear of frost on the inside,
it is probable that the window-clearing devices will be
mostly necessary at altitudes much lower than 25,000
to 30,000 feet.
In these experiments a t this low t emperature ap­proximately
100 cubic feet of air per minute at a tem­perature
of 130° Fahrenheit was directed over the
windows. This amount of air, however, was necessary
only when ice was being melted on the outside of the
windows. It required very little heat and very little
air vol'ume to keep the windows clear of frost on the
inside, probably about 15 to 20 cubic feet per minute.
Frosting of exposed metals on the inside of the pressure
cabin-Heat insulation.- During the experimental in­vestigations
carried out in the refrigerated room of the
Power Plant Branch, the metal surfaces of the skin and
accessories quickly became· excessively frosted. In a
period of 4 or 5 hours this fros t in some cases attained
a thi ckness of one-eighth to one-fourth inch. It was
noted that as the cabin became warmer much of the
frost melted, causing water to form and drip from the
metal surfaces. It was found that in the cases where
the felt insulation was pressed firmly against the cabin
wall there was no tendency to form frost . However,
in case the insulation did not press firmly against the
wall the formation of the frost was more excessive than
if the insulation were not present. These experiments
led to the following conclusions : The insulation of the
cabin should be glued firmly to the outer skin with a
glue which will not det eriorate or crack at very low
t emperatures, proba bly a bitumastic type of liquid will
serve for this purpose.
The insulation of the pressure-cabin walls should re­ceive
the attention of insulation engineers in connection
with the design of commercial jobs. The experiments
in the refrigerated room, insofar as this problem is con­cerned,
gave very little quantitative information con­cerning
the insulation necessary for the flight article.
The air conditions in this case were static while in con­nection
with the flight article the air conditions will be
aerodynamic, and the dissipation of heat from the cabin
will undoubtedly be much greater, due to the rapid
movement of the pressure cabin through the cold air.
It appears, however, that a scheme somewhat as follows
is desirable in insulating the cabin: First, glue a layer
of soft felt to the interior of the cabin skin by means of
an appropriate glue. On top of this felt, add the neces­sary
thickness of a lightweight heat insulator, as, for
example, dry-zero or blankets of kapok. It appears
that an experimental ship should be equipped with at
least approximately 1 inch of insulation.
Accessories connected to the cabin wall, such as
safety valves, control wire glands, etc., should be placed
well inside the cabin and insulated from the outside aR
much as possible to prevent frosting.
Automatic window flaps may be installed to hold the
air pressure in case of breakage of the windows, as
noted in connection with the description of experiments
on window flaps. (See fi g. 5 and description thereof.)
It is possible to design such a flap so that it will be quite
effective and inconspicuous even in a passenger pressure
cabin. However, t here is a question as to whether it is
des irable to provide windows with such flaps. On the
basis of experimental investigations such as one may
find in chapter V of this report , it is obvious that the
strength of a glass window can be det ermined as accu­rately
as the strength of a ny other part of the structure.
There appears to be no reason why a window should
fail any more than any other part of the structure
should fail. In a military airplane, of course, a window
m·ay be broken by a shell and hence this may be an
argument for a window flap. As a substitute for the
use of a flap, however, in this case it may be sufficiently
effective to carry two or three metal plates, padded on
one side with an inch or so of soft airproof sponge
rubber, which may be placed over the windows in case
of breakage. This maneuver, however, could not be
performed before the pressure is reduced to outside
atmospheric pressure in the cabin, since this takes place
in much less than a second. It seems probable, how-
ever, and experiments no\Y being conducted by Captain
Armstrong will show whether a few seconds at the re­duced
· atmospheric pressure ,Yill incapacitate the per­sonnel
sufficiently to prevent them from operating the
emergency flaps.
In a passenger airplane no need for the flaps is
apparent.
Utilization of dynamic air pressure and energy of air
discharged from the cabin.-Neither one of these ideas
appear to be practical, at least at the present stage of
development of the project. The dynamic a ir pressure
even at a speed of 300 miles per hour is only a small per­centage
of the cabin differential pressure required and
hence cannot be used to effectively assist in building up
the cabin pressure.
31
The kinetic energy of the a ir being discha rged from the
cabin is too small to be taken into consideration. The
question of using the compressed air in the super­charged
cabin for the engines is being considered.
However, the saving at most is very small and may not
be worth the additional weight of piping and apparatus
to make this practical. In general, the amount of air
required for a commercial supercharged cabin will be less
than one-tenth of the air required for the engines.
E.ffect of supercharging on instruments.- Instruments
which are affected by the pressure on the inside of the
cabin may be placed in a scaled compartment connected
with the outside atmosphere.
AIR FLOW AND REGULATION PROBLEM
Automatic and sensitive regulation of quantity and pres­sure,
f ree f rom noise and danger of f reezing at altitudes.­The
reader is referred to chapter VI of this report for a
complete discussion of this problem by Mr. R. R. Curtis.
Reference is also made to the description of the discharge
valve presented earlier in this report.
The valve tested in connection 'idth the refrigerated
chamber experiments had no means of regulating the
quantity of air but all the ot her r equirements were
satisfactory. The valve for pressures of 3 or 4 pounds
per square inch and above was unaffected by an increase
in differential pressure across t he valve. Different
simulations of a change in altitude of a bout 10,000 feet
at a constant simulated altit11dc, t he valve held an
absolutely nonfluctuating pressure. The muffl er shown
in figure 17 and described in adjacent paragraphs pro,·cd
perfectly effective in silencing the noise. The total
weight of the muffler was less than 1 pound. The
location of a discharge valve, as previo usly descri bed, on
the interior of the cabin absolutely prevents any chance
of freezing.
Regulation of the quantity of air flowing into the
cabin from the supercharging unit is necessary. If the
unit is separate from the engines, the regulation may be
affected by regulating the speed of the supercharger.
However, this would not be p ossible if the superchargers
were operated by the airplane motors. It would be
necessary in th is case to either regulate the amount of
intake air by bypass valves or allow the full capacity of
air from the superchargers to flow into the cabin at all
times and make proper adjustments in the discharge
valve mechanism to provide for changing volumes.
This would have Lo he done manually, or else automatic
regulating devices must be designed. (See ch. VI.)
The regulation of the quantity of a ir does not seem to
be a serious problem, except in sofar as a minimum
supply should always be provided. The discharge valve
should have sufficient capacity to properly care for the
entire maximum discharge of the superchargers into the
cabin without appreciable rise in pressure.
Automatic sealing of cabin and release of oxygen spray
in case of failure of superchargers.-The possibility of the
failure of the superchargers supplying the air to the pres­snre
cabin should be provid ed for. Effective check
valves should be installed in the line to prevent the
reverse flow of air out of the cabin into the super­chargers
; and in the case of a loss of pressure in the
cabin, provisions should be made for the automatic
spraying of oxygen from emergency tanks provided for
the purpose. The valves of these tanks may be operated
by a sylphon bellows in such a way that for a specified
equivalent altitude in the cabin the bellows trips a
spring valve of the oxygen container, allowing the proper
spray of oxygen into the compartment. For example,
several of these oxygen bottles may be installed in the
pressure cabin and the valve set so that the first bottle
will be opened at a pressure equivalent to 16,000 feet,
the second bottle will be opened at a pressure equivalent
to 17,000 feet, etc.
Safety valves .- The reader is referred to figure 6 of this
chapter and the adjacent descriptive paragraphs. The
valve therein illustrated proved to be quite effective
and simple in construction.
Since the valves are extremely light, it appears
advisable to use at least two of them on each pressure
cabin, each having a volume capacity greater than the
t otal volume capacity of the superchargers supplying
a ir to the cabin. The experiments show that the valves
should be placed 3 or 4 inches on the inside of the cabin
to prevent them from becoming inoperative because of
frostir> g.
PHYS10LOGICAL REQUIREMENTS
Air s upply.- The reader is referred to chapter VII of
this report by Capt. H . G. Armstrong and chapter VIII
by Mr. L . D. Bonham for a detailed treatment of this
subiect.
In general, it should be noted that the air supply
necessary for schoolroom conditions is 20 to 30 cubic
feet per minute. Tt appears probable that the required
air supply for a pressure cabin may be set at approxi­mately
10 cubic fee t per minute although it appears that
this amoun t is not absolutely necessary.
It may be possible that the personnel requirements of
air will not be the criterion for the supercharger capacity.
If the pressure cabin is to be heated through the
medium of the 'ntake air, it may be desirable to in­crease
the quantity rather than heat the air to a high
temperature in order to obtain the required number of
heat units for cabin heating. It appears desirable that
the cabin should be heated from this source, since this
makes available the heat supply for keeping the windows
clear of frost and ice. These, however, are some of the
problems which will have to be worked out thoroughly in
an experimental airplane designed for this purpose, as
the conditions cannot be very well · simulated in a
laboratory.
Pressure required.-In a passenger-carrying airplane
it is evident that sea-level pressure should be maintained
in the cabin for t he comfort of the passengers. In the
case of mili tary airplanes, however, t he personnel will
be in the proper p hysical condit ion for a lower pressure
in the pressure cabin. For example, it may be desir­able
to maintain a pressure in the cabin eq uivalent to
12,000 feet altitude. The cabin pressure t hen would
be the difference between t he barometric pressure at
the alt itude of flight and t he equivalent altitude main­tained
in the cabi n ; for example, if the altitude of flight
is to be 25,000 feet, the differential press ure on the
cabin walls would be the difference between the baro­metric
pressure at 25,000 feet and the barometric
pressure at 12,000 feet. This is approximately 3.6
pounds per square inch.
There are oth er t hings, however, to be taken into
consideration and, in general, if the weight of the struc­t
ure is not appreciably increased by a higher differen­tial
cabin wall press•.1re, the higher cabin wall pressure
is desirable because it gives a greater amount of air in
the cabin in t he case of an emergency for the use of the
personnel. For pressure cabins less than .5 or 6 feet
in diameter, t he differential design pressure may not
be t he criterion in the design of the cabin walls. The
flying stresses may dictate a thicker skin than is dic­tated
by the pressure requirements. In tl1is case it
appears desi rable to design for a sea-level pressure in­side
of the press ure ca!-J.in. An experimental a irplane
designed for sub-stratosphere pressure cabin experi­ments
should, of course, be designed for a sea-level
pressure in t he cabin.
Rate of discharge of air permissible in an emer­gency.-
As noted previously, in t he case of person­nel
whose ears are in good physical condition, there
appears to be no limit to t he rate of discharge of a ir
permissible. F urther experiments, however, may con­t
radict t his finding for la rger differential pressures or
for higher a ltitudes. However, it appears probable
that in the case of an emergency in the air the probable
effects of the sudden release of press ure in t he cabin
need not be given serious consideration .
Oxygen 8upp/y.-In the case of a pressure failure in
the pressure cabin at high altitudes oxygen supply
should be available for the operating personnel anu
the passengers. The release of t his oxygen supply
should be automatic. The automatic control is a sub­ject
for further experimentation, but it appears that
the simple device operated by t he pressure on a sylphon
bellows would be satiR[actor y. The sylp hon bellows
32
mechanism could be set to trip the oxygen control valve
to permit the proper supply of oxygen.
Contaminated air.-Care must be used to see that the
a ir di scharged into the p ressure cabin is not contami­nated;
for example, because of the extremely high speed
of a supercharger vane the oil from the supercharg•:r
may become mi xed with the air in t he form of a fine
mist. This must be either prevented or else the air
must be cleansed before it enters the cabin. In the
case of the investigations discussed above the air from
the supercharger was forced through one-half inch of
soft felt before it entered the cabin. However, this
did not keep the air clean and considerable mist from
the oil was injected into t he cabin.
Conditioning of the uir.-The reader is referred to
chap ter VIII of this report by Mr. L. D. Bonham for
a discussion of this phase of t he problem. Standard
methods are available for conditioning t he air. It ap­pears
only necessary to obtain the basic data for the
design of proper conditioning units. These data can­not
be obtained entirely under laboratory conditions
but must be obtained in the final analysis from actual
flight conditions. The author has given this problem
further consideration in connection with chapter X
in the consideration of " Design Req uirements for a
Pressure Cabin Airplane Laboratory and Training
Ship."
SUPERCHARGING REQUIREMENTS
The reader is referred to chapter I X of t his report by
Mr. A. L. Berger for a detailed discussion of this
subject.
There seems to be two schools of thought in this con­nection,
one is in favor of a supercharger unit entirely
separate from the power plant units. The super­charger
of this unit may be operated by a small aux­iliary
electric motor or a small gasoline motor. The
other is in favor of a supercharger uni t operating para]
lel with the engine supercharger units. It appears to
the author that the latter method will req uire less
weight, and since the units will be as reliable as the
motors themselves, it appears that th is method is more
desirable from a standpoint of reli ability. A separate
auxilia ry unit simply means another motor which may
cause tro uble.
It appears that the air supply required for a super­charged
cabin is so small in comparison to the other
supercharging requirements that t he problem of con­serving
the energy of the supercharged air in the pres­sure
cabin is not of great importance, at least at the
present stage of development of the project. Later
refinements may require a conser vation of this energy.
Chapter II !.- Development and Processes for Construction and Sealing of All-Metal Fuselages
and Compartments for Supercharged, High-Altitude Airplanes
(Prepared by Maj. D. G. Lingle, A.C.)
OBJECT
Development and processes for construction and
sealing of all-metal fuselages and compartments for
supercharged, high-altitude airplanes.
DISCUSSION
In accordance with approved Expenditure Order
No. 417-2-358, issued April 29, 1935, preliminary work
inch material with the dimensions as given below.
This tank was built with slightly concaved heads to
determine whether the metal would tear or the seams
fail under pressure when the pressure was applied on
the heads.
Method of construction.-The longitudinal seam, a lap
joint was spot welded with five spots per square inch
and at points "A" and " B," notch joints were con­structed
and torch welded. The concaved heads were
FIGURE 35.-T ank X at left; Tank Y at right of figure. Tank heads YY at right and Z at left. A at hottom and Bat lop of Tank X.
was started on or about May 1, 1935. Before that
date, very little information was available as to satis­factory
Il'eans of construction and sealing of joints to
withstand unequalized pressures that will result within
the airplane fuselage and compartments for high-alti­tude
operation. This problem was attacked as follows:
PROJECT I
seam welded in place with over-lapping, spots forming
an air-tight joint. The spot welded seam was coated
with a .020-inch layer of aluminum by using a metal
spray gun.
The tank was submerged in a tank of water and in­flated
with air to 5 pounds per square inch. No leaks
occurred.
Air pressure was again applied and the tank failed at
(a) Tank " X," figure No. 35, was constructed of 18 pounds per sq uare inch. The concaved heads were
aluminum alloy, A. C. Specification QQ-A-353, 0.050- forced outward as shown in figure 35, "X," and the
(33)
metal torn at the abrupt bend near the scam. Neither
the spot-welded or seam-welded joints failed.
(b) A second tank was constructed of the same ma­terial
and the same dimensions with the following
exceptions:
The heads were reinforced with discs of 0.050-inch
material, A. C. Specification QQ-A-353 to fit the inside
34
(c) A third tank (tank "Y," figs. 35 and 36), was
constructed in the same manner as tanks "X" with the
following exceptions:
Material for the one concaved end, 0.090 inch, A. C.
Specification QQ-A-353 and the cylindrical walls of
0.050-inch material, A. C. Specification QQ-A-353.
The second end was of a conical shape, with a spot-
FIGURE 36.- T ank z.
of the tank and then spot welded to the heads, thus giv­ing
added strength to the heads except at the seams.
These seams were identical with tank "X."
The inside of the tank was coated with Rubberoid
paint instead of metalizing the spot-welded seam as in
tank "X." The paint was allowed to dry, and the tank
then tested with 5 pounds of air pressure for leaks. No
leaks occurred.
Air pressure was applied and the tank failed at 47
pounds per square inch. The 0.050-inch metal was
torn loose at the seam weld as shown in "YY."
welded lapped seam as shown in figure 35 and of 0.020-
inch material, A. C. Specification 11072. This head
was seam welded in place and the tank tested with 5
pounds air pressure for leaks. No leaks occurred. It
was then inflated with air pressure and failed at 84
pounds per square inch. The failure started at the
seam weld, figure 36 (Material Branch photograph), and
the head was torn from the tank with an explosive force.
(d) Tank dimensions:
Tank "X," 12 inches diameter, 21 inches long, depth
of concaved heads~4 inches.
35
Tank "Y," 12 inches diameter, 21 inches long, depth
of concaved heads 4 inches.
Upon completion, it will be loaded, vibrated, and
static-tested. These results will be published at a later
Tank "Z," 12 inches diameter, 16 inches long, depth date.
of concaved head 4 inches, base to apex of conical head PROJECT IV
8 incnes.
PROJECT II
The sealing processes were employed to seal the
riveted joints in the fuselage of a C-11 airplane. The
paint was removed from the fuselage constructed of
aluminum alloy, A. C. Specification QQ-A-353. All
openings were closed by riveting plates of the
same material, 0.050 inch thick in place of the
openings. A heavy coat of Rubberoid paint
was applied at the joints before riveting. The
longitudinal seams were reinforced with addi­tional
rivets. The ends were sealed as shown
in figures 35, 36, 37, and 38, chapter JI, by
using soft rubber gaskets and a heavy coat of
Rubberoid paint and a wooden retaining ring
held in place by a series of bolts through the
bulkhead rings.
All the interior surfaces were coated with a
heavy coat of the above paint and allowed
to dry.
The front opening of the fuselage was closed
with a steel door, locked in place, by a series
of radial plungers. These plungers were bev­eled
at the locking joints and pivoted to a
revolving disc at center of the door. This disc
was revolved by means of an external handle,
thus forcing the tapered ends of the plungers
against the inner surface of the steel ring at
the openings of the conical section . A soft
rubber ring gasket was compressed between
the door and the retaining ring forming an air­tight
joint.
After completely closing the entire fuselage
compartment, air pressure was applied and
controlled by means of an air control valve (1 ),
figure 1, chapter II, and a pre-calibrated relief
valve (2). The fuselage withstood a press ure
of 10 pounds per square inch as regist ered on
the air gauge (3). No leaks occurred at any
of riveted or bolted joints.
A series of experiments and test s, discussed in other
reports, were conducted inside the sealed fuselage.
These tests extended over a period of approximately
2 months without any serious leaks at any of the sealed
seams or joints.
The fuselage was placed in pressure chamber at a
temperature of minus 40° Fahrenheit. It was found
that sealing paint hardened, became brittle, and flaked
l oose.
PROJECT III
The compartment cylinder assembly (drawing
X- 36N166 and det a iled drawings thereon) is now
under construction and is approximately 75 percent
complete.
The purpose of this project is to determine the
possi bilities of high resistance welding of aluminum
alloys and satisfactory bulkhead design and constru c­tion
for high altitude, supercharged compartments.
To determine the most satisfactory method of con­structing
and sealing riveted joints and seams.
A tank was built in accordance with figures 37 and
38. The tank was inflated with 5 pounds of air pressure
and tested for leaks. The joint at section D. D.,
A
A
11 ----- ---- 40 · -------+--15"
66 "-----------
S ec:fron A-A
Jo"Alum.inum. alloq pressure fa.nk
See F'ig. 38 for :section views of f oints.
FIGURE 37.
sealed with fabric and white lead, was the only section
withstanding the pressure without leaks.
After these tests, the interior of the tank was painted
with a thick coat of sealing compound, A. C. Specifica­tion
2- 86. This process completely sealed all joints
but the effects of abrupt temperat ure changes on the
compouud has not been determined to date.
CONCLUSIONS AND RECOMMENDATIONS
A leak-proof compartment or fuselage, of monocoque
construction, can be built for supercharging and high­altitude
operation with slight increase in weight.
These compartments or fuselages can be satisfactorily
sealed to maintain a constant pressure within them.
Further research and development work should be
conducted in connection with the development of seal ­ing
compounds to withstand abrupt temperature and
moisture changes without failures.
36
Further development work should be carried on to
develop a satisfactorily mechanically sealed seam or
joint with aluminum alloys.
Investigations in connection with various types of
rubbers and moisture-proof material s, used for sea ling
gaskets, should be continued so that a material may
be obtained that will r etain its sealing qualities when
subjected to pressure, moisture, and temperature
changes over a period of t ime.
.3 ROWS Ya' RIVE.TS AT % "
FABRIC AND RU5BER Ci:MENT
vrew A
,032.
7"
3Z
/"
l ll -:3-----
4
5"
16
.o3Z"
.2 ROWS t0
RIVET5 AT ~-(OUTER)
:z ROWS~~· RIVETS Al" o/e"6~NeR)
FABRIC AND MARlNE GLUE•
VI~W B
" .o3Z
~ .. -+ M~ ~ u;·..,,..
, Row Ya' RIVETS AT %-~ ..
I ROW ALTERNATE ~·' 8 ~l, RIVETS
AT 3/6" FABRIC IMPREGNATED WITH
DOPE• VIEW C
..
.03~
2 ROWS Ye R1vlrs AT 3/4- (ouTE~)
l ROW !-1li' RIVETS AT ~II (CENTER)
FASJttC ANP MARINE GLUE,
VIEW :0
1f1 ~~" f' 1~~
S" • ,t: .. t 15'.1'Y . os .OSS ,03~ .OS.~.
Z ROWS ~"RIVETS AT o/5' f ROWS Ys'' RIVETS AT %.
O·~t>'' ..
• v~ .o5'S'
FABRIC AND MARINE. GLUE., FABRIC AND RUBBER
z Rows Ye~ Rivers AT 7f FABRIC N/0 IMfflTf.
CEMENT. LEAD,
StC'l'tOl'h B-:B ,sEC'rlON C-C $1:(!'rLON D-D
,,
~
l/;1 ~ .oss~
• O'l-0" .sreEL
z i=tows l/a" RIVETS AT ,"
FABRIC ANO MARIME GLUf,
$EC"tloN E-E
----1-l",
7'' f' ${.
jfi V
.oss"
2 ROWS Ya" R IVfTS AT I''
fABJIC. AHO RU88'ER CEMENT.
.$EC'l'I0li ~ • ~
TE~T JOINTS FOR ·30" PRE.SSORE TAN,C.
SttOWING LOCATION OF Rl"E.T.S A~D COMPOSITfQN Of ~EAl­ALL
MAiE.RfAL IS ALUM, ALLOY E)(CEPT AS NOTEt)
FIGURE 38.
Chapter IV.- Stress Analysis of a Pressure Cabin
(Prepared by George D. Bogert)
OBJECT
To present the typical stress analysis problems that
enter into the design of a pressure cabin for a high­altitude
airplane.
CONCLUSIONS
Most of the problems encountered in pressure-cabin
design are analytically determinant and for a transport
airplane of medium size should not result in a structure
of prohibitive weight. The pressure cabin should
definitely be of circular cross section, or very nearly so,
and conical or hemisp herical ends are far superior to
flat closures. One of the principal considerations must
be the design of the riveted joints between skin and
bulkheads.
DISCUSSION
The design of a pressure rabin suitable for high­altitude
flying is influenced by many factors new to
aircraft structural analysis. Aside from the usual
aerodynamic and landing forces, the temperature and
pressure differentials must be considered in their effects
on the stresses in the cabin. The problem of tempera­ture
variation is important primarily in structures of
composite materials; changes in length of various com­ponents
of the cabin will possibly introduce stresses of
appreciable magnitude, and affect the operation of con­trols
which pass through the cabin walls. Tempera­ture
effects are considered in this report.
The most important problem is, however, the effect
of t he differential pressure on the cabin structure. This
effect is investigated in detail in order to indicate the
probable stresses that will be encountered in the various
components of a pressure cabin, and to form a basis for
future stress analyses of such cabins.
A. TEi\lPERA'l'URE EFFEC'l' ON A ME'l'AL FUSELAGE
The effect of lowering the temperatu re of a metal
fu selage is a uniform unit shrinkage of each dimension
of the structure. lf the construction embodies dis­similar
metals temperature stresses might be induced,
but in a structure of homogeneo us material this effect
is not present.
Within the expected range of temperature variation,
the coefficient of thermal expansion of duralumin is
nearly constant and equal to 12 X 10-6 per degree
Fahrenheit. The change in length of any dimension
of a structure is expressed by the formula:
new temperature, and C, is the coefficient of thermal
expansion.
Change in length of fuselage
Although the expected minimum air temperature is
approximately -54°, the lowest mean temperature of
the cabin walls will be probably not below 0° F ., due to
the temperature within the cabin being held at approxi­mately
+ 70° F. For an example, however, let us
assume that the mean metal temperature has changed
from + 70° to - 30°; for a fuselage 30 feet long, the
thermal change in length is
ti.l= l,(t2- t,) C,
lo= 360 inches
=360( - 30- (70)) 12X 10-6
=360X ( - 100) X 0.000012
= - 0.43 inches.
Diametral change
If. the fu selage is assumed to be 6 feet in diameter,
the diametral change due to the temperature variation
of - 100° is
ti.d = 72X (- 100) X 0.000012
= -0.0864 inches.
Equivalent stresses
To produce the same strains as those caused by the
temperature change of - 100°, the average stress would
be:
where E = modulus of elasticity, and the other terms
are as previously expla ined.
f= (- 100) X 10,500,000 X 0.000012
= - 12,600 pounds per square inch
This stress is the same in both longitudinal and ci r­cumferential
directions.
B. PRESSURE EFFECTS O