Professional Documents
Culture Documents
A
OYAL AIR I DLISKMENT
BEDFORD. REPORT 96
a.
O
Q_
ADVISORY GROUP FOR AERONAUTICAL
RESEARCH AND DEVELOPMENT
REPORT 96
TITANIUM FABRICATION
by
L P. SPALDING
APRIL 1957
TITANIUM FABRICATION
by
L.P. Spalding
This Report was presented at the Fifth Meeting of the Structures and Materials Panel,
held from 24th to 27th April, 1957, in Oslo, Norway
SUMMARY
SOMMAIRE
669.295.6
3c8b4:3c2blc
ii
CONTENTS
Page
SUMMARY i i
LIST OF FIGURES iv
1. INTRODUCTION 1
3. MATERIAL PREPARATION 3
5. MACHINING Q
6. CHEMICAL MILLING P
7. HEAT TREATING 10
9. WELDING 11
10.ASSEMBLY TECHNIQUES 13
12.CONCLUSION 14
FIGURES 16
DISTRIBUTION
111
1
LIST OF FIGURES
Page
iv
Page
L.P. Spalding*
1. INTRODUCTION
North American has used titanium in production since 1950. The first applications
were of unalloyed sheet for fire walls, shrouds, and ducts on an experimental model.
Both Sr% manganese alloy sheet and the unalloyed grade were used in limited quantities
on several versions of the F-86 Sabre Jet, beginning in 1952. The F-100 Super Sabre
airplane, whose production started in 1953 and has continued at a fairly high rate
to the present time, represents the major utilization of titanium by North American.
Other applications have been in Navy fighters and the Navaho missile. Annual
consumption of titanium is shown in Figure 1, which indicates a total of over 2%
million pounds used through 1956.
The P-100 airplane has about 600 lb of finished titanium parts, comprising some
350 lb of alloy sheet, 230 lb of unalloyed sheet, and a few pounds of forgings and
machined fittings. The low percentage of forgings, bar, and extrusions may be
explained in part by our preoccupation with the problems of getting sheet products
into production. However, we would expect that forms other than sheet would always
represent a relatively small part of airframe usage. Essentially all of this
titanium goes into the aft fuselage section (Fig.2) which surrounds the jet engine,
afterburner, and tail pipe. This section is built almost entirely of titanium, the
chief exception being two heavy steel forgings which support the horizontal and
vertical tail surfaces.
Structural members (longerons, frames, and most of the exterior skins, as shown
in Figure 3) are made from the 8% manganese alloy, while unalloyed titanium is used
for shrouds, ducting, and the exterior skins at the extreme aft end. These applica-
tions reflect the approximately 60 to 40% distribution by weight previously mentioned.
Since this paper is primarily concerned with fabrication, no attempt will be made
to explain the design concepts and structural analyses which led to the incorporation
of titanium in the original F-100 design. (This was not a 'substitution' program).
It suffices to say that these applications have paid their way in terms of weight
saving and performance.
(a) most of our experience has been with this product, and
(b) alloy sheet represents the largest and most Important application of titanium
in airframe construction. Where comments or data apply to other products,
this fact will be noted.
All of the sheet has been used in the annealed condition, with a standard size of
36 x 96 in. and a thickness range of 0.025 to 0.187 in.
Major sources of difficulty have been non-uniformity of gage and physical properties,
poor surface condition, and lack of flatness. Three years ago it was common to find
gage variations of 15% in a single sheet of material, in contrast to the 8 to 10%
allowed by specification for most gages. In the same sheet, yield and ultimate
strengths often varied as much as 15,000 lb/in.2 measured in a given direction - this
in addition to a spread of 15,000 lb/in.2 between longitudinal and transverse
properties. Gage and strength variations in a group of sheets were even greater.
These Inconsistencies caused extreme difficulties in such forming operations as
stretching. Departures from flatness of 2 in. in various areas of a standard size
sheet were not uncommon. Grinding marks, resulting from attempts to remove scale
during intermediate rolling, caused excessive breakage on formed parts. It should
be noted that, at the time, these conditions were typical of the best material
available - and we used it.
Current practice is to accept the vendor's test report, which certifies physical
properties (two sheets per lot, each tested in the longitudinal and transverse
direction), chemical analysis (three per ingot - top, middle, and bottom), and a
bend test on each sheet, with examination of the bend at 20X magnification for signs
of cracking. Finally, sheet thickness is measured at 2 ft intervals along the center
and both edges of every sheet.
• When a s t r a i g h t e d g e is laid across the sheet in any p o s i t i o n or over any length, the sheet
surface s h a l l not depart from the s t r a i g h t e d g e by more than 3% of the distance between the
points of contact of the s t r a i g h t e d g e . For sheet thicknesses of 0.025 in. and l e s s , the
allowance is 5%.
Upon receipt of the material, tensile tests are run on 10% of each lot. The
vendor's bend test samples, which are included in the shipment, are re-examined, and
repeat tests are run if deemed necessary. Each sheet Is checked for surface
condition, flatness, gages and dimensional tolerances. During a period when
difficulties were experienced with hydrogen contamination, analyses were run on
every sheet, but this has been reduced to a 10% sampling. Figure 4 shows the results
of hydrogen analyses for the year 1956.
In contrast to these rather elaborate procedures which have been evolved to control
the quality of titanium, the practices for aluminium and steel are quite simple.
For aluminium, the vendor's test report of chemical analysis and physical properties
is generally accepted without further checking. Precipitation-hardening stainless
steels, which are being extensively used in applications similar to titanium, are
checked (on the basis of one sample per heat) for chemistry, physicals, bend, and
heat-treatment response.
3. MATERIAL PREPARATION
This operation encompasses the preparation, by shearing and other means, of blanks
for subsequent steps in fabrication. Due to the high cost (about $20.00 per pound)
of titanium sheet, much thought and effort is put into the layout of blanks to attain
optimum utilization of stock sheets (see Figure 6). A two-year average utilization
of 85% has been achieved, based on actual weight of cut blanks.
The material preparation department has, for a time, the problem of flattening
a large percentage of the titanium sheets received. Even though the producers
have more recently met the 3% flatness specification, this is still not adequate
for some exterior skin parts. To overcome this difficulty, a process of creep
flattening was devised. A stack of five to ten sheets (or part blanks) is clamped
between two 1 \ in. thick steel plates and the assembly charged into a furnace at
1000°F. Due to its mass, the load requires several hours to reach temperature,
where it Is held for 30 minutes, removed from the furnace, and allowed to slow-cool.
The sheets come from this operation almost perfectly flat. In all, we have
salvaged over 50,000 lb of titanium by this process. Some of the producers are also
using it to supplement their more conventional flattening techniques of roller and
stretcher leveling.
Some blanks are jig-drilled at the material preparation stage for tooling hole
locations or for subsequent assembly riveting. In this operation, standard high-
speed steel drills of the shortest possible length are used with a total material
thickness not exceeding !4 in. Drills are air-driven at speeds of 1200 r.p.m. on
No. 40 drill size, down to 400 r.p.m. for the !4 in. size. A constant load and feed
are mandatory. Holes are deburred by means of flat belt sanders.
In summary, it may be said that the material preparation procedures for titanium
are quite like those for stainless steel, and differ from aluminium only in the
limitation on routing. Considerably more care is taken in deburring titanium blanks,
since edge imperfections frequently cause cracking in the subsequent forming
operations.
4. FORMING. G E N E R A L OBSERVATIONS
In contrast to the usual practice of cold-forming aluminium and steel sheet metal
parts, heat is very widely used on titanium. Many parts which cannot be cold-formed
are readily made on cold dies if the titanium blanks are preheated in air ovens to
900-1050°F. (At the time of forming, blanks may be at a temperature of 600°P to
800°F, depending on the. handling required, time of contact with dies, etc.). In
some cases, heated dies may be used, with either hot or cold parts. Hot forming of
titanium has been so successful that consideration is being given to the use of
similar techniques on precipitation hardening stainless steels. For this operation
to be successful, the hot forming treatment must be compatible with the heat treat-
ment cycle for the steel.
Figure 9 compares those physical properties of titanium, aluminium, and stainless
steel which are pertinent to forming operations. Attention is called to the high
yield strength of titanium and its narrow range between yield and ultimate, which
allows little room for error in working the material. Since the spring-back
characteristics of titanium cannot be predicted consistently, it is impossible to
provide a fixed allowance for spring-back, as is done with aluminium and some steels.
Uniform elongation is limited so that only about 3 to 4% over-all stretch can be
obtained cold. For those operations where hot stretching is not feasible, forming
in stages with intermediate anneals may be required. Titanium is not amenable to
shrinking, using normal fabrication techniques, even with the application of heat.
Cognizance of this limitation should be taken in the design of parts, in tooling,
and in manufacturing processes.
Practically all formed parts, with the exception of slightly contoured skins,
are stress-relieved or annealed after forming. This is described in more detail in
Section 6, "Heat Treating'.
4.1 Bending
Corresponding to the bend radius of 3T to 3&T required by material specifications
for titanium alloy sheet, part designs normally call for a 4T minimum bend radius.
For unalloyed titanium, the corresponding values are 2T to 2%T per specifications
and 3T minimum for design.
Angle and channel sections involving straight bends are formed on power brakes,
using a radiused punch and channel die. Degree of bend and allowance for variable
spring-back are governed by the depth of the punch stroke, at the control of the
operator. About 80% of the work is done with cold tools and with blanks heated to
900-1000°F. The remaining 20%, involving larger bend radii, are formed cold. Parts
are lubricated with a dispersion of molybdenum disulfide in mineral oil, lanolin, or
petroleum grease.
Some small parts (usually angles with a joggle at one or both ends) are made by a
two-stage process, using a mating *V block and punch. In the first step, a hot
blank is formed to approximate shape on cold tools. Then, the cold part is re-formed
on the same tools, which are heated by conduction from special holders on the platens
of a hydraulic press (see Figure 10). Cartridge-type or strip-type resistance heaters
are imbedded in the holders. Parts are held in the hot dies for from one to five
minutes at about 900°F; they emerge in proper shape, requiring no hand-finishing.
The bending operations described here for titanium are somewhat more complex than
those for steel and aluminium, which are both formed cold in a single operation.
Joggled steel parts will usually require some hand re-work.
Curved angle and channel sections are produced by stretching previously brake-
formed parts (Fig.11) which are annealed as an intermediate operation. Steel
stretcher dies are generally used with a light mineral oil lubricant, although
Kirksite dies (zinc base alloy with 4% aluminium and 3% copper) may be used for
short runs. On light gage sections, this work is done cold, and troubles have
arisen from material variability. Yield strength variations lead to uncontrolled
spring-back, while gage variation may cause local necking. Warm stretching has been
used on heavier gage material with fairly good success. The dies are preheated to
400°F in ovens, and electrically heated platens are used to maintain the temperature.
Forming is done slowly to permit heat conduction into the part as it wraps around
the die.
A maximum of about 10% stretch can be realized in the extreme outer fibers of
curved titanium sections; however, most parts can be successfully formed with
considerably less deformation. In contrast, steel and aluminium can tolerate a
maximum of 20% or more stretch, if formed in the soft condition.
An effective technique has been devised for the production of those sections
where spring-back is particularly troublesome. The parts are overcontoured using
one stretcher die, annealed, and then re-stretched on a second die to exact contour
(see Figure 12). Typically, on a part with a 20 in. radius of curvature the first
stage would overcontour & in. at the center of the arc.
Figure 14 compares the steps required to make a curved 'Z* section part from
alloy titanium, 18-8 CA hard) stainless steel, 17-7PH precipitation-hardening stain-
less, and 7075 aluminium alloy.
Exterior skin parts having curvature of large radius and in only one direction
are laid in place on assembly without preforming. More severely, or doubly, contoured
parts require stretch forming, which is usually done cold over Kirksite dies (see
Figures 15 and 16). A silicone-base lubricant is used. This operation is essentially
the same for titanium, stainless steel, or aluminium, except for material limitations.
The narrow working range and 4% limit on stretching titanium means that less severe
forming can be accomplished than with most other materials. Forming in stages with
intermediate anneals is a means of obtaining greater deformations. Parts made of
precipitation-hardening stainless are readily stretched in the annealed condition,
but they must be carefully supported during heat treating to minimize warpage.
Ordinary 18-8 type stainless is excellent for stretching when annealed, but quite
difficult to work in the Vz-hard condition. Aluminium alloy parts are stretched in the
heat-treated and aged condition wherever possible, to avoid the warpage attendant
upon the quench during heat treatment of annealed material.
As with stretching of sectional shapes, attempts have been made to hot-stretch skin
panels. This has been generally unsuccessful, due to the aforementioned difficulty
of getting sufficient, uniform heating. In some cases, the dies are warmed to about
400°P, using an open gas flame.
The grips used in stretch forming are made from 5% chromium, air-hardening steel,
heat-treated to 60 Rockwell 'C hardness. 60 degree 'V serrations are spaced ~ in»
on centers. Grips must be carefully matched or aligned by the use of shims. This is
largely a matter of operator experience and technique.
The poor shrinking characteristics of titanium make it necessary that all deforma-
tion be accomplished by stretching or bending. (This is good practice for all
materials). Controlled deformation is accomplished by the use of beads, or traps
(Pig.17), which restrain the edges of the blank, somewhat analogous to the use of
hold-down rings in deep drawing. Because of limited ductility, more stage dies may
be required for titanium than for other materials. Drop hammer parts are trimmed by
friction sawing at speeds of 5000 to 10,000 ft/min.
Rubber pads are conventionally used to impart extra pressure for localized
deformations. With the heated titanium blanks, a special heat-resistant rubber is
used. Due to the hardness of titanium, a 70 Shore hardness rubber is needed, in
contrast to the 40 Shore used on steel and aluminium.
Efforts have been made to reduce the large losses of material resulting from edge
trimming of drop-hammer-formed parts (especially where traps are used). Considerable
success has been achieved through the use of net size blanks (zero trim allowance),
which can be formed only to the approximate finished shape. These parts are then
finished using hot-form fixtures, as described in Section 4.2, 'Stretch Forming.
Sections'. There is an economic balance to be figured here between the cost of trim
losses versus the cost of the hot form fixtures plus the extra operation their use
entails. If a large run of parts with large trim loss is involved, the latter method
may be cheaper.
On heavy gage titanium alloy parts, the dies are heated to about 600°P, with the
blanks at 1000°F to 1050°F, and an aluminium overlay is used (see Figures 18 and 19).
This procedure gives fairly good parts, but one aluminium piece must be thrown away
for each titanium part made.
A new process has been devised to finish-form hydropressed and similar type parts.
This Involves the use of a hot straightening fixture which is an elaboration of the
process described earlier. In these machines (Fig. 20), the part is placed between
mating dies which are heated to 1000°P (by either gas or electricity), and hydrauli-
cally loaded both horizontally and vertically, maintaining about 2500 lb/in.2 pressure
on the parts. Pressure and temperature are maintained for from 5 to 15 minutes;
the parts, which need be only approximately to contour for placement in the dies,
come out in perfect shape. This equipment is expensive ($100,000) and usually
requires the use of 5% chromium hot work die steel inserts, since they are held at
operating temperature for extended times. (Recently, some success has been achieved
with cast iron dies). Use of the hot straightening machines has achieved a 75%
reduction in personnel assigned to this type of titanium forming, by reducing the need
for hand-working of improperly formed parts. At the same time, a reduction in
rejections from about 20% to 3% has been accomplished. However, a portion of these
improvements may properly be attributed to improvements in the titanium sheet.
We have had only limited applications for deep-drawn parts in titanium. Where
such parts have been produced, most of them have been made on a Cincinnati hydroform
press rather than conventional mechanical presses with double-acting draw dies.
The hydroforra process features simple and economical tooling and high forming forces.
This Is accomplished by developing 8000 lb/in.2 hydraulic pressure in a steel dome,
behind a tough rubber membrane. A male punch is forced vertically, by mechanical
means, through and beyond a heavy draw ring (see Figure 21). The flat bank,
previously placed over the face of the punch and draw ring, is slowly formed over
the male punch as it travels into the membrane. The 8000 lb/in.2 initial pressure
is increased to as much as 15,000 lb/in.2 by compression, as the punch travels into
the dome. It is possible to obtain very deep draws by this method when the work is
done in stages with intermediate anneals. Due to the high forming stresses, some
warpage occurs on stress relieving, which necessitates final straightening.
Titanium sheet has a tendency to crack (particularly on formed bend radii) if the
surface has a brittle oxide layer which was not removed in pickling. To ensure that
no formed parts are cracked, all primary structural parts and all parts with a crack
history are given a fluorescent penetrant Inspection. Any defects that are found
by inspection are removed by filing or grinding, and the parts are then re-inspected.
Figures 22 and 23 give examples of the types of defects found. Figure 24 is a plot
of the number of parts requiring re-work. The significant improvement achieved
during the past year can be attributed to two factors:
(a) Improved material is used, as a result of critically examining bend tests of all
incoming material at 20X magnification, and
Our experience with machining of titanium has been too limited to permit more than
a few general observations. The tendency for titanium to weld to the tool edge and
the frequent occurrence of a hard surface layer on plate and billet stock both
contribute to reduced tool life. Several types of high-speed steel have been tried
for cutters, with 6W-6Mo-4Cr-3V-lC proving very superior to other grades. Carbide
(80WC-20N1) cutters permit a more rapid rate of metal removal, but cannot be used
in many applications because of excessive wear and breakage. Cuts up to 0.3 in. in
depth have been made. It is preferable to finish on the first cut, but if a finish
cut is necessary, it should be of 0.025 in. minimum depth. Also, it is very
important that any cut go below the hard surface layer. Machine tools of adequate
power and rigid construction are required. A copious flow of coolant is desirable,
and a wide range of performance has been observed with different products. Since
most of these are proprietary materials, no specific description of recommended
coolants can be given.
As a broad generalization, it may be said that the rate of metal removal for
titanium is comparable to that for Cr-Ni-Mo steel heat-treated to 180,000 lb/in.2
tensile strength. The removal rate for annealed precipitation-hardening stainless
is about twice, and for aluminium alloys about ten times, that for titanium.
Conversely, on heat-treated precipitation-hardening stainless steels, the removal
rate is perhaps one half that for titanium. Under these conditions, tool life is
roughly the same in cutting any of the materials.
6. CHEMICAL MILLING
The Chem-Mill process, developed and patented by North American Aviation for use
on aluminium alloys, has recently been extended to steel and titanium. Essentially,
Chem-Milling is a method of selective, controlled metal removal by chemical dis-
solution. Pour steps are involved: cleaning, masking, etching, and mask removal.
The process is particularly adapted to the sculpturing of formed sheet metal parts
(see Figure 26). If these parts are first sculptured by milling, the subsequent
forming operations are unsatisfactory because of the varying section thicknesses.
If forming is accomplished first, the machining operation then becomes very complex
and costly. With Chera-Milling, the parts are formed and then etched to the desired
configuration. Based on pilot plant operations, the cost of titanium removal has
been estimated at $1.50 per pound.
10
On titanium, the only part of the Chem-Milling operation not completely solved is
that of hydrogen pickup during etching. Removal of 0.015 in. of material from each
side of an 8Mn titanium alloy sheet gave about 50 parts per million hydrogen pick-
up. On 6AI-4V, on 5AI-2& Sn, and on unalloyed titanium, only 70 parts per million
hydrogen pickup was observed in removing lk in. of metal. These results indicate
that vacuum annealing after Chem-Milling may be required in some cases, although
there is optimism over the possibility of reducing hydrogen pickup even further.
7. HEAT TREATING
The only heat treating operation performed to date on a production basis is that
of stress relief annealing, which is carried out in air atmosphere furnaces at 900°F
to 1050°F for 30 to 60 minutes. This accomplishes two things:
(b) the relief of forming stresses has been helpful in preventing delayed
cracking of parts which contain high hydrogen.
Three new alloys, now in the pilot production stage, show much promise as heat-
treatable sheet materials. Their compositions are: 3AI-6MO, 4AI-3MO-1V, and 2%
AI-16V. In the solution-treated condition, these alloys are expected to have yield
strengths around 65,000 lb/in.2, a yield-to-ulitmate ratio of perhaps 0.6, elongation
of 10%, and a 2T to 3T bend radius. This should provide excellent formability, and
after an aging treatment, ultimate and yield strengths of 180,000 and 160,000 lb/in2,
respectively, may be attained.
A major problem in heat treating any titanium sheet alloy is the warpage which
results from the quench during solution treatment. The titanium producers and the
airframe manufacturers are working on this problem, but no solution has yet been
proven on a production basis. Methods tried involve two approaches:
The latter method is restricted, since hot flattening may result in some aging, to
the detriment of forming capabilities.
11
Limited work has been done on the heat treatment of titanium alloy forgings of
4AI-4Mn and 5AI-l.3Fe-l.3Cr-l.3Mo. The 4AI-4Mn alloy is erratic in its behavior,
leading to conservative properties (155,000 lb/in.2 ultimate, 140,000 lb/in.2 yield,
and 7% elongation) after a cycle of 1450°F for 2 hours, water quench, and 8 hours
aging at 900°F. The AI-Pe-Cr-Mo alloy exhibits good hardenability and readily
attains an ultimate strength of 180,000 lb/in.2 and a 160,000 lb/in.2 yield with 7%
elongation after a cycle of 1765°F for & hour, water quench, and 8 to 12 hours
aging at 900°F to 1000°P.
Parts which have been fabricated without the use of heat are cleaned by vapor
degreasing to remove forming lubricants. Where hot forming or stress relieving has
been used, a light scale is formed. This can be removed by a two minute immersion
at 120°F in a bath of 2% HF and 48% HN03, a composition which gives negligible
hydrogen pickup, but is not always too effective in scale removal. Supplementary
treatments have been devised to facilitate descaling. Before heating, the titanium
is sprayed with a boric acid solution and dried. Subsequent heating produces a
modified scale which is readily attacked by the acid pickle. Another method is
caustic treatment of the scale between heating and acid pickling. This very effective
process employs a bath of NaOH and Na 2 Cr 2 0 7 used for 10 min at 275°F.
9. WELDING
Alloys with a larger amount of beta stabilizer, such as 8Mn, give brittle spot
welds because of transformation hardening. It is possible to restore ductility
by post-weld heat treatment, but this has not been regarded as a satisfactory
production process, due to warpage of the welded assemblies.
To produce sound welds by the shielded arc process, joints must be designed to
provide accessibility of the root side of the joint to inert backing gas. Even after
final assembly, the root side of the joint should be accessible, otherwise field
repair welds cannot be made. Weld tooling must provide good joint alignment and
fit-up of mating parts. Inert gas coverage at the root side of the joint, and maximum
chill. Chill is required to cool the weld deposit as rapidly as possible to reduce
contamination, which occurs to some extent even with the inert gas blanket. Figure
28 is a schematic illustration of the tooling for butt welds.
(b) maintenance of inert gas coverage of all surfaces until their temperature
falls below 1000°F, and
(c) avoidance of contamination resulting from movement of the filler rod in and
out of the gas shield.
Both argon and helium are used as shielding gases. Argon is most widely used,
while helium is preferred where a hotter arc and deeper penetration are desired.
In the consumable electrode process, a mixture of the two gases gives best results.
Direct-current welding is used exclusively on titanium.
Figure 29 illustrates a tank made of 5AI-2& Sn alloy, using both resistance and
fusion welding. Stiffeners were spot-welded to the skin. Fusion welds were made by
the non-consumable electrode inert gas shielded arc process. Procedures have recently
been established for making production welds in 0.090 in. thick 6AI-4V alloy, using
the consumable electrode process. The welds, made at a speed of 50 in./min, were
radiographically sound and has a bend radius equal to that of the parent material.
An automatic traveling welding head with automatic electrode feed Is used for this
work.
13
The other problem which plagued the assembly departments was delayed cracking.
At periods from a few hours to a few weeks after assembly of parts, they were found
to be cracked - usually at points of stress concentration, such as corners or rivet
holes. After lengthy Investigation, it was found that excessive hydrogen in the
titanium, combined with a state of stress, caused the cracking. It should be
remarked that this was no isolated occurrence. The entire titanium industry, both
producer and consumer, was faced with this problem. To overcome it, melting and
other mill processing techniques were modified to reduce hydrogen pickup, and
vacuum annealing was widely used. An arbitrary limit of 150 parts per million of
hydrogen was written into specifications. On our part, the efforts to reduce
built-in stresses by making better fitting parts was given strong emphasis. These
two actions together virtually eliminated the delayed cracking troubles.
Drilling and countersinking of titanium are about the same as for stainless steel.
High-speed steel (6W-6Mo-4Cr-3V-lC) twist drills have been found to give much better
life than other types. Keller air drills are used in preference to hand drills
wherever possible. The speed and positive feed control offered by the air drills
makes for much faster work and up to eight times the drill life. Jigs have built-in
provisions for the bayonet-type fittings on the air drills (see Figure 30).
Deburring is done with rotary files, chisels, and a variety of special tools.
Dimpling is accomplished by the hot coin dimpling process originally developed for
the harder aluminium alloys. A-2 sec hold time with the tools at 750°F is used on
titanium, in contrast to 450°P for aluminium. During the period when cracking
troubles were prevalent, a-6 sec hold time was used, but with no improvement.
Erco automatic punch riveting (Pig. 31) has been very successful on sub-assembly
work which can be fed to the machine. At present a maximum thickness of 0.100 in.
can be handled on titanium. The machine operates in two steps, at a rate of 15 rivets
per minute. In the first stage, the hole in punched and the punch retracts. In the
second stage, the rivet is fed in and upset against the punch, which forms part of
the upsetting tool.
14
Attempts have been made to use titanium instead of Monel rivets for increased
shear strength at less weight. These efforts have been unsuccessful, due to the
high percentage of cracked rivets encountered. Some very limited success has been
reported with hot upsetting of titanium rivets, but no production applications are
known.
Titanium bolts are receiving increased interest, again because of their attractive
properties. Acceptance was long delayed by their erratic behavior (in common with
all titanium products) especially in fatigue and notch brittleness. These difficulties
appear now to have been largely overcome.
Figure 32 shows what happens to all of the titanium sheet purchased for a single
F-100 airplane. It is seen that an over-all recovery of 53% is realized; however,
today only 2% of the total is lost because of material defects or workmanship. The
larger losses occur in preparing part blanks and trimming formed parts. This is
inherent in the type of parts being produced and, with the exception of receiving
inspection rejections, the figures given now compare favorably to our experience
with other materials.
The unalloyed scrap is sold to the titanium producers for $1.00 per pound and is
re-melted by them. Alloy scrap Is sold to scrap dealers for about $0.50 per pound.
It is hoped that the producers will develop better methods for scrap utilization,
with the goal of reducing the high price of titanium and, Incidentally, providing a
better market for scrap.
12. CONCLUSION
It will be evident from the foregoing presentation that the fabrication of major
airframe components from titanium is an expensive business - considerably more so
than with conventional materials. The high cost of raw material, increased inspection,
hot forming, and extra processing steps all contribute to the final bill. It
would be interesting to show a step-by-step comparison of the costs of producing
typical assemblies in titanium, stainless steel, and aluminium. However, such a
comparison cannot be made because, even today, the figures are constantly being
outdated by cost-reducing improvements in manufacturing techniques. Despite its
high cost, careful analyses have shown that titanium construction can be clearly
justified for many applications in the temperature range of 300°P to 700°F.
15
"THOUSANDS
o r POUNDS
1000,
NUMBEROF
SHEETS SAMPLES TESTED-6176
HIGHH 2 SAMPLES-266
%HIGH H2SAMPLES-4.3
SPEC MAXIMUM
150 PPM
PERCENT
REJECTED 3%-5% FLATNESS MAXIMUM HYDROGEN
REQUIREMENT REQUIREMENT
CEASE MILL
INSPECTION
20X BEND
1956
Fig.5 Titanium sheet rejected at receiving inspection
ULTIMATE
YIELD 120 130
%% MANGANESE 9
TITANIUM
ALLOY YIELD ULTIMATE
NEW HEAT 65 110
TREATABLE
TITANIUM ALLOYS
YIELD ULTIMATE
STAINLESS 40 130
STEEL
17-7PH ANNEALED
YIELD ULTIMATE
ALUMINUM 0 15 33
7075-SO
ANNEALED
iiii
£
Pig.10 Hot straightening on a hydraulic press
22
STRETCHING ACTION
4 1ST OPERATION
OVERCONTOURING PART'
NET STRETCHING
ACT,0N
CONTOUR "
2*B OPERATION ^
' N E T DIE''
•NET CONTOUR
Fig.12 Net die technique
23
1
i
i
±_
i
i
I ANNEAL
STRETCH FORM TO CONTOUR
• *
fc OR USE HOT ±
HEAT TREAT
±*
1 STRAIGHTENING
HOT
STRAIGHTENING FIXTURE
BE-STRETCH TO CONTOUR)
FIXTURE
TO AIRFRAME ASSEMBLY
OVERLAY
PART
FORMING
BLOCK
PART
ALUMINUM OVERLAY
DISCARDED AFTER EACH PART
REFRACTORY
BRICK
UNING
GAS
BURNER
HYDRAULIC ACTUATORS
Fig.20 Hot s t r a i g h t e n i n g press
27
STATIONARY
OIL UNDER RUBBER
PRESSURE DIAPHRAGM
PRESSURE
PAD TYPICAL PART
CRACKS
* V
oo
29
10,000
8,000
PAR"rs
REWORKE >
4,000
/ \
2,000
0
NOV D J F M A M J J A S 0 N D
1955 195<5
Pig.24 Titanium parts requiring rework
I *J- i
1^ J H HP*
I
w&wr
&
^ > •
V
11 A /
J B !
r
\M
V
* T,^-
iHi
>.,—>
fc \
1.
'l
• "
*
ii '•
St* 0 0
X^ - « >
W1
p.
Pig.31 Erco automatic punch r i v e t i n g
33
REJECTED IN
RECEIVING
LOST IN TRIMMING NSPECTION
FORMED PARTS
1 % SCRAPPED
DURING
FABRICATION
+ % SCRAPPED
DURING 580 LB AIRFRAME DELIVERED
ASSEMBLY TO CUSTOMER
Pig.32 Titanium losses during fabrication
DISTRIBUTION
FRANCE 0. N. E. R. A. (Direction)
25, avenue de la Division-Leclerc
Chatillon-sous-Bagneux (Seine)
TURKEY M. M. Vekaleti
TURQUIE Erkaniharbiyei Umumiye Riyaseti
Ilmi I s t i s a r e Kurulu Mudurliigii
Ankara.
Attn: Colonel Puat Ulug
*
Printed b> Technical Editing and Reproduction Ltd
95 Great Portland S t . London, 9 . 1 .