DIE CORRECTIONS FOR CHANGING

FLOW CHARACTERISTICS
Luis Bello
Garbell Tool & Die Company, Inc.

CORRECTING SOLID DIES

Y17
Some factors to always keep in mind when correcting
extrusion dies are die temperature, container
temperature, billet temperature, extrusion speed,
press alignment, deflection, and tool stock deflection.
Keeping these facts in mind, we are sure that in order FLOW _
to correct or build an extrusion die, 90?4. of the work
to be performed is based exclusively on experience
and vocation toward this type of work.
There is always more than one way to correct the
same problem on an extrusion die and no two die cor-
rections will approach a correction
way. The corrections will be different,
problem the same
but the end
+iPE
A TO SPEED
result will be the same “GOOD EXTRUSION”. UP FLOW
Aware of all these facts, we will proceed to mention
some typical experiences in die corrections.

Methods of Correction
The most common methods of correcting an extrusion
die are the following:
FLOW FILEON
1. Choking and relieving bearings, CHOKE TO
2. Correction by flow.
SLOW UP
FLOW
In Figure 1 sketches “A” and “B”, the die corrector
will balance the flow of aluminum by relief filing and
choke filing the die bearings.

Figure 2, sketches “A” and “B”, show one way of cor-

recting a “U” channel:
B
As you see in illustration “A”, the “U” channel extru- FIGURE 1.
sion in question is extruding with the side legs moving
extremely fast in relation to the top. To correct this In cases where angularity is off on legs, whether they
uneven balance of flow, we must speed up the bearing are in or out, the die corrector will file chock and relief
where the flow is slow by relief filing the bearing as in the directions in which he wants the leg to move.
shown on sketch “B”. (See Figure 3, sketches “A”, “B”, “C”, and “D”.)

EXTRUSION

E
INGS

\\ FAST FLOW ON
SIDE LEGS OF CHANNEL
LE(SS MOVING
TOO FAST
F/GURE 2.

89
 EXTRUSION FILE
ON CHOKE
A

LEG OUT / A

/
FILEON RELIEVE
INSIDE BEAR/NG

B
LEG IN /B
_._— —-

CHOKED

FLOW
c FILEON=
L
CHOKE FILEON
INSIDE RELIEVE
BEARING OUTSIDE
BEARING
RELIEVED FIGURE 4.

Per Figure 5, sketches “A” and “B”, the die corrector

will work the die strictlv by flow. The bearings will be
RELIEVED at 90° with the face of he die at all times. He will
always speed up by grinding the face of the de
around the areas he wants to speed up, or by cutting
excess bearing from the back of the die.
D
Both methods are widely used and both give good
results.
FLOW
The first step to take to correct an extrusion die is to
check all wall openings to make sure these walls
CHOKED
extrude within tolerance, especially those walls
located on deep tongues that will have extra
deflection.
The die corrector must make these inspections, for he
FI,GURE 3.
is familiar with the press or presses that he works
with daily. The die manufacturer has to apply a safety
factor on openings (especially on tongue openings) TO
MAKE SURE THE DIES WILL PRODUCE EXTRUSIONS
starting at bottom commercial tolerances, and conse-
Per Figure 4, sketch “A”, the corrector will file relief quently the extruder will get much more wear out of
in the inside bearing and chock on the outside bearing each die that he purchases.
of the leg on the right. By doing this the leg will tend
If after inspecting the die, the die corrector finds that
to go toward the inside of the “U” channel.
walls should be slightly open, he should open them
Per sketch “B”, the die corrector will file on choke on before extruding the first billet through the die. Once
the inside bearing and relief file the outside bearing the openings are to his satisfaction he must check all
of that leg. By doing this the leg will tend to go die bearings to make sure they are parallel and free of
toward the outside of the “U” channel. burrs and deep file marks, especially on exposed sur-

90
 SHORTER

L
GRINDING BEARING
FACE OF SURFACE“—/
DIE TO qo’

SPEEDUP /’ PARALLEL

Rr
qo” BEARINGS —
FLOW
/ A
DEFLECTEDTONGUE
+
UNDER PRESSURE
UJ WILL PRODUCE
PARALLELBEARINGS
SHORTER CUT EXCESS

J17-
BEARING BEARING
SURFACE FROM BACK Iil!l
\ OF DIE TO
SPEEDUP FLOW

B
FIGURE 6.
FIGURE 5.
The bearing opposite to the tongue should be sli!~htly
faces of the extrusion longer, so when the tongue deflects the exit of the
bearings will be on the same plane, If this is not done,
Deep Tongue Dies chances are that the extrusion will have a convex cur-
All bearings should be at 90° with the face of the die, vature, (See Figure 7, sketches ‘6A’”, “B” and “C”.)
with the exception of cases as are found in deep
tongues. Then we would rather have the upper bearing On the other hand, if the upper bearing is flat and the
on the tongue, at slightly more than a 90° angle. (See tongue bearing is filed on relief, the extrusion will
Figure 6, sketches ‘(A”, “’B” and “C” ) have the tendency to concave. (See sketches “D” “E”,
and “F”. )
A
It is important that the bearings be on the same plane
at the exit side of the bearings, especially on the
tongues which will have excessive deflection. The
closer you match the two opposing bearings on a
tongue the better contours you will obtain on your
extrusion.

Testing Dies
After some of the preceding steps have been checked,
then you can proceed to push one o! two billets
through the die for a test.
ANGLE MORE When running tests of new dies, always make sure
I
THAN 90° that the lead ends of the extrusions are marked pro-

/’
perly and cut. Similarly, the centers of the length of
AREA the extrusion must be marked and cut in pieces of
OF DEFLECTION about 12 inches in length. (See Figure 8.)

These steps must be carefully taken (especially on

multiple-hole dies), in order to avoid correcting the
wrong opening on the die,
Once you have the lead end and the center cuts of
your extrusion, you will proceed to study the tendency
of the flow of aluminum from the lead ends—that is,
the areas of the extrusion that tend to be slow or too
fast,
From the center cut you check all your overall and
wall dimensions. After checking your dimensions from
B I the center piece, you should open all walls that
require opening before doing anything else. Very often
DEFLECTED you will find that just by opening the walls to desired
TONGUE dimension, the flow of aluminum will change and the
UNDER PRESSURE extrusion will come out with the correct angularity. If

91
 EVEN LENGTH
\
BEARINGS
UNDER PRESSLJRE, CONVEX
m \ CURVE

GUE

B I
DEFLECTED
TONGUE

RELIEVED E F
BEARING_- CONCAVE
CURVE

Isa
EVEN
LENGTH T
BEARINGS
\
7’

D liil
1
UNDER
PRESSURE
BEARINGS
WILL
BECOME
&$ /
UNEVEN
FIGURE 7.

after opening the walls to allow commercial toierance, This does not mean a screw hole or any other critical
the extrusion is stiii out of angularity, then you wili design on a hollow dle mandrel wiil tear or starve by
proceed to correct the die openings according to the speed in the press. Sometimes this wili happen when
lead ends. these areas are not fed properly.

When testing dies, it should be remembered not to

use excess lubricant on your bearings. The excess Another thing that has to be considered in testing
lubricant will melt under extreme heat and will flow dies is the alignment of the press. A die which has
down to settle on the horizontal bearings. The excess been corrected for a press that is properly aligned will
lubricant on these bearings will make the aluminum not rerun properiy if the press is out of alignment. The
travel farther on the lower horizontal surfaces, causing die shouid be reworked to compensate for the press
what is known as a false breakthrough, or false lead being out of alignment. (See Figure 11 sketches “A”
end. (See Figure 9.) No greases should be used. and “B”. )
Graphite solutions resist melting best.
In sketch “A”, the die was corrected for a press that
The die corrector must keep aware of this fact at all was properiy aligned. The flow of aluminum wi~s
times. Often, he will find himself trying to slow down
a section of the extrusion that does not need it.
MARK AND CUT
Temperature and speed when testing dies are very CENTERS OF
important. The die should be run for production at the EXTRUDED ~ .,
LENGTHS
same temperature and extrusion speed that were used
at the time it was corrected. Often a die is corrected
at a given rate of speed, and when it comes up for pro-
duction the extrusion speed is increased. The
increased speed will increase the temperature of the
aluminum as it passes through the bearings, changing
its flow on critical openings where the flow should be
slower to keep up with the rest of the extrusion. This
is more often found when extruding hollow dies,
where you have screw holes or other critical designs
on the mandrel.
In the case of screw holes, there will be a tendency to
DIE J MARK AND CUT
LEAD ENDS OF
EXTRUSIONS “

tear. (See Figure 10.) FIGURIE 8.

92
 EXCESS /—- DIE
H LUBRICANT
—

AREAS

NO
LUBRICANT
HERE
FIGURE 9.

‘X’R”S’”YP9
SCREW HOLES
WILL TEAR
/
FIGURE 10. DIAMETER

balanced evenly. In sketch “B”, the same die was set /’

up to run on the same press, at the same billet DIE
temperature, with the same rate of speed, but this
FIGURE 11.
time. the press was out of alignment. Consequently,
the flow of metal will have changed. This, in most
container on the side of the container where the
cases, will require a slight correction to be able to run
elements are not working properly and are cold. These
acceptable metal.
will cause the flow of aluminum here to be slower
than on the hotter side of the container. Also, bad
Also in sketch “B”, you will notice that by the con- elements in the billet heater can cause one side of the
tainer being off to the left the billet is off just the billet to be at higher temperature than another. l-his
same. This condition will move a left leg of the shape will result in the cold side of the billet being slower;
toward the center of the diameter of the billet, where and even if the die was corrected properly, a billet in
the flow of aluminum is at a faster rate. When the flow this condition will make the flow change.
of aluminum is upset in such a manner, it is clear that
All these are problems that a die corrector must think
this die must be reworked at the press to compensate
of before making any final decisions of what to do to
for the unbalanced flow brought about by the press
correct any die. The examples we have used to
being out of alignment.
illustrate the problems you can run into because of
One correction in this particular case would be to bad elements in the container, bad elements in the
speed up the leg on the right which has been placed billet heater, and the press being out of alignment,
to the edge of the billet (Sketch “B”) where the flow is have been on one-hole dies. All these problems are
less due to the drag of the billet against the container. more complicated yet when encountered in running
multiple-hole dies. In multiple-hole dies the problem is
The alternative correction would be to choke up or
not to balance just one hold or one-half of the extru-
slow up the flow of metal on the left leg (Sketch “B”)
sion but to readjust the complete extrusion.
which has been placed towards the center of the billet
Otherwise, you will wind up with different IengtFs of
by the press being out of alignment.
extrusion which will cause your scrap rate to go up.
Before correcting a die that you have already cor- Also your stretcher jaws will have to be moved back
rected, be sure that the press where the die was run and forth to compensate for the different lengths of
was properly aligned. extrusions coming out of the die. This will increi~se
the cost of operation.
Other conditions that can cause changes in flow on
well-corrected dies are burned-out elements in the In Figure 12, sketches “A” and “B”, we have

93
 posed the same sample as we did with the one-hole
A
die in the previous illustrations. As can be seen, the I
problem here IS more complex, but the procedure to
follow to rework the die is the same as with the one-
hole die. The holes that fall on the outer edge of the
billet diameter will flow slower than the holes toward
the center of the billet diameter. Therefore, the slower
holes on these dies must be speeded up at the press
to compensate for the holes that were moved toward B
the center of the billet diameter.
BEARINGS
DIE FILED
A /- PARALLEL
BILLET
IA METER

d K

%1

IB
.T25

c BEAIUNGS]
FILED ON /77 ~A
RELIEF
(( 2 1
+3

<‘ ~f

E
/ BEARINGS
FILED
/ PARALLEL
.,,-.-/
FIGURE 12.

Other problems that may cause different runout on an

extrusion die are bearings filed on relief and bearings
filed on choke. On a two-hole die for instance, assume
one hole filed with flat bearings or on relief, and the
other hole filed on choke. (See Figure 13.
sketches “A”, “B” and “c”. ) L
As can be seen here, opening No. 1 and opening No. 2
are filed to 0.125 inch. The extrusions coming out of HOLE X2
L
.125
these two holes will be of the same wall thickness.
but by hole No. 2 having been filed parallel and hole
No. 1 having been filed on relief, the extrusion coming ?p
out of No. 1 will be considerably longer than the extru.
sion coming out of hole No. 2. The extrusion from
hole No. 1 will run longer and faster because the
BEARINGS FILED
aluminum finds less resistance and will have less
drag, due to the fact that hole No, 1 is filed on relief. FIGURE 13.

The same thing WIII happen if bearings on No. 2 hole

94
are filed parallel and bearings on No. 1 are filed on DIERING
choke. (See sketches “D”, “E”, “F”. )
In the sketches you will see that we have almost the TAP
COR
same condition we had in the previous illustration but
with the difference that hole No. 2 will be the fast
hole even though it is filed parallel, The reason is that
hole No. 1, having been filed on choke, will create
more resistance to the flow of aluminum traveling over
the choked bearing. Runout on hole No. 2 will be
longer than that from hole No. 1.
MINIMUM
WAUOf .061
These latter examples Illustrate some of the reasons r
why the die corrector should check the bearings on a
die before testing to make sure the bearings are as
parallel and as flat as possible. Checking these small
details will save a lot of time and guesswork for any
die corrector. It will also save tooling from being
ruined by over correcting, and most of all it will cut
down on the number of costly tests.

CORRECTING HOLLOW DIES

Having covered some of the more typical cases in cor- L--2. -.-1
recting solid extrusion dies, we WIII now discuss the
correction of hollow extrusion dies. The best defini-
tion of a hollow extrusion die IS that it is made for the MINIMUM
WA1lTHICKNESS
OF @22
purpose of a double extrusion. This fact makes it a
more complex dle to correct. [
We have defined these dies as double-extruded due to
the fact that first. the billet is divided into two, three,
or more sections as it hits the face of the die. After
the billet has been sectioned on the face of the die,
the sections will travel across and over the surfaces
of the mandrel until they come up to the opening on
the cap which will extrude the final form of the actual
extruded shape. (See Figure 14. )
Besides being more complex for correction, the
hollow extrusion dle has more varieties. You cannot
apply one type of hollow extrusion die to every extru-
sion shape to be produced. But rather, you should
pick the most suitable hollow die type for the extru-
sion that you will be trying to produce. 3.rwJ --l
Having a variety of hollow extrusion dies, the die cor-
rector must be familiar with the characteristics of FIGURE 15.
each type. Some WIII extrude colder, or with less
Pan-Cake Type Die
pressure and more or less heat than others, Some will
deflect more than others. Some can be lubricated after This type of hollow die is used to produce small extru-
each billet is extruded, and consequently permit sion shapes with excessively high extrusion ratios and
extrusion of more billets consecutively than others, which must be extruded on multiple hole dies.
CAP Usually this type of die is preferred for shapes starting
as small as 1/4 inch and running as large as 13A inch
for average size extrusion presses (from 700 tons to
2200 tons).

CORE
\ SECTIONED
BILLET AFTER BEING
I WILL TRAVEL OVER AND AROUND
THE CORE UNTIL IT REACHES FIGURE 16.
THE CAP OPENING.
FIGURE 14.
Housing Type Die Bridge Type Die
This type of hollow die is the most generally used in The bridge type hollow die is used mainly with the
aluminum extrusion operations. It is suitable for most idea of producing extrusions with better finished sur-
average extrusions with relatively low extrusion ratios face. It is a type of die that runs at cooler
such as shown in Figure 15. temperatures due to the fact that the die is stripped

95
 clean of aluminum each time that a billet is extruded. DiE RING O.D.
The die being clean as each billet is extruded, this
gives the operator a chance to lubricate the parts and
bearings on the die with lubricant spray mist.
Lubricating the die constantly maintains it at an even
temperature giving a better finish to the aluminum.
The disadvantage of running this die is that the scrap
rate is higher due to the face that a longer aluminum
butt must be allowed, to be able to pull back the butt
as the container starts the cycle going back.
Bridge dies are used to produce light wall extrusion
shapes as shown in Figure 17 and also multlple
hole dies.

DIE RING
PORT-HOLE TYPE
CK-UP
LATE
,//” ‘- ‘--y-
?+ \
.075 WALL

‘i
\

CORE \ 1,
I
B \{
\
‘\,

“’’L/’

FIGURE 18.

Taper Type Die

L-i moo

r
.050 WAU
The taper type hollow is mainly used on large shapes
with a low rate extrusion ratio. It is a die that will not
develop cracks on the mandrel easily due to the fact
that the small diameter of the die is slightly smaller
than billet diameter. For this reason part of the
pressure applied to the tooling is absorbed by the
mass of steel of the die ring and the die being a
separate part from the ring will tend to slide forward
as the billet is extruded and return to normal position
as the extrusion cycle is finished (acting as sort of a
piston).
Shapes like 13/4 x 4 in. are extruded on this type die.
The disadvantage of the taper type hollow die is that it
takes quite a long time to clean, check, polish and run
the die again,

FIGURE 17. DIE RlN6_”

Port Hole or Self. Contained Type Die

The port hole or self-contained type hollow die is used
only when the extrusion shape to be produced is
excessively large. Then to be able to put strength on
the die you must design the die to the full diameter of
your die ring and bolster. (See Figure 18.)

To produce this size of extrusion with a six inch billet

you must go to a port hole or self-contained hollow
die design.

96
 r .125 WALL
The chamber type die is a solid die, but with hollow
~:+.=.. -– -- ,1-. die features. It is used when deep critical tongues and
I ,1 shapes have to be extruded. It is a die built of a single
, ,fo
piece of steel, making it difficult to correct, file, clean
and polish.
See illustration below:
—,
I —.L 1

L— 4. -------
BACKER
125 WALL
-i/-”

——

-–.---J /

II I
II I

FIGURE 19. \ ~j~

Bridge Port Die FIGURE 22.

The bridge port hollow die is a die not commonly
used, It is used on high ratio extrusion shapes and is
built to fit a special ring with an opening slightly Places to Correct Hollow Dies
smaller than the billet so that some of the pressure The advantage of correcting hollow dies is that there
applied during extruding is absorbed by the ring are various places to apply corrections before you
itself, have to work the bearing itself, A well designed hollow
die can be corrected by just grinding or working the
DIE /?lfVG-’ flow in the ports and in many cases by not working
the bearings at all.
To correct the flow on a hollow die you should “first
examine the ports, to make sure that the volume of
aluminum to be carried by one port is balanced with
the volume carried by the rest of the ports, in relation
to the area of the side of the extrusion shape each
port should feed. (See Figure 23)

c
Taking this into consideration, we then know that by
opening or restricting the flow through the ports we
speed up or slow down the flow of aluminum in any
section of the shape being extruded.
FIGURE 20. A section of a hollow extrusion will hold back also if
that section is extruding light wall thickness. That is
Feeder Plate Die why you should look for other problems before lopen-
The feeder plate die is a solid die designed to produce ing” up the ports on a hollow die. The reasons a hollow
a shape larger than the billet size. It feeds the metal ex~rusion might not come out with light walls from a
into the die opening in a similar manner to what is hollow die are as follows:
done on a hollow die, and it forms Aluminum pockets
1. Mandrel shifting in relation to the cap.
toward both ends of the shape to be able to balance
the flow with the center and rest of the extrusion. (See 2. Poor feed on core
Figure 21.)
3, Walls filed lightly.
PORTS 10 FEED SHORT SIDES
OF EXTRUSION WIL1 HAVE LESS
VOLUME TO
fFfD LESS
AREA
T- \
-,
W3RTS TO FEED
— LONG SICIFS
y OF FXIRUSIONS
,,>,LLHAV,
MORE VOIUME
70 FEED MORE
a
.. . , /’” AREA

l~P
.125
WALL T /’
)
t DIE
L./
FIGURE 21. FIGURE 23.

97
 Shifted Dies obstacles on the mandrel to facilitate the flow of
In cases like the one illustrated in Figure 24 you aluminum while it travels along the surface of the
should not make any correction until you have taken mandrel. This is done by milling obstacles off. or by
care of the shifting on the die. grinding as shown before. (See Figure 25,
sketches “A” and “B’”. )
Another characteristic of hollow dies is that the ports
around the mandrels act as a long bearing, and the As can be seen by sketch “A”, the angle being steep
will create two obstacles on the side of the mandrel
SHIFTED which will slow down the flow of aluminum. On sketch
CORE “B”, you see the mandrel with both obstacles
removed, either by milling machine or by grind incj In
SHIFTED this case the angle is less, with no obstruction, and
CAP more volume of aluminum goes through. Therefore.
this will speed up these two sides of the extrusion.
The die makers will often build these obstructions on
hollow die manurels intentionally. Whenever a die
WALL EXTRUDING
HEAVY
1]]
l!--
1 WALL EXTRUDING
LIGHT
designer feels he should slow down the flow to feed
certain sections. he will do so and in many cases it IS
only a safety factor not to use a very heavy bearing. If
FIGURE 24. it is not needed it can be easily removed by grinclng

OBSTACLE TO FLOW GOOD EXTRUSION

\
A

STEEPANGLE
B CONCAVE WALLS
REMOVED CONVEX WALLS
OBSTACLES
T

c
B F1GuRE26.

Concave Walls
LESSANGLE
In cases of concave walls on hollow dies, the most
SPEEDS UP FLOW
common problem is that of excessive deflection on
the mandrel, The mandrel. under pressure. will give so
FIGURE 25.
much that the bearings around the core WIII moje past
aluminum traveling through the ports can create the entrance of the bearing on the cap. (See Flgllre
excessive heat. In some cases you should eliminate 27. sketches “A””. “B”’).

98
 EXTRA
HEIGHT

-J
CAP
A
SHIMS TO LIFT
BEARING THECORE
CORE UNDER FIGURE 28.

‘—/,PRESSURE

WASHOU~
f BEARING
I B
FIGURE 29.
CAP
BEARING As seen in the illustration, the bearings on the core
have been washed out after X number of billets
FIGURE 27.
pushed. In this case if the washout is not too deep
and your walls can be opened up some, then you will
Sketch “A” Figure 27 shows a hollow extrusion file, mill or surface grind the surfaces of the bearing
die not under pressure. As can be seen, the bearings on the core until the washout is no longer visible. If
of the core are a bit higher than the bearing on the the walls are already heavy, then the die should be
cap. This situation is normal, and if an extrusion replaced with a new one.
should be pushed it would most likely be satisfactory.
On sketch “B”, you see the die under extrusion Slit Corners on Extrusions
pressure. As you notice, the bearing around the man-
Some of the reasons why a hollow extrusion will
drel has gone beyond the entrance of the bearing on
develop split corners are inferior quality billets;
the cap. This excessive caving on the mandrel will
improper lubrication practices; quality of lubrication;
cause the extrusion to come out with concave walls.
and more than anything else, small welding chalmbers
on the die in relation to the wall thickness being fed.
To correct such problems in some cases, you should
shim up the core by placing spacers in between the When running into the problems of split corners, the
cap and the core. (See Figure 28.) first thing to do is to check to see that you have pro-
per welding chambers so that enough aluminum is
By placing the spacer between the cap and core, you getting into the welding chambers to permit better
will give the extra height to the bearing on the core in mixing before extruding the actual shape. (See Figure
relation with the bearing on the cap. So it will not go 30, sketches “A”, “B”, and “C”.)
beyond the cap when under extrusion pressure. If the To correct this problem, you should grind your
spacers needed to life the mandrel are too thick, then welding chambers deeper and wider if the size of the
whatever thickness of spacers you add, the same die will permit it.
thickness of material should be machined off from the
cap to maintain your total die thickness.

Convex Walls Twisted Extrusions

In the case of convex W4:S in hollow extrusion dies, There are a number af reasons why extrusions will
one of the contributor, d causes is that of bearing develop twists while being extruded, Some of th(?
washed out on tt,e core. (See Figure 29.) causes are imbalance of flow, imbalance of bearing

99
 In Figure 31 “A”, you will see two possible
reasons why the shape to be extruded on this par-
ticular design will tend to develop twists.
First, the port on the left is larger than the port on the
right. Consequently, by the port on the right being
small and restricting the flow of aluminum, the extru-
sion will tend to bend toward the side where the slow
CORNERS flow is located. (See sketch “B”. ) The correction to
this problem should be to balance the volume of
aluminum by grinding or machining the small port a
little larger. Second, the port at the bottom of the
illustration is not running parallel to the cap and the
core. The extrusion will tend to twist when something
like this happens. The way to correct this problem is
B
to grind or machine the port to be parallel to the cap
and the core.
As explained therefore, the problems indicated in
Figure 31, sketches “B” and “C”, can be over-
come by correcting the flow of aluminum around ‘[he
core of the die.

Bending and Twisting

I J When two or more walls have the same thickness, but
ADEQUATE WELDING the bearings are designed of different length or not
balanced, the extrusion will tend to bend and twist,
CHAMBER IN RELATION
just as it does when the core is not balanced properly.
TO WALL OPENING (See Figure 32.)

c As seen, the wall openings are the same, 0.125, but the
bearings are not equal. The bearing on the left is
shorter than the bearing on the right. This condition
will create bends or twists on the extrusion. In this
particular case the extrusion coming out of the die
will twist to the right. The larger bearing on the right
hand side will hold back while the bearing on the left
will have extra speed, which will cause a twist to the
slow side.

WELDING
CHAMBER TOO SMALL
IN RELATION TO
WALL OPENINGS B

FIGURE 30.

BEN1
design, cold spots on your billet, cold spots on your EXTRUSION
billet heaters, and certain types of die layouts.

UNEVEN
SIZES OF

“_ \ GRIND OR MACHINE
!0 MATCH
OPPOSITE
,~– POR7

;
/ GRIND OR MACHINE SO
—
- PORTWILL RUN PARALLEL
DISTANCE l\ TO EXTRUSION WALL

?KEK-X!H!!Y
F/GLJRE31.

100
 B

.!25 + 1+ &.12s

--u
SAME
WALL OPENING

FIGURE 32.
I , 1 I

Angularity Problems FIGURE 34.

When problems of flow and imbalance bearings are
present, you can expect extrusions out of angularity Radial Layouts
(See Figure 33, sketches “A”, “B” and “C”). Twists are also developed by the way the die openings
are laid out. Especially on the radial layouts (See
Figure 34).
The tendency of the extrusion to lay flat on the runout
table will develop twists on extrusions. It must be
pointed out that not all radial layouts will do what we
have just illustrated; but they are most likely to do so,
especially on heavy sections. Correcting this problem
is similar to correcting a one-hole die by flow. In this
case you have to balance the flow on each and every
B
# hole.

&
B
The causes of twisting on a hollow die extrusion are
more or less the same causes of twisting on solid die
extrusions. With the exception of mandrel drag, a
solid die will tend to twist by imbalance bearings
deflections and in some cases by die layouts.

BAD ANGULARITY
Canisters
CONVEX BOW The tendency of multiple-hole extrusion dies to twist
T and have bad or irregular runout can be lessened also
by using what are known in the industry as canisters.
“’z---- c
Canisters are used mainly on multiple-hole dies such
/7 as the three-hole extrusion die shown in Figure 35.
& r——7——7 \
As you see, the number of holes on the canister ml~st
be the same as the number of holes on the die. The
holes on the canister are mostly pie-shaped to con
form to the shape of the standard bolsters or sub-
bolsters and are slightly larger.
FIGURE 33. There are some canisters made to conform to certain
shapes, but this method is not used too much. When

r–– DIE

DIE RING SUPPORT CANISTER

FIGURE 35.

101
 used such canisters are made by assembling many
blocks of graphite. These are very easily broken in
operation. The most widely used type of canister is PLATE
one made out of brass. The initial cost of the brass
canister is higher, but it will pay off in the long run for ATEN
it will not chip or crack and will wear for years without
having to be repaired or replaced.
When canisters are used on multiple-hole dies,
especially on six-hole dies (or what are known through
the industry as “spaghetti dies”), the extrusions are
kept separated and moving straight forward while
traveling through the length of the canister. This gives
the extrusions time to cool, avoiding their sticking to
each other as soon as they leave the exit of the
bolster and before they reach the runout table.
The outside diameter of the canister must be slightly
I A
DIE SLIDE
smaller than the opening of the pressure plate. Its
length varies from 36 in, to 72 in. depending on the
i2° DIA.-
desired or individual needs.
The brass canister must be submerged in a solution
containing graphite so that it will become impreg-
nated with it and the extrusions coming in contact
with its walls will not become marked or scratched.
B I
Canisters must be clamped securely against the
bolster, and they must turn with it. Thus their open-
ings will always be in perfect alignment with the open-
ing of the bolster or sub-bolster.

T
PROBLEMS ORIGINATING WITH THE PRESS 8’/2”
To correct any type of die it is of greatest importance J_
that the die corrector be familiar with the press or
presses that he works dies for.
Some of the problems that you find with extrusion
presses are the following:
1, Improper tool stacks
2. Pressure ring not flat. FIGURE 36.

3. Improper seal of the container against the die the heavy bolster has been replaced by a thinner
bolster and two thin spacer rings. In most cases you
4. Washed out container liner, will get by using this setup, But when you have to run
5. Worn out dummy blocks. a critical section on a setup similar to this one, you
will have to work and rework the die in order to tallow
Improper Tool Stack for its deformation under pressure. A setup like this
The tool stack on an extrusion press is designed by one will cost extra time in die correction, as it will
the press manufacturers to suit the qualifications of increase the cost of production and the life of the dies
the size of the extrusion press built. Nevertheless, and support tooling will be at a minimum.
there are times when the extruder will make the error The spacer rings shown in Figure 37, sketch “A”,
of using spacer rings when they should not be used. should not be used. Instead, a sub-bolster should have
There is a limit to the size and number of spacer rings taken the place of the two spacers, which by being in
you can use in a single tool arrangement (See Figure one solid piece will provide much better supporl to
36, sketches “A” and “B”). the smaller bolster, backer and die (as shown in
Figure 37, sketch “B”), Cracks will develop on corners
As seen in Figure 36, sketches “A” and ‘LB”, the of your dies almost immediately when using tool
recommended total tool stack on this extrusion press
stacks like the one just mentioned (see Figure 38).
is 13VZ inches. The manufacturer set up the total
thickness of steel necessary to withstand the pressure Pressure Ring Not Flat
of that particular press. It is up to the extruder to make If your rvessure ring is not perfectly flat, you will have
the best use possible of that given distance in select- the proolem of excessive deflection even if you have a
ing his tool arrangement. It is also understandable that properly arranged tool stack.
the extruder will try to arrange his tooling sizes as Sk~tch “A”, Figure 39, shows the tooling with no
economically as possible. But he can overjo It, and in pressure applied to it. Th? pressure plate is worn out
the long run spend more money on tooling than if he on the edges aro~nd the opening. Consequently it
starts with more solid tool stack, even though such forms a gap between the bolster =md the pressure
might have a more expensive initial cost. plate itself. On sketch “B”, the tooling is under
In sketch “A”, Figure 36, you will see a very solid tool pressure and it is forced to fill the gap between the
stack made of the dies, backer, and a heavy bolster, In bolster and the pressure plate. When this happens,
Figure 37, sketch “A”, there is the same size tool dies will also develop cracks on corners and yOJ will
stack with the same pressure applied, but this time have a high rate of broken backers and bolsters also.

102
 SPACER RINGS
12” DIA. BOLSTER

1 PRESSURE

~ ~ 16;’’”1 GAP FORMED BY

W OUT PRESSURE
PLATE

B
SPACER RING
-1 TOOLING UNDER
B

PLATEN
6“ BILLET r
1

— (—

DIE
/ .

—
BACKER J J ‘ “
12“ DIA BOLSTER L PRESSUREPLATE
FIGURE 39,
FIGURE 37.

pRESSURE’ PLATE

r
A
DIE
‘e

177
n’ CRACKS
WILL
DEVELOP v- V
IMMEDIATELY
FROM
12 “ DIA. BOLSTER I OPENING

EXCESSIVE B
DEFLECTION
DEFLECTIO/V CAUSED BY
_o/v DIES 4
ING
FIGURE 38.
URE

rO
Relation of Platen Opening to Billet Diameter
r
Another situation that will cause caving and breaking ‘ER
of an excessive amount of tooling, and at the same s
time will require excessive die correction, is when the
opening on your pressure plate or platen is excessively
larger, than your
“A” and “B”, )
billet size. (See Figure 40, sketches UNDER
PRESSURE I
suPPokr
There are limitations to the size of opening that you IH OF ONLY
may have in the pressure plate or platen in relation to - 1 % “ INSTEAQ
the billet diameter that is going to be pushed. When OF 21/2“.
an extruder, in order to produce a larger extrusion, A DIFFERENCE
goes to an extreme by opening the pressure plate too OF % “ LESS
FIGURE 40.

103
 much, he is asking for trouble. In many cases when
this is done he will ruin not only the die he is using to LARGE DIE DIAMETER
push the extra large extrusion; but also the rest of his WILL OFFER
tooling at the same time. The extra deflection on the PROPER SURFACE
tooling will cause most of his dies to have to be
OF SUPPORT
corrected all over again, eventually.
TO THE DIE
Causes of Flare I CONTAINER f---
All of the deflection problems we have mentioned
because of deflection will increase the cost and lower
the quality of production. They will also cut the life of
your dies because of overcorrection of the same.
The improper seal of a container against the face of
the die will cause the billet to back-extrude or flare
around and between the die and die container (See
Figure 41).

BILLET

c
FIGURE 42.

Another cause of flares when extruding is the

pressure of the pumps not being great enough to hold
the container against the die. Flares will develclp in
this case, even if the liner is in good condition. In this
case the flare will start to come out at the end of the
extrusion cycle (See Figure 43, sketches “A”, “B”,
“C”, “D” and “E”) when the extrusion pressure
increases.
As seen in the sequence, the pumps on this example
are not developing sufficient pressure to hold the con-
tainer against the die when the billet is first sti~rted to
be extruded. The drag of the billet against the con-
FIGURE 41.
tainer liner will force the container forward against the
die, helping in this way to make a proper seal. As the
One of the causes of flare when extruding, is indenta- billet becomes shorter, toward the end of the extru-
tion or worn-out container liners. Indentation of a con- sion cycle, the drag against the liner will be at a
tainer liner happens when an extremely small diameter minimum. It is at this point that the die will separate
die is extruded in relation to the size of the opening from the liner and the flare will start to appear
on the container liner, (See Figure 42, sketch “A”. )
As seen, the excessive use of a small diameter die in Worn Out Liners and Dummy Blocks
relation to your container opening will eventually ruin
A washed-out container liner will give a die corrector
the container liner. When you set a large die, or the
problems on runout and in some cases the extrusion
pressure on the pumps that seals the die container
will tend to develop twists toward the end of the
against the die drops, the Aluminum will work its way
extrusion cycle.
out and cause flares.
Worn out or chipped dummy blocks will put indenta-
G tions on the liner, and eventually will ruin it if not
corrected.
When a dummy block starts to chip off from handling
or fatigue, the small particles that chip off eventually
will reach the bearings on the dies and consequently
ruin any bearing they come in contact with. When
ER chipping starts, the dummy block should be replaced
immediately.

Nitriding of Dies
One of the most important parts of die correction is to
get the tools set and ready for the nitriding process.
Usually, dies must be corrected to the point where

t’ A
you are positive they are producing
acceptable tolerances.
extrusions with

SMALL DIAMETER When sending dies to be nitrided you also m~lst be

DIE WILL WORK sure that bearings are free of file marks and that they
are highly polished.
ITSELF INTO THE
CONTAINER LINER After these steps have been taken, you should run

104
 whole surface of a die that is going to deflect
A
B excessively is to have this die develop corner cracks as
DUMMY
BILLET soon as you extrude ten or fifteen billets. There are
BLOCK
cases in which corner cracks will develop right from
r the first billet. (See Figure 44.)

II/ah he
Q;, crc#cks
devdop
on corlners
DECREASING due to the
MAXIMUM DRAG DRAG extremely
hardened
c
surkfee
CONTAINER by the
r- n ildding
process.
FIGURE 44.

Cracks develop on the surface of nitrided dies due to

the extremely hard surface created by the nitriding
1I 1 process, The hard crust of the nitrided die is usually
STEM 1 from 0.003 to 0.006 inch. The center of the die remains
DECREASING softer, thus having more elasticity than the crust.
DRAG When pressure is applied to the die and deflection
D occurs, the inside of the die yields but the surface
cracks due to its extreme hardness, See Figure 45,
DIE
sketches “A” and “B”,
When the nitrided penetration exceeds 0,006 inch then
the deeper it penetrates the more the danger of hi~ving
bearings flaking off. The flaking of the bearings
occurs when the excessively thick nitrided surface
cracks and peels off as the bearings pick up heat
while extruding.
‘-’-l As you can observe from Figure 45, sketches ‘(A” and
MINIMUM DRAG “B”, when pressure has not been applied to the de
the tongue and die plate are straight. When pressure
is applied and the plate is deformed, your nitrided
FLARE E
penetration (or crust) will tend to develop cracks, i~nd
in some cases these cracks will keep going deeper
EXTRUSION until the die breaks completely.

Nitriding of Hollow Dies

5 [ On hollow dies, the nitriding should be done only on
the bearings. The hollow die deflects more than a
solid die. The difference from front to back of an
extrusion coming out of a one-hole hollow will have
very little variation as far as dimension is concerned.
FIGURE 43.
The reason for the uniformity of dimension in hollow
die extrusion is the fact that when a one-hole hollow
anywhere from 500 to 1000 pounds of extrusions deflects, it is almost straight forward and not an
before the die is set and ready for nitriding. It should uneven deflection like the one you will find on solid
also be specified if the nitriding will be on the bear- dies due to different tongues, walls, etc.
ings only or on the whole die. Figure 46 shows deflection on a hollow die, sketch
If the nitriding is only going to be on the bearing sur- “A” shows a taper hollow without pressure appliecl.
faces, then you should specify it so the dies will be Sketch “B” shows the same die with pressure applied
copper-plated and the bearings polished, or what we to it. As you can see, the deflection is forward and
term as selectively plated, straight on this particular one-hole die.

When nitriding solid dies, most people will specify

nitriding of the complete die. If the dip is of an When you are dealing with multiple-hole hollow dies,
average shape with heavy enough wal, so that deflec- then the deflection (even though it is straight forward)
tion will be at a minimum, then this type of nitriding will tend to separate the two mandrels and create
will prove satisfactory. But on the other hand, if the what is called shifting, There are many cases,
die to be nitrided has light walls, high ratio, or deep especially when extruding sections with high ratio
critical tongues, then the nitriding should be on the extrusions, that you have to push at the maximum
bearings only. Nitriding of the complete die on this pressure. On these kinds of sections, when they are to
category would be to ask for trouble. To nitride the be run on multiple-hole hollow dies, the holes are built
105
 A B Pressure
Applied

Cracks will
Penetration of Die
develop due
nitrided surface 10 excessive 1 /--
deflection– ‘/ /
\
e

L
The whole
will cup and
di

tongue will t
L Die
applied.
with no pressure to cave in.
Sort of a Deep
I
cracks due to
diving board. hinge action movement

FIGURE 45.

B
Cap
~ Cap Core

Core r-
/’ /’ ./
/ N’

Pressure _.— —— ~
Deflection is
straight forward
so mosf extrusions
will keep good
contour or

L’ I ‘- ‘--–-
----1
‘ angularity
dimensions
ond

I more accurate.
Deflected tooling

FIGURE46

with the walls towards the center of the die smaller than the walls on the outside. As pressure is applled
than the walls towards the outside of the die. to this die (sketch “B”), and the deflection occurs, the
cores will move toward the outside and WIII clclse the
As you can see in Figure 47 sketch “A”. the walls on
outside walls At the same time this is happening, the
the tapered hollow die in question are open to the inside Wall will increase In thickness and inside and
same dimension. As soon as the die deflects, the outside walls WIII end up with the desired wall
inside walls are heavier than the outside ones. thickness.
Sketch “B”.
The examples I have mentioned about deflection and
shifting in hollow dies pertain to one type of hollow
Figure 48, sketch ‘“A” shows what is done to certain die only, which is the tapered type hollow die.
dies to avoid the difference in wall thickness when
extruding multiple-hole hollow dies that will shift
There are other types of hollow dies In wh(ch shifting
under excessive deflection.
IS even more a problem, due to excessive deflection
As you can see, the walls on the Inside are smaller These are: bridge type hollows, housing type tollows.

106
 A B

Core Cap
cop
/ -Core F“—

w
Pressure

t!i!i]
\
~
Heavy
Inside
walk
—

Lk
Light
Outside
walls

Uniform
Wall
openings

FIGURE 47.

~aperecf Type Ifollow Die

Cap J
Cap J
A B

F/GuRE48.

and porthole type hollows. The multiple-hole hollow larger until the core cracks completely. Actually, what
dies that present the least problem as far as deflec- happens here is that when extruding a shear action is
tion IS concerned, are the pancake type hollow dies created between the center of the core and the sup.
and the Insert type hollow dies. The pancake presents port of the pads against the cap and bolster (See
less problems because each hole on the die is built as Figure 50).
an independent opening and the insert die because it
IS built and located In a separate piece of steel called When dies have been set, corrected, and nitrided,
an insert holder (see Figure 49). whatever correction has been done to the die will
remain without necessity of change for a longer
period of time. This is due to the fact that a nitridl?d
Because of the excessive deflection mentioned, bearing is hard enough to resist wear from heat and
hollow dies should always be nitrided on the bearings
friction when extruding for longer periods of time.
only. If this is not done, cracks will develop under the
shoulders of the cores and every time the die is A nitrided die should be run much like any die that
p~shed these cracks will keep on getting larger and has not been nitrided; that is, you should only run a

107
 Pancake Die

Openings will deflect separately

Pressure Concentricity ccm be held with
more accuracy

FIGURE 49.

certain number of billets and then pull the die to of consecutive billets that you are able to push
replace it with a back-up or spare die. Or you should through a nitrided die before pulling it out froln the
just pull the die to clean it, polish it, and let it cool press to let it cool, is always greater than the number
before setting it up for production again. The number of billets you could push if the die was not nitrided.

Die Ring
Core ~ 1

—IVitriding should
only be done on
these surfaces,
/caving the rest of
the Die softer to
avoid cracks under
shoulders

Cap -/ ‘Bolster

FIGURE 50.

108
 When pressure is
applied and deflection Nitriding penetration
occurs fhe un-nitrided on bearing surfaces
surfaces will keep Iheir only
elasticity und give
more before cracking

pressure

FIGURE 57.

The number of billets to be pushed through any die If you have dies which you know will be of high pro-
before taking it out is determined by the size of the duction, then you must have one, two or even four
opening of the walls. If it is a light wall die the bear- back-up dies to increase die life and to maintain pro-
ings are so small that they will heat up fast as heat duction without having to wait too long for new
builds up while extruding. The condition will reduce corrections or for the production of a new die.
the number of billets that can be pushed. If the die is
To avoid cracking on corners of nitrided critical
of heavy wall openings, then the die bearings will run
tongue dies, some die correctors will “mask off” the
cooler because of the easier flow of the larger volume
corners that will have a tendency to crack, and in this
of aluminum going through the bearings. As you
manner avoid the nitriding of any particular corner. If
extruded such a die, heat will take longer to build up,
the nitriding does not penetrate on the corners, ihen
thus allowing a large number of billets to be pushed
they will remain softer and last longer before cracking
before having to pull the die to cool it down,
(See Figure 52).
Always keep in mind that even when a die is nitrided
there are rules to follow to keep a successful high Avoid Bad Practices
production per-die operation. There are bad practices in die correction that should

Nitrided Penetration

Mask Corners
and end of legs
to avoid premature
cracking

109
FIGURE 52.
 never be applied. One such practice is that of moving beads might interfere with the butt shear, the method
the die slide to the right or the left side in order to is not recommended for solid dies.
correct the flow on any given die. If the flow is fast to
the right, the die slide is moved to the left to make the How Chokes, Reliefs Affect Flow
flow of metal easier towards the slow side. If the fast As soon as a new die is received from the die shop, a
side is to the left, then the opposite is done with the test billet is run to determine if corrections will be
die slide to correct the flow. necessary. Indications usually show up immediately.
Always remember that when a die is running fast in One common trouble is failure of the metal to fill
one side and slow on the opposite side, the best thing sharp corners on small openings in the die. Choking
to do is to balance the flow by correcting the bear- the bearings on the larger portion or relieving the bear-
ings, at the press or on the bench. Never meddle with ings on the smaller will often correct the condition.
the die slide. Variously known as bevels or “speeds”, reliefs ci~n be
termed long or short. The degree of slope is deter-
CHANGING FLOW CHARACTERISTICS mined by the amount of correction required and ‘will
Acceleration of metal flow in specific areas will result range from a few thousandths to about 1/64 in. A long
when bearing surfaces are shortened. The latter is speed is sloped on a breakaway edge of the bearing,
accomplished by providing relief (speeds) at the break- and extends from about 1/3 the distance from the
away point, by various grinding techniques, and entering edge at the front to the back of the bearing.
other means. A long speed tends to draw the metal, directing the
flow toward the slope.
Shortening a bearing may be done with a carbide mill-
ing cutter (approximately 1/8 in. diam.) in a high speed If speed at the breakaway point is insufficient, a long
rotor to remove the metal at the breakaway point. This choke on the entering edge of the opposite wall will
speeds extruding metal in that area. Any milling on generally accomplish the required results, Such a
the back of the bearing should leave sufficient step choke is made from the entering edge to within l/3 of
for further correction if needed. As an example, if a the bearing length from the breakaway point. Thu:;, the
heavy section joins the bearing being cut away, the two overlap for about 1/3 the bearing length, creating
consequent speeded-up section may tear away from a channel pointed in the direction of flow, throu~lh the
the heavier area. In such instance. the bearing height bearing. As this results in an opening somewhat larger
must be sloped from the outer edge of the thin leg to than the original, the new extruded section must be
the area of increased speed. Such areas can also be checked for thickness as soon as a test run is made.
cleared by the use of conventional milling, using
Short chokes and speeds use bevels of approximately
tapered 11/2 in. to 7 in. cutters. 45°. These only go into the bearing about 1/32 in or
Another way to speed extrusions at certain points is less. A short choke has a tendency to slow the
to grind the face of the die in the areas that require aluminum flow through the bearing, directing the
faster flow, thus shortening the bearing length. metal toward the side on which it is paced. A short
However, die life also may be shortened by this speed will increase the flow on the bearing on which
procedure. it is filed and direct the extrusion away from thai side.
Such surfaces should never overlap as flow effects
The legs of a “U” channel tend to close in if the web
will become reversed.
is extruded faster. or open up if the web is slower
than the legs. A marked acceleration is obtained when Because of their short bearing lengths, short chc)kes
only a few thousandths of an inch are ground off the and speeds are difficult to use with sections that have
face of the die at the points where speedup is needed. thick walls. While such chokes provide a quick
However, this method is not as reliable nor can it be method of correction, they wear rather quickly.
predicted with the certainty of taper filing and the
For long or short choke or speed techniques, there
effect will not last as long. Additional corrections like-
should be at least a 0.015 in. step back of the bearing
ly will be required during a run, but grinding can be
before the relief begins, otherwise there may be dif-
helpful where an inadequate step exists back of a
ficulty in obtaining enough taper. Therefore, an
bearing or where filing is difficult. On wide, thin, flat
extremely thin, flexible file is required so that it can
pieces, or on multiple channels, care must be exer-
be bent at a great enough arc to reach the surface re-
cised to guard against taking off too much of the die
quiring correction.
face.

Slowing an area can also be accomplished by grinding Some dies need choking slightly to provide an even
several shallow grooves on the die face, parallel to the bearing where the die sags under pressure. (See
orifice and about 3/16 in. from it. The slope, depth and Figure 53,) This is accomplished with small die open-
distance from the die opening will determine groove ings by filing to size until the template enters, but
effect on the flow. Although not used often, this does not pass through. For rod and large sections,
technique can be employed when a bearing has been about 11/2 in. choke is prov!ded. Too much chok(~ can
altered to the point that there is insufficient metal left raise extrusion pressure considerably and shoul~ be
for further correction. avoided.
When correcting a die, it is often desirable to file so
Controlling Flow in Hollow Dies
that certain extruded surfaces are concave because
The juxtaposition of the spider and the die opening or subsequent stretching may tend to curve them irl the
openings and the thickness of the spider itself are opposite direction.
important factors in controlling flow in a hollow die.
Reducing thickness of the spider will generally speed Supporting Runout
up areas that are starved for metal.
Obtaining consistent results for the full length cf the
For slowing hollow dies, one improvisation sometimes runout is one of the main problems in die correction.
used is to weld two beads of metal to the tide face, Properly trained personnel is a requisite and
parallel to the hole that needs slowing. Because the sometimes mechanical aids are essential; however, a

110
 runout problem often may be solved by the simplest heating, fast or slow heating, water quench or varying
procedures. ram speeds. Because these are not constant the con-
Light shapes that have little rigidity are aided by tour of a shape can be different from the lead-out end
graphite canisters. These contact the back of the die of an extrusion to the butt end. These also contribute
and do not allow the soft extrusions to fuse. When to the spin or ripple appearing at the butt end of cer-
fusing does occur and only one string runs faster than tain shapes. The operator must work within these
the other, a plugged die and stoppage may result. limitations and seek to overcome difficulties by other
means than die adjustments alone.
Distortion can be held to a minimum, where the metal
is still soft, by using graphite separators or guides for
Twisting, Ripping from Uneven Fiow
multiple extrusions. Power-driven conveyors often are
When different legs of a section run at different
used to lessen drag on the runout table. However, the
speeds, a corkscrew action takes place. Simple ‘filing
conveyor and table top should be about the same level
of the lip, as shown in Figure 54, will choke the metal
or distortion may be caused rather than corrected. The
as it enters, thus slowing it down where requirecf.
use of carbon blocks or canisters to “bridge” up the
extrusion as it runs out from the press can do much Lubricating certain areas with graphite to speed up
to control the snake-like undulations that sometimes flow is sometimes sufficient, but not recommended
occur. These are caused by the tendency of the extru- except in extreme cases, results being generally
sion to maintain a straight-line relationship with the erratic. In cases where a leg rips off because it
die OP&IinQ. h is more pronounced on some of the emerges too slowly at the start, lubricating only the
heavier sections. The extrusion supports itself as a slow area often frees flow until the shape begins, run-
cantilever beam until its own weight pulls it down. ning smoothly. A leg rips because of (1) being thinner
Then it immediately starts to cantilever again. This than adjoining sections; (2) it has a sharp change in
goes on throughout the length of the extrusion in a cross sectional area, or (3), it runs too fast in the area
series of bows. A block of carbon used to prevent the where it fractures and too slow in an adjacent ar,aa. If
“falling” of the shape can very often eliminate the a leg comes off before emerging from the die, stop
situation by supporting the hot metal until it is suffi- ping the press and starting up a second time may
ciently cool to carry its own weight. Such support is bring it out.
also necessary in many cases to give satisfactory con-
The spin or ripple on the butt end of an extrusion can
tour results. As the metal is subject to distortion as it
be minimized by increasing the butt length, by cc)oler
leaves the die in a hot state, this can be used to
billets, or by slower extrusion speed at the end of the
advantage in correcting initial distortion. Experiments
runout. Spin or ripple is brought about by the increase
with various shapes will show which action is
in the heat of the billet and die due to friction and
necessary in each instance for best results. However,
metal working as the butt end is neared. At the same
sufficient support in line with the die opening may be
time, the flow pattern in the billet changes from a
required in addition to any corrective measures. Also,
relatively linear pattern, perpendicular to the die face,
notes should be kept of the bridging used to obtain
to a more curved and oblique pattern as the butt
desired results,
There will be considerable dishing in the extruding of
some shapes if there is no contoured bolster provided
for sufficient support. This is caused by extrusion
pressure on the face of the die and varies with the
temperature and direction of the metal flow entering Iil
the die. The metal temperature varies as the billet is
extruded, changing the flow pattern, which, in turn,
causes change in the die pressure and dish. Control
of the die dish, to some extent, is possible by chang-
ing metal temperature or flow pattern with taper

FIGURE 53.
Where the die sags under pressure, slight choking is neces-
sary to provide an even bearing surface. At the left is shown an FIGURE 54.
urschoked die opening, not under pressure. The seeond illustration
depicts an exaggerated result of sagging ( die deflection under pres- By choking the bearing as shown at “A” the metal flow
sure ). In the third drawing the aperture has been choked; under is slowed. By shortening the bearing, “B,” with relief ( sp~ed ),
pressure, as shown in the fourth cross section, the bearing surface “C,” the extruding is accelerated, Back taper is shown by “D.”
becomes level. Bearing system at right is sometimes termed a “crown” bearing.

111
 length is decreased. Controlling any one of the above quenching of the billet. The billet is heated to the
conditions may effect but small change, still, the pro- desired temperature for extruding and the back end of
per combination can correct twisting, ripple or ripping. the billet is quenched with water to produce a tempera-
ture gradient between front and back. In any event,
Temperature Factor in Extruding the corrections should not be made unless still
Excessive wear, deflection, springing, poor tolerance required after temperatures have been properly
and similar problems can be the result of extruding at adjusted.
unsuitable container, billet, and/or die temperatures.
Heat adjustment, on the other hand, often can Tool Failure Characteristics and
eliminate the need for die correction due to those Preventive Measures
factors, if caught in time. Although aluminum extrusion dies or componenl tools
Thickness and contour of the shape being extruded seldom fail in service, when the correct grades of
affect billet temperature. Trial runs will determine steels are used, there may be occasional difficulties
whether a hotter or colder billet is more satisfactory. encountered before an extrusion has run its course.
Containers should normally be 50 to 150” F cooler than Table 1 on page 116 gives probable causes and cor-
the billet. rective measures, from the standpoint of steel selec-
tion and treatment.
When extruding alloy 6063, billet temperatures are
determined by the exit temperature of the extrusion Most probable types of tool failure include:
from the die. To start the die when extruding 6063, it ● dies—cracking, caving, tongue breakage.
is recommended that either a billet of slightly elevated
temperature (50 to 75° F above normal operating ● backers—caving, tongue breakage.
temperature) or a short-length billet (1/2-2/3 standard ● bolsters—cracking, caving, tongue breakage.
length) be used. This will insure that the die will not
“plug” or “hang up” due to a cold billet. After stabiliz- Other failures in tools include splitting, wash (wear or
ing the die by running two or three billets, the temper- erosion of bearing surfaces resulting in opening of the
ature may be reduced to a point that will insure that die orifice), loss of size, heat checking, upsetting,
the exit temperature of the shape is 930”F or greater. closing in (closure is caused by flow of die material
This will produce the best mechanical properties for resulting in decrease of opening size), bending or
6063 alloy extrusions. Exit temperature requirements pulling.
vary, depending upon alloy, die, section, subsequent
heat treatment, other process factors and end-use Wear, closure, breakage and other conditions
specifications. influencing tool life are caused by tool hardness, tool
support, alignment, extrusion shape, billet tempera-
Tube @ctions may require higher billet temperatures ture, metal composition of the extrusion, extrusion
due to the higher reduction ratios and increased hot ratio, extrusion speed, surface conditions and
work. The ideal operating temperature for each die and lubrication.
alloy may be developed in the above manner.
In using a radial layout, slight trouble may be Sub-standard steel—containing inclusions, seams or
encountered in the flow differential between the cracks or a bad center, can cause die failure. When
center and outer edges. When the center is running hardness is held on the low side of the wide range
too fast, increasing the temperature of the container (about 42”C) premature failure may be the result of
will often balance the flow. When the outer edges of excessive wear. If hardness approaches the top side
the extrusion are fast, reducing the container of the range (about 52 °C), cracking or breakage may
temperature may help. result. Die hardness is more often a compromise
between high hardness, providing maximum wear
resistance, and low hardness, that is less likely to
Billet temperature control can often be employed to
check and break, but more likely to wash.
offset tongue deflection—a colder billet tends to
close and a hotter one to open a “U” channel, for
example. Taper billet heating can also help control Most dies fail because proper support is lacking when
contour or size of the shape. their elastic limit is reached. H is very important to
support a die properly for anticipated operating condi-
tions. Misalignment of extrusion tooling is a common
Taper Heating mechanical factor that results in premature failure,
During the extrusion cycle the shape tends to ending in poor tolerance, worn, sprung or broken dies.
increase in temperature due to the generation of heat
by frictional forces in the container and at the die. Many dies are worked out of size prior to actual ser-
This may cause a problem when extruding hollow vice, rather than worn out in operation. This is due to
shapes as the increase may be as much as 100” F or the operator’s endeavor to get the die started and
more. The increase in temperature is also affected by make it serviceable for immediate or special opera
the reduction ratio and speed of extrusion. If the tional conditions. Seldom can two “identical” ciies be
temperature increase is too great, hot shortness and “started” by any standard predetermined procedure.
tearing of the shape may be encountered. This This is one of the leading causes of wide variance in
increase in temperature may be overcome by taper die production.
heating the billet to allow for the heat buildup. The
typical practice is to taper heat the billet so that the EXTRUSION SHAPE can also have a major influence
back end of the billet is 100 to 150° F cooler than the
on the life of a die. When round bars or tubes are
front end. This can easily be accomplished in induc-
extruded, optimum service can be expected. TPle more
tion billet heaters with two or three stage coils by
complex a shape, the less die and mandrel life is like-
reducing the power input to the last stage or two. ly. Sharp corners and thin sections, especially, are
The same effect as taper heating can be accomplished more conducive to wear, partly because of the short
with gas-fired billet heaters by resorting to taper bearing lengths necessary.

112
EXTRUDING TEMPERATURE, a factor in the life of a SOLID SHAPE (FLAT DIES)
die, is determined somewhat’ by resistance to defor- B
● Improve surface finish
mation of an alloy and the extrusion ratio or percen-
tage of reduction for a particular shape. Temperature ● Reduce adherence of aluminum to steel
extremes are not necessarily detrimental to a die, as
long as variances are not pronounced or rapid during a . Caustic used to clean aluminum oxide from
bearing surfaces and bearing undercuts,
single push. Factors such as friction and upset cause
increased heat to be generated in a billet, uniform
heating of the latter is not always sought. Consistent
temperature at the face of the die is necessary for A lubricant is often used in starting an extrusion
optimum results, as well as maximum die life. Taper through the die, particularly where the shape com-
heating is often employed to attain both. Keeping the bines a relatively heavy element with light, thin
temperature differential low is also aided by proper elements. Metal moves more rapidly through the larger
preheating and cooling of the tooling components. For areas of the die aperture with a tendency to starve in
most alloys, tools are preheated from about 400° F to the smaller openings. Swabbing the die with Sllitable
about 900” F. For alloys in the higher temperature lubricant tends to prevent starving in the smaller
extruding ranges, tooling should be heated proportion- opening and once started the metal tends to flow
ately higher. evenly.
Improper use of die lubricant is one of the main
STRESS RELIEVING OF STEMS periodically, as well sources of trouble. This is often true where the shape
as of other high-hardness tools when used without is asymmetrical and of fairly large circle-size in rela-
preheating, will add to tool life. Frequent inspection of tion to the container size,
stems will decrease down time, ultrasonic and
magnetic-particle methods are often employed. Stems It is often the practice to clear the die of metal with
are sometimes stored at about 150” F to maintain their the new billet, which pushes the previous extrusion
properties. through (the butt of the old billet having been sheared
off and removed). Where the wall thickness and con-
Exit speed from the die is known as extrusion speed. tour of the shape are such that the extrusion WIII lead
When this speed is excessive die life is lowered due itself out straight, regardless of the uneven applica-
to the higher frictional temperatures, and its effect on tion of lubricant, this procedure is satisfactory.
die hardness. This condition, when present, may However, with an unbalanced shape of relatively large
cause lateral cracks or checks on extrusion finish. The circle-size and thin walls, the extrusion which is still
same type extrusion defects can result when extruding in the die from the previous push prevents even appli-
speeds are too low, which permit longer die contact cation of the lubricant.
with the hot billet.
Present standard practice is to spray die lubricant on
Ratio of the cross-sectional area of the billet to that of the surfaces or paint it on the billet face for even
the extruded section (extrusion ratio) influences distribution. This eliminates most difficulties that
pressure on the tools, and thus regulates @ life to a were formerly related to inadequate or partial lubrica-
great extent. Billet temperature is also increased with tion; a tendency was to make a straight pass across
higher ratios. the die with the lubricant paddle, one side often get-
ting more lubricant than the other. Unless the die was
PRESSURES needed for extruding aluminum vary with swabbed carefully with a circular motion of the pad-
alloy, temperature, speed of extrusion and area reduc- dle, metal pickup on the inadequately-lubricated side
tion. Pressures range from about 10,000 to 100,000 psi. of the die became great enough, after several runs, to
At the beginning, pressure is generally high, diminish- cause the metal to twist or spin out of the die. Some-
ing as the process proceeds, then returning to higher times a smaller paddle proved more satisfactory for
than starting pressure as the butt end of the billet controlling breakout action of the die.
reaches a thickness of about 1 in. to 1/2 in. The flow
Powdered graphite in an oil or water vehicle is a com-
then stops.
mon type of lubricant used for aluminum. Howe’(er,
unless applied sparingly, it can have adverse effects
CARBURIZATION or DECARBURIZATION of the tools on the surface finish of the extrusion.
are two of the most common surface conditions con-
tributing to decrease of their service life. Nitriding of
hot-work steels to prolong their life is sometimes
practiced but his makes die correction or salvage
Die Cleaning
more difficult.
The use of sodium hydroxide for die cleaning rather
than employing emery paper, can result in a decrease
Die Lubrication in extrusion pressure, permit lower temperatures,
Aluminum extrusion dies are lubricated to improve higher speeds and less metal pickup, partly because
surface finish of the extruded section, to recuce die of the superior finish on bearing surfaces. Solutions
wear, and to reduce pressure. The lubricant also Of 10% tO Z’s”to are CLIStOmWy.
lessens the tendency of aluminum to adhere to the When tanks are employed, “standard” solutions are
steel bearing surfaces. Among the lubricants used are: the rule, but a higher concentration is necessary if
colloidal graphite and water mix; colloidal graphite wiped on. Tank solutions are heated to about 140” F.
and oil mix; graphite and grease mix; various mixtures Thorough rinsing is necessary following cleaning to
of the foregoing; and mixtures of low flash point oils guard against caustic effects on the aluminum. Pre-
or greases which leave a deposit of graphite or carbon cautions against excessive breathing of fumes as well
on the die. Caustic solution (NaOH) is frequently used as direct contact with the skin, particularly the eyes,
as a cleaning agent for the die. should be taken. Potassium hydroxide also is used for
There are specific reasons for the lubrication of cleaning but the cost can be prohibitive. Cleaning
different types of dies. caustics may stain extrusions.

113
 Structural Streaking OXIDE STREAKS—While hardly discernible on the as-
Although structural streaking is usually not objec- extruded section, oxide streaks show up as dark
tionable in the mill product or in its final application streaking areas of varying width and intensity on the
and is seldom cause for rejection, some end uses etched or anodized part. The streak can entirely dis-
require “clearer” as-extruded finishes than do others. appear or “feather down” from a wide streak to a
Where appearance is vital, especially in sections that small line on an extrusion in a very short length.
are to be buffed, etched and/or anodized, extra care is The major cause of oxide streaks is press misalign-
taken to eliminate streaking conditions, since these ment. Misalignment of the press causes oxides from
processes often accentuate the streaking. the skull (residual aluminum from previous billet) ‘to be
Streaking is most often found in the form of bands of extruded into the shape. When this problem is
“color”, which are of different shades than the base encountered the press alignment should be checked
metal, In alloy 6063, for example, these bands are before working on the dies.
usually gray to black in color. In the majority of other There are several other causes for oxide streaks,
alloys, the streaking assumes various shades of gray. including:
This is caused by variations in grain size, grain orien-
tation, or by variations in these and other details of 1. Section too close to container wall.
metallurgical structure that are exposed as thin layers 2. Container too hot, causing oxides to be extruded;
of the metal are removed in surface finishing processes and
3. Heavy skull buildup in the container,
Types of Streaking
Causes of streaking often may be identified by appear- Eliminating the Die as a Cause of Streaking
ance. Although sometimes difficult to distinguish,
There are certain corrective measures that often can
there are 3 distinct types of streaks which may be due
be used to eliminate the die as a contributing factor to
to one or a combination of conditions. These 3 types
the conditions leading to streaking. (See Figure 55.)
of streaking are referred to as:
1. Bearing streaks.
~~
2, Variations
3, Oxide streaks.
in grain size or grain orientation.
T> INADEQUATE OR EXCESSIVE FEEDING

BEARING STREAKS—Light
face of an untreated extrusion
reflection from the sur-
may reveal “pressure
lines” or “heat lines”, These are shadows which
7-
appear and disappear when the extrusion
various angles. This condition

dips or depressions
is viewed at
is often due to uneven
surfaces on the die bearing, having created a series of
across the flat surface, reflecting
nll’ -DIE BEARING ANGLE OR METAL FEEDING

~n
light in different planes. These streaks sometimes INADEQUATE FLOW OVER MAN DRI:L OR
become more predominant when the extrusions are +> “., ..-
KtWt LU
-------
AKt A (C-X-)
.
3A4A1L

buffed and anodized.

SHARP CHANGES IN BEARING also may cause
+--l EGS RuNNING SLOW DUE TO DIE
streaks, due to uneven flow of the metal or to inade- BEARING OR POOR BILLET FEIDING
quate filling of the die opening. Care must be taken to
insure that all bearing changes are smooth. Blending
of bearings with full radii will reduce the tendency for
bearing streaks on extrusions. TD~J DIE BEARING ANGLE OR POOR FEEDING

ROUNDED BEARING EDGES at either entrance or

breakway points, may cause bearing streaks on wide
flat extrusions.

aluminum
Bearing edges should be sharp and
clean to insure good entrance and exit of the
from the die to eliminate

size of an extruded aluminum

streaks.
VARIATIONS IN GRAIN SIZE ORIENTATION—Grain
alloy depends upon
Ihirill]
FIGURE 55.
) —IMPROPER FEEDING

~ $,,
OR DIE BEARING

V’, ”,,!,,, 1,,

ANGLE

, ,,

temperature, amount of hot work, and the degree of

recrystallization, Because of these factors, variations
in grain size may be encountered in different elements
of the same shape.
Close checks and controls on the repair of critical
Raising or lowering the temperature may eliminate the dies. in order to insure satisfactory bearing conditions
grain problems. However, its solution is at times prior to production runs, are necessities. Bearings
extremely difficult with certain alloys and shapes. should be flattened and polished as uniformly a~,
possible and should be either square or slightly
Streaking also may be encountered on wide, flat sur-
choked in respect to the die face. Bearings should not
faces of hollows due to the fact that the long bearings
be rounded off on either die or mandrel. In brief, the
employed cause overheating. For best results it is
die should be maintained as nearly perfect as possible.
recommended that bearings be as short as practical,
with a minimum amount of choke. Using inverted or In die design, all possible effort should be channeled
reversed bearings on hollow shapes, so that the exit toward eliminating all areas of interference to metal
edges of the bearings are all in the same plane, helps flow. “Dead” areas in weld chambers should be faired
alleviate the grain size streaking. in. For example, in the case of porthole dies, ports

114
should effect maximum shear in order to reduce heat In the event that dies or other tools are welded without ,
due to friction, Weld chambers should be as large as prior annealing, the danger of cracking is greatly
possible in the weld area, and attempts to retain the increased and the temperatures for preheating and
linear flow of metal should include the deepest pocket post heating are limited. Where such is the case, the
area attainable. As a result, return to perpendicular hardened part should be preheated between 900 and
flow with respect to the die face, following re-welding 1050” F, or a comparable temperature consistent with
under the mandrel bridge, is more readily obtained. requirements for the particular steel. The area tc) be
welded should then be heated further with the torch.
The bridge or mandrel support should shade the entire
Throughout the procedure, the welding should be
hollow shape, as nearly as possible, so as to obtain a
interrupted, if necessary, to maintain a temperature of
uniform working in the metal,
at least 6000 F.
The port and circumscribing circles of the shape to be
The cooling rate should be retarded following the
extruded should always be well within the cylinder
completion of welding. This can be accomplished by
diameter. A minimum clearance of at least 1/2 in. on
furnace cooling from the preheated temperature or by
each side should be maintained; in some instances,
burying the die in insulating compound. Another good
however, this may be totally insufficient,
precaution is to follow up retarded cooling with a
tempering operation at as high a temperature as is
consistent with final hardness of the particular steel
being used, This not only lowers the hardness gradient
Welding Extrusion Dies across the weld area but also tends to relieve stresses,
Some extruders, because of the safety hazard, are
Ideal welding characteristics include a relatively quiet
hesitant to weld any die that has undergone damage.
weld puddle and no undercutting, with good bonding
However, many will concede that this type of repair is
characteristics and minimum porosity. Porosity
practical for tube dies,
increases with greater weld puddle activity unless
When repairing chromium hot-work steel extrusion steps are taken to peen or forge the weldment as the
dies and tools, it is preferable to weld them in the welding progresses. Free machining steels, showing
annealed condition. Annealing minimizes the danger more undercutting, sometimes require an extra layer
of cracking induced through thermal stresses. This of weld material or a gradual blending with the base
precaution increases in importance as the size of the metal,
unit repaired increases, or where the weld repair is
extensive. Such annealing should employ a protective CONCLUSION
pack using a temperature of 1550” F and a slow furnace The experiences which have been related in these
cool. writings are only to illustrate a few of the problems
Although preheating the area to be welded may be that will arise when dealing with extrusion dies.
However, we want to make clear that the few seh?cted
done with the torch, furnace preheating the entire tool
between 900 and 1200° F should precede any attempt are more or less basic problems. To try to enumerate
all the problems is an impossibility for there are many
at welding, provided annealing has been previously
that are known to this industry. There are very many
carried out.
that will come up, often no one in the industry will be
Shielded arc methods are recommended for repair able to correct them until they have been worked with
welding as these produce the most consistent results and experimented with for a while.
with steels containing appreciable amounts of chrom-
The main object of these writings is to familiarize as
ium in their composition, Tungsten inert gas welding
many people as possible with the art of correcting an
is particularly preferred for repairing the relatively free-
extrusion die, and at the same time give the
machining steels, such as H13, as this method is
newcomers in the industry something of an idea and
more effective in minimizing undercutting, boiling or
base for starting work towards becoming a die
other activity in the weld puddle.
corrector.
Shielded arc welding permits the use of uncoated
Up till now nothing has been written on this subj(?ct
rods, selection depending upon ultimate requirements,
making the choosing of new personnel by the
Where the weld is in a critical location that represents
executives of any given extrusion plant a hard task,
a working area, a composition similar to the base
being almost impossible to detect if an applicant was
metal is the logical choice. Rods of H13 are generally
capable of doing his job or not. These writings should
applicable, most extruders carrying the standard
enable anyone that familiarizes himself with them, to
1/16 in., 3/32 in, and 1/8 in. sizes in stock. When the
discuss the cited problems with any person that they
weld area is non-critical and performs no functional
might be considering hiring and thereby to determine
purpose, an austenitic 18-8 stainless rod is often
his knowledge and thus avoid the costly procedure of
employed.
hiring some inexperienced individual and waiting until
When shielded arc processes are not available, elec a number of dies are ruined before letting him go.
tric arc or acetylene welding is sometimes used. A
You will find that men capable of learning this job will
coated rod, to flux the oxides formed in heating, is
take three times as long to do it by watching and
necessary with either,
guessing at it, than if they had some basic explanatory
An important aspect is postheating the die. Equally material already written like this which I have offered.
critical is that of retarding the cooling rate. This is I want to emphasize that by reading these articles,
best accomplished by burying the die in insulating people will not be experts on the matter, but basically
material, such as lime or ashes, or by recharging into this is what they should know to start on the road to
the preheating furnace and cooling slowly with the fur- becoming a die corrector.
nace. After this retarded cool, annealing the die again
at 1550” F is recommended, after which it can be
reheat-treated in the customary manner for the
particular grade of steel involved.
115
 TABLE 1
EXTRUSIONTOOLTROUBLECHART

Problem Possible Cause Corrective Measures with Respect Steel not suitable 50/. Chromium steels aretoljgher
to Steel Selection than tungsten base; use former
when possible.
Heat Carburization Use proper atmosphere Hardness to high Temper to lower hardness;
Checks Decarburization check hardening conditions
Hardness too high Check carburization; Mechanical Check alignment, clearances,
temper higher temperature; loading; avoid sharp corners and
lower carbon steel; points of stress concentration;
check hardening temperature improper support, insufficient
and time. cross sectional area.
Exposure to Consider cooling dies; use higher Excessive internal Use proper quenching; stress
excessive heat alloyed steel. stress relieve after grinding, reharden-
Excessive internal Use less drastic quench, ing, welding or hard facing
stress aircoolingfor5% chrome hot Excessive gradients Heat as uniformly as possiblein
work steels and interrupted oil of temperature preheating and in service; do not
quench on loss of color for alloy cooling under 300F in
higher alloy tungsten and molyb- service,
denum base steels; stress relieve Wear Decarburization Enrich atmosphere.
particularly after grinding, scoring Low hardness Use lower tempering tempera-
rehardemng, welding or hard ture; check hardening; check for
facing.
decarburization.
Cooling too Tungsten steels generally should
Exposure to exces- Use steels with higher red hard-
drastic not be water cooled.
sive heat in service ness; use coolant; alternate tools,
Upsetting, Low hardness Use lower tempering tempera-
Cracking Carburization Use proper atmosphere. Bending ture; check hardening.
Decarburization Misleading hardness readings
are result of decarburization, Carburized Carburized surface gives rmis-
internal hardness too high; use Ieading hardness reading for
proper atmosphere; check surface; check internal hardness
internal hardness as well as as well as surface; use prcper
surface hardness. atmosphere.
Overheating Use lower hardening temperature Mechanical factors Check alignment, clearances,
or shorter time; anneal before loading; avoid sharp cornel sand
rehardening; steels with fine poipt of stress concentration:
grain characteristics will with- improper support; insufficient
stand overheating hefter buf cross sectional area.
proper heat treatment will bring Insufficient strength Use steels and heat treatment to
out best qualities in any hot insure full martensitic structure
work steel. on quench.

116