Documente Academic
Documente Profesional
Documente Cultură
Prestressed Concrete
The State of the Art
H. Kent Preston
Senior Consultant
Wiss, Janney, Elstner Associates, Inc.
Princeton, New Jersey
134
cussed under "Modulus of Elasticity."
Section 6.3 Elongation — In order to Synopsis
reach the required 3' percent elonga-
tion, it is almost always necessary to use Presents the state-of-the-art of
grips of the type described under testing procedures for stress-relieved
"Gripping a 7-Wire Strand." and low-relaxation 7-wire strand.
It is dangerous to leave sensitive ex- Specific information is given on
tensometer equipment on the strand material properties, gripping devices,
until a 3' percent elongation has been determining the modulus of elasticity,
reached. A premature failure due to a testing equipment, and short-term re-
wire weld or a gripping problem could laxation test.
ruin the extensometer. Since strands Definitions common to 7-wire strand
generally have an elongation of 4½ per- and prestressed concrete are in-
cent or more, perfect accuracy is not re- cluded.
quired for this measurement. One type Discussions of some of the testing
of device that can be used is shown in methods are given together with sug-
Fig. 8. gested procedures.
Section 6.4 — This section calls for a
retest if the specimen fails in the grip
without meeting the required ultimate
strength and/or elongation. Some
laboratories interpret this to mean that spotty or for the full length of the pack-
one retest is required and that the strand age of strand. If the loose wires can be
is rejected if the retest has a premature twisted back into place and held by
failure in the grip. Section 6.4 actually taping at the end, the strand should have
means that all tests which fail prema- its normal properties except for a prob-
turely in the grips without meeting the able foot at the beginning of the load-
required ultimate and elongation are in- elongation curve.
valid. Section 8.4 — Oil and grease prevent
Section 7.2 — The requirement that proper bond to the concrete and an oily
the center wire be at least so much center wire will often slide through the
larger than any outside wire is very im- outer wires and not carry its share of
portant. If a proper relation is not main- load. Many oils and greases become acidic
tained, the outer wires can form a pipe with the passage of time and invite cor-
around the center wire without bearing rosion.
on it and the center wire will slip and Light rust improves bond and is not
not carry its share of the load. detrimental in typical applications. A pit
Section 7.3 Permissible Variations in large enough to be seen without a mag-
Diameter — These limitations were es- nifying glass is a stress raiser. A strand
tablished many years ago in collabora- with a pitted wire can fail at 30,000 cy-
tion with the producers of strand an- cles of repeated load when a strand with
chors and gripping devices. A strand of a no pits will carry 2,000,000 cycles with-
given nominal diameter meeting these out failure.
tolerances should work in a grip
specified for the same nominal diame- ASTM A416 Supplement -
ter. Section 7.4 permits any size up to Low-Relaxation Strand
0.75 in. diameter as a special case.
Section 8.3 — If one or more wires fly The requirements for low-relaxation
out of position when cut and cannot be strand are the same as those for stress-
replaced by hand, that wire is under- relieved strand with the following two
preformed. This condition may be exceptions:
Standard carton
Strand Spliced Splice Rated
size, Catalog Color diameter, length, strength, Splice Weight, Units,
in. number code in. in, lbs sets lbs in.
3/8 TLS 1120 Yellow 0.615 52 23000 1@3, 26–b4 55 25
7116 TLS 1121 White 0.718 60 31000 1@3, 2@4 35 10
112 TLS 1122 Red 0.854 70 41300 2@3, 1@4 55 10
136
Li
ϗ
PITCH LENGTH
•
CENTER MARK
- - -y^^-- -- - - ^- THIN LAYER CF TAPE
tҟ SUBSET
I.D. TAG
of the strand being tested. This means section on gripping devices. It includes
that a reasonable number of failures a discussion of the use of cushioning
should occur in the clear between the material such as lead foil, aluminum foil,
grips. carborundum cloth, brass shims, etc.,
All of the reusable wedge grips cur- between the strand and the teeth in the
rently being used in casting yards cause testing machine grip. These methods
shear failure of wires in the grip before have been successful where special long
the required ultimate load andlor elon- jaws with fine teeth have been used.
gation has been reached. Four methods of gripping strand
None of the post-tensioned anchors which have given good results are dis-
available will consistently reach the re- cussed in the following paragraphs. The
quired 3.5 percent elongation before Sand Grip yields about 50 percent clear
failure. breaks. The PLP, the Tinius Olsen, and
Section VII of ASTM A370 includes a the aluminum insert method yield 90
L-- 3--------------
USE TESTING MACHINE "v "GRIPS FOR 3/4" DIA TO I" DIA BARS.
THESE DETAILS APPLY TO 3/8" AND 7/I6" DIA. STRANDS.
138
Fig. 4. Test assembly with Thin Line Splice in testing machine.
140
3" 3"
Iҟ ^ 7/32ҟ
Iҟ
Iҟ 1
ҟ
-SEE ILD STEEL
TABLE
3/8 23,,..
27/64.. ".025
7/16
1/2 , 3%64., .025
2 I/2' 3%4'•
9/I6
.600 19/32" .025
may be gripped directly in the U's with- the application of silicon carbide grit.
out any special preparation, it has been 2. A mixture of epoxy compound with
found that the number of clear breaks sand or grit is applied to the inner face of
decreases as the grips wear and the use the aluminum insert.
of aluminum oxide grit improves the A more recently installed procedure
performance of worn grips. uses a machine with hydraulic grips
which press the serrated teeth of the
90-deg V groove machine grips into the
The Aluminum Insert
outer faces of the aluminum angle. The
Aluminum insert angles have been pressure also forces the strand config-
used successfully in several ways. Two uration into the inner faces of the
procedures have been used in machines aluminum angles. With this procedure
that have tapered mechanical or wedge clear breaks are experienced all the
grips: time.
1. A thin layer of grease is applied to The testing machine being used to
the ends of the test piece followed by obtain these results is a 750 KN Avery
Fig. 7. Tinius Olsen SLS Grips, new liners for grips and strand that broke in clear
between grips.
142
quires a considerable degree of accuracy yield strength.
at the job site and in the laboratory. Sev- The MOE of a strand is affected by the
eral things, including variations in pis- total area of the seven wires and, to a
ton friction and gage accuracy, can cause lesser degree, by the pitch or lay of the
an error in the hydraulic jack's mea- six outside wires.
surement of load in the strand. If this ASTM A416 permits a pitch of any-
amounts to 2.5 percent of the load, then thing from 12 to 16 diameters but re-
the measured elongation must be within quires that whatever pitch is used be
2.5 percent of the computed elongation. constant for the full length of the strand.
Extremely good laboratory equipment A strand with a long pitch has a larger
and procedures are needed if the MOE MOE than a strand with a short pitch.
established in the laboratory is to always ASTM A416 gives a nominal area for
be within 2.5 percent of the actual MOE each strand but permits a considerable
of the strand being tensioned at the job tolerance with respect to actual area as
site. As indicated in the following dis- long as the specified minimum ultimate
cussion, strand fabricating tolerances load is met, Because of different rod
are such that the actual MOE of two properties and wire drawing proce-
lengths of strand fabricated on the same dures, there is a very appreciable differ-
equipment to the same specification can ence between the actual area of the
differ by as much as 2.4 percent. In ad- strand with the smallest area on the
dition, it is difficult to attach a strain market in the United States and that of
gage to a member composed of seven the strand with the largest area.
individual wires so that it will measure When wire is drawn for a prestressed
the MOE with complete accuracy across concrete strand, the shop practice of the
a relatively small gage length. strand manufacturer will specify a di-
In order to reduce the effect of shop ameter, but it must also permit some de-
tolerances to a minimum, some strand gree of tolerance. The wires of a viz-in.
fabricators furnish a typical curve which diameter strand have a diameter of
is the average of the last 30 to 50 tests on about 0.167 in. The minimum reason-
the specific size and grade being fur- able tolerance is ±0.001 in. This varia-
nished. If the MOE test of the strand tion in diameter of 0.001 in, represents a
being shipped is within 2.5 percent of change in area, and MOE, of 1.2 per-
the typical, it is within the accuracy that cent.
can be expected and is considered ade- The diameter of the bole in the last
quate. Tensioning to the typical curve die of the wire drawing operation de-
eliminates the large differences that can termines the diameter of the finished
come from testing a strand that happens wire. Since the diameter of this hole in-
to have a maximum MOE and then get- creases with wear as wire is drawn
ting a strand with a minimum MOE at through it, it is standard practice to start
the job site. a new die with a hole slightly under size
ASTM A416 requires a minimum and continue to use the die until the
yield strength of 85 percent of the ulti- hole is producing wire that meets the
mate for stress-relieved and 90 percent maximum permissible oversize toler-
of ultimate for low-relaxation strand. ance. The result is a deliberately created
This is defined as the load at which the variation in area, and MOE, of-_* 1.2 per-
strand reaches 1.0 percent elongation cent from the nominal.
and the straight line MOE does not ex- It is essential that a standard proce-
tend to these values. The load- dure for computing MOE be established
elongation curve furnished by the strand and adhered to. A complete series of
fabricator should be consulted for ten- tests on a strand includes measuring the
sions within 10 percent of the specified diameter of each wire to he sure that the
144
DIAMETER
I.2 OF STRAND
Ir^l
Iϗ O.I" I, 1
Iϗ
0.5"r > I
1
0TRj
20"
DRILL AND TAP FOR {/4"
HOLLOW CONE POIN
SET SCREW
< DRILL FOR 1/4"
CONE POINT
I--1 0.75 ^ 1
L0" 1 10..
I 4 / I
LFl 25"
A / A
Godfrey who is now a consultant in that the extensometer will hang verti-
Pennington, New Jersey. The following cally. The flexible fitting is fastened to
discussion and illustrations have been the strand by split collets which have an
furnished by Mr. Godfrey. internal diameter equal to the strand di-
ameter.
The dimensions of the collets shown
THE 50-IN. EXTENSOMETER in Fig. 9 are those of the solid sleeve
FOR PRESTRESSED before slitting with a milling cutter. The
CONCRETE STRAND cut allows the two halves of the split
collet to be pressed against the strand by
The 50-in. extensometer consists of an means of the 1'a in. hollow flat point
aluminum tube somewhat less than 50 setscrew which is held in a 1.3 in. out-
in. in length. A flexible fitting is at- side diameter (OD) steel ring.
tached to the upper end of the tube so The 1.3 in. OD ring is pivoted inside a
146
ALUMINUM
DISC
516"DRILL
y4
rite
1/4 DRILL B TAP
3/ ҟ21/4„
50 INCH EXTENSOMETER
LOWER FIXED GAGE POINT
Fig. 10, Details of extensometer for Double Dial method of measuring elongation.
does not result in any friction between is presented in Fig. 12 and a typical
the strand and the stopper. load-elongation curve is shown in Fig.
Either 1 in. or 2 in. travel dial gages 13.
may be used for measuring the elonga- Curves produced by the Double Dial
tion of the strand as the load is applied. method and a testing machine with a
The full range of 1 in. travel dial gages is load scale that permits the reading in-
equivalent to an elongation of 2.0 per- crements to be determined with accu-
cent. In general, load-elongation tests racy have given very good results. If it is
are discontinued when the elongation not possible to get a 50-in, gage length
has exceeded 1.0 percent which is the in a machine, the equipment can be de-
extension for determining the yield signed for a shorter length but some ac-
strength according to ASTM A421. curacy will be sacrificed.
The 50-in, extensometer is not de- The Double Dial test is run at a con-
signed to be left on the strand when stant speed. As each increment of load is
fracture takes place so it is necessary to reached the machine operator calls
remove the extensometer before testing "Mark" and the two assistants reading
the strand to failure. A view of the ex- the dials record the dial readings at that
tensometer attached to a test specimen instant. At the conclusion of the test, a
curve is plotted using the load incre- strand. There are several models from
ments and the average of the two dial which to choose and they can be used on
readings. Experience has shown that all types of specimens in addition to
this method gives more consistent and strand. Users report good results.
accurate results than any other method In the United States, Wallace Exten-
that requires attaching the gage points to someters are available through:
the strand. Testing Machines Inc.
Experience has shown that the exten- 400 Bayview Avenue
someters which are normally attached to Amityville, New York 11701
wires or steel bars for making auto- (516) 842-5400
graphic curves do not perform properly In 1984, for an extensometer suitable
when attached to a 7-wire strand. After a for use with 7-wire strand, Testing
considerable development effort Tinius Machines Inc. quoted a price of
Olsen Testing Machine Company, Inc., $16,565.00 plus freight and installation.
succeeded in developing and now mar- Since the test for yield strength begins
kets an extensometer that can be at- with a 10 percent load on the strand,
tached to a strand to produce good and some laboratories begin their load-
consistent autographic curves (see Figs. elongation curve at this point and then
14, 15, and 16). Fig. 17 shows a curve project it downward to zero load as a
made with Olsen equipment. projection of the line from 10 percent
The Wallace No-Contact Extensome- upward. If the strand is overpreformed,
ter eliminates the problem of slippage of the bottom part of the curve is not a
the contact points by using optical straight line and this procedure fails to
reading devices which do not touch the detect the condition. When load is first
148
24,000
% EXTENSION
22,000
20,000
8,000
16,000
14,000
In
J
4 12,000
O
J
ro,000
8,000
6,000
0
0 0.002 0.004 0.006 0008 0.010 0.012 0.014 0.016
ELONGATION, INCH PER INCH
applied to an overpreformed strand, a curve increases until the wires are all
small increase in load causes a large in- tight and the remaining curve is a
crease in elongation creating a nearly straight line to the point where yielding
flat line called a foot at the beginning of begins.
the curve. As the outer wires pull down In almost all tendon installations, both
against the center wire, the slope of the pretensioned and post-tensioned, each
150
strand is pulled to a tension of 1000 lbs
or more before elongation mea-
surements are begun. Since this load
will pull down all but very severely
overpreformed outside wires, failure to
detect the foot when plotting the load-
elongation curve is not critical. It should
be noted that the low-relaxation process
eliminates both the overpreformed and
underpreformed condition as a result of
the permanent stretching of the wires.
Under "Specimen Preparation"
ASTM A370 states "Wire slippage may
be minimized by fusing together the cut
ends of the specimen ..." ASTM A416
requires the center wire to be larger
than the outer wires by a specified
amount. This difference is required to
make sure each outer wire bears on the
center wire and grips it as there is no
other means of applying and keeping
tension on the center wire during its
service life as a tendon. If the wires are
properly dimensioned and are not lubri-
cated it should not be necessary to fuse
the ends of the specimen.
If the center wire slips during testing
of an unfused strand, the cause should
be investigated. It could be oily wires,
undersize center wire or over or under-
preforming. The author does not ap-
prove of fusing the cut ends of the
specimen. If one of the wires is in any
way out of place at the time of fusing,
the resultant load-elongation curve will
be inaccurate. Fig. 16. Tinius Olsen testing machine with
SLS grips and extensometer for strand in
place.
SHORT-TERM
RELAXATION TEST
sured as a percent of initial tension, in-
The purpose of the 30-minute short- creases as the initial tension is in-
term relaxation test is to determine creased, the short-term test is conducted
whether the sample of 7-wire strand at a high initial tension (80 percent of
being tested is of stress-relieved grade GUTS) to create as large a difference as
(SR) or low-relaxation grade (LR). It is possible during the short term of the
not intended to provide data that could test.
be used to project long-term losses In the 1000 hour relaxation test re-
under service life conditions. quired by ASTM A416, the specimen is
Since the difference in relaxation loaded to a prescribed tension and then
losses between SR and LR strand, mea- held at constant length for the duration
152
and other items of small mass will not he 9. At end of 30 minutes read and re-
the same as the effect on the testing ma- cord extensometer reading, load in
chine and other items of large mass. A strand and temperature.
maximum permissible temperature 10. Compute percentage of loss using
change, say ±3°F, should be established foregoing Egs.(1), (2), and (3),
and maintained. The specimen being
tested as well as the environment of the
testing machine must be maintained Notes on Short-Term Test
within the limited temperature range for It must be remembered that the func-
a considerable period of time prior to tion of the 30-minute test is solely to
the start of the test as well as for the determine whether or not the strand
duration of the test. being tested is the low-relaxation grade.
Test data used to make any sort of pre-
Procedure for Short-Term diction with regard to long-term losses
must be made in equipment capable of
Relaxation Test making the 1000-hour test and should be
1. Make sure that strand, testing ma- of at least 100 hours duration.
chine, extensometer, etc., have been in The P, used in the short-term test
the same constant temperature en- should be 80 percent of the GUTS of the
vironment long enough for all parts of all specimen being tested.
items to reach the same temperature. Since the change in load being re-
2. It is not permitted to perform plastic corded is a very small percentage of the
bending or straightening on the sample total load in the strand, it is important
to be tested for relaxation behavior. that the machine being used be able to
3. Place strand in testing machine and record small load increments with accu-
load to approximately 5 percent of spec- racy.
ified minimum ultimate strength. Use Maintaining a constant length for the
gripping devices with an efficiency that 30-minute duration of the test is of ex-
will provide a comfortable factor of treme importance. This means pre-
safety above the load to be applied and venting movement of the head of the
that will not allow the strand to slip testing machine. By the time the speci-
during the period of the relaxation test. men has been loaded to P, and held at
Slippage and/or seating of wedges dur- that load for 1 minute, all seating of
ing the application of P, is not detri- gripping wedges or slippage within the
mental to the test. grips should have taken place.
4. Attach extensometer to strand using The old style mechanical testing ma-
a long gage length, preferably 24 in. or chine is best suited for this purpose; as
more. long as the driving gears are not oper-
5. Load at a uniform rate to P,. Loading ated, the head remains motionless and
time should take not less than 3 minutes the operator simply balances the beam
and not more than 5 minutes and should to weigh the load as relaxation takes
be approximately the same for all tests. place. Depending on their design, hy-
6. Maintain the load atP, for 1 minute. draulic testing machines can hold a con-
During this period make any desirable stant length with anything from a fair to
adjustments to extensometer equipment a very poor degree of accuracy.
such as setting dial gages to zero, etc. Theoretically the results obtained
7. At end of I minute hold, read and using Eqs. (1), (2), and (3) should be the
record extensometer reading, load in same whether the value of P L, is rela-
strand and temperature. tively small and the value of P M is rela-
8. Maintain length constant for 30 tively large or the reverse is true. The
minutes. data from some tests suggest that the re-
154
seeds 1.20 percent, two additional tests of ASTM A416 Section 8.3 is underpre-
should be made of strand from the same formed.
pack and both tests must he equal to or Overpreformed — If the outside wires
less than 1.20 percent for acceptance. are too heavily preformed they do not fly
It is further suggested that any lab- open when the strand is cut, but there is
oratory planning to conduct 30-minute open space between some of the outside
tests obtain a length of low-relaxation wires and the center wire. This causes
strand with known relaxation properties an undesirable foot on the bottom of the
and conduct several trial runs to make load-elongation curve as discussed
sure that its equipment and procedure under "Modulus of Elasticity."
produce correct results. Stress-Relieved — When a 7-wire
strand emerges from the stranding ma-
DEFINITIONS chine which twists the six outer wires
around the center wire, all of the wires
Wire — A wire of the type used in have high internal stresses as a result of
7-wire strand is a single, round, high the wire drawing and stranding process.
tensile strength, high carbon steel unit This is called a "green" strand. The
of considerable length. green strand is passed through a heating
7-Wire Strand — A 7-wire strand is device which raises its temperature to
composed of six wires wrapped helically about 725°F for a few seconds before it
around a center wire. is water cooled. While it is at the ele-
Cable — A cable is a tension carrying vated temperature its yield strength is
tendon composed of two or more paral- appreciably reduced, the high internal
lel 7-wire strands (or wires) installed to- stresses cause local yielding and the
gether and functioning as one unit. stress-relieved strand emerges with far
Nominal Area — The nominal area of less internal stress than was present in
a strand as shown in ASTM A416 Table the green strand.
1 is the theoretical cross-sectional area Low-Relaxation — Low-relaxation
computed by dividing the minimum strand is made from the same green
specified ultimate strength by the unit strand as stress-relieved and subjected
strength of the grade specified. Thus, to the same stress-relieving elevated
the nominal area of a ½-in. diameter 270 temperature. While at the elevated tem-
Grade strand is 41,300/270,000 = 0.153 perature, it is tensioned to a high stress
sq in. Since the diameter tolerances which causes a permanent elongation of
specified in Section 7.3 of ASTM A416 approximately 1 percent. The resulting
are the only items which limit actual strand has very little remaining capacity
area, the actual area of the strand for creep or relaxation. The 1.0 percent
supplied may differ appreciably from elongation also eliminates any overpre-
the nominal area. formed or underpreformed condition.
Pitch or Lay — The pitch, specified in
Section 3.1 of ASTM A416, is the dis-
tance along the strand in which an out- ACKNOWLEDGMENT
side wire makes one complete turn
around the center wire. The information presented herein is
Underpreformed — During the based upon a study on "Testing 7-Wire
standing operation the outside wires are Strand for Prestressed Concrete," spon-
preformed so they will fit snugly around sored by the Post-Tensioning Institute,
the center wire. If the amount of pre- Phoenix, Arizona. This paper has been
forming is insufficient they do not re- reviewed and recommended for publi-
main in place when the strand is cut. A cation by the PT! Ad Hoc Committee on
strand that fails to meet the requirement Strand Identification and Testing.