Evaluation of Degree of Rusting On Prestressed Concrete Strand
Evaluation of Degree of Rusting On Prestressed Concrete Strand
Evaluation of Degree of Rusting On Prestressed Concrete Strand
he presence of rust on prestressing steel strand has The material standard specification for seven-wire pre-
May-June 1992 25
l 1 A
Photo 1. Strand surface before cleaning . Photo 1A. Strand surface after clean ing.
2 2 A
Photo 2. Strand surface before cleaning . Photo 2A. Strand surface after cleaning .
3 A
3
Photo 3. Strand surface before cleaning . Photo 3A. Strand surface after cleaning .
4 A
Photo 4. Strand surface before cleaning. Photo 4A. Strand surface after cleaning.
26 PCI JOURNAL
Photo 5. Strand surface before clean ing . Photo SA. Strand surface after cleaning .
6 6 A.
Photo 6. Strand surface before cleaning . Photo 6A. Strand surface after cleaning .
Photo 7. Strand surface before cleaning . Photo 7A. Strand surface after cleaning .
Photo 8. Strand surface before cleaning . Photo 8A. Strand surface after cleaning .
May-June 1992 27
to withstand repeated or fatigue loading. In many cases, a
heavily rusted strand with relatively large pits will still test to
an ultimate strength greater than specification requirements.
However, it will not meet the fatigue test requirements.
In order to evaluate the extent of pitting, the superficial
rust has to be removed. In the samples described herein,
care was taken to not abrade the strand surface below the
iron oxide or rust layer. This was accomplished by cleaning
the surface with Scotch Brite~ cleaning pads in order to ex-
pose the pits. Scotch Brite Cleaning Pad No. 96, made by
3M, or its equivalent, is a synthetic material which is non-
Photo 9. Scotch Brite Cleaning Pad. metallic. This material is available from cleaning supply re-
tailers or supermarkets for general purpose cleaning. A sam-
ple of this material is shown in Photo 9.
can expose only a very small area of the strand, and it will Cleaning is accomplished by holding a new pad by hand
take a long time to clean a wider area for better observation. and rubbing it against the strand surface longitudinally
Thus, in this paper, the use of a Scotch~ pad for cleaning the along the strand axis. The amount of pressure exerted on the
strand is recommended to remove the corrosion products pad against the strand is equivalent to that when cleaning
and expose the pits. pots and pans.
This paper describes a method to establish visual stan- After the samples were cleaned, additional photographs
dards which identify the degree of corrosion in a strand at were taken and these are marked as Photos lA through 6A.
which pitting occurs. This guide will help field inspectors AJJ pictures with the suffix "A" were taken after cleaning
make proper judgments on the suitability of strand. and are placed next to those taken prior to cleaning for ease
Visual standards are established by photographing samples in making the comparison.
of strand with varying degrees of rust as illustrated in Photos 1 These pictures can be used as visual standards from which
through 6. For publication purposes, the pictures are scaled the user and supplier may agree on the surface quality that is
down. However, the actual examination will be done by in- acceptable. Following the above-mentioned procedure, the
spectors using the unaided eye. strand in question may be accepted or rejected by comparing
Photo 1 represents a new strand with no rust and has a the cleaned surface with the picture that was previously
bright surface. Photos 2 through 6 illustrate various amounts agreed upon as the standard.
of corrosion on strand samples that were exposed to a corro- It is the opinion of the writer that Picture Sets 1 through 3
sive environment for different lengths of time and include are acceptable. Picture Set 4 is borderline and is subject to
some pits that are to be considered cause for rejection. discussion, agreement or compromise. Some engineers may
What is cause for rejection and why? find this level of rusting objectionable for critical applica-
Light rust does not harm any of the properties of the tions. Picture Sets 5 and 6 are pitted and unacceptable.
strand and it actually enhances bond. Rust alone is not a The corrosion and pitting in the center wire were exam-
cause for rejection. ined. A rusty strand sample shown in Photo 7 has corroded
A pit visible to the unaided eye, when examined as de- the outer wires with the same degree of pitting as in Picture
scribed herein , is cause for rejection. A pit of this magnitude Set 5. After cleaning with a Scotch Brite pad, the pitted sur-
is a stress raiser and greatly reduces the capacity of the strand face is revealed as shown in Photo 7 A.
1
2
3
4
5
6 .
Photo 10. Samples after testing for mechanical properties.
28 PCI JOURNAL
Table 1. Mechanical properties of strand . Table 2. Results of bend Table 3. Depth of pits on
I Load at I percent test on strand. strand. [Note : 1 in. = 25.4 mm.]
I Breaking
I Ultimate
strength, extension, elongation, Number Photomicrograph Depth,
Sample No.* lb lb
I percent Sample of90-deg. No.
-----+-
1 in.
-
I 43,800 40,000 I 5.00
No. I bends 1---
I *
2 43,700 39,800 4.95 I 15 2 *
3 43,500 39,700 5.73 2 15
4 43,300 39,600 5.2 1 II 3 *
3 14 4 0.0008
5 . 42,800 38,900 5.73
4 9 5 0.003 1
6 42,400 38,800 5.2 1
5 7 6 0.0077
* Designation corresponds to the surface conditions shown in Photos I th rough 6. 6 5
Note: I lb = 4.448 N. I * Not measurab le.
0 10 20 . 30 40 0 10 20 30 40 50
,,,,, 111111111111 I I I II I II I 111,111111 Ll ,I J ' ,, ljl I I, I 1
1 II
0 10 20 30 40 0 10 20 . 30 40 50 .
1
Jllljl'ldil!illllllllllllii 1 1111' H' IIIII IIIII I I 11111111 I I I I II I I I I! Iilii Llllt I Ill! I I II
0 10 20 30 40 0 10 20 . 30 40 50
!111 iJ II I llllllilllllll I lillll,lllll ILl I l,lillll!' !1111 llllllllllllldllll llldllllllwl1 11IH!ti
Photomicrograph 3. Sample from Photo 3. Photomicrograph 6. Sample from Photo 6.
May-June 1992 29
When the strand was opened by unstranding the outer right angles to the jaws of the vise. The speed of bending is
wires , the center wire showed that it was not pitted. A such as to avoid appreciable heating of the specimen.
closeup picture of the center wire after unstranding is shown This test is sometimes used as a measure of ductility or
in Photo 8. The center wire did not show any pitting, as toughness of the wire, but it does not lend itself to accurate
shown in Photo 8A, after cleaning with the Scotch Brite pad. duplication of values and its use is not recommended for
These two pictures indicate that pitting due to corrosion general application. 5 The results of the reverse bend test are
takes place at the outer wires , which are exposed to a humid shown in Table 2.
atmosphere, while the center wire is protected. A similar The micro-examination was done by cutting a transverse
condition exists on portions of strand located inside the reel , section of the outer wire. The cut specimen was then
which are protected by the outer layers. The outer layers get mounted and polished using standard metallographic tech-
full exposure to the atmosphere and have a heavier degree niques for specimen preparation. The polished section was
of rusting compared to the inner layers. examined under a metallographic optical microscope at a
The specimens shown in the photographs were not tested magnification of 75X, but scaled down for publication.
for mechanical properties. Another set of six samples of 1/z Photomicrographs l through 6 show a section of the wire
in. (13 mm) diameter, 270K low-relaxation strand was pre- circumference from samples corresponding to Photos I
pared and exposed in a similar environment as the speci- through 6. Measurable depths of pits are observed in Pho-
mens that were photographed. All samples were cut from tomicrographs 4, 5 and 6, and the corresponding measure-
the same reel and were exposed uncovered in an industrial ments are shown in Table 3. One small division of the
outdoor atmosphere in Jacksonville, Florida. scale imprinted on the photomicrographs is equivalent to
Samples were removed at different exposure times, such 0.00077 in. (0.02 mm) .
that the second set of samples has a comparable amount of Based on these examinations, it can be deduced that the
rusting as shown in Photos 2 through 6. It took 19 months of sample with the deepest pits will fail first when subjected to
exposure time to get the surface condition similar to that cyclic loading. The pits are stress raisers which will serve as
shown in Photo 6. initiation sites for fatigue failure.
The second set of samples was tested for mechanical The foregoing presentation and discussion show that
properties and the results are shown in Table I. The mini- strands which are pitted when observed by the unaided eye
mum requirements for mechanical properties according to should be rejected. The observation must be conducted after
ASTM A 416 are: removing the rust or the superficial iron oxide layer on the
strand surface.
Breaking strength ........ .. .............. 41 ,300 lbf (184 kN)
Several disputes in the field regarding the acceptance or
Load at 1 percent extension ............. 37,170 lbf (165 kN)
rejection of prestressing strand have been settled using this
Ultimate elongation in 24 in. (610 mm) gauge length ....... . method. Its adoption can eliminate much of the controversy
. . ................................ . ....... . . . ....... . 3.5 percent over the acceptability of lightly rusted strand.
Except for Sample 2, which broke with one wire fracture, The contents of this paper may be reduced to a simple
all samples showed maximum breaking strength with seven manual as a guide for comparing degree of rusting in pre-
wire fractures which occurred away from the grips of the stressing strands. In order to be effective, the pictures must
tensile testing machine. This explains why Sample 2 has a be reproduced clearly and accurately in color.
slightly lower elongation value. The samples after testing
are shown in Photo 10. ACKNOWLEDGMENT
The differences in values between Samples 1 through 4
The author wishes to express ills appreciation and thanks to
are within the normal variation in the testing procedure and
James Barker, Ronald Dull, Michael Muckenfuss, Don Pellow
equipment. Samples 5 and 6 showed a significant drop in
and Don Pfeifer for their constructive comments; and to H. Kent
breaking strength compared to Sample l because of exces-
Preston for his valuable comments in editing this paper.
sive pitting, as shown in Photos 5 and 6 and also in Pho-
tomicrographs 5 and 6. However, the breaking strengths are
high enough to meet the minimum requirement of ASTM REFERENCES
A 416 because the loading was axial and the decrease in l. ASM, Metals Handbook , 9th Edition, S. Y. , "Bright Wire Fin-
cross-sectional area was very small. ish," American Society for Metals, Metals Park, OH, 1978.
The effect of the degree of pitting is shown in the following 2. ASTM, ASTM Designation: A 416-90a , Standard Specification
tests for ductility, which are the reverse bend test and the for Steel Strand, Un coated Seven-Wire for Prestressed Con-
micro-examination of the wires. One sample was picked at ran- crete, V. 1.04, American Society for Testing and Materials,
dom from the six outer wires of each strand for these two tests. Philadelphia, PA, 1991.
3. PCI, Manual for Quality Control for Plants and Production of
The reverse bend test is conducted using a bend test machine.
Precast and Prestressed Concrete Products , Precast/Prestressed
One end of the wire is clamped in the jaws, which are rounded
Concrete Institute, Chicago, IL, 1985.
to a radius of 0.312 in. (7.92 mm). The wire is then bent back 4. FlP, Recommendations for Acceptance of Post-Tensioning Sys-
and forth at a uniform rate through a total of 180 deg. Each 90- tems, Federation Internationale de Ia Precontrainte, London ,
deg. movement in either direction is counted as one bend, and United Kingdom, May 17, 1991.
the bending operation is continued until the outer fibers fail. A 5. AISI , Steel Products Manual, Wire and Rods , Carbon Steel,
guide is placed on the lever so that the wire bends on a plane at American Iron and Steel Institute, Washington, D.C., 1984.
30 PCI JOURNAL