Technical Notes

From the PTI DC-70 Special Topics Committee Issue 16 • February 2012

Short Tendon Elongations

Primary author: Kenneth B. Bondy, S.E.

Elongations are difficult to accurately and consistently predict

when post-tensioning tendons are short. One reason for this is, in
short tendons, normal seating losses are a significant percentage
of the calculated elongation. Normal variations in seating losses
in short tendons can often result in differences between measured
and calculated elongation, which fall outside the range of ±7%
tolerance permitted in ACI 318-11, Section 18.20.1, which
requires that differences between measured and calculated elonga-
tions exceeding ±7% “…shall be ascertained and corrected.” For
normal-length tendons, seating loss, and particularly variations in
actual measured seating losses, are a much smaller percentage of
the total tendon elongation and do not present the same problem
as they do in short tendons. Another reason is that any inaccu-
racies in placing the initial reference mark on the tendon tail or
measuring the elongations have a larger effect on small elonga-
tions than on large elongations. Just the width of the line made
on the strand to mark the unstressed position can be a signifi-
cant percentage of the calculated elongation. Variations between
assumed and actual seating losses in short tendons can also result
in significant differences in effective tendon force.
There is no precise borderline defining the difference between
a “short tendon” and a tendon of “normal length.” Herein, the PTI
Special Topics Committee (PTI DC-70) defines a short tendon as
one whose length is less than the tendon length affected by wedge
travel after jack release (anchor set) Lset with an assumed anchor Fig. 1—Tendon profile and tendon stress diagram.
set distance of 0.25 in. Typically, this will result in “short tendon”
lengths up to 35 ft. The purpose of this paper is to examine the parabolic in profile, each in one-way concrete slabs with thick-
ramifications in seating loss variation in short tendons and recom- nesses of 5 in. for the 15 ft length, 6 in. for the 20 ft length, and
mend measures to ensure that short tendons satisfy their full 9 in. for the 30 ft length. Tendon stresses were calculated before
design intent. It should be noted that “short tendons” can result and after jack release using standard procedures* assuming four
not only from building geometry but also by the placement of a sample seating losses of 0.125, 0.25, 0.375, and 0.50 in. for each
stressing construction joint across a longer tendon. of the three tendons.
Two remedies for the problems inherent in short tendon elon- The Committee feels that the conditions studied realistically
gations have been discussed within the Committee. One remedy represent the range of conditions found in practice for short
involves modifying the ACI 318 tolerance for short tendons from tendons. Tendons shorter than 15 ft in length are rarely used;
a percentage (±7%) to a finite length, such as ±1/4 in. The other designers are encouraged to use non-prestressed reinforcement
involves retaining the same ±7% tolerance, but conservatively in such conditions. The elongation discrepancy problems greatly
reducing the effective tendon force in short tendons to reflect the diminish with tendon lengths longer than 35 ft. Thus, it is felt that
largest possible seating loss likely to occur. A combination of meaningful conclusions and recommendations can be drawn from
these two remedies has also been discussed. the study executed herein.
To study the ramifications of each remedy, an analysis was
performed examining various parameters for three tendon lengths: *Post-TensioningManual, fifth edition, Post-Tensioning Institute, Farmington Hills,
15, 20, and 30 ft. The three short tendons were assumed to be MI, 1990, 406 pp.

Technical Note No. 16 PTI Technical Note 1

 ANALYSIS Ti = steel stress at stressing end after jack is released (ksi)
For purposes of this analysis, Fig. 1 and the following notation
will be used—all specific to this discussion. In all cases, steel Ti = 2TL − 216
stress at the jacking end of the tendon, before release, is 0.8fpu =
0.8 × 270 = 216 ksi; low-point CGS = 1 in.; and long-term losses e = base of Naperian logarithms
are assumed to be 15 ksi. K = assumed friction wobble coefficient = 0.0014
μ = assumed friction curvature coefficient = 0.07
Notation and Definitions α = total tendon curvature in distance L (radians)
L = tendon length (ft)
h = slab thickness (in.) 16a
a = tendon drape (in.) α=
L
h
a= − CGS Δ = calculated tendon elongation (in.)
2
where CGS represents the distance between the bottom of the (216 + Td )
L
concrete member and the center of the steel strand. ∆= 2 − SL
Tgrad = friction loss gradient, average stress loss per unit of 28, 500
tendon length due to friction between stressing end and fixed end
(ksi/ft) Fe = effective force per 0.5 in. diameter, 270 ksi low-
relaxation strand (kips)
216 − Td
Tgrad =
L Fe = 0.153(Ti − 15)

Lset = tendon length affected by anchor set (ft)

where 15 represents the long-term losses in ksi.
Fjack = jacking force per 0.5 in. diameter, 270 ksi low-
28, 500(SL ) relaxation strand (kips)
Lset =
12Tgrad
Fjack = 0.153 × 0.8 × 270 = 33.04 kips
SL = seating loss; wedge travel distance in anchorage after
Fi = steel force at stressing end after jack is released (ksi)
jack is released (in.)
Fi = 0.153Ti
Td = steel stress at fixed end before jack is released (ksi)
The results of the analysis are presented in Table 1.
Td = 216e − ( KL + µα )
EXPERIMENTAL STUDY
A study†
was conducted in 2011 to measure actual seating
TL = steel stress at a distance Lset from stressing end (ksi) losses in short tendons. In the study, wedge travel after jack
release SL and initial anchor force Fi were measured in four
TL = Td − Tgrad ( Lset − L ) randomly selected industry-standard anchorages using a tendon
with a length of 20 ft, 11 in. between the stressing-end and
Te = steel stress at fixed end after jack is released (ksi) fixed-end anchorages and stressed with a jack having hydraulic

Te = 2TL − Td †Report 2011-1, available from PTI.

Table 1
L, ft 15 20 30
h, in. 5 6 9
SL, in. 0.125 0.25 0.375 0.50 0.125 0.25 0.375 0.50 0.125 0.25 0.375 0.50
Lset, ft 26.3 37.1 45.5 52.5 27.4 38.7 47.4 54.8 28.3 40.0 49.0 56.6
Td, ksi 209.6 209.6 209.6 209.6 208.1 208.1 208.1 208.1 204.9 204.9 204.9 204.9
TL, ksi 204.7 200.0 196.4 193.4 205.2 200.7 197.2 194.3 205.5 201.2 197.8 195.0
Ti, ksi 193.4 184.0 176.9 170.8 194.3 185.3 178.5 172.6 195.0 186.3 179.7 174.0
Te, ksi 199.9 190.5 183.3 177.3 202.2 193.3 186.4 180.6 206.1 197.5 190.8 185.2
D, in. 1.22 1.09 0.97 0.84 1.67 1.54 1.41 1.29 2.53 2.41 2.28 2.16
Fe, k 27.3 25.9 24.8 23.8 27.4 26.1 25.0 24.1 27.5 26.2 25.2 24.3
Notes: 1 in. = 25.4 mm; 1 ft = 305 mm; 1 ksi = 6.89 MPa.

2 PTI Technical Note Technical Note No. 16

wedge power-seating capability. Each tendon was installed in a Table 2
steel tube with a load cell at the fixed end. Because there was
no friction acting on the strand between anchorages, the force Test Fjack, kips SL, in. Fi, kips Fe, kips
in the tendon was constant at each point in the strand and the 1 33.04 0.225 29.20 26.90
load cell at the fixed end also measured the force in the strand Fi 2 33.04 0.245 28.82 26.52
immediately behind the stressing-end anchorage. Each tendon 3 33.04 0.350 26.87 24.57
was stressed to a jacking force Fjack of 33.04 kips; the wedges
4 33.04 0.225 28.48 26.18
were seated, and wedge position was measured before and after
jack release. As in the analytical study mentioned previously,
long-term losses were assumed to be 15 ksi for the determina-
tion of the effective tendon force Fe. Results of the study are
shown in Table 2. and measured elongations that exceed the ACI ±7% tolerance.
Note: Stressing jacks that do not have hydraulic wedge On the other hand, a tolerance of ±1/4 in. between measured and
power-seating capabilities and that use mechanical wedge seating calculated elongations would be satisfied for each tendon length
will have higher seating losses. This type of stressing jack is not within the reasonable range of expected seating losses (0.25 to
recommended to be used when stressing short tendons, unless 0.50 in.) for all three tendon lengths.
the effects of the higher seating losses are considered in the final
effective force calculation. SUMMARY AND RECOMMENDATIONS
Whenever variations between measured and calculated
DISCUSSION OF RESULTS tendon elongations are observed, it should be standard practice
The results of the experimental study presented in Table to confirm that there are no shortcomings in the original elonga-
2 show that effective tendon forces can be determined accu- tion calculations or the standard stressing procedures. However,
rately with the analytical methods used to develop Table 1. the following factors can have a particularly significant effect
For example, in Table 2, the effective tendon force Fe with the on the measured elongation and the final effective force in short
maximum measured seating loss of 0.35 in. was 24.57 kips. This tendons:
is within 2% of the calculated effective force for the 20 ft long • Variation in seating loss;
tendon with SL = 0.375 in. in Table 1 (25.0 kips). Similar corre- • Inaccuracies or inconsistencies in placing the initial
lation between experimental and analytical results can be found reference mark on the tendon tail; and
with all of the tendons studied. The experimental study also • Inaccuracies in measuring the elongation.
suggests that a realistic value for the maximum seating loss is Refer to Chapter 6 and 7 in the third edition of PTI’s “Field
3/8 in. when using a stressing jack with hydraulic wedge power- Procedures Manual” for additional potential causes of elongation
seating capability. variation. The PTI Special Topics Committee recommends the
Knowing that effective tendon forces can be reliably calcu- following to mitigate these factors and ensure that short tendons
lated for the full anticipated range of seating losses, the ramifica- satisfy their original design intent.
tions of variations in seating losses can be evaluated. To accom- At the design and detailing stage (addressed to the licensed
plish the evaluation it will be assumed, as is normally done, that design professional and the tendon supplier, as appropriate):
the calculated elongation for each tendon length was based on a • Perform friction and long-term loss calculations using a
seating loss of 0.25 in. Those values are highlighted in yellow conservative assumption for seating loss for short tendons,
in Table 1. Then, the ramifications of different actual measured similar to the analysis presented in Table 1. These
elongations assuming seating losses of 0.125, 0.375, and 0.50 in. calculations should be system-specific and should
were evaluated. Seating losses between 0.125 and 0.50 in. consider stressing equipment to be used (power seating,
represent the maximum range reasonably anticipated in normal spring-loaded seating devices, and so on). For example,
construction; in addition, the Committee feels that a seating for the specific geometry studied in Table 1, with the
loss of 0.50 in. represents the maximum seating loss physically largest seating loss of 0.5 in., the minimum effective
possible with commercially available single-strand unbonded tendon force for all tendon lengths is 23.8 kips/tendon.
wedge anchorage systems. This was supported by the results of Thus, for this geometry, limiting the effective tendon
the experimental study. force for short tendons to 23.8 kips would account for
For all three tendon lengths, any of the studied seating all reasonable variations in actual seating loss. If this
losses different from the 0.25 in. value assumed in the elongation approach is used, the effects on the structure of larger
calculation would result in measured elongations exceeding actual effective tendon force (resulting from smaller
the ±7% tolerance permitted by ACI between measured and actual seating loss) must be evaluated.
calculated elongations. For example, with a tendon length of 15 ft, • Indicate the assumed effective tendon force for short
an actual seating loss of 0.125 in. would result in a difference in tendons, and the length that defines a short tendon, on
measured and calculated elongation of 12% (1.22/1.09-1) × 100, the contract documents.
which exceeds the ACI tolerance. The percentage differences When tendon elongations are evaluated (addressed to the
are even greater for actual seating losses greater than 0.25 in. licensed design professional responsible for approving elongations):
Similar conclusions can be drawn for the 20 ft long tendon. As • Rather than a percentage tolerance, consider an arbitrary
expected, it is seen that even small differences between actual and length tolerance, such as ±1/4 in., for variations between
assumed seating losses will result in differences between actual measured and calculated elongations for short tendons.

Technical Note No. 16 PTI Technical Note 3

 • Consider a percentage tolerance larger than ±7% for If all of the aforementioned recommendations and conser-
variations between measured and calculated elongations vative assumptions are executed by the design professionals and
for short tendons, as done by the AASHTO LRFD the tendon supplier at the design and detailing stage, variations
Standard (1998). between measured and calculated elongations in short tendons
larger than 7% can be evaluated and deemed acceptable.

Technical Note, Feb. 2012

38800 Country Club Drive, Farmington Hills, MI 48331
Phone: (248) 848-3180 • Fax: (248) 848-3181 • Web: www.post-tensioning.org
This document is intended for the use of professionals competent to evaluate the significance and limitations of its contents and who will accept responsibility for the application of the materials it contains.
The Post-Tensioning Institute reports the foregoing material as a matter of information and therefore disclaims any and all responsibility for application of the stated principals or for the accuracy of the
sources other than material developed by the Institute. The Post-Tensioning Institute in publishing these Technical Notes makes no warranty regarding the recommendations contained herein, including
warranties of quality, workmanship, or safety, express or implied, further including, but not limited to, implied warranties or merchantability and fitness for a particular purpose. The Post-Tensioning Institute
and the author shall not be liable for any damages, including consequential damages beyond refund of the purchase price of this issue of Technical Notes. The incorporation by reference or quotation of
material in the Technical Notes in any specifications, contract documents, purchase orders, drawings, or job details shall be done at the risk of those making such reference or quotation and shall not subject
the Post-Tensioning Institute or the Author to any liability, direct or indirect, and those making such reference or quotation shall waive any claims against the Post-Tensioning Institute or the Author.

Copyright © 2012 by Post-Tensioning Institute

All Rights Reserved
Printed in U.S.A.