Duct friction in unbonded post-tensioning can leave the force at the fixed (dead) end up to 25% lower than the force at the live (jacking) end, which is why long tendons are routinely stressed from both ends to keep mid-length force near the design value [S4].
The threshold is not a single number on a tape; it is a friction-budget decision driven by tendon length, profile (the high/low points that drive wobble and curvature friction), duct material, and the maximum acceptable deviation between design and as-built effective force. In the United States, the dominant strand for building tendons is 1/2 in. (13 mm) diameter, 270 ksi (1862 MPa), low-relaxation seven-wire strand per ASTM A416/A416M, and the friction coefficient assigned to that strand combined with the profile geometry sets the trigger [S2].
The 25% Friction Rule That Drives Dual-End Stressing
When a single jack pulls from one end, friction along the duct steals force progressively: the strand at the live end sees the jacking pressure, and the strand at the dead end sees whatever is left after wobble and curvature losses. The textbook figure quoted in PT coursework is that the dead-end force can be as much as 25% lower than the live-end force in long tendons, with most of that loss attributed to friction between strand and duct [S4].
For a long tendon stressed from only one end, the residual force near the fixed anchorage can fall well below what the design assumed, which translates directly into under-prestressed concrete and larger long-term deflections. The fix is to jack from both ends so the friction losses accumulate from each side, and the lowest tendon force ends up near the mid-length rather than pinned at the anchorage [S2]. This mid-length minimum is the geometric consequence of two friction gradients meeting in the middle, and it is the central reason dual-end stressing exists as a standard field practice.
PTI Technical Note 17: The Force Distribution Behind the Decision
Post-Tensioning Institute Technical Note 17 (DC-70, November 2013) describes the bilinear force distribution that drives the rule. For a normal-length tendon stressed from one end, the force drops from the jack, rises again through the wedge-seating influence distance after the jack is released, and then decreases again toward the fixed end. For a short tendon, where the wedge-seating influence distance exceeds the tendon length, the post-release force profile actually rises from the stressing end to the fixed end [S2].
The note is explicit on the dual-end case: "In a tendon stressed from both ends, the lowest tendon force normally occurs near the middle of the tendon length" [S2]. This is the engineering justification for specifying the Final Effective Force (FEF) per tendon and checking it after all friction and long-term losses are subtracted, rather than only checking the jack gauge pressure at the live end. The same note flags that the "constant force" assumption along the entire length (the basis for most U.S. building designs) has been debated in the literature, and any structure where redistribution is uncertain should be specified with conservative FEF values and verified on the stressing record [S2].
Which End First, and How to Read the Elongation
Sequence matters as much as the decision to use both ends. Bijan O. Aalami's September 2026 Structure Magazine article frames the three recurring field questions on stressing: which tendon group goes first in a multi-tendon pour, which end goes first on a long tendon, and how to reconcile an elongation that falls outside the target band [S1]. The article's general notes treat one-way and two-way slabs differently: one-way slabs run uniformly distributed tendons first, then beams, then girders; two-way slabs run distributed tendons first, then banded tendons, with continuous multi-span tendons stressed before any non-continuous tendons in the same pour [S1].
For the dual-end long tendon, the correct order is to jack one end to a partial force (commonly around 50% of the final jacking force, with the exact split set by the project-specific general notes), measure elongation at that end, then jack the opposite end to its partial force and measure there, and finally return to the first end to bring it up to 100% of the design jacking force and re-measure elongation. Aalami's point is that the elongation values are only meaningful if they are paired with the correct force step at the correct end, because each pass redistributes friction losses through the tendon [S1]. On the structure side, the slab is still resting on forms at the time of stressing, and standard design typically targets uplift of 60% to 80% of the member self-weight, so the stressing sequence does not have to "balance" the structure; it only has to put the design force into each tendon in the right order [S1].
Material and Strand Factors That Push You Toward Dual-End
Three strand-side factors raise the friction budget and therefore raise the probability that a given tendon will be specified for dual-end stressing. First, strand diameter: the 1/2 in. (13 mm) seven-wire strand that dominates U.S. building work has a higher contact pressure against the duct wall per unit jacking force than the larger 0.6 in. (15 mm) strands used in some bridge applications, so for the same profile geometry the percentage friction loss is higher. Second, duct type: galvanized metal ducts have different wobble and curvature coefficients than plastic (HDPE or polypropylene) ducts, and field-bent transitions in the profile compound the loss. Third, profile geometry itself: the more high and low points the tendon has to follow, the more cumulative angular change the strand sees, and the higher the curvature friction term becomes. [S2]
[S2] notes that the three common ways licensed design professionals specify FEF in contract documents are: minimum required FEF after all friction and long-term losses (in kips or kips per foot), the number of tendons of a particular size and strength with a stated FEF per tendon, or only the number of tendons with no FEF stated. When the FEF is explicitly stated and the design has been run with a tight friction budget, dual-end stressing is the construction-period tool used to hit that FEF on the as-built tendon rather than just on paper. For an unusual strand selection outside the 270 ksi low-relaxation seven-wire default, the specifier should verify whether the alternative (3-wire or 19-wire strand, for example) carries meaningfully different friction behavior; the comparison piece on strand configurations in precast and electrical specs lays out where those alternatives win or lose on criteria outside friction.
When Dual-End Stressing Is the Wrong Call
Not every long tendon should be stressed from both ends. Short tendons, where the wedge-seating influence distance is greater than the distance between anchorages, show a post-release force profile that rises from the stressing end to the fixed end; dual-end stressing on those tendons does not add any mid-length force and just doubles the labor and the wedge-seating losses [S2]. Tendons with very low total design force (small FEF per tendon) often have such a small absolute friction loss that the 25% rule does not apply, and a single-end jack with a slightly higher gauge pressure hits the target.
There are also access constraints. Both-end stressing requires a stressing pocket, a clear zone behind the anchor for the jack, and a worker who can stand safely behind that jack when it is pressurized. On a congested slab edge or a tendon terminating against a column face, the second pocket may be physically impossible, and the design has to either accept the as-built force loss or specify a different profile. Insisting on dual-end stressing in a configuration that cannot physically host the second jack is a known path to elongation records that do not reconcile, which is the third question Aalami flags in his 2026 article and the one the follow-up piece will address [S1].
Field Verification and What the Crew Actually Marks
Before the jack touches the strand, the crew paints a reference mark at a uniform distance from the anchor plate on every strand; after stressing, the mark tells the inspector how much the strand elongated at that end, and a strand that did not move is a strand that was missed [S3]. The paint-mark step is what links the jack gauge reading to the elongation number that goes into the stressing record, and it is the only on-site check that a tendon actually received the design force, because gauge pressure alone can mask a bound strand, a slipped wedge, or a duct blockage upstream of the jack [S3].
The acceptable elongation band is typically ±7% of the theoretical elongation calculated from the strand area, modulus, and free length between anchorages, with the exact tolerance set by the project specifications. If the field elongation falls outside that band, the contractor has to reconcile the record, usually by rechecking the gauge calibration, the wedge seating, and the actual strand free length, before the tendon is grouted or covered. On a dual-end tendon, the reconciliation has to be done per pass: a low elongation at the first end after a partial-force pass and a high elongation at the second end are both normal friction effects, not defects, and they are expected to balance once both ends are at full force.
Decision Criteria in One Place
The decision to specify dual-end stressing can be lined up against four concrete criteria. Length: tendons longer than roughly 100 ft (30 m) are the common threshold where friction losses start to bite hard, with the exact number depending on profile. Friction budget: when the design friction loss between live end and fixed end exceeds about 10% to 15% of the jacking force, the project usually shifts to dual-end to keep mid-length force within the FEF tolerance. Profile geometry: tendons with multiple high and low points accumulate curvature friction faster than straight or gently profiled tendons, and they trigger dual-end sooner. Access: both anchorages need a working pocket for a hydraulic jack and a safe stand-back zone for the operator; if the second end cannot host that, the design has to be revised rather than the field operation. [S2]
On the strand side, the dominant choice is the 1/2 in. (13 mm), 270 ksi low-relaxation seven-wire strand per ASTM A416/A416M, with 0.6 in. (15 mm) strand reserved for bridge and heavy-civil work where the larger diameter and lower contact pressure justify the heavier anchor hardware [S2]. The strand pattern (3-wire, 7-wire, 19-wire) is a separate decision, driven by the application's flexibility, fatigue, and electrical-conductivity needs rather than friction; that comparison belongs in the strand configuration piece rather than in the dual-end-stress trigger.
Signals Worth Tracking Over the Next Construction Season
Two trackable signals will sharpen the rule over the next 12 months. First, Aalami's follow-up Structure Magazine article addressing the third field question (how to reconcile an out-of-tolerance elongation record on a dual-end tendon) is expected to publish in the 2026-2027 cycle, and it will set the industry working position on when a re-stress is required versus when a force-and-elongation record can be accepted as-built. Second, the Post-Tensioning Institute DC-70 committee is the standing body that issues the Technical Notes governing FEF specification, friction coefficients, and stressing records; any update to Technical Note 17 or a successor document will move the dual-end threshold and the elongation tolerance band for U.S. building work. [S1]
Spec-level background on the components involved: pressure transmitter, flow meter, and industrial valve.