SEVEN-WIRE LOW RELAXATION PRESTRESSING TENDON · PDF fileSEVEN-WIRE LOW RELAXATION PRESTRESSING ... Keywords: Low-relaxation seven-wire tendon, Elastic modulus, ... the strand, a specific

223

This article can be downloaded from http://www.ijerst.com/currentissue.php

Int. J. Engg. Res. & Sci. & Tech. 2015 John J Myers and Wendy L Bailey, 2015

SEVEN-WIRE LOW RELAXATION PRESTRESSINGTENDON SUBJECTED TO EXTREME

TEMPERATURES: RESIDUAL PROPERTIES

The use of cold-drawn prestressing steel as reinforcement in concrete is common amongbridge design throughout the world. This composite material is particularly useful for designsconsisting of large spans where the dead load will cause significant cracking and deflection.Unlike mild steel reinforcement, prestressing steel is stressed and cause a compression forcewithin the concrete. This prevents cracking and increases the structure’s capacity. A prestressedconcrete member will also have a longer life expectancy due to the prevention of cracks. Withoutcracks the steel will not be exposed to the environment and therefore will be at a reduced risk ofcorrosion. The increased capacity, ability to sustain longer spans, and durability make this typeof material an advantageous choice of construction. This paper investigates the residualproperties of seven wire, uncoated, 0.5 in. (12.7 mm) and 0.375 in. (9.5 mm) diameter lowrelaxation grade 270 ksi (1862 MPa) prestressing tendon subjected to extreme temperature.The temperatures selected for the study were 500°F (260 °C), 800°F (427 °C), 1000°F(538 °C), 1200°F (649 °C), and 1300°F (704 °C). The upper limit was defined by the furnace’scapability at Missouri S&T. In addition, control specimens were tested for each strand size. Acontrol was defined as exposure to approximately 68°F (20 °C). Two cooling methods werealso investigated, namely inside the furnace and outside the furnace. Test results presentedinclude visual observations, yield stress, ultimate load, and elastic modulus.

Keywords: Low-relaxation seven-wire tendon, Elastic modulus, Extreme temperature propertyeffects, Tensile strength, Ultimate load, Yield stress

INTRODUCTIONIt may be argued that bridges are the mosteffective way to move commerce across bodiesof water or low-lying elevations. They provide

1 Ph.D., PE, Missouri University of Science and Technology, Rolla, MO.2 Burns & McDonnell, Kansas City, MO.

Int. J. Engg. Res. & Sci. & Tech. 2015

ISSN 2319-5991 www.ijerst.comVol. 4, No. 3, August 2015

© 2015 IJERST. All Rights Reserved

Research Paper

means for trade and communication to travelacross land quickly and efficiently. However, aswith any structure there lies the risk of damageor destruction, which can be attributed to a

John J Myers1* and Wendy L Bailey2

*Corresponding Author: John J Myers [email protected]

224



number of sources. Natural disasters, such as ahurricane or tornado, accidents, such as spilledgasoline tankers, or terrorism are all possible andcommon causes for damage to a bridge’sstructural integrity. Often the damage due toextreme events to the bridge is quite severekeeping the bridge out of commission for a largeextent of time.

In particular, fire damage is a common andsevere cause of destruction caused by manydifferent disasters. It is difficult to recover quicklyfrom these incidents because very little is knownregarding the extent of damage caused by a fireto a bridge. Accidents such as the Bill WilliamsRiver Bridge in Arizona as well as a number ofexploding tankers in Iraq have brought to light thefrequency of fire on bridges and the cripplingresults the damage has on society afterwards.The result is often a complete repair of the bridgewhich proves to be costly and creates problemswith traffic flow. In some cases a trade route iscompletely closed requiring travelers to travel 100miles (161 km) or more out of their way to reachtheir destination.

There are a number of studies which havebeen performed on Prestressed Concrete (PC)bridges following fire damage. However, thisresearch is either limited to the exterior of thebridge or is performed by decomissioning thebridge and testing components of it in the lab.

Internal observations and flexural strengthtesting cannot be performed on existing bridges.However, with an increase in the understandingof how fire and extreme temperature affect thebridge an educated decision on the structuralintegrity of the bridge will be able to be madewithout laboratory testing. This will result in fewerrepairs and minimize the economic and

commercial implications typically caused by firedamage.

RESEARCH PROGRAMOBJECTIVESThe objectives of this overall research programwere undertaken in three primary phases. PhaseI was undertaken to determine properties forgrade 270 low relaxation seven-wire prestressingstrands before and after exposure to elevatedtemperatures. This task was undertaken toassemble and add to the present data base tohelp understand the extent of damage whenseven-wire prestressing strands are exposed tohigh temperatures. Phase II studied the bondstress between High-Strength Concrete (HSC)and grade 270 low relaxation seven-wireprestressing strands after exposure to elevatedtemperatures. HSC was selected since minimaldata existed regarding bond stress of HSCexposed to elevated temperatures. Phase IIIimplemented the results from the first two phasesto develop an improved understanding ofPrestressed Concrete (PC) bridge behavior afterexposure to fire including thermal heat transferusing Finite Element Modeling (FEM). Firedamage to PC bridges is an occasionaloccurrence, yet investigation of the fire is still verydifficult and time consuming. This paper detailsthe results of Phase I. Future publications willdisseminate Phase II and III results and findings.

SCOPELaboratory testing included two types ofprestressing strand testing, tension (Phase I) andpullout (Phase II). Tension results are presentedherein. Tests were performed on strands whichhad been exposed to elevated temperatures andallowed to cool. The data obtained from the

225



tension testing gave an understanding of thetensile strength and stiffness properties of theprestressing strands subjected to elevatedtemperatures used in subsequent phases ofstudy.

TENSION TESTINGGeneralTension tests for steel are all governed by ASTME8-04 “Standard Test Methods for Tension Testingof Metallic Materials.” This document providesspecific detail as to how the test shall beperformed and the results analyzed. ASTM A370-07a also provides information for all types of steeltesting (tension, bend, hardness and impact) andgives specific guidelines based on different typesof bar products (fasteners, round wire, multi-wire,etc.).

Prestressing StrandsIn addition to the general specifications for tensiontesting of steel, ASTM A 416/A 416M-06 and ASTMA370-07a Annex A7 have also been published asgoverning standards for the tension testing ofseven-wire prestressing strands. Within ASTMA370-07a Annex A7 a recommended procedureand apparatus are given. Due to the geometry ofthe strand, a specific method is not required andit is acceptable to employ a method of choice aslong as the strand meets the minimum breakingstrength given by ASTM A 416/A 416M-06.Guidelines for determining the yield strength andelongation are also given by both specifications.

LITERATURE REVIEWFire Damaged Materials Found in BridgesMaterials which are commonly affected by bridgefires include the concrete, prestressing strandsand mild steel reinforcement. The amount of

information regarding the fire damage propertiesvaries by material. Within this research residualproperties of any material are defined as theproperty of the material after it has been heatedand then cooled back to room temperature.

ConcreteConcrete damage caused by fire has been widelyresearched. A significant amount of data has beenpublished which allows engineers to understandthe compressive strength properties of concreteduring and after fire exposure. Since the focus ofthis paper deals with tendon characteristics,further discussion is not presented. However,Moore and Myers8 provide a detailed literaturereview on normal strength and high strengthconcrete as it relates to fire damage that may bereferenced for further detail.

Mild SteelSimilar to concrete, damage caused by fire tomild reinforcing steel has been widely researched.The reported properties3 for grade 60 mildreinforcing steel are illustrated in Figure 1. Themodulus of elasticity was found to remain thesame for elevated temperatures despite thedecrease in tensile strength4.

Figure 1: Residual Strength vs. Temperaturefor Mild Reinforcing Steel (adapted from Dias,

1992)

226



PRESTRESSING STRANDSIn contrast to concrete, much less research hasbeen performed to understand how elevatedtemperatures physically affect the properties ofprestressing strands.

Tensile Strength of Prestressing StrandsGuyon4 reported the earliest known data regardingthe tensile strength of prestressing strands(unreported strand type) exposed to elevatedtemperatures. The research consisted ofhotstressed, hot-unstressed, cold-stressed andcold-unstressed tests. Temperatures varied bytest scenario, but no more than four temperatureswere chosen per scenario. The type of strandalso varied, 0.2 in. (5.08 mm) cold drawn, 0.2 in.(5.08 mm) rolled, and 0.1 in. (2.54 mm) colddrawn. From the testing performed it was foundthat for stressed specimens tested while heatedthere is an initial increase in tensile strength upto 302°F-482°F (150 °C-250 °C). Thereafter asignificant loss of tensile strength occurs. Forunstressed specimens tested after cooling, aconstant loss in tensile strength occurs astemperature increases. However, the loss intensile strength is smaller than that of thestressed specimens for temperatures of 572°F(300 °C) and greater. For this test program theheat soak time was also varied. In these cases agreater loss of tensile strength was seen forspecimens heated longer.

Abrams and Cruz1 performed an in-depthinvestigation of the behavior of seven-wire,stressrelieved prestressing strands andtemperature. The test program consisted of threesevenwire strand sizes 0.25 in. (6.35 mm), 0.375in. (22.23 mm), and 0.438 in (11.11 mm). Duringtesting, failure modes were witnessed to be eithera few wires breaking, followed by the remainder

of the wires breaking singly or all the wiresbreaking at the same time. Abrams and Cruz1

noted that although the failure mode varied thedata did not differ significantly; therefore the failuremodes were acceptable.

Also addressed by the researchers was therate of heating and cooling. By heating severalstrands up at various rates and then testing, itwas determined that the failure was independentof the heating rate. For the cooling analysisseveral strands were also heated up and thenallowed some to cool “fast” and “slow”. Fastcooling was defined as removing the specimensand placing them under a stream of cold waterfor 10-20 seconds until they returned to normaltemperature. Slow cooling was where thespecimen was left in the furnace several hoursuntil it reached normal temperature. Based ontension testing following cooling, it was found thatthe failure was also independent of the methodof which it was cooled. Abrams and Cruz alsoperformed tension tests on specimens at elevatedtemperatures. It was found that the tensilestrength sharply decreases at 200°F (93 °C) andcontinues until reaching 5% residual tensilestrength at 1400°F (1860 °C).

In 1967 Abrams and Erlin2 performed a follow-up to the previous research where the effects dueto exposure time were examined and hot and coldtensile strengths were compared. For thisresearch 7-wire, stress-relieved prestressingstrands were also tested. Exposure times testedwere 1 hour, 4 hours and 8 hours. For theseexposure times, the residual tensile strengthswere found to slightly decrease as the exposuretime increased. Overall the 8 hour exposure timeproduced at residual tensile strength of 90%, 60%,41%, 32% and 29% at respective temperaturesof 752°F (400 °C), 932°F (500 °C), 1112°F

227



temperature, where T is in degrees Celsius andfu is the ratio of the ultimate tensile strength at agiven temperature T, to the ultimate tensilestrength at 68ºF (20 ºC).

...(1)

A summary of the published residual tensilestrength of prestressing strands collected fromthese various studies is shown in Figure 2. Thenotation NS, SR and LR refer to the type of strand.NS is for unspecified strands, SR is stress-relieved strands and LR is low-relaxation strands.

Modulus of Elasticity of PrestressingStrandsThe modulus of elasticity was found to beindependent of temperature by Holmes et al. andMcLean. The modulus of elasticity propertyincreased slightly as the temperature increasedbut then decreased back to the undamaged valuenear the end of testing.

RESEARCH PROGRAMFor Phase I of this experimental program tensiontests were performed. The tests were performedafter the strands had been exposed to different

(600 °C), 1300°F (704 °C) and 1589°F (865 °C).Despite the extended exposure time, residualtensile strengths were approximately 40% higherthan that of specimens tested at their respectiveelevated temperature.

Neves et al. heated a single wire which wascut from the center of the seven-wire prestressingstrand. Temperatures examined were inincrements of 212°F (100 °C) from 392°F-1652°F(200 °C-900 °C). The specimens were held attheir designated temperature for 60 minutes andthen were cooled one of two ways, naturally inthe furnace with the door opened or immediateimmersion in a vessel containing water. Thebehavior of the tensile strength of the strandsinitially decreased as reported by Abrams andGuyon. However, at 1472°F (800 °C) Neves8

reported an increase in tensile strength of 8percent for the specimens cooled naturally in thefurnace and an increase of 20% for the specimenscooled by water. This result is quite different fromthat reported by Abrams and Guyon. Nevesproposed the increase in tensile strength was dueto the differences in steel composition.

A recent study performed by MacLean7

replicated the procedure of Abrams and Neves’previous research. MacLean tested single wirescut from the center of seven-wire, lowrelaxationprestressing strands. The wires were heated totemperature increments of 212°F (100 °C) from392°F-1652°F (200 °C-700 °C) and a control68°F (20 °C) and then were held at theirdesignated temperature for 90 minutes. Thespecimens were then left in the furnace to cool.The results obtained were consistent with Abramsand Guyon. Based on the experimental data anddata previously published Equation 1 wasproposed as a method of determining the residualtensile strength of prestressing strands based on

Figure 2: Residual Tensile Strength ofPrestressing Strands vs. Temperature

Frac

tion

of O

rigin

al T

ensi

le S

tren

gth

228



levels of elevated temperature. The tension testingwas used to analyze the tensile strength andstiffness properties of the prestressing strandafter damage.

TENSION TESTSWithin the tension tests there were two phasesof testing. Phase IA was designed to understandthe tensile strength of the strand at elevatedtemperatures. It also considered effects due tothe method of cooling. Phase IB of the experimentexamined the tensile strength properties due to ashorter time of heat exposure (heat soak).

Test matrices for Phase IA and IB can be seenin Tables 1-4. The Specimen ID given in eachtable is given in the format of A-B-C-D, wherethe A designates the phase number, Bdesignates the strand size, C specifies thetemperature level and D gives the coolingmethod. The label B is used to designate strandsize with “1” for 0.375 in. and “2” for 0.5 in. TheC designation is given by numbers 1-6, whichrefer to the temperature levels beginning withthe control as “1” and continuing up to 1300°F(704 °C) which is given as “6”. The D designationfor the cooling method is denoted by “1” for

cooling outside the furnace and “2” for coolinginside the furnace.

Phase IAFor the control, 500°F (260 °C) and 800°F(427 °C), three (3) coupons per strand size wereheated and tested. These specimens were cooledby removing them from the furnace. For the highertemperatures, 1000°F (538 °C), 1200°F (649 °C),and 1300°F (704 °C), six (6) coupons were heatedfor the 0.5 in. diameter strands and four (4) wereheated for the 0.375 in. diameter strands. Theincrease in number of strands was to observethe effects due cooling. Three (3) of the 0.5 in.and two (2) of the 0.375 in. were cooled insidethe furnace and the remaining three (3) 0.5 in.and two (2) 0.375 in. were cooled by removingthem from the furnace. All strands were held attheir specific temperature for 60 minutes. ThisPhase IA length of soak time was consistent withprevious studies6,7.

Phase IBIn addition to Phase IA, additional testing was alsoperformed where the specimens were held attheir desired temperature for 35 minutes. Thedecrease in heat soak time was due to interest in

Table 1: Tension Test Matrix Phase IA: 0.375-in. Strand Diameter

229



the materials properties for specimens exposedto elevated temperatures for shorter periods oftime (i.e., more rapid emergency response). Thisalso raises the question of whether a 60 minutesoak time would appropriately address longduration fires. This issue is addressed in greaterdetail in Phase III of this study.

For Phase IB, three (3) temperatures werestudied, 1000°F (538 °C), 1200°F (649 °C) and1300°F (704 °C). Three (3) strands pertemperature were tested for both the 0.5 in.and 0.375 in. size strands. The specimens

were removed from the furnace and coolednaturally.

MATERIALSPrestressing StrandsThe specimens selected for the experiment wereuncoated seven-wire low-relaxation prestressingstrands of grade 270 ksi (1862 MPa). Two (2)sizes of wires were used, 0.5 in. (12.7 mm) and0.375 in. (9.53 mm) diameter, with cross-sectional areas of 0.153 in2 and 0.085 in2

respectively. ASTM A 416/A 416M-06 provides

Table 2: Tension Test Matrix Phase IA: 0.5-in. Strand Diameter

Table 3: Tension Test Matrix Phase IB: 0.375-in. Strand Diameter

Table 4: Tension Test Matrix Phase IB: 0.5-in. Strand Diameter

230



required properties for this type of prestressingstrand as illustrated in Table 5. In order for a strandto be acceptable for use in construction andcertified by its supplier the yield stress andminimum fracture strength must be met. Thesevalues are also used to verify testing proceduresused in experimental research.

DESCRIPTION OF TESTSPECIMENSThe coupon specimens were cut into lengths of18 in. (457.2 mm), a value based on ASTMA416M-06, the grip length of the jaws, and thefurnace dimensions. Nothing additional wasapplied or performed on the prestressing strandsprior to exposure to elevated temperatures. Aschematic and actual view of the specimen isgiven in Figure 3.

TEST SETUPFurnaceIn order to simulate fire damage, the specimenswere placed inside a cylindrical tube furnace and

heated to their designated temperature at a rateof approximately 8°F/min (4.4 °C/min). Thetemperature was measured using athermocouple which was directly linked to thetemperature controller.

For the first set of coupons, the temperaturewas increased until it reached its designated valueand then held for 60 minutes, allowing a uniformtemperature to be reached. The specimens thatwere to be cooled outside the furnace were thenremoved, placed at room temperature and allowedto cool. The furnace was turned off and theremaining specimens were left in furnace andcooled as the furnace naturally cooled down. Thesecond set of coupons were heated in the samemanner, but only held at their specific temperaturefor 35 minutes. They were cooled outside thefurnace after their heat soak was completed. Aspreviously noted, the heat soak time period of 60minutes was based on previous research6,7 andthe 35 minute period was chosen to research theeffects caused by a shorter period of exposuretime.

Table 5: Mechanical Properties of Prestressing Strands

Figure 3: Schematic and Actual View of Coupon Specimens

Conversion Units: 1 in. = 15.2 cm

231



When heating the coupon specimens, allreplicate specimens for each condition wereplaced in the furnace at the same time. Thefurnace and heating setup can be seen in Figures4 and 5. The white blocks shown in Figure 5 wereoven bricks which were used elevate the couponspecimens in the furnace and keep them fromtouching one another during heating.

Testing Equipment and ProceduresTensile testing was performed using a MTS880machine as shown in Figure 6. Load, strain, andstroke were electronically recorded for eachspecimen. In order to achieve equal grip strengtharound the strands, a 3 in. (76.2 mm) longaluminum tube made of aluminum alloy 6061 witha thickness of 0.049 in. was placed on both endsof the coupon. For the 0.5 in. dia. strands a 0.625in. (15.88) outside diameter, 0.527 in. (13.39 mm)inside diameter aluminum tube was used. The0.375 in. dia. strands employed a 0.5 in. (12.7mm) outside diameter, 0.402 in. (10.21 mm)inside diameter aluminum tube. This allowed thegrips to squeeze the aluminum into the gapsbetween the individual wires and prevent slippingor premature fracture. A small weld was alsoplaced at the ends of each specimen to ensurethe strands were loaded uniformly. Grippingstrength was set at 7.5 ksi (51.8 MPa) for the 0.5in. specimens and 6 ksi (41.4 MPa) for the 0.375in. specimens. A typical specimen placed in theMTS880 machine can be seen in Figure 7.

Figure 4: Cylindrical Tube Furnace

Figure 5: Typical Heating Setup

Figure 6: MTS880 Testing Machine Priorto Tension Tests

232



The procedure for the coupon testing beganby centering the specimen inside the testingmachine. The specimen was loaded to an initialload of 10% of the minimum breaking strengthas specified by ASTM A416M-06 and ASTM A370-07a. A Class-C extensometer was then placedon the strand and the gauge reading was set to0.001 in./in. (0.0254 mm/mm). Loading rates foreach strand diameter were selected to be 23%of the maximum acceptable load set by ASTMA370-07a. These values were based on thestandard’s allowable range and testing machine’scapabilities. Initial loads and load rates can beseen in Table 6.

Loading continued until yielding took place.The extensometer was then removed in order toprevent damage to itself during fracture. Forspecimens unexposed to the furnace the yieldwas taken at an elongation of 1% which wasrecorded by the machine as a strain value of 0.01in./in. (0.254 mm/mm) in accordance with ASTMA416M-06. For the heat-exposed specimens yieldoccurred much sooner and the extensometer wasremoved once the curve on the computer clearlychanged slope signifying a yield. During and afterthe removal of the extensometer, the loadingcontinued and was completed when thespecimen fractured. For certain cases,particularly the higher temperatures, a clearchange in slope was not recognizable andtherefore the extensometer was left on thespecimen until failure.

EXPERIMENTAL TESTRESULTSVisual ObservationsVisual observations of the prestressing strandswere made prior to testing and are presented inFigures 8-12. Noticeable changes to the strand’sappearance were first observed with thespecimens exposed to 1000°F (538 ºC). Thesecoupons’ shiny appearance was replaced by adark dull appearance which indicates thebeginning of steel oxidation. The strands heatedto 1200°F (649 °C) were also found to be dulland in addition their exterior coating began toslightly flake off. Finally the specimens of 1300°F(704 °C) showed significant flaking of the exteriorand dullness. The discolored areas in Figure 12are parts of the strand where the exterior flakedoff after heating during transport. Couponsexposed to 500°F (260 °C) and 800°F (427 °C)remained cosmetically the same as they were

Figure 7: Tension Test Setup

Table 6: Testing Properties of PrestressingStrands

233



prior to heating, with an exterior characterized bya shiny appearance. These observations werethe same for each temperature regardless of thetype of cooling method or length of heat soak.

Test ResultsThe results of the tension tests are presented inthis section. For each of the tests, stroke andload were recorded for the entire loading period.Strain was recorded until at least the yield pointas discussed earlier. A typical stress-strain plotproduced by a tension test is shown in Figure 13.Moore and Myers1 report the yield stress, theultimate load, the modulus of elasticity as well asthe standard deviation for each replicate andcondition of three previously mentionedproperties.

Failure ModeThe failure modes of the strands were directlyrelated to the heat damage experienced in thefurnace. As the exposure temperature increased,the failure mode moved closer to the lower gripwhere the machine was elongating the strand.Due to the irregular shape of prestressing strandsthis type of failure is considered acceptable byASTM A370-07a.

Figure 8: Strand After Exposure to 500ºF(260 ºC)




Figure 12: Strand after Exposure to 1300ºF(704 ºC)

Figure 13: Typical Stress-Strain Curvefor Tension Testing

234



Specimens exposed to elevated temperatureswere not expected to meet mechanical propertyASTM requirements due to mechanicalalterations by heat, however unexposed strandswere used to verify acceptance of the testingmethod. These specimens failed in an acceptablemanner stated by ASTM A370-07a by producinga yield greater than 243 ksi (1675 MPa) and abreaking strength greater than 270 ksi (1862MPa). These specimens also failed in the centerbetween the two jaws. Typical failure modes areshown in Figure 14.

TENSILE STRENGTHFrom the test results several importantobservations can be made. Tensile strength,modulus of elasticity and yield strength are allspecific properties which have been analyzed andreported in this section. Additional analysis andconclusions have been made regardingtemperature level, size of strand, cooling methodand heat exposure time.

The percent of original tensile strength vs.temperature for the specimens of Phase IA(heated for 60 minutes) can be seen in Figure15. The tensile strength of the strands decreasesby only 4% for the 0.5 in. strands and 1% for the0.375 in. strands between the temperatures of68°F (20 °C) and 500°F (260 °C). A slightly largerweakening occurs between the temperatures of500°F (260 °C) and 800°F (427 °C) as there isan 8% and 11% decrease for the 0.5 in. and 0.375in. strands, respectively. However, a significantloss in tensile strength occurs after 800°F (427°C). The curve begins a steep downward trenduntil it reaches 1200°F (649 °C) where it starts tolevel off. For temperatures 800°F (427 °C) to1200°F (649 °C) a total loss of 48% and 46%was experienced by the 0.5 in. and 0.375 in.strands respectively.

Based on Figure 15 it appears that the loss oftensile strength is proportional for both strandsizes. There is a small statistical differencebetween the strands sizes in the percent ofultimate tensile strength at 500°F (260 °C), 1000ºF(538 ºC) and 1200°F (649 °C). However, there isno indication that the size of the steel has anyeffect on the residual tensile strength of the steel,

Figure 14: Typical Failure Modeof Prestressing Strand in Tension

Figure 15: Residual Ultimate Tensile Strengthvs. Temperature

235



because the strand with the highest residualtensile strength varied by temperature. If therewas an increase in the specimen pool size it islikely that a statistical difference would not exist.

All data obtained from this study and previousstudies has been reprinted in Figure 16 with theaddition of the experimental results from PhaseIA of this study. The current results are quitesimilar to that of previous research and can beassumed to be accurate.

Figure 17 compares the ultimate tensilestrength of the strands which were left to coolinside the furnace (i.e., more gradually) and thosewhich were removed and cooled outside of thefurnace. Statistically there is a small differencebetween strands cooled inside the furnace andstrands cooled outside of the furnace. Thestandard deviation for various replicate testing foreach conditioning ranged from 0.10-0.87% and0.54-3.13% for the 0.375 and 0.50 in. strandsrespectively.

The length of heat soak time is shown in Figure18. At higher temperatures there is a greater lossin strength for specimens soaked for 60 minutesversus those only soaked for 35 minutes. Fortemperatures between 1000ºF (538 ºC) and

1300ºF (704 ºC), it appears to be more significantthan the lower temperatures. However, there isstill a measurable difference in strength loss. Thedata obtained for the specimens soaked for 35minutes is similar to that of Guyon4 who soakedspecimens for 20 minutes. However, Guyon didnot report the type of strand tested; therefore nodirect correlation can be made. Abrams and Erlin6

also noted a difference in tensile strength due tothe length of time the specimens were soaked.However, their study consisted of longer timeintervals (1 hour, 4 hours and 8 hours) whichresulted in small variances.

Figure 17: Comparison of Cooling Methods

Figure 18: Comparison of Heat Soak Time

Figure 16: Comparison of Results with OtherPrevious Research

Frac

tion

of O

rigin

al U

ltim

ate

Tens

ile S

tren

gth

236



MODULUS OF ELASTICITYThe modulus of elasticity based on temperatureexposure is given in Figure 19. The values forthis property were determined by measuring theslope of the initial linear section of the plot. Thisparticular property was found to be fairlyconstant despite the increase in temperature.The values actually increased for temperaturesof 500°F (260 °C) and 800°F (427 °C). They thendecreased for the remaining elevatedtemperatures, but only to 97% of the originalmodulus value. This compares similarly toMacLean8 and Holmes9. The behavior of theprestressing strand is much like that of mildsteel which has also been found to not changeafter heating and cooling3.

unrecognizable as fracture occurred before anyslope change occurred.

In the case of the 0.5 in. dia. strand heated to800ºF (427 ºC), the fracture occurred immediatelyafter the yield with little tensile strength increase.Strands heated to temperatures above 800ºF(427 ºC) fractured before an indication of yieldingoccurred. The 800ºF (427 ºC) temperature markis very close to the limit at which all non-linearbehavior is lost. You will note from the stress-strain curves shown in Figures 20-21, that the0.375 in. strand heated to 800ºF (427 ºC) didexhibit some non-linearality, but at 1000ºF (538ºC) did not exhibit any. Therefore, thetemperatures of 800ºF-1000ºF (427 ºC-538 ºC)are a critical temperature range for the strands inwhich all non-linear behavior is lost.

The loss of non-linear behavior is directlyrelated to the loss of ductility and prior indicationof failure. Strands heated to the 800ºF-1000ºF(427 ºC-538 ºC) temperature range will stillmaintain over 65% of their undamaged tensilestrength which in some applications will besufficient. However, in addition they will also losealmost or all of their non-linear behavior becoming

Figure 19: Residual Modulus of Elasticity

YIELD STRENGTHAs defined by ASTM A416M-06, the yield strengthis taken at 1 percent elongation. However, for thespecimens exposed to elevated temperatures,this elongation was not possible and the yield wastaken at the point of significant slope change.Furthermore, for very high temperaturespecimens, yield strength was often

Figure 20: Stress vs. Strain for 0.5 in.Specimens of Phase I

237



more brittle. Materials which fail without significantwarning are normally avoided for use in structuralcomponents. Some sign of distress such assignificant concrete cracking prior to failure isdesirable, which means the loss of ductility in thereinforcing is highly undesirable.

TENSILE STRENGTH ATELEVATED TEMPERATURESAnother topic of interest is how the residual tensilestrength of the strands compares with the tensile

strength of the strands at elevated temperatures.In Figure 22, the experimental results from thisstudy are compared with the tensile strength ofprestressing strands at elevated temperaturesreported by other researchers5,10. It can be notedthat there is a significant increase in tensilestrength upon cooling for all temperatures greaterthan 400ºF (204 ºC).

CONCLUSIONPrestressing strand properties were evaluatedafter exposure to temperatures ranging between68°F (20 °C) to 1300°F (704 °C) and then cooledeither inside or outside the furnace. Exposure timeperiods analyzed included 35 minutes and 60minutes of soak time in the furnace. Propertieswere determined by tension testing of the strands.Based on the experimental data, the followingconclusions can be drawn:

1. There is significant loss in prestressing strandtensile strength upon exposure to elevatedtemperatures greater than 500ºF (260 ºC).This significant loss for the increment of 500ºF-800ºF (260 ºC-427 ºC) is 9.6% and increasesto 21.5 and 26.0% for respective temperaturesincrements between 800ºF-1000ºF (427 ºC-538 ºC) and 1000ºF-1200ºF (538 ºC-649 °C).The final temperature range, 1200ºF-1300ºF(649 °C-704 ºC), which is the smallestincrement, experienced a tensile strength lossof 4.8%. A minimal tensile strength loss of 2.3%occurred at the initial temperature incrementof 68ºF-500ºF (20 ºC-260 ºC).

2. The duration of exposure to elevatedtemperatures is critical in the residual tensilestrength after cooling. Strands soaked at atemperature for 35 minutes performed betterthan those soaked for 60 minutes. A significantdifference in performance of 6-25% was found

Figure 21: Stress vs. Strain for 0.375 in.Specimens of Phase I

Figure 22: Comparison of Tensile Strengthfor Prestressing Strands at Elevated

Temperatures and Residual Tensile Strengthof Prestressing Strands

238



for temperatures 1000ºF-1300ºF (538 ºC-704 ºC).

3. Regardless of the cooling method employedin this study the prestressing strands behavedsimilarly.

4. The soak or exposure time investigated in thisstudy (35 and 60 minutes) did exhibitmeasurable dif ferences in the tensileproperties.

5. The modulus of elasticity post-conditioningwas largely unaffected by the exposuretemperature.

6. The non-linear behavior of the prestressingsteel is significantly affected upon reaching thecritical temperature range of 800ºF-1000ºF(427 ºC-538 ºC). Within these temperaturesthe steel becomes brittle, yielding at fractureor fracturing before yielding depending on thetemperature exposure.

ACKNOWLEDGMENTThe authors would like to acknowledge thef inancial support of the Center forTransportation Infrastructure and Safety (CTIS)at the Missouri University of Science andTechnology (Missour i S&T). CoreslabStructures, Inc. of Marshall Missouri is greatlyacknowledged for their contributions ofprestressing tendon for the study. Thanks arealso due to the Materials Research Laboratory(MRC) at Missouri S&T for assistance andusage of their high temperature furnace. Thetechnician and staff support from the Centerfor Infrastructure Engineering Studies (CIES)and Department of Civil, Architectural, andEnvironmental Engineering at the MissouriUniversity of Science and Technology has alsoprovided much appreciated assistance.

REFERENCES1. Abrams M S and Cruz C R (1961), “Behavior

of High Temperatures of Steel Strand forPrestressed Concrete”, Journal of the PCAResearch, Vol. 3, No. 3, pp. 8-19.

2. Abrams M S and Erlin B (1967), “EstimatingPost-Fire Strength and ExposureTemperature of Prestressting Steel byMetallographic Method”, Journal of the PCAResearch, Vol. 9, No. 3, pp. 23-33.

3. Dias W P S (1992), “Some Properties ofHardened Cement Paste and ReinforcingBars Upon Cooling from ElevatedTemperatures”, Fire and Materials, Vol. 16,No. 1, pp. 29-35.

4. Edwards W T and Gamble W L (1986),“Strength of Grade 60 Reinforcing Bars AfterExposure to Fire Temperatures”, ConcreteInternational: Design & Construction, Vol. 8,No. 10, pp. 17-19.

5. Guyon Y (1953), Prestressed Concrete, F.J.Parsons, Folkestone and Hastings, Limitedof London, London.

6. Holmes M, Anchor R D, Cook G M E andCrook R N (1982), “The Effect of ElevatedTemperatures on the Strength Propertiesof Reinforcing and Prestressing Steels”,The Structural Engineer, Vol. 60B, No. 1,pp. 7-13.

7. MacLean K (2007), “Post-Fire Assessmentof Unbonded Post-Tensioned ConcreteSlabs: Strand Deterioration and PrestressLoss”, Masters of Science Thesis,Department of Civil Engineering, Queen’sUniversity, Kingston, Ontario, Canada.

8. Moore W L and Myers J J (2008),“Performance of Fire-Damaged Prestressed

239



Concrete Bridges”, p. 139, MissouriUniversity of Science and Technology.

9. Neves I C, Rodrigues J P and Loureiro A(1996), “Mechanical Properties ofReinforcing and Prestressing Steels After

Heating”, Journal of Materials in CivilEngineering, pp. 189-194.

10. Precast/Prestressed Concrete Institute(PCI), (2004), PCI Design Handbook, 6th

Edition.

Documents

SEVEN-WIRE LOW RELAXATION PRESTRESSING TENDON · PDF fileSEVEN-WIRE LOW RELAXATION PRESTRESSING ... Keywords: Low-relaxation seven-wire tendon, Elastic modulus, ... the strand, a specific