Self-Developed Testing System for Determining the Temperature

materials Article

Self-Developed Testing System for Determining the Temperature Behavior of Concrete He Zhu, Qingbin Li and Yu Hu * Department of Hydraulic engineering, Tsinghua University, Beijing 100084, China; [email protected] (H.Z.); [email protected] (Q.L.) * Correspondence: [email protected]; Tel.: +86-010-62781161 Academic Editor: Ming Hu Received: 9 February 2017; Accepted: 11 April 2017; Published: 16 April 2017

Abstract: Cracking due to temperature and restraint in mass concrete is an important issue. A temperature stress testing machine (TSTM) is an effective test method to study the mechanism of temperature cracking. A synchronous closed loop federated control TSTM system has been developed by adopting the design concepts of a closed loop federated control, a detachable mold design, a direct measuring deformation method, and a temperature deformation compensation method. The results show that the self-developed system has the comprehensive ability of simulating different restraint degrees, multiple temperature and humidity modes, and closed-loop control of multi-TSTMs during one test period. Additionally, the direct measuring deformation method can obtain a more accurate deformation and restraint degree result with little local damage. The external temperature deformation affecting the concrete specimen can be eliminated by adopting the temperature deformation compensation method with different considerations of steel materials. The concrete quality of different TSTMs can be guaranteed by being vibrated on the vibrating stand synchronously. The detachable mold design and assembled method has greatly overcome the difficulty of eccentric force and deformation. Keywords: concrete; thermal behavior; mechanical behavior; cracking; temperature stress testing machine (TSTM)

1. Introduction Cracking due to temperature and restraint in mass concrete, such as concrete dams, highway pavement, and bridge decks, can be easily observed. It is a critical concern to study the mechanism of temperature cracks through the testing method [1,2]. To investigate the performance of concrete after it is casted, some testing programs, such as the so-called ring tests, plate tests, uniaxial restraint, and substrate restraint tests, were conducted [3]. The cracking reasons may be complex and most of the influencing factors on cracking sensitivity can be investigated at the same time by the uniaxial restraint tests [4]. The temperature stress test (TST) method, by adopting the uniaxial restraint test, is thus developed to determine the mechanical and thermal behavior of concrete. The most widely-used TST devices are cracking frame and temperature stress testing machines (TSTM). 1.1. Cracking Frame Both ends of the specimen for cracking frame are held by two cross-heads [5]. The cross-heads are connected by two longitudinal bars, called steel shafts, that provide restraint. The steel shafts are made of special invar steel whose thermal expansion coefficient is extremely low, approximately 1 × 10−6 /◦ C, to eliminate the thermal effect on the specimen. Currently, two types of the concrete temperature control methods are provided, primarily. One type uses a fluid-cooled temperature

Materials 2017, 10, 419; doi:10.3390/ma10040419

www.mdpi.com/journal/materials

Materials 2017, 10, 419

2 of 18

control formwork, which controls the temperature history of concrete by adjusting the temperature of the flowing fluid in the formwork. The other type uses an air environmental chamber, where the concrete specimen is placed into the small environmental chamber, and the temperature of the concrete specimen is affected by adjusting the temperature and flowing of the air in the environmental chamber [5,6]. The concrete deformation varies as the temperature changes. Temperature stress is produced as the concrete deformation is restrained by the cross-head and steel shafts. Deformation of the steel shaft is measured by the strain gauge to calculate longitudinal restraint stress. Researchers have carried out many tests for concrete cracking with the cracking frame [5–11]. Under the assumption that concrete is elastic, the degree of restraint provided by the steel shafts on the concrete can be defined by Equation (1) [6]. kr = 1 −

εr 100 = Ac εf 1 + EEcs A s

(1)

where kr is the degree of restraint (%); ε r is the strain under restrained condition; ε f is the strain of a free specimen (kr = 0); Ec is the concrete elastic modulus (MPa); Ac is the concrete cross-sectional area (m2 ); Es is the elastic modulus (MPa) of the steel shaft; and As is the cross-sectional area (m2 ) of the steel shaft. In reality, creep and damage could affect the concrete so that Ec should be modified to measure the restraint degree. As the concrete elastic modulus increases with the development of age, the degree of restraint is not a constant. Thus, the cracking frame can provide a high degree of restraint condition rather than a full restraint condition. 1.2. TSTM To achieve a certain restraint condition, the TSTM was developed by Gierlinger and Springenschmid [12]. Its schematic drawing is shown in Figure 1. Both ends of the specimen are held by two cross-heads, which are the same as the cracking frame. The concrete specimen casting region is surrounded by the cross-heads and the formworks. The difference with the cracking frame is that two fixed ends are designed and the electrical motor is fixed on one end. The movable cross-head is connected to the electrical motor in order to be adjustable. Two fixed ends are also connected by two longitudinal bars, called the steel shafts, that provide restraint. When a tiny deformation of shrinkage or expansion occurs in the concrete, the movable cross-head will return to the designated position, driven by the motor. The greatest distinction between the cracking frame and the TSTM is the simulation of the restraint degree. The restraint degree of the TSTM is controlled by a computer in a closed-loop manner. The length of the specimen is monitored at a short interval and if the concrete deformation exceeds the control threshold, such as 5 µm for the 1 m specimen, a corresponding controlling order is sent to the motor. Any restraint degree from free deformation to full restraint can be simulated.

Materials2017, 2017,10, 10,419 419 Materials

18 33ofof18

Figure1.1.The Theplan plan sketch sketch for for the (1)(1) actuating motor, (2) (2) left Figure the TSTM. TSTM. The Thenumbered numberedparts partsare areasasfollows: follows: actuating motor, fixed end, (3) load cell, (4) left snap joint, (5) sliding sleeve, (6) concrete specimen casting region, (61) left fixed end, (3) load cell, (4) left snap joint, (5) sliding sleeve, (6) concrete specimen casting region, movable cross-head, (62) (62) fixedfixed cross-head, (7) deformation measuring device, (71) deformation sensor, (61) movable cross-head, cross-head, (7) deformation measuring device, (71) deformation (72) sensor fixture, (73) embedded rod positioning fixture, (74) deformation transfer guide bar, (8) right sensor, (72) sensor fixture, (73) embedded rod positioning fixture, (74) deformation transfer guide bar, snap joint, (9)joint, right(9) fixed end, and (10)and steel shaft (developed from the from prototype of [12]). of [12]). (8) right snap right fixed end, (10) steel shaft (developed the prototype

1.3.The TheDevelopment DevelopmentofofTSTM TSTM 1.3. TheTSTM TSTMhas hasbeen beendeveloped developedby bymany manyresearchers researchersininrecent recentyears yearsand andthe thedevelopment developmenthistory history The ofthe the TSTM TSTM has has been been reported reported by by Staquet Staquet [13]. [13]. The Thedesigns designsofofTSTMs TSTMsfrom fromdifferent different university university of laboratories differ in some way and the characteristics are summarized below. laboratories differ in some way and the characteristics are summarized below. 1.1.

The method.ItItisis important to determine the value of shrinkage, Thedeformation deformation measurement measurement method. important to determine the value of shrinkage, elastic elastic modulus, and Several creep. methods Several methods been such adopted, such as the monitoring the modulus, and creep. have beenhave adopted, as monitoring displacement displacement of the movable grip [14–17], however, this method could generate some artifacts of the movable grip [14–17], however, this method could generate some artifacts by ignoring the by ignoring between the interaction grip Altoubat and the sample. Altoubat [18,19] used differential the linear interaction the gripbetween and the the sample. [18,19] used the linear variable variable differential transformer method demonstrate the difference transformer (LVDT) attachment(LVDT) methodattachment to demonstrate thetodifference from LVDT on thefrom grip. LVDT on the grip. However, only the displacement of one surface could be measured rather However, only the displacement of one surface could be measured rather than the whole sample. than the whole sample. A method using [20–23] embedded [20–23] was developed to concrete directly A method using embedded bars/rods wasbars/rods developed to directly measure the measure the concrete displacement. The embedded method has the disadvantage of displacement. The embedded bars/rods method hasbars/rods the disadvantage of creating local damage creating local damage inside the sample, and the strength would be affected. Thus, a inside the sample, and the strength would be affected. Thus, a deformation measurement deformation measurement method of directly measuringwith the little concrete displacement with could little method of directly measuring the concrete displacement influence on the sample influence on the sample could be developed. be developed. 2.2. Temperature Temperaturecontrol controlmethod. method.Some Somedevices deviceswithout withoutthermal thermalregulation regulationsystems systemshave have been been designed to meet the purpose of studying the shrinkage property [18,24,25]. For TSTMs, two designed to meet the purpose of studying the shrinkage property [18,24,25]. For TSTMs, types of temperature control methods, the fluid-cooled formwork [26] and environmental two types of temperature control methods, the fluid-cooled formwork [26] and environmental chamber chamber[27,28], [27,28],have havebeen beenemployed. employed.Only Onlya asingle singlespecimen specimencan canbebetested testedininthe the same same environment restrained by the present temperature control method. Concrete is an artificial environment restrained by the present temperature control method. Concrete is an artificial material materialhaving havingdiscrete discreteproperties, properties,sososome someparallel parallel specimens specimensshould should be be synchronously synchronously performed same environmental atmosphere to evaluate the material properties more performedininthethe same environmental atmosphere to evaluate the material properties representatively. more representatively. 3.3. Concrete Concretequality. quality.Concrete Concreteisispoured pouredinto intothe themold moldofofthe theTSTM TSTMdirectly directlyand andaavibrating vibratingrod rodisis used to guarantee the concrete quality. When multi-TSTMs are employed, the homogeneity used to guarantee the concrete quality. When multi-TSTMs are employed, the homogeneity of of all all samples is hardly toguaranteed be guaranteed by vibrating the sample separately with a handheld samples is hardly to be by vibrating the sample separately with a handheld vibration vibration The representation and applicability ofresults the testwill results will be influenced. rod. Therod. representation and applicability of the test be influenced. A certain number of concrete samples should be conducted to guarantee that the test results are A certain number of concrete samples should be conducted to guarantee that the test results are representative and acceptable. To develop a multi-TSTM system, some designs and methods could representative and acceptable. To develop a multi-TSTM system, some designs and methods could be developed. be developed.

Materials 2017, 10, 419

4 of 18

2. The Self-developed Synchronous Closed Loop Federated Control TSTM System Materials 2017, 10, 419 Materials 2017, 10, 419

4 of 18 4 of 18

To improve the shortcomings of the TSTM, a synchronous closed loop federated control TSTM system is developed and will be introduced inLoop detail as follows. 2. The Self-developed Synchronous Closed Federated Control TSTM System 2. The Self-developed Synchronous Closed Loop Federated Control TSTM System

To improve the shortcomings of the TSTM, aTSTM synchronous loop federated control TSTM To improve the shortcomings of the TSTM, a synchronous closed loop federated control TSTM 2.1. The Synchronous Closed Loop Federated Control System closed system is developed and will bebe introduced inin detail asas follows. system is developed and will introduced detail follows.

Referring to Figure 2, a natural environment simulation laboratory system is built, which 2.1. The Synchronous Closed Loop Federated Control TSTM System includes an outer laboratory, aLoop walk-in environment simulation laboratory located in the outer 2.1. The Synchronous Closed Federated Control TSTM System laboratory, a hosttocontrol anenvironment environment controllaboratory cabinet, and ais built, host computer. The Referring 2, natural environment simulation laboratory system is which built, which Referring toFigure Figurecabinet, 2, aa natural simulation system includes environment control cabinet is provided with an environment simulation control unit, which consists includes outer laboratory, walk-in environment simulationlocated laboratory in the outer an outeranlaboratory, a walk-in aenvironment simulation laboratory in thelocated outer laboratory, a host of some environmental factor simulation modules, such as temperature, humidity, a carbonization laboratory, a host control cabinet, an environment control cabinet, and a host computer. The control cabinet, an environment control cabinet, and a host computer. The environment control cabinet environment control with an environment simulation control unit, which3,consists simulation module, spray, and isanprovided illumination simulation module. Referring to Figure to prevent is provided with ancabinet environment simulation control unit, which consists of some environmental factor of simulation some environmental factor modules, such asand temperature, humidity, a carbonization modules, such as simulation temperature, humidity, a carbonization simulation module, spray, and an the foundation from transferring heat, a heat preservation waterproof structure is designed. The simulation module, spray, and an illumination simulation module. Referring to Figure 3, to prevent illumination simulation includes module. Referring to Figure 3, to prevent the foundation inner side of the foundation a ferroconcrete layer, a waterproof layer, a from heat transferring preservation theheat, foundation from transferring heat, a heat preservation and waterproof structure is designed. The a heat preservation and waterproof structure is designed. The inner side of the foundation layer, and a concrete layer, from top to bottom. To enhance the heat preservation effect of the inner side of the foundation includes a ferroconcrete layer, a waterproof layer, a heat preservation includes a ferroconcrete layer, a waterproof layer, a heat preservation layer, and a concrete layer, from laboratory, the wall is configured to be of polyurethane warehouse board. Two rubber isolation layers layer, and a concrete layer,the from to bottom.effect To enhance the heat the preservation effect oftothe top to bottom. To enhance heattop preservation of the laboratory, wall is configured be of are pasted to the inside and outside of the wall to shield the vibration from outside of the laboratory. laboratory, the wall is configured be of polyurethane board.toTwo isolation layers polyurethane warehouse board.toTwo rubber isolation warehouse layers are pasted the rubber inside and outside of the arewall pasted to the inside and outside of the wall to shield the vibration from outside of the laboratory. to shield the vibration from outside of the laboratory.

Figure 2. Thelayout layoutdiagram diagram of of the the laboratory. Figure The layout diagram Figure 2.2. The thelaboratory. laboratory.

Figure 3. The detailed foundation design.

Figure 3.3.The design. Figure Thedetailed detailed foundation foundation design. The walk-in environment simulation laboratory may simultaneously accommodate a plurality of TSTMs therein, and four TSTMs have been placed in the same environmental simulation The walk-in environment simulation laboratory may simultaneously accommodate a plurality The walk-in laboratory simultaneously accommodate a plurality atmosphere, whichenvironment can achieve simulation synchronous control may of multiple machines. The system is, thus, of of TSTMs therein, and four TSTMs have placed in the sameand environmental simulation TSTMs therein, and four TSTMs have beenbeen placed in theTSTM same environmental simulation atmosphere, named the synchronous closed loop federated control system, the system has some which can achieve synchronous control of multiple machines. The system is, thus, named atmosphere, which can First, achieve synchronous controlofoftemperature multiple machines. system is, the thus, obvious advantages. multi-concrete specimens stress tests The can be conducted synchronous closed loop federated control TSTM system, and the system has some obvious advantages. named the synchronous closed loop federated control TSTM system, and the system has some synchronously under the same environmental conditions, so that the discretization errors of results obvious multi-concrete of more temperature stress and testshave can be conducted can advantages. be decreased,First, enabling the testing specimens results to be representative a broader

synchronously under the same environmental conditions, so that the discretization errors of results can be decreased, enabling the testing results to be more representative and have a broader

Materials 2017, 10, 419

5 of 18

First, multi-concrete specimens of temperature stress tests can be conducted synchronously under the same environmental conditions, so that the discretization errors of results can be decreased, enabling the testing results to be more representative and have a broader application scope. The test set parameters of four TSTMs, such as the restraint degree, can be the same or different according to the test requirements. Many parallel tests can be carried out and the test efficiency is improved greatly. Second, no obvious fluctuation of the concrete temperature appears before or after the concrete specimen demolding, which is hardly to be realized by a fluid-cooled formwork [16]. Finally, the steel shaft, steel frame, and concrete specimen of the TSTM are in the same environment. The temperature gradient of the steel shaft and concrete specimen, and even the concrete specimen itself, could be Materials 2017, 10, 419 of 18 eliminated. However, a new problem is produced where both the concrete and the steel shaft 5are in the varied environmental atmosphere, and the temperature deformation is inconsistent as the material application scope. The test set parameters of fourand TSTMs, suchshaft as theare restraint degree, be thewill same thermal expansion coefficients of the concrete the steel different. The can solution be or different in according to the test requirements. Many parallel tests can be carried out and the test introduced Section 2.4 in detail. efficiency is improved greatly. Second, no obvious fluctuation of the concrete temperature appears 2.2. TheorDetachable Design of the TSTM before after the Mold concrete specimen demolding, which is hardly to be realized by a fluid-cooled formwork [16]. Finally, the steel shaft, steel and concrete specimen from of thethe TSTM are in thepart same The concrete specimen casting mold frame, is designed to be detachable TSTM host to environment. The temperature gradient of the steel shaft and concrete specimen, and even the modify the problems in Section 1.3. Multi-concrete specimens, together with molds, can be vibrated concrete specimenstand itself,simultaneously. could be eliminated. However, a new problem produced where both on the vibrating However, the centricity may be aisnew problem caused bythe the concrete and the steel shaft are in the varied environmental atmosphere, and the temperature detachable mold design. A mold assembly platform was designed as shown in Figure 4, on which deformation is inconsistent as the material thermal expansion coefficients of the concrete and the steel the cross-head can move along the longitudinal direction and the height of the horizontal direction shaft are different. The solution will be introduced in Section 2.4 in detail. can be kept constant. The cross-head positioning fixture is employed to fix the position of cross-heads. The formworks are then assembled between two cross-heads with bolts. Two supporting bars are 2.2. The Detachable Mold Design of the TSTM located between two side formworks to guarantee the formwork is non-deformable when the concrete The concrete mold is designed to is bevibrated, detachable host partand to specimen is beingspecimen vibrated. casting After the concrete specimen the from moldthe willTSTM be transferred modify the problems in Section 1.3. Multi-concrete specimens, together with molds, can be vibrated assembled on the TSTM host part by the hoisting device. Figure 5 shows the design diagram before on vibrating stand simultaneously. However, the centricity may be a new problem caused by the thethe mold is assembled. detachable mold design. A mold assembly platform was designed as shown inball Figure 4, on which The fixed ends are placed on the working platform. Eight height-adjustable supporting points the can move along the longitudinal direction and the height of can the allow horizontal direction andcross-head two bottom formwork supports are designed on the platforms, which the concrete to can be kept The cross-head fixture employed to fix of cross-heads. move on it.constant. The movable cross-headpositioning and the steel shaftisare connected bythe theposition sliding sleeve of which The are then assembled betweenistwo cross-heads with bolts. Two supporting bars are onlyformworks the longitudinal direction (X direction) allowed. located twomold sidedesign formworks to advantages. guarantee the formwork is non-deformable when the Thebetween detachable has some First, multi-concrete specimens can be vibrated concrete specimen is being vibrated. After concrete specimen isspecimens vibrated, can the be mold will be on the vibrating stand synchronously, so the the quality of multi-concrete guaranteed. transferred and assembled on the TSTM partare bymore the hoisting device. and Figures shows the Parallel tests could be conducted so thathost results representative test 5efficiency is design greatly diagram before the mold is assembled. improved. Second, formworks can be detached during the test, so that the influence of friction is eliminated. Finally, it is more convenient for the tester to adjust some operations during the experiment.

Figure Figure4.4.The Theplatform platformfor formold moldassembly. assembly.

The fixed ends are placed on the working platform. Eight height-adjustable ball supporting points and two bottom formwork supports are designed on the platforms, which can allow the concrete to move on it. The movable cross-head and the steel shaft are connected by the sliding sleeve of which only the longitudinal direction (X direction) is allowed. The detachable mold design has some advantages. First, multi-concrete specimens can be vibrated on the vibrating stand synchronously, so the quality of multi-concrete specimens can be guaranteed. Parallel tests could be conducted so that results are more representative and test efficiency is greatly improved. Second, formworks can be detached during the test, so that the influence of friction is eliminated. Finally, it is more convenient for the tester to adjust some operations during the experiment.

Materials 2017, 10, 419 Materials 2017, Materials 2017, 10,10, 419419

6 of 18 of 18 6 of6 18

Figure 5. The detachable mold design sketch of the The numbers in Figure 5 correspond with those Figure 5.detachable The detachable mold design sketch ofTSTM. theThe TSTM. The numbers in Figure 5 correspond with Figure 5. The mold design sketch of the TSTM. numbers in Figure 5 correspond with those of of Figure 1, so the numbers are discontinuous. The numbered parts are as follows: (1) actuating motor, (2) left those Figure 1, soare thediscontinuous. numbers are discontinuous. The numbered parts as follows: (1)(2) actuating Figure 1, soof the numbers The numbered parts are as follows: (1)are actuating motor, left motor, (2) left fixed end, (3)snap loadjoint, cell, (4) left snap joint, (6) (5)concrete sliding sleeve, (6) casting concreteregion, specimen casting fixed end, (3) load cell, (4) left (5) sliding sleeve, specimen (61) movable fixed end, (3) load cell, (4) left snap joint, (5) sliding sleeve, (6) concrete specimen casting region, (61) movable region, (61) movable cross-head, (62) fixed cross-head, (74) deformation transfer guide bar, (8) right cross-head, (62) fixed cross-head, (74) deformation transfer guide bar, right snap joint, (9) right fixed end, cross-head, (62) fixed cross-head, (74) deformation bar, (8)(8) right snap right end, snap joint, (9) right fixed end, (10) steel shaft,transfer (11) ballguide supporting points, (12) joint, roller,(9) (13) ballfixed supporting (10) steel shaft, (11) ball supporting points, (12) roller, (13) supporting points, and (14) bottom formwork (10) steel shaft, (11) ball supporting points, (12) roller, (13) ballball supporting points, and (14) bottom formwork points, and (14) bottom formwork support. support. support.

2.3. The Direct Measuring Deformation Method

Direct Measuring Deformation Method 2.3.2.3. TheThe Direct Measuring Deformation Method

The design program of the direct measuring deformation method is shown in Figures 1 and 6. The design program of direct measuring deformation method is shown in Figures 1 and program ofmeasuring thethe direct measuring method is shown inand Figures 1 and 6.at6.the ThreeThe setsdesign of deformation devices, atdeformation the most, could be employed are located Three sets of deformation measuring devices, at the most, could be employed and are located at the Three setsleft of deformation measuring devices, at deformation the most, could be employed and the are located at the top side, side, and right side. For the top measuring device, embedded rod is top side, left side, and right side. For the top deformation measuring device, the embedded rod top side,above left side, and right side. theby top deformation device, the embedded rod is is located the mold cavity andFor fixed the embedded measuring rod positioning fixture. The side embedded located above the mold cavity and fixed embedded rod positioning fixture. The side embedded located above mold cavity and fixed byby thethe embedded rod positioning fixture. side embedded rod is fixed bythe a spring adjusting component that comprises two threaded baffleThe rings and an adjusting rod is fixed by a spring adjusting component that comprises two threaded baffle rings and rod is so fixed by a side spring adjustingrod component that comprises twoside threaded baffleIt rings and anan to spring that the embedded that can be fixed against the formwork. is unnecessary adjusting spring so that the side embedded rod that can be fixed against the side formwork. It is adjusting spring so that the side embedded rod that can be fixed against the side formwork. It is adopt three setstoofadopt deformation measuring devices simultaneously, which depends on the purpose unnecessary three sets of deformation measuring devices simultaneously, which depends unnecessary to adopt three sets of deformation measuring devices simultaneously, which depends of test. For example, three sets of deformation measuring devices woulddevices be adopted to evaluate purpose test. example, three sets deformation measuring would adoptedthe onon thethe purpose of of test. ForFor example, three sets of of deformation measuring devices would bebe adopted deviation of different concrete sides. Only the top deformation measuring devices are allowable for evaluate deviation different concrete sides. Only deformation measuring devices to to evaluate thethe deviation of of different concrete sides. Only thethe toptop deformation measuring devices areare studying the restraint strength and the failure strain because too many embedded parts would cause allowable studying restraint strength and failure strain because many embedded parts allowable forfor studying thethe restraint strength and thethe failure strain because tootoo many embedded parts local damage. The embedded part is designed as a cross shape with steel wire, which has enough would cause local damage. The embedded part is designed as a cross shape with steel wire, which would cause local damage. The embedded part is designed as a cross shape with steel wire, which stiffness andstiffness little volume for reducing local damage. enough stiffness and little volume for reducing local damage. hashas enough and little volume forthe reducing thethe local damage.

Figure 6.Local Local detail of the the deformation measuring devices. Figure detail deformation measuring devices. Figure 6.6.Local detail of of the deformation measuring devices.

Materials 2017, 10, 419

7 of 18

2.4. The Compensation Method of Temperature Deformation All components of the TSTM are in the varied environmental atmosphere, and the temperature deformation is inconsistent as the material thermal expansion coefficients of the concrete and the steel are different. The deformation of the concrete may include the contribution of the temperature deformation caused by the steel shaft and other steel components. A compensation method of temperature deformation is proposed to eliminate the impact of the temperature deformation of the steel components on the concrete specimen deformation. Two different steel materials, 4J36 invar and 45# steel, are adopted. The thermal expansion coefficient is α (1.5 × 10−6 /◦ C for 4J36 invar) and α’ (12 × 10−6 /◦ C for 45# steel). The steel shaft is made from 4J36 invar and its length is L1 . The parts of the left fixed end, the right fixed end, and the snap joint are also 4J36 invar and the total longitudinal length is L2 . The fixed cross-head, the moveable cross-head, and the load cell are made from 45# steel, wherein the total length is L3 . Referring to Figure 1, if the longitudinal thermal deformation caused by the steel shaft is equal to that of other steel components, the longitudinal thermal deformation caused by the steel of TSTM cannot transfer onto the concrete and its effect could be eliminated. Thus, the concept was proposed that the relationship of the thermal expansion coefficient and the dimensions meets the demand of Equation (2). α × L1 = α × L2 + α’ × L3

(2)

Finally, the performance parameters of the self-developed TSTM system are listed in Table 1. Table 1. The performance parameters of the self-developed TSTM system. TSTM Performance

Capacity

Restraint degree Loading capacity Temperature control capacity Humidity control capacity The diameter of the steel shaft Specimen dimensions

0%–100% ±200 kN −20 to +80 ◦ C 0–100% RH 130 mm 150 mm × 150 mm × 2000 mm

3. Experimental Program To verify the ability of the self-developed TSTM system, three series of tests with the same concrete composition were conducted for the purpose of centricity assessment, temperature deformation compensation verification, and temperature stress testing. The raw materials for mixing the concrete are from the Xiangjiaba Dam, which is one of the largest hydropower stations in the world, with a power plant installed capacity of 7.84 million kW. The mixed ratio of the concrete for the experiments in Sections 3.1–3.3 is listed in Table 2, of which the water-cement ratio is 0.53. The constituents of the Portland cement are listed in Table 3. The fine aggregate (river sand) with a fineness modulus of 2.12 is used and the limestone gravel (with a diameter 5–20 mm) is employed as coarse aggregate. The aggregates are oven-dried at 100 ◦ C for 48 h before the concrete is mixed. The limestone gravel has a dry-bulk density of 2660 kg/m3 and a saturated water absorption value of 1.94% by mass, which is determined by the method in [29]. Table 2. The mixing ratio of concrete specimens for the TSTM experiment (kg/m3 ) Water

Cement

Sand

Gravel

200

380

900

975

Materials 2017, 2017, 10, 10, 419 419 Materials

of 18 18 88 of

3.1. The Centricity Assessment of the TSTM Table 3. The constituents of the Portland cement.

The TSTM has adopted the design of a detachable mold, which has never been reported in traditional TSTMs. The the specimen be3 assembled and at 2intervals, Compound CaOmold SiOand Al2 O MgO maySO Fe2 O3 Na2 Odisassembled K2 O TiO P2 O5 so 2 3 the Mass concentricity of the rig is vital to ensure that the load and deformation of the specimen are Percent/% 47.87 25.12 11.29 5.52 2.95 2.39 0.654 0.599 0.399 0.247 uniform. Since the designing scheme and the operation processes of all four TSTMs are identical, only one TSTM is conducted for assessing the centricity of loading. Three sets of deformation measuring 3.1. The Centricity Assessment of the TSTM devices are employed, as shown in Figure 6, and the specific implementation process of the TSTM The TSTM has adopted design of a is detachable mold, whichuntil has the never been reported in experiment is shown in Figurethe 7. The sample set as a free specimen concrete age reaches traditional mold and the may assembled and–35 disassembled intervals, so seven days,TSTMs. and theThe load is applied tospecimen the sample andbevaried between kN and 15 at kN. The 15 kN the of the is vital to ensure that the load and deformation are uniform. loadconcentricity represents that therigconcrete stress is in tension. The beginning load of is the –25specimen kN to reduce the gap Since and the processes of all four TSTMs are identical, only one TSTM effectthe anddesigning the strainscheme is recorded as 0operation με initially. is conducted for called assessing the centricity of loading. Threebysets of deformation measuring devices are A criterion the eccentricity degree is defined Equation (3) to evaluate the concentricity employed, as shown in Figure 6, and the specific implementation process of the TSTM experiment is of the TSTM. shown in Figure 7. The sample is set as a free specimen until the concrete age reaches seven days, and Eccentricity Degree =│dtop−dave│/dave (3) the load is applied to the sample and varied between –35 kN and 15 kN. The 15 kN load represents that thedconcrete stress is in tension. The loaddis kNaverage to reduce the gap effect the where top is the displacement value of thebeginning top side, and ave–25 is the displacements of and the top, strain is recorded as 0 µε initially. left, and right sides.

S1: Assemble

The mold is assembled on the platform as shown in Figure 4.

S2: Concreting and vibrating

Specimens are poured and vibrated on the vibrating stand synchronously to guarantee the quality.

S3: Hoisting on TSTM

Specimens are hoisted on the TSTM in 0.5 hour with the specific hoisting device as shown in Figure 5.

S4: Parameter setting

The parameters, such as restraint degree, deformation threshold, temperature, and humidity, are set.

Figure 7. The workflow of the TSTM experiment.

Figure 7. The workflow of the TSTM experiment.

A criterion called the eccentricity degree is defined by Equation (3) to evaluate the concentricity 3.2. Temperature Deformation Compensation Verification Experiment of the TSTM. Eccentricity Degree =|dtop −dave |/dave (3) To evaluate the effect of the temperature deformation compensation method, the temperature deformation test displacement is designed. The parts of top the side, steel and shaft, cross-head, and the snap joint inofFigure where dtop is the value of the dave is the average displacements the top,1 are wrapped left, and right with sides.the electric heating cable, as shown in Figure 8a, and then it is coated with the thermal insulating material, as shown in Figure 8b. Seven thermocouples are attached to the surface 3.2. Temperature Deformation Compensation Verification Experiment of the steel to monitor the temperature of the steel. The age of the concrete sample adopted in the test is three months so that the temperature field of the concrete is stable during the test. The environment To evaluate the effect of the temperature deformation compensation method, the temperature temperature is maintained at 20 °C to keep the concrete temperature constant. The steel shaft, crossdeformation test is designed. The parts of the steel shaft, cross-head, and the snap joint in Figure 1 are head, and the snap joint are heated by the electric heating cable from 20 °C to 70 °C to generate wrapped with the electric heating cable, as shown in Figure 8a, and then it is coated with the thermal temperature deformation and the heating process is shown in Figure 13. Thus, the concrete insulating material, as shown in Figure 8b. Seven thermocouples are attached to the surface of the deformation caused by the temperature deformation of the steel part can be measured.

Materials 2017, 10, 419

9 of 18

steel to monitor the temperature of the steel. The age of the concrete sample adopted in the test is three months so that the temperature field of the concrete is stable during the test. The environment temperature is maintained at 20 ◦ C to keep the concrete temperature constant. The steel shaft, cross-head, and the snap joint are heated by the electric heating cable from 20 ◦ C to 70 ◦ C to generate temperature deformation and the heating process is shown in Figure 13. Thus, the concrete deformation caused by the deformation of the steel part can be measured. Materials 2017,temperature 10, 419 9 of 18 Thermocouple-

Thermocouple-

4, 5, 6

1, 2, 3

Thermocouple-7

(a)

(b)

Figure 8. The verification of compensation the compensation method for temperature deformation. (a)steel A steel Figure 8. The verification test test of the method for temperature deformation. (a) A shaft wrapped the electric cable. (b) The TSTM coated with the thermal insulating shaft wrapped with with the electric heatingheating cable. (b) The TSTM coated with the thermal insulating material. material.

3.3. TSTM Performance Verification Experiment 3.3. TSTM Performance Verification Experiment The temperature stress test of concrete is conducted and the TSTM performances, such as The temperature stress test of concrete is conducted and the TSTM performances, such as simulating the different restraint degrees, multiple temperature regulation modes, and closed-loop simulating the different restraint degrees, multiple temperature regulation modes, and closed-loop control of multi-TSTMs, are verified. Four concrete specimens numbered as TSTM-A, TSTM-B, TSTM-C, control of multi-TSTMs, are verified. Four concrete specimens numbered as TSTM-A, TSTM-B, and TSTM-D are poured into the TSTMs, correspondingly. The thin film is placed between the concrete TSTM-C, and TSTM-D are poured into the TSTMs, correspondingly. The thin film is placed between and the formwork, and the lubricant is used for reducing the friction and making the demolding easier. the concrete and the formwork, and the lubricant is used for reducing the friction and making the easier. • demolding Test Start Time 

Test Start Time The method to determine the beginning of the test has been discussed by the researchers. The to determine beginning of time the test has been discussed Altoubat [18]method states the 12 h is the the earliest possible to apply instruments to by the the test researchers. without Altoubat statesStaquet the 12[13] h ischooses the earliest possible apply instruments thethe test without damaging the[18] sample. the start of testtime at thetoend of the setting timetoand sample damaging the sample. chooses the start of testbeing at the end of setting timeinside and the is initially restrained by theStaquet stiffness[13] of the frame with the motor turned offthe until the stress is reaches initially0.01 restrained by the stiffness of theisframe withenough the motor beingthe turned off of until thesample concrete MPa. The concrete strength not high to suffer weight thethe stress inside the concrete reaches 0.01 MPa. The concrete strength is not high enough to suffer deformation measurement device mentioned in Section 2.3 at the very early age. The displacement ofthe weight of the deformation measurement device mentioned in Section 2.3 at the very early age. the movable cross-head is measured [14] instead of measuring the concrete directly during the firstThe the movable cross-head is measured instead of measuring the concrete 25 hdisplacement to avoid localofdamage near the embedded part. This [14] method is acceptable at the early age asdirectly the during 25 h is tovery avoid local damage the embedded part. method is acceptable at the rigidity of the first concrete low. The motor near is also turned off until 25This h passes to avoid premature early and age the as the rigidity of the concrete is very low. The motor also turned damage sample is initially restrained by the stiffness of theisframe [21,25]. off until 25 h passes to avoid damage andare thedemolded sample is as initially restrained by thehas stiffness of the frame [21,25]. Afterpremature 25 h, the formworks the concrete strength increased and the direct After 25 h, the formworks are demolded the concrete strength has increased the direct measuring deformation method is adopted by the as upper side of the deformation sensor, asand described measuring deformation method isdemolding adopted byisthe upper of the deformation sensor, described in Section 2.3. The TSTM picture after shown in side Figure 9. As the formworks are as removed, in Section 2.3. The TSTM picture after demolding is shown in Figure 9. As the formworks are removed, the concrete is exposed in the same environment with the machine part of the TSTM. Additionally, the friction is eliminated as the surfaces are no longer restrained by formworks. The measured distance of the specimen is 1000 mm even though the actual length is 2000 mm, obtaining a uniform distribution of restraint stress. The resolution of the deformation sensor is 0.1 μm.

Materials 2017, 10, 419

10 of 18

the concrete is exposed in the same environment with the machine part of the TSTM. Additionally, the friction is eliminated as the surfaces are no longer restrained by formworks. The measured distance of the specimen is 1000 mm even though the actual length is 2000 mm, obtaining a uniform distribution of restraint Materials 2017,stress. 10, 419 The resolution of the deformation sensor is 0.1 µm. 10 of 18

Figure 9. 9. TSTM TSTM after after the the demolding. demolding. Figure

 •

Multi-TSTM Settings Multi-TSTM Settings Four TSTMs are all tested in the same environmental atmosphere, for which the air relative Four TSTMs all TSTM-A tested in is thesetsame environmental atmosphere, for whichofthe air relative humidity (RH) is are 80%. as the free specimen and the reference TSTM-C. The humidity (RH) is 80%. TSTM-A is set as the free specimen and the reference of TSTM-C. compensation cycle of TSTM-C begins when the increment value of the free specimen strain exceeds The cycle of TSTM-C begins when the increment of the free TSTM-D specimenis strain 2 με compensation (2 μm for 1 m measured distance) and the restraint degree of value TSTM-C is 50%. set as exceeds 2 µε (2 µm for 1 m measured distance) and the restraint degree of TSTM-C is 50%. TSTM-D 100% restrained, which is implemented by the motor adjusting the concrete to its original position is set as restrained, which is2 implemented by the motor adjusting the concrete to its original when the100% concrete strain exceeds με. position whenisthe concrete strain exceeds 2 µε. TSTM-B also set as an almost free specimen, of which the restraint degree is 5%. To measure TSTM-B is also set as an almost free specimen, of which the restraint 5%. 90 To h) measure the the elastic modulus (E), a compressive load about 45 kN (before 90 h) anddegree 60 kNis (after is applied elastic a compressive load about 4512 kNh,(before h) and“the 60 kN (after 90 h) isofapplied on the modulus specimen(E), to generate deformation every which 90 is called active method elastic on the specimen to generate every 12 h, which modulus determination” [16].deformation E is calculated by Equation (4): is called “the active method of elastic modulus determination” [16]. E is calculated by Equation  (4): E (4)  ∆σ E= (4) where  is the stress variation and  is the stain ∆ε variation. The method called standard method [29] is also employed to determine E. The specimen where ∆σ is the stress variation and ∆ε is the stain variation. dimension is 100 mm × 100 mm × 300 mm; and the specimens are all cured in the walk-in environment The method called standard method [29] is also employed to determine E. The specimen simulation laboratory which is the same as the TSTM. When the age is 24 h, 48 h, 120 h, and 160 h, dimension is 100 mm × 100 mm × 300 mm; and the specimens are all cured in the walk-in environment three specimens are tested and the average value is adopted as the elasticity modulus of the age. The simulation laboratory which is the same as the TSTM. When the age is 24 h, 48 h, 120 h, and 160 h, E of each specimen is calculated by Equation (5): three specimens are tested and the average value is adopted as the elasticity modulus of the age. The E    0.5MPa of each specimen is calculated by EquationE(5):  40% (5)



σ40% − σ0.5MPa where  40% is the stress correspondingEto=40% of the failure stress,  0.5MPa = 0.5 MPa, and  (5) is ∆ε the strain variation. where σ40% is the stress corresponding to 40% of the failure stress, σ0.5MPa = 0.5 MPa, and ∆ε is the  Temperature Modes strain variation. To evaluate the temperature control ability of the TSTM system, three temperature modes are •designed. Temperature Modes No. 1, from four hours to 70 h, the mode of simulating the semi-adiabatic condition for concrete is designed and the phase is ability namedofas temperature following phase,modes whichare is To evaluate the temperature control thethe TSTM system, three temperature implemented the environment the same the internal temperature of designed. No.by 1, controlling from four hours to 70 h, the temperature mode of simulating theassemi-adiabatic condition for the concrete. No. 2, to simulate the temperature history in the dam by water cooling, a designed cooling temperature curve is input and this phase is named as the design cooling phase (70–135 h). No. 3, the cold wave weather case, which is common in dam engineering, is designed by cooling down the environment temperature at the rate of 1 °C/h to obtain a crack (after 135 h). The specific processes of the temperature results are shown in Figure 10. Three thermocouples

Materials 2017, 10, 419

11 of 18

concrete is designed and the phase is named as the temperature following phase, which is implemented by controlling the environment temperature the same as the internal temperature of the concrete. No. 2, to simulate the temperature history in the dam by water cooling, a designed cooling temperature curve is input and this phase is named as the design cooling phase (70–135 h). No. 3, the cold wave weather case, which is common in dam engineering, is designed by cooling down the environment temperature at the rate of 1 ◦ C/h to obtain a crack (after 135 h). The specific Materials 2017, 10, 419processes of the temperature results are shown in Figure 10. Three thermocouples 11 of 18 are embedded into the concrete on each TSTM and the average values of each TSTM are named as T-B, T-B, T-C,T-C, and and T-D,T-D, as shown in Figure 10.10. The T-environment T-A, as shown in Figure The T-environmentrepresents representsthe theaverage average temperature temperature thermocouples located located 30 30 cm cm above above the the concrete concrete specimens. specimens. value measured by four thermocouples

Figure 10. The simulated temperature history curve of the experiment. Figure 10. The simulated temperature history curve of the experiment.

4. Results and Discussions 4. Results and Discussions 4.1. The Degree of of TSTM TSTM 4.1. The Deformation Deformation Results Results to to Assess Assess Eccentricity Eccentricity Degree The verify the the eccentric eccentric The concrete concrete specimen specimen was was pushed pushed and and pulled pulled by by the the actuating actuating motor motor to to verify degree of the self-developed TSTM. The results in Figure 11a show that the three deformation degree of the self-developed TSTM. The results in Figure 11a show that the three deformation curves curves of the the sensors sensors located and right right sides of located on on the the top, top, left, left, and sides appear appear to to have have the the same same shape. shape. Additionally, Additionally, the deformation magnitudes of of the the three three sensors sensors are are very very close. close. Figure 11b shows shows that the deformation magnitudes Figure 11b that the the maximum maximum eccentricity degree is 16%, and it occurs at the moment that the strain value is low and around 0 με. eccentricity degree is 16%, and it occurs at the moment that the strain value is low and around As we know, the calculation error will be amplified when the denominator magnitude is small. The 0 µε. As we know, the calculation error will be amplified when the denominator magnitude is small. eccentricity degree is below 10% when the strain is larger than 10 με. Meanwhile, the value of d top-dave The eccentricity degree is below 10% when the strain is larger than 10 µε. Meanwhile, the value of ranges between −2 με and 3 με in all loading processes. The difference between the top deformation d top -dave ranges between −2 µε and 3 µε in all loading processes. The difference between the top value and thevalue average value is quitevalue small, showswhich that only adopting theadopting top side deformation anddeformation the average deformation is which quite small, shows that only deformation measuring device during the test is proper to achieve the aim of directly measuring the the top side deformation measuring device during the test is proper to achieve the aim of directly concrete deformation. the centricity assessment the sample is pulled failure,toand the measuring the concrete After deformation. After the centricity test, assessment test, the sampleto is pulled failure, failure pattern is shown in Figure 12a. Concrete failed at the position near the embedded rods because and the failure pattern is shown in Figure 12a. Concrete failed at the position near the embedded rods three setsthree of embedded parts have affected integrity of the concrete caused local damage. because sets of embedded parts have the affected the integrity of the and concrete andthe caused the local The failure pattern of the sample with only the top embedded parts adopted in Section is shown damage. The failure pattern of the sample with only the top embedded parts adopted in3.3 Section 3.3 in shown Figure in 12b. The 12b. failure position far from the from embedded part andpart in the middle the sample is Figure The failure is position is far the embedded and in the of middle of the where the stress almost distributed evenly,evenly, whichwhich indicates that the part designed as a sample where theisstress is almost distributed indicates thatembedded the embedded part designed cross shape causes little local damage. The results show that the volume of the embedded part should as a cross shape causes little local damage. The results show that the volume of the embedded part be designed as small as as small possible avoid causing damage. the top deformation measuring should be designed as to possible to avoidlocal causing localOnly damage. Only the top deformation device is proposed if proposed the restraint strength andstrength the failure strain the research becausepoints, three measuring device is if the restraint and the are failure strain arepoints, the research sets of embedded parts will cause local damage and the failure stress will be decreased. because three sets of embedded parts will cause local damage and the failure stress will be decreased. In addition, the shape of the load curve in Figure 11a is also similar to the deformation curves. The deformation shows the synchronicity with the load changes, which indicates that the actuating motor connects tightly with the mold, and the detachable mold design of the TSTM shows good centering performance. With the consideration of a series of design concepts, such as the sliding sleeve on the steel shaft, the height-adjustable ball supporting points, cross-head positioning fixture, and the mold assembly platform, the self-developed TSTM has greatly overcome the difficulty of eccentric force and deformation.

Materials 2017, 10, 419 Materials 2017, 10, 419 Materials 2017, 10, 419

12 of 18 12 of 18 12 of 18

(a) (a)

(b) (b)

Figure 11. The deformation and eccentricity degree results of TSTM. (a) The load and deformation Figure11. 11.The Thedeformation deformation and eccentricity degree results of TSTM. (a) Theand loaddeformation and deformation Figure and eccentricity degree results of TSTM. TSTM. (a) curve curve of the concrete generated by the actuating motor of (b)The Theload eccentricity degree of the curve of the concrete generated by the actuating motor of TSTM. (b) The eccentricity degree of the of the concrete generated by the actuating motor of TSTM. (b) The eccentricity degree of the TSTM. TSTM. TSTM.

(a) (a)

(b) (b)

Figure 12. The failure patterns with different deformation measurement methods. (a) The sample with Figure Figure 12. 12. The The failure failure patterns patterns with with different different deformation deformation measurement measurementmethods. methods.(a) (a)The Thesample samplewith with top, left, and right embedded parts in Section 3.1. (b) The sample with top embedded part in Section 3.3. top, left, and right embedded parts in Section 3.1. (b) The sample with top embedded part in Section top, left, and right embedded parts in Section 3.1. (b) The sample with top embedded part in Section3.3. 3.3.

4.2. The Results of the Temperature of Deformation Compensation Verification 4.2. The Results ofthe the shape Temperature Deformation Verification In addition, of the of load curve in Compensation Figure 11a is also similar to the deformation curves. As shown inshows Figurethe 13,synchronicity the heating rate isthe difficult to control quantitatively by the adopting an The deformation with load changes, which indicates that actuating As shown in Figure 13, the heating rate is difficult to control quantitatively by adopting an electric heating cable. However, the temperature results of thermocouples 1–7 inTSTM Figureshows 8 are quite motor connects tightly with the mold, and the detachable mold design of the good electric heating cable. However, the temperature results of thermocouples 1–7 in Figure 8 are quite similar, especially underWith 60 °C, shows that steelofshaft is heated evenly coatedsleeve with centering performance. thewhich consideration of athe series design concepts, suchby asbeing the sliding similar, especially under 60 °C, which shows that the steel shaft is heated evenly by being coated with the thermal insulating material. The thermal expansion coefficient of Invar is not afixture, constant, on steel shaft, the height-adjustable ball supporting points, cross-head positioning and but the thethe thermal insulating material. The thermal expansion coefficient of Invar is not a constant, but rather a parameter related to the temperature. The average thermal expansion coefficient of the steel mold platform, self-developed has greatly overcome the difficulty ratherassembly a parameter relatedthe to the temperature.TSTM The average thermal expansion coefficientofofeccentric the steel shaft adopted by the TSTM system is 1.5 × 10−6−6/°C for the temperature ranging from −20 °C to 100 °C. force deformation. shaft and adopted by the TSTM system is 1.5 × 10 /°C for the temperature ranging from −20 °C to 100 °C. When the temperature of the steel shaft varies from 20 °C to 60 °C, the concrete deformation varies When the temperature of the steel shaft varies from 20 °C to 60 °C, the concrete deformation varies 4.2.only The Results themeasuring Temperaturedistance of Deformation Verification by 0.3 μm of (the is 1 m).Compensation However, the temperature deformation of concrete by only 0.3 μm (the measuring distance is 1 m). However, the temperature deformation of concrete variesAsfrom 0.3 in μm to 1.613, μmthe when the temperature increases from 60 °C to 70by °C.adopting an electric shown Figure heating is difficult to control quantitatively varies from 0.3 μm to 1.6 μm when therate temperature increases from 60 °C to 70 °C. The concrete temperature is kept constant at 20 °C and the testing time is approximately one heating However, the temperature resultsat of20 thermocouples 1–7 intime Figure 8 are quite similar, Thecable. concrete temperature is kept constant °C and the testing is approximately one hour, so that the concrete deformation is only caused by the external temperature deformation. The ◦ especially under 60 C, which shows that the steel shaft is heated evenly by being coated with the hour, so that the concrete deformation is only caused by the external temperature deformation. The temperature deformation of the steel shaftexpansion is approximately 150of μm as the temperature from thermal insulating material. thermal coefficient Invar is not a constant,varies but rather temperature deformation of The the steel shaft is approximately 150 μm as the temperature varies from 20 °C to 70 °C. However, the deformation magnitude of the concrete specimen is 1.6 μm, which is a20parameter related to the temperature. Themagnitude average thermal coefficient steel shaft °C to 70 °C. However, the deformation of the expansion concrete specimen is of 1.6the μm, which is only approximately 1% of the steel shaft, and negligibly small. − 6 ◦ ◦ ◦ adopted by the TSTM 1.5 shaft, × 10 and / C for the temperature ranging from −20 C to 100 C. only approximately 1%system of the is steel negligibly small. The of the TSTM, mainly the steel shaft, willvaries transfer When thetemperature temperaturedeformation of the steel shaft varies from 20 ◦ C contributed to 60 ◦ C, theby concrete deformation by The temperature deformation of the TSTM, mainly contributed by the steel shaft, will transfer into the concrete if the compensation method was not adopted. Consequently, the motor will start, only 0.3 µm (the measuring distance is 1 m). However, the temperature deformation of concrete varies into the concrete if the compensation method was not adopted. Consequently, the motor will start, caused pseudo-deformation. The results indicate compensation method of the ◦ Cthe ◦ C. from 0.3by µmthe to 1.6 µm when the temperature increases from that 60 to 70 caused by the pseudo-deformation. The results indicate that the compensation method of the temperature deformation with different steel materials combined with each other can eliminate the temperature deformation with different steel materials combined with each other can eliminate the vast majority of the effects of the external temperature deformation on the concrete specimen. vast majority of the effects of the external temperature deformation on the concrete specimen.

Materials 2017, 10, 419 Materials 2017, 10, 419

13 of 18 13 of 18

Figure 13. Verification test results of the compensation method for temperature deformation. Figure 13. Verification test results of the compensation method for temperature deformation.

4.3. The Results of the TSTM Performance Verification Experiment The concrete temperature is kept constant at 20 ◦ C and the testing time is approximately one  Temperature Regulation Results hour, so that the concrete deformation is only caused by the external temperature deformation. The temperature deformation of the steel shaft the is approximately 150 µm asofthe temperature varies During the temperature following phase, concrete temperatures four TSTMs develop ◦ C to 70 ◦ C. However, the deformation magnitude of the concrete specimen is 1.6 µm, which from 20 synchronization in Figure 10, which shows that the hydration degree and quality of the concrete of is of theItsteel and negligibly small. quality of the different TSTMs can theonly fourapproximately TSTMs differ 1% slightly. alsoshaft, indicates that the concrete The temperature deformation of the TSTM, mainly contributed by the steel transfer into be guaranteed to be similar to the greatest extent by being vibrated onshaft, the will vibrating stand the concrete if the compensation method was not adopted. Consequently, the motorresults will start, caused synchronously. During the design cooling and rapid cooling phases, temperature of the four by the pseudo-deformation. The results indicate that the compensation method of the temperature TSTMs and the environment show small differences. The maximal temperature difference between deformation withthe different steel materials combined withthan each1other can eliminate the vast of the surface and internal concrete can be smaller ˚C when the cooling ratemajority is 1 °C/h. the effects four of the external on the concrete specimen. Although TSTMs aretemperature placed in thedeformation walk-in environment simulation laboratory, covering an area of approximately 40 square meters, the temperature can be controlled effectively as a pre-set. The 4.3. The Results of the TSTM Performance Verification Experiment temperature gradient of the four TSTMs is small, at 1 °C. Multiple temperature conditions, such as •the semi-adiabatic Temperature Regulation Results temperature rise experiment for concrete, the design temperature history regulation, and rapid cooling can be implemented, which is particularly applicable for research and During the temperature following phase, the concrete temperatures of four TSTMs develop engineering. synchronization in Figure 10, which shows that the hydration degree and quality of the concrete of  four Temperature Stress Test Results of Multi-TSTMs the TSTMs differ slightly. It also indicates that the concrete quality of the different TSTMs can be guaranteed to be similar to the greatest extent by beingdifferent vibrated setting on the vibrating stand Four TSTMs are operated synchronously with conditions. Thesynchronously. specimen of During design cooling rapid coolingisphases, temperature of theand fourstrain TSTMs and the TSTM-Dthe cracks first when and the temperature reduced to 25.67 °C.results The stress curves are environment show small differences. The maximal temperature difference between the surface and the drawn in Figure 14. If the concrete is compressive, then the stress is expressed as a negative value. ◦ C when the cooling rate is 1 ◦ C/h. Although four TSTMs are internal concrete can be smaller than 1 The strain is expressed as a negative value when the concrete shrinks, and vice versa. During the placed in the walk-in environment laboratory, area of approximately 40 square temperature following phase, the simulation concrete swells as thecovering concretean temperature increases. However, meters, the temperature can be controlled effectively as a pre-set. The temperature gradient of the the stress increases only 0.1 MPa due to the effect of creep and relaxation. During the cooling phase, ◦ C. Multiple temperature conditions, such as the semi-adiabatic temperature four TSTMs is small, at 1 the concrete shrinks obviously. When the strain of the concrete exceeds 2 με, the motor adjusts the rise experiment for concrete, theand design temperature history regulation, rapida cooling can be concrete to its original position maintains the deformation at zero toand achieve 100% restraint implemented, which is particularly applicable for research and engineering. degree condition. The failure stress of TSTM-D is 1.35 MPa and the cumulative elastic strain is 37.6 με, while the strain of free specimen on TSTM-A is approximately 230 με. The failure strain is small • Temperature Stress Test Results of Multi-TSTMs because the shrinkage strain transforms to creep strain. Four TSTMs are operated synchronously with different setting conditions. The specimen of TSTM-D cracks first when the temperature is reduced to 25.67 ◦ C. The stress and strain curves are drawn in Figure 14. If the concrete is compressive, then the stress is expressed as a negative value. The strain is expressed as a negative value when the concrete shrinks, and vice versa. During the temperature following phase, the concrete swells as the concrete temperature increases. However, the stress increases only 0.1 MPa due to the effect of creep and relaxation. During the cooling phase, the concrete shrinks obviously. When the strain of the concrete exceeds 2 µε, the motor adjusts the

Materials 2017, 10, 419

14 of 18

concrete to its original position and maintains the deformation at zero to achieve a 100% restraint degree Materialscondition. 2017, 10, 419 The failure stress of TSTM-D is 1.35 MPa and the cumulative elastic strain is 37.6 14 ofµε, 18 while the strain of free specimen on TSTM-A is approximately 230 µε. The failure strain is small Materials 10, 419 strain transforms to creep strain. 14 of 18 because the 2017, shrinkage

Figure Thestress stressand and strain strain results Figure 14.14.The The resultsofof ofTSTM-D. TSTM-D. Figure 14. stress and strain results TSTM-D.

The restraint degree TSTM-Cisis50% 50%and and TSTM-A TSTM-A isisthe free specimen. The The cracking The restraint degree of of TSTM-C thereference reference free specimen. cracking The restraintofdegree of is TSTM-C 50% is and is the reference freeThe specimen. The cracking temperature TSTM-C 20.7 °C, is which 4.97TSTM-A ˚C less than that of TSTM-D. results show that temperature of TSTM-C is 20.7 ◦°C, which is 4.97 ◦˚C less than that of TSTM-D. The results show that temperature of TSTM-C is 20.7 C,under whichthe is strong 4.97 Crestraint less than that of The TSTM-D. results the concrete cracks more easily condition. failureThe stress is 1.02show MPa that the concrete cracks more easily under the strong restraint condition. The failure stress is 1.02 MPa and the cracks failure more strain easily is 115.75 με. The of TSTM-C is far greater than that isof1.02 TSTM-D the concrete under the failure strong strain restraint condition. The failure stress MPa and and the strain isrestraint 115.75 με. The The failure strain ofis TSTM-C is far than that of TSTM-D duefailure tostrain the different degree. creep smaller thatgreater of TSTM-D. The specific the failure is 115.75 µε. The failure strain ofstrain TSTM-C is farthan greater than that of TSTM-D due to due to the and different degree. TheFigure creep 15, strain smaller than that of TSTM-D. The specific stress strain restraint curves drawn andisthan the cracking results areThe listed in Table 4. and the different restraint degree.are The creepinstrain is smaller that of TSTM-D. specific stress stressFigure and 15 strain curves are drawn in Figure 15, 50% andofthe cracking results are listed in that Table 4. shows that the strain of TSTM-C is almost TSTM-A at any time, which indicates strain curves are drawn in Figure 15, and the cracking results are listed in Table 4. Figure 15 shows that Figure shows that the strain of TSTM-C is almost 50%restraint of TSTM-A at any time, which specimen indicates that the15self-developed TSTM system can simulate different degrees and the reference the strain of TSTM-C is almost 50% of TSTM-A at any time, which indicates that the self-developed can be selected with other parallel the self-developed TSTM system canspecimens. simulate different restraint degrees and the reference specimen TSTM system can simulate different restraint degrees and the reference specimen can be selected with can be selected with other parallel specimens. other parallel specimens.

Figure 15. The stress and strain results of TSTM-A and TSTM-C.

Figure stress strain results of TSTM-A TSTM-C. Figure 15. 15. TheThe stress andand strain results of TSTM-A andand TSTM-C.

derived from Figure 17 is approximately 8.8 GPa, but the compression elastic modulus of concrete is 26.1 GPa, determined by the active method on TSTM-B, shown in Figure 16. The difference between the results in tension and in compression may be because the sample is damaged during the active push process every 12 h. This can be supported by the phenomenon that the failure stress is only 0.93 MPa andMaterials only 68.8% of419 the failure stress of TSTM-D (1.35 MPa). However, the stress ratio is 68.8% of the 2017, 10, 15 of 18 free specimen/restraint specimen, which conflicts with the universally accepted research conclusion that the failure stress of the restraint sample is smaller than the free sample in the same experiment Cracking results of TSTM experiments. condition. The ratio of the stressTable to the4. tensile strength at failure is 0.6–0.9 for the restraint specimen [17,30–32]. TSTM B C D The results indicate that the “active method of determining elastic modulus” may damage the Tensile (MPa) 0.93 1.02 be greatly1.35 sample. The micro-damage is Strength not obvious in compression, but could affected in tension. Failure Strain (µε) 111.1 115.75 37.6 The difference of E values derived from the active method and the standard method Cracking Temperature (◦ C) 20.7 25.67 is small, so the active method is also an effective method. However, the active method is not recommended to be applied on the restraint specimen of the TSTM because the frequent tension and compression will • E Results influence the thermal and mechanical properties of the concrete under the effects of restraint and temperature. The E values of the active method and the standard method show little difference, TSTM-B is set as an almost free specimen to evaluate the elasticity modulus every 12 h. The E–time indicating that the E value does not depend on the dimension of the sample [33], but could depend relationship is shown in Figure 16 and a fitting equation is listed as Equation (6): on the curing temperature of the concrete [34]. Thus, a specific TSTM, without any other application, such as a restraint test, could used the160 active The E =be5.2 × lnfor + 0.6165 (20Ehwith ≤t≤ h; r method. = 0.94285 (t)determining ) standard method (6) is also practical for temperature stress testing under the same temperature and humidity conditions.

Figure Figure16. 16.The Thecompressive compressivestress-strain stress-straincurve curvefor fordetermining determiningthe theelasticity elasticitymodulus modulusof ofTSTM-B. TSTM-B.

The specimen is pulled to failure at the rate of 0.3 mm/min after the specimen of TSTM-C cracks; meanwhile, the concrete age is approximately six days. The stress-strain curve is drawn in Figure 17, in which the failure stress is 0.93 MPa and the failure strain is 111.1 µε. The tensile elastic modulus derived from Figure 17 is approximately 8.8 GPa, but the compression elastic modulus of concrete is 26.1 GPa, determined by the active method on TSTM-B, shown in Figure 16. The difference between the results in tension and in compression may be because the sample is damaged during the active push process every 12 h. This can be supported by the phenomenon that the failure stress is only 0.93 MPa and only 68.8% of the failure stress of TSTM-D (1.35 MPa). However, the stress ratio is 68.8% of the free specimen/restraint specimen, which conflicts with the universally accepted research conclusion that the failure stress of the restraint sample is smaller than the free sample in the same experiment condition. The ratio of the stress to the tensile strength at failure is 0.6–0.9 for the restraint specimen [17,30–32]. The results indicate that the “active method of determining elastic modulus” may damage the sample. The micro-damage is not obvious in compression, but could be greatly affected in tension. The difference of E values derived from the active method and the standard method is small, so the active method is also an effective method. However, the active method is not recommended to be applied on the restraint specimen of the TSTM because the frequent tension and compression will influence the thermal and mechanical properties of the concrete under the effects of restraint and temperature. The E values of the active method and the standard method show little difference, indicating that the E value does not depend on the dimension of the sample [33], but could depend on the curing temperature of the concrete [34]. Thus, a specific TSTM, without any other application, such

Materials 2017, 10, 419

16 of 18

as a restraint test, could be used for determining E with the active method. The standard method is 16 of 18 temperature stress testing under the same temperature and humidity conditions.

Materials 2017, 10, 419 also practical for

Figure of aa specimen specimen in in TSTM-B. TSTM-B. Figure 17. 17. The The tensile tensile stress-strain stress-strain curve curve of

5. Conclusions 5. Conclusions TheThe self-developed federatedcontrol controlTSTM TSTM system is effective an effective self-developedsynchronous synchronousclosed closed loop loop federated system is an device device for researching the mechanical and thermal properties, and it has the following characteristics: for researching the mechanical and thermal properties, and it has the following characteristics: 1. 1. It has thethe comprehensive ability of simulating different restraint degrees, multiple temperature It has comprehensive ability of simulating different restraint degrees, multiple temperature andand humidity modes, and closed-loop control of multi-TSTMs during one test period. humidity modes, and closed-loop control of multi-TSTMs during one test period. 2. The environment simulation laboratory system, including a walk-in environment simulation 2. The environment simulation laboratory system, including a walk-in environment simulation laboratory, allow for multi-concrete specimens of temperature stress tests to be conducted laboratory, allow for multi-concrete specimens of temperature stress tests to be conducted synchronously under the same environmental conditions. Additionally, no obvious fluctuation synchronously under the same environmental conditions. Additionally, no obvious fluctuation of of the concrete temperature appears before or after concrete specimen demolding. The operation the concrete temperature appears before or after concrete specimen demolding. The operation is more flexible during the test as the TSTM is no more restrained by the small chamber and is more flexible during the test as the TSTM is no more restrained by the small chamber and fluid-cooled formwork. Finally, all parts of the TSTM are in the same environment, which fluid-cooled formwork. Finally, all parts of the TSTM are in the same environment, which reduces reduces the disturbance caused by the external temperature. the disturbance caused by the external temperature. 3. The deformation measuring design with embedded cross part can obtain more accurate The deformation measuring design with embedded cross part can obtain more accurate 3. deformation and restraint degree results with little damage. deformation and restraint degree results with little damage. 4. The compensation method of temperature deformation with different considerations of steel 4. The compensation method of temperature deformation with different considerations of steel materials can eliminate the external temperature deformation’s effects on the concrete specimen. materials can eliminate the external temperature deformation’s effects on the concrete specimen. 5. Some design concepts, such as the sliding sleeve on the steel shaft, height-adjustable ball 5. supporting Some design concepts, such as the sliding sleeve on mold the steel shaft, platform, height-adjustable points, cross-head positioning fixture, and the assembly have beenball supporting points, cross-head positioning fixture, and the mold assembly platform, have been taken into consideration so that the self-developed TSTM has greatly overcome the difficulty of taken into so that the self-developed TSTM has greatly overcome the difficulty of eccentric forceconsideration and deformation. force anddesign deformation. 6. Theeccentric detachable mold of TSTM guarantees that the concrete quality of different TSTMs is 6. similar The detachable mold design TSTM guarantees the concrete of different TSTMs is to the greatest extent by of being vibrated on thethat vibrating stand quality synchronously. similar to the greatest extent by being vibrated on the vibrating stand synchronously. 7. The active method should be used for determining E on a specific TSTM without any other 7. application, The active method shouldtesting, be used for determining E on a specific TSTM without other such as restraint because the frequent tension and compression willany cause application, such as restraint testing, because the frequent tension and compression will cause damage which will influence the mechanical properties in tension. damage which will influence the mechanical properties in tension. Supplementary Materials: The following are available online at http://www.mdpi.com/1996-1944/10/4/419/s1, Video S1: Mold assembling process, S2: Hoisting process,online S3: Demolding process. Supplementary Materials: The following are available at http://www.mdpi.com/1996-1944/10/4/419/s1, Video S1: Mold assembling process, S2: Hoisting process, S3: Demolding process. Acknowledgments: This work was supported by the National Natural Science Foundation of China (No. Acknowledgments: This work supported by of theTsinghua National Natural Science Foundation of China (No. 51579134 51579134 and No. 51339003), thewas Research Fund University (Grant No. 20161080079) and the and No. 51339003), the Research Fund of Tsinghua University (Grant No. 20161080079) and the National Basic National Basic Research of China (973Grant Program, Grant No. 2013CB035902). The costs toinpublish in open Research Program of Program China (973 Program, No. 2013CB035902). The costs to publish open access have access beenby covered by the funding. beenhave covered the funding. Author Contributions: He Zhu, Qingbin Li, and Yu Hu developed the TSTM system. He Zhu, Qingbin Li, and Yu Hu conceived and designed the experiments. He Zhu performed the experiments, analyzed the data, and wrote the initial draft of the manuscript; and Qingbin Li and Yu Hu reviewed and contributed to the final manuscript.

Materials 2017, 10, 419

17 of 18

Author Contributions: He Zhu, Qingbin Li, and Yu Hu developed the TSTM system. He Zhu, Qingbin Li, and Yu Hu conceived and designed the experiments. He Zhu performed the experiments, analyzed the data, and wrote the initial draft of the manuscript; and Qingbin Li and Yu Hu reviewed and contributed to the final manuscript. Conflicts of Interest: The authors declare no conflict of interest.

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.

17. 18. 19. 20.

21.

American Concrete Institute. Report on Thermal and Volume Change Effects on Cracking of Mass Concrete; American Concrete Institute (ACI): Farmington Hills, MI, USA, 2007. Claisse, P.A. Chapter 23—Creep, shrinkage, and cracking of concrete. In Civil Engineering Materials; Butterworth-Heinemann: Boston, MA, USA, 2016; pp. 241–249. Bentur, A.; Kovler, K. Evaluation of early age cracking characteristics in cementitious systems. Mater. Struct. 2003, 36, 183–190. [CrossRef] Chen, B.; Sun, W.; Ding, J. Concrete crack resistance research based on temperature stress test—A short review. J. Chin. Ceramic Soc. 2013, 41, 1124–1133. Rilem, T.C. Avoidance of Thermal Cracking in Concrete at Early Ages; E&FN Spon: London, UK, 1998; pp. 461–464. Riding, K.A.; Poole, J.L.; Schindler, A.K.; Juenger, M.C.G.; Folliard, K.J. Quantification of effects of fly ash type on concrete early-age cracking. Aci. Mater. J. 2008, 105, 149–155. Thermal Cracking in Concrete at Early Ages: Proceedings of the International RILEM Symposium; Springenschmid, R. (Ed.) E&FN Spon: London, UK, 1994. Prevention of Thermal Cracking in Concrete at Early Ages; Springenschmid, R. (Ed.) E&FN Spon: London, UK, 1998. Breitenbücher, R. Investigation of thermal cracking with the cracking-frame. Mater. Struct. 1990, 23, 172–177. [CrossRef] Kim, J.J.; Jeon, S.; Kim, J. Development of new device for measuring thermal stresses. Cem. Concr. Res. 2002, 32, 1645–1651. [CrossRef] Amin, M.N.; Kim, J.; Lee, Y.; Kim, J. Simulation of the thermal stress in mass concrete using a thermal stress measuring device. Cem. Concr. Res. 2009, 39, 154–164. [CrossRef] Springenschmid, R.; Breitenbiicher, R. Are low heat cements the most favourable cements for the prevention of cracks due to heat of hydration? Concr. Precasting Plant Technol. 1986, 52, 704–711. Staquet, S.; Delsaute, B.; Darquennes, A.; Espion, B. Design of a Revisited TSTM System for Testing Concrete Since Setting Time under Free and Restraint Conditions; RILEM Publications SARL: Paris, France, 2012. Kovler, K. Testing system for determining the mechanical behaviour of early age concrete under restrained and free uniaxial shrinkage. Mater. Struct. 1994, 27, 324–330. [CrossRef] Bloom, R.; Bentur, A. Free and restrained shrinkage of normal and high-strength concretes. Aci Mater. J. 1995, 92, 211–217. Zhihai, L. Quantitative Evaluation of the Effectiveness of Expansive Concrete as a Countermeasure for Thermal Cracking and the Development of its Practical Application. Ph.D. Thesis, University of Tokyo, Tokyo, Japan, September 2006. Zhang, T.; Qin, W. Tensile creep due to restraining stresses in high-strength concrete at early ages. Cem. Concr. Res. 2006, 36, 584–591. Altoubat, S.A.; Lange, D.A. Creep, shrinkage, and cracking of restrained concrete at early age. Aci. Mater. J. 2001, 98, 323–331. Altoubat, S.A. Early Age Stresses and Creep-shrinkage Interaction of Restrained Concrete. Ph.D. Thesis, University of Illinois at Urbana-Champaign, Urbana, IL, USA, 2000. Van Breugel, K.; Lokhorst, S.J.; Kovler, K.; Bentur, A. Stress-based crack criterion as a basis for prevention of through-cracks in concrete structures at early ages. In International RILEM Conference on Early Age Cracking in Cementitious Systems; RILEM Publications SARL: London, UK, 2003. Bjøntegaard, Ø.; Sellevold, E.J. The temperature-stress testing machine (TSTM): Capabilities and limitations. In First International Rilem Symposium on Advances in Concrete through Science and Engineering; E&FN Spon: London, UK, 2004.

Materials 2017, 10, 419

22.

23. 24.

25.

26. 27. 28. 29. 30. 31.

32.

33.

34.

18 of 18

Springenschmid, R.; Breitenbücher, R.; Mangold, M. Development of the cracking frame and the temperature-stress testing machine. In Thermal Cracking in Concrete at Early Ages; E&FN Spon: London, UK, 1994. Lura, P. Autogenous Deformation and Internal Curing of Concrete. Ph.D. Thesis, Delft University of Technology, Delft, Netherlands, April 2003. Pigeon, M.; Toma, G.; Delagrave, A.; Bissonnette, B.; Marchand, J.; Prince, J.C. Equipment for the analysis of the behaviour of concrete under restrained shrinkage at early ages. Mag. Concr. Res. 2000, 52, 297–302. [CrossRef] Cusson, D.; Hoogeveen, T. An experimental approach for the analysis of early-age behaviour of high-performance concrete structures under restrained shrinkage. Cem. Concr. Res. 2007, 37, 200–209. [CrossRef] Wei, Y.; Hansen, W. Tensile creep behavior of concrete subject to constant restraint at very early ages. J. Mater. Civil. Eng. 2013, 25, 1277–1284. [CrossRef] Hu, S.; Chen, J. Principle and Equipment for Concrete Temperature Stress Detection; The Defense Industry Publication Agency: Beijing, China, 2008. (In Chinese) American Concrete Institute. Report on Early-age Cracking: Causes, Measurement, and Mitigation; American Concrete Institute (ACI): Farmington Hills, MI, USA, 2010. China, Ministry of Water Resources. Test code for hydraulic concrete. In SL 352-2006; China Water Power Press: Beijing, China, 2006. Riding, K.A.; Poole, J.L.; Schindler, A.K.; Juenger, M.C.; Folliard, K.J. Effects of construction time and coarse aggregate on bridge deck cracking. Aci. Mater. J. 2009, 106, 448–454. Shen, D.; Jiang, J.; Shen, J.; Yao, P.; Jiang, G. Influence of curing temperature on autogenous shrinkage and cracking resistance of high-performance concrete at an early age. Constr. Build. Mater. 2016, 103, 67–76. [CrossRef] Klausen, A.E.; Kanstad, T.; Bjøntegaard, Ø.; Kollegger, J.; Hellmich, C.; Pichler, B. Updated temperature-stress testing machine (TSTM): Introductory tests, calculations, verification, and Investigation of variable fly ash content. In Proceedings of 10th International Conference on Mechanics and Physics of Creep, Shrinkage, and Durability of Concrete and Concrete Structures, Vienna, Austria, 21–23 September 2015. Delsaute, B.; Boulay, C.; Granja, J.; Carette, J.; Azenha, M.; Dumoulin, C.; Karaiskos, G.; Deraemaeker, A.; Staquet, S. Testing concrete E-modulus at very early ages through several techniques: An Inter-laboratory comparison. Strain 2016, 52, 91–109. [CrossRef] Li, Q.; Guan, J.; Wu, Z.; Dong, W.; Zhou, S. Equivalent maturity for ambient temperature effect on fracture parameters of site-casting dam concrete. Constr. Build. Mater. 2016, 120, 293–308. [CrossRef] © 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).