Mix Proportioning and Casting of Beams

Unlike conventional concrete, GPC does not have a standard mix design procedure. Hence, the final mixes were arrived based on numerous trials. To obtain the mass of NaOH and Na₂SiO₃ solutions, the ratio of NaOH to Na₂SiO₃ were used in the different proportions. Around 10 trial mixes were made by varying the proportion of various raw materials. In all the mixes a working time of at least 20 minutes was available before they started losing their fluidity. The final mix details obtained from trial studies for M30 grade concrete are given in Table 3 and the strength results are tabulated in Table 4. From Table 4 it is noted that the rate of strength gain is more in case of GPC when compared to OPC concrete.

Using these mixes, reinforced concrete beams were cast for corrosion studies. The reinforcement detail of the beams is shown in Fig. 1. The two mixes are designated as GPC1 and GPC2. Control beams were cast for comparison and validation using conventional OPC concrete with grade M30 and designated as OPC, OPC 10 and OPC 20 representing without corrosion and corrosion for 10 and 20 years respectively. The beams are designated as GPC1, GPC1-10 and GPC1-20 refers to GPC beam without corrosion and with activated corrosion of 10 and 20 years.

Accelerated Corrosion Test on Concrete Beams

To study the effect of corrosion, the beams were subjected to accelerated corrosion by impressing current to the reinforcement bars and immersing the beams in sodium chloride solution. The accelerated corrosion test setup is shown in Fig. 2. Based on the chemical composition of the reinforcement obtained from the manufacturer, the equivalent weight was obtained as 361.01 g eq. The corrosion rate is assumed as 0.2 mm/year which is followed for very severe corrosion. Hence the mass loss is calculated as below. Mass Loss = Area * Thickness loss * Density

= 1*0.02*7.840

= 0.1568 g / year / cm² Mass Loss for 10 yrs = 1.568 g / cm² Mass Loss for 20 yrs = 3.136 g / cm²

The mass of rust produced per unit surface area of the metal due to applied current over a given time can be determined theoretically using the Faraday’s Law as given below:

Theoretical Mass loss,

Where,

M_th - Mass of rust loss (g/cm²)

W - Equivalent Weight of steel (g)

I - Current applied (Amp/cm²)

T - Duration (sec)

F - Faraday’s Constant (96487 Amp-sec)

For simulating 10 years corrosion to occur in 7 days, from equation 1, the impressed current is calculated as 0.00069 Amp/cm² in 7 days.

Hence the current for simulating 10 years corrosion is = 4208.40*0.00069 = 2.9038 ~3 Amp.

Similarly, for 20 years corrosion to occur in 7 days is 6 Amp.

Special curing tanks were cast for keeping the specimens in sodium chloride solution. The tanks were filled with 3.5 % NaCl solution and beams were immersed in the solution. Prior to that, the reinforcement bars were exposed and were connected to a constant current source for impressing current to the reinforcement. The tank with beam specimens during accelerated corrosion is shown in Fig. 2b). The accelerated corrosion test was stopped after the predesignated time period as per the above calculation.

Fig. 1.

Details of size and reinforcement in beams.

Fig. 2.

Accelerated Corrosion Test Setup: a) Typical Schematic Representation, b) Experimental Test Setup.

Test Setup and Instrumentation

After the accelerated corrosion test, the beam specimens were taken out of the tank, completely dried and white washed before carrying out the bending test. The beams were subjected to two-point bending test for evaluating the structural performance before and after corrosion. The beams were instrumented with strain gages and displacement transducers for measuring the various structural parameters. The beams were instrumented with strain gages at mid span and at quarter span locations at top and bottom to find the strain variations during the load testing. Displacement transducers were fixed at bottom of the beam specimen at mid span during load testing. A calibrated load cell was placed between the hydraulic jack and beam specimen for measuring the applied load during the testing. All the sensors were connected to a high speed data acquisition system (data logger) for continuous recording of the responses. The load was given in equal increments and at each increment the initiation of crack and propagation of crack was keenly observed and marked. The specimens were loaded till complete failure. After testing, the beams were completely broken and the reinforcements were taken out. The weight of the reinforcement after corrosion was measured and compared with the weight of the reinforcement before corrosion testing.

Visual Inspection

The test set-up was plugged off and beams were removed after 168 hours from solution tank. Around 100 hours of testing, the GPC and OPC specimens were found to be corroded. Rust formation initiated in OPC specimen after 60 hours, whereas GPC specimens were still resisting the rust formation. Brown stain due to rust was the first obvious evidence of corrosion initiation in the steel rebar and the corrosion products were also seen floating in the chloride solution tank. A crack in the OPC beams was discovered after over 100 hours. The crack continued to propagate along the longitudinal direction. After 160 hours of testing, a vertical crack was observed, which grew bigger under flexure. The GPC beams had no cracks, clearly showcasing their better corrosion resistance compared to OPC. GPC specimens showed better durability characteristics than OPC specimens. GPC’s corrosion resistance to accelerated corrosion exposure is considerably enhanced by several variables such as high alkalinity, low permeability and better microstructure.

Load vs. Deflection Behaviour

The maximum load carried by GPC beams was higher than that of OPC beam. However, a distinct difference is evident in the shape of the load-deflection curve. The load vs. deflection behaviour of the OPC beam can be categorised as a trilinear curve, with a very distinct yield, maximum and ultimate load points that can be located graphically itself. In case of GPC beams, the shape of the curve was curvilinear in nature, similar to a high strength steel reinforcement under tension. This could be as a result of the inherent behaviour of GPC under tension. In addition, the energy absorption of the GPC beams is similar to the OPC beam, though the peak load is greater in GPC. Once the peak load is attained, there is a curvilinear drop in the load with the increase in deflection. Hence it may be better to restrict the use at post peak strength in GPC beams to a lower limit that a OPC beam of corresponding strength and reinforcement.

On comparing the mass loss of steel, the OPC beam has more loss than GPC beams in both the mixes. Fig. 3 shows the load vs. mid span deflection curve for the tested beams. It was found that the structural capacity of the OPC beam reduced greatly with corrosion as compared to GPC beams. From Fig. 3a) it is seen that at 10 yrs. corrosion, the peak load reduces by around 50%, however it retained a load carrying capacity of 35% only. On comparing the load carrying capacity of GPC- 1 and GPC-2 beams, it reduced by 24.7% and 28.3% at 10 years corrosion. At 20 yrs. corrosion, GPC-1 and GPC-2 retained a capacity of 74.1% and 75.4% respectively in comparison to 35 % capacity retention by OPC beam. Table 5 shows the results of the tested beams.

After the tests were completed, the beams were broke opened and the reinforcements extracted. The reinforcements were cleaned as per standard procedure and weighed for the weight loss due to corrosion. These values are given in Table 6. From the table it can be seen that the loss in the steel mass in GPC mixes were lesser than OPC mix at all ages. This establishes that GPC mixes offered better resistance to corrosion in comparison to OPC mix.

Fig. 3.

Load Vs. Deflection: a) OPC, b) GPC 1, c) GPC 2, d) Control, e) 10 years, f) 20 years.

Fig. 4.

Load Vs. Strain: a) OPC, b) GPC 1, c) GPC 2, d) Control, e) 10 years, f) 20 years.

Load vs. Strain Behaviour

Fig. 4 shows the load vs. microstrain curve for the OPC, GPC-1 and GPC-2 beams tested. From the graph, it can be seen GPC shows better performance than OPC. On comparing GPC 1 and GPC 2 plots, during the initial test the ultimate load capacity of GPC 2 is 180 kN which is superior. As the corrosion level of the beam increases, GPC 1 is observed to produce stable results. The maximum strain response of 2750 microstrain is recorded by GPC 2 specimen during test at corrosion level of 20 years whereas GPC 1 recorded a 2200 microstrain. The control specimen made up of OPC recorded less than 1000 microstrain in all the tests. These results clearly show GPC has higher strain rate compared to OPC specimens.

Mass Loss Measurements

Before the casting phase, the initial mass of rebar was noted. Following the corrosion exposure, and load testing the beams, the rebars were extracted. Metal brush and deionized water were used to remove the corrosion products and clean the rebar. The ultimate mass was calculated by weighing the rebar taken from the beams. After that, the proportion of mass loss for rebar in both OPC and GPC beams was computed. The comparative mass loss of GPC and OPC specimens are shown in percentage in Table 6. It was found that the steel rebar in the OPC beams had significant corrosion damage, but the rebar in the GPC beams had less corrosion damage than the OPC beams. The OPC beams lost a large amount of mass compared to GPC beams due to their early corrosion and crack propagation which infused chloride ions to penetrate easily. This mass loss % clearly depicts the better corrosion resistant behaviour of GPC.

Fig. 5.

Failure modes of tested specimens.

Failure Modes

Figs 5a) and b) show the typical crack patterns observed during the testing of GPC specimens. Fig. 5c) shows the broke open specimen after testing. Cracking was initiated only after the applied moment reached the cracking moment. Initial crack in GPC specimen was originated in the middle tension side of the beam under the loading point. The initial crack in GPC specimen was observed below the loading point. The initial crack load of GPC is higher than the initial crack load of OPC. Sometime after the initial crack was formed, many new cracks were emerged which propagated on continuous loading. The cracks observed in the specimens were vertical flexural cracks, runs normal to the longitudinal axis of the beam. The width of the crack was enlarged gradually due to the continuous loading and the crack propagates to the compression side of the specimen. The broke owing to concrete crushing in the middle top of beam (compression side) just under the loading point. The failure is observed at centre of the span and the point of loading. In Fig. 5b) the crushing of concrete under the load can be seen, which propagated instantaneously, followed by the rupture of the rebar (tension-compression failure). The failure of GPC specimen was ductile failure because the crack propagate from the bottom side to top of the specimen (tension to compression) and crushing of the concrete happened at the bottom side of the specimen.