High strain rate response of UHP ( FR ) C in compression

The objective of this study was to investigate the compression behaviour of the UHPFRC and its matrix (UHPC) under high strain rate. Two experimental set-ups were used for compression testing: a traditional Split Hopkinson Pressure Bars and a compression version of the Modified Hopkinson Bar. The tests were conducted, in the range of 100–500 s−1 on cylindrical specimens with both diameter and height of 20 mm. Results show significant increases in peak strength and dissipated energy.

The objective of this study was to investigate the compression behaviour of the UHPFRC and its matrix under high strain rate.Two experimental set-ups were used for compression testing: a traditional Split Hopkinson Pressure Bars [12] and a compression version of the Modified Hopkinson Bar.The tests were conducted in the range of 100-600 s −1 on cylindrical specimens with both diameter and height of 20 mm.

Materials
The materials here analysed were commercial products characterized by high strength and durability, thanks to the exceptional properties of the matrix (UHPC).UHPFRC was obtained adding to the UHPC an elevate percentage of high strength steel fibres reinforcement (3% in volume) having a diameter of 0.16 mm and a length of 13 mm.The UHP(FR)C specimens had a cylindrical shape with diameter and height of 20 mm.All specimens were drilled out from standard cube (150 mm side).The results of the quasi-static compressive tests were f C = 104 ± 22 MPa for UHPC and f C = 127 ± 21 MPa for UHPFRC.The Young's modulus was 51 GPa.Moreover UH-PFRC had high durability guaranteed by high water and low gas permeability, very high resistance to chloride penetration, carbonation, acid attack and abrasion.

Experimental set-ups
The high strain rate compressive behaviour was examined by means of two set-ups.

Split Hopkinson Pressure Bar
The dynamic compression tests were performed using a Split Hopkinson Pressure Bar (SHPB), installed at the Dynamic Testing of Materials Laboratory of the Research Institute of Mechanics -Lobachevsky State University of Nizhny Novgorod (Russia), which consists of an input (2) and an output (5) bars with the specimen (4) sandwiched between them as schematically shown in Fig. 1.When the strike bar (1) impacts onto the input bar a compressive incident wave ε I (t) travels along the input bar.Once it reaches the specimen, a reflected wave ε R (t) and a transmitted wave ε T (t) are generated, propagating along the input and output bar respectively.According to the onedimension wave propagation theory the forces and particle velocities/displacements at the two faces of specimen can be determined by those three waves recorded.In Fig. 1a is also shown the Lagrangian graph describing the strain history of the two bars.
In Fig. 2 the incident, reflected and transmitted pulses measured on input and output bars in a dynamic compression test of UHPC specimen are shown.It can be observed as at high strain rate the specimen failure is reached just in the first cycle of loading.The time necessary to bring the specimen at failure in compression is about 100 µs.
With this set-up were carried out the test with higher strain rates.

Modified Hopkinson Bar in Compression
The Modified Hopkinson Bar for compression test installed at the DynaMat Laboratory (Fig. 3) consists of a hydraulic actuator (1) that put in tension a pre-stressed loading bar (2) thanks to a blocking ring (3) placed at the other extremity of this bar (see Fig.  bar was a high strength steel bar having a diameter of 12 mm and a length of 6 m.This pre-stressed bar is directly connected to an aluminium bar with a diameter of 30 mm and a length of 3 m working as input bar (4).The specimen (5) was sandwiched between the input bar and another identical bar used as output bar (6).Pulling the pre-stressed bar it is possible to drive the test by the energy stored in it.
The test starts when a fragile bolt (see Fig. 4b) positioned between the pre-stressed bar and the hydraulic actuator (1) by suddenly breaks.Consequently a rectangular stress wave pulse is generated and propagates through the input bar, the specimen (5) and the output bar ( 6), causing a

Results
The tests performed by means of SHPB were carried out using two (aluminium and steel) strikers having a diameter of 20 mm and a length of 300 mm.To obtain the strain rates three different pressure values were used (1, 3 and 6 bars).In the MHB only one velocity were used in order to compare the UHPFRC and UHPC (the preload was 50 kN).

SHPB results
The tests on UHPC specimen were performed using the aluminium striker at 1 and 3 bars as pressure in the gas gun obtaining an impact velocity equal to 15.4 and 27.2 m/s respectively.The strain rate obtained was in average 181 and 462 s −1 .As well known also in this case it can be observed as the compressive strength increases with increasing strain rate.
In Tables 1 and 2 are collected the results at the two strain rates while in Figs. 6 and 7 the stress versus time curves are shown.

MHB results
The compression test using the MHB was performed imposing a 50 kN of preload that produce an equivalent impact of 11.3 m/s.In Table 3 and Fig. 8 the results are shown.

SHPB results
For the tests on UHPFRC the pressure in the gas gun was incremented at 3 and 6 bars so impact velocities equal to 27.2 m/s (for aluminium striker) and 20.8 m/s (for steel striker) were obtained.The resulting strain rates were respectively.466 and 688 s −1 .
The results on UHPFRC are resumed in Tables 4 and 5 and in Figs. 9 and 10.

MHB results
The tests performed by means of MHB had exactly the same condition used for UHPC.
The strain rate obtained was in average equal to 111s −1 .

Discussion
UHPC is a brittle material while UHPFRC can reduce the brittleness thanks to the addition of fibers in the mix.In particular the presence of the fibres enhances the post-peak behaviour.In Figs. 12 and 13 the stress versus displacement curves of UHPC and UHPFRC specimens tested in tension at about 450 and 700 s −1 are shown.From these results it can be easily recognized the fibres role in the post-peak behaviour.The energy dissipation capacity is enhanced because of higher stress level for the same strain level.The peak strength seems not be 01020-p.4 DYMAT 2015  influenced by the fibres presence.This can be reasonable because the principal role of the fibres is strictly related to the bridging effect.The evolution of the strength in function of the stress rate (measured as the slope of the stress versus time curve) is shown in Fig. 14.The scattering of the results can be ascribed to several motivations.Firstly, the boundary conditions of the specimens such as flatness, presence of grease or copper gasket.Secondly, the presence of the fibres can be considered a discontinuity in the materials causing the premature failure.Finally, the grade of dispersion of the fibres in the cross section is definitely different in each specimen.

Concluding remarks
The results of dynamic experimental investigations on ultra-high performance fibre reinforced concrete in compression have been described.The experiments were carried out using a Split Hopkinson Pressure Bar and a modified Hopkinson bar device.UHPFRC shows an increase of the mechanical features as post-peak strength, failure time, and absorption energy respect to UHPC.The reason for this increase in the performance can be related to the contribution of the fibres distribution.01020-p.5

EPJ Web of Conferences
The results obtained at high strain rates are the basis to interpret the results acquired during blast experiments and to assess the goodness of the modelling.

Figure 1 .
Figure 1.Dynamic compression testing set-up: a) scheme and Lagrangian graph; b) view.

Figure 2 .
Figure 2. Signals measured on the input and output of UHPC.

Figure 5 .
Figure 5. Compression test: a) input and output signals; b) incident, reflected and transmitted signals.