Properties in fresh state

All tests aiming at evaluation of properties in fresh state were initiated soon after mixing.

In examination of concrete consistence in terms of flowability, being the first of workability properties tested, two conceptually similar measuring techniques having relation to yield stress were utilized. The first was a so-called mini-cone test following DIN EN 1015-3, which is a method commonly used for examining fresh mortars. Herein, it was applied to study the properties of finely grained UHPC acknowledging its mortar-like aggregate composition, but also of UHPC mixtures with coarse aggregate21. The apparatus is shown in Figure 3.4. In contrast to standardized equipment, a Haegermann cone was used but without flow table which was replaced by a base plate, altogether bringing about exclusion of jolting/beats during testing.

In parallel to the test or as the test alternative, some batches of both types of UHPC were studied using the larger and fully automatic custom-built set-up, see Figure 3.6. This apparatus was prepared so as to fulfil the requirements of DAfStb code of practice for self-

21 _{It was taken into account the low number of coarse-grained UHPC mixtures tested and success of application}

in previous study on UHPC [Dud 08b].

Figure 3.5: Details of testing equipment geometry and information record (to the left) and typical concrete cake

obtained from the mini-cone test (to the right).

70 [mm]

100

60

d1 d2

compacting concrete [DAfStb 03] and method sometimes referred to as inverted slump flow test method. However, the blocking ring (so-called J-ring) was not used.

During the experiments, two testing conditions were applied. Main condition used for both kinds of tests is referred hereafter as ‘moist’, according to which the testing equipment was moisturized prior to testing. Such preparation step is intended whenever testing of a self- compacting concrete is considered, see e.g. [DIN EN 12350-8][DAfStb 03][EFNARC 02]. In the procedure, treatment with damp sponge was chosen to avoid the aquaplaning phenomenon. For comparison, the condition addressed as ‘dry’ was used as foreseen in DIN EN 1015-3 standard, i.e. using dry equipment. According to [Sta 04a], distinction between different states of moisture of the testing equipment might not be necessary for mixtures of high viscosity (therefore UHPC as well) since having little impact on slump flow results. However, this sometimes changes for viscous concretes containing fibres and depends on the type of fibre used. Thus in own study, second slump flow test using dry equipment was performed simultaneously, although mainly for the mini-cone test due to small amount of material needed for the test.

The tests were completed when no particular change of the slump-flow (= the flow spread) was recorded. Subsequently, the final value of spread was read and calculated either manually or automatically from the two perpendicular diameters of the concrete ‘cake’ obtained.

A visual observation was carried out on occasion of each test performed to determine the

potential for bleeding and segregation. No other measures were taken in this respect.

Figure 3.6: Custom-built set-up used in measurement of slump flow with large cone.

Hollow truncated cone (Abrams cone), acc. to DIN EN 12350-2, with

dtop : dbottom : h = 200 : 100 : 300 mm Data aquisition system

Base plate 1000 x 1000 mm² One of the four sensor-equipped spread lines

The rheological behaviour of selected mixtures, that is which presented target slump-flow

after production from one premix (i.e. matching to that of control mix), was studied by employing a concrete rheometer called HAAKE MARS II (Modular Advanced Rheometer System). This Thermo Electron Corporation product is a multicomponent device, the main removable and operational parts of which are measuring cell and a rotor, see. Figure 3.7. The former referred to as unit cell is internally ribbed with demountable longitudinal lamellas. The rheometer itself can operate under one of two main regimes, i.e. defined rotational speed/shear rate (so-called CR mode) or specified torque/shear stress (so-called CS mode); in return, it gives corresponding values of torque force/shear stress or rotational speed/shear rate, respectively. The manual control over apparatus is taken with help of software called RheoWin. With this tool, the test can be performed in one of the two modes, either rotational mode or oscillatory test. The selfsame software can be used for data processing as well.

Because distinction between the two test modes is important for rheological characterization, keeping the goals of study in mind, the experimental programme as given in Table 11 was set. Figure 3.7: Rheometer test set-up.

Measuring head with engine and measuring unit Control panel Holder for temperature control unit Measuring elements (rotor and temperature gauge) Cf. bottom image Rotor Unit cell Frame

Table 3.11: Measuring profile for the rheological testing.

Seg-

ment Mode type Mode type

Duration

[s] Further settings

0 • Lift • • • Setting the rotor in the measurement position

1 • Pre-shear • CR-Time • 30 • γ
=1.7s-1 (corresponding to = 5.2 [min-1]), 10 steps 2 • CR-Time • 60 • γ
=0s-1, 10 steps

3 • Rotation (1st) • CR-RS • 120 • Controlled shear rate test, γ
=0 −3.2s-1 (corresponding to = 0-10 [min-1_{]), 60 steps}

4 • CR-Time • 30 • Controlled constant shear rate γ
=3.2 s-1 (corresponding to = 10 [min-1_{]), 10 steps}

5 • CR-RS • 150 • Controlled shear rate test, γ
=3 −.2 0s-1 (corresponding to = 10-0 [min-1_{]), 75 steps}

6 • Oscillation • CS-Time • 45x60 • Oscillation time test, torque M = 0.5 [mNm], f = 1.0 [Hz], 90 steps

7 • Pre-shear • CR-Time • 30 • γ
=1.7s-1 (corresponding to = 5.2 [min-1]), 10 steps 8 • CR-Time • 60 • γ
=0s-1, 10 steps

9 • Rotation (2nd) • CR-RS • 120 • Controlled shear rate test, γ
=0 −3.2s-1 (corresponding to = 0-10 [min-1_{]), 60 steps}

10 • CR-Time • 30 • Controlled constant shear rate γ
=3.2 s-1 (corresponding to = 10 [min-1]), 10 steps

11 • CR-RS • 150 • Controlled shear rate test, γ
=3 −.2 0s-1 (corresponding to = 10-0 [min-1_{]), 75 steps}

Generally speaking, to test effect of IC but also to trace change of rheological parameters, the test profile was composed of two identical rotational modes with separation by a mode of uninterrupted structure reconstruction. During both rotational modes, rather standard measuring sequence known more for cement paste was applied. This means, among other, execution of a so-called pre-shear (i.e. shearing at constant rate) before the main measurement. Such step and in fact stabilization time was necessary to reduce thixotropy effects on one hand and, as for first rotation, remove any traces of the UHPC characteristic elephant skin or agglomerations on the other hand. The main difference to the cement paste was usage of other, lower range of rotational speed (shear rates) leading to higher viscosities obtained. In comparison, the settings of latter test, i.e. second testing mode, were meant to maintain oscillation in the linear elastic range and were fixed based on preliminary studies on this material (amplitude sweep test). With this procedure, all measurements were conducted precisely 12 minutes after addition of water and meaning 5 minutes rest from mix production as it was minimal time for sample casting and transportation to the rheometer. The tests were then processed at a constant temperature of 23 ºC, although the temperature of the material inside the cell varied slightly from the walls to the centre.

The results acquired included torque as a function of rotational speed, or complex viscosity and other parameters of oscillation test in the time scale, presentation of which can be found elsewhere, see Appendix F or otherwise [TC 225-SAP] for the former part. After including calibration of test equipment done by Secrieru [Sec 12], the principal data have become shear stress and shear rate. The ratio of the two at a given shear rate, often addressed as so-called apparent or dynamic viscosity could be then calculated, see corresponding equation in Table 3.10. To determine the parameters yield stress (τ0) and plastic viscosity ( ), being less direct

procedure, Bingham model was assumed. In accordance, the regression analysis was performed for both up-curves (increasing shear rate) and down-curves (decreasing shear rate) from the shear stress-shear rate data obtained; for the purpose, only main data, i.e. without pre-shear underwent processing. The slope of curves was used to calculate the plastic viscosity, while the intercept on the stress axis at zero shear rate was utilized to derive the yield stress. Since so obtained yield stress values were found always positive, application of other fitting models was not considered.

Table 3.10: Calculation basis for finding rheological parameters.

Parameter Calculation formula

Shear stress (τ), Pa ₁₀−6 ⋅ ⋅ =Md Arheo τ Where: Md – torque/torsional moment [µNm]

Arheo- shear stress (calibration) factor (= 4750 Pa/Nm [Sec 12])

Mrheo – geometry (calibration) factor (= 3.0441 (1/s)/(rad/s) [Sec

12]) Shear (strain) rate (γ
), 1/s

rheo M ⋅ Ω ⋅ Π = 60 2 γ
Where:

– rotation/angular velocity [1/min] Mrheo – as in foregoing parameter

Apparent/dynamic viscosity η, Pa s γ τ η
= Complex viscosity ǀη*ǀ, Pa s ω η * * ₌ G Where:

ǀG*ǀ – complex shear modulus [Pa]

ǀG*ǀ = [(τ/(Φ*Mrheo)*cos δ)2 +(τ/( Φ*Mrheo)*sin δ)2]0.5

Φ – deformation corresponding arch [-] Mrheo – as in foregoing parameters

δ – phase difference angle [rad] ω – angular frequency [rad/s]

The air content of finely grained and coarse-grained mixes was initially determined by involving pressure method in accordance with the standards DIN EN 1015-7 and DIN EN 12350-7, respectively. The corresponding results obtained were presented in publications [Dud 10b][Dud 10c]. However, this method has met with some criticism and doubts regarding precision of the test, including researchers investigating SAP-enriched mixtures [Has 10]. To stay on the safe side, rival two-step and density-based approach called gravimetric method [Min 81] was therefore chosen. Accordingly, the bulk density was determined first, for which DIN EN 1015-6 and DIN EN 12350-6 for concretes without and with coarse aggregates were followed, respectively. For each measurement a minimum of two samples of 1 litre volume was produced whereas, assuming concrete is self-compacting, only in special cases using vibrating table. One such exception was concrete F-S.4 (0.4 % SAP but no extra water), for which intensive vibrating had to be applied to neutralize expected loss of workability. Compaction furthermore was applied for mixes F-R-vac, F-R.04-vac and F-S.3.04-vac as one step in process of air bubble removal (the full procedure involved mixing with vacuum, casting fresh concrete in at least two layers, a heavy compaction following each concrete pouring, excessive material removal and finalization with opening of the surface air pockets). Having done so, the weight of material of known volume was used to assess the bulk density. The air content was then calculated according to Eq. 3.3 [Min 81]:

100 ⋅ − = T W T A (3.3)

where A is air content [%], T is theoretical weight of the concrete based on an air-free basis, computed from the proportions and the specific gravities of the mix components and W is unit weight of fresh concrete.

Usage of fresh concrete density instead of that presented in hardened state yielded an obvious advantage given that effect of shrinkage on concrete volume was eliminated. The error made on this occasion owed to changeable specific gravities of some ingredients, especially cement, was insignificant (< 0.5 %). The average from all test performed on particular composition is reported.

Eventually, having the knowledge about all important properties in fresh state, the self- compacting ability of concretes investigated was decided. For acceptance criteria being referred to, please see Table F.1 in Appendix F.