TBM s backfill mortars Overview Introduction to Rheological Index

Size: px

Start display at page:

Download "TBM s backfill mortars Overview Introduction to Rheological Index"

Marion Wells
6 years ago
Views:

1 Tailor Made Concrete Structures Walraven & Stoelhorst (eds) 2008 Taylor & Francis Group, London, ISBN TBM s backfill mortars Overview Introduction to Rheological Index L. Linger, M. Cayrol & L. Boutillon VINCI Construction Grands Projets, Rueil-Malmaison cedex, France ABSTRACT: The aim of this article is to give an overview of the tail void grout mortars, which are used to backfill the gap between the tunnel lining and the ground, at the rear of the Tunnelling Boring Machine s (TBM) tail skin. These apparently rustic products are actually a key issue for a TBM s Project progress. They must fulfil several criteria more or less antagonistic, regarding both fresh and hardened mortar properties. The proposed approach is built up from the outcome of a large know-how gathered on many different projects in which VINCI has been involved. In particular, it will introduce an original and promising approach: the Rheological Index. This index is a kind of generalized volumetric Water/Cement ratio, which easily provides fruitful and faithful indications regarding fresh mortar behaviour, but also concrete mixes. In other words, this original approach aims to put in equation the workability of the backfill mortars (and more generally cementitious materials). The first part of this article gives a general overview of the TBM s backfill mortars problematic. The second part details the original Rheological Index approach. The main characteristics of so-called semi-inert mortars, and associated laboratory trials, are then developed and illustrated by several examples of successful applications. 1 INTRODUCTION The ring void annulus behind the TBM tail skin can be filled by inert, semi-inert (also called semi-active) or active backfill mortars. Apart from economical considerations, the main difference between them relies on the expected achieved long term strength properties, or their possibility to be chemically accelerated just after injection. This article describes the main characteristics of the semi-inert backfill mortar type. The proposed approach results from a long experience gathered on many different projects. The grout mortar is generally batched on the surface, placed in a specialized rail car, transported to the trailing gear, pumped into a holding tank, and from there pumped into the annular void.the essential function of the mortar is to avoid destruction of the soil confinement and to lock the ring. The efficiency of the injection is linked to the quickness of the process. The injection system must maintain the pressure and carry the soil load over the lining. That s why backfill mortar pressure should be equal or greater than the face confinement pressure. The mortar must provide support and completely contain the segments. In the long term, it must provide definitive stability in confining the ring. 2 SEMI-INERT TAIL VOID MORTARS 2.1 Main principles The main principle of semi-inert mortars relies on the stability of the slice of grout between ring and soil. This stability is provided when sufficient cohesion and friction coefficient at both interfaces mortar/ring and mortar/soil is reached. Mortar, as long as it is well formulated, can pass from a fluid and thixotropic behaviour to a solid behaviour quasi-instantaneously. The wringing effect by which the mortar under pressure expels part of its water obviously participates to this transformation. But even in impermeable grounds, the sole effect of rearrangement and interlocking of the grains inside the mortar (when injected under pressure) is sufficient to provide the necessary friction at the contact surfaces. Consequently, within seconds stability of the ring can be achieved. This phenomenon depends mostly on the sand granulometry (skeleton) selection and paste (fine particles + water) volume. It is to be noted that shear parameters at the interface is difficult to measure, therefore derived criteria assessing indicative early age mortar strength and cohesiveness, based on experience, are used. 271

2 As most of the ring movements, such as the uplift, generally occur as the tail skin leaves the ring while the TBM goes forward, the stabilizing cohesion of the mortar shall therefore be provided during the injection itself and not after 50 hours nor even 3 or 4 hours later. Consequently, the problematic of backfill mortars is far more a question of fresh state than hardened state (setting time) behaviour. Therefore, seeking long term properties forgetting short term properties may seriously affect the workability limits and the overall quality of the final mortar. The mortar should not indeed be seen as a segment extension, but far more as a mean to lock the segment against the soil. Consequently, the properties to be reached by the mortar must match the soil properties in terms of cohesion, pressure and modulus, rather than micro-concrete. 2.2 Main criteria The backfill mortar main objectives and associated criteria are summarized hereafter: 1. Perfect filling of the annular void between concrete segments and excavated ground, and limited heaving of the lining during placing: Mortar must be fluid enough to transit easily all around the lining. 2. Easy-working conditions: mortar must keep its pump ability for duration compatible with the works schedule. A target value between 12 to 24 hours must be taken into account. 3. Mortar should have a short-term stiffness, which is reached by contraction of the grains after its injection under pressure in the annular space between the concrete segments and the ground. Favourable shape of the grains, consistency of the mortar, water squeezing from fresh mortar under injection pressure may ease grain contraction. 4. Long term hardening in order to prevent slow movements. It can be reached by cement setting but also by pozzolanic effect between pulverized fly ashes and Ca(OH) 2 (hydraulic lime of cement). The expected value of strength (without wringing effect) is generally around 1 MPa at 90 days. These criteria are generally reached by: The choice of a well graded sands curve in order to get the necessary mortar pump ability. The sands curve (coarse particles) is also very important in order to reach the short-term stiffness after injection behind concrete segments. The choice of a binder (and, if necessary of a specific retarder admixture) which allows a long setting time of grout mortar, and limit natural bleeding of fresh mortar in order to prevent problems during mortar transportation in steel pipes. 2.3 Components The main components, which can be used for backfill mortar mix designs, are briefly described here below Aggregates Siliceous or calcareous, rounded or crushed aggregates can be used. Nevertheless, it is mentioned that the use of a major proportion of crushed aggregates is generally better for the backfill mortar mix design (short term shear strength). On the other hand, the use of a few proportions of rounded sands is helpful for mortar pump ability. As a matter of fact, it is often necessary to mix together two or three different aggregates in order to get the most favorable granular skeleton, and increase as much as possible the fines content in aggregates (<0.080 mm, but also <0.16 and mm). The initial amount of inert fine elements brought by aggregates will determine the choice (kinds and quantities) of the other components (active or inert) for the backfill mortar mix design. Global investigations must be made in quarries or borrow pits located close to each concerned project site in terms of rubbish materials, aggregates for road construction, etc...sands or coarse aggregates suitable for classical concrete production are not necessarily the best materials for backfill mortars. It must also be emphasized that the regularity and the homogeneity of onsite deliveries toward works is of similar importance than the intrinsic characteristic of each component, defined at the beginning of trials. At last, the batching plant must be set up considering the possibility of using sand with a high proportion of fines particles (stock piles protection against rain, use of a belt extracting system rather than typical aggregate bin trapdoors) Cementitious materials Many fine powders could be used for the backfill mortar mix design. The choice will be made according to detailed characteristics, cost considerations, and preliminary laboratory tests results. Pulverized fuel ash (PFA) Hydraulic hydrated lime Limestone fillers (LF) Other fillers produced by hard rocks crushing Micro silica (MS) Metakaolin Cement. Mortar formulated with cement starts setting and hardening a few hours after batching (depends on cement type and dosage, and additional admixtures if any). On the other hand, it is mentioned that, with the use of hydraulic hydrated lime, backfill mortar longterm compressive strength development will be slower 272

and lower than in the case of a mix design formulated with Portland cement. Mortar hardening (by chemical effect) will start only some days (depends on lime nature and dosage) after batching.

Increase of fresh mortar workability and pump ability, thanks to spherical shape (PFA, MS) Production toward time of stable constituents (C-S-H) in hardened mortar by hydraulic properties (cement)

This pozzolanic effect is associated with an increase of mortar long-term strength.

The use of bentonite can also be envisaged, especially because of its beneficial effect regarding pump ability, particularly in small diameter pipes.

3 and lower than in the case of a mix design formulated with Portland cement. Mortar hardening (by chemical effect) will start only some days (depends on lime nature and dosage) after batching. Cementitious materials have different effects on grout mortar mix design: Input of high amount of fine particles (<80 µm), even ultrafine particles (<1 µm) in case of MS, in mortar mix design. Increase of fresh mortar workability and pump ability, thanks to spherical shape (PFA, MS) Production toward time of stable constituents (C-S-H) in hardened mortar by hydraulic properties (cement) and/or pozzolanic effect (combination of PFA, MS, metakaolin with calcium hydroxides Ca(OH) 2 produced during cement hydration or directly brought by hydraulic lime). This pozzolanic effect is associated with an increase of mortar long-term strength. The batching plant must be designed considering the possibility to use several powders locally available, and it should have enough silos. The use of bentonite can also be envisaged, especially because of its beneficial effect regarding pump ability, particularly in small diameter pipes. Nevertheless, backfill mortar formulated with bentonite is very sensitive to the variation of mixing time during fresh mortar batching. In order to prevent a variable behaviour of the fresh mortar, it is recommended to use slurry rather than powder in order to achieve a better dispersion of bentonite in fresh mortar. It is highlighted that the bentonite has a major negative effect on long-term mortar compressive strength Chemical admixtures According to the mortar expected properties and to the quality of materials locally available, chemical admixtures could be necessary in order to reach all the grout mortar objectives. High Retarder (stabilizer) agent Plasticizer with retarder effect. It can also been envisaged to add polypropylene fibres in backfill mortar mix design to improve specific characteristics of tail void grout: Thixotropic properties in order to minimize the leakage of fresh mortar in front of the TBM, Ability to prevent washing out after injection. Range of initial slump: 100 to 110 mm (performed with small mortar cone). Fresh mortar consistency can also been measured with mortar flow test. Preservation of slump after 24 hours: 70 to 90 mm Natural bleeding in a cylinder (V > 500 ml) after 2 hours: <1% Loss of water under pressure 1 bar during 1/2 hour: around 5%. This criterion is assessed thanks to an in house test ( small perforated and drained cube). The following pictures illustrates the high achieved cohesion in a cube of fresh mortar mm 3 just after pressurization during 1 / 2 hour, then 90 days aged (extracted by coring). 2.4 Laboratory criteria The range of criteria target values generally taken into account for a semi-active backfill mortar is summarized hereafter. 273

4 Strength at 28 days: 0.5 MPa/Strength at 90 days: 1.0 MPa. Mortar long-term strength can be measured on cylinder or/and cubes without dewatering in order to get conservative values. 3 RHEOLOGICAL INDEX (IR) PHILOSOPHY Most of existing specifications and qualification processes dealing with cementitious materials properties refer to hardened product characteristics. They mainly assess their ability to achieve long term strength and durability, after completion. Site problematic is fully different: concrete, mortar or grout must fit client specifications, but must also be designed to be easily placed. 3.1 Water/Cement ratio (W/C) This ratio (effective water content to cement content by mass in the fresh concrete) is commonly internationally used to specify concrete, especially with regards to their expected durability with a typical range corresponding to [ ]. It is less commonly used for mortars and grouts. The only way to adjust the W/C ratio to an expected value is to increase the cement content and/or decrease water content, thanks to the use of high water reducer admixtures (super plasticizer). In order to take standardized mineral additives into account in W/C, the European standard EN defines the notion of k-value concept by replacing the term water/cement ratio with water/(cement + k addition) ratio. The actual value of k depends on the specific addition. As a matter of fact, the current approach does not take into account all the introduced mineral additives nor fines particles eventually brought by aggregates. Moreover, a mass ratio approach is not able to provide users with accurate data regarding fresh concrete behaviour. This matter is illustrated by the example corresponding to two concrete mix designs with the same W/(C + ka) ratio given in 4. However, the EN introduces the equivalent concrete performance concept, which permits amendments to the requirements in this standard as far as minimum cement content and maximum water/cement ratio are concerned. 3.2 Rheological Index (IR) Context In order to easily assess a concrete, mortar or grout mix design with regards to both fresh and hardened expected properties, an original and promising tool called Rheogical Index, or Indice Rhéologique (IR) in French, has been developed, thanks to the outcome of a long experience gathered on many different civil works projects (tunnels, dams, bridges, and others...). This method can be easily applied onsite in addition to existing empiric tools currently used to assess concrete strength (as Bolomey or Féret methods) Definition The Rheological index corresponds to the volumetric ratio between all the fine particles (including sands) and water. Pi = dry mass of fine particles <0.080 mm for each individual constituent i γ i = specific gravity of each constituent i V = water (Surface Saturated Dry Aggregate) According to the expected precision, cement and mineral additives can be considered as <0.080 mm solid materials and admixtures are not taken into account for this calculation Main benefits The Rheological Index must be considered as complementary of the well-known Water/Cement ratio. Similar fine particles (in terms of density and shape) assumed active or inert will have the same behaviour regarding workability. The main benefit is the possibility to closely examine the potential accuracy of a product s properties for a dedicated application, thanks to indicative scales given in the following paragraphs (like a so-called screening analysis). Rheological Index values give global indications regarding: Rheology of the mix, which correspond fresh cementitious material behavior (i.e. plus or minus fluid, viscous, thixotropic, sticky, cohesive). Compactness of the mix, which is highly influenced by total volume of fine particles with regards to water content, and consequently durability and strength classes. The more IR is high, the more the associated product will be sticky, dense and compact IR indicative scale for typical concretes (D max = 20 mm) IR > Concrete >C100/115 C60/75 C50/60 C40/50 C30/37 C20/25 grade + <compactness/strength> <bleeding/segregation> + In some cases, an over-rate can be envisaged in order to increase workability (very fluid concrete like 274

5 Self-Compacting concrete) or compactness/strength ( dry concrete). Illustrations and limits of approach are given in paragraph IR indicative scale for TBM backfill mortars IR Target dewatering and pressure < Standard > Fluidity Better behavior (grouts) Illustrations of backfill mortar mix design successfully used on several international projects are given in paragraph IR indicative scale for grout IR Target Backfill grout thick grout Bentonite grout fluid grout Grout effective strength will depend both on IR value and cementitious material s selection. 4 ILLUSTRATION FOR CONCRETE The benefit of the Rheological Index is clearly illustrated by the following example corresponding to two concrete mix designs with the same W/(C + ka) ratio: The proposed concrete mix designs (with identical W/(C + ka) ratio defined in European standard EN 206-1) will be different regarding flow ability and hardened concrete compactness (and associated durability). Concrete A will correspond to a grade C30/37, and concrete B will be closer to a grade C45/55, according to IR scale. Nevertheless, mechanical properties proposed classification is obviously valid in the limits of hydraulic and/or pozzolanic properties of cementitious constituents considered. Minimal cement content (as stated in relevant Codes, Standards, and States of the Art) is obviously necessary to achieve the expected concrete strength as illustrated in example given in Table 2. The two concrete mix designs will have similar rheological behaviour and compactness, but the lack of cement of concrete B, not formulated according to usual rules (limitation of cement partial substitution by PFA) will assign significantly the achieved compressive strength. This comparative example clearly illustrates the IR scale main benefit for specific materials (like backfill or substitution materials) requiring workability properties but no conventional strength criteria. Table 1. IR and W/C calculation and comparison for two different concrete mix designs. Density Concrete A Concrete B CEM I kg/m 3 CEM II/A-V kg/m 3 PFA kg/m 3 Fines (sand) kg/m 3 Water 170 l/m l/m 3 W/(C + 0,5 A) Fines volume l/m l/m 3 Paste volume l/m l/m 3 IR Table 2. Illustration of the importance of cementitious material selection. Density Concrete A Concrete B CEM I kg/m 3 50 kg/m 3 PFA kg/m kg/m 3 Fines (sand) kg/m 3 25 kg/m 3 Water 190 l/m l/m 3 C + 0,5 A 340 kg/m kg/m 3 W/(C + 0,5 A) Fines volume l/m l/m 3 Paste volume l/m l/m 3 IR Concrete grade C35/45 C35/45 Expected fc28 (cyl) 45 MPa <10 MPa According to indicative IR scale. 5 EXAMPLE OF TBM S TAIL VOID MORTARS This paragraph summarizes some of the many backfill mortar mix designs (with IR = 0.60 ± 0.05) successfully used in international projects in the last 10 years. Other target IR values (>0.65) have also been considered for projects facing particular ground conditions and/or higher groundwater pressure. 6 CONCLUSION Tunnelling Boring Machine s tail void backfill mortars look apparently rustic products but are actually a key issue for a TBM s Project progress. They must fulfil several criteria, regarding both fresh and hardened mortar properties, more or less antagonists. The approach given in this paper is built up from the outcome of large know-how gathered on many different projects. The main characteristics of so-called semiinert backfill mortars, and associated laboratory trials, 275

6 Table 3. Mortars used for SOCATOP (A86 motorway France). Semi-inert Active CEM I kg/m 3 45 kg/m 3 PFA 130 kg/m kg/m 3 LF 55 kg/m 3 50 kg/m 3 Crushed sand 0/6 720 kg/m kg/m 3 Crushed sand 0/4 840 kg/m kg/m 3 Retarder 0 to 1.2 kg/m kg/m 3 Water 280 kg/m kg/m 3 IR % fine particles. 8% fine particles. Table 4. Mortar used for Pannerdensch (The Netherlands). Hydraulic lime 72 kg/m 3 PFA 251 kg/m 3 LF 152 kg/m 3 Natural sand 0/ kg/m 3 Natural sand 1.4/ kg/m 3 Natural sand 3.4/ kg/m 3 Natural sand 5.6/ kg/m 3 Water 295 kg/m 3 IR 0.65 Table 5. Mortar used for S.M.A.R.T project (Malaysia). CEM I kg/m 3 Hydrated lime 40 kg/m 3 LF 340 kg/m 3 Natural sand 0/4 693 kg/m 3 Crushed dust sand 0/8 693 kg/m 3 Retarder 2.5 kg/m 3 Water 310 kg/m 3 IR % fine particles. 8% fine particles. Table 6. Mortar used for ASDAM project (Belgium). Table 7. Mortar used for projects (Russia). Hydraulic lime 70 kg/m 3 PFA 320 kg/m 3 MS 30 kg/m 3 Natural sand 0/4 820 kg/m 3 Crushed dust sand 0/6 540 kg/m 3 Water 365 kg/m 3 IR % fine particles. have been developed and illustrated by several examples of successful applications in which VINCI has been involved. It must be noted that backfill mortar volumes are in general very important (several tenth of thousand cubic meters). In the course of this overview, an original and promising approach and/or tool: the Rheogical Index has been introduced. This index is a kind of generalized volumetric Water/Cement ratio, which easily provides fruitful and faithful indications regarding fresh cementitious materials properties. In other words, this original approach aims to put in equation the workability of cementitious materials such as concrete, mortars or grouts. For typical concrete, the IR scale is an interesting tool to cross-check and validate a given mix design. The main benefit concerns the design of non-conventional products with specific properties, such as high fluidity, not necessarily associated to high strength. Indeed conventional Water/Cement ratio appears not to be adapted for this kind of materials, and more generally to concretes mixes, which are more and more formulated with high amount of different mineral additives and/or admixtures. It has been demonstrated on several project that a better consideration of the effective influence of all the fine particles (including those brought by sands and mineral additives) for cementitious materials mix designs was an interesting approach regarding both technical and economical concerns, and was associated to a better natural resource valorisation. CEM III/B kg/m 3 PFA 250 kg/m 3 Natural dune sand 0/1 154 kg/m 3 Crushed dust sand 0/4 693 kg/m 3 Crushed dust sand 0/6 693 kg/m 3 Retarder 1 kg/m 3 Water 335 kg/m 3 IR % fine particles. 13% fine particles. 276

NEW PERFORMANCE CRITERIA FOR FRESH TREMIE CONCRETE

NEW PERFORMANCE CRITERIA FOR FRESH TREMIE CONCRETE Karsten Beckhaus (1), Martin Larisch (2), Habib Alehossein (3,4) (1) BAUER Spezialtiefbau GmbH, Deutschland (2) Keller Australia Pty Ltd, Australia (3)