Reliability of the grouting of tendons under consideration of uncertainties under on-site conditions

. This paper contributes to the body of knowledge from German research work Deutscher Ausschuss für Stahlbeton and from author’s experiences, to improve the reliability of cable ducts grouting. The complete and reliable filling of post-tensioned tendons with grout is necessary to protect the prestressing steel against corrosion and for bond between prestressing steel and the concrete structure, which is in both limit states applied. Diagnostics of bridges built in the last century showed the cases of faulty grouting of cable ducts. These faults are cause of corrosion prestressing steel often. Based on these diagnostics, bridges are after 35 to 70 years of service life demolished, although service life of these bridges was to 100 years declared. In the studied research work were investigated the effects of individual components, procedures and conditions of the grout and results of grouting investigated. Tests have been performed both on the experiment cable ducts and the real structures. Based on this paper, certain principles have been recommended that aim to prevent grouting faults and to improve the reliability of the grouting. In these recommendations are suggestions of controls of grouting included.


Introduction
Injection of cable ducts with grout is very important process when constructing construction from post-tensioned concrete.Any faults during grouting are unacceptable, as the loadbearing capacity, functionality and consequently the lifetime of the structure would not be ensured.This paper responds to increasing of faults during the grouting of wire prestressing cables, which were used in beams KA-61, KA-67, KA-73 and I-62, I-67, I-73, IS-73 from 1950 to 1993 in Czechoslovakia.These faults are cause of corrosion prestressing steel often.This corrosion is closely monitored and then these constructions are demolished after 35 to 70 years of service life, although service life of these constructions was to 100 years declared.Principles of grouting have not changed, and risk of faults persists.It is important to know these risks and do everything to avoid these faults.The purpose of this report is to draw attention to possible risks and consequences bad grouting cable ducts, repeat basic knowledge of grouting from design of cable duct diameter, mixing mortar to injection, and indicate possible directions for increasing the reliability of injection works, including possible control.

Risk of bad grouting
Grout ensures permanent anti-corrosion protection of prestressing steel and bond between the prestressing steel and concrete structure, which is in both limit states applied.Although the knowledge of bad grouting increased, diagnostics of existing construction repeatedly show faults of grouting.Badly grouted cable ducts are often a source of corrosion of the prestressing steel because the prestressing steel is not protected in any way against corrosive substances.These substances entering to prestressing steel mainly from the area of anchors, through bad expansion joints, see Fig. 1.Reactions of bridge managers on bad expansion joints are often delayed.Bridge managers or bridge owners can act after recommendations of bridge inspections according to standard ČSN 73 6221.These inspections do not provoke an appropriate reaction in stage of construction condition from 2 to 5. Constructions are often operated for several years to decades.In this time rainwater or water saturated with chlorides (from road salt) penetrates to anchor.Bad grouting can be caused by e.g., inappropriate properties of the grout, impermeability of the cable duct, inappropriate arrangement of drain and vent, insufficient knowledge of the workers, unfavourable conditions during the grouting, insufficient grouting technology, etc.To this day, there are no sensors for performing quality of grouting and status of grouting during lifetime of the structure.Checking whether the cable duct is filled with grout is crucial for a reliable statement about intactness prestressing steel.Such a statement is only possible using a semi-destructive method.The procedure for this implementation of the method is as follows: 1. Diagnostics of the position of the prestressing steel -localization of the prestressing steel in the longitudinal and transverse direction, from the project documentation and on site using non-destructive methods, see Fig. 3.

Fig. 2.
Diagnostics of the position of the prestressing steel.
2. Probe -drilling of the cable duct, first with a drill of smaller diameter (stabilizing hole) and then drilling with a larger diameter, e.g., with a drill with a diameter of 40 mm, see Fig. 3.
3. Inspection of the grout -visual inspection of the condition of the grouting of the cable duct and the condition of the grout, see Fig. 4. 4. Extending the probe -drilling the probe through the grout up to the prestressing steel, possible mechanical loosening of the grouting, see Fig. 5.The possible results of the semi-destructive method are shown in the Table 1.The difference between correctly and incorrectly injected cable channels is shown in Fig. 8.Other examples of poor grouting are shown in Fig. 9 and Fig. 10.Table 1 Possible results of the semi-destructive method.
In the Fig. 8 (left) the grouting was performed incorrectly.Individual wires are not covered with grout and corrosion is obvious here.Also using LFT (lever force test), any tension stress was not found in the individual wires.The wires were loose in the cable duct and could be moved freely.The parallelism of individual wires is not evident here.Fig. 8 (right) is an example of a properly grouted cable duct.The individual wires are completely covered with grout and the corrosion of the prestressing steel is not visible here.
In Fig. 9 (left) a partially grouted cable duct can be seen.The individual ropes are partially covered with grout.Partial corrosion due to poor grouting can be seen here.Fig. 9 (right) is almost a deterrent case.The cable duct is without grout and the individual ropes lie in the water.One of the causes is poorly made hydro isolation, and therefore water can flow up to the prestressing steel.
Lumpy grout can be seen in Fig. 10 (left), this can be caused by e.g., a bad mixing technique or a wrongly chosen plasticizer.In Fig. 10 (right) completely wet grout can be seen.In this case, it leaks somewhere into the structure and the amount of water in the cable duct is as large as in Fig. 9 (right) but here the cable duct is filled with grout.

Suggested method of repair
Dry    Current knowledge about the grouting is based on extensive experiments of grout, grouting of cable ducts and various experiences.Grout represents a significant factor of uncertainty in the grouting of cable duct.One of the reasons is the fact that the grout must be often prepared under unfavourable on-site conditions.The properties of the grout strongly depend on the chemical-physical properties of cement and the environmental conditions during the grouting that affect them.This paper contributes knowledge, a selection of knowledge, experiments, and experiences both from German literature [1] and from the authors' own experience and knowledge and recommendations of Czech standards [2], [3] and [4] to improve the future reliability of cable duct grouting.

Description of the current state of problematic
Characteristics of grout depend on the used materials (cement, water coefficient w/c and used admixtures or additions), mixing techniques and injections techniques and mainly conditions during the grouting (temperature, processing time), see Fig. 11.As the length of the cable and the amount of prestressing steel increases, so does the volume of grout and possible of any failures.For example, water can be drained away between individual strands (bleeding), in this case cable duct can be blocked.Furthermore, if there is a too many prestressing strands in the cable duct, this makes it difficult for the grout to flow transversely between the individual strands, so that the individual strands do not have to be coated on all sides with grout.Therefore, requirements must be placed on the grout for its fluidity, prevention of segregation and limitation of bleeding [4].Grouting is complex process into which a lot of unknowns enter.Fig. 11.Elements of grouting of cable duct and their interrelationship [1].
Grout mainly consists of cement and water.According to the standard [4], the water coefficient (w/c) must be maximum of 0.40.Water coefficient is decisive for the properties of fresh and hardened grout.It must be confined at the top to minimize bleeding.This also ensures certain strength requirements.If the w/c ratio is too high, water can accumulate in the top of cable duct and then result in spacing in the grout.However, the w/c ratio cannot be reduced arbitrarily, as fluidity is limited.Additional additives or admixtures can be added to grout to improve the properties of grout, especially workability.These are mainly plasticisers or micro-silika.
A mixer and injection pump are required for the grouting itself.The mixing process is decisive for the properties of the grout (homogeneity, fluidity).The intensity of mixing is a basic prerequisite to produce homogeneous mixture.The correct grouting procedure must be selected (injection pressure, injection rate, injection procedure in the structure), the location of vents and drains, and the workers must be properly trained.
The quality of the grouting itself can also be influenced by the structural engineer when designing the structure.The ratio of the areas of prestressing steel and the cable duct is very important.If this ratio were too large, there would be no space for the grouting itself and thus proper grouting of the cable duct is not guaranteed.The probability of blockage the cable duct is much bigger.In addition, there can be a bleeding, where water is drained away through the individual strands.On the other side, with a decreasing ratio, the grouted cross-section increases, and this promotes sedimentation, which can create cavities from rising air bubbles and separated water.The optimal ratio resulted from many years of experience in the range of 0.20 to 0.45 [5].Another important aspect in the design is the location of the cable duct in the structure.If the cable duct were too close to the surface and the volume changes were large, the structure could be damaged.
A very important factor is the conditions during the grouting.We must monitor the temperatures in the structure, the grout and in the air during the grouting.According to the standard [3], the temperature must be between +5 and +35 °C, during the grouting and in the following 48 hours.The reasons will be explained later.
Another monitored variable is the injection pressure, which must not exceed 20 bars, primarily to ensure the safety of the workers.It is important to monitor the injection speed.The maximum speed is usually given by the injection device, but speeds above 30m/min lead to segregation of the grout.Speeds of 5 to 15 m/min are recommended.

Literature review
As already mentioned, some experiments, procedures and results presented in this chapter are taken from German literature [1] and the aim is to pass this information on.

Range of experiments
A total of 14 types of CEM I cement from different manufacturers were used for the experiments according to the literature.This type of cement is preferred in practice for grouting.All cements were subjected to chemical analysis.First were produce and tested grout with a standard formula (w/c = 0.38, T = 20 °C).Then had some selected grouts their formula modified, for example by adding water, micro-silica, or plasticiser.In another series of experiments, the properties of the grout were tested at different temperatures or using different mixing equipment.Grout with ultrafine cements were tested also.
The grout tests were carried out according to the then valid standards.Fluidity, volume changes, compressive strength and bleeding were mainly determined.Fluidity was verified using a dip test as well as a funnel flow test.Volume changes and bleeding were determined on a measuring cylinder with a volume of 1000 ml and a height of 950 mm.Compressive strengths were tested on typical beams.One of the results of the tests was the suitability of the test methods under on-site conditions, this topic is not presented in any way.
Injection experiments were also performed on a model cable without prestressing steel and with prestressing steel.The model was formed in a top arch with a plan length of 15.0 m and a top height of 1.40 m.Vents were located at the top and 1.5 m to the left and right of the top.Various types of material for cable ducts (steel, PVC, PE) were also tested.The shape of the grout during grouting was studied here.

Results of experiments
First, the connection between ingested cement and the properties of fresh grout was sought.The cements used were therefore investigated about their chemical and physical properties.From the results, it could be said that the fluidity decreased both with increasing C2S (belite), see Fig. 12 (left), and alkali content.Also, with increasing heat of hydration, the fluidity of the grout decreased.When choosing cement, it should be considered that the mentioned properties interact with each other, and the individual effects could add up or interfere with each other.Cements with good fluidity are suitable for grouting, and bleeding can be minimized by a suitable proportion of fine particles.Slow-reacting Portland cements should be preferred.Therefore, the content of C3A (characterized by very rapid hydration immediately after mixing) in the cement should be reduced, or the reactivity of the cement should be reduced.

Fig. 12.
Liquidity depending on the proportion of C3A (left) and Liquidity depending on the proportion of C2S (right) [1].Furthermore, tests were carried out on grouts with the addition of additives and admixtures.The results showed that the plasticiser can caused a loss of compressive strength of the grout (the strength values still met the minimum permissible values), but the processing time of the grout was significantly increased, see Fig. 13 (left), as the consistency of the grout was more fluid.Although the figure shows fluidity as immersion time and this test is not used here, a shorter immersion time indicates a more fluid mixture.A more liquid mixture, due to the addition of a plasticiser, enabled a reduction in the water coefficient and thus a better increase in the strength of the grout.With the addition of micro-silica to the mixture, the fluidity of the mixture increased, and the bleeding decreased, see Fig. 13 (right).

Fig. 13.
Comparison of flowability of grout with and without plasticiser (left) and bleeding depending on the volume of micro-silica (right) [1].
Ultrafine cement was also considered, but the grout was usually too liquid, and the compressive strength was below the minimum permissible value.Positive features were small volume changes and almost zero bleeding.Ultrafine cement is therefore unsuitable for grouting cable ducts but has found application in remediation or re-injection of cable ducts.
From other test results, it was evident that the volume changes did not depend on the cement but were to some extent determined by the swelling effect of the additive.In addition, the mixing technique and the ambient temperature of the grout influenced the volume change in the first hours of setting.As the temperature increased, so did the volume change.Volume changes also increased with high mixing energy.Furthermore, it follows from the conducted tests that bleeding depended on the cement, the mixing principle, and the temperature of the surrounding environment during the setting of the grout.Bleeding was also investigated under elevated pressure, but the results were not significantly different from conventional tests.The test results show that the volume change of the grout was completed about 3 hours after mixing, see Fig. 14.And as can be seen from this figure, bleeding took place at approximately the same time.Bleeding was completed between 3 and 4 hours after mixing, with only a small amount of water being separated from the third to the fourth hour.After 24 hours, the water was reabsorbed by the grout, which is the cause of the voids in the grout.Fig. 14.Volume changes and bleeding of grout [1].
Investigation of the effect of low and high temperatures was the subject of further tests, see Fig. 15.The results showed that the fluidity of the mortar decreased with increasing temperature.Furthermore, it was found that volume changes increased with increasing temperature for all investigated grouts.On the other hand, bleeding increased with decreasing ∆ temperature, with solid grouts bleeding was significantly lower than with liquid grouts.A possible reason for bigger bleeding at low temperatures is that the reaction between cement and water is slower.Thanks to this, water does not bind to the cement grains as quickly and excess water can settle.Fig. 15.Volume changes depending on the temperature (left) and Fluidity depending on temperature (right) [1].Furthermore, the shape of the grout during the grouting was studied on the model cable, see Fig. 16.The grouting took place from the bottom upwards.Grout moved smoothly upwards.In the case of grouts with low viscosity, the level was almost horizontal, if the viscosity was higher, the face of the grout was slightly convex.Once the top was passed, the thin grout flowed downwards and the section behind the top was gradually filled upwards and the air escaped through the vent in the top and behind the top.This can cause air voids in the grout.As the stiffness of the grout increased, the level of the grout in the descending part was no longer horizontal.The face of the grout kept its shape, and the protector was filled with grout from top to bottom.In this case, the vent at and near the top was filled with grout and air could only escape at the end of the cable duct.This ensured the correct grouting of the cable duct with grout without the formation of air cavities.The advantage was also the smaller area on the face of the grout, where residual water could mix with the grout.The grout thus remained more homogeneous.Fig. 16.Flow conditions of grout in the rising and falling cable duct depending on the consistency of the grout [1].
Another thing that was investigated in this test was bleeding and volume changes, see Fig. 17.It was noticeable that the separation of water was usually about 4 hours after grouting.Water collected in the ridge of the cable duct, in the peaks and specific points prone to the formation of cavities (expansions, areas behind the anchor, etc.) from where it could not escape anywhere.After 24 hours, the accumulated water completely disappeared or was absorbed by the grout and thus a cavity was created in the cable duct.Additional vents should therefore be placed at these points.Volume changes must also be paid attention to, because when combined with an inappropriate location of the cable duct in the structure (too close to the surface), tearing of the structure may occur.

Conclusion
The reliable grouting of prestressing steel in cable ducts with grout depends on many factors, some of which interact with each other.The presented results show that the properties of the mortar strongly depend on the cement and its properties.Based on experiments, it was possible to determine to a certain extent the connection between the physic-chemical properties of cement and the properties of fresh grout.The properties of the grout depended not only on the starting materials, but also on the mixing process and the temperature.
Grouting tests have shown that in grout with a strong tendency to sedimentation, voids are formed due to the bleeding, therefore it is important to pay attention to the separation of water in grout.It was also found that the bleeding takes place approximately 3 hours after mixing the mixture and the volume change is almost at its maximum at this time.During the next 24 hours, water is absorbed by the grout, creating voids.
It is not so easy to say what the optimal water factor is.The final properties pretty much depend on a lot of things.According to [4], the value of the water coefficient must be less than 0.4 and it must also meet the requirements according to [2].From the authors' own experience, the given requirements are met when using a water coefficient of 0.32 and the corresponding amount of plasticizer.
Most of the results were verified and carried out according to the then valid standards.It is certainly necessary to carry out similar tests again and adapt the conclusions to today's requirements and trends.Special attention should also be paid to the still undeveloped diagnostic methods of grouting reliability.

Fig. 1 .
Fig. 1.Demonstration of corrosion of the prestressing steel in the anchors, due to a bad bridge joint.

Fig. 5 .
Fig. 5. Extending the probe.5. Inspection of prestressing steel -visual inspection of the state of the prestressing steel and mechanical inspection of the state of tension (LFT = lever force test), see Fig. 6.

Fig. 17 .
Fig. 17.The formation of voids due to the water bleeding during solidification [1].