Aus Darmstadt Concrete Vol. 13 1998

CONSTRUCTION OF AN OFFICE BUILDING USING CONCRETE MADE FROM RECYCLED DEMOLITION MATERIAL

Peter Grübl, Andrew Nealen

SUMMARY
The use of recycled aggregate derived from concrete rubble in the production of ready mixed concrete is demonstrated in the example of an office building, erected in Darmstadt, Germany. The implemented quality management System secures a constant high quality regarding the composition and dry density of the recycled concrete aggregate, thus leading to a predictable water absorption. Due to varying core moisture and the inability to measure this moisture during dosage, the mix proportions during the three month production time were not constant, resulting in a Variation of the actual water/cement ratio and the initial consistency value of the concrete mix. A consistency correction is performed on site using superplasticiser. Finally, the statistical analysis of achievable compressive strength shows, that a high concrete quality can be maintained, provided properly processed and controlled recycled aggregate is in use.

1 INTRODUCTION
The Bauverein AG in Darmstadt, Germany, is the owner of the building project Vilbeler Weg [1], in which an office building with parking areas was erected (Figure 1). A complete section of this building is constructed from concrete made with recycled aggregate derived from concrete rubble.

Concrete technology, logistics and quality management were covered by the Institut für Massivbau of the Darmstadt University of Technology. The aim of this project was to demonstrate the general applicability of the durrent research results and to define the basies of a practical quality management system. It was decided, that in the production of concrete for indoor and outdoor components only recycled aggregates were to be used.

Fig. 1: Ground plan (A) and cross section (B) of the building project ,Vilbeler Weg‘

2 PREPARATION AND PRODUCTION PROCEDURE

2.1 Mixing Plant, Logistics and Quality Management
The concrete mixing plant was built between October 1996 and June 1997 in Darmstadt, Germany. It measures the moisture of the aggregate <2 mm and automatically corrects the water and aggregate dosage online. A moisture measurement for the recycled aggregate ³ 2 mm is not implemented because all current indirect moisture measurement methods evaluate the complete water content and do not differ between surface and core moisture as it would be necessary in the case of a water absorbing aggregate. Furthermore, a recycling plant for concrete residue was implemented, which allows the reuse of not used concrete in the fresh state, as well as an aggregate heating system.

For the use of concrete with recycled aggregate it is currently necessary to gain Individual Approval by the local building supervisory authority. For this reason, in the second quarter of 1997, numerous tests were conducted using concrete designed towards a B 35 according to German Code DIN 1045 to analyse the properties of the freshly mixed and hardened concrete (B 35 is equivalent to a C30/37 according to EN 206). These tests formed the basis of the application for Individual Approval, which was turned in in July 1997 and approved in September 1997. The approval is currently linked to concrete class B 35 with a maximum grain size of 16mm and 32mm respectively.

The implemented quality management system‘s main objects were to control the properties and the composition of the recycled aggregate in use as well as the quality of concrete made with this aggregate. The plant responsible for the preparation process in Büttelborn, near Darmstadt, was ordered to prepare and to deliver the recycled aggregate made from concrete rubble in fractions of 2/8mm, 8/16mm and 16/32mm, as well as an officially permitted recycled sand of the fraction 0/2a according to German Code DIN 4226. The acceptance inspection on delivery to the concrete mixing plant consisted of determining the grading curve, the composition, the dry density and the core moisture of all grain fractions. To evaluate the properties of concrete such as consistency, compressive strength, density and modulus of elasticity, samples were taken at the mixing plant and on site six times more than required by regulations (one sample per 25 m3 of concrete and on each production day).

The first construction members made from concrete with recycled aggregate (ground floor ceiling) were casted on the 11^th of November, 1997. Concrete construction ended mid February 1998. The amount of concrete with recycled aggregate built in adds up to 480 m³.

The concrete had a maximum grain size of 16mm.

2.2 Mix Proportions
The mix proportions of concrete formulation (M21) used during construction are shown in Table 1. The latest draft of the German guideline Concrete with Recycled Aggregate [2] allows to exchange up to 33% by volume of the total natural aggregate with recycled aggregate derived from concrete rubble for indoor concrete components.

In the project Vilbeler Weg, the complete aggregate consists of recycled mineral material. All grain fractions > 2 mm were derived from concrete rubble. Figure 2 displays the development of compressive strength in laboratory tests. The formulation with a maximum grain size of 32mm was not in use due to a high percentage of reinforcement of the elements.

Table 1: Concrete mix proportions according to the initial test for the Individual Approval

During the laboratory tests, the aggregate of the fractions 2/8 and 8/16 were used in water saturated state (core moisture saturation, no surface moisture), to simulate extreme weather conditions at the mixing facility.

After storing the dry aggregate under water for 24 hours, the surface moisture was removed. Then, the remaining core moisture was measured by weight and amounted to approximately 4,7 % by mass on dry basis. This is defined as the water absorption after 24 hours (w_24h). The water absorption of the aggregate <2 mm (the sand) was very small (<0,5 % by mass) and Therefore neglected.

Measurements of the water absorption beyond 24 hours showed no significant change in value, therefore the 24 hour water absorption value can be defined as the aggregates maximum water absorption (wa24h = wa,max). This behaviour is typical for aggregate derived from concrete rubble.

The aggregate grading curve was adjusted in the area of curves A16 and B16 in accordance with German Code DIN 4226.

Fig. 2: Development of compressive strength of the concrete during initial test

2.3 Water Dosage
The water amount dosed at the mixing plant (w_d) was always constant at 170 kg/m³ (taking sand moisture into account). The aggregate‘s present core moisture (w_a,p) at the mixing plant was variable due to climatic change. In the case of core moisture saturation on wet days (w_a,p = w_a,max), the actual water/cement ratio of the concrete mix was approximately w_act/c = 170/310 = 0,55. This value depicts the worst case with the highest actual water/cement ratio. In all other weather situations (w_a,p < w_a,max), the water absorption of the aggregate reduced the free water (wact) in the cement mortar therefore reducing the actual water/cement ratio. The altering core moisture (w_a,p) led to a variation of free water, consistency value, development of rigidity (faster development using dry aggregate) and compressive strength. The volume and mix proportions of the concrete mixture were also effected by the absorptive behaviour of the aggregate.

2.4 Consistency Variation and Consistency Correction
After mixing (at 10 minutes), the concrete had a flow table value between 380 mm and 500 mm, according to DIN 1048 and EN 206, depending on the present amount of core moisture Then it was transported to the building site, arriving there about 20-30 minutes after filling the trucks (t = 30-40 min.). It was then necessary to perform a consistency correction on site, since a flow table test value of 550 mm and above has tube guaranteed fur an excellent workability. The improvement of workability was performed by dosing amounts of superplasticiser. The amount depended on the consistency value of the concrete when it arrived on site. On arrival at the construction site, the flow table value varied between 340 mm and 490 mm due to core moisture content. Therefore, the amount of superplasticiser necessary to achieve the flow table value required varied between 5 ml/kg cement and 18 ml/kg cement (see Figure 3).

For wall castings the concrete was transported using a crane and tub on site, for ceilings it was pumped in up to 43 m height using a concrete pump.

Fig. 3: Superplasticiser dusage and change of flow table test value Da, cm

3 PERFORMED TESTS AND RESULTS

3.1 Cube Sample Frequency
Every production day and every 25 m³ of concrete produced, a series of three sample cubes (150 mm edge length) were taken both at the mixing plant and on site for testing compressive strength. At the mixing plant, the samples were taken directly after tilling the concrete mixing trucks (approximately t = 10 minutes after mixing). On site, the samples were taken after dosing the superplasticiser and mixing the concrete for another 10 minutes in the concrete mixing truck (approximately t = 60 minutes after initial mixing).

3.2 Composition of the Recycled Aggregate
The mean values of the aggregate composition used for concrete production are shown in Table 2.

Table 2: Aggregate compositiun

1) Concrete matrix, natural aggregate bound together with mortar residue
2) Natural aggregate from die processed concrete rubble without mortar residue

A sample of every aggregate delivery to the mixing plant was checked according to Table 2. The results of the separation process are then compared with the values listed within the ,Individual Approval‘. The aggregate composition was of a constant high quality and never differed significantly from the values of the initial tests.

3.3 Dry Density
The dry density of the recycled aggregate remained very constant during the period of construction, so the maximum amount of core moisture was predictable, as the amount is directly linked to the dry density of the aggregate.

Table 3: Dry density

The determined dry density of the recycled fractions and its small variation range is shown in Table 3.

3.4 Compressive Strength
Within the three months of concrete production, a sum of 152 sample cubes were tested for achievable compressive strength and evaluated statistically. Since three cubes were taken from each concrete sample, both at the mixing plant and on the construction site, the sample size fur each distribution adds up to n = 26 independent samples [3]. Figure 4 shows die Gaussian distributions of test results both at the mixing plant (t = 10 minutes) and on the construction site after adding superplasticiser (t = 60 minutes).

The variation of measured compressive strength within a series of three cubes (max. 2,1 N/mm²) is small compared to the variation of mean values of all series. As the recycled aggregate‘s composition and dry density were nearly constant, the influence of material quality on the variation of compressive strength is also not of larger significance.

Fig. 4: Gaussian distribution of the achieved compressive strength

Table 4: Results of statistical analysis of the compressive strength of concrete sort M21

The greater part of the distribution is related to the varying core moisture (w_a,p) of the aggregate fractions above 2 mm. The variation of the amount of free water (w_act) in the cement mortar influences the consistency value, which, measured at a fixed time, is a criterion for the achievable compressive strength of the concrete mixture.

Figure 5: Relation between flow table value (t 30 min.) and compressive strength of the samples taken on site

This is displayed in Figure 5, which shows the direct correlation between the distribution of compressive strength and flow table value at t = 30 minutes. Although the core moisture was not considered in the dosage, the Gaussian distribution shows a standard deviation of only 3,01-4,23 N/mm² (Table 4) which is considered as very good quality [4].

In analogy to the relation between compressive strength and flow table value, the amount of superplasticiser necessary for consistency correction correlates with the compressive strength: The higher the compressive strength, the lower the flow table value and therefore the more superplasticiser necessary to achieve a flow table value of 550 mm, as seen in Figure 6.

Figure 6: Compressive strength and superplasticiser dosage for all 26 concrete samples

4 CONCLUSIONS

The demonstration project Vilbeler Weg shows, that the application of the presented production concept is possible, if it is combined with an appropriate quality management system. To ensure a concrete quality equal to that of concrete made from natural aggregate, certain parameters and boundaries have to be considered:

1. As the water absorption of the aggregate in use was low, due to it‘s high density, the workability could always be restored by adding superplasticiser. This would not be possible if the aggregate in use had a lower dry density and therefore higher water absorption. The span in flow table value between the aggregate‘s dry and water saturated state could not be compensated fur by superplasticiser alone.
2. The variation of the actual water/cement ratio in the concrete mix led to a deviation of compressive strength results. As in the case of workability, an aggregate with a greater water absorption would make it necessary to consider the core moisture in the dosage of aggregate and water to guarantee a constant actual water/cement ratio and ensure, that the standard deviation of compressive strength remains within reasonable parameters.
3. The volume and mix proportions of the produced concrete also varied, due to different amounts of absorbed water. This fact persisted throughout the entire production phase and must be considered in the dosage of concrete mix proportions, so the quantity of concrete ordered at the construction site is guaranteed.

The difference in free water (w_act) per m³ of concrete between completely dry and water saturated aggregate was:

So, to guarantee a successful application of the presented production concept, the value of Dwact was limited to w_act,max = 53 kg, in this case, leading to a minimum free water content of w_act = 170 - 53 = 128 kg/m3. If this boundary value should for any reason be exceeded, then the water dosage (w_d) must be adjusted towards a constant flow table value at a fixed time.(i.e. 10 minutes). Further research towards alternative concepts, applicable fur concrete mixtures exceeding Dw_act,max are in progress.

References

[1] GRÜBL, P. Baustoffkreislauf im Massivbau. Bauingenieur 72(1997), 5. 425/430

[2] DEUTSCHER AUSSCHUß FÜR STAHLBETON, DAfStb: Richtlinie ,,Beton mit rezykliertem Zuschlag". Entwurf Stand Juli 1998

[3] BIELAK, E. Bewertung von Ergebnissen aus der Zement- und Betonprüfung. Beton 44(1994) H.6, 5. 3 18/323

[4] WEIGLER H. AND KARL 5. Beton: Arten — Herstellung — Eigenschaften. Verlag Ernst & Sohn, 1989

[5] GRÜBL. P. Die Erstellung von Bauwerken unter Verwendung von industriell gefertigtem Beton mit rezykliertem Zuschlag. 18. Darmstädter Massivbau Seminar, Vol. 18, 1997