GB2231042A - Pavement material - Google Patents

Description

1 - C 456/S Title: Improved Pavement Material

Field of invention

This invention concerns an improved material for constructing pavements for the take-off and landing of aircraft, particularly vertical take-off aircraft.

Background to the invention

Vertical take-off and landing pavements constructed from concrete have been subjected to extensive tests to determine the stresses experienced by such structures during take-off These tests showed that during the takeoff of a vertical take-off aircraft such as the Harrier jet, the landing pad concrete attained surface temperatures of up to 300 3500 C Take-off times for such aircraft can vary from between 5 seconds for an experienced pilot up to 7 seconds for a pilot in training.

Observations of the damage to the surface of the pad indicated that surface deterioration noted in such pads is due to a heat spalling problem rather than weather induced damage.

From a study of the effects of rapid heating of concrete to high temperatures it is apparent that there are two mechanisms linked to the spalling of the concrete.

In the first place moisture within the concrete will vapourise at such a rate that the resulting pressures created within the concrete can cause the material to 2 - fracture.

In the second place differential thermal expansion of the material matrix can cause internal stresses which will cause the concrete to burst apart.

In the first case, water is retained in set concrete in three ways, ( 1) as water chemically combined in the cement hydrates ( 2) as water adsorbed in the gel of cement compounds (ie in the gel pores), and ( 3) as water in the capillary pores (usually referred to as free water) of the cement matrix The effect of increasing temperature on the different water contents varies, but in general it can assumed that once 100 'C is attained most of the free water will turn to steam whilst the chemically bonded water and the adsorbed water will not vapourise until higher temperatures are attained, when the compounds in which the water is bound, disassociate and dehydrate.

Differential thermal expansion can result either from different coefficients of thermal expansion of the aggregate and cement paste or from the temperature gradient through the concrete when the surface is caused to rise in temperature rapidly In this connection a study of the temperature gradient in concrete subject to rapid surface heating, reveals that the gradient is very steep and only a very thin (a few millimetres at the most) skin of the material actually attains very high temperatures.

Ordinary portland cement (OPC) consists of four principal 3 - compounds, dicalcium silicate (C 2 s), tricalcium silicate (C 3 s), tricalcium aluminate (C 3 A) and tetracalcium aluminoferrite (C 4 AF), which hydrate to form calcium silicate hydrates plus calcium hydroxide, and calcium aluminate hydrate (C 3 AH 6), together with various other hydrates.

On heating, the OPC will first lose its capillary water at around 1000 C and then its absorbed and combined water in certain higher temperature ranges before eventually reaching a water-free state, beyond which new ceramic phases form (at temperatures above 1000 C) A study of differential thermoanalysis curves of hydrated cement minerals reveals that ettringite (if present) dehydrates at around 150 to 200 C Subsequently hydrogarnet (C 3 AH 6), the hydrate of tricalcium aluminate and tetracalcium aluminoferrite, breaks down at around 3500 C At around 5000 C calcium hydroxide breaks down to form free lime and the calcium silicate hydrates lose their water at temperatures above 800 C It is important to note that the reactions giving off water occur intermittently and this can be significant in causing cracking The formation of lime from the dehydration of calcium hydroxide is a weakness of OPC paste, because subsequent wetting can cause severe disruption due to the large volumetric expansion ( 44 %) as the lime rehydrates to calcium hydroxide Thus, though an OPC concrete can withstand high temperatures on first heating, it is unsuitable for applications involving a regime of repeated heating, cooling and exposure to moisture.

The major component of High Alumina Cement (HAC) is monocalcium aluminate (CA), together with small amounts of C 12 A 7 and other compounds dependent on the oxide 4 - composition In simplified terms CA hydrates to form CA H 110, C 2 AH 8 and alumina gel (AH 3) No calcium hydroxide exists It is now well known that these hexagonal hydrates are metastable and convert, particularly at higher temperatures and in the presence of moisture, to the cubic hydrate hydrogarnet (CAH 6) and alumina gel.

Molecular volumes of the conversion products are less than those of the original hydrates, so conversion results in increased porosity and strength loss Consequently, the conversion process has been a cause of concern when HAC concrete is used in structural applications.

On heating, HAC paste will also lose its capillary water at around 1000 C Between room temperature and 300 'C adsorbed water is lost and the calcium aluminate hydrates and alumina gel break down; this process is usually complete by around 400 500 'C Unlike OPC, the hydrates break down in a slower and more controlled manner causing less internal disruption There is also no calcium hydroxide available to form potentially destructive free lime At very high temperatures (typically in excess of 900 'C) a refractory HAC concrete will start to increase in strength again as ceramic phases form within the matrix.

Although an HAC concrete is more spall resistant when heated, because of the way it loses its water and the absence of free lime, it can suffer from the disadvantage in ambient structural applications of conversion of the hydrates with a consequent reduction in strength.

It is an object of the present invention to provide improved concrete mixes for take-off pads for vertical take-off aircraft.

- Summary of the invention

According to one aspect of the invention a proportion of the cement in an otherwise conventional pavement quality concrete mix is replaced by ground granular blast furnace slag (GGBFS) to increase the resistance of the concrete to high temperatures.

Up to 50 % of the concrete mix may be replaced with the preferred material, but a preferred replacement percentage is 40 %.

It is to be noted that GGBFS is normally used in concrete mixes to reduce the heat of hydration in large pours and/or increase the resistance of the material to sulphates in the ground It is surprising to discover that selective substitution of the cement by GGBFS increases the temperature resistance of the concrete.

According to a second aspect of the invention the resistance of a pavement quality concrete (PQC) mix is improved by replacing at least a proportion of the naturally occuring aggregates in the mix with pulverized fuel ash (PFA).

The PFA may be in the form of pellets to replace coarse naturally occurring aggregate normally employed in the POC mix A preferred material is that marketed under the name LYTAG Lby Booral Lytag Limited.

The pellets may be of 10 mm nominal size.

The pellets may be ground into a lightweight fine 6 - aggregate and used as a replacement for sand in the PQC mix.

The preferred PFA material is the waste product from the generation of electricity at coal burning power stations, pelletised and sintered at 1200 'C to produce hard spherical nodules with a 40 % void ratio.

The material has hitherto been employed as a light weight aggregate for high strength, low density cast concrete structures and can produce structural concrete with strengths in the range up to 60 N/mm 2 However it has never previously been realised that by incorporating this material in a concrete mix, the resulting material can be rendered greatly resistant to spalling when the surface is subjected to high temperatures during take-off of a vertical take-off aircraft.

According to a third aspect of the invention the high temperature resistance of a concrete mix is improved by incorporating lightweight coarse aggregate, as described in connection with the second aspect of the invention, in combination with natural sand and cement.

The invention will now be described with reference to the data relating to a series of tests undertaken to determine the spalling resistance of different concrete mixes.

As a first stage of the programme of investigation it was necessary to examine in some detail the possible mechanisms associated with spalling in order to identify alternative mix designs with potential for improved spalling resistance It should be pointed out that there appears to be very little published information associated 7 - with this type of problem because ordinary Portland cement (OPC) concretes are not normally used in high temperature/spalling situations Related to the problem, however, is the behaviour of cementitious materials in refractory applications (where OPC concrete is not used) and also the behaviour of structural concrete following fire damage.

There are two possible causes of spalling of concrete caused by rapid heating Firstly moisture vapour may form within the matrix at such a rate that resulting pressures fracture the concrete In a set cement, water may be held in three ways:

( 1) as water chemically combined in the cement hydrates, ( 2) as water adsorbed in the gel of cement compounds (ie in the gel pores), and ( 3) as water in the capillary pores (commonly referred to as free water) of the cement matrix Whilst the effect of increasing temperature on these different waters is complex, we can say that generally speaking the capillary water will turn to steam at around 1000 C and the combined waters (chemically bonded and adsorbed) will form moisture vapour at higher temperatures, dependent upon the temperature at which the various compounds dehydrate.

The second cause involves the differential thermal expansion of the matrix, which consequently bursts apart the concrete This differential movement can be the result of either different coefficients of thermal expansion of the aggregate and cement paste or the temperature gradient down through the concrete as the 8 - surface is rapidly heated; previous work has already shown that only the top few millimetres experience the very high temperatures.

A concrete's resistance to these two mechanisms will clearly depend on many factors Related to water vapour induced failure will be the water content of the mix, the chemical compounds of hydration and the temperature at which they dehydrate, the permeability of the matrix, the moisture content of the concrete, and the concrete's strength Related to the second mechanism will be the coefficients of thermal expansion of the materials, their thermal conductivity, and again the strength of the concrete.

In an attempt to increase the spalling resistance of the existing OPC pavement quality concrete mix, a number of options were examined:- (i) reducing the free water in the hardened concrete by lowering the water/cement ratio of the mix (using a superplasticizer); (ii) increasing the permeability of the concrete (using a coarser grained cement) in an attempt to allow the water vapour formed on heating to escape more easily; (iii) altering the hydrate compounds of the mix (utilising cement replacement materials) and hence its behaviour on heating; and (iv) reducing the effect of differential thermal expansion within the concrete (using man-made aggregates, both coarse and/or fine).

9 - In addition, the use of a high alumina cement was investigated as a replacement for the portland cement in the standard PQ concrete mix A plasticizer/retarder admixture was used in this mix to extend its working time since early loss of workability is known to be a potential problem in the field, preventing proper finishing of an

HAC concrete.

Cements Three cements were used in this investigation These were:

(i) ordinary portland cement complying with BS 12; (ii) coarse ordinary portland cement supplied by Castle Cement This cement had a specific surface of around 280 m 2/kg compared with around 380 m 2/kg for a normal portland cement, and (iii) ordinary high alumina cement (Fondu) supplied by Lafarge.

Aggregates Standard PQ concrete aggregates were used and consisted of washed river sand fine aggregate and a quartzitic limestone coarse aggregate, supplied in two maximum aggregate sizes ( 10 mm and 20 mm).

Two man-made aggregates were also used Lytag aggregates were utilised in two forms: a 10 mm granular coarse aggregate and a graded fine aggregate (approximate 4 mm - down) Lytag consists of pelletized and sintered pulverized fuel ash and is marketed as a structural lightweight aggregate.

An expanded fireclay grog (EFG) material, supplied by Alfa Aggregate Ltd, in three sizes ( 5-10 mm, 0-5 mm, and 0-2 5 mm) was also checked EFG is a medium duty lightweight chamotte manufactured by sintering and crushing fireclay materials, used in the refractory industry.

Cement replacement materials Two cement replacement materials were examined: ground granulated blast furnace slag (GGBFS) to BS 6699 and pulverized fuel ash (PFA) to BS 3892.

Admixtures An air entraining agent was used namely Fro-be supplied by Sika Ltd, and this was used in all mixes for the production of test panels with the exception of HAC limestone mix.

A superplasticizer, Sikament, supplied by Sika Ltd was used in one mix to produce a PQ mix with a low water content.

A retarding plasticizer, Plastiment VZ, supplied by Sika Ltd was used in the high alumina cement limestone concrete mixes.

In addition to the test panels manufactured as will be described later, using the aforementioned ingredients, three further panels were obtained from Proderite Ltd 11 - using their Hypol HTC mix as a comparison This is a specialist high temperature concrete containing cementitious and polymer materials which is finished with an aluminium based surface coating.

The concrete for the test panels was made to a specification which contained the following requirements:

( 1) maximum aggregate/cement ratio of 5 8; ( 2) maximum water/cement ratio of 0 45; ( 3) fine aggregate content to be 32-37 % by weight of total aggregate content; ( 4) volume of entrained air to be 5 % + 1 %; and ( 5) slump of 30-60 mm.

For trial mixes, the strength requirement is based on 150 mm cube compressive tests, the mean of three samples to exceed the target mean strength of 32 3 M Pa at 7 days.

This specification was generally followed for all the limestone aggregate/OPC based panel mixes Where man-made aggregates or HAC were used, the mix proportions were based on the suppliers' recommendations and experience from earlier trials Throughout the panel production the requirements on maximum water/cement ratio, workability and 7 day strength were followed and achieved It should be noted however that based on our previous work, workability was assessed using the Vebe method as opposed to slump test, a time of around 10 seconds having been found to give satisfactory workability for our method of panel manufacture.

For comparison purposes, a simple spalling test was devised in order to establish the relative performance of 12 - different mix designs, the intention being to identify suitable mixes for panel production for final testing on an aircraft lift-off pad This test consisted of heating a 405 by 255 mm slab under an oxy-acetylene flame operating at a fixed height of 145 mm above the surface.

This height was chosen because it produced a spall in an OPC PQ control mix in around 12 seconds; this is similar to the residence times to first spall achieved under test conditions of a lift-off pad.

The trial mixes were principally concerned with obtaining a satisfactory workability (ie Vebe around 10 seconds) and an acceptable surface finish 100 mm cube strengths were also measured in some instances.

A total of ten trial mixes were made for small slab manufacture and testing The mixes were chosen to cover the range of materials described in Table 1.

Tests were performed on either one or two slabs of each mix design A minimum of six spalling tests was carried out on both top and bottom faces of the slabs The maximum exposure time was generally 40 seconds The results are summarised in Table 2, where ( 40) indicates no spall occurring and, for the purpose of averaging, a no- spall result was treated as having a time to first spall of 40 seconds.

Results of simple spall test The first observation is that this simple spall test produced significantly different results with different types of concrete The tests indicated that:

13 - (i) using a coarse OPC (mix 23) had little effect on spalling time; (ii) using a superplasticizer to produce a low water content (mix 22) produced a very poor top surface but a significantly more spall-resistant bottom surface This difference may be due to the excessively wet surface produced during vibration and finishing; (iii) using a cement replacement material in an otherwise standard PQ mix (mixes 20 and 21) significantly improved spalling resistance, PFA appearing to be marginally better than slag; (iv) using a lightweight aggregate in combination with limestone sand fines also improved spall resistance, particularly on the top surface (mixes 23 A and 23 C), and the use of PFA in addition resulted in further improvement (mix 32); (v) using HAC (mix 23 D) produced good spalling resistance, but not exceptional; and (vi) using Lytag lightweight aggregates entirely (mix 23 C) produced a concrete of exceptional spall resistance, only one test out of eighteen causing a spall.

The test results indicated that a more spall resistant concrete than the standard PSA PQ mix design could be produced, and test panels using nine of these mix designs were manufactured for full scale evaluation.

Production of Test Panels for Ground Erosion Testing Mix details The nine mix designs used to manufacture panels for full scale testing are described in Table 3 (mixes 24-32), the low water/cement ratio PQ trial mix being excluded at this 14 - stage All mixes, with the exception of the HAC concrete, contained an air-entraining agent It should be noted that the use of PFA substantially reduced the air content (to 2 %) and it was impossible to record the air contents of the mixes containing man-made aggregates at the operating pressure used.

The workabilities of all mixes were satisfactory, being in the range of 6 to 9 seconds The 7 day 150 mm cube strengths were all above the minimum required margin ( 32 M Pa) and were all in the range 35 7 to 40 M Pa (with the exception of the HAC mix).

Panel manufacture Generally, three 600 x 600 x 100 mm panels were made from each mix, the exceptions being mix 31 (two 800 x 600 x 100 mm panels) and mix 32 (one panel of each size) The compacting and finishing technique used on all test panels were the same as in the previous test programme ( 2), that is:

( 1) fill mould completely, with small surcharge; ( 2) compact with vibrating poker; ( 3) level with 2 passes of vibrating beam; ( 4) remove fat with 2 passes of scraper (channel section); ( 5) close surface with 1 or 2 passes of vibrating beam; and ( 6) trowel surface to finish.

The panels were cured in the laboratory under plastic sheeting for approximately 28 days No curing compound was used but the surface of each panel was occasionally - wetted.

Test procedure The aim of the test programme was to evaluate the performance of the different mix designs under jet blast conditions The rig was therefore run with a Harrier jet nozzle configuration with a jet exit temperature of 700 'C at a height of three nozzle diameters above the panel surface and at a pressure ratio of 1 5.

Each slab (held in its steel mould) was mounted on the rails of a ground erosion rig and then positioned centrally under the jet nozzle, with its surface normal to the axis of the jet and at a specified height Prior to exposure, the test pressure and temperature were established on a heat shield protecting the specimen.

Once the correct run conditions were achieved, the heat shield was removed using the ground erosion rig's linear motor powered trolley, to expose the concrete to the jet blast.

Each slab was tested for a minimum jet residence time of seconds If no spalling had occurred the test was continued for up to 120 seconds before shut-down.

Complete visual and thermal imaging data for all test runs was recorded on video tapes for analysis.

Relationship between surface temperature and residence The maximum surface temperature of the panels, obtained from thermal image recordings, is plotted against residence time in Figures 1, 2 and 3, the mixes being 16 grouped in terms of OPC/cement replacement, aggregate, and HAC variants.

Comparison with the performance of the OPC PQ control mix shows that the cement replacement materials reduced the rate of temperature increase (Figure 1) The rate of temperature gain of the EFG aggregate based mix was similar to OPC PQ control, whilst that of the Lytag mixes was greater (Figure 2) The Hypol HTC and one HAC panel were similar in performance to the OPC control; however, the other HAC panel gained temperature at a much reduced rate (Figure 3).

Residence time to first-spall The time to first spall for each panel is given in Table 4, and shown graphically in Figure 4 (where a 120 second residence time indicates no spall at the end of this test period) The results show:

(i) the OPC PQ concrete mix performed slightly better that in the previous tests, with an average residence time of 15 seconds to first spall (compared with around 10 seconds previously, Figure 5); (ii) the coarse OPC limestone mix was slightly worse than the control mix, with an average of 10 seconds; (iii) the performance of the mix containing PFA as a cement replacement was disappointing, producing an average of only 15 seconds to first spall; (iv) in contrast, the panels with GGBFS as a cement replacement performed significantly better than the PQ 17 - control with an average of 32 seconds, over a 40 second test duration (no spall being taken as a 40 second value); (v) the two HAC concrete panels tested performed significantly better than the aforementioned OPC mixes, giving an average residence time to first spall of 77 seconds The spalls, when they did eventually occur, were spectacular in their suddenness and size; (vi) of the three EFG based panels, one failed after 15 seconds and the other two had not spalled after 120 seconds; (vii) all the Lytag mixes tested under Harrier conditions lasted the full 120 seconds without spalling (one Lytag/sand panel had some very small spalls located above large sand particles, which were discounted); and (viii) the two Hypol HTC panels were unspalled after 120 seconds.

Temperatures at first spall The surface temperatures at first spall for those panels that did spall are given in Table 4 and are shown graphically in Figure 6 The OPC PQ control mix panels produced first spall temperatures in the range 320 -370 'C, which is very much in line with values observed in previous work (Figure 7) The two coarse OPC panels for which the thermal images were successfully recorded produced first spall temperatures of 350 'C and 4150 C.

The first spall temperatures for the panels made with an 18 OPC/limestone concrete containing either of the two cement replacement materials used (PFA and GGBFS) were all in the range 390 'C to 4250 C The average first spall temperature for the HAC panels was 4350 C.

By comparison, the maximum surface temperatures reached by Lytag based panels, which did not spall before termination of the test, lay in the range 510 540 'C.

The spalled areas of those test panels which failed within seconds are given in Table 4 and shown graphically in Figure 8 Whilst the range of spall size is large the average spalled area are similar, being between 53 and 96 cm 2 Thus for these materials, whilst the residence time to first spall varies significantly, the amount of damage sustained is of a similar magnitude.

Because of the successful performance of some of the mix designs under Harrier jet blast conditions, it was decided to test four panels under more severe conditions Here the test rig was operated at a jet exit temperature of 1000 'C and a pressure ratio of 4 0 These tests served two purposes: firstly to try to differentiate between the spalling resistance of the four best mixes which were unspalled when exposed to the earlier jet blast test, and secondly to gather some information on the suitability of these materials for the future.

The four mixes tested were HAC (mix 30), Hypol HTC (mix 33), the OPC Lytag/sand (mix 28), and OPC/all Lytag (mix 31) The results of these four test runs are summarised in Table 5.

The increase in surface temperature with residence time is 19 - shown in Figure 9 and the residence times to first spall in Figure 10 From the results it can be seen the HAC and Lytag/sand mixes performed similarly in terms of both rate of temperature increase and time to first spall (around 1 second) Both tests were terminated within 20 seconds because of excessive spalling.

The Hypol HTC panel's surface heated up more slowly and spalled after 6 seconds By the end of the test ( 40 seconds) the panel had suffered considerable spalling and erosion.

In contrast, the OPC based mix with Lytag coarse and fine aggregates did not spall within the 40 second test duration After testing, the surface of the panel showed no discolouration although fine cracking was visible, as might be expected.

Simple laboratory spalling tests The simple laboratory spalling test using an oxy-acetylene flame produced significantly different spalling times for the range of concrete mixes studied The mix designs, in decreasing order of spalling resistance were:

(i) excellent OPC/Lytag fine agg/Lytag coarse agg (ii) very good OPC PQC with PFA OPC/pfa/sand/Lytag coarse OPC PQC with slag (iii) good HAC/limestone aggregate OPC/sand/Lytag coarse OPC/EFG aggregates/sand (iv) medium OPC PQC control Coarse OPC PQC - (v) poor OPC PQC with superlasticizer The correlation with the performance of the mixes tested under jet blast conditions in the ground erosion rig was generally good, the principal exception being the OPC PQC mix with PFA, which performed poorly under jet blast conditions.

Comparative performance of mixes under jet blast conditions The jet erosion tests confirmed that the mix designs had a wide range of spalling resistance, all but two of which were significantly better than the OPC pavement quality concrete control The most relevant indicator of spalling resistance is the residence time to first spall Based on the tests at Harrier conditions ( 700 C, pressure ratio 1.5) and the subsequent tests at 1000 C, with a pressure ratio of 4 0 (ASTOVL) on the most resistant materials, the assessment of spalling resistance of the mix design is as follows:

(i) excellent OPC/Lytag fine and coarse aggregates unspalled after 40 sec at ASTOVL conditions (ii) very good OPC/sand/Lytag coarse unspalled OPC/pfa/sand/Lytag coarse after 40 sec HAC/limestone aggregate at Harrier conditions (iii) good OPC/EFG/ aggregate plus sand most panels unspalled 21 - OPC PQC with slag after 20 sec at Harrier conditions (iv) medium OPC PQC with PFA spalled at OPC PQC control around 15 sec at Harrier conditions (v) poor Coarse OPC PQC spalled at around 10 sec at Harrier conditions The Hypol HTC panel performance could be classed as very good (being unspalled after 40 secs at Harrier conditions) but surface material had clearly deteriorated after 120 secs, it being discoloured and friable (easily removed with a fingernail).

Time/temperature profiles showed that generally the surface of the man-made aggregate panels heated up more quickly that the OPC PQC control concrete; this can be attributed to their general insulating properties.

Conversely, the concretes with cement replacements heated up more slowlythan the OPC PQC control, suggesting an increase in their thermal conductivity.

The OPC PQ control concrete produced temperatures at first spall in line with values observed in our previous work (ie 320 370 'C) The OPC/limestone aggregate mixes incorporating either the two cement replacement materials (PFA and GGBFS) produced higher first spall temperatures (in the range 390 4250 C) The HAC concrete did not 22 - spall until around 4350 C had been reached, and the mixes containing Lytag aggregates, reached surface temperatures of over 5000 C without spalling.

These surface temperatures and associated residence times at first spall should be compared with conditions observed previously in the field from which we conclude that the

HAC and Lytag based mixes would perform well under Harrier field conditions The OPC/slag PQ concrete and EFG aggregate based mixes might also be expected to perform satisfactorily, although these materials are clearly not as spall resistant.

Mechanisms of spalling Earlier it was suggested that two possible mechanisms caused spalling: one associated with water vapour formation, and the other with a differential thermal expansion effect.

In practice it is likely to be a combination of these mechanisms which cause the spalling of the PQ concrete.

A comparison of the spalling performance of the various mix designs with that of the OPC PQ concrete shows that whilst modifying the cementitious matrix (using PFA or GGBFS) produced some improvements, by far the most significant increases in spalling resistance were achieved when the natural aggregates were replaced with man-made ones (particularly the Lytag lightweight aggregate) This could be because these porous aggregates can absorb the released water vapour or, perhaps more likely, because they induce less differential thermal expansion (or withstand it better).

23 - It is of interest to note that concretes often exhibit a significant increase in their coefficient of thermal expansion at temperatures above 320 'C, and this has been attributed to dehydration of the cement paste A value of around 8 x 10-6 per O C below, and 20 x 10-6 per O C above this temperature, for a calcareous gravel aggregate concrete is available from the literature, whilst interestingly an expanded shale aggregate showed an increase from only 6 1 to 7 5 x 10-6 per O C.

In view of the fact that the PQ/limestone concrete under test spalled at a temperature of 320 370 'C (ie just above the critical transition temperature of 320 WC) and that the man-made Lytag aggregate concrete (also with a coefficient of thermal expansion of around 7 x 10-6 per OC) did not spall, it appears that the most likely principal mechanism causing the spalling is a differential thermal expansion effect between the top very hot layer of concrete and the cooler layers below This effect will be aggravated over any natural aggregate particles which are close to the surface, as the particles will have a much lower coefficient of thermal expansion than the paste at the surface.

The results clearly indicate the general superiority of the all Lytag mix for resisting jet blast conditions.

They also suggest that it may be possible to develop from this mix a pavement concrete suitable for the next generation of vertical take-off aircraft.

Mix Description Proportions estimated slump Vebe 7 day cube strength

No free w/c (mm) (s) (M Pa) cement fine coarse coarse ratio 19 OPC 20 mm PQ Control 1 2 1 1 4 2 2 0 39 10 12 51 0 as 19, with slag at 40 % cement replacement 1 2 1 1 4 2 2 0 38 10 12 35 6 21 as 19, with pfa at 40 % cement replacement 1 2 1 1 4 2 2 0 38 25 5 34 3 22 as 19, with Sikament s/plasticizer 1 2 1 1 4 2 2 0 31 15 16 61 5 23 as 19, with coarse OPC 1 2 1 1 4 2 2 0 36 13 8 35 6 23 A OPC, EFG aggregates + sand 1 0 3/0 3 0 7 0 5 ( O 44) 53 7 23 B OPC, Lytag fine, Lytag coarse 1 0 9 1 3 ( 0 42) 23 C OPC, sand, Lytag coarse 1 1 2 1 6 ( 0 35) - 23 D HAC, limestone, plastiment VZ 1 1 8 1 1 1 9 0 26 32 OPC, pfa( 30 %), sand, Lytag coarse 1 1 1 1 5 ( 0 45) 20 7 38 4 ( not allowing for absorption of lightweight aggregates) Table 1 Details of trial mixes I lto Time to first spall under 'oxy-acet jet blast' Mix Description

No Top surface Bottom surface Range (s) Average (s) Range (s) Average (s) 19 OPC 20 mm PQ Control as 19, with slag at 40 % cement replacement 21 as 19, with pfa at 40 % cement replacement 22 as 19, with Sikament s/plasticizer 23 as 19, with coarse OPC 23 A OPC, EFG aggregates + sand 23 B OPC, Lytag fine, Lytage coarse 23 C OPC, sand, Lytag coarse 23 D HAC, limestone, plastiment VZ 32 OPC, pfa( 30 %), sand, Lytag coarse 10-18 13 ( 40) 24-( 40) 3-6 9-( 40) 17-( 40) ( 40) 10-( 40) 4-( 40) 8-,( 40) 5-( 40) 5-4 ( 40) 2-,( 40) 30 ( 40) 3-,( 40) 4-4 ( 40) 10-425 10-,( 40) 4-49 8-4 ( 40) 6-,( 40) Table 2 Spalling test results for trial mixes I kun I Mix Date Description Proportions Estimated Slump Vebe Air 7 day cube strength

No Cast free w/c (mm) (s) (%) (M Pa) cement fine coarse coarse ratio 24 17 May OPC 20 mm PQ control 1 2 1 1 4 2 2 0 36 20 8 5 3 37 7 GGBFS 17 May as 24, with,,slag at 40 % cement replacement 1 2 1 1 4 2 2 0 37 15 6 4 2 35 7 26 19 May Coarse cement, limestone aggregates 1 1 8 1 2 1 9 0 31 10 8 5 7 36 2 27 19 May OPC, EFG aggregates + sand 1 0 3/0 3 0 7 0 5 ( 0 41) 0 7 5 40 0 28 24 May OPC, Lytag coarse, river sand 1 1 1 1 5 ( 0 43) 0 9 39 8 29 24 May as 24, with pfa at 40 % cement replacement 1 2 1 1 4 2 2 0 31 0 9 5 2 0 39 9 26 May HAC, limestone,plastiment VZ 1 1 8 1 1 1 8 0 29 0 8 1 3 98 1 31 26 May OPC,Lytagcoarse,Lytagfine 1 0 9 1 3 ( 0 51) 0 9 5 36 8 32 26 May OPC, pfn, Lytag coarse, river sand1 1 1 1 5 ( O 43) 20) 7 38 4 33 Hypol HTC absorption of lightweight aggregates not allowed for Table 3 Mix details for Warton test panels I 27 - Mix/panel B Ae run Mix description Time to Surface Test Spalled area

No No first spall temp at duration at end of test first spall (s) ( C) (s) (cm 2) coarse OPC/limestone OPC PQ control HAC/limestone OPC PQ/slag/limestone OPC PQ/pfa/limestone OPC/EFG/sand 10.2 15.2 5.7 14.0 18.1 12.2 92 22 33 17 17 11 415 350 370 320 345 390 480 395 425 400 410 390 425 OPC/Lytag/sand OPC/pfa/Lytag/sand OPC/Lytag Hypol HTC 223 14 36 102 32 128 548 623 0 179 33 48 7 96 0 0 0 0 0 0 0 0 Table 4 Summary of results for tests carried out at a jet exit temperature of 700 C and pressure ratio of 1 5 26/1 26/2 26/2 24/1 24/2 24/3 30/1 30/2 25/1 25/2 25/3 29/1 29/2 29/3 27/1 27/2 27/3 28/1 28/2 32/1 32/2 31/1 33/1 33/2 1642 1543 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 28 - Mix/panel B Ae run Mix description Time to Surface Test Spalled area

No No first spall temp at duration at end of test first spall (s) ( C) (s) (cm 2) 33/3 1666 Hypol HTC 6 500 40 ( 1) 31/2 1667 OPC/Lytag 40 0 29/3 1668 OPC/Lytag/sand 1 2 540 15 ( 1) 30/3 1669 HAC/limestone 1 0 460 20 ( 1) Table 5 Summary of results for tests carried out at a jet exit temperature of 1000 C and pressure ratio of 4 0 29 - C 456/S

Claims (13)

Claims

1 A pavement quality concrete mix wherein a proportion of the cement contained in the mix is replaced by ground granular blast furnace slag (GGBFS) so as to increase the resistance to high temperatures of the concrete produced from the mix.

2 A concrete mix according to Claim 1, wherein up to 50 % of the mix is replaced by GGBFS.

3 A concrete mix according to Claim 2, wherein 40 % of the mix is replaced by GGBFS.

4 A pavement quality concrete mix containing naturally occurring aggregates of limestone and sand, wherein at least a proportion of the said naturally occurring aggregates is replaced by pulverized fuel ash (PFA) in the form of a lightweight aggregate, so as to increase the resistance to high temperatures of the concrete produced from the mix.

A concrete mix according to Claim 4, wherein the PFA is in the form of pellets.

6 A concrete mix according to Claim 5, wherein the pellets are of 10 mm aggregate size.

7 A concrete mix according to Claim 5, wherein the pellets have been ground into a lightweight fine - aggregate.

8 A concrete mix according to any one of Claims 4 to 7, wherein the PFA is that marketed under the name LYTAG by Boral Lytag Limited.

9 A concrete mix according to any one of Claims 4 to 7, wherein the PFA comprises a waste product from the generation of electricity at coal burning power stations, which waste product has been pelletised and sintered at 1200 C to produce hard spherical nodules with a 40 % void ratio.

A concrete mix according to any one of Claims 4 to 9, wherein the PFA is incorporated into the mix in combination with natural sand and cement.

11 A concrete mix according to any one of the preceding claims, additionally comprising an air-entraining agent.

12 Concrete produced from a concrete mix according to any one of the preceding claims.

13 A concrete mix substantially as herein described with reference to the accompanying drawings and tables.

Published 1990 as The Patent Office, State House, 6 f 71 High Holborn, London W Cl R 4 TP Further copies maybe obtaunedfrom The Patent Office.

Sales Br nch, St l Mary Cray, Orpington, Kent BR 5 3RD Printed by Multiplex techniques ltd St Mary Cray Kent, Con 1/87