Dicata.ing.unibs.it

Concrete Repair, Rehabilitation and Retrofitting II – Alexander et al (eds) 2009 Taylor & Francis Group, London, ISBN 978-0-415-46850-3 Sulphate resistance of high volume fly ash cement paste composites E. AydinCyprus International University, Mersin 10 Turkey, North Cyprus ABSTRACT: This research was carried out to evaluate the effects of using high volume Class C fly ash on strength and sulphate resistance of construction materials. Physical, mechanical and sulphate resistance tests were conducted on the Φ50 mm/100 mm specimens. Physical tests considered were apparent specific gravity, water absorption and dry unit weight. Mechanical properties considered were compressive strength, and flexu-ral strength. The durability properties considered was: sulphate resistance. In general strength and sulphate resistance of High Volume Fly Ash (HVFA) were considerably affected by amount of fly ash. Also, the strength properties for the 20% fly ash mixtures were either comparable or superior to the no-fly ash concrete. The sul-phate resistance of HVFA composites was either comparable to or better than the no-fly ash composites. All the mixtures, with and without fly ash, tested in this investigation conformed to the strength and durability require-ments for excellent quality structural grade concretes. Based on the sulphate test results of this study indicates that the engineering performance of the final product can be adequate for using them in the manufacturing of construction materials (brick, ceramic tile, paving stone and briquette) and various civil engineering applica-tions such as construction of structural fills, embankments, grouting injection, road bases and sub-bases.
Alkali, magnesium and calcium sulphates present in alkali soils and ground waters react chemically The consumption of aggregates of all types has been with the hydrated lime, hydrated calcium aluminate increasing in recent years. The continued and expand- and calcium silicate hydrates to form calcium sul- ing extraction of natural aggregate is accompanied phate and calcium sulphoaluminate. These products by serious environmental problems. Often it leads have a considerably greater volume than the com- to irremediable deterioration of the countryside. The pound they replace, so that the reactions with the sul- use of waste materials such as fly ash to modify the phates lead to expansion and disruption of concrete properties of concrete and reducing the proliferation (Neville, 1995). On the other hand, the conversion of this potential pollutant also adds desirability to its of soluble calcium hydroxide to cementitious com- use in concrete structures. The engineering proper- pounds decreases bleed channels, capillary channels ties of concrete, a heterogeneous material composed and void spaces and thereby reduces permeability. At of cement paste and aggregate, are greatly affected the same time, the above chemical reaction reduces by the pore structure and the moisture content of the the amount of calcium hydroxide susceptible to attack medium. Porosity is an intrinsic property of cement by weak acids, salts or other sulfates. Replacing a paste and influences the strength and permeability portion of Portland cement with fly ash reduces the of cement paste, mortar, and concrete, as well as amount of reactive aluminates (tricalcium aluminate) its mechanical properties and durability (Erdogan, available for sulfate. Fly ash chemically binds free 1997). The porosity of the paste depends on many lime in cementitious compounds, rendering it una- factors and is typically decreased by decreasing the vailable for sulfate reaction. Fly ash activity reduces water to cement ratio. The type of cement also has concrete permeability, keeping sulfates from pene- major role on the amount of hydration products and trating concrete (Naik, 1992). ACI Committee 226 thereby the pore to solid ratio as a result of the rate affirms that, the sulphate resistance of either plain and degree of hydration within the time domain. The concrete or fly ash concrete is controlled by the same influence becomes less significant for mortar and factors which are: curing conditions, exposure and w/c concrete proportional to the increase in the aggregate ratio. The effect of fly ash on sulphate resistance is dependent upon the amount and individual chemical and physical characteristics of fly ash and cement dominance over the years (Neville, 1995). Despite used (Ramyar, 1993). Many investigators stated that the high energy consumption in cement manufacture, the sulphate resistance of fly ash pastes was largely the total energy requirement to make a cubic meter controlled by the rate of diffusion of sulphate ions of concrete is much lower than any other structural into the paste, being a function of pore size distri- material (Uygar & Aydin, 2005). This also makes bution. It has been shown that entry size of sulphate it very attractive in an age where energy conserva- susceptible pores within the fly ash cement paste tion and environmental protection considerations are decreases with an adequate curing (Mather, 1982, paramount. Furthermore the use of waste materials Mehta, 1985 & Naik, 1992). Mehta found a corre- such as fly ash to modify the properties of concrete lation between the alumina content of fly ash and while at the same time reducing the proliferation of sulphate resistance of Portland cement products con- this potential pollutant also adds desirability to its use taining the fly ash. He conducted an experiment on in concrete structures (Baker et al. 1991).
fly ash cement mortar specimens with 20% cement Based on the Electric Power Research Insti- replacement by high and low calcium fly ashes. After tute (EPRI) report, more than 600 million tons of accelerated curing in sulphate solution, he observed coal ashes are produced annually in the world, with that low alumina content specimen exhibited 8% 105 million tons per year produced in the United strength gain, whereas high alumina content speci- States. The disposal cost of fly ash is $10–20 per ton. men showed 23% strength decrease (Mehta, 1985). Therefore, $6 billion is needed per annum. It is neces- Mather has reported that, 30% replacement of several sary to utilize a large volume of fly ash for the future high C A content cements with different high lime activities (EPRI, 2003). Utilization of high volume fly ashes made the system less sulphate resistant. fly ash as a resource has been studied for decades On the other hand, Mehta, Hooton and Manz et al. in many areas such as cement/concrete applications, showed that, some high lime fly ashes may satisfacto- brick, ceramic tile, lightweight aggregate, highway rily improve sulphate resistance of concrete (Mather, pavements (Aydin et al. 2004). Based on the durabil- 1982 & Hooton, 1986). Scholz found that, for high ity tests of this study indicates that the engineering quality ash, replacement levels of 40–45% in OPC/ performance of the final product can be adequate PFA mortars resulted in about same level of resist- for using them in the manufacturing of construc- ance as for mortar made solely with sulphate resisting tion materials (brick, ceramic tile, paving stone and Portland cement. Scholz also noted that the quality briquette) and various civil engineering applications of the ash, determined by its pozzolanic index, par- such as construction of structural fills, embankments, ticle size distribution, glass content and surface area, grouting injection, road bases and sub-bases.
is important to the durability obtained. Wesche and Schubert replaced cement with ash to the 50% level and observed increased resistance, especially when a cement of low sulphate resistance was used. One of the reasons suggested for the beneficial action of fly It is necessary to utilize larger volumes of fly ash in ash in sulphates is the resultant decrease in pH in the the construction industry. End-products made with pore solution due to removal of calcium hydroxide; if HVFA have superior engineering properties, as well the pH is reduced below about 10, ettringite is unsta- as economic benefits. The objective of this study is: ble and thus the large expansions associated with its formation cannot occur (Wesche, 1991).
• To investigate the physical, mechanical, sulphate The engineering properties of the material both resistance properties of HVFA fly ash (>75%) m in fresh and hardened state are highly influenced by ainly composed of silica fume, lime, cement and the physical (fineness, grain size distribution, par- ticle shape) and chemical (pozzolanic activity/rate • To produce cost-effective and environmental and degree of hydration) properties of the mix ingre- friendly building products. These HVFA cement dients, mainly by the properties of fly ash being the paste composites can satisfactorily be used in main constituent (Poon et al. 1997). The design water manufacturing building materials such as bricks, content in fresh state, thereby the porosity in hard- briquettes, tiles and paving stones.
ened state are highly influenced by the physical prop-erties of the mix ingredients (Hewlett, 1998). The Based on the experimental results obtained in chemical oxide composition provided in the medium cold bonded engineering properties, five of the of the mixes is also of high importance in the char- mix groups were selected to manufacture the build- acterization of CSH formation, thereby, the strength ing materials. Design mix groups were prepared at development and the formation of pore structure. 100 mm slump for the manufacturing of briquettes, Concrete has been the most significant building mate- paving stones, bricks, tiles. Briquettes were prepared rial for many generations, and has not lost its market by using briquette molds and the others were by using 50 mm cubic, 40 mm * 40 mm * 160 mm prismatic of silica fume is 2.20. The chemical composition is and Φ50/100 mm molds. The workability of the mix- ture combinations were measured by using slump test Lime: The specific gravity of lime is 2.17 and its according to the ASTM C143-90a and flow table test chemical composition is presented in Table 1.
Admixture: A melamine based polymer dispersion water reducing admixture (WRA) was used in all of Cement: BS EN 197-1 CEM I 42.5 N cement was used. The chemical composition is presented in Table 1. The Blaine fineness of the cement is 2905 Five mixture groups with three different slump val- cm2/gr, and its specific gravity is 3.15. The grain size ues (0 mm, 100 mm and 200 mm) were studied dur- distribution of cement is given in Figure 1.
ing this study. The mix designs were chosen based on Fly Ash: Fly ash from Soma Thermal Power Plant the water to binder (w/b) ratio of the composites. The in Turkey was used in this study. Blaine fineness is numerical analyses results indicate that the relation- 2050 cm2/gr, the specific gravity is 2.08, and mean ship between the consistency (w/b, slump, flow), the grain size is 27 mm. The chemical composition is physical (apparent specific gravity, dry unit weight) presented in Table 1. The grain size distribution of fly and mechanical properties (unconfined compressive and flexural strength) of fly ash mix groups of differ- Silica Fume: It is obtained from the ferrochrome ent physicochemical properties are highly correlated factory located in Antalya-Turkey. The specific gravity with each other. Mixture groups by weight of cement were mixed in the Hobart mixer having 2.5 liters capacity. First, fly ash, silica fume, lime and cement Table 1. Chemical compositions of Soma fly ash, cement, were mixed together in dry form for 30 seconds and then tap water was added to the mix. The water to cementitious material ratio was kept as 0.39 to which the corresponding slump value was 100 mm. The w/c ratio was determined from the previous laboratory studies based on the relationship obtained for flow table—slump test. Slump and flow values were meas- ured and relationship between slump-flow was used After the mixes were cast and consolidated by vibration for 1 minute in briquette molds as described in the ASTM C109M-02 were taken. The samples were extracted from the molds after 1–2 days and continued to be kept in the curing room till the time of Dry bulk density, apparent specific gravity and water absorption experiments were performed accord-ing to ASTM C127-73.
The compressive strength tests were carried out on Φ50 mm/100 mm specimens according to the require-ments of ASTM C109M-02. The flexural strength tests were carried out on 40 mm * 40 mm * 160 mm prisms according to the requirements of ASTM C348-02.
The soundness of all mix groups were determined according to the ASTM C88-90 on the specimens fractured in the flexural strength, tensile strength and compressive strength tests. The purpose of conduct-ing the experiment on cement paste was to obtain the relevant index values to have an idea of the durabil- ity of the mix groups against adverse environmental conditions, though the method is used specifically for aggregates. The specimens were subjected to two cycles wetting in Na SO solution and drying in oven Figure 1. Grain size distribution of soma fly ash and ± 5ºC temperatures; the losses in weight then were found. The mixture proportions of the groups Table 3. 28—day UCS and 28-day FS values of mix groups.
are presented in Table 2. As the w/b increases, the slump increases as a consequence of the decrease in the interparticle friction. The spherical shape of fly and for Grade MS compressive strength of 15 MPa ash particles as well as the presence of glassy phase on the fly ash surface improves significantly the Based on the flexural strength (FS) test results structure of paste. Therefore the paste is effectively similar behavior to that in UCS is observed. Silica densified and the water content can be reduced. On fume replacement increases the FS of Soma fly ash the other hand the reverse effect occurs, because of mix groups at early ages probably due to the decrease the lower fly ash density as compared with cement. in volume of voids, however relatively decreases the Consequently, at higher fly ash content in cement the FS at late ages, since unhydrated silica fume, in the lack of adequate amount of calcium hydroxide, acts as bond barrier within the CSH network (Hewlet, 1998).
Proper curing is extremely important for fly ash mix groups. The pozzolanic reaction is very slow and any change in the environment can cause an adverse effect on strength (Neville, 1995). Expansion of gel due to absorption of pore water or contraction of gel The dry unit weight (DUW), apparent specific grav- due to extraction of water reduces the FS values more ity (ASG) and water absorption values of varies from in fly ash mix groups. The reduction is eliminated by 13.1 kN/m3 to 16.5 kN/m3, 2.02 to 2.39 and 10.22% to 31.80% for Soma fly ash mix groups.
Weight loss by sodium sulphate solution of HCP com- The 28-day unconfined compressive strength (UCS) posites vary between 9.19–22.75%, 3.75–15.32%, and and 28-day flexural strength (FS) values varies from 3.50–14.48% for 0 mm, 100 mm and 200 mm slump 3.6 MPa to 18.2 MPa and 0.82 MPa to 3.82 MPa for values of 7 days curing periods and 8.74–22.55%, Soma fly ash mix groups. The results are shown in 3.16–15.14%, and 2.49–15.27% for 0 mm, 100 mm and 200 mm slump values of 28 days curing periods.
Based on the UCS values of HCP (hardened cement Based on the results of sodium sulphate solu- paste) composite; the final composites are adequate tion indicate that a moderate to high perform- for manufacturing precast/prestressed elements, con- ance in terms of the durability of the composite is struction of structural fill, base and sub-base course, expected. The cement replacement for Soma fly ash construction of catch basins and manholes and both mix groups decreases the weight loss by enhancing non-load bearing and load bearing elements. The the cementitious compound formation. The silica final composites can also be used in manufacturing fume replacement also reduces the weight loss for Class MX and Class NX paving bricks, standard and fly ash type C which was provided that it does not special type tile production and manhole brick Grade increase the demand for mixing water, leading to MS and Grade MM; those bricks are intended to be the increase in the volume of voids. The WRA addi- used in manholes and catch basins not requiring high tion also reduces the weight loss to a greater extent degrees of abrasive resistance. The final composites provided that the volume of voids is not increased are not suitable for heavy vehicular brick applica- due to the decreased amount of mixing water for the tions. The compressive strength requirements of those corresponding compactive effort. The results indi- applications are not satisfied (i.e. for Class MX, Class cate that a moderate to high performance in terms NX, Grade MS compressive strength of 17.2 MPa of the durability of the composite is expected, when the relatively severe laboratory simulation conditions achieved. For this reason the best results in durability are considered. In Figure 2, Briquette samples after is obtained in superplasticized enriched fly ash mix durability test-Oven-dry and in Figure 3, Briquette groups (Aydin et al. 2004). As the cement content samples after durability test-prior to end of the cycle in the concrete mixture increases, hydration product Ca(OH) will also increase and hence the amount of The use of fly ash influences the physico—chemical Ca(OH) with which the fly ash will enter into reac- effects associated with pozzolanic and cementitious tion will increase, then an increased amount of C-S-H reactions that result in pore size reduction and grain will result. Consequently, in this way, fly ash will be size reduction phenomena. This affects the strength used more efficiently (Aimin & Sarkar, 1994). Super- and durability of hardened cement paste (Aydin & plasticizer causes a better hydration of fly ash Port- Doven, 2006). The durability properties of the final land cement pastes (Papadakis, 1999). Based on the composites are directly related with its porosity. Lower sodium sulphate test results; the final composites are the porosity higher the durability properties could be classified as medium to high sulfate resistant (stand-ard specification of weight loss should be in between 6–16%). With the continuous hydration and reactions of the HVFA pastes and the improvement of the bond between the fly ash particles and matrix, the fly ash particles become more effective inclusions in the sys-tem. The filler effect leads to reduction in porosity of the matrix and provides a dense microstructure and thus increases the strength of the system. Also since a part of the paste volume is occupied by fly ash particles even after long term curing, the hydra-tion and reaction products are required to fill a much smaller space. However; addition of silica fume or WRA into the system increases the resistance against sodium sulphate (Aydin & Doven, 2006). The higher strength developments and reduction in weight loss with silica fume are due to pore size refinement and matrix densification, reduction in content of Ca(OH) and CH (Hewlett, 1998). As the silica fume is finer than the cement particles, the finer particles of silica fume fill the gap between cement particles resulting in impermeable microstructure of the cement paste (Montgomery et al. 1981).
Figure 2. Briquette samples after durability test-Oven-dry.
Utilization of high volume fly ash as a resource has been studied for decades in many areas such as cement/concrete applications, brick, ceramic tile, light-weight aggregate, highway pavements. Based on the physi-cal and mechanical tests of this study indicates that the engineering performance of the final product can be adequate for using them in the manufacturing of construction materials (brick, ceramic tile, paving stone and briquette) and various civil engineering Table 4. Minimum compressive strength values of base, Sub-base and Sub-grade courses (ACI, 1985).
Minimum unconfined compressive strength at 7 days, psi (MPa) Figure 3. Briquette samples after durability test-prior to Table 5. Compressive strength requirements of tiles the physical and mechanical properties of mix groups are highly correlated with each other, good enough to construct a nomograph enabling target slump and strength based mix design. However; combination of WRA, silica fume together with HVFA shows better engineering properties than HVFA cement paste mix- tures. Higher ash replacements are become acceptable and some ashes which normally do not meet activity standards is become acceptable. Published research on mixtures containing both a fly ash and WRA is limited. The importance of making trial mixes with each source of fly ash is emphasized. The fly ash used Table 6. Compressive strength requirements of paving in this study is a Class C (high lime, Soma) fly ash, and the mix proportions shown are for this specific fly ash.
The properties of cement–based materials are pri- marily affected by the w/c ratio, chemical and mineral composition of binder material, microstructure and pore geometry of the cementitious materials. When siliceous by products are introduced, they change the behavior of cementitious composites significantly. The addition of fly ash can increase the fluidity of the fresh mixture. The effect of water reduction and the contribution of the fly ash to the mixture com- applications such as construction of structural fills, bination much depend on the source of the fly ash embankments, grouting injection, road bases and or the pozzolanicity. The w/b affects the strength and sub-bases (Table 4, Table 5 and Table 6). the durability of the material to a great extent due to its influence on the physicochemical (porosity and degree of hydration) properties of the cement paste in hardened state. Slump values of Soma fly ash mix groups are more sensitive to change in w/b ratio are The originality of this study is to manufacturing of due to its well graded grain size distribution. As the cost-effective environmental friendly building prod- w/b increases, the slump increases as a consequence ucts by using HVFA cement paste. Based on the of the decrease in the friction between the cement and report published by the Governmental Development of Turkish Republic, for 1 tone of brick production The engineering properties of the material both 1.3 tones, and for 1 brick/tile production, 4 kilogram in fresh and hardened state are highly influenced raw material (clay, shale, etc.) is needed. The replace- by the physical (fineness, grain size distribution, ment of fireclay and shale material by fly ash could particle shape) and chemical (pozzolanic activ- save about $ 10 per tone for materials, the cost of the ity/rate and degree of hydration) properties of the energy, and the time required to complete burnout of mix ingredients, mainly by the properties of fly ash the clay component is replaced by fly ash. The clay being the main constituent. The design water content minerals in coals are fired during coal combustion, in fresh state, thereby the porosity in hardened state so the energy consumption from firing during brick are highly influenced by the physical properties of manufacture is not needed, resulting in energy sav- the mix ingredients; the Soma fly ash, having a well ings. Few published literature is available on this sub- graded grain size distribution is more sensitive to the ject and neither of them is included HVFA cement change in w/b in terms of variation in the slump and HVFA composites show adequate sulphate resist- As an admixture, fly ash functions as either a ance and the trend is exponentially increasing by fly partial replacement for or an addition to Portland ash content for all slump classes. Use of silica fume cement and is added directly into ready-mix concrete together with WRA shows better sulphate resist- at the batch plant. Researches on high volume fly ance than ordinary Portland cement-fly ash system. ash utilization have been started by CANMET since Removal of calcium hydrates (CH) and decrease of 1980 s. Chemical compositions and physical proper- porosity by C-S-H formation causes an improvement ties of all kinds of fly ashes and their various usages in sulphate resistance. The numerical analyses results in concretes have been studied by many researchers. indicate that the relationship between the consistency, Although, using high volume fly ash (HVFA) has numerous advantages from the structural as well Aydin, E. & Doven, A.G. 2006. The Influence Of Water as the economic point of view (Naik, 1992). Nev- Content On The Ultrasonic Pulse Echo Measurements ertheless, there is a lack of commercial products Through High Volume Fly Ash Cement Paste;—Physi- containing high volumes of Class C and Class F fly comechanical Characterization. Research in Non-Destructive Evaluation 17: 177–189.
ash. If a large quantity of fly ash can be used in the Aydin, E, Pekrioglu, A, Uygar, E, Atak, C.E. & Doven, manufacture of fired bricks and other related prod- A.G. 2004. 6th International Congress on Advances in ucts, the disposal problem will be decreased, and a Civil Engineering, Istanbul, Turkey, 6–8 October 2004. value-added construction product will be created. The 28 days UCS results of the basic and fiber mix Baker, J.M., Nixon, P.J., Majumdar, A.J., & Davies, H. design groups of Soma fly ash shows that the final 1991. Durability of Building Materials and Components. composites can be classified as C16/C25 concrete Electric Power Research Institute (EPRI) Report. 2003. A fundamental approach, in this manner, is to use Combustion by-product use: 1–8.
Erdogan, T.Y. 1997. Admixtures for Concrete. The Middle neat high volume fly ash cement paste in the construc- tion/production of semi–structural/isolatin materials. Hewlett, P.C. 1998. Lea’s Chemistry of Cement and Con- The strength and the durability of the fly ash can be crete, 4th edition, John Wiley and Sons Inc, New York, improved to a great extent by addition of consider- ably low amounts of cement and/or lime and other Hooton, R.D. 1986. Properties of a High Alkali Lignite Fly mineral admixtures, letting the final product still be Ash in Concrete. Proceedings of the Second International within economical limits in comparison to the same Conference on Fly Ash, Silica Fume, Slag and Natural kind conventional construction materials.
Pozzolans in Concrete. Madrid, Spain, V.M. Malhotra, Based on the DUW values; final product is con- Ed., ACI Publication SP-91, 1: 333–345.
Malhotra, V.M. 1989. Fly Ash, Silica Fume, Slag, and Natu- sidered as a light weight material and can be used ral Pozzolans in Concrete, Proceedings Third Interna- satisfactorily in the manufacturing of light weight tional Conference, Trondheim, Norway, Vol. 1–2.
aggregates and semi-isolating materials. The abil- Mather, K. 1982. Current Research in Sulfate Resistance at ity to proportion mixtures having low unit weights the Waterways Experiment Station. George Verbeck Sym- is especially advantageous where weak soil condi- posium on Sulfate Resistance of Concrete, ACI, Detroit, tions are encountered and weight of the fill must be minimized. The use of high volume fly ash produced Mehta, P.K. 1985. Influence of Fly Ash Characteristics on denser and stronger matrix properties and resulted in the Strength of Portland Fly Ash Mixtures. Cement and a higher ability to resist the sulfate attack.
Concrete Research 5(3): 669–674.
Montgomery, D.G., Hughes, D.C. & Williams, R.I.T. 1981. Based on the compressive strength tests results; Fly Ash in Concrete: A Microstructure Study. Cement final products are adequate for using them in low and Concrete Research 11: 591–603.
to medium technology applications such as in road Naik, T.R. 1992. The State of Art Report: High-Volume Fly bases, manufacturing of bricks, tiles, and ceramic Ash Concrete Technology. Report on CBU-1992–15.
applications (Refer to Table 4, Table 5 and Table 6).
Neville, A.M. 1995. Properties of Concrete, Addison For structural fill applications; required minimum compressive strength may vary from 0.7 MPa to 8.3 Papadakis, V.G. 1999. Effect of Fly Ash on Portland Cement MPa (ACI 230, 1985). The 28-day UCS values of the Systems Part I: Low-Calcium Fly Ash. Cement and Con- final products are also adequate for those applica- crete Research 29: 1727–1736.
Poon, C.S. Wong, Y.L. & Lam, L. 1997. The Influence tions. Depending on the strength requirements final of Different Curing Conditions on the Pore Structure product can be used for foundation support.
and Related Properties of Fly Ash Cement Pastes and Mortars. Construction and Building Materials 11(7–8): 383–393.
Ramyar, K. 1993. Effects of Turkish Fly Ashes on the Portland Cement Fly Ash Systems. PhD Thesis.
Uygar, E. & Aydin, E. 2005. Effect of Silica Fume on the ACI 230.1 R, 1985. American Concrete Institute Report on Fresh and Hardened Properties of High Performance Concrete. 3rd International Conference, Construction Aimin, X & Sarkar, S.L. 1994. Microstructural Develop- Materials, Performance, Innovations and Structural ment in High Volume Fly Ash Cement System. Journal Implications, Vancouver, Canada, August 22–24.
of materials in Civil Engineering 6(1):117–136.
Wesche, K. 1991. Fly Ash in Concrete: Properties and ASTM C212. 1996. Standard Specification for Structural Performance, Chapman and Hall, Newyork.
ASTM C902. 1995. Standard Specification for Pedestrian

Source: http://dicata.ing.unibs.it/plizzari/CD/Pdf/013.pdf