- Email: [email protected]
Optimization of the polymer concrete used for manufacturing bases for precision tool machines
Composites: Part B 43 (2012) 3061–3068
Contents lists available at SciVerse ScienceDirect
Composites: Part B journal homepage: www.elsevier.com/locate/compositesb
Optimization of the polymer concrete used for manufacturing bases for precision tool machines Header Haddad a,⇑, Mohammad Al Kobaisi b a b
Faculty of Engineering and Industrial Sciences, Swinburne University of Technology, Melbourne, VIC 3122, Australia School of Applied Sciences, RMIT University, Melbourne, VIC 3000, Australia
a r t i c l e
i n f o
Article history: Received 19 September 2011 Received in revised form 3 April 2012 Accepted 4 May 2012 Available online 18 May 2012 Keywords: A. Thermosetting resin A. Polymer–matrix composites (PMCs) B. Adhesion C. Finite element analysis (FEA)
a b s t r a c t Due to its superior damping ratio, high adhesion and fast curing, polymer concrete is used in manufacturing bases for a wide range of precision machines. The coefficient of thermal expansion for polymer concrete is one of the main parameters that can affect the level of accuracy in precision tool machines. Flexural strength is a fundamental strength of the base. In this study six aggregates (basalt, spodumene, fly ash, river gravel, sand and chalk) were investigated. Polymer concrete samples were prepared with different compositions of aggregates containing the same resin volume fraction (aggregates 83% and risen 17%). A four points flexural test was employed to measure the flexural strength of the polymer concrete samples. The coefficient of thermal expansion for polymer concrete was measured using a custom built device. The preliminary optimum composition, with the highest flexural strength and lowest thermal expansion coefficient, was found to be basalt, spodumene and fly ash. Basalt, sand and fly ash composition was the second in the rank. The second composition was nominated for further optimization in terms of resin volume fraction in consideration of its ability to adapt a smaller amount of resin. Different samples of polymer concrete were prepared with a variety of resin volume fractions as follows; 17%, 15% and 13%. The resin volume fraction has been demonstrated to have a significant effect on the coefficient of thermal expansion and flexural strength for polymer concrete. The final optimized composition was basalt, sand and fly ash (filler 87% and resin 13%). ANSYS 13 software was employed in visualizing the influence of polymer concrete compositions on the thermal expansion of the base and how it affected the level of precision of the tool machine. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction Polymer concrete (PC) is a composite material consisting of well-graded inorganic aggregates bonded together by a resin instead of the water and cement binder typically used in traditional cement concretes [1]. Unsaturated polyester resin (UPR) is one of the binders used in polymer concretes providing excellent adhesion to solid particulate fillers, good stiffness, dimensional stability and a high damping ratio [2]. UPR is a thermosetting material created by combining two commonly used polymerizing constituents, namely unsaturated polyester (UP) and a vinyl based monomer such as methyl methacrylate (MMA) [3]. This combination of inorganic fillers and a polymeric matrix allows for partial control of the mechanical and thermal properties of the produced composite system. PC can be tailor-made for specific applications, based on the conditions and functional requirements. PC is used for a variety of applications, such as industrial flooring [4], underground piping, ⇑ Corresponding author. E-mail addresses: [email protected] (H. Haddad), mohammad.alkobaisi@ rmit.edu.au (M. Al Kobaisi). 1359-8368/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.compositesb.2012.05.003
frame supports for MRI, X-ray, and CT-Scanners, for CNC manufacturing centers [5] and for CNC grinding tool machines. This variety of applications reflects the diversity of PC properties, such as good mechanical strength, high damping ratio, fast curing, good adhesive strength, long term durability, low permeability of water and chemical resistance [6]. The level of accuracy for precision machines is governed by the machine base-frame upon which the positioning and drive subsystems are installed. In a precision machine, the base-frame ensures the positioning of subsystems under the complex dynamic and static conditions. In these operational conditions, forces, torques and heat gradients are dynamically produced throughout the system to satisfy complex product geometry and elevated levels of tool precision requirements [7]. A milling tool is a good example to demonstrate the effect of the level of tool precision on product accuracy through the manufacturing sequence. Once a milling tool is produced it will then be used in a CNC milling machine to manufacture other products. Some of these products can be considered as tools for a particular application, such as micro injection molding that can be used, for producing the micro gears used in a micro medical device. Each process in the manufacturing sequence of the milling tool will have a manu-
3062
H. Haddad, M. Al Kobaisi / Composites: Part B 43 (2012) 3061–3068
facturing influential effect on the precision of the micro injection mold (tool). Based on this statement, the accuracy and functionality of the end product (micro plastic gears) will be affected by the post process levels of accuracy. In this case micro plastic gears inherit the level of accuracy from the original tool (milling tool) through the process sequence of manufacturing the micro mold. This depends greatly on the original milling tool produced by the CNC grinding tool machine and the milling machine used for the milling tool. The level of accuracy of both machines determines the precision level for the rest of the products being dealt with, and hence, the importance of polymer concrete base for both the CNC grinding machine tool and the milling machine [5]. It has a tremendous effect on the precision of the milling tool by damping the unwanted vibration created by servo motors during the manufacturing operation [8]. PC suffers from some drawbacks such as shrinkage [9] and low creep resistance [10]. The biggest drawback on operational effectiveness is the coefficient of thermal expansion (CTE) of the PC used in precision machinery applications. According to Valore, CTE is high for PC containing UPR as a binder [11] in comparison to cast iron and other metal inserts in the base. A complex combination of metal inserts and connecting plastic pipes for fluids are essential in the PC base of CNC grinding tool machines. This condition leads to a non-uniform distribution of thermal expansion for the base during operation. The non-uniform thermal expansion distribution on the base may reduce the target accuracy of the precision tool machine. Achieving a CTE for the PC as close as possible to the CNC grinding machine metal components (inserts) could enhance the distribution of thermal expansion for the base. It has been reported that by increasing the resin volumetric ratio in the PC, leads to an increase in the CTE of the PC and other mechanical properties, such as compression strength and thermal conductivity [12]. The properties of a PC composite are also determined by the properties of the aggregates, such as particle shape and size distribution, particle surface properties and aggregate volume fractions. The compressive strength of PC is elevated with an increase in the compressive strength of the course aggregate and decreased linearly with an increase in its volume fraction [13]. The main purpose of this study is to optimize the aggregate composition with a sufficient amount of resin for the PC to reduce the CTE and increase the flexural strength, which will result in an improvement in the level of accuracy for the precision machine base. This will have a positive impact on the precision level of the end product. The precision level of end product depends on the accuracy the tools and machine that it dealt with, through manufacturing sequence.
2. Materials and method Commercial general purpose, unsaturated polyester (AROPOL) (67% unsaturated polyester dissolved in 33% styrene) was obtained from Huntsman Chemical Company (Australia), methyl methacrylate (MMA) from Degussa (Australia), cobalt octoate and Dimethyl aniline (DMA) from Alfa Aesar (USA), and methyl ethyl kenton peroxide (MEKP) – the commercial name is NR20 from Nuplex Industries (Australia). A resin sample was produced by mixing a 3:2 volumetric ratio of UP to MMA. To this mixture, 0.8% cobalt octoate (promoter), 0.2% DMA (accelerator) and 2% (v/v) MEKP (initiator) were added and mixed in subsequent order. The basalt, river gravel, sand, and chalk used as aggregates were from Roca. Fly ash was from Cement Independent Australia and spodumene from Talison Minerals, Western Australia. The aggregates true density qTrue (g/ cm3) was measured using a pycnometer, their bulk density qbulk (g/cm3) was measured using a measuring cylinder, and the packing factor (Vp) was calculated based on the ratio of the obtained true and bulk densities. Table 1 summarizes the measured properties
of the aggregates and referenced aggregate properties, such as CTE for each filler (as°C1) [14]. The fine aggregate (chalk, fly ash) particle size distributions were obtained using Mastersizer X (Malvern Instruments, USA). Middle and course aggregate size distributions were obtained using sieve analysis. Fig. 1 shows the particle size distributions for all aggregates. The BET (Brunauer, Emmett and Teller, 1938) active surface area of the aggregates was determined by nitrogen gas adsorption/desorption. Nitrogen sorption measurements were performed using Micromeritics ASAP 2010 (USA). The adsorption and desorption isotherms were recorded using a 89-point pressure table with 15 s equilibration intervals. The packing system of the aggregates was determined according to the Furnace method [15], starting with the number of aggregates based upon the ratio of the biggest particle diameter and the smallest particle diameter. The diameter ratio was calculated using data from Fig. 1. Volume of voids, which is equal to 100 – Vp (%), was calculated using Vp of aggregates in Table 1. By applying the diameter ratio and volume of voids in a special graph provided by the Furnace method [15], the number of aggregates generations was obtained. Both the diameter ratio and the void volume were too small, leading to three aggregate generations. According to the Furnas method, the diameter of intermediate generation D2 was calculated using the following expression:
D2 ¼
pffiffiffiffiffiffiffiffiffiffiffi D1 D3
ð1Þ
where D1 and D3 are the biggest and the smallest generations of particle diameters. Total absolute volume of fillers V 0 was calculated using the following formula:
V0 ¼
1 V1 V2 þ þ 1 þ V1 1 þ V2 1 þ V3
ð2Þ
where V1, V2, V3 are volumes of voids for the biggest, intermediate and the smallest generations. Relative volume of voids can be calculated using 1 Vp. The total volume fraction occupied by the solid particles V f was calculated using the following formula:
V f ¼ V 0 ð1 V 21 Þ
ð3Þ
The volume fraction of each aggregates component Pi was calculated using the below formula:
Pi ¼ V p =V f
ð4Þ
This approach takes into account volume of voids and provides a good indication of possible filler composition. However, it does not take into account particle shapes and can be used mainly for spherical bodies. To prepare the PC mixture fine aggregate 8.3%, coarse aggregate 49.8% and middle size aggregate 24.9% (v/v) was added to 17% (v/v) resin in subsequent order and mixed using traditional concrete mixer. All aggregate proportions were obtained using the Furnace method. The mixing continued for 15 min to ensure the best wetting of all particles with resin, in order to maximize the resin interface with the pebbles and obtain sufficient adhesion. The mixture of resin-aggregates was poured in a rectangular concrete mold (100 100 300 mm), and coated with a release agent that contained a gel coat and PVA to avoid adhesion of PC to steel mold. The mold containing the PC mixture was fixed on a vibrating table for 10 min for better packing. The samples were cured for 28 days at ambient temperature. In terms of resin volume fraction optimization, the changes were in the amount of the resin and aggregate proportions. PC samples were prepared with 17%, 15% and 13% of resin volume fraction. Aggregates have compatible proportions as resin proportions change accordingly.
3063
H. Haddad, M. Al Kobaisi / Composites: Part B 43 (2012) 3061–3068 Table 1 Thermo mechanical properties of all sizes of aggregates. Material Gravel Basalt Spodumene Sand Chalk Fly ash
qBulk (g/cm3) 1.63 1.6 1.56 1.71 1.35 1.33
qTrue (g/cm3) 2.61 2.77 2.879 2.63 2.69 2.71
Fig. 1. Particles size distributions for all aggregates.
3. Experimental The flexural strength of the PC samples was measured utilizing a four points test using Snitch 60/D universal according to Australian standard AS 1012.1. Coefficient of thermal expansion was measured using a custom built device. The device includes a heating chamber (Thumler Model number TH2700-26, Germany) with two displacement probes attached to a small digital display unit (SYLVAC DSOS, Swiss) and a thermostat connected to temperature control microprocessor. The PC sample temperature was obtained using a data acquisition system equipped with a computer as illustrated in Fig. 2. Two rods made of invar are used in this setup. Invar is a 36% nickel – iron alloy with the lowest CTE among metals alloys in 20–230 °C range, ai = 1.2 106 °C1. One of the rods is used as a reference and the other is placed above the PC sample. The expansion of the reference rod and the sample with the second rod is detected by the probes on the top of the heating chamber
Vp (%) 62 58 54 64 50 49
as (°C1) [14] 6
11.2 10 5.5 106 3 106 11 106 10 106 5.2 106
BET (m2/g) 0.13 0.22 0.11 0.29 1.61 1.02
which touch the invar rods. Fig. 2 shows the arrangement inside the heating chamber. The PC sample has a hole in the top center in order to place the thermocouple sensor (SE00 type K thermocouple, Pico technology, UK) to monitor and control the temperature inside the sample. The Pico data acquisition system contains TC08 thermocouple data logger (Pico technology, UK) connected to the computer to display the temperature. Due to the low thermal conductivity of polymer concrete, approximately 1 h is required to achieve thermal equilibrium. Thermal expansion was measured at five temperatures; 25, 35, 40, 50 and 60 °C. Some temperatures are beyond the temperature range of the base during the operational condition. Extra precautions were taken to cover the selected temperatures to this extent. The CTE of each PC sample was calculated using the following Eq. (8), derived according to the physical principle of thermal expansion as follows:
DL ¼ ai lr DT þ as ls DT DR ¼ a i l r DT þ a i l s DT DS ¼ DL DR DS as ¼ þ ai ls DT
ð5Þ ð6Þ ð7Þ ð8Þ
where DL is the change of the length for the Invar rod and the sample within the temperature differences DT, and DR is the change in length of the reference Invar rod within the same temperature difference DT. DS is the difference between the readings of the two probes, i.e. the difference between DL and DR. lr is the length of the Invar rod, which is located on top of the PC sample, and ls is the length of the PC sample. Substituting Eqs. (5) and (6) in Eq. (7), and using simple algebra simplifications, results in Eq. (8), to be used in the calculation of CTE polymer concrete as for the sample.
Fig. 2. Custom build device for measuring CTE of PC sample.
3064
H. Haddad, M. Al Kobaisi / Composites: Part B 43 (2012) 3061–3068
Fig. 3. SEM of failed PC sample through the interfacial bonding.
4. Results and discussion The aggregate surface morphology, particle size distribution and resin volume fraction are synergic and fatalistic parameters in governing the thickness and shape of interfacial adhesion bond-
ing between particles. The main driver of the mechanical properties for PC, including CET and flexural strength, is the interfacial adhesion bonding between particles. The thermal properties of particles and aggregate proportions effect the overall polymer concrete CTE [16]. Fig. 3 exhibits the PC sample demonstrating the failure mechanism that went through the interfacial adhesion bonding between filler particles, i.e. a polymeric matrix. The first suggested optimum composition for a precision machine base application is basalt, spodumene, and fly ash, as shown in Table 2. The flexural strength was 22.85 MPa, due to the fact that crushed basalt increases the flexural strength of concrete [17]. Fly ash also increases the flexural strength of polymer concrete [18]. The basalt surface contains rough textures (rough surface) due to the crushing process. Spodumene, which comes from lithium mining, has textures attributable to the mining process. Fig. 4A shows the spodumene particles. The microstructure of spodumene is in the form of crystallite layers [19]. The layers may partially separate to some extent at the beginning and the end of the layer, yet differently for each layer in the spodumene particle. This situation allows the mixture of fly ash resin to flow in between the spodumene layers. This condition results in higher adhesion bonding between the polymeric matrix and the aggregates. In addition fly ash particles have a spherical shape, a smooth surface and a micro size, as shown in Figs. 1 and 4C. This leads to low viscosity at the time of mixing fly ash with the resin compared to the chalk, as illustrated in Fig. 5. This indicates the high flow ability and filling capability of
Table 2 CTE and flextural strength for various PC compositions.
1 2 3 4 5 6 7 8
Compositions
Flexural strength (MPa)
Deflection (mm)
CTE (°C1)
Basalt, spodumene, fly ash Basalt, sand, fly ash Basalt, spodumene, chalk Basalt, sand, chalk Gravel, sand, fly ash Gravel, spodumene, chalk Gravel, spodumene, fly ash Gravel, sand, chalk
22.85 21.90 21.10 20.05 18.92 18.22 16.27 16.92
1.458 1.05 1.007 1.00 0.94 0.90 0.97 0.92
10.0 106 14.9 106 10.3 106 16.1 106 14.6 106 12.6 106 12.5 106 18.8 106
Fig. 4. SEM (A) spodumene, (B) sand, (C) fly ash, and (D) chalk.
H. Haddad, M. Al Kobaisi / Composites: Part B 43 (2012) 3061–3068
3065
Fig. 5. Mixture viscosity resin fly ash (ow), chalk (high) versus curing time. Fig. 7. Flexural strength of PC versus resin amount.
Fig. 8. Temperature versus time for CNC grinder tool machine.
Fig. 6. CTE of PC composite versus resin amount.
fly ash. The resin-fly ash mixture provides a better reach for the microspore cavities. The basalt texture and the in-between layers of spodumene facilitate the formation of a highly adhesive interfacial bonding between the polymeric matrix and the aggregate. This is a possible explanation for the high flexural strength for the first composition. The first composition scores the lowest CTE 10.0 106 °C1 as all the aggregates have a low CTE [16], and the lowest value is for the spodumene as illustrated in Table 1. The flexural strength of the second composition (basalt, sand and fly ash) is 21.40 MPa. Sand has a higher BET active surface area than spodumene, as shown in Table 1, and less textural roughness on the surface, as shown in Fig. 4B. In addition, there are no layers in the microstructure of the sand particles to give extra grip, which may be the reason for the slightly reduced flexural strength. The second composite has a higher CTE, by approximately 40%, than the first one. This is due to the sand CTE effect, which is almost three times more than spodumene CTE and two times more than fly ash CTE, as shown in Table 1. The main reason for the reduced flexural strength in the third composition (basalt, spodumene and chalk) is that the chalk particles have a more active surface area than fly ash, as shown in Table 1. This indicates that more resin is required for efficient wetting. Added to that, the chalk resin mixture provides less flow ability and filling capability than the fly ash resin mixture, as shown in Fig. 5. The rationale behind this is the chalk particle shape is irregular and has several sharp corners
(rough surface), as shown in Fig. 4D. This results in a higher viscosity during the mixing, which leads to the formation of a lower adhesive bonding mechanism on the surface of the aggregates compared to fly ash. This level of gripping adhesion leads to a reduction of the flexural strength. The CTE for the third composition was close to the first composition due to spodumene participation, as shown in Table 2. The fourth composition (basalt, sand and chalk) suffered because of the chalk and sand that both share responsibility for low flexural strength and high CTE for the reasons explained previously. Replacing basalt with river gravel in the fifth composition resulted in a reduction in flexural strength, since the crushed basalt increased the flexural strength of the concrete [17]. River gravel has less BET surface area and a smoother surface compared to basalt, which may lead to less adhesion and a reduction in flexural strength. River gravel increased the CTE of the fifth composition, as shown in Table 2. It has a higher CTE (river gravel and sand) than other aggregates, as shown in Table 1. The presence of river gravel, chalk and sand in PC composite system reduced the flexural strength and increased the CTE at various levels for different compositions. This condition was due to the CTE of the particles as well as their morphology and may have resulted in an interfacial bonding that has less adhesion between particles than basalt, spodumene and fly ash. The second composition was nominated for further optimization in terms of resin volume fraction. This step was due to its ability to accommodate a resin volume fraction of less than 17%. The logic behind this is that sand has high packing volume due to its rounded particle shape. In addition, the particle shape of fly ash is spherical. Fig. 6 shows the effect of a
3066
H. Haddad, M. Al Kobaisi / Composites: Part B 43 (2012) 3061–3068
Fig. 9. (A) First composition and (B) last composition.
reduction in the resin volume fraction on the CTE of polymer concrete. Decreasing the resin volume fraction results in a decrease of the CTE of PC since the resin has the highest CTE (80 106 °C1) in PC components. PC composite follows the rule of the mixture regarding composite mechanical properties. Reducing the resin by 4% reduced the CTE of the polymer concrete by approximately 31%. The other side effect of resin reduction is flexural strength. Fig. 7 demonstrates that a reduction in the amount of resin by 3% reduced the flexural strength by approximately 40%. Since the amount of resin that fulfilled the interfacial adhesion bonding in between particles, reduced [20]. The lowest flexural strength was 15.7 MPa for 13%, which is still acceptable in building the base. The base cross section is sufficient to cope with the relatively low flexural stress, considering the small machine weight in comparison to the base size.
5. Results validation It is essential to categorize the optimization of composition based on CTE and flexural strength since both properties are key controls for the thermal and mechanical stability of a precision tool machine base in operational conditions. To have a realistic visualized understanding of filler compositions and resin volume fraction reflections on the base, Finite Element Analysis (FEA) was conducted by using CAE software ANSYS 12.1 to reveal the effects of optimized properties on the functional performance of the base within the boundary of an industrial thermal load. Fig. 8 shows temperature with time for the inlet coolant, outlet coolant and environmental temperature during operational conditions. The framed area in Fig. 8 has been selected as the boundary conditions for the environmental temperature, inlet coolant temperature and
3067
H. Haddad, M. Al Kobaisi / Composites: Part B 43 (2012) 3061–3068
Fig. 10. Basalt, sand and fly ash (A) 15% resin and (B) 13% resin.
Table 3 Effect PC composition and resin amount on rail deflection. Compositions
Resin volume fraction (%)
Maximum deflection on the base (lm)
Variation in deformation on the rails (lm)
Variation in deformation between two sides of rails (lm)
Basalt, spodumene, fly ash Gravel, sand, chalk Basalt, sand, fly ash Basalt, sand, fly ash
17 17 15 13
25 45.5 30 14.6
2 3–5 0.43–1.6 0.13–0.31
0.1–0.5 1–3 0.2–0.34 0.011–0.06
outlet coolant temperature, which were all to be applied at the peak in the simulation. Fig. 9A illustrates the directional structural deformation of the base for the first composition (basalt, spodumene, fly ash). The most effective part of the base is the rails that
hold the moving components and drive the operational function in CNC grinder tool machine base. Maximum deformation is 25 lm on the base, concentrated on an ineffective, small portion of the base located on the walls that is approximately 2% of the base.
3068
H. Haddad, M. Al Kobaisi / Composites: Part B 43 (2012) 3061–3068
Variations in deformation on the rails on one side is approximately 2 lm. Variations in deformation between the two sides of the rails are approximately 0.1–0.5 lm. The last composition in ranking river gravel, sand and chalk has almost double the thermal expansion of the first composition. Fig. 9B exhibits the directional deformation on the base. Maximum deformation is 45.5 lm and 75% of the upper half of the base that has been deformed nearest to the maximum. Variations in deformation on rails are 3–5 lm. Variations in deformation between the two sides of the rails are approximately 1–3 lm. Fig. 10A shows the directional structural deformation of the base for the second composition (basalt, sand, fly ash, resin volume fraction is 15%) and the maximum deformation in Z direction is 30.2 lm. Rail variations on one side are 1.5– 2 lm. Variations in deformation between two sides of the rails are 0.2–0.34 lm. Fig. 10B exhibited the directional structural deformation of the base for the composition (basalt, sand and fly ash) with a 4% reduction in resin volume fraction. Maximum deformation in Z direction for the base is 18 lm. Rail variations are 0.13–0.34 lm. Variations in deformation between two sides of the rails are 0.011–0.06 lm. These variations are very small, and hence can be ignored since they are less than 1 lm for both one side and two sides of rails variations in deformation. Table 3 illustrates all measures in one domain for a clear comparison. 6. Conclusions The utilization of aggregate morphology, thermal and mechanical properties of particles resulted in the optimum polymer concrete being obtained for the base of a precision machine. In conclusions: The optimum composition is basalt, sand and fly ash (87% filler and 13% resin), which has the least CTE and acceptable flexural strength. Reduction in resin has diminished the negative effect of sand in relation to the final CTE composition and the resulting flexural strength was acceptable for the application. When any aggregate of the optimum composition is replaced with another aggregate (gravel, sand and chalk) it will result in the composition suffering a further reduction in flexural strength and an increase in CTE. The main cause for this is that particle properties reflect on interfacial adhesion bonding behavior.
The optimum composition is extremely cost effective compared to the first composition. Evidently sand is 90% less expensive than spodumene and 4% less resin, that the most expensive component in PC polymer concrete composite system. The optimum composition has reduced the variations in deformation of rail bases to a sufficiently lower level for it to be ignored. This induces further enhancement for precision tool machine accuracy. This enhancement of the operational conditions for precision tool machining accelerates the level of precision to a higher peak in achieving precise products.
References [1] ACI, Guide for use of polymers in concrete. Detroit: American Concrete Institute; 1986 [2] Ferreira AJM, Ribeiro MCS, Marques AT. Analysis of hybrid beams composed of GFRP profiles and polymer concrete. Int J Mech Mater Des 2004;1(2):143–55. [3] Vipulanandan C, Paul E. Characterisation of polyester polymer and polymer concrete. J Mater Civil Eng 1993;5(1). [4] Yuasa N, et al. Effects of moisture content and pore structure of subset concrete on adhesive strength of epoxy coating. In: Ohama, editor. Polymers in concretes. London: E&FN SPON; 1997. [5] Suh J, Lee D. Design and manufacture of hybrid polymer concrete bed for highspeed CNC milling machine. Int J Mech Mater Des 2008;4(2):113–21. [6] Hojo H et al. Corrosion behaviour of epoxy and unsaturated polyester resins in alkaline solution. Polym Mater Corros Contr 1986:314–26. [7] Bruni C et al. Hard turning of an alloy steel on a machine tool with a polymer concrete bed. J Mater Process Technol 2008;202(1–3):493–9. [8] Sezan O. Investigation of vibration damping on polymer concrete with polyester resin. Cem Concr Res 2000;30(2):171–4. [9] Haque E, Armeniades CD. Montmorillonite polymer concrete: zero shrinkage and expanding polymer concrete with enhanced strength; 1985. [10] Khristova Y, Aniskevich K. Prediction of creep of polymer concrete. Mech Compos Mater 1995;31(3):216–9. [11] Valore RC et al. Resin bound aggregate material systems. In: International congress on polymer concretes, London; 1975. [12] Kim HS, Park KY, Lee DG. A study on the epoxy resin concrete for the ultraprecision machine tool bed. J Mater Process Technol 1995;48(1–4):649–55. [13] Ohama Y et al. Effect of coarse aggregate on compressive strength of polyester resin concrete. Int J Cem Compos 1979;3(1):111–5. [14] Katz HS, Milewski JV. Handbook of fillers for plastics. New York: Van Nostrand Reinhold; 1987. [15] Furnas CC. Grading aggregates. Ind Eng Chem 1931(23):1053–9. [16] Xu ZR et al. Effect of particle size on the thermal expansion of TiCAl XD™ composites. Script Metall Mater 1994;31(11):1525–30. [17] Kılıç A et al. The influence of aggregate type on the strength and abrasion resistance of high strength concrete. Cem Concr Compos 2008;30(4):290–6. [18] Rebeiz KS, Craft AP. Polymer concrete using coal fly ash. J Energy Eng 2002;128(3):62–73. [19] Graham J. Mineralogical notes on a-spodumene. Am Mineral 1975;60:919–23. [20] Knab LI, Flexural behaviour of conventionally reinforced polyester concrete beams. University of Cincinnati; 1972.
Contents lists available at SciVerse ScienceDirect
Composites: Part B journal homepage: www.elsevier.com/locate/compositesb
Optimization of the polymer concrete used for manufacturing bases for precision tool machines Header Haddad a,⇑, Mohammad Al Kobaisi b a b
Faculty of Engineering and Industrial Sciences, Swinburne University of Technology, Melbourne, VIC 3122, Australia School of Applied Sciences, RMIT University, Melbourne, VIC 3000, Australia
a r t i c l e
i n f o
Article history: Received 19 September 2011 Received in revised form 3 April 2012 Accepted 4 May 2012 Available online 18 May 2012 Keywords: A. Thermosetting resin A. Polymer–matrix composites (PMCs) B. Adhesion C. Finite element analysis (FEA)
a b s t r a c t Due to its superior damping ratio, high adhesion and fast curing, polymer concrete is used in manufacturing bases for a wide range of precision machines. The coefficient of thermal expansion for polymer concrete is one of the main parameters that can affect the level of accuracy in precision tool machines. Flexural strength is a fundamental strength of the base. In this study six aggregates (basalt, spodumene, fly ash, river gravel, sand and chalk) were investigated. Polymer concrete samples were prepared with different compositions of aggregates containing the same resin volume fraction (aggregates 83% and risen 17%). A four points flexural test was employed to measure the flexural strength of the polymer concrete samples. The coefficient of thermal expansion for polymer concrete was measured using a custom built device. The preliminary optimum composition, with the highest flexural strength and lowest thermal expansion coefficient, was found to be basalt, spodumene and fly ash. Basalt, sand and fly ash composition was the second in the rank. The second composition was nominated for further optimization in terms of resin volume fraction in consideration of its ability to adapt a smaller amount of resin. Different samples of polymer concrete were prepared with a variety of resin volume fractions as follows; 17%, 15% and 13%. The resin volume fraction has been demonstrated to have a significant effect on the coefficient of thermal expansion and flexural strength for polymer concrete. The final optimized composition was basalt, sand and fly ash (filler 87% and resin 13%). ANSYS 13 software was employed in visualizing the influence of polymer concrete compositions on the thermal expansion of the base and how it affected the level of precision of the tool machine. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction Polymer concrete (PC) is a composite material consisting of well-graded inorganic aggregates bonded together by a resin instead of the water and cement binder typically used in traditional cement concretes [1]. Unsaturated polyester resin (UPR) is one of the binders used in polymer concretes providing excellent adhesion to solid particulate fillers, good stiffness, dimensional stability and a high damping ratio [2]. UPR is a thermosetting material created by combining two commonly used polymerizing constituents, namely unsaturated polyester (UP) and a vinyl based monomer such as methyl methacrylate (MMA) [3]. This combination of inorganic fillers and a polymeric matrix allows for partial control of the mechanical and thermal properties of the produced composite system. PC can be tailor-made for specific applications, based on the conditions and functional requirements. PC is used for a variety of applications, such as industrial flooring [4], underground piping, ⇑ Corresponding author. E-mail addresses: [email protected] (H. Haddad), mohammad.alkobaisi@ rmit.edu.au (M. Al Kobaisi). 1359-8368/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.compositesb.2012.05.003
frame supports for MRI, X-ray, and CT-Scanners, for CNC manufacturing centers [5] and for CNC grinding tool machines. This variety of applications reflects the diversity of PC properties, such as good mechanical strength, high damping ratio, fast curing, good adhesive strength, long term durability, low permeability of water and chemical resistance [6]. The level of accuracy for precision machines is governed by the machine base-frame upon which the positioning and drive subsystems are installed. In a precision machine, the base-frame ensures the positioning of subsystems under the complex dynamic and static conditions. In these operational conditions, forces, torques and heat gradients are dynamically produced throughout the system to satisfy complex product geometry and elevated levels of tool precision requirements [7]. A milling tool is a good example to demonstrate the effect of the level of tool precision on product accuracy through the manufacturing sequence. Once a milling tool is produced it will then be used in a CNC milling machine to manufacture other products. Some of these products can be considered as tools for a particular application, such as micro injection molding that can be used, for producing the micro gears used in a micro medical device. Each process in the manufacturing sequence of the milling tool will have a manu-
3062
H. Haddad, M. Al Kobaisi / Composites: Part B 43 (2012) 3061–3068
facturing influential effect on the precision of the micro injection mold (tool). Based on this statement, the accuracy and functionality of the end product (micro plastic gears) will be affected by the post process levels of accuracy. In this case micro plastic gears inherit the level of accuracy from the original tool (milling tool) through the process sequence of manufacturing the micro mold. This depends greatly on the original milling tool produced by the CNC grinding tool machine and the milling machine used for the milling tool. The level of accuracy of both machines determines the precision level for the rest of the products being dealt with, and hence, the importance of polymer concrete base for both the CNC grinding machine tool and the milling machine [5]. It has a tremendous effect on the precision of the milling tool by damping the unwanted vibration created by servo motors during the manufacturing operation [8]. PC suffers from some drawbacks such as shrinkage [9] and low creep resistance [10]. The biggest drawback on operational effectiveness is the coefficient of thermal expansion (CTE) of the PC used in precision machinery applications. According to Valore, CTE is high for PC containing UPR as a binder [11] in comparison to cast iron and other metal inserts in the base. A complex combination of metal inserts and connecting plastic pipes for fluids are essential in the PC base of CNC grinding tool machines. This condition leads to a non-uniform distribution of thermal expansion for the base during operation. The non-uniform thermal expansion distribution on the base may reduce the target accuracy of the precision tool machine. Achieving a CTE for the PC as close as possible to the CNC grinding machine metal components (inserts) could enhance the distribution of thermal expansion for the base. It has been reported that by increasing the resin volumetric ratio in the PC, leads to an increase in the CTE of the PC and other mechanical properties, such as compression strength and thermal conductivity [12]. The properties of a PC composite are also determined by the properties of the aggregates, such as particle shape and size distribution, particle surface properties and aggregate volume fractions. The compressive strength of PC is elevated with an increase in the compressive strength of the course aggregate and decreased linearly with an increase in its volume fraction [13]. The main purpose of this study is to optimize the aggregate composition with a sufficient amount of resin for the PC to reduce the CTE and increase the flexural strength, which will result in an improvement in the level of accuracy for the precision machine base. This will have a positive impact on the precision level of the end product. The precision level of end product depends on the accuracy the tools and machine that it dealt with, through manufacturing sequence.
2. Materials and method Commercial general purpose, unsaturated polyester (AROPOL) (67% unsaturated polyester dissolved in 33% styrene) was obtained from Huntsman Chemical Company (Australia), methyl methacrylate (MMA) from Degussa (Australia), cobalt octoate and Dimethyl aniline (DMA) from Alfa Aesar (USA), and methyl ethyl kenton peroxide (MEKP) – the commercial name is NR20 from Nuplex Industries (Australia). A resin sample was produced by mixing a 3:2 volumetric ratio of UP to MMA. To this mixture, 0.8% cobalt octoate (promoter), 0.2% DMA (accelerator) and 2% (v/v) MEKP (initiator) were added and mixed in subsequent order. The basalt, river gravel, sand, and chalk used as aggregates were from Roca. Fly ash was from Cement Independent Australia and spodumene from Talison Minerals, Western Australia. The aggregates true density qTrue (g/ cm3) was measured using a pycnometer, their bulk density qbulk (g/cm3) was measured using a measuring cylinder, and the packing factor (Vp) was calculated based on the ratio of the obtained true and bulk densities. Table 1 summarizes the measured properties
of the aggregates and referenced aggregate properties, such as CTE for each filler (as°C1) [14]. The fine aggregate (chalk, fly ash) particle size distributions were obtained using Mastersizer X (Malvern Instruments, USA). Middle and course aggregate size distributions were obtained using sieve analysis. Fig. 1 shows the particle size distributions for all aggregates. The BET (Brunauer, Emmett and Teller, 1938) active surface area of the aggregates was determined by nitrogen gas adsorption/desorption. Nitrogen sorption measurements were performed using Micromeritics ASAP 2010 (USA). The adsorption and desorption isotherms were recorded using a 89-point pressure table with 15 s equilibration intervals. The packing system of the aggregates was determined according to the Furnace method [15], starting with the number of aggregates based upon the ratio of the biggest particle diameter and the smallest particle diameter. The diameter ratio was calculated using data from Fig. 1. Volume of voids, which is equal to 100 – Vp (%), was calculated using Vp of aggregates in Table 1. By applying the diameter ratio and volume of voids in a special graph provided by the Furnace method [15], the number of aggregates generations was obtained. Both the diameter ratio and the void volume were too small, leading to three aggregate generations. According to the Furnas method, the diameter of intermediate generation D2 was calculated using the following expression:
D2 ¼
pffiffiffiffiffiffiffiffiffiffiffi D1 D3
ð1Þ
where D1 and D3 are the biggest and the smallest generations of particle diameters. Total absolute volume of fillers V 0 was calculated using the following formula:
V0 ¼
1 V1 V2 þ þ 1 þ V1 1 þ V2 1 þ V3
ð2Þ
where V1, V2, V3 are volumes of voids for the biggest, intermediate and the smallest generations. Relative volume of voids can be calculated using 1 Vp. The total volume fraction occupied by the solid particles V f was calculated using the following formula:
V f ¼ V 0 ð1 V 21 Þ
ð3Þ
The volume fraction of each aggregates component Pi was calculated using the below formula:
Pi ¼ V p =V f
ð4Þ
This approach takes into account volume of voids and provides a good indication of possible filler composition. However, it does not take into account particle shapes and can be used mainly for spherical bodies. To prepare the PC mixture fine aggregate 8.3%, coarse aggregate 49.8% and middle size aggregate 24.9% (v/v) was added to 17% (v/v) resin in subsequent order and mixed using traditional concrete mixer. All aggregate proportions were obtained using the Furnace method. The mixing continued for 15 min to ensure the best wetting of all particles with resin, in order to maximize the resin interface with the pebbles and obtain sufficient adhesion. The mixture of resin-aggregates was poured in a rectangular concrete mold (100 100 300 mm), and coated with a release agent that contained a gel coat and PVA to avoid adhesion of PC to steel mold. The mold containing the PC mixture was fixed on a vibrating table for 10 min for better packing. The samples were cured for 28 days at ambient temperature. In terms of resin volume fraction optimization, the changes were in the amount of the resin and aggregate proportions. PC samples were prepared with 17%, 15% and 13% of resin volume fraction. Aggregates have compatible proportions as resin proportions change accordingly.
3063
H. Haddad, M. Al Kobaisi / Composites: Part B 43 (2012) 3061–3068 Table 1 Thermo mechanical properties of all sizes of aggregates. Material Gravel Basalt Spodumene Sand Chalk Fly ash
qBulk (g/cm3) 1.63 1.6 1.56 1.71 1.35 1.33
qTrue (g/cm3) 2.61 2.77 2.879 2.63 2.69 2.71
Fig. 1. Particles size distributions for all aggregates.
3. Experimental The flexural strength of the PC samples was measured utilizing a four points test using Snitch 60/D universal according to Australian standard AS 1012.1. Coefficient of thermal expansion was measured using a custom built device. The device includes a heating chamber (Thumler Model number TH2700-26, Germany) with two displacement probes attached to a small digital display unit (SYLVAC DSOS, Swiss) and a thermostat connected to temperature control microprocessor. The PC sample temperature was obtained using a data acquisition system equipped with a computer as illustrated in Fig. 2. Two rods made of invar are used in this setup. Invar is a 36% nickel – iron alloy with the lowest CTE among metals alloys in 20–230 °C range, ai = 1.2 106 °C1. One of the rods is used as a reference and the other is placed above the PC sample. The expansion of the reference rod and the sample with the second rod is detected by the probes on the top of the heating chamber
Vp (%) 62 58 54 64 50 49
as (°C1) [14] 6
11.2 10 5.5 106 3 106 11 106 10 106 5.2 106
BET (m2/g) 0.13 0.22 0.11 0.29 1.61 1.02
which touch the invar rods. Fig. 2 shows the arrangement inside the heating chamber. The PC sample has a hole in the top center in order to place the thermocouple sensor (SE00 type K thermocouple, Pico technology, UK) to monitor and control the temperature inside the sample. The Pico data acquisition system contains TC08 thermocouple data logger (Pico technology, UK) connected to the computer to display the temperature. Due to the low thermal conductivity of polymer concrete, approximately 1 h is required to achieve thermal equilibrium. Thermal expansion was measured at five temperatures; 25, 35, 40, 50 and 60 °C. Some temperatures are beyond the temperature range of the base during the operational condition. Extra precautions were taken to cover the selected temperatures to this extent. The CTE of each PC sample was calculated using the following Eq. (8), derived according to the physical principle of thermal expansion as follows:
DL ¼ ai lr DT þ as ls DT DR ¼ a i l r DT þ a i l s DT DS ¼ DL DR DS as ¼ þ ai ls DT
ð5Þ ð6Þ ð7Þ ð8Þ
where DL is the change of the length for the Invar rod and the sample within the temperature differences DT, and DR is the change in length of the reference Invar rod within the same temperature difference DT. DS is the difference between the readings of the two probes, i.e. the difference between DL and DR. lr is the length of the Invar rod, which is located on top of the PC sample, and ls is the length of the PC sample. Substituting Eqs. (5) and (6) in Eq. (7), and using simple algebra simplifications, results in Eq. (8), to be used in the calculation of CTE polymer concrete as for the sample.
Fig. 2. Custom build device for measuring CTE of PC sample.
3064
H. Haddad, M. Al Kobaisi / Composites: Part B 43 (2012) 3061–3068
Fig. 3. SEM of failed PC sample through the interfacial bonding.
4. Results and discussion The aggregate surface morphology, particle size distribution and resin volume fraction are synergic and fatalistic parameters in governing the thickness and shape of interfacial adhesion bond-
ing between particles. The main driver of the mechanical properties for PC, including CET and flexural strength, is the interfacial adhesion bonding between particles. The thermal properties of particles and aggregate proportions effect the overall polymer concrete CTE [16]. Fig. 3 exhibits the PC sample demonstrating the failure mechanism that went through the interfacial adhesion bonding between filler particles, i.e. a polymeric matrix. The first suggested optimum composition for a precision machine base application is basalt, spodumene, and fly ash, as shown in Table 2. The flexural strength was 22.85 MPa, due to the fact that crushed basalt increases the flexural strength of concrete [17]. Fly ash also increases the flexural strength of polymer concrete [18]. The basalt surface contains rough textures (rough surface) due to the crushing process. Spodumene, which comes from lithium mining, has textures attributable to the mining process. Fig. 4A shows the spodumene particles. The microstructure of spodumene is in the form of crystallite layers [19]. The layers may partially separate to some extent at the beginning and the end of the layer, yet differently for each layer in the spodumene particle. This situation allows the mixture of fly ash resin to flow in between the spodumene layers. This condition results in higher adhesion bonding between the polymeric matrix and the aggregates. In addition fly ash particles have a spherical shape, a smooth surface and a micro size, as shown in Figs. 1 and 4C. This leads to low viscosity at the time of mixing fly ash with the resin compared to the chalk, as illustrated in Fig. 5. This indicates the high flow ability and filling capability of
Table 2 CTE and flextural strength for various PC compositions.
1 2 3 4 5 6 7 8
Compositions
Flexural strength (MPa)
Deflection (mm)
CTE (°C1)
Basalt, spodumene, fly ash Basalt, sand, fly ash Basalt, spodumene, chalk Basalt, sand, chalk Gravel, sand, fly ash Gravel, spodumene, chalk Gravel, spodumene, fly ash Gravel, sand, chalk
22.85 21.90 21.10 20.05 18.92 18.22 16.27 16.92
1.458 1.05 1.007 1.00 0.94 0.90 0.97 0.92
10.0 106 14.9 106 10.3 106 16.1 106 14.6 106 12.6 106 12.5 106 18.8 106
Fig. 4. SEM (A) spodumene, (B) sand, (C) fly ash, and (D) chalk.
H. Haddad, M. Al Kobaisi / Composites: Part B 43 (2012) 3061–3068
3065
Fig. 5. Mixture viscosity resin fly ash (ow), chalk (high) versus curing time. Fig. 7. Flexural strength of PC versus resin amount.
Fig. 8. Temperature versus time for CNC grinder tool machine.
Fig. 6. CTE of PC composite versus resin amount.
fly ash. The resin-fly ash mixture provides a better reach for the microspore cavities. The basalt texture and the in-between layers of spodumene facilitate the formation of a highly adhesive interfacial bonding between the polymeric matrix and the aggregate. This is a possible explanation for the high flexural strength for the first composition. The first composition scores the lowest CTE 10.0 106 °C1 as all the aggregates have a low CTE [16], and the lowest value is for the spodumene as illustrated in Table 1. The flexural strength of the second composition (basalt, sand and fly ash) is 21.40 MPa. Sand has a higher BET active surface area than spodumene, as shown in Table 1, and less textural roughness on the surface, as shown in Fig. 4B. In addition, there are no layers in the microstructure of the sand particles to give extra grip, which may be the reason for the slightly reduced flexural strength. The second composite has a higher CTE, by approximately 40%, than the first one. This is due to the sand CTE effect, which is almost three times more than spodumene CTE and two times more than fly ash CTE, as shown in Table 1. The main reason for the reduced flexural strength in the third composition (basalt, spodumene and chalk) is that the chalk particles have a more active surface area than fly ash, as shown in Table 1. This indicates that more resin is required for efficient wetting. Added to that, the chalk resin mixture provides less flow ability and filling capability than the fly ash resin mixture, as shown in Fig. 5. The rationale behind this is the chalk particle shape is irregular and has several sharp corners
(rough surface), as shown in Fig. 4D. This results in a higher viscosity during the mixing, which leads to the formation of a lower adhesive bonding mechanism on the surface of the aggregates compared to fly ash. This level of gripping adhesion leads to a reduction of the flexural strength. The CTE for the third composition was close to the first composition due to spodumene participation, as shown in Table 2. The fourth composition (basalt, sand and chalk) suffered because of the chalk and sand that both share responsibility for low flexural strength and high CTE for the reasons explained previously. Replacing basalt with river gravel in the fifth composition resulted in a reduction in flexural strength, since the crushed basalt increased the flexural strength of the concrete [17]. River gravel has less BET surface area and a smoother surface compared to basalt, which may lead to less adhesion and a reduction in flexural strength. River gravel increased the CTE of the fifth composition, as shown in Table 2. It has a higher CTE (river gravel and sand) than other aggregates, as shown in Table 1. The presence of river gravel, chalk and sand in PC composite system reduced the flexural strength and increased the CTE at various levels for different compositions. This condition was due to the CTE of the particles as well as their morphology and may have resulted in an interfacial bonding that has less adhesion between particles than basalt, spodumene and fly ash. The second composition was nominated for further optimization in terms of resin volume fraction. This step was due to its ability to accommodate a resin volume fraction of less than 17%. The logic behind this is that sand has high packing volume due to its rounded particle shape. In addition, the particle shape of fly ash is spherical. Fig. 6 shows the effect of a
3066
H. Haddad, M. Al Kobaisi / Composites: Part B 43 (2012) 3061–3068
Fig. 9. (A) First composition and (B) last composition.
reduction in the resin volume fraction on the CTE of polymer concrete. Decreasing the resin volume fraction results in a decrease of the CTE of PC since the resin has the highest CTE (80 106 °C1) in PC components. PC composite follows the rule of the mixture regarding composite mechanical properties. Reducing the resin by 4% reduced the CTE of the polymer concrete by approximately 31%. The other side effect of resin reduction is flexural strength. Fig. 7 demonstrates that a reduction in the amount of resin by 3% reduced the flexural strength by approximately 40%. Since the amount of resin that fulfilled the interfacial adhesion bonding in between particles, reduced [20]. The lowest flexural strength was 15.7 MPa for 13%, which is still acceptable in building the base. The base cross section is sufficient to cope with the relatively low flexural stress, considering the small machine weight in comparison to the base size.
5. Results validation It is essential to categorize the optimization of composition based on CTE and flexural strength since both properties are key controls for the thermal and mechanical stability of a precision tool machine base in operational conditions. To have a realistic visualized understanding of filler compositions and resin volume fraction reflections on the base, Finite Element Analysis (FEA) was conducted by using CAE software ANSYS 12.1 to reveal the effects of optimized properties on the functional performance of the base within the boundary of an industrial thermal load. Fig. 8 shows temperature with time for the inlet coolant, outlet coolant and environmental temperature during operational conditions. The framed area in Fig. 8 has been selected as the boundary conditions for the environmental temperature, inlet coolant temperature and
3067
H. Haddad, M. Al Kobaisi / Composites: Part B 43 (2012) 3061–3068
Fig. 10. Basalt, sand and fly ash (A) 15% resin and (B) 13% resin.
Table 3 Effect PC composition and resin amount on rail deflection. Compositions
Resin volume fraction (%)
Maximum deflection on the base (lm)
Variation in deformation on the rails (lm)
Variation in deformation between two sides of rails (lm)
Basalt, spodumene, fly ash Gravel, sand, chalk Basalt, sand, fly ash Basalt, sand, fly ash
17 17 15 13
25 45.5 30 14.6
2 3–5 0.43–1.6 0.13–0.31
0.1–0.5 1–3 0.2–0.34 0.011–0.06
outlet coolant temperature, which were all to be applied at the peak in the simulation. Fig. 9A illustrates the directional structural deformation of the base for the first composition (basalt, spodumene, fly ash). The most effective part of the base is the rails that
hold the moving components and drive the operational function in CNC grinder tool machine base. Maximum deformation is 25 lm on the base, concentrated on an ineffective, small portion of the base located on the walls that is approximately 2% of the base.
3068
H. Haddad, M. Al Kobaisi / Composites: Part B 43 (2012) 3061–3068
Variations in deformation on the rails on one side is approximately 2 lm. Variations in deformation between the two sides of the rails are approximately 0.1–0.5 lm. The last composition in ranking river gravel, sand and chalk has almost double the thermal expansion of the first composition. Fig. 9B exhibits the directional deformation on the base. Maximum deformation is 45.5 lm and 75% of the upper half of the base that has been deformed nearest to the maximum. Variations in deformation on rails are 3–5 lm. Variations in deformation between the two sides of the rails are approximately 1–3 lm. Fig. 10A shows the directional structural deformation of the base for the second composition (basalt, sand, fly ash, resin volume fraction is 15%) and the maximum deformation in Z direction is 30.2 lm. Rail variations on one side are 1.5– 2 lm. Variations in deformation between two sides of the rails are 0.2–0.34 lm. Fig. 10B exhibited the directional structural deformation of the base for the composition (basalt, sand and fly ash) with a 4% reduction in resin volume fraction. Maximum deformation in Z direction for the base is 18 lm. Rail variations are 0.13–0.34 lm. Variations in deformation between two sides of the rails are 0.011–0.06 lm. These variations are very small, and hence can be ignored since they are less than 1 lm for both one side and two sides of rails variations in deformation. Table 3 illustrates all measures in one domain for a clear comparison. 6. Conclusions The utilization of aggregate morphology, thermal and mechanical properties of particles resulted in the optimum polymer concrete being obtained for the base of a precision machine. In conclusions: The optimum composition is basalt, sand and fly ash (87% filler and 13% resin), which has the least CTE and acceptable flexural strength. Reduction in resin has diminished the negative effect of sand in relation to the final CTE composition and the resulting flexural strength was acceptable for the application. When any aggregate of the optimum composition is replaced with another aggregate (gravel, sand and chalk) it will result in the composition suffering a further reduction in flexural strength and an increase in CTE. The main cause for this is that particle properties reflect on interfacial adhesion bonding behavior.
The optimum composition is extremely cost effective compared to the first composition. Evidently sand is 90% less expensive than spodumene and 4% less resin, that the most expensive component in PC polymer concrete composite system. The optimum composition has reduced the variations in deformation of rail bases to a sufficiently lower level for it to be ignored. This induces further enhancement for precision tool machine accuracy. This enhancement of the operational conditions for precision tool machining accelerates the level of precision to a higher peak in achieving precise products.
References [1] ACI, Guide for use of polymers in concrete. Detroit: American Concrete Institute; 1986 [2] Ferreira AJM, Ribeiro MCS, Marques AT. Analysis of hybrid beams composed of GFRP profiles and polymer concrete. Int J Mech Mater Des 2004;1(2):143–55. [3] Vipulanandan C, Paul E. Characterisation of polyester polymer and polymer concrete. J Mater Civil Eng 1993;5(1). [4] Yuasa N, et al. Effects of moisture content and pore structure of subset concrete on adhesive strength of epoxy coating. In: Ohama, editor. Polymers in concretes. London: E&FN SPON; 1997. [5] Suh J, Lee D. Design and manufacture of hybrid polymer concrete bed for highspeed CNC milling machine. Int J Mech Mater Des 2008;4(2):113–21. [6] Hojo H et al. Corrosion behaviour of epoxy and unsaturated polyester resins in alkaline solution. Polym Mater Corros Contr 1986:314–26. [7] Bruni C et al. Hard turning of an alloy steel on a machine tool with a polymer concrete bed. J Mater Process Technol 2008;202(1–3):493–9. [8] Sezan O. Investigation of vibration damping on polymer concrete with polyester resin. Cem Concr Res 2000;30(2):171–4. [9] Haque E, Armeniades CD. Montmorillonite polymer concrete: zero shrinkage and expanding polymer concrete with enhanced strength; 1985. [10] Khristova Y, Aniskevich K. Prediction of creep of polymer concrete. Mech Compos Mater 1995;31(3):216–9. [11] Valore RC et al. Resin bound aggregate material systems. In: International congress on polymer concretes, London; 1975. [12] Kim HS, Park KY, Lee DG. A study on the epoxy resin concrete for the ultraprecision machine tool bed. J Mater Process Technol 1995;48(1–4):649–55. [13] Ohama Y et al. Effect of coarse aggregate on compressive strength of polyester resin concrete. Int J Cem Compos 1979;3(1):111–5. [14] Katz HS, Milewski JV. Handbook of fillers for plastics. New York: Van Nostrand Reinhold; 1987. [15] Furnas CC. Grading aggregates. Ind Eng Chem 1931(23):1053–9. [16] Xu ZR et al. Effect of particle size on the thermal expansion of TiCAl XD™ composites. Script Metall Mater 1994;31(11):1525–30. [17] Kılıç A et al. The influence of aggregate type on the strength and abrasion resistance of high strength concrete. Cem Concr Compos 2008;30(4):290–6. [18] Rebeiz KS, Craft AP. Polymer concrete using coal fly ash. J Energy Eng 2002;128(3):62–73. [19] Graham J. Mineralogical notes on a-spodumene. Am Mineral 1975;60:919–23. [20] Knab LI, Flexural behaviour of conventionally reinforced polyester concrete beams. University of Cincinnati; 1972.