cement production - an overview | sciencedirect topics
Cement production has undergone tremendous developments since its beginnings some 2,000years ago. While the use of cement in concrete has a very long history (Malinowsky, 1991), the industrial production of cements started in the middle of the 19 th century, first with shaft kilns, which were later replaced by rotary kilns as standard equipment worldwide. Today s annual global cement production has reached 2.8 billion tonnes, and is expected to increase to some 4 billion tonnes per year in 2050 (Schneider et al., 2011). Major growth is foreseen in countries such as China and India as well as in regions like the Middle East and Northern Africa. At the same time, the cement industry is facing challenges such as cost increases in energy supply (Lund, 2007), requirements to reduce CO2 emissions, and the supply of raw materials in sufficient qualities and amounts (WBCSD, 2008).
In this chapter, the environmental impact of cement production will be described. The chapter will first focus on the most common cementitious product: ordinary Portland cement (OPC) and will then evaluate the main perspective in terms of reduction of cement production s environmental impacts. To do so, we will describe the improvement perspective in the cement sector as well as the alternative products that could, at least partially, replace OPC.
Cement production is a thermal energy intensive process, which requires heating solid particles up to 1450C and cooling it down. The process generates hot and CO2 rich exhaust streams. Fin order to study energy efficiency of the process, authors like Mujumdar et al. (2006, 2007) developed detail models of process units, while others highlighted energy and exergy performances of plants (Kolip et al 2010). Some authors have studied the integration of cogeneration units in cement industry as an option of waste heat recovery (Wang et al 2008). The previously mentioned studies are considering, in detail, thermodynamic performances, but they are not investigating the heat integration inside the process. Some other authors, such as Murray et al (2008), Kaantee (2004), Mokrzycki (2003) have focused their analysis in the use of alternative/waste fuels, but they did not show the impact and the potential of these fuels on the heat integration between streams.
This paper explores the use of process integration techniques to improve the energy efficiency of cement plants, focusing on the dry route cement production and the integration of alternative fuels. Flow sheeting modeling, Pinch Analysis and mixed integer linear optimization techniques are applied to study an existing cement production facility.
Cement production is one of the largest industries in the world. Annual world production in 2013 was approximately 4 GT (of which, about half was in China). It is produced in kilns at around 1400oC (2500oF), and approximately 750 kg (1650 lb) of CO2 are released for each tonne (2205 lb.) that is made. This is 400 m3 (524 yd3) of gas; and CO2 normally forms only 0.04% of air. Cement production is responsible for around 7% of worldwide greenhouse gas emissions. The only larger contributors are electricity generation and transport.
Most of the CO2 released when cement is made comes from the chemical reaction. The limestone from which it is made is calcium carbonate, and when this is heated it gives off carbon dioxide, and forms calcium oxide. Thus, while some benefit has been obtained by making the process more energy efficient, there is little scope for any further reductions.
Numerous different solutions to this problem have been proposed, including carbon capture and inventing completely new cements; but the only effective method currently used to achieve significant reductions in greenhouse gas emissions is the use of cement replacements.
One process that has not been extensively investigated is carbon sequestration. In this process, the hydrated cement reacts with CO2 in the air, slowly reversing some of the processes that took place in the kiln when the cement was made (this is the carbonation process; it also causes reinforcement corrosion, see Section 25.3.2). It is estimated that this may reduce the carbon footprint of the cement industry by 35%.
Cement production is an energy-intensive process. The cost of energy constitutes more than 60% of the cost of the cement; hence cement plants have to consider minimizing the cost of energy when planning production. However, there are several challenging issues regarding the production plan. First, there are some operational constraints for the production process itself, such as keeping a proper level of inventory. Second, electricity prices are not fixed; they change on an hourly basis. Hence the cost of the electricity consumed may change substantially depending on the time of production. In this study, we developed a mixed integer programming model to minimize the cost of production (including energy, labor, and storage) by shifting working times to hours with low energy costs and trying to maintain activities during hours with high energy costs.
Cement production processes can be categorized as dry, semidry, semiwet, and wet processes depending on the handling of raw material before being fed to the rotary kiln. Nowadays, almost all new plants are based on the dry process and many old wet plants are also remodeled to dry or semidry processes. Dry cement manufacture has three fundamental stages: preparation of feedstocks, production of clinker, and preparation of cement [15,16].
Preparation of feedstock. This stage includes the process of siege, crushing, and prehomogenization. Typical raw materials used for cement production have 85% cayenne, 13% clay or blackboard, and under 1% each of materials such as silica, alumina, and iron ore. These feedstocks are crushed into particles with a diameter of less than 20mm and mixed with a prehomogenization pile .
Clinker production. As shown in Fig.6.6, the feedstock first enters the raw material to make a fine powder (raw feed), where 85% of the material is slightly smaller than 88m. Then, the feed is transferred to the homogenizing silos to impair the material difference. After that,the meal is thrown into the precalciner tower to start the chemical change to cement. Precyclone towers intermix with raw feed and almost 1000C of exhaust gas to recover energy, preheat feed, and initiate chemical reactions that result in cement. Gases straight from the kiln, but in precalciner facilities, gasoline, and air are provided by a combustion vessel inside the tower and kiln. Typically, 60% of burned calciner and more than 90% calcination have been reached before the material enters the rotary kiln. Inside the kiln, temperatures reach approximately 1400C to complete the process of chemical reactions and produce calcium silicates, called clinkers, with a diameter of 1025mm. The gas from the preheater tower is usually blended in a rawmill, which will help stabilize the future feedstock. After flowing through the rawmill, the gases are eventually released by a dust collector, which also obtains good particles when feedstuffs are milled. The dust will then be recycled into homogenizing silos and served as part of the kiln feed.
Preparation of cement. This stage completes the manufacturing process where clinker nodules are milled into cement. Following clinker milling, the cement is ready for use as a binder in various concrete mixes.
The cement production process, for example, starts with mining of limestone, which is then crushed and ground to powder. It is then preheated to save energy before being transferred to the kiln, the heart of the process. The kiln is then heated to a high temperature of up to 1480 degrees to convert the material to a molten form called clinker. The clinker is then cooled and ground to a fine powder with other additives and transferred to storage silos for bagging or bulk transportation (Portland Cement Association, 2014). The production of cement is either through the wet or dry process with the dry process as the preferred option because of the lower energy intensity. Cement production accounts for about 5% of total anthropogenic emissions (IFC, 2017). Cement-based structures constitute the largest surface area of all man-made structures (Odigure, 2009). World cement demand was about 2.283 billion tons in 2005, 2035 million tons in 2007, and 2836 million tons in 2010 with an annual estimated increase of about 130 million tons (Madlool, Saidur, Hossaina & Rahim, 2011; Odigure, 2009). World total cement production for 2016 was about 4.2 billion tons with emerging markets playing a dominant role (IFC, 2017). The energy intensity of cement production ranges from 3.6 to 6.5 GJ/ton depending on production process and location of the production (Hammond and Jones, 2011; Ohunakin et al., 2013; Worrell et al., 2000).
Cement production is one of the largest contributors to CO2 emissions. SCMs have been partially or completely used as replacement of cement or fine aggregates in construction to reduce the demand of cement and corresponding CO2 emissions (Al-Harthy etal., 2003; Babu and Kumar, 2000; Bondar and Coakley, 2014; Cheng etal., 2005; Jia, 2012; Khan and Siddique, 2011; Kunal etal., 2012; Limbachiya and Roberts, 2004; Lothenbach etal., 2011; Maslehuddin etal., 2009; Najim etal., 2014; Nochaiya etal., 2010; Siddique, 2011; Siddique and Bennacer, 2012; Toutanji etal., 2004). Some of the established SCMs are fly ash, silica fume, blast furnace slag, steel slag, etc. Pozzolanic materials, such as fly ash, steel slag, and cement kiln dust (CKD) when used as replacement to cement, improve the long-term performance of concrete as the pozzolanic reaction takes time. But, the early age strength of SCMs is a concern, as the reduction in cement content causes lesser hydration and, consequently, lesser formation of CSH gel (Lothenbach etal., 2011). The problem of low early strength of SCMs can be solved by using carbonation curing at early ages.
Apart from CO2 sequestration, carbonation curing has also been found to act as an activation mechanism for SCMs (Monkman etal., 2018). Many studies have tried to assess the effect of ACC on use of SCMs (Monkman and Shao, 2006; Sharma and Goyal, 2018; Zhan etal., 2016; Zhang etal., 2016; Zhang and Shao, 2018). ACC not only enhances the hydration degree of alternative cementitious materials but also improves the early age performance of concrete. Monkman and Shao (2006) assessed the carbonation behavior of blast furnace slag, fly ash, electric arc furnace (EAF) slag, and lime. All four materials reacted differently when subjected to carbonation curing of 2h. Fly ash and lime showed highest degrees of carbonation, followed by EAF slag, whereas ground granulated blast slag (GGBS) showed least reactivity towards CO2. Calcite was the major reaction product from fly ash, lime, and EAF slag, whereas aragonite was produced by carbonation of GGBS. Sharma and Goyal (2018) studied the effect of ACC on cement mortars made with CKD as cement replacement. ACC was found to improve the early age strength of cement mortars by 20%, even for mortars with higher CKD content. Several studies tried to assess the CO2 sequestration ability of steel slag binders (Bonenfant etal., 2008; He etal., 2013; Huijgen etal., 2005; Huijgen and Comans, 2006; Ukwattage etal., 2017). Presence of C2S component in steel slag makes it a potential cementitious material that could act as a carbon sink for CO2 sequestration (Johnson etal., 2003).
Zhang etal. (2016) in their study found that fly ash concrete was more reactive to CO2 as compared with OPC concrete. With the reduction in OPC content, a porous microstructure was generated due to insufficient hydration reaction. The enlarged distance between cement grains facilitated higher possibility of reaction with CO2, and hence, a higher degree of CO2 sequestration. The performance of SCMs subjected to carbonation curing is majorly dependent upon fineness of material and water content postcarbonation. Finer particle size of SCMs provides a higher specific area for effective carbonation reaction. Due to this, it was observed in many studies that concrete made with SCMs had better reactivity toward CO2 than OPC (Monkman and Shao, 2006). Water content postcarbonation also plays a dominant role in determining the performance of SCMs. Sufficient water content postcarbonation is necessary for complete hydration and pozzolanic reaction of SCMs (Monkman and Shao, 2006).
In Europe, cement production decreased by 26.9% from 1990 to 2012, whereas CO2 emissions decreased by 38.6%, showing an improvement in the cement production (CEMBUREAU, 2014). However, to reach the objectives of various sustainability programs, further efforts must be made in order to improve every step in the concrete production line. For concrete, the main solutions for reducing the environmental impact of modern construction are (Flatt et al., 2012):
Some other technical solutions to improve cement production are also of interest, including CO2 capture and storage. If this solution were to present a major impact on the final objective, it would need technological breakthroughs that are still in the research phase and are not planned to be ready for users before 2030, but they could potentially capture up to 45% of the CO2 produced by cement (IEA and WBCSD, 2010). By 2050 the baseline emissions will be 2.34 gigatonne (Gt) but will be reduced to blue emissions of 1.55Gt by the contribution of energy efficiency (10%), use of alternative fuels and other fuel switching (24%), clinker substitution (10%) and carbon capture and storage (56%) (IEA and WBCSD, 2010).
The worldwide use of blended cement in the production of cement is already a commonly used improvement, with significant investments for research made by cement producers (Table 15.7). Many blended cement types exist because of the great variability of usable SCMs. These cement types are produced by replacing part of clinker with SCMs, which leads to cement with unchanged or improved properties for both general and special applications. The same properties must be maintained; otherwise, for a functional unit (concrete beam) and on a product scale (cement), more cement may be necessary to obtain the same durability. Therefore the beneficial impact of replacing a part of clinker can be useless (Li et al., 2004). In the European context, binary cement incorporates up to 35% of each type of SCM even in the case of inert mineral fillers, and ternary cement uses up to 80% in mass replacement of clinker (EN 197-1, 2001). A final trend is the development of quaternary binder, but a lot of improvements remain necessary in order to ensure a better understanding of the interactions between clinker and SCMs.
As the CO2 resulting from the decarbonation of limestone during the calcination process is a fixed amount by clinker volume, the two major solutions are optimisation of the heating process to reduce the energy needed to reach 1450C and the use of blended cements.
The improvement of a specific burning process leads to no or very little change in the reduction of CO2 emissions, but a change from a wet process to a dry process with preheating and precalciner kilns can lead to significant improvements (WBCSD, 2013). The energy used by a wet kiln is estimated to be between 5.9 and 6.7GJ/tonne of clinker, whereas a kiln with preheating and precalciner kilns uses only 2.9 to 3.3GJ/tonne of clinker (reduction by 50%). Shifting the cement production from wet to dry with preheating and precalciner kilns can lead to a reduction of 20% (International Energy Agency, 2010) of the energy needed and 17% in the amount of CO2 emitted per tonne of clinker (Damtoft et al., 2008). It can be seen that the increase in part of the clinker produced by this technology could partially overcome the increase in cement production (WBCSD, 2013). These savings could be increased by the use of more alternative fuels (Nielsen and Glavind, 2007) (to coal, which actually accounts for 60% of the fuel used in cement production).
The use of SCMs in concrete is different from using them to produce blended cements. There is no guarantee on the potential strength achieved by a mixture of cement and SCM because of the great variability in the physicochemical properties of SCM. In order to control the amount used in concrete, use of SCM has been standardised through the k-value concept (Smith, 1967). This concept is based on the potential reactivity of each SCM, which helps in fixing a maximal replacement rate of cement to achieve the same mechanical and durability properties. The k-value for each addition differs depending on the type, on the concrete exposure conditions (frost, salt, sulphate, etc.)and on the local national standards. Depending on the type of SCM, the volume used and the targeted concrete strength, the savings in terms of CO2 emissions can be more than 20% (Table 15.9).
The previous solutions for improving the environmental aspect of concrete are widely used and remain of interest for further development, and some new, challenging solutions are also being studied but will need either breakthrough technology or huge investments in development. So far emissions of CO2 are inherent to cement production, so finding ways to prevent this gas from getting into the atmosphere need to be explored, as illustrated next.
Carbon capture and sequestration (CCS): This process consists basically of capturing the CO2 before it is released into the atmosphere and then compress it and store it underground (in mines, caves, oceans). Unfortunately, currently this process remains quite expensive and energy-consuming. Nevertheless, it is a new technology which can be improved in the future. On the other hand, it can also be argued that far from dealing with the problem, CCS is just a way to avoid the problem and leave it for later. Therefore it is totally in contradiction with sustainable development, which is defined by the Brundtland report (United Nations, 1987) as a development that meets the needs of the present without compromising the ability of future generations to meet their own needs.
Biofuel production: Although this approach is not yet totally explored, some researches have shown that captured carbon can be used in the production of algae that can then be transformed to biofuel, agricultural fertilizer or even animal protein (Potgieter, 2012).
Electrochemical carbon reduction (ECR): This process involves creating a reaction which will transform carbon into formic acid. This product is often used in the pharmaceutical industry. It is usually manufactured with high energy consumption but ERC requires less energy (Potgieter, 2012).
The first approach will be to reduce the volume of CO2 in the production of cement which is emitted into the atmosphere. But the most interesting solution in the short-term will be the development of alternative binders with a lower environmental impact, and some proposed solutions already provide excellent results on a laboratory scale or event in pilot projects (Aldred and Day, 2012; Duxson and Provis, 2008; Owens et al., 2010). Alkali-activated cements are aluminosilicatepozzolana-based materials (glass furnace slag, natural volcanic glass, manmade glass, fly ash, metakaolin). The activation occurs between the alumina-rich pozzolana and a strong alkali base, which dissolves the silicate and aluminate groups to form a cementitious gel to form the structure of the matrix. The so-called geopolymer is proven to have similar mechanical properties as cement-based concrete with a lower environmental impact (Duxson et al., 2007). This solution is already implemented in Australia by using fly ash and slag (Duxson and Provis, 2008). This solution is, however, hindered by some practical aspects, including the robustness of the design of the mix, the heat needed for a proper reaction to occur in a reasonable amount of time, the limited amount of available aluminosilicate compounds, the handling of the alkaline activator on the job site, the cost, and the environmental impact of the activator (Flatt et al., 2012).
Belite-rich cement, with a combination of calcium sulphoaluminate and calcium sulphoferrite, have been tested with success. Compared to ordinary Portland cement, belite-rich cement contains relatively higher belite (C2S) and lower alite (C3S) contents (Chatterjee, 1996). Belite-rich cement produces performances comparable to cement with the production of 2030% less CO2 (Li et al., 2007).
Some specially designed concretes have been developed with the eco-efficiency of products at a structural level in mind. This is particularly the case of HPC, including SCC (Hossain et al., 2013; Sonebi and Bartos, 1999, 2002). First, SCC will not act directly on the environmental aspect of sustainability, but rather on economical and societal aspects. When properly designed from the beginning, a construction made of SCC brings economic benefits by increasing the productivity because of the higher casting rate and the reduction in manpower which results in the elimination of vibration. This latter aspect also has a major societal aspect because it reduces the noise at construction sites and concrete factories and removes the risk of injury to workers related to crowded construction sites and vibration (Nielsen, 2007). The use of SCC in concrete factories (Yahia et al., 2011) can also be a solution for increasing mould service life and saving energy. In the case of concrete pipes, this reduction can be about 1.0GWh (De Schutter et al., 2010). Note also the reduced need for post-treatment of a surface by plaster, which accounts for 0.57kg of equivalent CO2 per m2, or by paint because of the better finish of concrete surfaces (Witkowski, 2015). This latter aspect can also increase the CO2 uptake by concrete carbonation because the concrete surface has no applied post-treatment. The major issue of SCC is that proportioning of a mix involves a significant amount of cementitious material in order to improve the workability of the concrete. This problem can be overcome with the development of high-performance chemical admixtures, use of high SCM content (Diederich et al., 2013), and an adequate selection of the granular skeleton of solid particles to achieve proper rheological properties of concrete. The latter approach has allowed the production of Eco-SCC with 40% fewer CO2 emissions than standard vibrated concrete, along with a 150-year service life (Mansour et al., 2013).
The development of high-strength concrete and high-strength, fibre-reinforced concrete is also a new approach to sustainability in construction materials. These concrete types are designed with a relatively greater amount of binder (cement and SCM) and chemical admixture, which achieves better performance than conventional concrete, including 316 times more compressive strength, 10 times more flexural strength and 10100 times more durability than conventional concrete does (Wang et al., 2015). These high performances allow a reduction in the size of elements for similar structural performance, therefore leading to a lower volume of concrete for the same structure. This reduction in volume can lead to a reduction of 65% in raw materials consumed, 51% in the primary energy used, and 47% in CO2 emissions (Batoz and Rivallain, 2009).
Currently in the United Kingdom, concrete debris is not sent to landfills but is treated so it can have a second life in the construction sector. Indeed, concrete is a material that can be fully recycled. In 2011 recycled and secondary aggregates represented about 5.3% of all aggregates used in concrete (Concrete Centre, 2011). This material is used mainly in road construction, but recycled concrete aggregates (RCA) can count for up to 30% of the aggregates in a concrete mix. In the United Kingdom recycled and secondary materials represent 28% of the total aggregates used in the marketplace, the highest in the Europe. The precast concrete industry provides greater opportunities for using recycled aggregates over to 20% (Concrete Centre, 2011). However, the following issues need to be considered when using RCA:
CO2 emissions from cement production are incurred through the consumption of fossil fuels, the use of electricity, and the chemical decomposition of limestone during clinkerization, which can take place at around 1400C. The decarbonation of limestone to give the calcium required to form silicates and aluminates in clinker releases roughly 0.53t CO2 per ton of clinker . In 2005, cement production (total cementitious sales including ordinary Portland cement (OPC) and OPC blends) had an average emission intensity of 0.89 with a range of 0.650.92t CO2 per ton of cement binder . Therefore, the decarbonation of limestone contributes about 60% of the carbon emissions of Portland cement, with the remaining 40% attributed to energy consumption, most of which is related to clinker kiln operations; the WWF-Lafarge Conservation Partnership  estimated that the production of clinker is responsible for over 90% of total cement production emissions.
In view of the fact that the requirement for decarbonation of limestone presents a lower limit on CO2 emissions in clinker production, and that there exist technical issues associated with the addition of supplementary cementitious materials (SCMs, including fly ash and ground granulated blast furnace slag), which restrict the viability of direct Portland cement supplementation by SCM above certain limits, the possibility to reduce CO2 emissions using Portland chemistry is limited. The WWF-Lafarge Conservation Partnership  expects that the emissions intensity of cement, including SCM, could be reduced to 0.70t CO2 per ton of cement by 2030, which still amounts to around 2 billion tons of CO2 per annum worldwide, even if cement production does not increase from its current level.
Fig. 10.6 shows the CO2 emissions of various binder designs as a function of Portland cement content. There have been a limited number of life-cycle analyses (LCA) of geopolymer technology. One reasonably extensive research program carried out in Germany  has provided information regarding the selection of precursors and mix designs for a range of geopolymer-based materials. However, geographic specificity plays a significant role in a full LCA, so there is the need for further studies considering different locations in addition to a wider range of mix designs spanning the broader spectrum of geopolymers. The main carbon-intensive and also the most expensive ingredient in geopolymer cement is the alkali activator, which should be minimized in mix design. McLellan et al.  provided further detail, while Habert et al.  concluded that geopolymer cement does not offer any reduction in carbon emissions; such a conclusion needs to be drawn with caution.
Sodium carbonate is the usual Na source for the production of sodium silicate. The different processes for conversion of Na2CO3 (or NaOH) and SiO2 to sodium silicate, via either furnace or hydrothermal routes, differ by a factor of 23 in CO2 emissions, and up to a factor of 800 in other emissions categories . It is therefore essential to state which of these processes is used as the basis of any LCA. Moreover, the best available data for emissions due to sodium silicate production were published in the mid-1990s , so improvements in emissions since that time have not been considered. Sodium carbonate itself can be produced via two main routes, which vary greatly in terms of CO2 emissions. The Solvay process, which converts CaCO3 and NaCl to Na2CO3 and CaCl2, has emissions between 2 and 4t CO2 per ton of Na2CO3, depending on the energy source used. Conversely, the mining and thermal treatment of trona for conversion to Na2CO3 has emissions of around 0.14t CO2 per ton of Na2CO3 produced plus a similar level of emissions attributed to the electricity used. This indicates an overall factor of 510 difference in emissions between the two sources of Na2CO3 .
A commercial LCA was conducted by the NetBalance Foundation, Australia, on Zeobonds E-Crete geopolymer cement, as reported in the Factor Five report published by the Club of Rome . This LCA compared the geopolymer binder to the standard Portland blended cement available in Australia in 2007 on the basis of both binder-to-binder comparison and concrete-to-concrete comparison. The binder-to-binder comparison showed an 80% reduction in CO2 emissions, whereas the comparison on a concrete-to-concrete basis showed slightly greater than 60% savings, as the energy cost of aggregate production and transport was identical for the two materials. However, this study was again specific to a single location and a specific product, and it will be necessary to conduct further analyses of new products as they reach development and marketing stages internationally. Fig. 10.6 shows a comparison of the CO2 emissions of four different E-Crete products against the Business as Usual, Best Practice 2011, and a Stretch/Aspirational target for OPC blends. It is noted that in some parts of the world (particularly Europe), some of the blends shown here in the Stretch/Aspirational category are in relatively common use for specific applications, particularly CEM III-type Portland cement/slag blends, but this is neither achievable on a routine scale worldwide at present, nor across the full range of applications in which Portland cement is used in large volumes.
Environmental impact of cement production is calculated based on the data provided by Ecoinvent (2012). Detailed information on the production process and on all inputs can be taken from Knniger et al. (2001). The functional unit is the production of 1kg of Portland cement strength class 42.5 (CEM I 42.5 R). Average fuel for clinker production is composed of 6.81103MJ natural gas (high pressure), 3.74104kg light fuel oil, 2.55102kg heavy fuel oil, 3.54102kg hard coal and 3.91103kg petroleum coke. In addition 5.80102 kWh electricity (medium voltage) is considered. Besides 0.912kg of Portland cement clinker, an input of 0.063kg gypsum (not balanced, origin from flue gas desulphurization), 0.025kg additional milling substances (not balanced as it is taken as waste without environmental burdens, e.g. dust from the cement rotary kiln, fly ash, silica dust, limestone) and 3.50104kg ethylene glycol (process material for grinding) is taken into account. A total energy for grinding and packing of 4.85102 kWh (electricity, medium voltage) as well as transport processes corresponding to 4.40103 tkm (lorry, 16t) is necessary. The production of 1kg Portland cement CEM I 42.5R yields an environmental impact of 0.833kg CO2-eq. (GWP100), 2.241108kg CFC-11-eq (ODP), 4.211105kg C2H4-eq. (POCP), 1.138103kg SO2-eq (AP) and 1.702104kg PO43-eq. (NP).
seven pcb milling tips
After a lot of weekends of trial and error I finally have my PCB milling process working. In this post I share some things I have learned along the way - hopefully it will make it easier for anyone else trying to develop the same process.
This is one of the most important factors to get repeatably reliable boards. It's very tempting to just tape a board down randomly and start milling - but the next board might not be in the same place and you wind up with very different results.
Your goal should be to minimise the number of variables as much as possible - use the same size blank boards, always mount them in the same place and at the same height. My solution was to use a piece of pine for my milling base which I bolt to the bed of the CNC and mount the blank boards on to the pine itself with M3 bolts.
This means the blank PCB is in the same place for every session and I can be sure that changes in results are due to changes I've made, not random differences caused by different placement or other factors.
If your machine doesn't have limit switches already you really need to install them. Apart from the increased safety provided by ensuring you don't exceed your cutting area it makes it easier to position your work co-ordinates by using the homing function.
When you run a homing cycle the CNC will move the gantry until the limit switchs are triggered and then set the machine zero co-ordinates to that point. This allows you to accurately reposition the tool head even after a power cycle.
Limit switchs are not difficult to implement, essentially they are normally open switches that are closed when the gantry makes contact with them. On my machine I just used standard push buttons mounted in 3D printed attachments.
Milling a PCB involves cutting away a very thin (0.05mm or less) layer of copper. Having the board as level as possible during the process is absolutely essential - cut too deep and the engraving tool will cut away too much copper leaving you with very thing tracks; cut too high and the tracks won't be completely isolated and the PCB will have short circuits.
Although this helps (a lot) it still won't give you a perfectly flat surface. The solution is to use probing to measure the actual surface of the board and then an auto-levelling function to adjust for the differences in it.
The probe itself is simply a normally open switch, usually you just use the tool tip itself and the PCB as the two contacts. A probe operation simply moves the tool head (at a very low feed rate) until it comes into contact with the surface and measures the height at that location.
Most CNC controller software allows you to probe an area of the board and then automatically adjust the Z level as it processes your gcode file to take into account any differences in height. This process gives you the most consistant depth across the entire board.
Blank PCB material is not as rigid as it appears, depending on its size and how it was stored (not to mention how you mount it) there will be some bowing which results in a curved surface under your engraver.
Unfortunately, the probing process doesn't put enough pressure on the board to push it down while the actual engraving process will. This means you will get shallow cuts even with an autolevelling process.
The trick is to ensure that the probe pushes the board down to it's fully flat level before measuring the height at that point. My solution was to put a small ball on the spindle that is 2mm lower than the probe tip - as the probe moves downwards the ball pushes the PCB to the underlying bed and compresses before the probe makes contact. This gives you a far more accurate depth reading.
The tool tips used to drill and cut the board in the final steps can tear the copper away which damages the pads and tracks (and it looks horrible). I found that if I engraved away the copper at the edge of the cuts (circles around the drill holes and an engraved path along the board outline) this tear was reduced significantly.
There are a few simple changes you can make to your board layouts to make the milling process more reliable. The isolation cuts are between 0.2mm and 0.3mm wide so you should ensure that the minimum distance between pads and tracks is about 0.25mm.
Footprints that use circular pads can be problematic - if the annular ring (the portion of the pad that surrounds the drill hole) is too small the entire ring can lift off during drilling. Applying tip 5 can alleviate this somewhat but a better solution is to use rectangular pads and use slightly larger pads than you would normally.
Rectangular pads help with another issue as well - with a 'normal' PCB the pads and tracks are surrounded by blank space, in a milled PCB there is still copper surrounding them. This makes soldering a milled board more difficult, you have to carefully avoid letting solder short the pad and the surrounding copper. Using larger, rectangular pads gives you a large surface to solder to and makes things a little easier.
You should also avoid using tracks that are too thin - if the engraver cuts too deep you can cut the track completely, I have been using tracks that are at least 0.6mm thick (twice the width of the isolation cuts).
Your PCB layout software will most likely set these parameters as design rules and you may have to modify some component footprints to adhere to them. Changing the layout this way will not adversly affect other production methods - you can use the same gerber files to mill prototype boards and then send them off for fabrication without issues.
One of the more frustating parts of milling PCBs is the number of dead boards you generate as you develop the process. I have a drawer full of failed early attempts and at times, after a sequence of failures, I was ready to give up on the process altogether.
Eventually I developed a work around that let me use the resulting boards even if the process wasn't successful. By painting the blank PCBs with a light coat of spray paint before milling them I could soak them in etchant afterwards to correct for any shallow cuts. Even extremely shallow cuts that didn't even remove the paint layer could be corrected by simply scratching the isolation lines by hand.
Having the ability to quickly and reliably make PCBs for my projects has had a dramatic effect - because the process is so much faster and reliable I don't have to worry about the time it takes to prepare a board. As a result I have no problems making small boards for very specific purposes that I would have avoided (or simply breadboarded) previously.
There are still things to improve though - milling footprints for surface mount devices is my next goal. The accuracy of the milling process is good enough for a range of SMD parts, how difficult it will be to solder them is the big question.
old mill restaurant | private functions | westminster, ma 01473
The Old Mill is the perfect place for a wedding or baby shower. We have rooms to accommodate 10 to 150 and we are committed to excellence in everything we do. For an elegant or casual dining experience, join us in our charming rustic setting to provide memories that will last a lifetime.
TheOld Mill is noted for tempting foods served in a setting of rare charm. Voices of diners mingle with the merry music of water rushing gaily over the mill dam as it dances its way to the sea. Thus the Old Mill is reborn - its attractive vistas doubled in splendor by their reflection in the mill pond. It has become a shrine to the epicure for its delicious food... to the art lover for its rustic beauty.
milling your own custom titanium abutments & split-file crowns
Ill admit, Ive been somewhat obsessed with ti-bases. especially angle screw channels! As a result, I havent had a huge reason to want to mill my own custom titanium abutments because ti-bases rock and work very well. That being said, there is a time and place within my practice and laboratory where its tempting to use custom abutments. You may know what Im talking about those cases where you just want a bit of extra strength or possibly an abutment with a bit more of a subgingival convexity as compared to the steep angled ti-base. OK, so we are convinced that there is a benefit of using both custom and ti-base abutments, so lets get over that and just look at whats the innovative and exciting part milling titanium abutments and zirconia split-file designs in a combined clinical-laboratory dental office!!
Part of the reason why I invested in a VHF R5 mill is so I could mill titanium discs and blocks. Why would I need a special mill to be able to do that? Turns out milling titanium requires significant cooling as the contact between a bur spinning at 80,000 RPM tends to cause sparks and heat generation. To overcome that limitation, the milling machine sprays coolant, typically water with a bit of lubricant, to help cool the metal and minimize sparking. As a result, the R5 has the native ability to mill both zirconia and titanium in a compact machine ideal for my combined clinical-laboratory office.
A patient came in requesting an implant & tooth replacement for missing tooth #3. From the intraoral scan using my 3Shape TRIOS intraoral scanner, the clinical presentation looked pretty ideal, but lets take a look at it on a CBCT.
The CBCT scan was made with a VATECH Green CT and brought into 3Shape Implant Studio software where it was combined with the intraoral scan and the tooth was planned according to the ideal position within the arch. The occlusion was checked, contours verified, and a virtual Hiossen ET implant placed into the software for analysis. We planned a 4.5mmD x 10mmL implant with a surgical guide.
The implant was placed and a healing abutment placed at the time of surgery. Approximately 8 weeks later the patient returned for final impressions. The healing abutment was removed, the TRIOS scan of the arch and tissues was made without the healing abutment in place to capture the emergence form of the tissues. A DESS Scanbody for Hiossen was placed onto the implant and tightened down. An intraoral scan was made of the scanbody merging the two scans together automatically. The opposing arch scan was made and MICP/bite record made with the scanner. The healing abutment was replaced and patient dismissed.
The optical files were imported into 3Shape Dental System software and using the step-by-step wizard function I aligned the virtual scan body to the intraoral scan of the scan body and the software picked it up right away. Using the wizard functions, the 3Shape software took me through designing a custom abutment following the shape of the soft tissues and emergence form.
The software allowed me to adjust the size of the custom abutment to fit within the contours of the planned restoration. One great thing about using the right DME files in software like 3Shape to do this task is that you can view the shape of the designed abutment within a virtual blank to ensure that the designed abutment will fit properly within the DESS Ti Pre-Mill Blank for the custom abutment.
3Shape has a feature that allows you to create a screw-mentable crown with a screw access channel through the crown above the custom abutment!! My goal was to deliver an abutment, torque it down, and use a small amount of luting agent to cement the crown intraorally. Essentially I built a split-file abutment and crown so that way it was one design and its ready for milling!
The R5 automatically can switch between wet and dry milling without user input or intervention so I was able to program this milling procedure to occur overnight. I started the machine at 6pm, went home, and came back in the next morning and everythings done! In the morning, I removed the discs from the machine with the completed abutment and zirconia crown and the machined worked overnight.
I printed models and gingiva using a NextDent 5100 3D printer using NextDent Model 2.0 Peach and NextDent Gingiva Mask. I fitted a DESS analog into the printed model and articulated the two together using a 3Shape articulator. After removing the sprue from the abutment, I air abraded the abutment using a Danville Microetcher and fitted the crown. Crown fit right on the abutment, although some minor adjustments was needed, it took all of 3 minutes to adjust for small manufacturing tolerance errors.
After completing the staining and glazing procedure for the crown, fitting the contacts, and checking occlusion, I anodized the abutment everywhere except the implant connection to turn the abutment a slightly gold hue color.
The patient returned, healing abutment removed, and the abutment and crown fitted onto the implant. Literally no adjustments needed, the patient was thrilled the with the final result. The contours of the milled abutment and crown allowed everything to fit in effortlessly.
I torqued the abutment down to manufacturers recommended values, air dried the abutment, placed PTFE over the screw channel, and used a small amount of RGMI cement to cement the crown to the abutment. A nice feature of creating a screw channel through the crown was this screw-mentable design allows for excess cement used during this process to easily escape from the top of the crown with minimal chance of subgingival cement.
The entire delivery appointment was fast, efficient, and leaves a very happy patient with the beautiful result of the final restoration. Digital dentistry doesnt need to be complicated or laborious. The key is to make our technology work for us in the right way!
Interested in learning more about CBCT, implant planning, 3D printing, and some amazing step-by-step protocols in your office and/or laboratory? Check out our innovative step-by-step online digital dentistry course at www.LearnDental3D.com
why machine this boxy part on a lathe? |
modern machine shop
South Morgan Technologies has long thrived on its ability to mill boxy parts on a turning platform. Today, newer technology and techniques enable saying yes to more work than ever before.
Machineable pie jaws can clamp virtually any profile shape, including boxy geometry, in a manner that enables turning rounded features and leaving the rest to the Y-axis tool turret.
Although more capable milling and the ability to clamp multiple parts on the table lead the VMCs to cut more metal, this small contract manufacturers live-tool lathes handle a greater number of unique jobs.
Kevin Ames, shop owner, explains the advantages of Okumas IGF conversational programming system, citing the systems ease of use and production of tweakable G-code programs among the reasons why the majority of the shops workhorse machines over the years came from that builder.
The shop considers extra offsets critical to its flexibility. Likewise for this machine-mounted touch sensor, which saves significant time compared to setting tools by hand.
Plant manager Mike Lueck replaces tools in Okuma LB3000EX-IIs Y-axis turret. Leveraging offsets and not skimping on cutters and holders helps make the most of every tool station. Options for a larger subspindle chuck (an 8-inch diameter rather than the standard 6-inch) also help maximize the machines capacity (the main spindle chuck diameter is 10 inches).
A garage, a mill and a lathe. These are common elements in the origin stories of many manufacturers profiled in this magazine over the years, since before handwheels and tape readers gave way to CNCs. However, the lines between machine functionalities have blurred in recent decades. Consider the case of Kevin Ames, who began his own shop-in-a-garage origin story in 2011 not with two machines, but one: a 1996-model turning center with Y-axis live tooling.
Granted, it wasnt long before Mr. Ames purchased a dedicated milling machine as well. Nonetheless, theres good reason why a flexible turning platform was his first priority. Anticipating his role as production manager, machinist and shopfloor handnot to mention salesman, accountant and janitor, to name a few othershe knew that time and resources would be limited, and that versatility would be key to success. You generally cant do any turning on a milling machine, he explains. With a turning machine, you dont have to spend a whole lot more to start drilling holes, too, and a Y-axis will enable you to make some parts complete.
A lot has changed in five-odd years. Having outgrown the garage, South Morgan Technologies now occupies a 6,000-square-foot leased space in Girard, Pennsylvania, where five total machine tools provide more processing options than ever before. And although Mr. Ames regularly gets his hands dirty, a staff of three leaves him with more time to focus on running the overall business. Nonetheless, this small contract manufacturers capacity to make the most of its resources and adapt nimbly to ever-changing shopfloor realities remains rooted in its capability to mill on a turning platform. The difference is that the shop is now leaning on a more capable platform, one that drives a greater variety of live tools at higher power and faster speeds while leveraging a subspindle to consolidate setups even further.
With a more versatile machine, the team at South Morgan is finding it easier to get creative and to say yes to more work than ever before, he says. Among other keys to turning on that creativity, Mr. Ames cites the right tooling and options, making the most of offsets, and treating every setup like a puzzle to solve, often through the use of machinable pie jaws.
This turning center isnt the shops most productive machine, points out Mike Lueck, plant manager. At least, not in terms of raw numbers of parts produced during any given time period. It isnt the newest, either. Both of these honors go to a Genos M560-V three-axis VMC from Okuma America, purchased last February to complement a 1989-model CNC milling machine. That old workhorse is still running, and it is still more than capable of holding the 0.001-inch tolerances typical of most of the shops work. However, the older machines days are likely numbered. And if the recent past is any indication, the future of South Morgan will involve a lot more milling, particularly in difficult materials like titanium. Even now, some parts demand too much horsepower for the shops two live-tool lathes, and the Genos is far more productive than the older mill, Mr. Lueck says.
Still, its telling that the shop maintained only one dedicated (and aging) milling resource through multiple advances in turning center and turning center milling capability over the years. Its perhaps even more telling that Y-axis turning centers handle more unique jobs than the VMCs, despite the latter machines advantage in raw number of parts produced. After all, driven-tool lathes can often provide more efficient machining options than traditional three-axis milling strategies, Mr. Lueck says. When thats not the case, less-aggressive cutting can make them a viable alternative to an occupied mill. For South Morgan, operating in this way helps stretch productive capacity (and financial and human resources) as far as possible in an environment where anything more than 50 parts is considered high volume. Even if we quote whats obviously a milled part, that doesnt necessarily mean its going to go on a VMC, Mr. Ames adds. If there are any round features, and we have enough speed and power and tooling for the milling, well usually do whatever we can to put it on a lathe.
The shops newest turning center ensures that this approach will likely remain critical to stretching capacity well into the future, he says, even amid the active pursuit of more VMC work. Installed in March 2014, this Okuma LB3000EX-II horizontal lathe features the shops first subspindle and backworking live tools, which enable working both sides of a part in one setup. It can also take beefier cuts in more difficult materials than the shops other live-tool turning center, a year-2000-model Okuma LB300-M that drives tools at 4.4 hp (continuous) versus the new machines 7.5 (peak hp ratings are 9.5 and 10, respectively). Weve driven 5/8 taps in 316 stainless with no problem, Mr. Ames notes. Meanwhile, driven-tool speeds ranging to 6,000 rpm (the older, non-subspindle machine offers 4,500) facilitate the use of both smaller and higher-quality tools that enable more precise, more intricate milling.
Although every job is different, one strategy is common to virtually all boxier geometries produced on a turning center: using machineable aluminum chuck jaws to clamp virtually any part-periphery shape. Even if a part requires no turning at all, indexing the lathe spindle (or subspindle) can present different faces to the Y-axis turret without refixturing. That said, leveraging a turning machines primary function is always a goal, and Mr. Ames and Mr. Lueck specifically seeks out rounded geometry when evaluating any potential new job. Custom jaws enable orienting work to keep these features on centerthat is, aligned with the middle of the chuck and the machines Z axisso that they can be turned.
Consider the part on the cover of this magazine. One of two mating, rectangular components, it was initially quoted for a two-setup VMC operation but moved to the LB3000EX-II because the VMC was occupied, Mr. Ames says. After facing, profiling, drilling, tapping and chamfering operations, the roughly 6-by-6-inch workpiece was flipped over and remounted, this time in a separate, overlapping jaw profile machined at a 45-degree angle to the first (a clearance issue prevented the obvious approach of handing the part off to the subspindle instead). This second setup put the parts off-center, round geometry at the center of the chuck for turning.
However, this approach would have been more difficult if the rounded boss stuck out much farther, Mr. Lueck says. Thats because an out-of-balance setup can create a wobbling effect as the spindle turns. The extra vibration impacts the workpiece long before it affects spindle bearings or other machine components (barring a part flying out of the chuck), but of course the South Morgan team strives to avoid adverse effects on either. To that end, stick-outs longer than three times diameter generally arent tolerated, and machining parameters for longer parts are often conservative. Another common approach is bolting additional material onto the jaws or machining some away to compensate for off-center part mountings.
Beyond these go-to strategies, Mr. Ames says its all about putting together the pieces of the puzzle that each job becomes when capacity must flex as needed to keep machines occupied and profits in the black. In cases when turned features are offset from workpiece center by a considerable distance, performing a significant portion of the milling work first might help balance things out. Similarly, pre-roughing a turned feature with live tools might temper some of an off-center parts pull against the chuck jaws. We might mill (a turned feature) into a shape like a stop sign first, he explains. Then, when we go in with a turning tool, the interruptions are less severe.
However creative, a machinist is limited at some point by the capabilities and processing options on the shop floor. Beyond a subspindle and greater speed and power, plenty of tooling was a top priority for the shops newest Y-axis turning center, Mr. Ames says. Eight live toolholders, four axial and four radial, offer a variety of processing options, although most jobs require fewer than that. The shop invested significantly in the tools themselves, too, outfitting the lathe (and in some cases, other machines) with a fresh package of cutters from Sandvik Coromant that are better suited for its speed and power.
Versatile tooling is a particular priority for the light, fast milling passes favored at South Morgan. Five-flute, 0.5-inch-diameter end mills with corner radii rounded off to 0.03 inch to prevent edge breakdown are standard for almost every milling operation that doesnt involve aluminum, which favors a bigger, three-flute bite, Mr. Ames says. Tools can also be leveraged in multiple ways when necessary. For instance, a chamfer might be machined with a spot drill instead of a chamfer tool if the application calls for more tooling than can be mounted in the machine. Similarly, relying more on boring tools to reach final hole size is considered a small sacrifice for stocking fewer drills in larger-diameter increments.
Extra offsets have also been critical to leveraging the tool turrets full capacity. One reason is that two of the axial toolholders are double-sided, enabling the same turret station to work on both main and subspindle. These stations require offsets for two tools, not one, and each tool is commonly associated with multiple offsets. Consider a simple grooving operation. For anything wider than the tool that machines it, the IGF conversational programming system that the shop purchased with the machines THINC OSP P300L CNC automatically assigns two offsets to opposite corners of the tooltip, one for each side of the groove. Controlling the position of each wall independently reduces the risk of a single offset leaving one side of the groove within tolerance and the other out of specification. In a more complex operation involving particularly sensitive finishing, the right combination of offset slots can enable a single tool to leave different levels of stock on different areas of the workpiece.
Certain offset slots are never overwritten, remaining linked to specific tools, Mr. Lueck adds. That way, if a tool is removed from the turret, it (or another of the same make, model and size) can be replaced without any additional measurements. Granted, he emphasizes that this works only for simpler tools, most notably drills (unlike, say, a grooving tool, a drill cuts at spindle centerline and is relatively forgiving in terms of precision). Nonetheless, this strategy saves time by making it fast and easy to re-install a drill that had been temporarily replaced by a more specialized cutter.
Static turning tools can also be doubled up on individual turret stations, one right above the other (say, a rougher and a finisher). Thats thanks to the machines optional offset turning function, which enables turning at Y-axis offsets other than zero. Via a slight Y-axis adjustment, the second tool can move in to the work right after the first. This saves a tool slot and eliminates the need to index the turret. This function is particularly useful during threading operations, Mr. Lueck says, explaining that a slight Y-axis offset can make a world of difference in tool life by applying additional, chatter-reducing load on longer parts that are subject to flex. Mr. Ames adds that the function might someday come in handy as a stop-gap if the machine were somehow knocked out of alignment.
Conversational, shopfloor programming has always been the way at South Morgan Technologies. After all, writing 200-odd part programs per year would be far more time-consuming with a CAM package than with Okumas IGF system and its graphical, step-by-step interface, Mr. Ames says.
Even so, offline programming has recently found a home here as well. Based on trial-cutting results, the shop purchased a seat of SolidCAM high-speed milling (HSM) CAM software along with its new Genos M560-V VMC. The software has increased milling productivity through significantly higher machining parameters and more efficient toolpath strategies, he says. With this system in action, employing the same HSM software on the LB3000EX-II turning center has become a tempting proposition as well. The CAM developers turning package promises similar gains.
This isnt to suggest that South Morgan would abandon conversational programming if it does indeed dive deeper into CAM software. Offline programming of the LB3000EX-II would provide yet another option for this evolving shop, an option that could further that evolution by facilitating faster cycles on more complex work. Meanwhile, a quickly drawn-up conversational program will always be superior in some casessay, completing emergency work on the subspindle while another job runs in the main spindle.
Just as conversational and CAM each have their place, its important to keep in mind that a shops success with one type of machine tool shouldnt be taken as a universal endorsement over other processing options. For instance, its not hard to imagine a rotary-fourth-axis VMC making a significant impact on the shops approach to quoting and machining at any point in its history, had it ever opted to go down that road. Whats more, no one can fully predict future needs. For his part, Mr. Ames says he didnt fully anticipate the pace of the shops recent plunge into more advanced milling even as he pursued that very goal. Versatile machining turns not only on versatile technology, but also on creativityon the drive and the ability to make the most of whatever resources are available.
With macros and canned cycles resident in the CNC on most contemporary turning centers, single point turning of OD threads can seem like almost a default process decision. However, for numerous applications, OD thread rolling has inherent advantages as an alternative to cutting threads.