In the third part of his ongoing series on lead manufacturing back to basics Michael McDonagh reviews the technology commonly used to make battery grids.
There have been many attempts by alchemists throughout the ages to turn lead into gold. Whilst we now understand why that was not possible, battery technologists still persist in attempting to change the natural properties of metallic lead into that of a different chemical species. Far from the romantic notion of transforming something dull and uninteresting into a glittering thing of beauty, we have to content ourselves with a much more mundane transition: that is the transition of something quite chemically inert, into something quite chemically reactive. And that I’m afraid is about as good as it gets in the lead-acid battery world.
In the first part of this article I will be looking at the remaining techniques of moulding and shaping lead which utilises its unique metallic properties for the conducting structures used in plate manufacture (grids). The second part will examine the first stage of the transition of lead into the reactive substance used to store and provide electrons, i.e. the manufacture of lead oxide.
Pressure die casting
Whilst the majority of lead-acid batteries have the construction of a flat grid with active material applied from a roller, there is a group of traction batteries which have a positive plate produced in an entirely different way. This group called tubular plate batteries are designed for heavy duty deep discharge applications, where batteries can expect a daily hammering to at least 80% of their rated capacity. For this reason their construction is designed to stop active material from shedding due to the stresses of the volume changes in the active material during the charge discharge reaction at the positive:
The volume change associated with these reactions leads to cracking and spalling of the active material, shedding in fact. This accounts for most of the sludge accumulation in the mud space at the bottom of traction battery cases during their life.
In order to retain the active material and keep it in contact with the grid conductor, the cast spine is inserted into a gauntlet consisting of an array of tubes, Fig 1 shows the spine cast from the machine. These tubes are designed to keep the active material in and in close contact with the spine. The properties and material of this gauntlet will be described in a later article discussing the methods of applying the active material to the grid.
One problem with this construction is that it promotes plate growth. The charging of these batteries inevitably puts in sufficient overcharge to corrode the grid. The oxide PbO2 has a greater volume than the metal. This causes expansion of the substrate metal. Being a rod it expands along its linear axis to grow upwards through the lid. There is limited allowance for this growth in the design of the lid and pillar seal, however the resistance to this slow growth i.e. the creep strength of the material is of vital importance.
The process is different from the gravity casting methods in that the alloy is injected into a mould under pressure. Due to the length of the spines it is impractical to attempt to gravity cast this shape, although for smaller plates as in 6 or 12 volt traction monoblocs, casting is still common practice. However, there are still companies who will manufacture long grids by casting short spines then, cropping spines from other castings, will fuse these together to make a longer spine. These spines are then used to manufacture 2 volt traction DIN and BS ranges. This, of course, is a practice requiring extreme care and strict quality control methods.
A typical pressure die casting utilises a hydraulic piston rod which provides the pressure and velocity of the molten lead alloy to be injected at high speed into the grid mould which is clamped in place under pressure. The grid mould is a steel block containing an injection nozzle and feed chamber designed to ensure that the molten alloy is distributed evenly across the width of the plate (Fig 2). Even so, the normal pattern of filling is that the sides are filled first and some of the lead flows backwards from the end to meet the incoming lead, in the middle, from the last stages of the filling shot. This occurs in less than a second and it is important to ensure that there is adequate air venting in the area where the upwardly mobile latter, meets the downwardly flowing former (similar to the executive toilets in a merchant bank, which these days, also require good ventilation).
The second part of the process is to ensure there are no voids or defects in the top bar of the casting which joins the feed chamber. This is a slower piston movement, which pushes a further amount of molten alloy into the top section of the frame, which in this casting arrangement is the bottom section of the mould.
The main property of lead required for this process is obviously fluidity, so antimony and tin would be obvious alloying ingredients. However, we have also to consider the properties required during battery use and also the materials used for the mould.
Fig 3 shows the fluidity of lead antimony, as the antimony content varies from 0 to 11%. The maximum fluidity is around 11% which is the Eutectic point and therefore no pasty region, (as discussed in the last article). Older machines are based on using alloys of this composition and without extensive modifications they are generally unable to cast with alloys below 9% antimony.
Whilst many companies still use higher antimony alloys there is a growing trend to use very low antimony and sometimes lead calcium alloys in order to reduce gassing, particularly for valve regulated traction batteries. For these alloys there are two problems: the lower creep strength and the lower fluidity. The former problem is mitigated by use of higher arsenic levels, and in any case their use will include a low overcharge to reduce gassing, which means lower levels of corrosion of the spine thereby reducing the plate growth. The lower fluidity is compensated by use of higher casting temperatures and nozzle pressures.
In the effort to reduce antimony content, I have known some manufacturers who use a 5% Sb alloy. As can be seen from Fig 3, this actually gives a low fluidity and makes pressure die casting quite difficult.
Fluidity can be improved by alloy specification and by spraying the mould with lubricant. However, the choice of lubricant is a critical issue. Silicon based sprays actually leave a coating on the grid which is resistant to corrosion during the formation process. It can act as a barrier layer which reduces the conduction between the grid and active material at the interface. This is a real danger, and I have had personal experience of this with one company in the early nineties. Whole production batches of Industrial 2 volt cells giving only 50% capacity were traced to a helpful well-meaning operator who wanted to improve his casting figures and reduce his reject rate. He had found that by using a silicon spray left by the maintenance department, he could increase his daily total from 850 to around a 1 000 per shift with minimal scrap.
For those production managers who believe perfect batteries can be produced with perfect instructions, please take note: nowhere did it say in the written procedures ‘you cannot use a can of lubricant left by the maintenance department to spray a spine mould.’
The use of lubricant is important in obtaining a good productivity rate and most manufacturers fit an automatic spray system to the mould opening section. The most common lubricant is stearic acid (candlewax), I have also seen (and smelt) animal fat being used, mixed as a spray with hot water. The same method is also used for beeswax. Generally anything providing a thin liquid at the casting temperatures, which reduces the surface contact angle of lead on the steel substrate, but will be broken down by the formation process into harmless components, is acceptable.
A summary of the casting conditions is given in Table 1, however it is worth noting that at too high a temperature, the alloy remaining in the feed section of the mould may still be molten. In this case when the piston finishes its down-stroke and returns to its up position, liquid alloy can be sucked back out of the nozzle section and will leave voids in the top bar or lug of the casting. It is worth instructing the operators to achieve a condition in the feed section where there is a cavity in the untrimmed casting to a depth of around 1 – 1.5 centimetres. Less than this and the casting may contain defects in the spine due to low temperature conditions. More than this and there may be voids created in the top section of the casting.
Rolling, rolling, rolling, the continuous future of automotive battery production methods
The need to produce automotive batteries with higher cold crank performances, better charge acceptance, lower weight, lower price and longer life has led to various methods of grid production which rely on providing a strip form of lead, which is flexible and can be shaped into a grid which is thinner and lighter than a conventional cast grid. This then not only gives the ability to increase the total plate area for performance improvements, but also can reduce the weight of the non-contributing grid support structure. The three main contenders for this method are wrought strip, continuous casting and rolled punched systems.
In the continuous rolled system, a strip is cast, rolled to the required thickness within tight tolerances, and then follows a shaping operation either by slitting and expanding or by punching and flattening at the same time. However, the Concast system of Wirtz continually casts the strip and the finished grid in one operation. The cooling rolls provide the grid shape to give a continuous strip containing the final grid with no further processing other than cropping.
There are advantages and disadvantages to each, depending on the end use and the material structure requirements. All have the advantage of being a continuous system capable of using orifice pasting methods and formation as a coil prior to being cut and assembled into batteries. The alloys predominantly used are based on lead calcium tin, and the relative merits and uses of these alloys is listed in Table 3 of my second article in issue 36 of BEST magazine (also available to download).
A common misconception is that alloy formulations should contain additives which provide a grid which is as corrosion resistant as possible in order to prolong the life of a battery. Corrosion resistance is important, but also the necessity of forming a good bond, using the same corrosion mechanism, at the grid and active material interface is paramount. This bond is created initially in the curing reactions then during the plate formation processes. These reactions and their significance for performance and life require optimised processes to produce the best results.
The effecton grid microstructure of these processes is shown in Fig 4. With greater mechanical deformation comes a finer crystal structure. Again there is a debate concerning the ideal structure, a small crystal structure will provide a high grain boundary area which will improve the tensile strength and improving creep resistance, but under some charging parameters may in fact increase the volume of corrosion product and cause additional grid growth.
It is however, quite obvious that if the corrosion mechanism is via grain boundary attack, then a large grain structure will provide an easier path for the corrosion path to penetrate into, and possibly through, the grid. This is the basis of early catastrophic failure. A fine grain structure has always been preferred in cast systems and should also be suitable for rolled structures. It is quite clear that the rolling process can provide an effective method of controlling and obtaining an optimum grain size.
The main advantage of all three methods is the ability to produce thinner grids suitable for high performance high power batteries by packing more plates in the box for the same or even lower weight than with conventional cast grids. However, the high capital cost means that what you save in labour is more than offset in depreciation. Downtime becomes an expensive luxury and increased engineering complexity makes excessive downtime even more of a probability. For this reason a significant number of companies prefer the flexibility of gravity casting. If you have four casting machines instead of one strip caster, the likelihood of your production being completely stopped is very small.
The expanded grid and roll punched system have some common features, namely they both rely on lead calcium strip which is rolled to the appropriate thickness and then shaped in the final process to provide the grid for the battery plate. Both are capable of continuously producing thin grids down to around 0.85 mm at high production rates. These two processes and their relative merits will be compared with the continuous casting method.
Expanded metal systems
These are based on the method of slitting strip then pulling the edges to expand the width of the strip and thereby create the familiar diamond pattern. This is basically the method used by manufacturers of galvanised fencing and fan grills from steel strip, and the basic engineering principle is very similar. Early attempts to produce grids by this method suffered from severe corrosion at the node points when used as a positive grid. This was traced to microcracks introduced into the structure at these points due to the high deformation caused by the pulling apart of the strip after slitting. This allowed access of electrolyte and therefore corrosion reactions, deep into the section of the grid. The effect was to reduce the cross sectional area of the grid section and accelerate the mode of failure due to positive grid corrosion.
Fig 5 shows a typical example of an expanded grid. The lug or flag is cropped from the sides which have not been slit. The process of expanding the strip after cutting creates the diamond pattern. However, it introduces deformation at the nodes, which can produce cracks, particularly if the material is already age hardened. This deformation can also lead to recrystallisation and produce a crystal structure which is only partially transformed and therefore corrosion sensitive when it is put into the battery and charged.
This problem has now been largely overcome by control of alloy content, control of deformation, and controlled aging of the strip after deformation. However it is important to understand that the alloy specification and process conditions must be very carefully controlled. Work by David Prengeman has shown that some concentration ratios of tin to calcium are more sensitive to producing semi-transformed grains. A partially transformed grain was found to be corrosion sensitive and was responsible for creating premature battery failure due to positive grid corrosion in the initial years of this technology.
The main disadvantage of this process is the lack of design options and the inability to optimise the electrical conducting path and therefore reducing the internal resistance of the battery. This means it is limited if we are to maximise the cold cranking performance and charge acceptance in starter batteries. However, the system is probably the least expensive of the rolled systems and the technology is now mature, with most of the technical snags now ironed out. Table 2 gives a comparison of the two rolling technologies, and the continuous casting system.
This method developed by Wirtz, principally for the positive plate, provides the advantage of flexibility of grid design whilst retaining the advantage of a continuous production system and very thin controlled thickness grids. As with the expanded metal system, lead calcium alloy strip is rolled to the desired thickness then given a final stamp which produces three things: the final grid shape, the final thickness and also the final surface finish. Fig 6 shows a part of a typical stamped grid with a roughened surface finish. This is considered an important aspect of ensuring that a good bond is created between the paste and grid, particularly during the curing process.
Again this process has its roots in the steel industry, where molten steel is poured into a tundish then allowed to flow initially through cooling rollers to form a continuous semi‑solid billet which is then further rolled through a series of reducing rollers which also turn the continuous billet through 90º and onto the rolling table to continue the process of making the steel strip.
The Concast process from Wirtz uses this principle, but the cooling rolls also contain the grid pattern which is formed as part of a continuous strip which is transferred to a coiler for storage, age hardening and subsequent pasting. With this process, grids can be made thinner than with conventional casting methods. The immediate advantage of the concast system is that there is no scrap lead to recycle from either trimmings as with conventional casting and punching out the grid and lug shapes from the continuous strip systems.
The other big advantage is that there is more scope for complexity of the grid design being produced. The shape of the grid wire can be optimised for corrosion resistance and conductivity and the grid wires can be distributed to minimise the internal resistance of the battery and maximise the cold cranking capability. This flexibility for component shape and design also mean that the use of lead for the conducting element of the electrodes can be minimised. This is an important feature not only for cost but also for increasing the power and energy densities of the lead-acid battery.
In summary, there are significant economic, design and environmental advantages to these methods. However, the high capital cost and financial consequences of downtime are powerful disincentives to encouraging their use in small and medium sized enterprises.
Let’s get chemical- lead oxide for battery electrodes
The previous processes have used the properties of lead as a metal and alloy, basically its conductivity, malleability and fluidity. These properties have been exploited in the processes used to manufacture the conducting grid and to provide a stable supporting structure for the active mass, which will not only give reasonable life but also enable maximum possible battery performance.
For the active material which provides the electron generator and acceptor, we must exploit the properties of lead in a different chemical form. To change lead from being relatively inert to being relatively reactive takes a change of state. This change is achieved in several energy intensive stages, oxidation to create a reactive powder with a high surface area, paste mixing which provides the basic physical and chemical structure to apply to the grid, followed by curing and formation.
The concept of turning lead into gold is not new and even recently I have heard battery manufacturers (usually from sales or commercial departments) who regard the production of batteries as a modern allegory of this ancient aspiration. Enthusiastic managing directors and the plethora of marketing executives that you meet during years of working in the lead-acid battery industry, often use the idea of making batteries as analogous to turning lead into gold.
Whilst this may be at least partly applicable to the selling agents and retailers, it is very far from the mark for lead-acid battery manufacturers. When you consider that the selling price from manufacturers to agents or OEMs is often little more than 50% of the LME value of the lead inside the battery, this phrase should be revised. A more accurate adaptation would be ‘Turning lead into something slightly more valuable, such as wood or potatoes’.
Refined, recycled or primary lead, different starting points- same outcome?
Most manufacturers of lead-acid batteries have strict standards for the level of impurities in the refined lead used for oxide production, ranging from 99.99% purity to 99.97% purity. The specific levels of impurities are a matter for speculation and will be marginally different for each manufacturer. This is based partly upon past experiences but usually, and more likely, it is the result of past influences from conferences, literature, and more than likely, hearsay involving technical guys networking. However, considering the genuine past disasters concerning impurities in electrolyte and paste mixtures, and the importance of minimal gas evolution in the ever increasing use of recombinant valve regulated technology, it is hardly surprising that manufacturers are nervous.
At the moment there are two distinct camps in the field of impurity levels. Research into the effects of known impurity elements, such as bismuth, have actually lifted this well-known impurity to the status of an additive, claiming beneficial effects on life, performance and gassing rates in VRLA batteries. The research carried out by the ALABC suggests that far from being impurities some of the long standing elements whose presence in refined lead was considered detrimental, might now be considered as beneficial.
If we accept these findings, then we can drastically revise our ideas of both the total purity required and the individual levels of some elements. Table 3 gives the main elements considered and the individual levels which can be tolerated. It is not suggested by any of the researchers involved that the maximum levels for each element are to be permitted in the refined lead. This would lead to a lead purity of less than 99.90%, I do not believe this would be acceptable to most manufacturers.
With regards to total impurity levels, there has been work undertaken which claims that any impurity level in the active material can spell disaster for VRLA Batteries. One thing which has not been researched thoroughly is the effect of these impurity levels on the manufacturing processes. There may be contributing effects to the efficiency of oxide production and also on battery performance which result from impurity levels affecting the oxide structure.
The transition – from heavy metal to energy store
Lead is a heavy metal with an atomic number of 82. This means that the outer orbitals which contain the 4 valence electrons do not have such well-defined energy levels due to the screening effects of the electron orbitals closer to the nucleus. In effect the atomic structure is as follows:
Electrons per shell:
2, 8, 18, 32, 18, 4
4f14 5d10 6s2 6p2
Lead is a member of Group 14 (IVA) of the Periodic Table because it has four electrons in its outer, or valence, shell. However, because of the overlapping 6s and 6p orbitals the usual valence of lead is +2, rather than +4. The two s electrons have higher ionisation energies. As a result, tetravalent lead is not easily formed and usually is found in low concentrations as a free ion.
The basic reaction for production of lead oxide is simply:
2Pb + O2 = 2PbO
PbO exists in two allotropic forms: tetragonal α-PbO which is red, and orthorhombic β -PbO which is yellow. The allotropic form has an influence on the products of subsequent chemical processes, notably the paste mixing and curing processes. The structure of the oxides is an important factor in influencing the reaction products which are formed during subsequent processes. Fig 7 shows the crystal structure of α-PbO. The ratio of α to β which is formed can be controlled to some extent in one of the standard production methods, since this does have an influence on the subsequent processes and final battery performance, it is important to understand the parameters which affect this outcome.
The two standard methods which provide a practical minimum cost method of producing lead oxide are by using a ball mill with heated air flow and a high temperature reactor or Barton Pot processes. In both cases the conversion efficiency is very low and typically the yield of oxide is less than 80% with variations of 75% to 35% from the same equipment being considered as acceptable.
This variation is the starting point in establishing the uniqueness of lead-acid batteries in the manufacturing world. It is also the source of many of the subsequent process compromises and tolerance limits which would be considered as simple insanity in most other industries. It is vitally important therefore, that a full understanding of these limits and their consequences in the manufacturing processes are thoroughly understood. The methods of oxide production, the control and the optimisation of oxide properties will be examined in the next article.