Implications of lead alloy solidification characteristics

Dr Mike McDonagh elucidates the problems faced by manufacturers of lead alloy grid-casting machines for PbA battery manufacture with a focus on the implications of the solidification characteristics of lead alloys. In this article he reviews the gravity book-mould method of grid casting.

There are two main alloys; hypo-eutectic Pb/Sb and Pb/Ca/Sn versions generally used for battery grid manufacture. Due to their very different metallurgical treatments, I will be reviewing only the gravity book-mould method of grid casting, the rolled strip method will be dealt with as a separate article.

The first step is to examine the binary phase diagrams of both the antimonial and calcium-based lead alloys. A binary phase diagram is a graphical “map” used in metallurgy to show which phases (liquid, solid, or both) are stable at varying temperatures and compositions for a mixture of two metals. It shows how an alloy will melt, solidify, or transform from one phase to another during heating and cooling.

Depending on the solubility and reactivity, electronegativity etc. of the two elements to be alloyed, phase diagrams will have different forms. In this case, the Pb/Sb and Pb/Ca binary phase diagrams will be used to predict the solidification characteristics of these alloys to predict the problems of obtaining good defect-free grid castings using the gravity book-mould casting method.

For simplicity of illustrating the casting difficulties, we will only consider the binary versions for lead-antimony and lead-calcium alloys.

Lead-antimony alloys

The alloying of these metals results in a binary eutectic phase diagram, Fig 1. This shows the lines of stability for various compositions for the α and β phases from 0–400°C. The eutectic point at 11.1 wt% Sb represents the composition where the melting (or freezing) point is at a minimum. It is also the point where the solidification process is a straight transition from liquid directly to solid at a single temperature.

At Sb concentration before or after the eutectic point, there are zones represented by the liquidus to solidus lines, where freezing occurs over a range of temperatures. The α and β phases are lead containing antimony in solid solution, and antimony containing lead in solid solution respectively. In each case the solubility limit is defined the solidus lines, their solubility limit varies with temperature. The alloy concentration chosen is a hypo-eutectic alloy of 1.8% Sb. This is a common alloy used for monobloc, industrial and SLI batteries. Larger 2V cells used for deep cyclic applications will vary from 2.5–9wt% Sb. The reason for this is also the subject of a future article.

Based on the equilibrium phase diagram, we can predict the proportion of α and liquid phase present in the solidifying metal. Without going into detail, we use the lever arm rule to measure the relative proportions of the horizontal isothermal lines on either side of the vertical concentration (isopleth) line. Using this rule predicts the relative amounts of α phase and liquid phase in equilibrium Fig 2.

As the cooling progresses, the relative proportion of liquid reduces until it reaches the eutectic solidification temperature of 252°C. At this point all remaining liquid converts directly into eutectic solid, i.e. 87% α phase (lead with antimony in solid solution), and 13% β phase (antimony with lead in solid solution).

However, this is under equilibrium conditions. With the rapid cooling of a gravity cast mould, the molten alloy may be chilled to below the eutectic melting point in less than a second. Under these conditions the phase changes and solidification process do not follow the equilibrium pattern.

The speed of solidification causes Sb solute atoms to be rejected from the solid α phase faster than they can diffuse away from the surface. This imbalance leads to a higher than predicted concentration of Sb in the liquid compared to equilibrium cooling condition of the phase diagram.

This effect is known as constitutional undercooling Fig 3. It provides a greater than anticipated volume of liquid eutectic in the final moments of solidification. This greater volume of eutectic should theoretically improve the alloy flow into the smaller sections of the grid structure. Unfortunately for us humans, nature seldom gives us that kind of assistance. As you might expect, there are a couple of other consequences to the freezing pattern just described:

• Dendrite growth. A binary eutectic alloy will grow from a seed crystal in a dendritic (tree like) structure, Fig 4. The diagram shows the formation of arms or branches that follow a thermodynamically favourable path along the advancing concentration gradient. Nucleation sites on the initial branches of the dendrite are the source of further branch growth. These branches essentially impede the flow of the rapidly crystallising molten alloy by both blocking passage via entanglement and trapping any remaining low melting point liquid in the dendrite branches. Modern alloys use a grain refiner such as selenium that nucleates more solidification sites for the α phase, by precipitation of higher melting point particles of PbSe. The high number of sites means that the dendrite growth will be greatly reduced and a greater number of smaller, equiaxed α crystals will be formed within a low melting point liquid.

  

• After-flow of the remaining liquid phase. This is a consequence of the different cross-sectional areas of the grid members, particularly the difference between the grid frame and the much thinner grid wires. The thinner sections will freeze more quickly than the frame due to their lower mass and the higher surface area contact with the mould. The thinner wires of the grid then freeze in advance of the larger frame sections. The wires will shrink and essentially suck any remaining liquid alloy out of the frames. If there is entanglement of the α phase dendrite arms this will impede the flow and can result in local pockets where liquid has been pulled out into the shrinking wires but nearby micro pools of molten alloy are prevented from flowing into the pockets by the entangled dendritic grains. This dendrite entanglement can be avoided by changing the crystal size and morphology by use of grain refiners such as selenium. Selenium also reduces the interface energy to change the shape of the grains Fig 5.

  

• Grid mould design factors. There are many engineering and design problems to overcome. However, the three big issues can be listed as:

1. The mould gate and ladle pour
2. The rapid heat conduction and freezing of the molten alloy
3. Cross sectional area differences in the grid design
4. Internal mould temperature balance and production rate.

The first issue has two competing elements. The first is the need to have a high enough pour temperature and alloy mass to enable the metal to flow around the entire mould in fractions of a second. The second is to remove the heat from the gate section to enable the remaining lead mass above the attached grid to freeze at a fast enough rate to achieve a high productivity.

This second point relies on mitigating the high conduction rate of the steel surface to enable the alloy to rapidly flow through the entire mould, including the thinner sections, without freezing. This is done by including in the grid production process an insulating-cork, mould spray.

This strategy reduces the temperature gradient by providing a barrier to the heat conduction. The effects of the freezing-induced shrinkage and voids/frame sinkage due to cross sectional area differences can be mitigated by incorporating a radius at the wire/frame intersections. This basically reduces the immediate difference in freezing rates between the two sections and permits some levelling out of the differing rates of solidification.

To get this right there must be a precise balance of temperatures from top to bottom of the mould. By this level of control, grids can be ejected quickly without cracks, internal voids or missing wire sections due to localised overchilling.

Lead-calcium-tin alloys

For simplicity, we can consider the freezing characteristics of the binary Pb/Ca equilibrium phase diagram, Fig 6. The solidification pattern for this alloy, follows a peritectic profile. This is very different from the Pb/Sb eutectic diagram.

At a concentration of 0.8% Ca, it can be seen by following the isopleth line as it transitions from liquid to solid, That the liquidus/solidus region spans just a fraction of one degree Celsius – Fig 7. Likewise, the concentration range is less than 0.05 wt%. The main differences between this and the Pb/Sb phase diagram are:

• There is a very small temperature range for the liquidus and solidus lines up to 0.12% Ca. The transition from liquid to α phase is rapid with no real pasty region.
• There is no low melting point eutectic phase. This means no after-flow of low melting point alloy. It also means there is no dendritic grain growth.
• The second phase formed is Pb₃Ca which has a higher melting point than the α lead matrix. The control of this precipitation and mitigation of the potential for corrosion is a major consideration in the alloy composition Fig 6.
• Due to the relatively sharp transition between the solid and liquid phases, it is essential that the mould temperature control prevents hot and cold spots that may cause casting to be ejected with some liquid within the structure. This would lead to hot tearing. The mould design strategies for overcoming this are similar to those outlined in the Pb/Sb section above.

The considerations for a gravity casting mould design are similar for both alloys. Simply put, these are:

• Controlled heat extraction via control of conduction through the steel walls.
• Ensuring the upper and lower mould parts solidify uniformly.
• Differences in the frame and wire cross sections where they meet, should be gradual. This is achieved with a radius at the joining sections.
• The balance of ensuring a liquid alloy after-flow from the gate section (to avoid defects due to volume differential) and the grid freezing rate, has an additional role of enabling maximum productivity.
• Residual air has to be removed to ensure that there are no air locks to impede lead alloy flow. For this reason, removeable air-escape slots are incorporated into strategic locations in the mould block.

In essence a lot of the engineering requirement is all about heat dissipation. There are two heat burdens to consider when designing the cooling requirements of a mould:

• Qs – The sensible heat or enthalpy – This is the heat content that is contained in the alloy that pushes the temperature beyond the melting point.
• L_f – The latent heat of fusion. This is a thermodynamic property of materials that is released when transitioning from a liquid to a solid state.

As explained, the heat needs to be taken out of the lead in a relatively uniform manner from top to bottom. The mould design has both heaters and water-cooling jackets. The heaters initially warm up the mould to the casting temperatures and the strategically placed cooling jackets will help to regulate the mould surface temperature to maintain optimum casting conditions.

The time to solidify can be calculated using this formula:

t_s = (ƿ*x* L_f)/h*(T_f – T_m)

Where:

t_s = time to solidify

ƿ = density of the lead alloy

x = ½ thickness of the cross section

L_f = latent heat of fusion

h = mould coating heat transfer coefficient

T_f = alloy freezing temperature

T_m = mould temperature

Using known values of heat transfer rates for the mould cork spray and the steel surface, the latent heat of fusion and the alloy density, it is possible to calculate the freezing time of different parts of the mould. It is also necessary to plug in the temperature and frame-section values, which are exclusive to individual designs. This is particularly important for balancing the freezing rates of the thinner sections such as the wires against the thicker frame sections.

In my opinion (not universal), it is probably better for the frame to solidify a little later than the wires. This would ensure that the tendency to produce defects, due to faster cooling and shrinkage of the grid wires, is mitigated by a liquid reservoir in the frame sections. However, too fast a wire shrinkage combined with a solid outer frame with excessive inner liquid, can suck liquid from the reservoir in the frame to create an internal void at the cross section.

If a condition arises where there is an internal void in the frame and its surface is sufficiently soft due to its temperature, this frame section will be depressed by atmospheric pressure, to give a noticeable depression, or even a crack. If the surface of the frame has cooled to a point where it is more rigid, this will result in an internal void. Either case is not good news.

Clearly, this balancing act is a vital part of the design and operation of a well-engineered grid casting block. However, there is a further operation for book-mould casting: the cork spraying technique.

A significant part of obtaining the heat transfer balances outlined above, is the application of an insulating cork powder to the mould surface. If applied correctly, the thickness of the layer can be varied to adjust the freezing rates of different sections.

A company at the forefront of battery grid casting technology is Wirtz. They have developed cork spray formulations, spraying techniques and operator training courses that are second to none. With the relevant skill level and the right cork mixture, an operator can spray a mould once in a shift and continue casting for a full eight hours, with only minor tweaking of the sprayed surface. This is important, as the significant downtime experienced by many companies due to operators cleaning and respraying a mould two or three times a shift, can seriously reduce productivity.

I hope this article has given insight into the complexity of book-mould gravity casting and the factors to be considered when designing the moulding equipment. A knowledge of the differences between the solidification characteristics of an antimonial and a calcium-based lead alloy is essential in providing a tool design that ensures good quality and efficient productivity.

That being said, the strong reliance on operator skill, the limitations to grid thickness and the number of possible routes to failure with this method, have been a driving force to find a better and more cost-effective route. The next article in this series will examine rolled-strip casting and compare it to the gravity book-mould method.