- THE MAGAZINE
As an extruded or cast ceramic body with 15 to 20% water dries, the body loses water and shrinks. It is relatively easy to measure the dimensions of the shape with a ruler and to measure its weight with a scale. These manual measurements can be plotted as a function of time. The resulting curve would show that the body ceases shrinking at some point while it continues to lose weight. Eventually the body ceases losing weight and is dry.
A simple drying test can be performed by placing a wet sample into a heated chamber and measuring its dimensions and weight at certain intervals as the temperature of the chamber is changed to simulate a production drying schedule. Every time a data point is taken, the sample must be removed from the dryer, measured and returned to the dryer. Figure 1 is a graph of such a test using a wet, extruded structural clay brick. Note how the shrinkage progresses quickly and ceases while the weight continues to drop, as predicted by drying theory.
From drying theory, we know that the temperature, the percent relative humidity and the air velocity (speed and direction) all influence the rate of drying. These conditions are present in every drying chamber, and should be recorded every time the dimensions and weight are measured. However, not every heated chamber can monitor the percent relative humidity and air velocity, so these critically important factors may not be known.
The ability to change and control the factors that influence drying rates is helpful in studying, characterizing and optimizing drying cycles. Many conventional heating chambers or dryers cannot alter or control the relative humidity and air velocity, so they limit the ability to characterize drying. Instead, a more sophisticated piece of equipment, such as a TGDA dryer, is needed to gain a better understanding of the drying process.
The TGDA DryerThe TGDA dryer is a laboratory instrument designed to quantify the drying characteristics (shrinkage and weight loss) of a ceramic body when exposed to controlled temperatures, humidities and air circulation. The dryer combines a thermal gravimetric analyzer (TGA) and a thermal dilatometric analyzer (TDA) into one unit that simultaneously and continuously monitors the weight loss (DW) and dimensional change (DL) of a ceramic body during a controlled temperature and humidity drying cycle.
The body to be tested (an extruded structural clay brick, a cast whiteware slab or a section cut from a larger formed shape) is placed on the sample pan inside the sealed chamber. This pan is suspended from an electronic balance that monitors the weight change of the shape. The linear variable displacement transducer (LVDT) probe rod contacts the shape to monitor its dimension change. One thermocouple is placed on the surface of the sample, and another inside. A PID controller is programmed to control the chamber air temperature, the chamber humidity, and the air circulation as a function of time.
Once the run has begun, the signals from the electronic balance, the LVDT, the three thermocouples and the relative humidity sensor are continuously saved in digital format for later analysis. After the run is complete, the operator examines the shape for cracks and dries the sample in a separate chamber at a higher temperature to determine the final “bone dry” size and weight.
Examining TGDA DataFigure 3 is a typical graph of TGDA data. The sample was an electrical porcelain, and the run was one of a series from a project to reduce the drying time of a 72-hour production cycle. The data from the run was imported into a spreadsheet and plotted using the spreadsheet’s dual Y-axis graph application. The three temperatures (chamber, internal and surface) and relative humidity are plotted on the left Y-axis, the percent length and weight changes are plotted on the right Y-axis, and both Y-axes are plotted against time on the X-axis.
The first eight hours were programmed for a slow drying rate. The drying conditions were then abruptly changed (the air temperature was increased and the percent relative humidity was reduced) to observe how the body would respond to an accelerated drying environment.
Measuring Weight LossIn Figure 4, the percent weight change is plotted against time, along with its first derivative, the rate of change of the weight loss. The first derivative was a simple rolling average algorithm that was normalized for the surface area of evaporation. Even though this simple arithmetic technique generated a jagged first derivative curve, the nature of the curve was apparent. It was not necessary to make the curve aesthetically clean by hand drawing a smooth curve, or by applying a smoothing technique that shifts the curve or diminishes the magnitudes of the body’s behavior.
The first observation was the slow rate of weight loss (water loss) at the beginning of the run when the drying conditions were mild. Immediately upon changing to more aggressive drying conditions (an increase in temperature and decrease in relative humidity) at 8.3 hours, the rate of water loss increased. A vertical dotted line at 8.3 hours highlights the time when the chamber conditions were abruptly changed.
As the run progressed, the sample continued losing weight at a faster rate. The rate of weight loss reached a maximum during the 10 to 11 hour period. Another vertical dotted line at 10.5 hours highlights the time period when the sample was losing water at a maximum rate. After hour 11, the sample continued to lose weight but at a decreasing rate. By hour 18, the rate of weight loss was almost zero.
For clay bodies with the normal amount of forming water (about 15 to 20%), an old “rule of thumb” for drying is that the shrinkage is complete when the body loses half its water. Since the final bone-dry weight was measured after the test, it was easy to determine that the 50% weight loss amount was approximately 8.25%. A dotted horizontal line at -8.25% was added to the graph, along with a dotted vertical line at 10.9 hours where this occurred. Note that the 50% weight loss occurred at the end of the maximum rate of water loss period.
Measuring ShrinkageIn Figure 5, the percent length change of the piece is plotted against time, along with its first derivative, the rate of change of the shrinkage. This first derivative was the same, simple rolling average algorithm used for the rate of weight change determination. Even though this simple arithmetic technique generated a jagged first derivative curve, the nature of the curve was apparent. As with the weight loss measurements, it was not necessary to make the curve aesthetically clean by hand drawing a smooth curve, or by applying a smoothing technique that shifts the curve or diminishes the magnitudes of the body’s behavior.
The first observation was the slow rate of shrinkage at the beginning of the run when the drying conditions were mild. Immediately upon changing to the more aggressive drying conditions (higher temperature and lower relative humidity) at 8.3 hours, the rate of shrinkage increased. A vertical dotted line at 8.3 hours highlights the time when the chamber conditions were abruptly changed.
As the run progressed, the sample continued shrinking at a faster rate until it reached a maximum just before hour 10, the beginning of the maximum rate of weight loss period. After hour 10, the sample continued to shrink but at a slower rate. By hour 11, the end of the maximum rate of weight loss period, the rate of shrinkage approached zero. For reference, the vertical dashed lines at hour 10.5 (mid-point for the maximum rate of water loss period) and hour 10.9 (50% of total weight loss) were added to the graph.
By hour 11, the body had lost half of its forming water, and the shrinkage was essentially complete.
Surface and Internal TemperaturesFor the first eight hours of the run, the heat transfer from the surface of the sample to the interior was fast enough so that no temperature difference existed between the surface and the interior of the sample. The surface and internal temperatures were identical for this initial, slow period of drying. At hour 8.3, when the drying conditions changed abruptly, the surface and internal temperatures also changed abruptly.
Figure 6 is an enlarged version of the graph in Figure 3 that shows hours 7 through 12. Note how the surface and internal temperature curves are identical up to hour 8.3, and how they are almost parallel to the chamber temperature curve. The rates of temperature change for all three are practically the same. At hour 8.3, the chamber temperature abruptly increased, and the rate of temperature change dramatically increased, as seen by the dramatic change in slope. Shortly after hour 8.3, the rate of temperature increase of the surface and internal temperatures increased, but not as fast as the chamber temperature.
Evaporative drying is an endothermic process that requires energy from its surroundings. In the initial stage of drying, evaporation occurs on the surface of the body, where there is a continuous film of water. The energy for evaporation is taken from the surrounding air, from the liquid in the body, and from the grains in the body. As the rate of evaporative drying increased, the rate of weight loss increased, and there was an increased demand for energy to support this faster evaporation rate.
The evaporation phenomena that occurred on the body surface were the same that occur on a wet-bulb thermometer. Evaporation on the surface of a wet-bulb thermometer takes energy from the liquid in the wick surrounding the bulb, thus reducing the temperature of the liquid surrounding the bulb and the temperature of the bulb itself. The evaporation on the body’s surface took energy from the liquid on the surface and reduced its temperature. The temperature inside the body increased even slower, as seen by the gap between the surface and internal temperatures. Even though the surface and internal temperatures were still rising, the rate of temperature rise was much lower than the surrounding air due to the evaporative cooling effect. This cooling effect was so strong that during the period of the highest rate of weight loss, hours 10 to 11, the rate of increase of the surface and internal temperatures slowed to the extent that the surface and internal temperatures remained constant.
Once the weight loss rate and demand for energy began to slow, the temperatures of the surface and the interior began to rise again to close the gap with the chamber air temperature.
Converting TGDA Data to the Kingery FormatFigure 7 is the TGDA graph with vertical dashed lines at the critical time periods. The percent weight and length changes are based upon the initial wet conditions, and they are plotted as a function of time. The drying process progresses from left to right.
The simple, arithmetic conversion process for determining the rate of moisture loss generated an erratic curve. Despite the jagged appearance, the trends were visible, so no sophisticated mathematical smoothing routines were applied. Instead, manually drawn straight lines were added to highlight the trends.
Starting at the right, note the relatively low rate of water loss at the beginning of the process when the drying conditions were mild. At hour 8.3, the chamber conditions abruptly accelerated, and the rate of water loss abruptly increased at a linear rate. This was the time period when the gap between the surface temperature and the chamber temperature increased.
As the drying process continued, the rate of water loss increased linearly until the 10 to 11 hour period. During hours 10 to 11, the body was losing weight at approximately the same maximum rate. The horizontal line highlights this period of constant rate of moisture loss, which is defined as stage 1 drying. This was the same period as the maximum rate of weight loss, and the surface and internal temperatures appeared to remain constant.
Stage 1 drying is the critical stage of drying, since it is where the majority of shrinking occurs. By hour 10.5, the body had shrunk from about 3% to less than 0.1%, and the shrinkage was essentially complete.
After 11 hours, the rate of water loss began slowing and the internal and surface temperatures began rising. After a brief transition period, the rate of water loss declined linearly. This linear falling rate period is highlighted by another straight line, and is defined as stage 2 drying. The amount of shrinkage was negligible, which is typical of stage 2 drying.
Test ResultAfter 18 hours, the TGDA stopped collecting data. The shrinkage was complete, no cracks were visible and less than 1% water remained in the body. The sample was considered dry enough to proceed to the kiln for firing.
This drying test showed that by carefully monitoring and controlling the parameters of the air temperature, air circulation and percent relative humidity, the body could safely withstand a drying schedule that was reduced from 72 hours to 18 hours. Additional TGDA tests that adjusted these same parameters showed that the drying schedule could be shortened even more.
Optimizing the Drying ScheduleBased upon a series of runs using green, structural clay and electrical porcelain bodies with normal amounts of forming water, a general method of shortening drying schedules has been found. First, the body should be heated without drying to improve the ability of water to migrate through the structure to the surface. A critical parameter is to keep the percent relative humidity of the chamber air high so that evaporation from the surface is minimized. Once the body is heated so that the water inside it is less viscous (more fluid), drying should begin by increasing the chamber temperature and reducing the percent relative humidity. Once the body has lost approximately 50% of its forming water, the critical moisture content has been reached, and the majority of the shrinkage has occurred. After the 50% total weight loss, the drying cycle can be significantly accelerated without damaging the product’s shape.
However, developing an optimized drying schedule requires more than just a basic understanding of the drying process. Instead, a precise method of observing and measuring the effects of drying on the ceramic body is needed. Tools such as the TGDA dryer can provide this level of control. By manipulating and monitoring the air temperature, air circulation and percent relative humidity of the chamber air, the TGDA can characterize the nature of a ceramic body during the drying process.
While the TGDA can indicate optimum conditions for a fast, safe drying schedule of a single shape, the actual production cycle must take into consideration the differences in temperature, relative humidity and air circulation conditions at various positions inside the dryer chamber, or inside the setting pattern of the shapes to be dried. A series of runs with the TGDA using a range of test conditions will reveal how sensitive a ceramic body will be to the real-world differentials inside the dryer or setting pattern. With this information, the user can determine if changes to the body, the setting pattern, the dryer schedule or the dryer equipment are required to solve dryer problems or increase output.