Thermal Analysis - Part 2

Before I get into the ATAS testing I want to show some TA data we collected a few years ago to illustrate how TA data can be used as a metallurgical fingerprint and this fingerprint can be used to compare iron produced during different time periods and different conditions.

Figure 24 shows statistics from a series of 35 non-tellurium cup samples poured from a base Ductile Iron. There is no meaning to the shape of this curve, rather it only provides a convenient means of looking at the average and standard deviation of any number of samples. The curve was produced by entering the average arrest temperatures into a spreadsheet, plotting the points and connecting the points with a smoothed curve. At the time we only monitored four temperatures, the maximum temperature of the sample (MAX), the liquidus (TAL), the undercooling (TEU) and the recalescence (TER). The reason we track the maximum temperature is to get a feel for the consistency of the sampling, which we think has some effect on the results. The next slide, Figure 25 shows the same iron after inoculation with CS-3 and 75% FeSi. We use the standard deviations of each of these temperatures to determine the consistency between the base and final irons. Here it is difficult to tell which is more consistent due to two arrest temperatures having higher standard deviations for the final iron and two arrests having higher standard deviations for the base iron.

The next two slides, Figures 26-27, show the same type of information for a time period using a different inoculant. Here the final iron is more consistent on the average as shown by the lower standard deviations of the arrest temperatures. Figures 28-29 show the same base and final averages as the previous sets of slides, but for a different time period. The final iron is more consistent here as well.

Here on the next slide, Figure 30, the averages from the previous curves are displayed to make it easier for comparison. It appears that the irons poured during 1996 were similar in terms of all four temperatures, but the iron from a period in 1995 was somewhat different as shown by the lower max and liquidus values for the base as well as the final iron.

A few months ago we finished testing with an ATAS unit in five of our member foundries. 1 would like to share some of the preliminary results. Five separate foundries decided to participate in this testing. Four of these foundries make large castings with section thicknesses of 3-4 inches and larger. All of the testing was done with Ductile Iron only.

We initially established some project goals which you see here. We knew from some of our other member's experiences with the ATAS unit that it could determine metallurgical properties of small castings, but we did not have any experience with large castings. Although shrinkage is not much of a problem with large castings, we wanted to see if the ATAS unit could predict when the iron would be more shrink prone.

The next slide shows the test plan we tried to follow for a period of six weeks. The first thing we did after training melting personnel in how to identify samples was to pour a lot of samples to build a history of normal iron. We also poured test castings at some of the foundries.

Before showing the data I want to review a few definitions of acronyms so that everybody will not wonder about the terms they will see on the next few slides.

One of the foundries participating in the test program tested the ATAS unit for two different time periods. This gave us the opportunity of comparing the iron fingerprint for both of these time periods to see if anything had changed. Figure 31 shows that irons from these two different time periods were essentially the same.

Figure 32 shows the results from base samples taken at each foundry. Notice there is a similarity among the foundries in terms of averages, but some of the foundries were more consistent than other as measured by their standard deviations.

When we looked at base parameters compared to treated parameters we were interested in consistency again. As you can see from Figure 33 which shows the details from one of the participating foundries, some of the treated iron parameters were more consistent and some were less consistent. if we had taken fully inoculated samples we would have expected parameters that were more consistent than either those from base or treated irons.

The next slide Figure 34 shows the treated Ductile results from four of the five foundries that participated in the testing program. One of the foundries did not pour any treated samples. It is important to note that most of the inoculation for this iron was from in-the-mold inoculants and our sampling was from freshly treated iron. The rightmost column shows the target levels for these selected parameters for shrink and chill free iron according to the ATAS documentation. As can be seen some foundries fared better than others in attaining the target values.

The test phase is now complete, there is some analysis work yet to be done before we can make our final conclusions. We learned much in terms of metallurgical consistency of the iron and look forward to working with this unit more in the future.

We learned

1. TA data can be used to compare the metallurgical condition of iron from one time period to another.
2. New charge materials, alloys and inoculants can be evaluated using TA data.
3. There are metallurgical differences in irons produced at different foundries.

To make the most of thermal analysis as a metallurgical tool we recommend collecting process information that can be correlated with TA data. This means that foundries should be monitoring and storing this information electronically so that when changes are noticed, time will not be wasted entering data into a computer; the data will already be there. A list of the minimum type of information that is necessary for this effort can be seen in this slide.

In conclusion, we believe that thermal analysis is another tool that foundry men can use to control the metallurgical consistency and properties of irons they produce. We have encouraged our members to incorporate TA in their process controls for many years and with the development of the ATAS and other TA products we are renewing our efforts to encourage the use of this technology.

Part 3 of Thermal Analysis
Suggestions for Improved Reliability in Thermal Analysis of Cast Irons