Electrical conductivity in aluminum: possible alternative to thermal analysis.

Electrical Conductivity in Aluminum: Possible Alternative to Thermal Analysis 

In recent years, thermal analysis has become popular as a method to assess the grain size and degree of eutectic modification in aluminum-silicon casting alloys. In applications that do not require information about grain size, measurements of electrical conductivity can provide the same degree of control of the eutectic structure at substantially reduced costs.

Surprisingly overlooked as a testing procedure, the method is so simple that a foundryman can purchase a standard piece of off-the-shelf equipment for a modest cost and begin to use it immediately, without a great deal of instruction, to accurately inspect modified aluminum castings.

The Principle

In electrical conductivity testing, the ease with which electrons flow through a cast structure can be measured. That is because most metals are excellent conductors of electricity, and microstructural changes have only a very small effect on their electrical properties.

Aluminum-silicon alloys are somewhat of an exception since they contain large quantities of almost pure silicon, which is a semi-conductor with an electrical conductivity about one millionth that of aluminum. Further, the conductivity depends on how the silicon is arranged within the alloy, i.e., on the microstructure of the alloy.

Unmodified alloys contain silicon in a coarse, plate-like form (Fig. 1a) while the addition of a modifier, such as sodium or strontium, causes the silicon to solidify as a fine fibrous network (Fig. 1b). In 1959, German researchers showed that a modified alloy possessed a higher electrical conductivity than the same alloy in unmodified form. [1] This effect has since been verified by others, [2-4] and it has been conclusively demonstrated that the conductivity changes are due to microstructural changes brought on by modification. [5]

A simplified view of how modification affects electrical conductivity is presented in Fig. 2. In the unmodified alloy, the large plates of eutectic silicon impede the flow of electrons, which bounce off the nonconducting plates and some even try to flow in the opposite direction. The result is an alloy with a low conductivity. When the alloy is properly modified, the silicon plates change into fibers, which present a much smaller impediment to electron flow, and the conductivity is increased.

The Equipment

In the foundry, the simplest way to measure electrical conductivity is to use the eddy current technique. In this method, a small electric current is generated within a volume of metal by a hand-held probe placed on the surface of a solid sample. Figure 3 is a photograph of a typical commercially-available testing device which gives the conductivity expressed as a percentage of the International Annealed Copper Standard (% IACS).

These instruments are not expensive, and are available from several sources. To our knowledge, none has been built specifically for aluminum-silicon alloys, but those designed for use with nonferous metals work well. The depth of penetration of the induced electric current can vary from instrument to instrument. In order to avoid surface effects, a depth of penetration of from 0.25 in. and 0.5 in. is recommended.

The Method

Figure 4 shows how the electrical conductivity of A356 alloy changes with strontium concentration. As the strontium level increases, modification of the eutectic occurs and the conductivity rises. There is an increase of about 10% from the unmodified to the modified state. Smaller increases of about 5% are associated with undermodified structures.

To assess the degree of modification, buttons were cast two inches in diameter by three inches thick in an insulated graphite mold. One button is cast before the modification treatment is done, the second afterward.

Evaluation of the modification is based on the difference in electrical conductivity between these two samples. A difference of approximately 10% implies a well-modified structure and a successful treatment. Smaller difference, on the order of 5%, means that the structure is only partially modified and that more modifier should be added. Of course, if there is no difference in conductivity, the modification treatment was unsuccessful.

No special surface preparation of the sample is required. Because the current penetrates at least 0.25 in. into the sample, the condition of the surface is relatively unimportant. The method works well on both sand cast and permanent mold cast surfaces. Provided the melt is reasonably gas-free, porosity is also not a problem. For a more detailed discussion of the factors that can influence the electrical conductivity, refer to the article by Argo, Drew and Gruzleski. [3]

The button dimensions are not critical as long as they are large enough to avoid edge effects. [3] In addition to graphite molds, permanent copper molds, usually employed for taking spectrochemical samples, have been used with complete success.

The most important part of sampling is to ensure that the sample always cools at the same rate after it has solidified. This is necessary because the quantity of alloying elements retained in solid solution has an important bearing on the electrical properties. The sample can be allowed to cool in air to room temperature before testing, or it can be quenched in water at a fixed time after it has been poured. The time is not important, but it should always be the same.

Examples

In order to demonstrate the use of this technique, we have treated A356 melts with sodium to produce well modified, overmodified and undermodified structures. This was done by making an excessively large sodium addition, and then degassing with nitrogen for 20 minutes to remove some of the sodium. An additional three hours was allowed for sodium fading to take place.

The evolution of electrical conductivity with time, of samples poured from this melt, is presented in Fig. 5. The results are summarized in Table 1, and the microstructures of samples taken at six times throughout the melt treatment are shown in Fig. 6.

When the sodium level is approximately 0.01%, the microstructure is well modified and increases in conductivity of 8-10% are observed. Partially modified and overmodified structures result in increases of from 5-6%. It is not possible to distinguish these two structures, but it is possible to tell that they are not well modified.

The method works equally well when strontium is the modifier. A nominal addition of 0.025% Sr was made to an A356 alloy, and samples were poured at intervals up to two hours after the addition. Table 2 summarizes some of the conductivities obtained and the microstructures observed.

Once again, partial modification leads to an increase of about 5-6% in conductivity, while changes of 10% are indicative of full modification. It is interesting to note that the highest conductivity occurred about two hours after the strontium addition. This is a reflection of strontium's incubation time in the melt, and its resistance to fading.

Summary

Electrical conductivity measurements can provide the aluminum caster with a quick check on the effectiveness of a modification treatment. Eddy current testing equipment is inexpensive and readily available.

The method operates on the basis of a difference in conductivity between a sample cast from an untreated melt, and one cast from a modifier-treated melt. Since the comparison is between samples cast from the same basic alloy chemistry, compositional variations within the alloy specification are not important.

This article deals with tests based on the very popular 356 alloy type. The method, however, is applicable to many other aluminum-silicon alloys. As the silicon concentration increases, more eutectic appears in the alloy microstructure and the differences between modified and unmodified structures are enhanced. [6]

References

[1] W. Patterson, D. Ammann, "Die elektrische Lietfahigkeit veredelter und unverdelter Aluminium-Zilizium Legierungen," Aluminium, pp 139-140, Mar 1959.

[2] S. Jacob, A. Remy, "Conductivite electrique et morphologie du silicium dans l'AS13 et l'AS7G," Fonderie-Fondeur d'Aujourd'Hui, vol 22, pp 33-41 (1983).

[3] D. Argo, R. A. L. Drew, J. E. Gruzleski, "A Simple Electrical Conductivity Technique for Measurement of Modification and Dendrite Arm Spacing in Al-Si Alloys," AFS Transactions, vol 95, pp 455-464 (1987).

[4] H. Oger, B. Closset, J. E. Gruzleski, "Characterization of the Eutectic Microstructure in Al-Si Foundry Alloys by Electrical Resistivity," AFS Transactions, vol 91, pp 17-20 (1983).

[5] M. H. Mulazimoglu, R. A. L. Drew, J. E. Gruzleski, "The Effect of Strontium on the Electrical Resistivity and Conductivity of Aluminum-Silicon Alloys," Metallurgy Transactions, vol 18A, pp 941-947 (1987).

[6] M. H. Mulazimoglu, R. A. L. Drew, J. E. Gruzleski, "The Electrical Conductivity of Cast Al-Si Alloys in the Range 2 to 12.6 wt. Pct. Silicon," Metallurgical Transactions, vol 20A, pp 83-389 (1989).