Improve your vacuum test: the results of a vacuum test for aluminum melt quality are only as good as the accuracy of the test itself.

Of all the customer reasons for returning finished castings, porosity is the leading cause of rejection. The varied causes for porosity include contained inclusion matter, dissolved hydrogen in the melt, shrinkage, air- or mold/core-generated gases trapped in the pouring stream, and combinations of all of the above. Once the true cause of the porosity has been determined, the task becomes one of evolving and implementing solutions through process modifications. This ultimately will include testing to ensure the manufacturing processes remain free of the conditions causing the porosity.

One of the more common problems resulting in porous aluminum castings is the tendency of molten alloys to dissolve hydrogen from a number of sources. Relatively large quantities of hydrogen can be present in the molten metal at normal melting and pouring temperatures, and the excess will be rejected upon solidification of the alloy because of lower gas solubility in the solid metal. In the worst cases, the rejected gas forms large pores in the castings (Fig. 1).

[FIGURE 1 OMITTED]

In addition to being a source of poor surface finish on machined or polished castings, pores caused by hydrogen gas can be a source of poor tensile and fatigue properties. Structural castings requiring high mechanical properties almost always are made to strict standards for low gas content and an absence of detrimental inclusions. Nevertheless, with proper control of melt levels of gas and pore-nucleation sites, hydrogen can be beneficial in minimizing shrinkage in areas of castings that are inaccessible to feeding by risers (Fig. 2). Maintaining hydrogen addition levels at some intermediate level of gas and nucleants can be an important manufacturing tool in aluminum permanent mold and sand casting facilities.

[FIGURE 2 OMITTED]

Vacuum testing has become the standard method for controlling gas levels in castings. The relative ease of operation, low cost of equipment and rapid availability of quantitative results are deciding factors in the adoption of this test procedure. Unfortunately, errors in technique and control of the basic process frequently reduce the value of the results.

Increased awareness and understanding of the process variables influencing the results has the potential to generate the desired level of consistent process control. While current equipment is capable of achieving that level of control when properly calibrated and maintained, new equipment is now available with enhanced automatic capabilities.

Testing Melt Quality

Over the years, metalcasters have repeatedly attempted to relate reduced pressure or vacuum solidification test results on small solidified samples to the actual hydrogen level in melts. While general trends exist, the correlations are somewhat marginal. Vacuum testing methods are usually considered to be capable of estimating hydrogen down to a level of 0.1-0.15 cu. cm/per 100g of sample weight. The problem generally is traced to the quantity, size, shape and type of pore-nucleating inclusions present in the melt (Fig. 3). These exercise control over the number of pores that will develop during solidification of the test sample and have a definite influence on the distribution and size of the gas pores in the castings being manufactured.

[FIGURE 3 OMITTED]

For the wrought and extrusion industry, which requires absolute control of the hydrogen level in its melts, other more sophisticated test methods, such as the recirculating gas, First Bubble, subfusion and vacuum fusion extraction methods, have evolved. All have been evaluated and used for metalcasting applications but generally are used infrequently because of the cost of the equipment and time delays in quantifying batch melting operations.

Many metalcasters, diecasters and some wrought manufacturers have come to the conclusion that the accuracy of hydrogen analysis is not as important as being able to predict the potential for porosity in their final product.

Basic Equipment

The necessary equipment for a vacuum evaluation of an aluminum melt consists of a pump, a gage to monitor the vacuum level of the test chamber and a control system to regulate the test vacuum (Fig. 4). In practice, the molten sample is placed on the pedestal, a chamber with a viewing port is placed over the sample, the chamber is evacuated to the desired level and the sample is allowed to solidify under the reduced pressure.

[FIGURE 4 OMITTED]

The reduced pressure in the chamber causes the dissolved hydrogen to come out of solution in the melt and either escape from the sample through the molten surface or form bubbles in the solidifying sample. Under the low pressure in the chamber, some of the precipitating hydrogen bubbles will float to the surface of the sample and escape. However, with proper control of the pressure level, cooling rate and sample size, the majority of bubbles will form later during the mushy stage of solidification and be trapped inside the sample. Thus, they cause the sample volume to expand. This generates the familiar, visibly-porous cross-sections commonly found in gassy samples and may result in the sample overflowing the cup when high gas levels are present in the melt.

Improving Vacuum Evaluations of Melt Quality

Over the years, a number of metalcasting personnel and university associates have worked to identify the importance of the many variables influencing vacuum testing of aluminum melts. Management of these variables is possible with even the simplest form of testing equipment, but the task can be made more operator-friendly with the use of more advanced available tools. A few basic steps should be taken to provide the best results.

1. Control of Vacuum Levels--One of the most important variables in vacuum testing is the accurate control of the vacuum level in the test chamber. Variations in that pressure level can create large differences in the size of the vacuum-induced porosity in the samples (Fig. 5).

[FIGURE 5 OMITTED]

A problem in test accuracy arises with the differential gages used on most current vacuum testers that measure the departure from current ambient pressure levels. At sea level, the average pressure will support a 30-in. (760-mm) column of mercury. The height of that column will change as high- and low-pressure weather systems move through the area and can vary more than 3 in. (75 mm) around the average level. That variation can be tracked with a manometer and is a common feature of weather reports.

While the terms "vacuum level" and "pressure" are often used interchangeably, it is important to understand the true meaning of the terms. Atmospheric pressure is the load created by the volume and density of the air mass above a specific point on earth, and in North America it is expressed in inches of height of a mercury column. Vacuum levels are not traditionally quantified with reference to atmospheric pressure but rather to an absolute zero pressure. They are expressed as Torr units and are pressure readings measured by the height of a column of mercury in a manometer, one side of which is zero pressure. One Torr is equal to 0.039 in. (1 mm) of mercury (Hg). Thus, 50 Torr is approximately equivalent to 2 in. Hg (30-28 in.), 75 Torr is about 3 in. Hg (30-27 in.), and 100 Torr is about 4 in. Hg (30-26 in. Hg).

When a conventional differential gage is used to measure the pressure level in the test chamber, the chamber pressure is not regulated to a constant absolute zero reference point, but rather to current atmospheric pressure. As noted above, differences in excess of 75 Torr (3 in. Hg) can occur which then will affect the volume of porosity.

The impact of these changes (Fig. 6) upon the process testing can be minimized by making an adjustment to the test chamber pressure that reflects the variation in current ambient pressure. Local barometric pressures can be used to adjust the target gage reading, monitoring the chamber by adding or subtracting the deviation from standard sea level pressure (29.92 in. Hg). Once these corrections have been made, the test reproducibility will be improved, assuming that accuracy of the gage calibration is maintained.

[FIGURE 6 OMITTED]

2. Control of the Vacuum Pump--Maintenance of the vacuum equipment must include all aspects of the system. Thus, good chamber seals, constant pump-down rates and timing of the test to avoid premature release of the vacuum are important variables in achieving reproducible measurements. Further, vibration of the sample during solidification can vary the nucleation and pore size of bubbles. Operators should isolate the pump from the chamber mounting or turn off the pump once the target chamber pressure level has been attained to avoid such vibration.

Selecting the proper vacuum to evaluate individual samples is normally a function of the sample porosity level that is targeted for the castings being manufactured. If high purity is targeted, vacuum levels are chosen that will emphasize small amounts of residual gas in the melt (-28 in. Hg or lower). Conversely, if medium to high gas levels are desired in the end product, less critical vacuum levels commonly are chosen (-26 in. Hg). The objective is to produce a range of specific gravities in the test coupons that will define working limits for the process.

3. Control of Sample Cooling Rate--Since the sample can lose some gas through the mechanism of vacuum degassing, control of sample temperature at the start of the test and the cooling rate through the mushy zone can be critical to the volume of gas retained in the pores formed. Practices should be designed to provide as much uniformity to the start-of-test temperature and sample weight as practical. Optimum sample sizes are generally believed to be 100 to 150 g (3.5 to 5.3 oz.), with a profile that minimizes the temperature range in the solidifying sample. It also has been found that test reproducibility and accuracy benefit from some sample pedestal and chamber preheat, and the sample cup should be insulated from the platform.

Once the sample has solidified, the treatment of the sample and cup are important to the assessment of the percent porosity present in the sample. Neither the cup nor sample should be immersed in water because of the potential for water absorption in the sample and moisture retention in the cup.

4. Observation of the Solidifying Sample--Operators should not ignore the vacuum test's ability to assess the inclusion content of a melt. This may be estimated by watching the surface of the solidifying sample. Melts with a high level of inclusions will produce an early bubbling in the sample as it solidifies. The earlier the bubbles form, the higher the inclusion content in the melt and the less gas retained in the final sample.

For those operators desiring maximum mechanical properties in their product, low levels of gas and inclusion are desired in the melts and in the final casting. This can be confirmed by the near complete lack of bubbles observed during sample solidification and a smooth surface on the top of the sample. If intermediate levels of gas are desired, as noted before, some nominal level of small inclusions will be required to generate a uniform bubble distribution. Melts with high gas levels and inclusion content are undesirable and render achieving a uniform final product difficult. The visual appearance of a range of samples is shown in Fig. 7.

[FIGURE 7 OMITTED]

5. Sample Evaluation--In the past, one of the most common methods of evaluating sample porosity has been to compare a casting's surface to a visual chart depicting a range of sample porosity. This method has multiple problems, including possible variation in bubble density across a sectioned sample and the extremely subjective nature of the visual estimates or rankings of porosity. Comparisons of operator rankings using the visual method have demonstrated a range in estimates of two or three levels, with an additional error possible in the variation of bubble density.

The preferred method of evaluating porosity is to measure the specific gravity of a sample by weighing it in air and water and calculating the value. However, it should be recognized that the specific gravity will vary with shrinkage in the sample, as well as variations in chemical composition of the alloy, and adjustments may be required.

For More Information

"Development of a Statistically Optimized Test Method for the RPT," 1998 AFS Transactions (98-18).

D. Groteke, Q. C. Designs Inc., St. Joseph, Mich.

Steps for an Accurate Vacuum Test Measurement

* Use a thin-walled, coated cup for sample collection

* Skim oxides from the melt surface

* Preheat cup in an adjacent surface area

* Backfill cup to minimize collection of surface oxides

* Collect 4.4-5.2 oz. (125-150 g) sample from the surface maintained at process temperature

* Move sample rapidly to a vacuum unit with good seals and pump capacity

* Place on an insulated pedestal

* Cover with a chamber that permits visual observation of the solidifying sample

* Start the pump and minimize vibration of the sample during solidification

* Control the vacuum to target level adjusted to an absolute vacuum reference

* Observe the sample while solidifying to develop an estimate of metal cleanliness 

* Adjust the time of the cycle to allow for full solidification of the sample

* Upon cycle completion, remove sample and air cool both cup and sample

* Measure specific gravity of the sample and compare to process specifications

* Adjust process as necessary

Daniel Groteke is a past chair of AFS Committee 2G. Aluminum Molten Metal Processing. He may be reached at [e-mail] or via the webpage www.qcdesignsinc.com.