RPT measures hydrogen gas, effects on casting quality.
It is well known that liquids absorb gases as illustrated by the carbon dioxide contained in a bottle of beer. Similarly, liquid aluminum has a high affinity for hydrogen gas, entrapping it in the form of internal pores and surface blisters when the metal solidifies. Too much hydrogen in aluminum can seriously degrade the casting quality.
The principal source of contaminating hydrogen is from water vapor or steam present in the atmosphere in and surrounding the furnace. Moisture is present in refractories, scrap and a host of other environmental sources, even the crew servicing the furnace. Because excess hydrogen in the melt causes casting defects, knowing the amount of the gas present in a melt before castings are made is an important economic concern to the aluminum foundryman.
Hydrogen gas solubility in aluminum rises as the temperature of the aluminum rises. As the temperature of molten aluminum falls, unstable nascent hydrogen is rejected from solution and subsequently combines to form molecular hydrogen gas bubbles.
The rapidly evolved bubbles become entrapped within castings in the form of small, usually visible, voids as solidification of the metal proceeds. If the quantity of gas is high, the top of the casting will erupt in blisters caused by the enormous pressure exerted by the hydrogen as it is ejected by the metal. Evaluating levels of hydrogen in liquid aluminum is an important quality control device for foundrymen casting aluminum parts.
Reduced Pressure Testing
After several decades of use, the reduced pressure test (RPT) remains one of the most widely used procedures to prevent potentially damaging hydrogen porosity in aluminum alloy castings. This is because the required equipment is inexpensive, durable, quick and simple to use. The results, when used with regularly calibrated RPT systems, are easily repeatable and correlate consistently with casting quality.
Three elements directly determine the utility of test results:
* the quality and temperature of the test sample;
* exact vacuum control during solidification;
* consistent interpretation of test results.
The test sample must be representative of the melt in terms of limiting turbulence and inclusions during sample preparation and pouring. Limiting the temperature loss in the sample from furnace to vacuum chamber, assuring that bell jar chamber seals are tight and vacuum gauge readings are constant and accurate are important to RPT accuracy and repeatability from melt to melt.
Reduced pressure tests have qualitative and semi-quantitative evaluation capabilities. The former uses visual means to evaluate surface or section porosity and dye penetrant to detect subliminal porosity. The latter applies bulk density and specific gravity methods to determine the volume of voids present. Qualitative results are less specific with the simple RPT equipment, but still reasonably accurate to support the results achievable with this test.
The main objective of the qualitative aspect of RPT is to quickly and easily gauge the effect of entrained hydrogen in an aluminum melt. It is simply a matter of observing the surface bubble formation or the interior porosity of an aluminum alloy sample under controlled reduced pressure conditions. These observations can be correlated to specific foundry-acceptable standards for bubble or pore formation in an alloy sample.
The results are an indication of the amount of hydrogen in the melt and provide guidance in mitigating the porosity damage to castings that might occur if the gas content in the melt was left unmanaged. There are tests that are more precise evaluators of hydrogen content but their cost is considerably higher than that for a RPT system and they require longer test times and more skill to complete.
The basic visual bubble test involves placing a molten aluminum alloy sample in a bell jar, pulling a specified vacuum and observing any surface bubbling or blistering occurring as the sample solidifies. This test can be used in alloys with a small amount of gas present; 2-5 bubbles on a relatively flat sample surface indicate a low level of porosity. If the sample blisters and mushrooms up amid a large number of bubbles, a high level of hydrogen is indicated. A set of 3-4 separate gas test levels are possible in this simplest RPT analysis.
A visual comparison test, a step beyond the simple visual test, involves the same test procedure without the obligation of watching the sample solidify. Instead, the solidified sample is cut in half and examined for porosity as shown in Fig. 1. The bottom section shows gross porosity, some porosity was formed in the sample second from the bottom and the top two samples exhibit no visible bubbles. This kind of test is useful when high levels of gas are suspected.
Another step beyond the visual comparison test involves finishing the sectioned samples shown in Fig. 1 using 125 grit paper. This eliminates the sawing roughness and makes the sectioned surfaces easier to inspect for porosity. Figure 2 offers a clearer view of the bubbles, the resulting hydrogen porosity and indicates some probable, localized shrinkage in the top specimen.
A constant vacuum level during RPT is critical to the usefulness of the test. A comparison of different levels of reduced pressure on samples from the same melt are illustrated in Fig. 3. All of the samples contained the same amount of hydrogen but the differences caused by variations in vacuum are evident in the wide variations in their physical appearance.
When working with alloys with a low level of gas, subjecting the sectioned samples to dye penetrant testing tends to show porosity otherwise nearly undetectable. Figures 4 and 5 illustrate how porosity shows up under dye penetrant inspection. Figure 4 shows even distribution of pores, while Fig. 5 shows a small concentration of pores possibly caused by shrinkage induced by the lack of gas in the sample.
The use of RPT to quantify the amount of hydrogen in a sample is effective as a measure of the degree of porosity when done by calculating the density of a sample and correlating that with the dissolved hydrogen content of the melt. This produces accurate information as long as the test is standardized and carefully controlled, though the results are semiquantitative at best. This is due in part to the bubble nucleation effects of oxides, inclusions and alloy modification, plus the hydrogen lost during solidification.
RPT can be an effective quality control tool if the if the influencing variables are controlled and understood. A constant vacuum level, of course, is a requisite for the most consistent results. Frequent calibration of the vacuum gauge and maintenance of the vacuum pump and regulator and the bell jar seal are important because constant use of the test system in a hostile environment can alter results. Another factor affecting accuracy is sample cleanliness, metal free of oxides from the melt surface. The sample crucible should be clean, preheated and coated with core wash. The sample should be free (as possible) of shrinkage porosity lest it bias the test results.
RPT measurements can be used to make statistical process control (SPC) charts if the testing conditions are controlled and held relatively constant. What is important in this process is to arrive at a number that can function as a sample constant. This constant provides a base for a statistical control chart from which deviations can be recognized. The foundry should strive to get the best information from actual test results to establish its statistical chart rather than make up arbitrary or speculative porosity specifications.
The bulk density method is done by simply weighing the known volume sample and dividing it by its volume (weight per unit of volume), arriving at gr/ml or lb/cu in. (density). There are several ways to measure the bulk volume, the simplest being the water displacement method in which the sample is submerged in water and the displaced water measured. The only equipment needed is a graduated cylinder. In contrast to the simple displacement method, which has an accuracy of |+ or -~1%, is a $3000 helium system requiring a laboratory. Accurate to 0.01%, a test takes about five minutes.
Specific gravity is another quantitative method that is simple and quick. The sample is weighed in air and then in water; the weight in air is divided by the weight in air less its weight in water. All that is needed is a platform scale with a suspending arrangement and a water tank below it. It is accurate to about 5% but it must be recognized that accuracy is influenced by water density. There is about a 2% change in water density from 40-100F.
It is suggested that foundries using this method try to regulate water baths to a constant temperature, and avoid suspended solids in the bath because they can affect sample buoyancy. It also is suggested that the water bath be changed frequently and that distilled water be used rather than deionized water because of the large number of suspended solids in the latter. Because the standard density of water is 1 gm/ml, there is a relationship between specific gravity and bulk density.
The porosity method requires taking a sectioned sample, rough grinding and then polishing it with a 120 grit paper prior to examination and rating or measuring the area fraction of voids in the cross-sectioned piece. The assumptions are that the voids in the sample are somewhat spherical, evenly dispersed and macroscopic, or visible. The interior voids are 3-D as opposed to surface porosity that is 2-D, but the numbers and size of surface porosity on the cross section are related by extension to the porosity existing in the entire sample.
Thus, by using an SQC standards chart of acceptable porosity, cross section porosity can be correlated to the contained hydrogen in any given melt. This requires a visual estimate, but the human eye is a good integrator of visual information. From person to person, evaluation of the same set of sample standards would vary by not more than 10%.
By way of comparison to visual estimation of alloy samples, Fig. 6 illustrates an expensive (roughly $45,000) image analyzer. It requires examining several fields of view, but it is an extremely accurate method of calibrating standardized SQC charts for any foundry's melt practices.
There is a relationship between density and porosity and the difference between the two can give a close approximation of the amount of internal hydrogen porosity present in a melt. The product of the bulk density or specific gravity divided by the nominal (theoretical) density or specific gravity gives a percentage of volume porosity as noted in the equation in Table 1. This would indicate that a plot of the voids in a sectioned sample correlates to the total volume of voids in a sample.
It should be noted that the nominal density of a sample will vary according to the chemical composition of the aluminum alloy being cast. If a foundry is producing a melt requiring a particularly low hydrogen gas level, the nominal specific gravity value must also be changed to take into account the chemistry of the alloy being processed.
Table 1. The porosity and density relationship is evident in this equation for determining the percentage of volume porosity. P = |1-(D bulk/D nom)~ x 100 P = % VOLUME POROSITY D bulk = BULK DENSITY OR SPECIFIC GRAVITY D nom = NOMINAL (THEORETICAL) DENSITY OR SPECIFIC GRAVITY