| Oxidation Control in Vacuum Furnaces Question: Why are we seeing so much discoloration on the parts we nickel braze in vacuum at 2000°F (1093°C)? We are brazing 304stainless parts with Nicrobraz LM (BNi-2) filler metal. We know the parts were clean to begin with, and we operate the fumaceat2x 10-5 Torr prior to cooling the parts in argon. Answer:
1. Furnace Cleanliness As commercial loads are processed through successive furnace cycles, outgassing will gradually cause contamination of the furnace walls and cage structures. We have seen that some of this contamination can later re-deposit onto parts being run in subsequent cycles (especially when later cycles are run at higher temperatures), causing discoloration of parts and even inhibiting brazing action! In some cases, these outgassing occurrences will be identified during the cycle as unexpected "spikes" that are observed (or recorded) by one of the vacuum gauges. Such "virtual leaks" are easily preventable. This is why most commercial shops do periodic "furnace clean-up runs" to eliminate contaminants. Don't wait until problems arise before implementing your clean-up program. A typical clean-up cycle (with no load in the furnace) might be 2100°F (1150°C) for 2 hours at temperature. Some shops then lower the furnace to 1600°F (870°C) and hold it there for an additional 8 hours. Another suggestion is to run a 2-hour clean-up cycle at a minimum of 1 00°F (55°C) above the highest furnace operating temperature you plan to use for your upcoming heat treating or brazing work. Running "hot leak-up rates" during these clean-up cycles will insure that clean-up is complete. 2. Vacuum leak-up rate When a furnace begins to pump down, it must be sealed to prevent outside air from leaking into the chamber. However, it is not uncommon for some vacuum furnaces to have enough pumping capacity to easily bring the furnace down to the desired vacuum level, even with an air leak. For example, a few years ago, two of our customers were vacuum brazing similar parts, one at 10-5 Torr (which produced badly discolored parts) and the other at 10-3 Torr (these parts emerged bright and shiny). The furnace pulling the "harder" vacuum (10-5 Torr) had an air leak in a door seal, whereas the "softer" vacuum (10-3 Torr) furnace had no leaks. The one furnace was able to pull a "hard" vacuum in spite of the leak because it had such powerful pumps. Yet it was literally "whistling" air through the furnace, and all that air carried oxygen which oxidized (discolored) the parts. The 10-3 Torr furnace run did not have that problem and therefore was able to keep the parts clean. Furnace Leak-Up Rate (FLUR) is a critical quality control factor that every vacuum brazing shop should monitor regularly. Some shops monitor FLUR at the start of each day, others once a week, etc. You will need to determine the best frequency for your own plant operations. The test should be run on a clean, empty, pumped-down furnace. Helpful Hint: An excellent tool to monitor potential problems in the brazing furnace chamber when using high-temperature filler metals, such as nickel
base alloys, is the 304 stainless T-specimen, shown in Figure 1. This is a small, very inexpensive QC tool that should be placed in every furnace load. (See discussion in Summer 1992 issue. A data sheet and order form are included with this issue.) Because the T-specimen is very sensitive to the presence of oxygen, it will quickly show if you have a problem with the furnace, the atmosphere, the base metals being brazed, or with the brazing filler metal. Figure 1 shows two T-specimens that have been exposed to unsatisfactory furnace conditions. The one in the background was exposed to a furnace leak. The surface is discolored and the brazing filler metal has not flowed. If the base metal parts being brazed are in the same condition, then the T-specimen is showing you that there is a problem with the furnace (a leak?) or the atmosphere (high oxygen?) during the heating cycle. The T-specimen in the foreground was exposed to a slightly oxidized atmosphere after brazing was complete. Even though the surface is discolored, the filler metal has flowed, which indicates a problem during the cooling cycle; perhaps a possible air leak in the argon cooling system. We recommend placing one T-specimen into each and every furnace load. This ensures that if trouble does occur in that load, the conclusions you draw from examining your "on-site QC inspector" will be highly accurate. 3. Vacuum operating levels The article on pp. 3-4 of this issue discusses the effects of high vacuum on the outgassing of various metals. A vacuum that is too "hard" can hurt your brazing operation rather than help it! You do not need to operate at 10-5 Torr for good brazing. As discussed earlier, perfectly good results are common in the 10-3 to 10-4 range. The key is leak-up rate, not the absolute level of vacuum being shown on the gauges! If you do not have an ongoing program for monitoring FLUR in your plant, request our data sheet entitled "Monitoring Vacuum Furnace Leak-Up Rates." |
Your question suggests three areas that should be investigated: (1) furnace cleanliness, (2) vacuum leak-up rate, and (3) vacuum operating levels.