Fortunately, the problem is usually quite controllable, and regular furnace inspections can usually keep those leaks completely under control.

Leaks most typically occur through some of the sealing-surfaces in the furnace, the most common leak-source being the O-ring seal-surface in the furnace door itself. As the door is opened and closed everyday, the light coating of vacuum-grease on the door seal and on the O-ring can pick up dust and dirt, which, if not properly removed regularly, might begin to initiate small holes/cracks in the O-ring seal, which can eventually open up enough to allow air to start to leak into the furnace during furnace operation. Additionally, seals between the furnace and the pumps, valves, pipe-connections, etc., all can become sources of potential air-leaks into the furnace.

Rarely is the furnace material itself at fault, but it cannot be overlooked. It is possible that some gas/air might get trapped between double-welds used in the initial building of the chamber, which could then very slowly outgas during subsequent furnace operation. Another potential source of difficult-to-find leaks might be any cast-metals that are used in the construction of furnace-flanges or other components that need to be vacuum-tight. Some aluminum castings used in flanges have apparently had this issue in the past.

As mentioned above, all vacuum furnaces will leak air into the vacuum chamber from the outside atmosphere in the factory during vacuum-brazing operations, and then the pressure inside the furnace will start to go back up towards atmospheric pressure. That is why it is a called the “leak-up” rate of the vacuum chamber, and must be monitored, recorded, and controlled on a regular basis.

Typically these leaks are very small. In fact, they are so small that they are measured in millionths of atmospheric pressure leaking into the furnace over a one-HOUR period! That’s right, only millionths of atmospheric-pressure per hour! Yes, the leaks are indeed tiny.

Important: All vacuum-brazing furnaces must be regularly checked for vacuum-leaks, and the “leak-up rate” determined and controlled!

There are a number of “dimensions” that folks use to specify this leak-up rate, but here in North America it is most typically measured in “Microns per hour” (often also referred to as “millitorrs per hour”, since the word “micron” and “millitorr” are used interchangeably for the same measurement).


Leak-up rates are typically performed after a furnace “clean-up run” (furnace bake-out), in which an empty vacuum furnace is put through it’s normal high-temperature clean-up/bake-out cycle, then allowed to cool to room temperature, and the leak-up test then performed prior to opening the furnace.

To perform a leak-up test (usually conducted at room temperature when the furnace has cooled), the furnace is literally being asked to “hold its breath”, so to speak. The furnace valves are all closed, and then the vacuum gages are watched for a minimum of 20-minutes to see what rate of vacuum-decay is happening. The result is then extrapolated to a full hour, in order to specify a “microns per hour” leak-up rate for that furnace. Thus, if the reading is taken after only 20-minutes the result would be multiplied by three; if taken after a half-hour, the result would be doubled, etc.

Note: It is NOT recommended to take a reading after only 10-minutes and multiply it by six in order to get the leak-up rate in microns per hour. Such results have often proven to be too inaccurate. Even 15-minute readings may be a bit questionable. My personal recommendation is to wait a minimum of 20-minutes before taking a reading, then multiply that reading by 3.

Please note that the longer you wait to take a reading, not only will the reading be more accurate, but usually it gets lower! People have found that a 20-minute reading of, let’s say for example, 3-microns vacuum loss, when extrapolated to 1-hour, would give a leak-up rate of 9-microns per hour. However, when they took a reading after 30-minutes, the degradation of vacuum was 4-microns, which extrapolates to an 8-micron per hour leak-up rate. And, when actually measured after one-hour, the total micron-loss in the chamber was just over 6-microns, thus a 6-7 micron-per-hour leak-up rate.

What leak-up rate is acceptable?

The leak-up rate you should be looking for depends primarily on the type of base-metals you are trying to braze. As mentioned at the beginning of this article, any air-leak into the furnace represents an influx of oxygen into the furnace. Metals vary considerably in their tolerance of oxygen, with nickel and gold being examples of metals that have a large tolerance for the presence of oxygen. That is, they will not oxidize to the point that there is any negative effect on brazing.

However, other metals, such as chromium, or base metals which contain titanium and/or aluminum in their chemistry, will be highly sensitive to the presence of any oxygen in the furnace. Thus, may I suggest the following as a guideline for vacuum-furnace brazing:

Tolerable Vacuum Furnace Leak-Up Rates

  • For base metals that contain any titanium or aluminum: less than five (5) microns per hour
  • For base metals containing chromium (Cr), or manganese (Mn) but no titanium and/or aluminum: less than ten (10) microns per hour
  • For base metals containing nickel (Ni), gold (Au), etc., but without any Cr or Mn or Ti or Al: less than fifteen (15) microns per hour.

Please understand that any good vacuum furnace being manufactured today should easily be able to perform to these requirements. If the testing of your furnace reveals leak-up rates in excess of 15-20 microns per hour, then you really need to track down those leaks and repair them, or consider updating your vacuum-furnace equipment.

Next month: In next month’s article I’ll describe how a helium leak-detector can be used for finding any leaks in your vacuum furnace. In a subsequent article to that, we’ll look at other sources of “virtual leaks” in your brazing furnace that may come from outgassing of things such as the binders from brazing paste, or and contamination on the furnace walls.

Problem solve and improve with our technical experts.