Fig. 1 A typical furnace brazing cycle as perhaps might be recorded on a furnace’s temperature recording chart. Note that this chart is recording the temperatures from four TC’s in the furnace, i.e., the main furnace TC, and then three (3) load-TCs.

When parts are brazed in a vacuum furnace, distortion of those brazed assemblies can easily happen. To prevent parts from distorting, people have tried a variety of things, including extended stress-relieving of components prior to assembling those parts for brazing, the use of rigid fixturing to try to keep parts from moving during a brazing cycle, and even making components heavier and more massive in order to make them more distortion-resistant. Distortion still occurs.

Yes, some of these things, such as stress-relief heat-treatment prior to brazing might help to some extent, but it is not the answer to controlling distortion during any furnace brazing cycle. The real key to controlling distortion is to control the heating and cooling rates used in the brazing cycle.

Shown in Fig. 1 is an illustration of what a typical furnace brazing cycle might look like, in which the furnace temp is controlled by the “Furnace thermocouple (TC)” shown on the left side of the chart, and three (3) load TC’s are used on one part to see the temperature-spread (temp-differential, or delta-T) within that one part.

Load-TC1 is placed on the thinnest section of a single assembly being brazed, TC2 may be in the center of the assembly, and TC3 is buried in the heaviest (most massive) part of that same assembly. Please understand that you can, and should, run additional TC’s in your furnace load, so that you can also be measuring the temperature-spread in the entire furnace load (front to back, top to bottom), etc., but it is important when developing ANY furnace brazing cycle that you place multiple TC’s on at least one of the parts/assemblies in that load to determine what the heating rate (and cooling rate) does to the temperature spread within that one part/assembly.

NOTE: Distortion of a part occurs because of the temperature-differential (delta-T) in that one part, and is NOT due to the temp- differential (delta-T) throughout the entire furnace load! This is VERY important to understand!

In prior articles written for this site, I’ve discussed the need to closely control the heating rate used, and to use the fastest heating rate you can that will allow you to go up to brazing temp without having to use any “holds”! We discussed at length that the ONLY reason for using “holds” is because the heating-rate is too fast, causing the heavier sections of parts to lag well behind the temperature in the thinner sections of a component, causing a large enough delta-T (temperature differential) in that part to cause distortion. “Holds” are designed only for the purpose of allowing heavier sections in a component (or assembly) to “catch up” in temp to the thinner sections in that part until the temperature-spread in each part has narrowed down to an acceptably small range. When that narrowed temp-spread within a part has been finally achieved, after holding the furnace at a constant temperature for many minutes (often ranging from a minimum of 30-minutes to as long as a couple of hours) the heating can start up again, and will continue to heat up the furnace until the delta-T within the part once again becomes too wide and another “hold” is needed, etc.

Please understand that — IF — you (the person doing the brazing) – if YOU do not have multiple TC’s placed on at least one part in that load, recording the temp of the thin section of that part and the temp of the heavy section of that same part (so that you can measure the delta-T in that one part), then your heating process is not really in good control! Yes, you may be using a “controlled heating rate”, but you will NOT know the effect of that heating-rate on the individual components in the load, and thus your PROCESS is NOT in proper control! Control is key to your success in brazing!

Now look at the right side of Fig. 1, where you see the second and third parts of that furnace brazing cycle, namely the “hold at brazing temp” until the part has been fully brazed, and then the cooling cycle once the brazing is completed, and the furnace-heating is either turned off, or a rapid-cooling/quench gas is added to the furnace to bring the parts down to room-temp as quickly as desired.

When a component is being brazed, at what point in the brazing cycle is it the weakest? At room-temp, or at brazing-temp? Obviously, it is weakest at brazing temp. And when during its hold-time at brazing temp is the part the weakest, as soon as it reaches brazing temp, or at the end of the time it has been sitting at braze temp? Again, the obvious answer is that it would be weakest at the end of that hold-time, and just prior to going into the cooling portion of the furnace-cycle. Thus, when the part is at its weakest, we’re going to ask it to do something very dangerous — namely, cool down!

As mentioned above, sometimes people think that they are being “gentle” with the parts by implementing a so-called “furnace cool”, in which they merely turn off the heat but let the parts “cool down naturally inside the closed furnace”. This can be anything but “gentle”, and is responsible for more distortion in furnace brazing than any other factor. Again, look at the right side of Fig. 1. Notice that as soon as the heat is turned off, the parts actually begin to quite rapidly cool down. Why does that happen?

Let’s take a look at the amount of power required to heat the load and bring it up to brazing temperature. Let’s further assume (just for discussion purposes) that you had a “power meter” on your furnace instrument panel, like the imaginary one shown in Fig. 2, that is showing the amount of electrical energy being used to heat the parts as a percent of total available power that can be sent into the heating elements in the furnace.

Fig. 2 Theoretical power-settings, as a percentage of available furnace heating-power for (A) when the furnace is heating up to brazing temperature, and (B) when the parts are “holding at brazing temp” for a certain amount of time. Heating rate in (B) is such that the heat going into the parts equals the heat being radiated out of the parts.

Fig. 2 Theoretical power-settings, as a percentage of available furnace heating-power for (A) when the furnace is heating up to brazing temperature, and (B) when the parts are “holding at brazing temp” for a certain amount of time. Heating rate in (B) is such that the heat going into the parts equals the heat being radiated out of the parts.

When the cycle begins and the heat is turned on, let’s say that the meter is showing a power input of “95” percent (Fig. 2A), just for our discussion purposes. As the parts absorb all that heat and get closer to brazing temp, that power-reading on the meter drops slowly, causing less heat to be thrown at the parts, so that the heating rate slows down a bit. Then, during the “hold at brazing temp”, where the parts are being held at a constant temperature, what does the “power meter” show, percentage-wise, in this example? Is it reading “zero power”, since the parts are no longer rising in temp? Of course not! It will, in fact, show a fairly high amount of power still being sent into the heating elements of the hot-zone (for our discussion, let’s just say that the meter is reading 55%, as shown in Fig. 2B) because the amount of heat being sent to the parts is just equal to the amount of heat being radiated from the parts, so the temperature of the parts will remain level!

Then, when the furnace-heat is turned off to initiate the beginning of the cooling cycle there is obviously no more heat being thrown at the parts from the heating elements, and the parts will quickly dump their heat (by radiation) and quickly start to cool. The thin sections will dump their heat the fastest (as shown by TC1 on the right side of Fig. 1), whereas the heavier sections of that same brazed component can only shed its heat at a much slower rate, as shown by TC3 on the right side of that chart.

Notice the huge temp-differential (delta-T) between TC1 and TC3 very soon after the cooling cycle has begun. This large delta-T can quickly cause distortion in that “weak/soft” brazed-component. Please note that delta-T within an assembly can be a bad enough problem when the assembly is being heated up to brazing temp, but once the part has actually been brazed and is now a single, fully-bonded component, any significant delta-T on the way down can cause far worse distortion problems. This is usually the case since cooling rates are usually much quicker than heating rates.

IMPORTANT: I teach in my seminars that distortion does NOT have to happen when furnace brazing and the way to prevent it is to control the heating rates used and to also very carefully control the rate of cooling used. When done properly, parts should be able to go into the furnace within desired dimensional tolerances, go through the brazing cycle, and come out of the furnace still within those desired dimensional tolerances.

An experienced furnace-brazing operator (he had more than 25-yrs of brazing experience in his shop), attended one of my brazing-training seminars a few years ago, and he objected strongly to this statement, because he “KNEW” that distortion will ALWAYS happen during furnace brazing based on his own experiences over those 25-yrs of furnace brazing. However, after he had learned about proper temperature control during the 3-days of the brazing seminar, he told me: “I found out that I had 1-year of bad brazing experience, repeated 25 times!” He wished that he would have been able to attend one of my seminars 25-years earlier so that he could have avoided making those same mistakes year after year for 25-years!

How to Furnace Cool “Correctly”?

Take a look now at Fig. 3. Notice the end of the brazing cycle is actually called a “Controlled Heat-Down”, rather than a cooling cycle. To me, this is very important. When the time for the “hold-at-brazing-temp” is over, don’t just turn off the power to the heating elements! Build into your cycle a gradual reduction in the amount of power going to the heating elements, rather than just turning them off.

Fig. 3 When the brazing has been completed, and it’s time to cool the load back down to room temp, do not just turn off the heat, but instead, slowly lower the heating rate, to achieve a “controlled heat-down” of the parts instead.

Fig. 3 When the brazing has been completed, and it’s time to cool the load back down to room temp, do not just turn off the heat, but instead, slowly lower the heating rate, to achieve a “controlled heat-down” of the parts instead.

Illustration

Remember our example earlier in which we stated that when the parts were being held at brazing temp the “Power Meter” was reading about 55% power? What would happen to the temperature of the parts if you turned the power down to 48%, instead of just turning it off? Obviously, since the amount of heat you would be throwing at the parts would be LESS than the amount of heat the parts were radiating, their temperature would start to fall a bit. Then, when you lowered the power even further, let’s say down to 40%, more heat would drain from the parts, etc. Thus you would, in fact, be letting the parts “gently” lower their temperature (cool-down), and, by some experimentation, you would be able to keep the delta-T within any one of those brazed-assemblies such that no distortion would occur in that load, or in any subsequent load, in your furnace.

What delta-T within a component is needed to prevent distortion?

That needs to be determined experimentally for your parts in your furnace, but —- for starters, look at Fig. 4, a chart developed many years ago and published in the article: “Control of Distortion During the Furnace Cycle” published by the American Welding Society (AWS) in their Welding Journal in October 1971.

When the temperature difference (delta-T) within that metal part is below the curve for that metal, then the thermal-stresses will cause that metal to begin to yield (distort) due to permanent “plastic deformation” of that metal. At temps below any of the curves, the thermal-stresses are less than that needed to cause plastic-deformation (yielding) of the metal, and thus, distortion will not occur.

Fig. 4 Experimental results showing the “yield-point curves” for several different metals resulting from significant temperature differentials (delta-T) within a piece of that metal. (See: “Control of Distortion During the Furnace Cycle”, published in the AWS Welding Journal in October 1971.)

Fig. 4 Experimental results showing the “yield-point curves” for several different metals resulting from significant temperature differentials (delta-T) within a piece of that metal. (See: “Control of Distortion During the Furnace Cycle”, published in the AWS Welding Journal in October 1971.)

Notice that the lines on the chart go from upper-left down toward the lower-right, meaning that the allowable delta-T gets less and less as the temp in the furnace gets higher and higher. Obviously, only a few metals are shown on the chart, and the curves only represent the results of that specific series of tests in that lab furnace at that time. It would be good to have these experiments done by other labs in their furnaces to make a more complete chart with more metals. Can any companies reading this article take on that task?

Conclusions

It is VERY important that the temperature throughout the entire brazing cycle be properly and fully controlled if you want to have consistent results in your furnace brazing operations, and significantly reduce (or eliminate) scrap and rework. Too many shops do not use thermocouples properly, or try to minimize the use of TC’s as much as possible because they think that “we know what we’re doing, and TC’s aren’t really helpful to us anymore”. Proper temperature control is absolutely essential if you want to eliminate distortion of brazed assemblies, and this, therefore, requires the use of multiple thermocouples (TC’s) in each furnace run, and the use of multiple TCs on at least one brazed assembly in each of those furnace loads. Then, by modifying the heat-up rates in your furnace so as to eliminate any “holds” on the way up to brazing temp, and especially by introducing “controlled heat-down” programming into your furnace brazing-cycles you will find that your process control is more consistent, the parts will braze better, and any distortion of parts being brazed can be effectively eliminated (or reduced significantly). The decision is yours.

Problem solve and improve with our technical experts.