All metals expand when they are heated, and contract when they cooled. This fact has been thoroughly explored over the years, and data-tables have been published showing the coefficients of thermal expansion (CTE’s) for each of the many metals available for use in product design and construction. But carbon-steels present a unique situation to designers and brazing companies, because when being heated all the way up to copper-brazing temperature, the metals will actually go through a temperature-range where the steel will actually be contracting (shrinking) while being heated, and then do just the opposite when cooling, thus potentially causing distortion and/or fracturing of brazements during a high-temp brazing cycle. Such a scenario is illustrated in Fig. 1 where an automotive fuel rail brazement failed to braze properly, because some unique CTE problems associated with carbon-steels was not properly taken into account during the furnace brazing cycle.
Notice in Fig. 1 how the projection-welded bracket has been pulled away from the lower tube due to different rates of expansion/contraction between the lower and upper tubes during the copper-BFM brazing cycle in which the assembly was brought to a furnace temperature of 2030°F (1110°C).
Let’s take a closer look at this situation.
By examining the diameters of the two tubes shown in the photo in Fig. 1, it can be seen that the upper tube has a smaller diameter than the lower tube. This was intentionally designed this way because the designers understood that they did not require as much fuel flow through that upper tube as compared to the lower tube, and so, had the opportunity to design that upper tube in the photo to have a smaller diameter in order to save money. Earlier fuel-rails had used same-diameter tubes for both portions of the fuel-rail.
IMPORTANT NOTE: Whenever a smaller diameter tube is specified in a design, the wall thickness of that smaller tube will generally also be thinner than the wall-thickness of a larger-diameter tube. Thus, as in the case of this particular fuel-rail design, the cross-sectional mass of the smaller tube was actually a lot less than the cross-sectional mass of the larger diameter tubing shown in the lower part of the photo in Fig. 1. Thus, during the heating portion of the brazing cycle in the furnace, the smaller-diameter tube expanded more quickly than the larger-diameter tube, since the smaller mass of the smaller tube allowed it to heat more quickly than the heavier mass of the larger tube.
Is this a problem? YES! Let’s see why.
“Hiccups” in the CTE curves of carbon steel
Fig. 2 shows typical thermal expansion data for some common base metals. It is apparent from these curves that different metals expand at different rates when heated. It can be seen in Fig. 2 that aluminum expands very rapidly, whereas graphite is a low expanding material compared to the others on that chart.
Notice in Fig. 2 the strange curve for 1018 steel (typical of all carbon and alloy steels, and certain stainless steels such as 17-4PH), in which the 1018-steel, when heated, continues to expand until it reaches a temperature of about 1300°F (700°C), where it begins to shrink (get smaller) while still being heated. This shrinking of the metal continues until it reaches a temperature of about 1550°F (850°C), above which temperature the steel begins to expand once again.
This strange behavior of 1018 steel when heated is due to what is called “polymorphism” (meaning “many forms”) of iron, i.e., iron, when heated, changes from one crystallographic arrangement of atoms to another, as shown in Fig. 3.
Notice in Fig. 3 that pure iron, at room temperature, has a body-centered cubic (BCC) arrangement of atoms. But when pure iron is heated above about 1670°F (910°C) it will change from BCC to a face-centered cubic (FCC) arrangement of atoms, which has closer spacing between atoms than does BCC. Notice in the illustrations in Fig. 3 that the atom in the center of the room-temp BCC structure moves from the center of the cube out to the center of the face of the FCC structure. Notice that the distance from any corner atom to the center atom in the BCC structure becomes less when that center atom moves out to center of any given face of the FCC structure, and in actuality, the entire “cube” gets smaller (unlike in the drawing shown in Fig. 3 where I’ve made the FCC cube bigger, only to be able to show the FCC structure more clearly for the reader).
Therefore, when pure iron is slowly heated, it gets smaller when it crosses 1670F (910C), and will once again start to get larger when it cools down below that temp.
Please note that carbon-steels and alloy steels have a number of additional metal constituents in their alloy chemistry other than just pure iron. Consequently, the polymorphic transformations in various steels will occur over a temperature-range (as shown in Fig. 2), rather than just at one temp as with pure iron (Fig. 3).
Heat-treaters and brazing personnel must take this polymorphism-of-steel into consideration during any thermal cycles to which the steel is submitted, or the effects of this polymorphism may result in distortion of the parts being heat-treated or brazed.
Back to the photo of the fuel-rail shown in Fig. 1. How did ferrous-polymorphism cause this assembly to break-apart during the brazing cycle, literally breaking the projection welds on the bracket shown in the left of the photo? Let’s look at Fig. 4 for the answer.
As can be seen in Fig. 4, the smaller-diameter tube, having a thinner wall thickness, and thus a much lighter cross-sectional mass than the larger-diameter tube, will absorb heat faster and transform sooner than the heavier tube which will take longer to reach its transformation temp. When the smaller-diameter tube reaches its transformation temperature and begins to shrink, the larger tube to which it is attached is still growing. It should become apparent that when the smaller tube is shrinking and the larger tube is growing, something needs to “give-way”. Who will win — the small tube or the larger tube? They BOTH will win!
The smaller-diameter tube WILL continue to shrink as it transforms from BCC to FCC. The larger-diameter tube WILL continue to expand because it hasn’t yet reached its transformation temperature. The “loser” is the projection-welded bracket! It’s trying to hold onto both tubes as one is shrinking and the other one is still expanding, and the stress will be too much for the bracket, which then has to break apart from one side or the other.
How to solve this problem?
To prevent this distortion problem in your brazing furnace, you must either slow-down your heating and cooling rates in your furnace programing as you approach the transformation temps for the particular steel with which you are working, so that the smaller-diameter tube and the larger-diameter tube will go through their transform together, at the same time, or you must make the two tubes equivalent in diameter, or you must somehow find a way to make two different diameter tubes go through their metallurgical transformation together (at the same time).
The most logical solution would be to use the same diameter tubes for both sides of this rail assembly. But in the world of automotive brazing, once a commitment to a design is made, you may be stuck with this kind of situation for a year or two before any design correction could be made. Thus, you must find another way in which to try to make these two different-diameter tubes behave the same in your brazing furnace.
By placing a strip of ceramic-fiber blanket material along the smaller tube during brazing (as illustrated in Fig. 5), you could slow down the heat-absorption of the smaller tube so that it absorbs heat at about the identical rate as the larger-diameter tube, thereby allowing the two different tubes to transform together, eliminating the distortion problems that will otherwise occur.
Experimentation is required to find the size and amount of ceramic-blanket material to use (or a piece of solid alumina-ceramic) so that both tubes heat and cool at the same rate, thus eliminating this distortion issue. Typical ceramic-fiber materials go by such trade names as FiberFrax, Refrasil, Kaowool, etc., just to name a few.
CONCLUSION — Always remember that different metals expand and contract at different rates. Different masses of the same metal will likewise expand/contract at different rates. This must ALWAYS be taken into account when designing components for brazing, and when setting up the heating/cooling cycle to use for your brazing operations.
This becomes especially critical when any of the metals used in a brazed assembly go through a polymorphic phase-transformation during the heating/cooling cycle, as discussed in this article. If it is not carefully taken into consideration, a lot of scrap and rework can result which is quite non-productive.
Dan Kay – Tel: 860-651-5595: – Dan Kay operates his own brazing consulting/training company, and has been involved full-time in brazing for more than 40-years. Dan regularly consults in areas of vacuum and atmosphere brazing, as well as in torch (flame) and induction brazing. His brazing seminars, held a number of times each year help people learn how to apply the fundamentals of brazing to improve their productivity and lower their costs. Contact information for Dan Kay (e-mail, phone, fax, etc.), can be found by visiting his company’s website at: http://www.kaybrazing.com/
To View a listing of all Dan’s articles please click here
Read Dan Kay’s Biography
© Copyright Dan Kay 2015