Brazing Q+A June Layout 1 5/16/13 2:14 PM Page 24 BRAZING Q&A BY

BRAZINGQ&A BY TIM P. HIRTHE

Q: We manufacture an aluminum heat ex-changer for an automotive application. Itis of a round tube-to-fin design. The openends of the tubes in the circuit are joinedwith a U-bend, and are brazed with a fluxbearing aluminum-zinc braze alloy. Weuse a relatively simple fixture with fourtorches. We have set up to use natural gasand oxygen. The design of the heat ex-changer includes a mounting bracket thatmakes it difficult to heat the entire cir-cumference of one of the braze joints. Weare having significant leak issues in thearea where accessibility is hindered. Isthere equipment available or some tech-nique that will allow us to get a qualitybraze in this joint? We are consideringswitching to an aluminum-silicon brazealloy, which we understand is a highertemperature, but which we are hearing isa better braze alloy to use. Early trialswith it, however, have resulted in meltingof the tubing. A photo of the joint in ques-tion is shown in Fig. 1.

A: To state the obvious, the best situationwould be to not have the bracket in theway in the first place. You mentionedthat it is an automotive application, sothe odds of removing it from the designare probably quite low. An alternativewould be to redesign the assembly tohave the bracket able to be out of the wayprior to brazing and bent into place afterbrazing. Using perforations that allowbracket bending is one method. This iscommon practice in a variety of heat ex-changer applications. This type of ap-proach is perhaps obvious, and you prob-ably would have done it already if a de-sign change was possible. For the pur-pose of the remainder of my response, Iwill assume that no design change is possible.

By not being able to get heat onto theback side of the joint where the bracketis located, you are relying on the thermalconductivity of the aluminum to transferheat to obtain proper melting and wet-ting of the braze alloy on that side of thejoint. I can’t determine from Fig. 1 or theinformation provided what the jointclearances are. The problem is that thealuminum-zinc braze alloy requires aloose fit and a rather deep joint. The rea-son for this is that these joints tend to beporous due to the zinc content. Zinc is ahigh vapor pressure element that willboil off at these braze temperatures.Since the braze alloy requires a loose fit,the heat transfer characteristics will bepoor. You usually end up overheatingone part of the joint to get another parthot enough to accept the braze alloy

properly. Even if you can achieve mini-mally acceptable brazing temperaturesaround the circumference of the joint,you will most likely overheat it such thatyou will get zinc vaporization.

Another difficulty I see is that you areusing natural gas and oxygen. Using oxy-gen to combust the natural gas results ina flame that is relatively hot and difficultto control when brazing aluminum.When brazing aluminum, we prefer thefuel gas be combusted with air ratherthan oxygen. It produces a cooler, morecontrollable flame. We suggest a blowerbe used rather than plant compressed airas the latter typically contains a lot ofmoisture. If you are going to use plantair, you must assure that it is being driedbefore combusting the fuel gas. I am con-cerned about using natural gas also. Nat-ural gas fluctuates greatly in Btu contentfrom the supplier and, depending on thetime of year, it also can contain a greatdeal of moisture. Propane or a similarfuel is preferred.

You mention you are consideringmoving to the aluminum-silicon brazealloy because you believe it is a betterchoice. It is attractive because it does nothave the vapor pressure problem, and itis a stronger material than the alu-minum-zinc. There are several issues,though, that must be addressed in con-sidering switching to an aluminum-sili-con braze alloy. The most obvious one isthat the aluminum-silicon braze alloymelts at a significantly higher tempera-ture. I am not sure which aluminum-zincbraze alloy you are using, but for refer-ence, the 98% zinc/2% aluminum brazealloy has a melt range of 715° to 725°F(379° to 385°C). The aluminum-siliconalloy most commonly used is the 88%aluminum/12% silicon, which has a meltrange of 1070° to 1080°F (577° to 582°C).I don’t know which aluminum tubing ma-terial you are using, but I am sure it meltsin the vicinity of 1200°F/649°C. There issignificantly less room for overheatingwith the aluminum-silicon.

If you are trying to use the aluminum-silicon on a joint designed for aluminum-zinc, you will have a substantial problem.As mentioned earlier, aluminum-zinc re-quires a loose fitting joint (~0.005 in./0.127 mm) that has considerable depth(~0.250 in./6.4 mm). The way to com-pensate for the tendency to vaporize thezinc is to provide this very deep joint.The aluminum-silicon requires a tightfitting joint (~0.001 in./0.025 mm) withless depth (~3.0 mm). Your trials withaluminum-silicon ran into the fit prob-lem. You are having a struggle getting

heat to the back side of the joint with alu-minum-zinc because of the loose fit sinceyou must overheat the front to getproper heat transferred to the back. Youget away with this due to the low meltingpoint of the aluminum-zinc. When youuse aluminum-silicon, you do not havethat margin of error with the tempera-ture. The result is melting of the alu-minum tube as the difference betweenthe melting point of the tubing and thatof the braze alloy is too small to toleratethis overheating.

Assuming again that no designchanges can be made, it would seem thatchanging to the 88% aluminum/12% sil-icon alloy is not feasible. Consideringthat the bracket cannot be taken out ofthe equation, overheating will occur andmost likely result in tube melting. If thejoint could be redesigned to accept thisalloy, i.e., tighter fit and shallower depth,it may be possible; but I have ruled outthe option of changing the prints. Goingforward then it seems the alloy must re-main aluminum-zinc.

I would look at modifying your exist-ing brazing setup to find a way to getmore heat into the joint blocked by thebracket. Switching to air to combust thefuel will help. Using different torch tipsthat have a softer, broader, and longerflame condition that will allow heat toreach the backside of the joint will help.Adding more torches may help. In orderto heat this particular joint, it may re-quire a longer cycle time as you are al-

Fig. 1 — View of the customer’s brazedaluminum heat exchanger.

JUNE 201324

Brazing Q+A June_Layout 1 5/16/13 2:14 PM Page 24

2525WELDING JOURNAL

lowing for the heat to transfer to theproblematic area of the joint. You mayalso want to contact induction heatingmanufacturers as they may have heatingcoils that can preferentially heat differ-ent locations on the joint circumference.The lesson to be learned here is to takeinto account the joint accessibility whendesigning the assembly.♦

This column is written sequentially byTIM P. HIRTHE, ALEXANDER E.SHAPIRO, and DAN KAY. Hirthe andShapiro are members of and Kay is an ad-visor to the C3 Committee on Brazing andSoldering. All three have contributed to the5th edition of AWS Brazing Handbook.

Hirthe (timhirthe@aol.com) currentlyserves as a BSMC vice chair and owns hisown consulting business.

Shapiro (ashapiro@titanium-braz-ing.com) is brazing products manager atTitanium Brazing, Inc., Columbus, Ohio.

Kay (Dan@kaybrazing.com), with 40years of experience in the industry, operateshis own brazing training and consultingbusiness.

Readers are requested to post theirquestions for use in this column on theBrazing Forum section of the BSMC Website www.brazingandsoldering.com.

COMMERCIAL DIVE TRAINING

Education That Works

AIM HIGH. DIVE DEEP. Call Today!

1.800.238.DIVE (3483)www.diversacademy.com

Near Atlantic City, New Jersey

For info go to www.aws.org/ad-index

For info go to www.aws.org/ad-index

Brazing Q+A June_Layout 1 5/16/13 2:18 PM Page 25