Brazing vs. soldering

[DaBit]

There is a question that I have never seen answered completely. Why are all tube connections in a refrigerating system done with brazing, and never with normal soldering?

To be sure that we are talking about the same thing (English is not my natural language, and I am not very good in foreign languages besides Verilog and C++):

- With brazing I mean connecting two metal surfaces with a filler metal, melting point > 650 °C
- With soldering I mean connecting two metal surfaces with a filler metal, melting point < 350 &deg;C

When I am 'brazing', I am using a powder-like flux and brazing rods containing about 40% silver. When I am soldering, I am using a liquid flux and SnPb or SnAg solder.

Now, both types of connecting metals provide a strong joint. Now I wonder: why is the more expensive and possible more risky method (easier to damage TEV's, for example) of brazing always used? I can imagine a few reasons:

- Strength. Joints at the high pressure side of the circuit must be strong to resist peak pressures. Now, I have already seen hydraulic systems where pipes are conencted using normal solder joints. The pressures in these systems are usually higher than those in any common refrigeration system. A pipe-to-pipe connection using SnAg solder should easily withstand 30 bar peak pressure.

- Heat. Joints in the discharge line must withstand heat and high pressure. I can imagine that normal solder joint gets weak at high temperatures, but those temps are not that high, are they?

- Tightness. Maybe a normal solder joint leaks refrigerant?

Well, what is the reason that those connections need brazing instead of normal soldering?

[RogGoetsch]

Short answer: halocarbon-based refrigerants are incompatible with lead-based solders. They will eventually leak.

Why not SnAg solders? Habit, mostly. When you've seen things done only one way for your entire life, it's difficult to open to a new idea.

Also, flux in refrigeration systems is very bad. Safer and faster to set employees up with the self-fluxing 15% SilPhos brazing rod and a nitrogen purge than expect them to flux carefully enough to do the job without contaminating the system.

Also, some technicians have told me that they stay with 15% SilPhos because they don't have to prep the joint, a position that drives me nuts.

I use SnAg 6% Ag ("Stay Brite") for most joints now. Most of my work is small commercial repairs where clearances are tight and it is a joy to get in and out quickly with a small, cool (air-acetylene) flame without rigging nitrogen, flame shields and heat sinks.

I still use 45-55% Ag brazing alloy with paste flux for ferrous-nonferrous joints (or when I don't like the looks of the copper plating on the compressor stub tube.) When I have to do that, once I have the nitrogen purge dialed in, I'll go ahead and SilPhos the rest of the job, clearances allowing, because in my gut it still feels better! (I still have the visual from tech school of pulling our joints apart with a hydraulic device and seeing the brazed joints hold past the failure point of the tubing, while the silver-soldered joints popped apart.)

Rog

[DaBit]

Short answer: halocarbon-based refrigerants are incompatible with lead-based solders. They will eventually leak.

Right, that was the answer I was looking for.

I use SnAg 6% Ag ("Stay Brite") for most joints now. Most of my work is small commercial repairs where clearances are tight and it is a joy to get in and out quickly with a small, cool (air-acetylene) flame without rigging nitrogen, flame shields and heat sinks.

If SnAg is suitable, and people are actually using it, I might consider using it for non-critical connections like a T-connection. I lack an oxy/acetylene torch, so the pure fact that I can use a normal propane torch is invalueable. Flowing with nitrogen was also not possible for me, so I used butane for that.

I still don't feel good about it, but if I consider using SnAg for some joints, what kind of flux would you recommend me? The normal S39?

[lindsay jahnke]

I have worked with Silver Solder and Soft Solder for years and would prefer to walk away from a silver Solder Joint any day. (Providing its done wright by cleaning excess flux away) It concerns me useing Soft Solder because its not as strong, and I have repaired cracked Joints before.

[Dan]

With brazing I mean connecting two metal surfaces with a filler
metal, melting point > 650 ?C
- With soldering I mean connecting two metal surfaces with a
filler metal, melting point < 350 ?C

Those temperature ranges and definitions are something I have seen discussed before. I can say this much, however.

A typical supermarket uses between 50 and 75 pounds of silphos (15% silver). We call it silver solder, perhaps loosely. It is by far the most common joining process for copper-to-copper connections in use for commercial hcfc installations. Stay-brite is used for disimilar metals, such as brass-to-copper. 45% solder(we call it "silver-solder", also) is used for dissimilar joining much more often than stay-brite is. It works well with steel to copper joints. It requires a paste flux.

I don't have much experience with stay-brite, but I think I might require a paste flux, too.

But 15% silver-solder (LOL, we use so many different words for the same thing!) is the mainstay of commercial installations, mostly because it doesn't require additional flux.

Sadly, it often works even if the installer does not properly prepare the connections. But, then again, that is also its blessing.

[DaBit]

I normally use the SilPhos rods you are referring to, and for copper->iron connections I use rods containing 40-45% silver, combined with a natriumtetraborate based flux. This stuff is a powder which you should mix with water to form a paste.

The disadvantage for me is that I need an oxy-acetylene torch to make the connections to the compressor etc. For normal piping I can get away with using a propane torch; it only takes longer to make the connection. I have been trying to buy a decent second hand oxy-acetylene set, but they are are hard to find. New equipment is way too expensive.

If the brazing is done with pipes not connected to anything else, I rinse the pipes afterwards with sulphuric acid, distilled water, and finally I flow the pipe with butane and heat the pipe to get rid of the water. Then, I cap it and I can be sure that there are no traces of water or non-condensibles left in the piping.

Sure, this method is truly different than what the professionals are used to.

Using SnAg would have advantages. The copper won't oxidize that much, and SnAg is easy and quick to handle with a normal air-propane torch. I had a bad feeling about it, but now I am seriously considering to use it. It would make my life as a DIY refrigeration 'engineer' so much easier.

Sure, I have to take care of the flux, but when you clean your piping well before conencting (which I always do. Even with SilPhos the joint is nicer to see and the solder flows better), you need only a very little bit of it, and preventing flux or flux residues to get in the system is easy.

Besides the flux problem (which I can handle), what drawbacks are there regarding SnAg?
Which connections should I still do with SilPhos/45% Ag brazing rods?

[staycool]

Try out the oxy turbo kit small and recommended to every engineer, look at countyweld.co.uk for what your paying its an amazing piece of kit.. Plus silver solder is alot safer, knowing the jobs done properly aswell

[pendlesteve]

Some interesting articles here, but be aware that here in Europe we have to consider both PED and EN:378 where all joints made or altered on site should be pressure tested to 1.43 times maximum working pressure. So for a brazed or flare joint we use the following, 32 C for the low side of a system, 43 C for the high side of a water cooled system & 55 C for the high side of an air cooled system. So the majority of times we would look at 32 & 55 C. This gives us Maximum Working Pressure.

A couple examples for UK latitudes would be R134a at 32 C has a pressure of 7 bar (103 psi) & at 55 C, 14.9 bar (200 psi). These figures are then multiplied by 1.43 for the strength (pressure) test giving 10 & 21.3 respecively (148 & 286 psi). At these pressures there should be no rupture or deformation of the pipe and joints. Leak testing is carried out at 10% above MWP.
If you look at R410A:eek: 32 C = 19.7 bar (272 psi) and 55 C = 33.3 bar (485 psi) X these by 1.43 & you get 28.2 & 47.7 bar respectively. All pressure testing should be performed using an inert gas mainly oxygen free nitrogen (OFN).
Please be very careful with the lines you use connecting the cylinder to the system and avoid using your gauge manifold particularly if it has a sight glass fitted.
So don't even consider soft solder.
As for brazing, you should always "drift" ofn through the pipes to avoid oxidisation on the inner surfaces of the pipe or fittings etc. If oxides do form and a system components fails during the warranty period it has been known for the mauufacturers to reject the warranty because of tyhe contamination.
The other point here is flux. Powder only goes where it's put so when for example brazing in, a new compressor it is possible for the pipe inside the joint to oxydise resulting in a potentialy weak spot. But if the flux is mixed to a paste and carefully applied to the outside of the pipe it will keep the joint clean. Fluxed rods suffer the same as powder application, so use clean rods & a small tub of flux which you mix with clean water as and when needed.

[Robearbam]

I always refer to this site...someone posted it here. The first lesson tells it all. Figured it would be worth posting it again. Cheers!



http://www.refrigtech.com/bigblumanual3.html#EVERTHING%20LEAKS

[sumant mathure]

I believe that solders give mostly Keying type joints,, where as in Brazing you can get some good diffusion joints, which work better with vibrations, due to compressor on off, This is one of the major reason for brazing joints, rather than soldering joints.

sumant mathure

[semb]

Short answer: halocarbon-based refrigerants are incompatible with lead-based solders. They will eventually leak.


So if I'm using a hydrocarbon based refrigerant, is there a reason to not use soft solder?

[botrous]

If SnAg is suitable, and people are actually using it, I might consider using it for non-critical connections like a T-connection. I lack an oxy/acetylene torch, so the pure fact that I can use a normal propane torch is invalueable. Flowing with nitrogen was also not possible for me, so I used butane for that.

DaBit as I understood from the posted that I quoted you are using butane as a non-oxydation agent in the circuit while you are soldering ?????
Isn't that dangerous . . .
If I miss understood your post please explain your point of view.

Respects and regards