I've struggled with getting good results in the past with Silver soldering so thought I would jot down the things I've learnt as I go along. Note I'm not an expert but these are the steps I use to get reasonable results. For further reading I can recommend "Soldering and Brazing" by Tubal Cain (Workshop Practice Series) which is a thoughtful and complete treatise on the subject, and "Model Locomotive Boilermaking" by Alex Farmer for those wishing to make a boiler themselves, this is aimed at those using oxy-acetylene gear but much of it it can be used when employing Propane.
Our clubs boiler inspector suggested I make some practice pieces before I start boilermaking proper. Wise advice as my attempts in the past have been a bit poor. This was my first attempt.
After much head scratching and reference to Tubal Cain I managed this, which I am pretty pleased with.
To summarize you need the following for a successful joint.
Correct joint design, most Silver solders need at least a 0.15mm gap between the joint surfaces. AT BRAZING TEMPERATURE!
Cleanliness. Pickle copper parts in Dilute Sulphuric Acid, and rinse and dry just before fluxing. I use rubber exam gloves to keep oily prints off.
Flux. You must flux all joint surfaces as well as any existing joint you may disturb. The flux should be freshly mixed before you start.
Heat. Use a gentle heat to dry out the flux then bring both parts to temperature as fast as possible, making sure both sides are at equal temps. Requires adequate torch. Apply heat on the far side to where you will apply the silver solder. It will flow to the hot area.
Solder. Use the right amount of the correct grade for the job. I'm using SilverFlo-24, then SilverFlo-55 and finally EasyFlo No.2 so I can solder without worrying about previous joints. I would recommend starting with EasyFlo No2 as it melts at the lowest temperature.
Lighting. I found it much easier to gauge temperature when the ambient light was very low. As I Silver-Solder outside this means waiting to late evening.
In this example I am simulating a joint between a 3/4" dia Boiler flue pipe and the fire box tube plate of Spencer The hole was drilled 3/4" and then filed to make it 0.1mm larger in diameter (2 times 0.05mm) This is the correct clearance for SilverFlo-55. I use a 0.1mm feeler gauge to check this.
I have diluted 1Litre of 98% Pure Sulphuric Acid (OneShot Drain cleaner) with 10 litres of tap water. ALWAYS ADD ACID as my old chemistry teacher Mr. Fuller said, the reaction generates a lot of heat initially so add it in stages. I keep the solution in a close lidded bucket inside a second container. I used a full face guard, plus glasses and rubber gloves to protect myself. Note I keep a couple of bricks on the lid to keep our cats out of it.
To prevent me losing small parts I have cut up an old milk carton which I lower into the acid.
It has some holes in it to let the acid in/out.
Cleaning will remove any oxide from the copper, plus grease and old flux.
The Flux I'm using is Easy-Flo from Johnson Matthey. You should always mix it fresh. I use a small spray bottle to add water as it is difficult to mix small quantities otherwise. Usually I only need a teaspoon or two of paste. I have been told it is possible to use a soft (B) graphite pencil on areas you don't want silver solder. First the joint area on one side of the plate is fluxed.
Then the other side of the plate is fluxed.
The pipe itself is also fluxed.
A common mistake is to apply the solder at the wrong time or apply the wrong amount. See the mess I made on an earlier piece.
My approach now is to place the correct amount of Silver Solder on the joint before heating then you only need to worry about getting the heat in the right place and in the right amount. (Though it is useful to have a pre-fluxed rod to hand if you have been too stingy at this stage). Here is a good example of placing solder.
Here is not so good. The solder is only touching the pipe in two points.
It's a good idea to add some more flux to protect the solder.
I'm using a Sievert 86 Pro Handle with a short neck tube and 2943 burner. The max output of this burner is a bit over the top for this example but it does allow the heat to be increased rapidly.
I had to think carefully about how to arrange the part so I could apply heat evenly to both parts and on the opposite side to the solder ring. This is how I arranged the fire bricks.
This allows access to the torch underneath and a good view of the joint to judge heating.
The Sievert handle has a multi turn knob so you can accurately adjust the size of flame. I start with a gentle heat to remove the water from the flux, too hot and the flux will bubble all over the shop. Once it settles down I increase the heat and concentrate on making sure both sides are heated equally. If one side is substantially bigger, favour that. Eventually the flux will turn to a clear liquid which indicates it is almost at working temperature the silver solder will flow pretty soon after. Once it flows all of the way round remove the heat.
Non-fluxed areas will have oxidised black and the joint will be covered with the glassy flux residues.
An hour in the pickle bath will remove the worst of the mess so you can inspect the results.
This test piece was made to show our boiler inspector so I sectioned the joint and polished it so it is easy to see the penetration.
Another common joint is one which is loosely riveted to hold components in alignment. Here is a test piece I made which simulates the Firebox Tubeplate to wrapper joint. I drill each hole one at a time inserting the rivets as I go to keep alignment.
You'll need a dolly which can reach inside the fire box. I used a handy bit of flat bar and drilled a shallow hole the same size as the head.
This shows how the dolly is used, note I haven't trimmed the rivet yet.
Before making the joint the first plate is fluxed.
Then the top plate is added and they are riveted together.
More flux is added to both sides.
A piece of SilverFlo-24 is placed on the joint.
The parts are arranged so that they can be heated from below.
This was heated from below with a Sievert 2943 nozzle as shown above.
This was then pickled for an hour.
I have sawn through the joint here to check the penetration.
Back to Home Page
This site and all images Copyright 2008. No unauthorised reproduction.