Techniques Posts

Brass, Silver and Flux

Brass, Silver and Flux

Thus far I have used the following filler rods and fluxes on my projects and suspect they are the most widely used:

BRASS:

SIF 101, relatively low melting point brass rods in variety of diameters, 1.5, 2, 2.5 and 3mm. For lugs and braze-ons (1.5mm) and Fillet brazing (larger diameters). SIF do their own fluxes but I favour the Cycle Design Low Fuming Bronze which I find much easier to clean off and generally better in use.
See Cycle Design Products at: http://cycledesignusa.com/wp/

IMG_0945 IMG_0944

SIF No 2, slightly higher melting point and harder result. I have used this for small fillets but its main benefit is it can be used to braze stainless steel. Unfortunately in this case it needs to be used with the specific flux viz. Tool Tip/Braze Stainless flux which does not seem to melt as well as other fluxes. That said I have used it to braze a lugged stainless fork crown with no problems. Mine has come as a crystalline substance like sugar and I find grinding it to a powder with a pestle and mortar or hitting it with a hammer in a plastic bag enables mixing a much smoother paste more quickly as otherwise it takes an age to mix. It takes up water only gradually and therefore if used too quickly will harden so I even found leaving it overnight to ensure full water take up was useful. Equally when heated it dries rapidly and has a habit of dropping off the work. I spoke to someone who had used it extensively and he found it was effective used dry by warming the tubes and covering them in dry flux and also warming the rods and constantly dipping them, the heat enabling the flux to adhere.

SILVER:

SIF No 43, 55% silver solder in 1.5mm rods, 0.5 metre lengths, for lugs and braze-ons.

CYCLE DESIGN FILLET PRO, Silver solder for fillet “brazing”, also useful for filling gaps such as slotted dropouts. 2.35mm diameter sold in coils by the troy ounce. 1 troy ounce = 31.1grams, which equates to about 0.78 metres of fillet pro 2.35mm and a little over 1.8 metres of the 1.6mm diameter.
I use both the above solders with the Cycle Design Stainless Light flux which is very smooth to apply and washes off very easily compared to other fluxes I have used. Like all fluxes it come with health warnings about toxicity and it is the one flux I have used that I would wear an appropriate mask for when fillet brazing because in this situation you use quite a lot of flux and hence a lot of fumes are generated which are pretty acrid. Using it with lugs does not generate as much fume.

IMG_1036 IMG_0943

Hot and Cold setting

Hot and Cold setting

The only adjustment needed on the first frame I built on my course was to adjust the chain stays for wheel twist before fitting the seat stays. This was achieved by clamping the bottom bracket in a vice and pulling on the chain stay until it moved. If it were to break then it wasn’t fixed properly in the first place!
I have since needed to develop some methods of adjusting frame twists and movements which are to some extent inevitable.
The ultimate seems to be a truing table which I have never seen and I am sure would be prohibitively expensive for most hobbyists. Some framebuilders use a surface plate to use as a reference but these are also expensive and take up a lot of room, but frequently are available on Ebay.
Novice framebuilders will probably have to get a frame alignment checker at some point though if you have an accurate jig you can adjust the frame to sit perfectly in the jig again, which is largely what I have done, though I also used the “string method” for a few frames. Given that mainstream manufacturers are alleged to accept a tolerance from front to back of the frame of 5mm I suspect most framebuilders can get better than that using their eye. A dropout alignment tool is probably a necessity also.

Cold Setting can be done largely with the frame clamped by the bottom bracket in a vice (Paterek recommends fitting bottom bracket cups into the threads first to prevent distorting the bottom bracket). I use a 4ft plus length of 3×3 (69 x 69mm) timber for leverage. Whilst this is unwieldy I am certain there is no flex in it and all the force is transmitted directly to the frame, so be gentle. By inserting the timber end on between the chain stays with the shaft abutting the dropouts I can lever the dropouts apart either side as required. This method does tend to have an effect on both dropouts rather the just the one you intend to move, you can correct this by pushing the overcorrected one back. The alternative method is to slide the timber between the dropout you intend to widen and the same side of the lower seat tube and lever outwards. I fancy this is more risky to the integrity of the seat tube but it’s only an opinion. You could insert half a tube block between lever and frame.

IMG_1059
Chamfering the end of the timber helps

IMG_1064
Insert firmly between dropouts

IMG_1067
Gently lever out the dropout, there will be slight movement of the opposite side also

IMG_1068
Or lever one side only

I prefer to close the dropouts if required by hand by pushing against the dropout which usually requires all my body weight. The idea is to check how far out the dropouts are from the centre first using the frame alignment checker or string with reference to the desired axle width which I always aim to be 2mm more than the wheel axle length to make wheel insertion easier.

IMG_1087
The dropout should touch the tip of the frame alignment checker, which is adjustable, on both sides, keeping the axle width required.

IMG_1078
Or adjust the position of the dropouts so the measurement to the string is the same either side

I have managed to adjust a small amount of front triangle twist by inserting the timber through the frame and levering. Care needs to be taken with thin tubes. Better to insert the timber close to the tube joints in this case. With jig built lugged frames I have experienced very little in the way of front triangle twist. However with fillet brazing outside of a jig this has become more noticeable. I had great success with Hot Setting in this instance. This involves heating the portion of tubing/joint at the perceived point of twist on the side to which the frame has pulled i.e. the contracted side. Use a softer flame than for brazing and waft the torch back and forth either side of the joint until the metal goes dark red, avoiding melting the brass, then stop and leave to cool. Hopefully the frame will be restored to its former position, mine was! The same method can be used on chain stays but I have stuck to cold setting here. Final dropout adjustment is done with a dropout alignment tool with which the dropouts can be bent back into alignment and checked.

IMG_1092 IMG_1089

Designing Frames

Designing Frames

After completing an initial frame building course I did not really have much idea about how to design a frame and have muddled along by trial and error. What does become apparent is that conventional designs work which is why they are still around. A conventional rear triangle is always stiff enough and strong enough for most things you throw at it and therefore most attention needs to be paid to the front end design. Always check you are likely to have enough mudguard clearance, toe clearance unless you are not bothered about toe overlap and heel clearance if you are going to fit panniers. If you are going to be fitting large chainrings make sure they will clear the chain stay. This issue has caused me most problems and close shaves when building the frames up after the fact.
The Paterek manual teaches the use full sized drawings, offering up completed sections of the frame to the drawings to check accuracy and even brazing sections over the top of the drawings on a large table. Perhaps most amateur builders will have the same problems as I have had in not having a large enough space to accommodate said table or being able to find paper large enough! I sometimes do full size drawings of lug sections which helps when sanding out lugs to adjust the angle. I guess building without any form of jig would require the use of full size drawings.