Installing bridges and effect on rear dropout width

[Paul Jacobs]

I apologise if this topic has already been covered. I did a search on here and did not find the answers I was looking for. Please can someone explain to me why brazing in a brake bridge and/or a chainstay bridge brings the rear dropouts closer together. This has not come as a surprise to me, I read about the phenomenon before I built my first frame and took steps to compensate with fair success on a few frames. My latest frame, with an intended rear dropout width (spacing) of 130mm has ended up 129.6mm. Somehow this irritates me more than if it had settled at 130.4mm. The latest frame has longer (44.5mm) chainstays than my other frames so I accept that I have under-compensated but I wish I understood the physics better.

Paul Jacobs.

[Paul Jacobs]

I apologise if this topic has already been covered. I did a search on here and did not find the answers I was looking for. Please can someone explain to me why brazing in a brake bridge and/or a chainstay bridge brings the rear dropouts closer together. This has not come as a surprise to me, I read about the phenomenon before I built my first frame and took steps to compensate with fair success on a few frames. My latest frame, with an intended rear dropout width (spacing) of 130mm has ended up 129.6mm. Somehow this irritates me more than if it had settled at 130.4mm. The latest frame has longer (44.5mm) chainstays than my other frames so I accept that I have under-compensated but I wish I understood the physics better.

Paul Jacobs.

I meant 445mm chainstays

[e-RICHIE]

Do you do the work with a surrogate axle in the dropouts?
I hope so.
Also after each braze, wave the torch on the outside of the stay where the bridge is, maybe for a 10-15 seconds.
That should neutralize any desire for the stays to contract.

[Paul Jacobs]

Do you do the work with a surrogate axle in the dropouts?
I hope so.
Also after each braze, wave the torch on the outside of the stay where the bridge is, maybe for a 10-15 seconds.
That should neutralize any desire for the stays to contract.

I do use a surrogate axle and for my latest frame the surrogate axle was 132.5. This is longer than I would normally use. I chose this empirically based on my previous experience because of the long (445mm) chainstays and because I was placing a chainstay bridge some way from the bottom bracket to support a fender for a 650B wheel and 42mm tyre. I did not put the bridge all the way out, I plan to bolt the mudguard to the bridge through a spacer. A lot of the other frames I have built using a Sachs bottom bracket shell and no chainstay bridge which works well and for which I thank you. Thank you also for the tip about heating the outside of the stay which I did not know and did not do but will try next time.

[e-RICHIE]

I do use a surrogate axle and for my latest frame the surrogate axle was 132.5. This is longer than I would normally use. I chose this empirically based on my previous experience because of the long (445mm) chainstays and because I was placing a chainstay bridge some way from the bottom bracket to support a fender for a 650B wheel and 42mm tyre. I did not put the bridge all the way out, I plan to bolt the mudguard to the bridge through a spacer. A lot of the other frames I have built using a Sachs bottom bracket shell and no chainstay bridge which works well and for which I thank you. Thank you also for the tip about heating the outside of the stay which I did not know and did not do but will try next time.

The chain stay length won't matter for the expansion/contraction issue.
PS I've never used a bridge with oval stays, though others have.

[Paul Jacobs]

The chain stay length won't matter for the expansion/contraction issue.
PS I've never used a bridge with oval stays, though others have.

They were round oval round stays, only the second time I have used these. They were also stainless steel which may have an influence although I have used oval stainless chainstays before with no apparent problems. As for the length, I was thinking along the lines that if a dropout moves 1mm with a 410mm chainstay I could expect a movement of 445/410 = 1.09mm because of the extra length in the stay for the same expansion/contraction. More exactly I suppose 370/335 = 1.104mm as the bridge is 75mm from the centre of the bottom bracket.

[e-RICHIE]

As for the length, I was thinking along the lines that if a dropout moves 1mm <cut>

Best not to think. Simply place the surrogate axle in, braze carefully, and wave some heat on the exteriors after each braze. PS If you're on the new end of all this, I'd ask why stainless. That's a question that shouldn't alarm or surprise you coming from me.

[Paul Jacobs]

PS If you're on the new end of all this, I'd ask why stainless. That's a question that shouldn't alarm or surprise you coming from me.

I am definitely a novice and a mature novice at that so I am not trying to make a career of this, just trying to improve my efficiency and expand my knowledgeDSC01302.jpgDSC01303.jpg. One reason for using stainless is the same one that George Mallory gave for why he wanted to climb Mount Everest - "because it's there". I suppose I am trying to achieve the same aesthetic that formerly drove people to chromium plating. It is only the chain stays and seat stays which are stainless and they will be left largely unpainted. I am using Reynolds 921 stainless tubes which are thicker walled than 953 and can be cold set.

[e-RICHIE]

One reason for using stainless is the same one that George Mallory gave for why he wanted to climb Mount Everest - "because it's there". I suppose I am trying to achieve the same aesthetic that formerly drove people to chromium plating. It is only the chain stays and seat stays which are stainless and they will be left largely unpainted. I am using Reynolds 921 stainless tubes which are thicker walled than 953 and can be cold set.

Mallory was experienced, and Everest is a climb among climbs, or near it. To be a better maker, keep it simple. Walk slowly and deliberately, then ascend.

[Eric Estlund]

Mallory did die on Everest.

Essentially it's just uneven heating and cooling between two stays, a bridge and the filler betwixt. I'm sure there is math to correct it, but do as Richard says and keep on practicing and exploring the technique. Don't get overly upset when materials respond to a dynamic process.

[Curt Goodrich]

My addition to this already good information is to flux the perimeter of the stay where the bridge is going. Also, start wider than your intended width. That amount will vary according to the individual. Fit the bridge well. Braze it into place. Then heat the outside of the stay for approximately the same duration as it took to braze the bridge in. Heat to approximately the same temp. The flux will help in illustrating this. Then allow to cool. By heating 360° of the stay, you essentially stress relieve the area thereby reducing warpage. With regard to stainless, this step is even more critical because most stainless steels warp more than cromo.

[Paul Jacobs]

Thank you everyone, I have found the advice and discussion really helpful.

My understanding is that when you heat one side of a tube the part heated wants to expand but is held in check, to a degree, by the cooler opposite side of the tube. This induces micro distortions, or stress if you like, in the heated area which make it end up shorter when cooled than it was at the start. This bends the tube inwards at the point of heating. By heating both sides of the tube, this phenomenon can be minimised. A similar argument explains why a bracket shell can end up narrower at the top than the bottom after brazing in the seat tube.

This did not seem a big issue to me. I had a way of working which, even with the dreaded stainless only got me 0.4mm out but I like the idea of stress relieving the whole tube instead of starting off well over the intended width to compensate for the distortion.

Although I am relatively new to frame building, I am not new to metal work (although it is not my profession). I first used a lathe, milling machine, bench and brazing torch in 1967. In the interim, a lot of my interest has been on woodwind instruments including silver soldering of silver for keys and tubes etc. Silver conducts heat so well it is difficult not to heat it evenly all the way through.

[VertigoCycles]

I admit to only scanning this thread quickly, but I didn't see mention of the fact that the filler is contracting as it cools. Your miter tolerance also plays a role. If the bridge fits loosely on the obtuse side, it'll have more opportunity to pull the gap closed dummy axle in place or no.

[Amaro Bikes]

Wise experience has already been talk here, and I guess brazing works slightly different from TIG, but I add to Sean's advice.
Not only the tighter the miter the better to avoid stays narrowing (as potentially filling the miter gaps), but I do also kind of miter the bridge slightly wider than needed, so the bridge goes slightly "forced" (so if I took the dummy axle out the dropouts would go couple of mm wider), this results in both questions compensating themselves (the welding narrowing vs the wide miter opening) therefore the dropout spacing ends up as it was before welding.

Cheers

[crumpton]

And I'll add, do not stuff the bridge in under load. everything relaxed. It should just fit.