repechage

There are a number of silver fillers now.

Cycledesigns (Fred Parr's old brand) might still be around. They had a filler one could even create a fillet with.

You have designed precise matching parts, but they won't be matching that perfectly due to the slight taper. ( I really don't think you are going to get enough "spring" flex without permanent deformation if you design the dropout for the max bore of the taper)

I will return to my previous concern, four pointy tips per stay...and that square slot that I also just don't like while I am giving a crit. think stress riser... (yes, you could radius both parts but you still have those tips- heat risers, pretty easy to overheat... you need a practice assembly)

The problem is you have a very thin stay adjacent to a thick hunk of dropout... it is going to be a trick to get the inside core up to temp and NOT overheat the tube points. ( am I making a decent argument yet?)

I would really consider not slotting the stay for the dropout. but slot the dropout for the stay, if you use "SLA" deposition (I know slang terms)

you can create those slots in the dropout to match the stay shape on the outside and even one surface of the inside.


I don't like the look of the vertical dropout earlier in this thread either... 3Rensho Super Ends for one did a better job- Confente designed a set too that never were realized, he was going to bore out the stay a wee bit to help make up for the taper mismatch, not full contact but reasonable. Slotting? no stinking slotting.

(And everybody thinks he liked hand work, he was fast and effective but wanted to "plug and play" like the industry eventually did.)

iab

I have no issue changing stuff, it's just ones and zeros at this point. Give me a week or two for a rev. I'll likely make plastic prototypes too.

rhm

Since you're revisiting and updating this thread, I'd like to revise my comments about the belt drive, which was in these two posts:

Il Gregario - a custom project

Il Gregario - a custom project

The bike I ride, with the Gates belt drive, is a folding bike that I use mainly in the summer. When I posted the above comments I had had the bike only a few months, having bought it at the end of the previous summer. It was still 'new.' Since then it has been my main commuter for three summers; this summer I put about 550 miles on it; and previous summers I probably rode it that much or a bit more. Total mileage on the bike is probably around 2000 miles.

Anyway, over that time and use, the belt drive has softened up nicely. It rides more smoothly, and I no longer notice the drag I noticed before. All in all, I think I can withdraw my misgivings about the belt drive. I'm still not very fond of the Shimano hub, but that's another story.

iab

I changed my mind again. Going back to a fixed/free flip-flop hub, chain drive. It will only get city use and I don't plan on living in a hilly city. I have a really cool crank and I love FB 3-piece high flange hubs.

iab

Updated with square ends.


55_Gregario_Rear_Dropouts by iabisdb, on Flickr



56_Gregario_Rear_Dropouts by iabisdb, on Flickr

unterhausen

I like slotted dropouts. The slots are cut at an angle to the stay so nothing needs to be bent. You can use 45 percent silver to fill big gaps or get some Cycle Designs fillet pro. Any metal powder is painfully expensive, but I guess the prices have gone down quite a bit. I know there are people that are printing dropouts, mostly Ti.

I wouldn't think lug thickness would bother an experienced builder. There are always differences in thickness. I think the importance of thin lugs is overblown. I mostly see failures at places where the builder has thinned the tube or made some other error.

I must not have been involved in the framebuilding forum discussion about chain stay length. People mostly go after short chainstays for a perceived advantage in steering quickness. Otherwise, you can't tell the difference. Even the quickness advantage is questionable in my mind. I do have to admit that I like shorter chainstays on bikes I design. I consider it a moral weakness.

gugie 

0.5mm is a pretty big gap for silver to fill. Here's a thickness (gap) vs tensile strength graph for one type of silver alloy filler:



At half a millimeter, you'd make up for the lower strength with a longer socket.

I'd make the socket nominally the size of the tubing, then sand/file it down until you've got essentailly a slip fit.

gugie 

I'm wondering if your plan is to 3D direct print or to do a lost wax casting. @rhm has used shapeways.com, they'll print in wax, then use it for a lost wax casting. This eliminates the issue of porosity. I'd also warn about accounting for shrinkage (insert Seinfeld joke here). As I mentioned above, I'd err on the side of too much material, and file/sand to fit. This will make for a stronger, cleaner joint.

iab

Due to the tubing taper, what you suggest cannot be done.

iab

My old bikes have chainstay length that cannot be purchased for a "road" tubing set. I like it. I have no need for a short wheelbase, quick turns.

gugie 

Why? How much does the tubing taper over the length of the socket?

Don't let perfection get in the way of good enough.

iab

As I wrote before, the chainstay & fork blades taper most. On the dropout (narrow) end of the chainstay, the ID is approximately 10.2mm. 25mm moving toward the bottom bracket, the chainstay ID is approximately 11.4mm in diameter. If the dropout socket OD is greater than 10.2mm, the chainstay will never slip over it. At one end, it will be match metal, on the other end of the 25mm socket, there will be a 0.6mm gap. At the midpoint of the socket, 12.5mm, the gap will be 0.3mm. And so on.

I cannot have a reverse taper on the plug. My idea before was to cut the stay so the ends would flex, hopefully elastically. But the couple three builders talked me down from that idea to the straight diameter socket.

iab

Lost wax is much more spendy over printing or rapid machining. More suited for doing multiple parts, not one-offs.

Shapeways "steel" is some weird steel/bronze mix.

repechage

Time to buy the tube set, or at least the stays. I am not sure how the Stainless sets are created, but the "classic" sets from Reynolds and Columbus (even Reynolds Taper gauge) had the wall thickness slightly increasing at the minor diameter end. The termination wall thickness could well be over what you need, so, filing or boring out the stay could yield a constant ID for a bit of length, if it was 10 mm, you would be in a terrific place.


As to the latest design, I do not think your plug insert ends need be so long, and I would consider giving them a relief... trying to come up with an analogy.. like a shortened single part clothes pin from yesteryear. The clothes were dried on the line. (still popular in Oz, btw) Not shown but they could be hollow. based on the planned creation method.

I would not set aside investment casting, a big part of the cost driver is the creation of the wax mold pattern, create those wax masters one off, or near one off by "analog" methods and the cost to do what would essentially be prototype samples may not be out of line.

One aspect of the previous design I liked was the location of the visible silver edge of the joint, this latest design will create an off color ring however thin that interrupts the surface.

You have kept the "plate" thickness of the dropout constant... and the edge square... a designer's choice. Send me a PM with an email address and I can email you an image of the track dropout of my Confente. He did a few details that might push you to change your approach a small amount.
When I view a "plate" dropout I just immediately think stamped. My prejudice.

The latest configuration will have benefit of the builder not cursing you though.

unterhausen

you probably want to go a little large on the stay plugs so they can be shaved a little. The taper isn't that big of a deal for a plug dropout. It's not going to pull loose unless the brazing is done horribly wrong. I use one of the inexpensive inspection cameras (look for borescope on ebay) and check the brazing from the inside. The plugs on your dropouts look very long to me. For an example, look at this one https://www.cycle-frames.com/bicycle...lug-style.html

iab

I thought about that. My concern is the brazing method I was told used with sockets. I was told you coil the rod and attach it to end of the socket, place tube over coil and socket and braze from the inside out. If the socket is hollow, I would worry the molten silver could run down the "inside" and not the "outside".

Also, all diameters will have a 0.2mm interference. It is easier to remove than add material.

I'm also not keen on the transition from the tube to the dropout. My plan is to put an inner bevel on the tube so there is a sharp edge butting up to the dropout. I think that will make a crisper line.

How long should the sockets be? The ones unterhausen posted look to be about 10mm. Tengrainbreads' look about 20mm. My model is 25, knowing I can always cut them off.

And finally, I'll print everything in plastic and post pics.

repechage

Since the creation of these ends will be an additive process, and undercuts allowed...

Why not create a reduced down region to "seat" the braze ring and encourage the material to go where you want it, keep the hollow, a rim with a reduced diameter essentially.

One thing I wonder about the process, is how "clean" it is after forming, Like aluminum, stainless likes to be very clean just prior to joining, this part requiring to be abraded clean? Like Investment cast parts do?

Yes, I would place a ring of silver at the end, braze the sub assembly oriented in a way that gravity is your friend.

Really not very different than many investment cast fork crowns with internal spigots.

I think 12-15 mm should be enough engagement. 20 mm for the seat stays just seems overkill.

repechage

not part of this project, but I liked short chain stays, even had a bike that had chain stays so short to get the rear wheel out without deflating the tire the forward 4 mm of the horizontal dropout was ground away. That bike was super short, wheelbase of 37.25". overlap? Lots. with a high bottom bracket 43 years ago it was a terrific criterium bike. With 165mm cranks I could pedal through corners that my competition would try and do the same only to dig in a pedal. Made winning easier. Having written that, I think short chain stays are not necessary, 415mm is just fine for a "race bike", only now "gravel" bikes are returning to that attribute.

I do wish from time to time I still had that bike, it was just Fun.

iab

I like the idea of building a recess in the socket end for the brazing rod. Is 1.6mm typical for rod diameter?

unterhausen

yes, silver rod is all 1/16", 1.6mm.

You can also build half moons into the shoreline so silver can be fed in from the outside. Lots of builders notch the stays/fork blades to allow this. If you do put a groove in for the filler, make sure it stops before it gets to the shoreline.

clasher

Might have good luck filling the gaps with nickel-silver rods like all-state 11.

iab

I'm fairly certain that 90% of the time I won't like it 100% of the time. Onto round 2.



20181026_081347 by iabisdb, on Flickr

iab

I wrote that over 7 years ago. Whoopsie daisy.

Any how, progress! Rapid CNC is much cheaper now, and I make more money, so win-win. I had these bits and bobs made. Obviously there needs to be some welding and polishing done, but the stem/bars/levers are coming along nicely in my opinion.

Bars 02 by iabisdb, on Flickr

Bars 01 by iabisdb, on Flickr

Bars 03 by iabisdb, on Flickr

Bars 04 by iabisdb, on Flickr

Bars 05 by iabisdb, on Flickr

Bars 06 by iabisdb, on Flickr

rccardr 

Very impressive, as usual.

smontanaro 

A perhaps better place to investigate this would be on the framebuilders forum, but in my case, I filed the bejeezus out of the tabs on my dropouts. (Stainless steel is HARD!)



It's a lather, rinse, repeat process. Note that my chainstays weren't slotted. I filed the ends at the appropriate angle (7º comes to mind) to complement the filing of the dropout tangs and achieve the correct angle. Once the chainstays and dropouts are brazed, the assembled chainstays are brazed into the bottom bracket. Note in the description all the time required to get everything aligned just so before committing to the final brazing. Yours probably won't be as critical given you are planning to use horizontal dropouts. Vertical dropouts leave little room for error. OTOH, if you are 3D printing your dropouts, you should be able to just bake in the desired angle.



No matter what direction you eventually go, you might find it worthwhile to spend a day at Doug Fattic's shop (or another framebuilder) to figure out some of these details.