Jig Cutting Aftermath

Stopped at my friend's house to cut the window in the fork jig and the curve for the blade bending jig; no pictures yet, but overall I think it worked out alright.  There's still some finishing work to be done, but here are some thoughts about how I decided to do this:

• We used a jigsaw, which may or may not have been the best tool for the job.  Kind of a no-brainer, but it was far easier to cut through the MDF than the 1"-thick piece of maple.
• The jigsaw's turning radius was not very tight, so the window on the fork jig is a little larger than I wanted.  This isn't terribly critical, but I'm a little worried that the stiffness of the board may have been compromised somewhat.
• The cut on the maple somehow ended up not perpendicular to the thickness of the board.  The board is actually very slightly warped, it seems.  Again, I don't think this is a huge deal, but I'm going to try to sand it flat anyway.

I still need to attach the clamping blocks to the fork jig and cut a trench in the maple to hold the blades.  I think there's enough scrap left from the maple to make a separate tube block out of it, which should make the bending process simpler.

A scroll saw (or anything with a narrower blade and turning radius) would have worked better for the hole in the frame jig, and a bandsaw probably would have been the best tool for the maple.  At any rate, you use what you have at hand.

I'll be traveling for the next week or so, so there won't be more construction going on until next weekend at the earliest. 

Gettin' Jiggy With It

Time to start cuttin'!  I need two jigs for the bicycle fork - a blade bender and a layout jig.  You can purchase professional ones for several hundred dollars, but for a home workshop that's overkill.  Keep in mind that the total cost of your bicycle includes more than the raw materials; every tool, all the consumables, even the electricity to run the lights add to the cost.

I found a perfect piece for a layout jig at Home Depot:

On the left is a pre-cut shelf made of MDF, about 15"x36".  On the right is a 1"-ish thick piece of maple, about 8"x12" that I got at a local woodworking shop.  The maple will become the blade bending jig, since it's strong enough not to deform while exerting a concentrated compressive load against the grain - as I will when I'm trying to wrench a metal pipe around it.  Using a string compass, I scribed an 8" radius at one end for the bend on the fork blades.  Hopefully there will be enough scrap left over that I can make a tube block to hold the blades.

The MDF shelf needs three operations done on it: attach a clamping block for the steerer, cut a window out for the crown, and attach a block at the bottom for the dropouts.  The dropout block will have a threaded rod through it so that the dropouts can be "clamped" in place with a set of nuts; this will hopefully properly set the distance between the dropouts and keep the whole fork in alignment while it's being brazed.

I scribed marks on the shelf:

The line down the board marks the center of the shelf.  To be more "proper" with my marking, I should have scrawled something in the window - a zig-zag line or the word "CUT" to show what's being removed.  The line at the bottom is where the dropout block will be placed.

There's some scrap wood in the basement, so I took a piece and cut the blocks out of it:

It's not large enough to provide the proper height for the dropout block, so those two small pieces get glued together:

Some wood glue and a clamp, then leave to dry.  The cuts were all made with a hand saw I had lying around and are a little crooked, so I have to go back and sand them flat.  The dimensions of the blocks aren't terribly critical - the spacing for the dropouts will be accomplished by moving nuts on the threaded rod that will be in the dropout block, and the steerer block only needs to be long enough to securely clamp the steerer to it.

Mocking up the jig:

There's a problem with this.  The steerer block actually should not be adjacent to the centerline of the board like that; because the centerline of the steerer tube should be coincident with the centerline of the board, the block actually needs to be 1/2" to the left (I have a 1" steerer).  Luckily, nothing is attached yet.  I did actually make a drawing for the jig with the correct offset, but failed to copy it over to the board.

I don't actually have the tools to cut the window in the jig (or the radius on the fork bender), so it's off to a friend's house to borrow his scroll saw.

Order of Operations

Building stuff is so much more than the act of joining - putting a nail into a board, welding a seam, tightening bolts.  There's a lot of extra work that goes in to making that is too often behind the scenes; that's one of the main reasons I started this blog.  Framebuilder sites tend towards shiny photos of fire! and glowing red metal! and the you have the finished frame, and the next photo in the series is the happy builder riding around on a completed bicycle.

There's a bit more involved.  I've started into a familiar loop, and it goes something like this:

• Well, it's time to get started with Task X (cutting tubing/welding/whatever)
• Assemble all the parts and tools
• But... I can't do Task X without doing Task Y first!
• GOTO 10

I bought tools, cleaned up the basement, made an impromptu workshop, and bought the components for the fork.  Time to get cuttin', right?!

Not so fast.

With this in mind (and because I haven't uploaded jig photos to my computer yet), here's my to-do list for the fork:

1. Set up workshop
2. Design the fork
3. Purchase tools
4. Purchase material
5. Design fork jig
6. Design fork blade bending jig
7. Acquire jig materials
8. Layout jigs
9. Cut/assemble jigs
10. Clean preservative off fork parts
11. Bend fork blades
12. Cut blades to length
13. Cut blades for dropouts
14. Dome blade ends
15. Cut steerer to length
16. Fit blades to fork crown
17. Fit steerer to fork crown
18. Partially finish fork crown
19. Mock up fork on jig
20. Thorough polishing and cleaning of all joint areas
21. Braze dropouts to fork blades
22. Braze steerer to fork crown
23. Braze blades to fork crown
24. Remove excess flux
25. Finish all joints
26. Thin dropouts
27. Finish fork crown
28. Final alignment

I'm at step eight right now.  I still haven't purchased brazing consumables yet.  Notice that there are three steps out of 28 that directly involve applying fire to metal.

All of these steps are important for a quality product.  If you want evidence yourself, go find a welding class nearby.  I took a MIG course a few weeks ago that was designed for sculpture artists which was very valuable.  Try to take two rusty pieces of iron and weld them together - it works, but only kind of.  It's difficult to strike an arc, you're more prone to blowing out the metal, and the joint isn't very strong.  If you're making a garden sculpture maybe that's ok, but  I'm going to be trusting my ass to this bike, so quality and attention to detail are important.  The last thing I want is the fork to collapse underneath me while I'm riding.

Once I get over to a friend's house to borrow his scroll saw, I'll describe how I went about building my jigs.

Tools

My order of tools arrived in the mail the other day, and I think I have just about everything I need to make this bike.  Here's everything I've purchased so far:

Tools!

 As I mentioned before, I'm starting with basically nothing; there was a not-insignificant amount of money put into simply getting a usable work space set up.  What I already owned includes:

• Assorted screwdrivers in various sizes and condition
• A set of Park Tool box wrenches (for working on my daily rider)
• 12" crescent wrench
• Leatherman multitool
• Other random bits and bobs - paint scraper, wire strippers, etc.

I purchased most of this equipment from Harbor Freight; their prices are very reasonable (especially when there's a sale, like those ratchet clamps for two bucks each) and most of the tools seem of decent quality.  What I ordered from them (left-to-right in the photo):

• Framing square (for laying out jigs and measuring parts)
• 6" digital calipers (measuring parts)
• Protractor/angle finder (verification of the frame layout)
• Claw hammer (...I have never owned a real hammer.  I needed one.  Probably won't use it for this project)
• Automatic center punch (marking hole locations on parts)
• 12" hacksaw (cutting tubing)
• Needle file set (tube mitering and joint cleanup in tight areas)
• 2 6" C-clamps (clamping to jigs)
• 4 ratchet clamps/spreader bars (can't have too many clamps, and they were on sale)
• 12" file set (tube mitering and rough joint finishing)
• Wire brushes (cleaning tubes and removing flux)

Total cost: about $100.

The work bench and MAPP gas torch (not shown) came from Sears.  The bench is regularly $70, but I bought mine as an open-box item for $45.  The torch head was $50 - much more than I expected, but it can accept different tips and is designed for brazing/silver soldering.  MAPP gas tank cost was about seven dollars.

Also shown in the photo are a set of emery paper in various grits and a file card, which I picked up at Home Depot when I went to get lighting.  I spent another hundred dollars or so for the lights, light bulbs, emery paper, file card, bucket (for washing parts and - for now - holding tools), hook bolts to hang the lights, nails (needed more nails to hang picture frames), extra hacksaw blades, and the base for my fork jig.

I still have not managed to find layout fluid, which is used to dye metal and allows you to scratch cutting marks into it.  Harbor Freight's website is pretty thin on consumables and the ever-helpful Home Depot employees had no idea what I was talking about.  I can probably finish the fork without it, but once I need to start filing miters into the frame tubes I'll need some way to mark lines on to the tube ends.

If I never make another bicycle after this, I think the only "wasted" tools I've bought will be the files and the torch.  Everything else should be useful on any other kind of project.

Work Space

Here is where my projects usually start to break down.  Thinking, researching, laying out designs on paper - I could do that forever.  One problem I consistently have is that, once I have the conceptual part of the project done, the rest (as in, actually building) is just "details".

One reason this happens is that I have never had very good space to work in.  I had access to a machine shop in college, but since graduating I've been living in apartments with very little space.  Once I start wondering, Where am I going to build this?  How am I going to set up a work space? I tend to get discouraged.  After all, to have a successful project I need to have a workshop, and a bench, a vise, expensive tools,...  And it would take so much effort to get these things set up.

Fast forward to a few weeks ago.  We live in a triple-decker apartment now, and one of the benefits of the apartment is some basement storage space.  The main room of the basement is split down the middle by a partition, and we have one half of that room to ourselves.  It's been filthy and cluttered since the day I moved in - the landlord has been keeping old building materials in our space, and previous tenants left whatever they didn't want when they moved out.  I only had a handful of boxes and a bike that I wanted to keep in the basement, so I never bothered to clean.

It finally got bad enough that my girlfriend insisted we at least sweep the floor.  I bought a broom and a shelving unit, we threw out all the old boxes and other crap that had accumulated, and we reorganized the rest of it.

There's a lot of space down there!  And power outlets that I didn't know existed!  This is when I decided I could actually have a workshop.  And even though I decided to start the bike project earlier than this afternoon of cleaning, I was planning on working outside and hoping that I'd finish before the summer ended.

One last challenge existed: there's almost no light in the basement.  Our whole side of the basement is lit by a single light bulb with a switch upstairs in the kitchen.  This weekend, then, we made our way to Home Depot and bought a bunch of shop lights.  I got several different types - a couple of "cage" style lamps, a clip light with a metal reflector, and a halogen shop light - and strung them up all over the basement.  The cables are draped between the exposed beams and over water pipes, but it seems like there's enough light down there to work, for surprisingly little money - the cage lights only cost about five dollars each!

The basement is small but usable.

I may discover I need more lighting as I get my hands dirty, but for now the change is so dramatic that it's like, well... night and day.

If you need lights, also remember some extension cords and power strips.  I bought one of each and probably needed at least one extension cord per light.  Right now they're mostly clustered around the power outlet.

So, now there's some space that's been swept, lit, and ready to go.  I needed a surface to work on, though.  I bought a Black and Decker portable work bench (this one, actually) before cleaning out the basement, figuring that I would need to carry it outside every time I wanted to work.  I think it will be especially useful for working on tubing because the two halves of the benchtop split apart to work as a vise and the bench includes some pegs that fit into the tabletop that are shaped to hold tubes.  I should be able to hold jigs in between the two halves of the tabletop and tubing in the pegs.

I'll take some pictures and add them here to show the workspace; in the meantime, I need some tools.

Detail Design - Part 2

With the main triangle designed, it's time to think about the rear triangles and the fork.

The rear triangles are made of the seat stays and the chain stays.  The chain stays connect the rear wheel dropout to the bottom bracket, and the seat stays connect the dropout to the top of the seat tube.  The distance between the dropouts is dependent on the wheel and gearing that are desired, and 130 mm is a fairly standard width.  My design calls for a 130 mm distance between the two dropouts.

The chain stay length is determined by the size of the wheel and any other accessories, i.e. a fender.  With the 700c wheels I've chosen and the desire for a fender, BikeCAD tells me the chain stays need to be 425 mm long.  The bottom bracket drop (the vertical distance between the rear wheel axle and the center line of the bottom bracket) for this design will be 73 mm; this is calculated as the difference between the height above ground of the rear wheel axles and the height above ground of the bottom bracket (the bottom bracket height).  The BB height on this design is 267 mm and was determined by BikeCAD based on my lower body length, length of the cranks, and to provide for adequate ground clearance in a turn.

Since the seat stays do not fit into a socket at the top of the seat tube, their lengths are not quite as critical as the other tubes.  I plan to size them during construction and cut to length.  Some more trigonometry yields a seat stay length of 519 mm or so.

The seat stay and chain stay lengths are measured from the axis of the rear wheel - when these tubes get cut, I have to take into account the size of the dropout as well as (as with the other tubes) the length that gets inserted into the lug sockets.

Chain Stay Dimensions provided by BikeCAD

The fork is the final part of the bicycle to be sized; the critical lengths here are the fork blade lengths, the rake, and the steerer length.

BikeCAD provides the length of the fork blades; this length is again determined by the size of the wheel and the included fender.  This length is defined as "the straight-line distance between the front wheel axle and the front brake hole" (Talbot).  BikeCad provides a dimension from the top of the fork crown to the vertical height of the wheel axle - 375 mm in my case.

The rake is the offset distance from the front wheel axle to the center line of the fork blade.  This offset affects the ride quality and steering; less of an offset provides quicker turning but a stiffer ride.  I opted for a 50 mm rake, hopefully for a relatively smooth ride with adequate steering quickness.  The rake length is a design choice.

The steerer tube connects the fork crown - where the fork blades connect and the (caliper) brake assembly is mounted - to the upper half of the headset.  It must be long enough to fit into the crown, through the bearing between the crown and the head tube, through the entire head tube, and the upper bearing assembly of the headset.  At this point, I need to decide on a headset and measure the depth of the hole in the fork crown for the steerer; since I haven't done this yet I don't have a dimension for the steerer.  I have a set of calipers on order and will measure the crown when they arrive.

Fork Dimensions provided by BikeCAD

That's the entirety of the design process for now.  Talbot suggests making a half-scale or full-scale drawing, but I think that having the BikeCAD model on a computer in the shop should be sufficient for my needs.  I may sketch out full-size drawings of the joint areas if I think I need them for fitting the tubing together.

Detail Design - Part 1

With a general design complete, the next step is to start taking some measurements and laying out the dimensions of the bike.  This is where custom frame building really shines; a custom frame is like buying a tailored suit - it fits perfectly for your body type.

I used two resources for sizing my frame, Richard Talbot's book Designing and Building Your Own Frameset and BikeCAD, a java applet for designing bicycle frames.  Talbot provides some instructions and tables for sizing a frame according to the C.O.N.1 method, which was developed for sizing racing frames by the Italian Central Sports School.  Talbot states that the method is also applicable to touring frames, and the measurements will provide you with the dimensions for your seat tube and top tube.  BikeCAD has the ability to take just about any of your body measurements - you can give it only your height, or all of the C.O.N.1 measurements - and spit out a properly sized frame.

The three measurements required for the C.O.N.1 method are:

• Distance from the floor to your pubis (A)
• Distance from the shoulder joint to the wrist "fold" (B)
• Distance from the pubis to the top of the sternum (C)

 The "lower limb" measurement (A) determines the seat tube length, and the "upper torso" measurements (B+C) determine the top tube length.  Measurement (A) is slightly longer than your inseam, as it is measured directly to the base of the pubis bone - you may need to feel around a little bit to get the right position.  The arm measurement is taken from the back; raise your arm slightly and feel for a depression by your shoulder blade, then measure to the fold of skin made by bending your wrist back (as if you were doing push ups).  The top of the sternum can easily be felt by hand.


Talbot provides a table for determining the tube lengths from these measurements; you can also put all of these measurements into BikeCAD and get the lengths.  I did both and found that BikeCAD suggested a longer seat tube length than the C.O.N.1 method, but roughly equal top tube lengths; if you already own a bicycle that you find comfortable, you may wish to measure its dimensions.  Another alternative is to go down to your local bike shop and get fitted there.

My measurements provide me with a seat tube length of 53.7 cm and a top tube length of 56.8 cm.

The head tube and the down tube are sized based on the wheels you've chosen, the top tube angle you want, and the head and seat tube angles.  If you haven't thought about wheels yet, now would be a good time.  I picked 700c x 28 wheels because they're fairly standard and not too beefy.  I defined the head tube and seat tube angles as 74 degrees each.  I also wanted a flat top tube, so I put all of this into BikeCAD and got a head tube length of 12.7 cm.

Why did I pick those particular head tube and seat tube angles?  Talbot provides some guidance: the seat tube angle "...usually falls between 68 degrees and 75 degrees..." with time-trial bikes going even steeper.  Seat tube angle affects the stiffness of the ride and the center of gravity.  Head tube angle affects the steering qualities of the bike and the front-end stiffness; I want a little snappier handling than my Raleigh provides, so a 74 degree angle with an associated rake on the fork (more on this later) seems like it will provide fairly responsive steering while still having a comfortable ride.  This head tube angle also shortens the wheelbase a little, which affects the center of gravity.  Since I'm expecting to have weight over the back wheels fairly often (loaded panniers on the rack from grocery shopping), this will hopefully minimize the effects of the rearward center of gravity while carting goods home.

Also, lugs with these angles are very easy to obtain.  I don't want to go fiddling around with bending lugs at this point.

Determining the down tube length is left to geometry and bit of math.  The free version of BikeCAD does not provide this dimension, unfortunately.  You can idealize the frame with a set of lines that make a quadrilateral, then split it into a series of triangles and use trigonometry to determine the down tube length; my result is a 60.2 cm down tube.

 

 Main Triangle dimensions provided by BikeCAD

That's the main triangle of the frame; the rest of the frame and the fork are left to draft up.

On Design

Always start a project with a set of requirements. They don't need to be specific, but they will help you constrain the project sufficiently enought that it won't overwhelm you. Especially for a purpose-built item like a bicycle or anything that may require tight tolerances... like a bicycle. Art projects can be a little more free-form, but in the past I've at least had a list of requirements in my head for those.

The questions I asked myself before starting this project included: What do I want to do with this bicycle? How far will I be riding it? Will I be carrying anything?

The first question - What do I want to do with it? - is the most important to ask yourself before starting anything else. Again, the answer doesn't need to be very specific. My answer was, "I want to tool around the city and go to the grocery store." Other answers could include rock-hopping, century rides, multi-day rides, "I want to carry my dog with me wherever I go!". Think about where you live, and where you will use it.

Based on this answer, I determined a few basic "shalls" for the bicycle:
-The user (me) shall be capable of riding the bicycle for up to five miles at a time without significant fatigue.
-The bicycle shall be capable of carrying a full load of groceries for two miles.
-The bicycle shall incorporate safety and comfort equipment suitable for city riding (essentially, lights and fenders).
-The bicycle shall be light enough to be carried from its storage location in the basement to the street.

That's basically it - I want to carry it up a short flight of stairs, ride around town comfortably, and carry a load of groceries. This helps determine a basic shape and a basic set of components.

What will it look like, then? A very basic triangle frame with an upright riding position, a rack, some lights, and fenders. A step-through frame may be more suitable for what I want the bike for, but the triangle is easier to build for a first frame (maybe that should be another requirement).

I have a bicycle that I can borrow the design from - a 1966 Raleigh Superbe. It's a great city bike, but it's a little large for me and doesn't have a rack. I'll try to make a slightly smaller version of it and include the necessary frame components to mount all the gear I need.

The next step is to perform a detailed design based on my body dimensions and the riding qualities I'm looking for.

Welcome!

The purpose of this blog is to document my attempt to build a bicycle from nothing. I will buy tubing, construct a frame, and then construct the rest of the bike around that frame.

The idea for this project came about mostly as a lark, but was spurred on by some discussion on MetaFilter. Posts will occur on an as-shit-gets-done basis and will include everything from initial design, development of the workspace and tooling, all the way through to spit-shining the chrome.

Why this site? Well, while researching for this project, I found many personal sites that detail people’s completed frames, but very little geared towards a bare-bones start. Usually, the sites include a few sketches of the design, maybe a photograph of the tubeset that got delivered, and – voila! the bike is finished, painted, and on the road. These sites are great for inspiration, but not entirely useful for a first-timer trying to learn how to build stuff. Hopefully, someone else can take what I hope to document here and use it to build their own bicycle.

My credentials include a B.S. in Aeronautical and Mechanical Engineering, a few years’ experience in a machine shop at school, and several years in the aircraft industry. I have never built a bicycle before, and previous projects have not included the level of precision required to fabricate a ride-able bike.

Stay tuned, and I hope this site becomes a helpful resource!

-Mike