October 2007 - Posts

Tikit conversion to internal hub gear

I've always thought that the Bike Friday Tikit should have an internal hub gear.  The range with the derailleur is less than most 7sp and 8sp hub gears and the hub gear is cleaner and better protected from the environment, all good things on a commute oriented folder.  I expect that Bike Friday released it with a derailleur because the hub gear would have added a couple of hundred dollars to the cost.

Yesterday I built a Tikit rear wheel with a SRAM S7 hub.  It was probably the most annoying wheel that I've ever built...there are 36 very short spokes (around 6" each) in a tiny little rim.  A Park spoke wrench just fit between each of the spokes.  It was very comforting to finish it.  I chose the SRAM S7 over the Shimano Nexus 8 for a few reasons:

  • A LBS (Aaron's Bike Repair) stocks the SRAM S7 and spare parts.  I've overhauled SRAM S7 hubs before and they are pretty easy to dismantle and put back together.  I probably won't need to, but it is nice to know that I can.
  • The Nexus 8 is pretty hard to find right now (specifically the shifter is hard to find right now).
  • The SRAM S7 has a similar gear range to the Nexus 8sp.
  • I prefer the clickbox to the shifter cable arrangement on the Shimano Nexus hubs.

This morning I put the rest of the bike together:

The hub gear makes the bike look a lot cleaner.  There is no derailleur hanging down low to pick up dirt and road grime.  You might ask how I got away without a chain tensioner since the Tikit has vertical dropouts.  The answer is that I was lucky -- the stock 53t chainring and the SRAM 18t cog work perfectly with the chainstay length on this Tikit.  They also give good gearing: 27", 32", 38", 47", 58", 70", 82".  I'd like something in the 60s between the 58 and 70, but this is good enough and the range is exactly what I wanted.

I bought the drum brake version of the hub and removed the stock V-brake.  The drum brake should work well in this application and will allow the rim to last nearly forever.

I wanted to experiment with some other flat bars, so I ordered a pair of "Origin-8 Space Bars".  They are sort of a hybrid between Moustache bars and Albatross bars, hopefully it is a combo that works well for me.  The make the folded bike smaller than drop bars and work more easily with the hub's rotary shifter.  They also have some rise, so I'll probably need to cut down the stem mast a little bit.

The bike folds to about the same size as the stock Tikit:

I spent all day working on projects, so I didn't get to give the new setup much of a test ride (just a few laps around the block).  I'll report back on how the bike works in this configuration.  I think it'll work very well. 

Posted 29 October 2007 04:38 AM by AlexWetmore | 11 comment(s)
More on bicycle lanes

Portland had a bicycle accident on Thursday that resulted from a right turning cement truck hitting a cyclist who was going straight in the bicycle lane.  This has prompted a response from many of my favorite blogs.

Beth Hamon wrote a great piece questioning the use of bicycle lanes on very busy arterials.  This resonated with me:

Striping West Burnside -- one of Portland's busiest streets -- with a bike lane is a bad idea. Allowing bicyclists to occupy a driver's blind spot (alongside the car) means they run the risk of getting hit if the driver doesn't see them. This is not a question of fault. This is a question of poor planning that can and should be fixed. Lose the bike lane, post a lower speed limit and instruct bicyclists to *take the lane* in front of or behind the motor vehicle so that they can be seen. If this doesn't work -- either because cars and bikes won't play nice there, or because West Burnside eventually becomes the higher-speed backdoor to Beaverton known as Cornell Road, then reconsider the function of such higher-volume, higher-speed roads. Bikes belong some places better than cars do. Cars belong some places better than bikes do. This isn't rocket science. It's planning, and until there are no more cars in the world that planning has to work both ways.

Portland is in a unique position (for a medium/large city in the US) of having bicycle lanes everywhere.  Perhaps they have too many?  Does it do a disservice to cyclists to try and isolate them from all car traffic?  I'd argue that it does -- cyclists need to understand how to safely get through intersections.  Bike lanes don't help here.

Kent Peterson followed up with an entry coining the term suicide slot:

If you look at Michael Bluejay's spot (and I really hope I've convinced you to take the time to look at it), you'll see that several of those ten common crashes involve the cyclist being where the driver isn't looking and/or being in the driver's blind spot. I call this the "suicide slot", being to the right of a right turning car.

Now you may say "suicide slot" is a loaded term, that I'm blaming the victim, that the driver should see the cyclist. Well, we can talk about what drivers should do, but as near as I can tell not everybody does what they "should" do. So even even they "should" look to the right, I'm thinking some won't. And if I'm off to their right anyway, well that strikes me as suicidal. But maybe it is a loaded term. Loaded like a gun. And like a gun, it can kill you.

A couple of months ago John Speare posted a photo of a great sign in Redmond that tries to warn about the Suicide Slot:

 

I like this sign.  It is better to have the lanes properly stripped (or not there at all), but this gets across the message in a easy to understand graphic.  I don't think that having this sign everywhere would help, but having it at a few key intersections will pass along some education in an easy to digest form.  I'd like to see this sign at Eastlake and Furhman.

At work we have an internal mailing list that cyclists communicate on.  Yesterday someone posted a slide deck from New York City showing their new bike lane plan.  It puts parked cars as a buffer between the bike lane and car traffic.  This is a common practice in some parts of Europe too.

I don't like this design.  It assumes that there are no intersections. 

My first observation when I see this graphic is that the bike lane just before 18th is lining up to put bicycles in the suicide slot to the left of left turning cars.  My second observation is that bicycles are completely trapped if they want to turn right onto 19th.

The deck says that NYC solves this problem with bicycle only signals.  That is a solution that doesn't scale.  When you start having roads with this design intersect other roads with the same design you end up with twice as many signals.  For every car signal (straight, turn only in each direction) you end up with a bicycle signal. 

It isn't practical to isolate bicycles from cars in most situations.  Bicycle Paths (like Seattle's Burke Gilman Trail) work because they are generally located on rights of way that already had few intersections and where most of the remaining intersections are at different grades (the bike path passes above or below the road).  That isn't something that can be built along every road.

Drivers and cyclists need basic education on how to ride safely on the same roads.  Bike lanes need to be striped to avoid the suicide slot.  Speed limits need to be enforced so that 30mph arterials don't turn into 50mph highways.

Ride safely out there.

Posted 13 October 2007 03:31 PM by AlexWetmore | 8 comment(s)
Filed under:
Rack Building Basics -- Tubing Selection

 

Basics 

Selecting tubing for a rack is an important first step in building a rack.  The number of options is fairly limited based on weight and the accepted standards to work with things like pannier hooks.  Selecting tubing sizes will also help you decide which bender to buy.

The three basic outside diameters to choose from are: 3/8" (about 10mm), 5/16" (about 8mm), and 1/4" (about 6mm).  There are two wall thicknesses that I've worked with: 0.028" (about 0.7mm) and 0.035" (about 0.9mm).

I made this spreadsheet to help figure it all out.  For these three sizes of tubing and two wall thicknesses it provides the weight of a foot of tubing and the deflection when there is 10 pound load on a cantilevered beam.  That second measurement isn't specifically useful by itself, but it gives an idea of the stiffness of the tubing.  I hope that I got the deflection math correct, I didn't do a test to verify it.

Name Outside Diameter (inches) Wall Thickness (inches) Inside Diameter (inches) Weight of one foot tube (oz) Deflection of one foot tube (inches)
3/8 x 0.035 0.375 0.035 0.305 2.0357 0.3552
3/8 x 0.028 0.375 0.028 0.319 1.6621 0.4194
5/16 x 0.035 0.3125 0.035 0.2425 1.6615 0.6500
5/16 x 0.028 0.3125 0.028 0.2565 1.3627 0.7586
1/4 x 0.035 0.25 0.035 0.18 1.2873 1.3831
1/4 x 0.028 0.25 0.028 0.194 1.0633 1.5868

There are some interesting properties to note:

  • The 0.035" wall thickness tube of one size is close to the weight of the 0.028" wall thickness tube of the next size up.
  • Outside diameter has a bigger influence on stiffness than the wall thickness, but both matter.
  • 1/4" OD tubing slips nicely into 5/16 x 0.35 tubing, and 5/16" OD tubing slips nicely into 3/8 x 0.028 tubing.  This is very handy to know when you are splicing two sections of tubing together.
  • The heaviest tubing is over 4x stiffer than the lightest tubing, but less than twice as heavy.

Handlebar Bag Rack

A handlebar bag rack is designed to support the bottom of a handlebar bag.  It is smaller than the bottom of the bag.  The bag isn't tightly connected to it in most cases, so the rack doesn't need to be designed for high lateral loads.  They connect to the fork blades at the mid-point or higher, so there aren't long unsupported beams.  As a result of these design parameters they can be made with fairly lightweight tubing.  So 1/4" x 0.028" tubing everywhere is probably enough.  Some of the classic French racks are made with even smaller tubing (4mm outside diameter).

A standard sized handlebar bag rack has a platform of about 4x7" or 5x7".  There is one cross member at the middle of the platform and two stays going to the fork blades.  For a 5x7" platform with 8" stays this gives us a total of (5*2)+(7*2)+5+(8*2)=45 linear inches of material.  There will be a little more for the backstop and fork crown mount.

Using 1/4" tubing we can build the basic rack (no backstop, no fork crown mount) in 4oz of tubing.  Using 3/8 x 0.035" tubing the rack would weigh 7.5oz, a pretty large weight gain since we don't need the stiffness.  Going with 5/16" only adds an ounce, so that isn't too big of a deal (especially if you don't have 1/4" tubing or bender).

Porteur Rack

A porteur rack is a much larger platform rack that is designed to have the load strapped directly to the rack.  The rack is expected to work with loads of 50lbs or even more.  The stays on the rack usually connect directly to the fork dropouts, so there is a longer unsupported span (from the fork dropout to the front of the rack).

On these racks lateral stiffness is important.  We can get some lateral stiffness from smart design (good triangulation), but we also need to get some of it from the metal.  The stays are over a foot long, and we wouldn't want the rack to shift by an inch or more with a moderate side load (as you'd find when going around corners). 

Porteur racks also have a lot more metal.  I've been building them with 4' rack circumference (this is a roughly 14" by 10" platform), 4 cross members (10" each), and two stays per side (around 15" each).  That is about 148" linear inches of metal.  One made with 3/8 x 0.035" tubing would have about 25oz (still under 2lbs) of steel.

You can optimize the weight be mixing and matching tubing.  The circumference can be made with 3/8 x 0.028" tubing and still handle most loads.  3/8" x 0.035" tubing makes sense on the stays to minimize side deflections.  The cross members can probably be made with 5/16 x 0.028" since they aren't that long and each cross member doesn't carry a lot of weight.  By mixing and matching we can get the weight down to 21oz (4oz savings) without compromising the integrity of the rack.

Recommendations

  • 1/4" or 5/16" for handlebar bag racks (5/16" if you plan to zip tie a basket to the rack or carry heavier loads)
  • 3/8" for heavy duty porteur racks
  • 5/16" for cross members on porteur racks or lighter duty porteur racks

If I were starting out I think I'd get a 5/16" tubing bender.  5/16" is a good balance of being big enough to make moderate duty porteur racks and cargo racks (Bruce Gordon racks are 5/16") and 5/16" x 0.028" is light enough to make a decent handlebar bag rack.  The benders are cheaper too, a Ridgid 405 5/16" bender is about $60, while the Ridgid 506 3/8" and Swagelok 3/8" benders are about $200ea (Ridgid also makes a cheaper 406 3/8" bender, but I don't know how well it performs with cromoly steel). 

You can go too flexible.  Rory and I built a handlebar bag rack for his bike out of 1/4 x 0.028" tubing.  The stays connect to the fork dropouts and the bag connects just to the rack.  It works, but he says that it is pretty flexible with heavy loads.  I think that we'd both use 5/16" tubing if we were building it again.  We're going to build a couple of small porteur racks with 5/16", and I'll report back on how that works once they are done.

Tubing Sources

It's hard to find tubing in these dimensions in the US.  I know of two reliable sources, Aircraft Spruce and Dillsburg Metal.  Aircraft Spruce has much easier ordering (you can order online), but Dillsburg is a bit cheaper, especially if you need high quantities.

I'm going to be away for about a week, so it'll be a little while before the next entry.  Leave comments or email me if you have any requests.  I know this one didn't have the interesting photos, next week we'll be back to photo oriented entries.

Posted 12 October 2007 02:25 AM by AlexWetmore | 5 comment(s)
Filed under: , ,
Rack Building Basics -- Vice grips and spokes

I find the combo of vice grips (or other clamps) and spokes to be really useful for holding small brazeons in place.  Today I installed mid-fork eyelets onto a fork, small guides for headlight cable routing, and a headlight mount. 

Another favorite tool is the flat stainless stock that is used to secure Blackburn-style racks to bikes.  This stuff is really handy for making temporary fixtures.

I used a fender stay here instead of an old spoke, but a spoke really is superior.  Brass won't stick to it and it bends more easily.  These guides are cut for 1/4" tubing, but you can also pull the rollers out of used chain.

Watch out for the vice grips, they can distort tubing pretty easily.  Used with care they are very handy for temporary clamps.

Posted 08 October 2007 12:28 AM by AlexWetmore | 4 comment(s)
Filed under: , ,
Rack Building Basics -- Making cross members

Most racks have some sort of cross member, straight tubing that goes across the rack platform.  They are pretty easy to build.

One of the tricks is getting the two miters to be parallel to each other.  I find it simplest to put some scrap metal in the miters and then eyeball them for parallel.  You want to check early and often because you do want to be parallel when you have the tube the correct length.  In this case you can see that I'm a little off:

It is important to clean the inside and outside of the metal before brazing.  To clean the inside of the tubing I make a simple bit for the drill.  This is a piece of 1/4" (or smaller) tubing with a slot cut in it to hold a small piece of shop cloth.  Scraps of shop cloth are always somewhere on the workbench.

Support the tube in a vise and put the bit inside of it.  The sand paper will remove any deposits on the metal and make for better joint.

Once you have the cross members cut, mitered, and cleaned you need to lay them out.  I find that it is best to number which position each one is in and draw alignment marks on the cross members and the rack.  I measure to get even spacing across the rack, but use my eye to determine when they are parallel to each other and the sides of the rack.  The rack platform may not be a true rectangle and it's more important that everything look good than that everything is perfectly measured and square.

The rack is parallel to the ground and held in the vise after the last step.  With everything laid out I tack each of the joints with a small blob of brass.  I don't use any clamps or fixtures for this, the cross members stay in place on their own.  This isn't the only possible option -- Alistair likes to position the rack vertically and use clamps to prevent the stays from sliding down.  Once everything is tacked I rotate the rack in the vise and make sure that the valley of each miter is pointing down.  This lets me use gravity to flow the brass where I want it to go and to get the best fillet.

Once you've brazed everything on one side you flip the rack and do the other side.

Posted 06 October 2007 01:46 AM by AlexWetmore | 1 comment(s)
Filed under: , ,
Rack Building Basics -- Fork Crown Mount

In this entry I'll show how I build the fork crown mount for a rack.  This is the part that goes from the back of the rack through the fork crown.

On a rack that is designed for very heavy loads I would recommend connecting the rack to the fork crown using two stays, one over each fork blade.  This single stay setup is not as good at resisting high lateral loads.  It does work with almost every fork though, where the two stay method only works with forks that have extra eyelets on the tops of the fork blades. 

I don't have a lathe, so I use a bolt as the threaded stud that runs through the fork crown.  This is a M6 bolt that I'm removing the head from:

The bolt will be held inside some larger diameter tubing.  First I need to cut the tubing square.  Note that to make sure that it is square I need to check the cut in two orientations 90 degrees from each other.

This is an exploded view of what is going on.  I have a 6mm (close to 1/4") threaded stud that will be inserted into some 5/16 x 0.035" tubing.  This is then inserted into some 3/8 x 0.035" tubing.  The 3/8" tubing is bent and will be brazed into the fork crown.  The 5/16" tubing is pulled out from the 3/8" tubing in this view, but when brazed together the 3/8" and 5/16" tubing will be flush.  This creates a nice face to sit against the fork's crown.

Everything is loaded up with flux and ready for brazing:

A view after the three parts are brazed together. 

Here is a shot showing how I hold the piece in place while brazing it to the back of the rack.  I'm using vice grips on the rack and a piece of scrap metal (PCI slot cover from a PC) to hold the piece in place.  I checked with a square to make sure that the threaded stud is square with the back of the rack. 

Here I am checking the angle.  The bike that this is going on has a 73 degree head tube angle, so the face of the fork crown is also 73 degrees.  I want it to sit flat (or close to it), so I need to angle my mount at 17 degrees (73 + 17 = 90).  I'm using a simple engineers protractor to check the angle.  My reference is one of the stays on the rack.  I'm a little shallow here, but that is okay.  It will make the front of the rack slightly higher than the rear, and that is acceptable.

The other option for doing this is to mount the fork crown mount into a fork and then use a fixture to hold the rack in place.  The advantage of this method is that you can see how everything will look before it is brazed together.  Alistair designed this nice little jig out of a test tube holder for holding the rack in place while brazing.  I like the jig, but find that things are a little more secure using my method.  Give both a try and see what works best.

The fork crown mount is half brazed onto the rack.  I flipped the rack over to finish the brazing on this side.

All done

Posted 05 October 2007 11:43 PM by AlexWetmore | no comments
Filed under: , ,
Rack Building Basics -- 90 degree miter

This is mitering for a basic 90 degree joint with 3/8" tubing.

The tube directly after cutting with a hacksaw.  It's a little messy:

Make a notch in the centerline of the tubing using a file or hacksaw.  This keeps the file centered in the next step:

File.  A handy hint is that a 12" Nicholson *** file has a 3/8" diameter, so it makes a nice miter into 3/8" tubing.  A 10" file works for 5/16", a 8" for 1/4" tubing.

Once filed you have a miter, but it's pretty ugly:

 

Clean up the inside of the miter with production cloth wrapped around a piece of tubing:

Then clean up the outside with production cloth:

Check the miter.  It looks pretty good to me:

Posted 03 October 2007 05:01 AM by AlexWetmore | 2 comment(s)
Filed under: ,
Rack Building Basics -- Tools

This will be the first in a multipart set of blog entries with some basic information on rack building.  I'm going to start by discussing tools.  If you have any requested subjects please let me know in the comments and I'll concentrate on them in future entries.  The next entry will be about mitering tubing.  I don't plan on doing any entries around the use of a brazing torch, I think that you should learn that hands on from another person who is skilled in brazing.

My existing toolkit already had some of what I needed, such as a good vise and a hacksaw.  However I also needed to buy a number of more specialized tools such as files, clamps, and machinists squares.

My most used tool is the vise.  It supports tubing when I'm cutting it, mitering it, and brazing it.  I have a pretty basic Japanese made vise that has 4" wide jaws.  I haven't found the need for anything bigger yet when building racks, but this one is probably too small for mitering larger tubing used in frames.  I often clamp rack sized tubing directly into the jaws of the vise, but it is better to make tubing blocks.  You can see a homemade behind the vise.  I have my vise mounted in front of the workbench so that I can access it from three sides.  Framebuilders often have a vise mounted on a pedestal for 360 degree access.

 

My torch kit started out as a Victor Superrange II, but I quickly replaced a number of the parts.  It would have been cheaper to buy a kit from scratch.  I use Oxy/Propane (Oxy/Acetylene is more common).  My welding shop swapped the stock acetylene regulator for a propane one.  Propane doesn't burn as hot as acetylene, but it is hot enough for brazing.  It's nice being able to use the same fuel tank as my propane BBQ.  My oxygen cylinder has a 55 cubic foot capacity and that seems to last me about 6 months (maybe 10 racks?). 

I've upgraded the hoses to the Smith Kevlar hoses (around $50) and replaced the torch with a smaller and lighter Victor J-28.  I primarily use a W-1J tip when making racks.

A great early project is a torch stand.  This one is made from pieces of a BMX frame that I found in the trash.  The stand lets me keep the torch turned on when I need to put it down for a minute and also holds short pieces of brazing rod and my striker.

A lot of small tools are involved in making racks.  I laid out some of them on my workbench.  From left to right we have:

  • Drill and bits
  • Tubing bender
  • Clamps (Kant Klamps are the specific brand)
  • Pliers
  • Shop cloth or production cloth.  This is sandpaper on a roll and 1" or 1.5" wide.  I buy 80 grit shop cloth and a roll seems to last a long time.  Enco has it pretty cheaply.
  • Sharpies are really useful for marking on steel. 
  • Machinists squares in a couple of sizes are useful for keeping things square.  Enco has a small kit for $20 with 4 sizes.
  • Vise grips are useful for clamping fixtures.
  • Round files are useful for mitering.  More on this in a future entry.
  • A welding magnet can also be useful for clamping.
  • A center punch makes drilling holes in tubing much easier.  You need to drill small vent holes in most of the tubes that you braze.
  • Sitting under the center punch is a deburring tool.  It cleans up the inside of tubing nicely after you cut it.
  • Brazing flux and a flux brush.

You need to wear some sort of eye protection when brazing.  These safety glasses are special in that they have didymium lenses with a flip-up #3 brazing lens.  They are large enough to fit over my regular glasses (yup, I look like a dork with three pairs of lenses on my eyes).  You can get them from Sundance Art Glass.  Without didymium the flame and flux produce a bright orange sodium flare that is very difficult to see through.  It makes it hard to see the underlying metal, which is necessary to know when the flux is getting hot enough (it turns glassy) or the steel is getting too hot (it turns orange).  Here is an example without the glasses:

Can you see anything under that huge orange flame?  I can't.  This is what happens when you put the glasses on:

Whoa.  That tube is too hot.

The glasses are expensive and a luxury, but they are a really useful luxury.  I think that they help my brazing quite a bit.

Posted 02 October 2007 02:51 AM by AlexWetmore | 11 comment(s)
Filed under: ,
More Posts