A few members on BikeGuide have started discussing how to improve axial coasters. The idea is to come up with a design that can be easily made and will at least improve the issues faced with axial freecoasters. Here are some requirements (see the thread for more details):
Low cost (i.e. using as many existing components as possible)
Low(er) weight; sub 20 oz
Suitable for peg use, not necessarily abuse.
Better bearing system, especially in the driver
Easy, and safe maintenance (i.e. no little pieces to damage)
If you have suggestions, or would just like to participate in the process, please post your comments in the BG thread, or here and I will forward to the thread. Once we come up with something suitable, I will post the dimensions, CAD files, and any other info on BMX-U.
I was pleasantly surprised to find this in my Google Reader list this morning. If you don't know about Reader, it is a web based application for aggregating RSS feeds, updated search results, and other periodicals. The recommendation and searching functions are pretty good, but some of the BMX RSS feeds are surrounded by noise (i.e. one good feed with a bunch of dead ones).
Director Forks have interested me since they were first unveiled. Integrating the dropouts with the fork legs is a logical step that was taken by road and some mountain quite a while ago. Eliminating the weld at the dropout avoids cracks there as well. And using a simple insert gives good compression support, while maintaining a hollow dropout for weight. Overall a nice design. Unfortunately, since this is included in the patent, any other company who wishes to go the integrated dropout route will have to find a manufacturing method that does not violate George's patent.
The crown junction is another unique feature of the directors (unless you look at sunday's chainstay wishbone that G had a hand in as well). The claim is that it reduces localized stress and extra material found in conventional open ended steerers.
I'll dig into this deeper after some discussion, but I really have to head off to class, so this simple post is all that is going up for now :)
Some random feeds I subscribe to (post yours up!):
- Bicycle Design is not updated all that often, but can be interesting.
- Sweeter Skins for those of you with Android phones, this is really the future in UI.
- Roues Artisanales mostly posts in French about the road bike weight weenie scene.
- Cool Tools updates often with some neat little productivity tools.
- Scheier on Security is one of my favorite blogs. Written by a security expert who is often contacted to comment on security issues by media, government, and private agencies. He takes a very logical approach to dissecting issues. His post "Fear and Over-reaction" really struck a chord in me.
- Google Patent "Bicycle" Search feed updates every so often with new results (usually patent applications or updates). You can use Google patents to get customized results.
- Endless Innovation has some good posts, but can be a bit repetitive at times.
- Physics World has some neat, informative updates about research findings in classical, relativistic, and quantum physics.
- The Big Picture has great photography, and is updated very regularly. Each post contains a link to 40 high-def photos about a certain topic.
- F-Secure has regular, detailed computer security reports.
- Cycling Tech hasn't been updated in 2 months, but it has some cool stuff when they do.
- Industrial Design has a lot of design *cough* Crap *cough*, but you can find some real gems in there. Be weary of hub-less bicycles, there is a reason they haven't gone to production.
- Ted Talks are a great way to look at a situation from a particular (speaker's) perspective. They are all free videos on a very wide range of topics.
Over the past 5 years, KHE’s freecoaster system has taken over the niche market. This was not because their product is the best we can do, but it is better than other options that were available (see my post on radial vs. axial freecoaster systems). In this post we will look at how the KHE system has evolved in its lifespan.
In the beginning was the Geisha. It offered a stud system (EAS) that allowed easy changing between 3/8” and 14mm in three different materials. That’s right, when it first came out, there were aluminum, steel, and titanium studs. They even went so far as to have a hybrid aluminum/steel clutch in their flatland version. The major advantage of the KHE was it’s ball-spring resistor, and low weight (Poverty coasters had the same resistor with a roller thrust bearing behind the clutch. But the resistor was eventually welded into place and the hub was pretty heavy in comparison)
Quickly after their initial run, the aluminum and ti studs, as well as the hybrid clutch were discontinued. To most people, this didn’t matter as the studs would unscrew with the axle regardless of material. Many would have preferred to use a bolt-on system (female axle), but KHE used a 3/8-26 thread in the aluminum spindle, which is very difficult to find bolts for. Blue6 now sells a bolt-on conversion kit for the KHE EAS system with titanium bolts and aluminum collars.
Now I am not sure if this next change came before the 1-piece axle or after, but it was an improvement none the less. KHE replaced the brass ball retainers with steel. Again, most people were unaffected, but to KHE’s defense, it was an attempt to fix the issue of customers tapping their axles out the wrong way (to remove any KHE axle, tap the drive side). Unfortunately, some people, when they encounter resistance, just hit the axle harder. Thus, customers still damage the retainers, and must purchase a new axle to replace them. I was always a bit annoyed that they didn’t make the retainers themselves available for the customers who damage them.
The next change marks the inflection point; when all of a sudden KHE took over the market, expanding horizontally to prevent competition. They finally came out with a 1-piece steel axle just before releasing the Federal and Reverse freecoasters. Both hubs were the same, but with different hubshells, and marketed under different brands. They both featured steel-one piece axle, but at a 4 ozweight penalty.
At this point, the Odyssey Reloader was out of the market, and the Poverty was rumored to be on it’s last run. KHE looked to fill the niche. They introduced LHD versions of both the Federal and Reverse. In order to do so, they would have to replace the threaded engagement ring with a splined one (if you run a threaded hubshell with LHD driver and clutch, the engagement ring will unscrew).There have been some issues with the splined ring coming loose, either by initial fit or compression of the aluminum splines. If you have this issue, retaining compound or JB Weld on the splines will fix it.
Some customers had issues blowing out the drive-side hubshell bearing. KHE attempted to fix this by switching the 6905 deep groove bearing with a 7905 angular contact bearing. Although not yet confirmed, it is believed the 7905 they are using is a 15 degree contact angle. Even this modest change can make a difference to someone who persistently blows out their drive side bearing. Unfortunately, the bearing seat is too thin to allow it to be bored out to accept an angular contact roller bearing, which would eliminate the issue all together (see Rob Ridge’s modified Reloader).
The drive side hubshell bearing was small potatoes compared to the driver bearing issues inherent in axial coaster systems (especially those with smaller drivers, unsealed bearings, and 14mm axles). KHE’s first attempt to resolve this issue was in the switch to steel in their 1-piece axle. The hope was that steel would better resist deformation of the measly 0.5mm bearing shoulder on the axle. It didn’t work. Their next attempt was to install a thrust bushing behind the 6902 bearings in the driver. I do not know how this has affected the number of driver bearing issues, but can’t help see it as an attempt to again fix the symptom (blown bearings) rather than the problem (axial wheel loads on the driver bearings).
Their most recent attempt to fix the driver issue was to have a custom bearing made by reducing the inner diameter from 15mm to 14mm. This gives a 1.5mm shoulder as opposed to a 0.5mm one. Although I like that they are addressing the issue, custom bearings always create headaches (such as buying a new axle that won’t work with your existing bearings). And again, it is not a proper solution to the underlying problem.
And finally, the one I skipped: somewhere along the line, KHE made a modification to the 1-piece axle to allow external slack adjustment. This was the first coaster to do so. It uses a simple screw attached to the internal stop-washer. A nice improvement indeed, but I have felt no inclination to change my slack, once set to my preference, in any of the coasters I have ridden.
So what can you take out of all of this?
KHE has made incremental improvements in their design over the years. This extends the lifespan of the product line.
KHE has committed to their system by expanding horizontally, now having 5 production models as nearly 50% of coasters currently available (Astern, Geisha lt, Reverse, Federal, Alienation).
Pushing KHE to make specific improvements may actually manifest them. Though look for incremental change, as they will likely milk this system for as long as they can before releasing something new (I hope they are working on a radial system, since they are still the only company to have released one, even if it was a failure).
Here is a very in-depth look at the history of the lathe. He does a good job of explaining what advancements were made, and how those affected the materials used and the quality of finished turnings.
Gun Drills are used for deep bores. And are usually equipped with channels for feeding coolant from the base to the end of the drill. Using a normal drill would not work because the flutes are not long enough to evacuate chips and flood the cutting surface with coolant. Drills are available in straight or spiral flutes (spiral drills are made by taking a two flute straight drill, heating, and twisting).
Here is a link that talks about making air gun barrels. Kind of a good description.
300+ pages, but with short naration, packed full of side stories, short biographies, and historical perspective. It is a great, easy read (i.e. to be left out for guests or customers) that explained my love affair with bicycles better than I could. In a word, my bicycle represents Freedom. There is nothing I have felt better than the silence of a bicycle focused mind, hard-worked muscles, and a smile that just wont fade. This book kept me in the right state of mind.
In recent years we have seen a shift in sprocket/crank interface paradigm... Ok, that may be a bit exaggerated... At least there has been a change in design.
Traditional, Perpendicular Bolt
My first introduction to crank torque transfer was with the integrated stud on one piece cranks. Although effective, tolorances were often poor, and low-end sprockets were usually thin steel, further compounding any issues.
As BMX progressed, we moved to a bolt to secure the sprocket to the crankset (usually 3/8" or 10mm). Today, virtually all cranks use the bolt-on design. This helped solve some of the tolerance issues and has served as the standard for bmx. From a structural perspective, this design requires either a larger spindle boss or a separate hole cut in the crank arm for a sprocket boss.
A slight hybrid of one and three piece mounting systems, that has resurfaced recently is the use of a headless bolt. Mechanik makes a bolt called the Pin Head. The purpose of this design is not to hold the sprocket to the crank, but only to transfer torque. Using this design allows small spacers to be inserted between the arm and the sprocket to dial in the chainline.
Nano-drive
As Sprockets reduced in size (with flatland at the forefront), a need for less offset in the mounting position. Enter the nano-drive. Most nano-drive bosses use smaller bolts than the standard bosses.
Spline Drive
When faced with the challenge of small sprockets, Tree Bicycle Co. decided to redesign the mounting design. What they came up with was transfering torque through the spindle, rather than the crank arm. With most cranks now using 19mm, 48 spline, Tree's spline drive has a large potential market. Recently Profile decided to release their own version of spline drive (releasing a spline drive sprocket was logical for Profile as they are known for their 48 spline cranks). Unfortunately, Profile decided it was best to use 8 large splines on the adapter, rather than the 9 available on the Tree. I don't know why this happened, but it may be just to avoid any patent issues.
There is one big problem with spline drive: it is limited to 19mm, 48 spline cranks.
Socket Drive/ Nonogram
Flatware's (Odyssey) response to shrinking sprockets was to standardize. Well, sort of. They are the only company making a crank that works with the 9-sided mating surface, but do provide the specs and drawings(PDF) on their site. If this design is adopted, adapters can be made to adapt socket drive to a 48-spline crank, in effect making it another spline drive. The big advantage to this over standard spline drive is it does not limit the spindle design.