Waltly Titanium Disc-Brake Road (Race) Frame

A couple years after I built up my FM166 (described in posts one, two, three) carbon disc road bike, I decided to change things up once again and replace the frame with a titanium frame.  Yes, a frame that would weigh more, cost more, and probably be no more comfortable to ride.  I have really enjoyed this foray into carbon, but there is definitely something I really like about riding metal bikes.  I do love my [titanium] Habanero commuter "cx/touring" bike to death.  (Literally: it has died twice; I'm riding my 3rd frame, all wonderfully replaced under warranty.)  That is a fantastic multi-surface bike, but I also love my road (race) bike, which happens to be carbon.  Sometimes it's nice to ride something purpose-built for going fast on pavement.

While I love the handling of my road bike, something about the carbon frames has never fully made peace with me.  I don't know if it's the hollow plastic sound when road debris hits the bike or just a personal aesthetic preference for the simplicity of welded tubes.  I take a lot of pleasure with the tangible, mechanical nature of the bicycle as it contrasts with sitting in front of a computer all day, so I suspect this is more about the constructed aesthetics than anything else.  I guess I'm a little bit hipster like that.

So, I set out to build (have built) a titanium frame that would synthesize modern frame features with a more classic frame aesthetic, and with road-race geometry.  Disc brakes, new dropout standards, round straight tubing, 73.5/73.5 angles, short-ish chainstays, etc.

Choosing a builder

I decided early on that I'd have this built by one of the mainland Chinese titanium builders.  I can't bring myself to spend more than $1k on a frame, so that was really my only option if I wanted a custom titanium frame.  There are a few players in this arena that will deal direct to customers, most notable are XACD, Titan Product, and Waltly Titanium. 

The key to helping me decide which builder to use (and really the key to getting into this project at all) was Andrew's www.spanner.org.uk blog.  I traded a number of emails with Andrew both while choosing factory/fabricator and then for comments on designs and to share final result.  (Thanks, Andrew!)  Update: Andrew just published his own account of this build.

After getting initial quotes from XACD, Titan, and Waltly, I ultimately decided to go with Waltly.  Waltly was not the cheapest option (that would have been Titan), but they left me with the most confidence in their ability to produce the frame I wanted.  Of the three only XACD was willing to work with butted tubing (for additional cost), and while that was originally something on my list I decided that it wasn't a requirement to save a couple hundred grams (and my Lemond Victoire that had butted tubing did have a few dents in the tubes, so straight tubing seems like a more robust answer).  I would consider using XACD in the future, though their sales rep, Porter, is quote notorious and they way they charge extra for every frame feature would have made this frame a fair bit more expensive.

What to build?

At a high level I knew I wanted a titanium road disc frame with a race-oriented geometry.  I provided a basic set of features I was looking for to get the original quotes (since some builders charge extra for some of these features).  But I was also open to changing specs depending on what the builders could offer for dropout design, cable routing, etc.

Frame Features / Design

My basic requirements:

• Disc brakes (of course!)
• ... but newer flat-mount disc brake road standard
• 142x12 rear dropouts
• 44mm head tube for tapered fork
• 27.2mm seatpost
• Threaded (BSA) bottom bracket

And my high-level design goals were pretty basic:

• round tubes
• straight tubes, clean lines (so some internal cable routing, probably)

Geometry

One upside to changing my road bike frame every year or couple years is that I've had a lot of experience riding different frames.  And I've had a bike fitting which basically confirmed that experience.  For sake of simplicity, I knew that my current carbon frame is configured with a good stack/reach setup for me.  I then used the excellent Stack and Reach Calculator spreadsheet to plug in my current frame and stem/spacers setup and then find measurements that would work well for the new frame (a slightly steeper seat tube angle fits me better, so that needed to be factored in too).

CALCULATE BASED ON GEOMETRY
GEOMETRY	FM166-58	Ti	DIFFERENCE
TOP TUBE LENGTH (EFF)	581	580	1.00
HEAD TUBE LENGTH	175	175	0.00
HEAD TUBE ANGLE	73.5	73.5	0.00
SEAT TUBE ANGLE	73	73.5	-0.50
BB DROP	68	68	0.00
FORK LENGTH	369	369	0.00
FORK RAKE	43	43	0.00
STEM LENGTH	110	110	0.00
STEM ANGLE	-6	-6	0.00
SPACERS + HEADSET + STEM STACK*0.5	39	43.26	-4.26
STACK	576.36	576.36	0.00
REACH	404.79	409.27	-4.49
STACK WITH SPACERS	613.76	617.84	-4.08
REACH WITH SPACERS	393.71	396.99	-3.28
STACK WITH SPACERS + STEM	633.80	637.89	-4.08
REACH WITH SPACERS + STEM	501.87	505.14	-3.28

After providing the various frame measurements, I also indicated that I'd be using a Whisky No. 9 fork with a 367mm axle-to-crown length. I wanted to call out that I'd need the head tube angle/length to compensate for the fact that I'd have an extra 14mm lower stack height on account of the external (Hope) headset lower bearing.

Lower bearing stack height matters.

Adjusting so that the effective angle would be correct despite additional stack height under the head tube wasn't something that I had fully appreciated when starting this project.  But as you can quickly tell using the bikegeo.muha.cc calculator, adding 15mm to the effective axle-to-crown measurement of a fork makes a significant difference in the effective head tube angle.

Other geometry dimensions (BB drop, chainstay length, etc.) were based on my current frame.

The Process

My sales rep at Waltly was Amy Lv.  She was very courteous and communicated very well and very promptly (obviously there was a timezone difference, so typically email exchanges would take 24 hours).  There wasn't a language barrier per se, but I did learn that it was most efficient to be as clear as possible (and use standard/non-idiomatic vocabulary and phrasing) and include example photos or drawings to save back-and-forth emails.

But there were still a lot of emails, even when there was no misunderstanding.  Around a hundred, last I counted.   My own indecision was in no small part to blame there.  Also, an important thing that Andrew (spanner.org.uk) relayed early on was that these builders are more fabricators than designers.  They will happily make small design decisions to fill in unspecified gaps, but typically they will implement whatever they are told to implement -- even if it is a bad design.  This is probably the biggest risk with building a Chinese custom ti frame.  So I spent a ridiculous amount of time looking at other frames for ideas and reading the endless debates over the structural value of chain stay and seat stay bridges, etc.  This all made me decide that someday I'd like to build my own frame -- probably with steel.

So first step was to work with Amy to specify as exactly as I could what I was looking for.  I ended up also sharing with her a requirements/specification document to help me keep that growing list of specs organized.  It was also a useful way to embed photos/diagrams and seemed to work pretty well as a means of enhancing email communications.  Once we had specified everything, it went off to design department.  Within a couple  of days I had the first design back.

Version 1

So that looked like a good start.  There were a couple of things that had gotten lost from requirements, though, and a couple of things that I had requested but were impossible with my design.  And I was talked out of such a small gauge down tube by Andrew -- and then this was reiterated by Amy. (At first I suggested that the 31.8mm downtube on my Habanero was just fine, but this argument was blunted a bit by the fact that my Habanero down tube cracked after 25k miles...)   So changes to this design included:

1. Straight 44mm head tube (not a tapered head tube).
2. Larger 40mm down tube.
3. Straighten those chain stays! -- which meant lengthening to 415mm and decreasing tire clearance.

This came back quite promptly in the second design iteration:

Version 2

This looked much better, though I started worrying about the internal routing -- and specifically the gaping hole to accommodate 3 cables (housing) exiting near the BB.  This didn't seem wise and I wasn't entirely sure that they had run any sort of structural integrity calculations on this.  So after many different design iterations routing internally through various tubes and moving around caliper mounting etc. (mostly just iterating to myself, luckily), I settled on the obvious option to run the brake housing/hose externally using zip-tie mounts and internal routing for just the derailleur cables.  In the back of my mind was the idea that if I drop my FD at some point I'll just route the brake internally making it a bit cleaner.

So after that was worked out I had the final version:

Version 3 (Final Verison)

I also requested a sandblasted logo (on the otherwise brushed ti frame).  I didn't actually have a design, but in an effort to have my frames be more forthcoming about their origins, I asked them to "paint" the Waltly logo on the head tube and the company name in Chinese characters (that you can see in the design document) on the front of seat tube.  They were a bit suspicious as to why I wanted their name on the frame, but I explained that to me it just made sense since they were the ones building it.  Of course, later I saw that for bike shows they have a standard branding they use on their frames; I would have just requested that had I known it existed.  Adding the sandblasted logos added $30.

A rough attempt to communicate desired logo placement (on a generic frame)

Once all that sas settled, I approved the design and paid the 50% deposit (so a bit over $500) and it entered the production pipeline from which it was estimated it would emerge after 35-45 days.

Result

After just a bit longer than estimated time (frame was finished but logo sandblasting added a bit of time), I received a bunch of photos of the finished product, all staged against a beautiful shrubbery. 

Obviously one can only tell so much from photos (e.g. was it built to spec? any issues with head tube or BB threads?), but it looked good to me so I paid the remaining 50% of balance and received the frame 5-7 business days later (shipped via EMS).

Despite knowing this would not be a weight-weenie build, I still of course had to weight it immediately upon receipt. Our kitchen scale has weighed many more bike components than food products.

Well, that was heavier than I had expected -- I think the original estimate (but before we had really nailed down lengths and dimensions) was 1300-1500g.  So for a big frame with relatively large, unbutted tubes and flat mount, 44mm HT, etc. I guess this is probably just how it is.  This is ~750g heavier than my carbon frame.  Which is about the weight of one full bottle.  So now I know at least how much weight I need to lose before I can complain that the bike is slowing down my hill climbs.

I had accumulated all the little bits I would need to move parts over from my current road frame (e.g. headset, FD clamp, flat-mount to post-mount adapters, new cables/housing).  So it was a pretty quick job to swap it over.

Most importantly, though, the frame built up perfectly -- no issues with ovalized head tube (had read horror stories, but not with Waltly), no issues with BB shell, rear dropout, or caliper alignment. No seatpost slipping.  And the final bike is exactly what I wanted.  I did not try to take my own measurements of the tubes or angles, but everything seems to be correct (and fit was exactly right, compared to other bike).  Waltly did an excellent job building out exactly what the final design indicated.

One of the first rides.

(I was actually riding on the road adjacent to this trail.)

And a few detail closeups:

142x12 thru-axle dropouts

Internal cables exiting (and un-crossing) near BB.

Internal cables entering near head tube and externally routed brake cable.

The logo effect is subtle (and hard to capture in photo), but I think looks great.

Ridden

So having now ridden it for a few hundred miles, I can say that I'm definitely enjoying it.  At first I noticed that it was heavier than the carbon bike, but now that I have nothing to compare it to, I don't notice that anymore.  It's plenty stiff in the bottom bracket.  I did not notice any flexibility, though I'm sure that if I jumped on a stiff carbon bike I could tell the difference.  I love the way the metal frames ride over the road, though; it feels very solid, which is great.  The rear thru axle is really nice.  Not sure I can tell a difference in stiffness, but being able to mount the wheel while in the stand (without needing to recenter it in dropouts while on the ground) is a nice perk.  No creaks anywhere is nice too.  I hope this frame will last for a long time.  Andrew had (has) some concerns about the internal routing (adding holes to highly-stressed areas of a frame); I guess time will tell.  This is an area (one of many) where I wish I understood the physics at work here better.

Lessons Learned

I don't have any regrets with this frame.  Sure, a titanium bike is heavier than carbon, but to me it feels like so much more of a bicycle.  And Waltly did a first-rate job building to match the drawings.

There was, however, something I overlooked in the final drawing; I had intended to have the rear derailleur cable stops under the chain stay be simply zip-tie stops instead of traditional cable stops -- primarily because I wanted to run full-length housing.  This was always "wrong" in the drawings and I only noticed it after that part of the frame was complete, so it's no one's fault but mine.  That said, I think I prefer the traditional stops in the end anyway -- and this isn't my rain bike, so having the full-length housing is a bit unnecessary.  I was thinking that zip-tie stops would look better when someday I move to eTap and get rid of the cable .... but that's probably a long, long time away (and I can always dremel off the cable stops if it bugs me).  So the lesson there is just to check every single detail of the design multiple times.  It boggles me that I missed that on 3 designs, but then it's a pretty subtle drawing difference between the zip-tie stop and the traditional full cable stop.

One "lesson" that did come out of this is that I should have been clearer initially with regard to chainstay spacing and not talk about the more indirect measurement of tire clearance. I originally requested clearance for 30mm tires (with intent of having option of running a 28mm tire on a wide rim).  The v1 drawings indicated 46mm of spacing which seemed to be way more clearance than I needed (especially with no fender mounts); I was expecting maybe 5mm of diameter wiggle room (above requested tire clearance size) since there are no fender mounts on this frame.  When I backed that down and just asked for 26mm clearance, I still have a frame with 42mm spacing, so I suspect I could comfortably fit a 32mm tire in there if I were so inclined.  While right now the idea of even 28s seems a little redundant with my commuter/cx/gravel/adventure bike, it's nice to have the option.  So clearly there was a very large (16mm) markup when translating "tire clearance" into "chainstay spacing".

Next Steps

Next up, the groupset is getting an upgrade to hydro disc -- and flat-mount calipers.  I'll probably try a new Hongfu flat-mount 12x100 fork I have.  And I'm currently planning to switch to a 1x system (50t ring with 11-40t cassette), though I keep wondering if that is really the right move for this more purpose-built bike.

If (when) my Habanero adventure bike breaks next, I will strongly consider getting another Waltly frame to replace it.  I have loved my Habanero frames, but I really want the versatility to use a tapered fork (the vast majority of disc-brake cx forks are tapered) and if the next break is also at the down tube then I think I'll have enough data points to suggest that a larger down tube would suit my riding style better.  I might also design that frame around a different fork spec (e.g. a suspension-corrected 26" MTB axle-to-crown length) so that I could have flexibility to run 2.1" or larger 27.5" tires.  Hmmm ... sounds like I have already started my next project :-)

Update

A little while after writing this, I switched over to a new flat-mount fork from Hongfu:

HongFu FK-079-F 12mm thru-axle road fork.

And, more significantly, Force1/Force22 hdyro brakes/shifter.  The new fork is really nice.

I have been enjoying this new bike immensely.  It still feels heavy when I pick it up, but I enjoy the ride.  It is plenty stiff but plenty comfortable.

I did decide to replace my commuter/gravel/cx frame!  The requirements took a bit of a twist, but I'll write up a blog post on that shortly.  (I'm working on the build now!)

24" MTB Build: "Kona Queso"

It all started with a free bike, a 24" Kona Hula (12" frame) that my friend's son had outgrown.  The frame is really on the large side for my son, who is only a year into his Isla Bein 20" Small.  While the bike as it was was a perfectly ridable bike, I decided that we would turn this into a bike-building project.  My son loved this idea.  And we set right to collecting parts for this too-large bicycle, because to do anything else would have required patience and deliberation.

The project goals were pretty simple:

1. To have fun personalizing and customizing the bike.
2. To work on it together (hopefully he'd learn a bit about the mechanics of it)
3. And to drop 10lbs from the finished product (original weighed in at 30lbs; goal was 20lbs). 10lbs is pretty significant to a 50lb rider.

The constraint was to try to stay within budget of what his next (24" bike) would have been.  Luckily my wife and I have separate allowance funds; this way the discretionary expenses (the majority of the costs here) wouldn't cause marital discord :)

The plan for the bike was:

• Convert to disc brakes.  This was bold, but research suggested there were some options that didn't involve welding mounting tabs to the [aluminum] frame.

• Convert to rigid fork.  The affordable kids suspension forks don't actually compress for little 50lb children.  The high-end forks would exceed the budget (and how!).

• Convert to 1x10 drivetrain for simplicity (and weight savings) but with enough range to ride any terrain around here.

The Frame

Here's the original frame.  It's a Kona Hula, a 12" frame size according to the sticker.  Made of 7005 aluminum alloy.  Thankfully, very standard specs: 27.2mm seatpost, 1 1/8" headtube, 68mm BSA (threaded) bottom bracket. 

"We" (I) decided early on that we'd convert this to disc brakes, so we ensured we couldn't change our mind by sawing off the brake bosses.

I removed the material with a dremel.  Luckily my friend Kurt has a pro-grade workshop and helped grind this down to finish this like a pro:

I drilled out the cable stops so we could run full-length housing for the brakes and derailleur.

I let my son choose the color.  We used Nova Powder Coat which uses (among others) Prismatic Powders; my son chose Hot Yellow.  Good choice, kid.  (I encouraged a high-viz color choice.)

Part of the fun was getting to bike down to Chantilly to pick it up.

The finished product really looked fantastic:

We got some Kona stickers off eBay.  They're not the same as the original ones, but they look good.

The Brakes

Before starting this project, I had no idea that people actually converted rim-brake frames to disc-brakes.  (Often the preference is to run a "mullet" with disc front and rim-brake rear.)  I decided to investigate, though, since I had a set of cable brakes from my commuter (before switching to hydro) that I wanted to use. I learned that there are several ways to convert a rim-brake bike to disc-brakes; the one I opted for was the A2Z DM-UNI adapter.  Found one for $20 from the UK, which was a lot cheaper than the domestic options (mostly ebay).

Before we sawed off the brake bosses, needed to make sure that this idea would actually work.  After spending some time with a file flattening out the dropout lip on the inside of the dropout, the adapter sits flat against the dropout and fits quite well. However, you can see that there's some space at the bottom (between the lower left bolt and the frame):

While this wouldn't be a problem for regular braking from forward motion, it would be a problem if applying the brake while rolling backwards.  That didn't seem like a great state of affairs.

Without any fancy tools, my solution was to hunt down some big aluminum washers in just the right diameter that would eat up the extra space.  This actually looked like it would work fine, if a little ugly:

Kurt rescued me again here with a much better idea, involving drilling through the frame and tapping out the thick aluminum backplate to the adapter for an M6 bolt.  This is a much swankier solution.  You can see the extra M6 bolt right above the QR skewer:

(A little preview of the finished product there.)  You can see in the photo above that I'm using TRP Spyre disc brake calipers.  This is worth mentioning, because these are road brakes, not mountain bike.  This actually is perfect for a build for a little person, though, since BMX brake levers are (typically) short-throw, just like road levers.

The Fork

I spent awhile investigating the options for forks.  There are a few suspension forks that are decent for this size wheel, but they're pricey.  The off-the-shelf suspension forks don't compress under the weight of tiny people, so they're really just good for weighing down the bike.  Many people also use 26" forks.  I measured the axle-to-crown distance on the suspension fork that was on the bike; it was around 425mm.  After scouring the various carbon fiber forks on ebay, I found a 26" fork with a 415mm axle-to-crown measurement, so while this might be a little taller than perfect for the 24" wheelsize, it was going to be 1cm lower than the RST fork it was replacing.

Critically, this fork was available in tapered or straight steerer.  We chose the green logo, which locked in our color scheme.

I am really impressed with this fork, actually.  The finish is top-notch and the weight is pretty incredible at 421g.

I would consider using this on a CX build (only slightly taller than the standard 400mm axle-to-crown measurement of a CX fork).  If only it came with thru-axle.

The Drivetrain

After doing a bit of research, I settled on a Shimano Deore SL-M591 (10-speed).  My son has SRAM (X4) grip shift on his current bike, which works fine, but folks on the interweb suggested that Shimano trigger shifters were easy for little hands -- and I found someone selling a take-off shifter and derailleur combo for $60.  Done!

I perplexed for a long time over the crankset.  I wanted one with short arms.  Ideally something like 145mm, which is what he has on his Isla 20" bike now.  Most cranksets with arms that short are BMX cranksets.  While a BMX crankset would work, it is hard to find BMX cranksets with 4-bolt 104mm BCD, since typically BMX chainrings are much bigger than the 30-32t I was hoping to fit.  Also BMX chainline is different from MTB; BMX bikes have narrower rear spacing.

In the end, I decided to splurge and buy the Trailcraft 1x crankset which includes a 30t ring.  The arms are a little longer, at 152mm, but this seemed fine.  Of course, we went with a green chainring to keep our color theme going!

Now this is a crankset that needs a square-taper bottom bracket.  The standard Shimano square taper bottom brackets are really heavy (300+grams), so I was really fortunate to find someone selling a used Race Face Taperlock BB on eBay for $25.  This has a titanium spindle and weighs in at a paltry 165g.

For the cassette, I really went overboard and picked up a (used) XTR 11-36t cassette.  While I would never buy myself a titanium wear-and-tear part like that, I figured that as lightly as he rides it'll probably last through both kids and while [the used one] cost me twice the price as a [new] standard cassette, the weight savings are huge (230g for the XTR vs ~400g for a standard SRAM PG1030).

The Wheels

I should really have led out with a discussion of the wheels, because this was really the keystone of the build: or at least the reason we were converting to disc-brakes.  I really wanted to build a wheelset using Stans Crest 24" rims and the leftover 28-hole Novatec D712SB / D771 hubset I had (the rear couldn't be converted to 11sp). 

We used Laser spokes and alloy nipples.  This is a very light wheelset.

This was one of the more fun parts of the build for me too -- and I think for my son; he did the bulk of the wheel lacing (with help & instruction).

We ordered Schwalbe Rocket Ron tires from Germany (it was cheaper than domestic prices, even paying 20EUR for shipping).  There aren't many tires options for 24" wheels; the Rocket Rons are generally agreed to be the best.  Mounting them tubeless was super easy, thanks to the excellent Stans rims.

Cockpit & Other Bits

Not too much to say here.  I traded a Ritchey WCS alloy road bar from my parts bin for a Easton EA90 SL carbon flat bar.  Another example of something I'd never buy myself.  My son helped cut them down a little narrower.

Found some lower profile ODI grips with green endcaps.  Probably could have gotten something even narrower for his hands, but these seem to fit fine.  Picked up some green BMX levers. 

For the seatpost, I was gonna use my old carbon Bontrager XXX seatpost, but it doesn't slam down far enough for him right now (yeah, the bike is a bit large), so we're using an Easton EA70 seatpost I had in the bin.

For saddle, I floundered around a bit.  Originally I found a great deal on a Bontrager Kovee Elite (ti rails), but this proved to be way too long a saddle to work.  Eventually I found the Selle SMP Junior Lite saddle for 50% off in the UK and ordered it. A bit pricey, but I think it'll be moving between bikes (and between kids) for many years.  It's probably more tailored to road riding, but my son says he likes it (we haven't been on any long rides yet, to be fair).

Finished Product

My son named his bike "Kona Queso", since the yellow/green/red color scheme was like District Taco and apparently my son can't get enough of DT queso :)

The weight came out better than expected at 17lbs 12oz (8.05kg).

Osiris (co-owner of District Taco) hooked us up with a DT sticker, which we cut out and placed on the toptube:

We haven't done any long rides yet, but it's been ridden a whole bunch since we finished.  Including some light mountain biking (with little brother who has just started riding the little Isla CNOC 14).