I’ve been building custom bicycles for a bit more than a decade now. Hard to imagine 10 years have gone by but here I am! When I first started out, there was so many variables to overcome to obtain a level of proficiency I could then master. I came to the bench with a level of design, fabrication experience and skill however applying those concepts and methodologies to a bicycle frame was a new challenge that required mastering new skills and further honing old ones. One of the variables that seemed like a massive hill to climb was geometry, fit and handling and how those 3 interact between rider, bicycle and terrain. Working towards a full suspension platform felt like putting the cart before the horse. But in 2019, after years applying the “build, ride, repeat” mantra, I felt like I was actually gaining a greater understanding of that trifecta’s interaction and finally felt like I was ready to apply all that I had learned out on the trail and at the bench to a bicycle with moving parts: enter full suspension and the Snakedriver Heavy Project was born.
As I’ve mentioned in previous posts, I had not ridden an FS bike since about 1998. That’s another point that’s hard to imagine. Also in 1998, I had graduated from Penn State with a BA in Jewelry and Light Metals from their College of Arts and Architecture and upon graduation had visited the Rhode Island School of Design to possibly go to graduate school for Industrial Design. A quick conversation with one of the professors answered my questions with a simple pair of questions: Did I have the degree and job I wanted and was looking to move up the pay scale and take on more of a leadership position or did I lack the degree to get the job I desired? The answer was the former. So undergraduate it was. During my senior studio, I was going to build a full suspension bike but I quickly realized I was WAY in over my head. I knew this had to be shelved and I needed to learn a heck of a lot more before I was ready to tackle that project. So after 3 years of intense design study, I had a BFA in Industrial Design and an offer from Reebok as a shoe designer… All of this from a crossroads decision back in 1998 to go back to school when I was fresh out of PSU and landed an internship at Bicycling Magazine by chance while out for a ride on a day off from my temporary bike shop wrench job. A slight sidestep from starting my own bicycle company! Fast forward to 2019: So much had changed by the time I felt confident enough to actually tackle a project of this size and complexity. Motivation was not lacking but where was I to start? There in lay the question.
So the journey began with a single pivot. I wanted to start with a known platform that was fairly simple to fabricate, was low maintenance and didn’t have too many moving parts. I also needed a starting point to understand two basic design parameters:
1. What kind of bicycle do I want to ride?
2. What kind of bicycle do I want to service and support?
Furthermore, I wanted to fabricate everything in-house which meant all of the pivot parts, sub-assemblies and of course tooling. I did this on purpose to keep the cost low and to immerse myself solely in the design and fabrication process so I did not have to manage any additional suppliers. This acted as a set of design guardrails which can be important restrictions. And it was a proud moment when everything was wrapped up and back from powder! I recall that night very clearly:
The result was a 150mm Front / 130mm Rear single pivot test platform which is below:
This bike was pretty straight forward in design and geometry. I wanted something relatively quick and snappy that could serve as a starting point to assess pivot points, front/rear center, details such as cable routing, dropper port location, water bottle placement, hardware, torque settings, bearing choices, pivot hardware design and much more. Important cliff notes below:
· 65° HT Angle
· 75° ST Angle
· 25mm Drop
· 837.5mm Front Center
· 438mm Rear Center
· 500mm Reach
· 631mm Stack
· 150mm Front Travel
· 130mm Rear Travel
· 1275.5mm Wheelbase
The swing arm was asymmetric and was a bit tricky to fabricate which resulted in losing a few millimeters in chainstay length so the rear tire under full compression just buzzed the seat tube. Here’s a close up of that fit up:
And welded. That was actually a lot of fun to weld!
The seat tube was admittedly too slack and contributed to that fabrication shortfall and immediately I realized that I needed to steepen the seat tube angle for a bit better climbing position when seated. But other than that, the bike performed rather well out on the trail. The bike was quick, had pop and was a lot of fun to ride. Front triangle storage space was clearly lacking and I really do not like the aesthetics of a bottle mount under the top tube. But the essence of what I was after was present in how the bike performed overall. The build weight was 30 pounds on the money. Subsequent builds were heavier weighing in at 32lbs which is still reasonable but I really enjoyed that round number of 30 pounds which is a target weight for the production bike. Slowly but surely, after successive rides, tuning and tweaks, a clearer picture started to form in what I wanted out of a bike of this nature. But I wasn’t done here. This was only a starting point to learn, build, ride and repeat. So on to prototype number 2.
Enter prototype number 2: A 160mm front / 140mm rear travel Horst Link big boy bike. Admittedly, I wanted to go the opposite direction of what I started with so I went super long, low and big on this next prototype. A goal of this next phase was to have a complete solid model assembly in Fusion 360 of the entire frame. All of the linkage parts, shock mounts and various bits were to be turned in house and contracted out to a local CNC machinist. Unfortunately, upon delivering part drawings and STEP files to a local machinist, about 1/2 way through machining, they lost interest and went silent. I was on a timeline to have a prototype to ride and test by mid-summmer/early fall and was losing time. Right about this time in 2020, things were starting to evolve in the SLM / 3D printed metal industry. I made the decision to completely scrap everything I had just worked an entire winter designing in Fusion 360 and completely redesign the platforms linkage components, swingarm assemblies and pretty much everything else leveraging 3D printing. So that meant no prototype for 2020. I worked all fall spending long hours again in Fusion 360 and that spring of 2021, I was ready to send files to a 3D printer. Parts arrived within a relatively short turn around and I was now able to start the fabrication process throughout the summer fitting it in between client builds here and there. By the end of July/early August, I had a completed prototype ready to head to powder:
Assembly began in earnest once the frame and swingarm returned from powder:
The geometry cliff notes of prototype number two are below:
· 64° HT Angle
· 77° ST Angle
· 25mm Drop
· 864mm Front Center
· 450mm Rear Center
· 516mm Reach
· 640mm Stack
· 160mm Front Travel
· 140mm Rear Travel
· 1304mm Wheelbase
This bike was indeed big and long and had a stance all of its own:
Physically building the bike was probably the quickest to date out of any bike I’ve built. This has much to do with all of the front end design work I performed as well as the nature of the tube and socket design of the 3D printed parts. I was quite surprised just how quickly it came together and that design time up front paid off. A horizontal shock orientation remained but I added a radius to the down tube to open up more space to be able to run a large bottle in the lower half of the front triangle. I also moved away from aluminum carriers for cable management and used zip tie style (more subtle experimentation). Steepening the seat tube angle solved seated climbing and the bike proved to be quite adept at climbing given its overall length. The bike really loved to go down hill and ate up chunk. The long 450mm chainstays were a bit of a bear to pick up and loft the front end. I did however shorten my cockpit by a good margin to get myself more upright which proved a really good choice moving forward in overall fit/feel of the bike producing a very familiar stance. The testing process was going really smoothly with tuning wrapped up but 10 days in, something I wasn’t expecting brought everything to a halt. The swing arm failed.
After some analysis, speaking with a few fellow builders who are well versed in SLM as well as contacting a few engineers, I believe the issue was two fold: The design created a confluences of forces in a high fatigue prone region of the swing arm yoke and the material (316L) is not up to the task of shelled printed bicycle components that see high stress and fatigue. The SLM process has many variables which include powder alloy type, the number of times the powder is recycled, protocol for handling the powder, part print orientation, resolution of the print, presence of inclusions in the print, and finally strength loss of the parent material due to the printing process. Material choice and post print tensile strength are two very important factors to consider. Generally, post print you’re looking at approximately 70% of tensile strength of the parent material. So for example, if the parent material’s tensile strength is 100,000 psi, post print you’re looking at about 70,000 psi tensile strength post print. So you ideally want a parent material who’s tensile strength is relatively high. So this meant looking for a new printer who provides tensile bar tests for each batch of parts so they have a record of known results and tensile strengths they’re getting post print and starting with a parent material who’s tensile strength is high to compensate for the loss post print. This means moving from 316L (which has a tensile strength of 74,000 psi and a yield strength of 29,700 psi with 60% elongation) to 15-5 which has tensile strength of 200,000 psi / yield of 184,900 psi and elongation of 10% or 17-4 with a tensile strength of 200,000 psi, yield strength of 195,100 psi and an elongation of 8%.
Of note was how quickly the 316L failed at 10 days into testing. With an elongation rate of 60%, it seems odd that the part failed that quickly without actually deformation. I will own any faulty design on my part but I still wonder if there may have also been some inclusions/imperfections in the print.
With all this said, I reached out to New Zealand based RAM3D who is the largest 3D printer of metals in the southern hemisphere. They’ve been a joy to work with, they really know the 3D printing process inside and out and have a thoroughly dialed process along with providing tensile bar test results on request (they do this regardless with each batch of parts).
So with that failure, back to the drawing board to redesign the yoke and make a 3rd prototype.
Parts arrived quickly from RAM3D and I got straight to work. Here’s a shot of all the new parts, the majority of which had internal latticing.
I decided to start fresh and not just replace the section of the swingarm that failed as now it called into question the integrity of all the 316L parts. Plus I was already thinking about revisions, so this gave me a chance to not only start with fresh parts but to revisit aspects of the design to further refine geometry and handling. I realized I went a tad too long in wheelbase with front/rear centers needing to be reeled in a touch. SO the following is what changed between V2 and V3:
· 64° HT Angle
· 77° ST Angle
· 25mm Drop
· 847mm Front Center
· 440mm Rear Center
· 500mm Reach
· 640mm Stack
· 160mm Front Travel
· 140mm Rear Travel
· 1297mm Wheelbase
These changes made a big difference out on the trail between V2 and V3. The bike still lacked some of that pop and snap that I really liked about V1 and by the time I had wrapped up design revisions I already knew I wanted to try a similar set of parameters as V1 in a 140/120mm flex pivot design. Parts count for the Horst platform was an interesting education. The more parts means the more you need to keep up with when maintenance called. We got a lot of rain this past summer and what was great about this experience was I needed to maintain the bike throughout that rainy test period. It became abundantly clear I needed to reduce parts count for maintenance and upkeep, but also for overall costing. So simplicity came to the forefront of design parameters when I began V4’s design process. Here’s a look at the Horst links parts count:
The funny part is there is a pair of bearings missing (I just noted this!) from this photo. That’s a lot for a small brand like my own to manage and for a customer to then maintain. Failure is the best way to learn and improve any design and although this was a tough part of the overall process to get through, I’m really thankful this happened now vs later during ISO testing or even worse to a potential customer. This really forced me to reassess my design process, identify shortcomings, create new goals and make improvements to the broader strokes that help guide the overall process. V3 certainly worked well but one of its short comings along with V2 is it tended to sit into its travel a bit more than I’d like. V1 sat a bit higher in its travel initially so I had more clearance under pedal load of obstacles but I also had more working travel.
While I was waiting parts for V3, I had built up V1 again and by happenstance, this was another great turn of events as it again forced me to reassess what I liked from a previous design, compare that to the current version and then apply those changes and what I had learned to V4. During V3’s extended test period, I was noting under braking load, a lot of flex in the seat stay subassembly of the swing arm. Ironically and once again, as I was assessing whether to update that section of the swing arm from 5/8 OD tubing to 3/4 OD tubing… The seat stay just above the ISO mount failed. More learning process. I thought I was safe without adding any braces under the ISO tab and most of the braking forces would be directed back up into the frame but I was wrong. I’ve actually redesigned this sub-assembly and had parts printed to use .75″ OD tubing for seat stays along with adding a pair of seat stay braces that tie into the link by the dropouts. But again, this is really important to get these failures now rather than later and they appear to be confined to the swing arm. So back on V1 again while I was designing V4. This happened on a friday, and by Sunday, I had actually repaired the swing arm because I wanted to see how the 3D printed parts could be treated in the event of a needed repair where parts were fine, but it was a tube that was damaged. Everything came together quickly and very smoothly. Here’s a look at that C3PO repair:
But that brings us to V4 and what I feel is very close to a final offering.
What I really appreciated about V1 was its pop and snappy feel. Relatively quick handling, light weight and no nonsense feel. Clean, understated simplicity. The V4’s geometry cheat sheet is below:
· 65° HT Angle
· 76° ST Angle
· 25mm Drop
· 813mm Front Center
· 435mm Rear Center
· 479mm Reach (Size Large-ish)
· 620mm Stack (Size Large-ish)
· 140mm Front Travel
· 120mm Rear Travel
· 1248mm Wheelbase
This version is built around a 140mm fork and is designed around a 185x50mm trunnion shock delivering 120mm of rear travel via a flex pivot. The idea here is a customer can choose between a Fox 36 (for example) that uses their 140mm travel kit and a Float X2 for a highly tunable and aggressive trail build or shave weight by going with a Fox 34 and say a Float X for a slightly lighter trail bike.
Additional details are as follows:
· Fits Large Water Bottle
· Tool Mount Under TT
· UDH Compatible
· ISCG05 Equipped
· External Cable Routing
· Internal ST Dropper
All of which is designed around a 30mm stem giving the option for a potential rider to size up on cockpit length with the choice of 30, 35, 40 and 50mm stems. This is pretty close to what I’ll be offering publicly. I missed the pop and snap of V1 and pretty much have nailed how I want the bike to ride and perform which is fast, aggressive trail with pop. Straight forward simplicity with an eye to serviceability and ease of maintenance in a tight package of 140mm F/ 120mm R via a flex pivot. So you like to climb and earn your descents? Jib on features, bunny hops over logs and boost off of stumps? Lean hard and carve turns? Manual and foot plant over rock walls? V4 has it. Here’s a close view of parts count for comparison between V2/V3 and the current version. SO much less to maintain!
One thing to note here is I want to make as much of these parts stock and off the shelf. So whether you’re on the road or at home, you can simply place an order with a supplier like McMaster-Carr or head to your local hardware store to pick up what you need. I worked long and hard on this aspect and I’m pretty stoked that with the exception of the two pivot axles, everything pictured above is readily available online and/or at a local Ace or True Value hardware store.
Bearing count is down from 10 to 8. I moved the bearings from the swing arm yoke to the frame which allowed me to narrow up that area, stiffen the feel of the swing arm laterally and allow room for an ISCG05 mount. Parts count is night and day. I’m considering changing the rockers main pivot width to match the swing arm pivot width of 42mm so I can use the same pivot axle width and cut down of further parts count. I’ve already sent out revisions for a 2 piece rocker design (the current version is a 3 piece design). A 32t chainring JUST fits on the current yoke design. There were some things that I just didn’t “see” in the CAD drawing that became apparent once the parts were a real assembly in hand. So I’ve been working on refining chainring clearance and considering if I max out for 34 or 32t chainrings. 34t chainrings require a bit more thinning of the profile in critical areas of the yoke to gain clearances. 30t and 32t chainrings seem a bit more the average of chainrings for this type of terrain and style of riding I’m designing around but it also allows me to shorten up the yoke a tad, and retain material in a critical area for additional strength and stiffness.
Cable routing is via aluminum clamps again (which seem to retain the cables better on an FS vs the zip tie versions on the down tube. Bottle placement is mid-down tube which actually makes reaching it much easier. The climb switch on the shock is a bit of a stretch to reach, but it has become really obvious that Fox’s line of shocks were designed to be oriented vertically. All of the controls face up towards the rider or are facing the drive side for sit down / in the stand maintenance and tuning. The bike is actually really easy to work on, assembles quickly and upkeep so far has been minimal if anything at all.
The main pivot uses a set of internal snap rings that serve as bearing stops. I’ve turned a bearing spacer so that once torqued, the pre-load on the bearings is centered around the pivot center vs the perimeter of the bearing without one. On a recent ride the M12 pivot bolt loosened up a touch, so I’ve increased torque to 20Nm and went from blue Locktite 242 to red Locktite 271 and all has been holding aok now. And that’s one of the fine details that seems lost in translation: All the small little things that I’ve had to consider in order to deliver a super reliable and easy to maintain mountain bike. What I’m really enjoying now is that I’m seeing light at the end of the tunnel so I can now focus on these smaller details and with a known design direction and suspension platform, I can further refine and hone all the key aspects and details of the bike. And this is the destination I’ve really been looking forward to arriving because I really enjoy the refinement process! Here’s a few shots of the rocker and associated pivots:
As the bike sits, its a tad over 32 lbs. One of the key areas I’m now revising is where and how to apply manufacturing techniques. So CNC vs 3D SLM vs Laser/Water Jet cutting parts that are post machined and welded into assemblies. This has been allowing me to assess each of these manufacturing techniques strengths and weaknesses and then leverage their individual processes and approaches to design uniquely to deliver the most cost effect part as well as the most easily manufactured part with an eye to strength, durability and weight savings. So I’m pretty confident through all of this refinement process I can shave out those 2 pounds to deliver a sub 30 pound complete bike.
The addition of an ISCG05 tab has been a long time coming and it really has made a big difference.
It seems with 11 and 12 speed, I’ve not broken any chains until the FS project got under way. I’ve broken 3 so far and they’ve all been preventable rock strikes and hits. This project has not only challenged my design process and forced me to reassess my own process and methodologies in the shop but its also changed the way I ride. No longer am I seeking the smoothest line but now I’m looking for the fastest and most fun line, which sometimes isn’t necessarily the smoothest line!
One last detail I’d like to touch on is the brake mount. For now, I’ve designed a proprietary brake mount that uses a custom post mount adapter. Basically I’m calling it ISO44 which is 11mm in the Y axis above the axle centerline, 22mm in the X axis to the first mounting bolt and 44mm between mounting center points. Super easy to remember. But this allows for a mount that is directly on the chainstay, is small, relatively low with good clearance and has a small footprint. One big update to this style ISO mount is there is a recessed step that the Post Mount adapter keys into on the backside of the mount on the frame. That means the bolts are no longer in sheer. Whether I formalize this and make it widely available is up for debate. This was something I wanted to try so I could separate the chainstay and seat stay from each other and allow the seat stays to act as the flex portion of the flex pivot without the need for bracing (I could add them if I want of course). I am working on a post mount part that allows me to weld it directly on to the chainstay and not need any proprietary mounting standards (the post mount part would start at 180mm in either case and be integral to the frame if welded vs the machined proprietary mount). Here’s a look
This tucks the caliper between the chainstay and seat stay and getting it out of the way of debris. It also makes the rear end incredibly clean and compact aesthetically (something I really dig). But we’ll see as things take shape and I further refine this phase of the Snakedriver’s journey.
So there you have it. For those interested, here’s a master list of the albums via Flickr of all 4 builds:
As I said before, what I’ve been after in a full suspension bike is clean understated simplicity with an eye to easy to access maintenance and upkeep in a light and tight package that is fun, poppy and has snap. So whether you’re hitting local trails or out for a longer ride or on a road trip, this bike will perform. This isn’t a park bike or meant as a burly enduro rig but rather a fun, aggressive and playful trail bike. Build it up with a bit more burly parts for more aggressive riding or strip it down and focus more on weight savings for a fun and playful trail bike.
More soon as I continue to refine this final stage of the Snakedriver project!
Snakedriver Design and Development
I’ve been building custom bicycles for a bit more than a decade now. Hard to imagine 10 years have gone by but here I am! When I first started out, there was so many variables to overcome to obtain a level of proficiency I could then master. I came to the bench with a level of design, fabrication experience and skill however applying those concepts and methodologies to a bicycle frame was a new challenge that required mastering new skills and further honing old ones. One of the variables that seemed like a massive hill to climb was geometry, fit and handling and how those 3 interact between rider, bicycle and terrain. Working towards a full suspension platform felt like putting the cart before the horse. But in 2019, after years applying the “build, ride, repeat” mantra, I felt like I was actually gaining a greater understanding of that trifecta’s interaction and finally felt like I was ready to apply all that I had learned out on the trail and at the bench to a bicycle with moving parts: enter full suspension and the Snakedriver Heavy Project was born.
As I’ve mentioned in previous posts, I had not ridden an FS bike since about 1998. That’s another point that’s hard to imagine. Also in 1998, I had graduated from Penn State with a BA in Jewelry and Light Metals from their College of Arts and Architecture and upon graduation had visited the Rhode Island School of Design to possibly go to graduate school for Industrial Design. A quick conversation with one of the professors answered my questions with a simple pair of questions: Did I have the degree and job I wanted and was looking to move up the pay scale and take on more of a leadership position or did I lack the degree to get the job I desired? The answer was the former. So undergraduate it was. During my senior studio, I was going to build a full suspension bike but I quickly realized I was WAY in over my head. I knew this had to be shelved and I needed to learn a heck of a lot more before I was ready to tackle that project. So after 3 years of intense design study, I had a BFA in Industrial Design and an offer from Reebok as a shoe designer… All of this from a crossroads decision back in 1998 to go back to school when I was fresh out of PSU and landed an internship at Bicycling Magazine by chance while out for a ride on a day off from my temporary bike shop wrench job. A slight sidestep from starting my own bicycle company! Fast forward to 2019: So much had changed by the time I felt confident enough to actually tackle a project of this size and complexity. Motivation was not lacking but where was I to start? There in lay the question.
So the journey began with a single pivot. I wanted to start with a known platform that was fairly simple to fabricate, was low maintenance and didn’t have too many moving parts. I also needed a starting point to understand two basic design parameters:
1. What kind of bicycle do I want to ride?
2. What kind of bicycle do I want to service and support?
Furthermore, I wanted to fabricate everything in-house which meant all of the pivot parts, sub-assemblies and of course tooling. I did this on purpose to keep the cost low and to immerse myself solely in the design and fabrication process so I did not have to manage any additional suppliers. This acted as a set of design guardrails which can be important restrictions. And it was a proud moment when everything was wrapped up and back from powder! I recall that night very clearly:
The result was a 150mm Front / 130mm Rear single pivot test platform which is below:
This bike was pretty straight forward in design and geometry. I wanted something relatively quick and snappy that could serve as a starting point to assess pivot points, front/rear center, details such as cable routing, dropper port location, water bottle placement, hardware, torque settings, bearing choices, pivot hardware design and much more. Important cliff notes below:
· 65° HT Angle
· 75° ST Angle
· 25mm Drop
· 837.5mm Front Center
· 438mm Rear Center
· 500mm Reach
· 631mm Stack
· 150mm Front Travel
· 130mm Rear Travel
· 1275.5mm Wheelbase
The swing arm was asymmetric and was a bit tricky to fabricate which resulted in losing a few millimeters in chainstay length so the rear tire under full compression just buzzed the seat tube. Here’s a close up of that fit up:
And welded. That was actually a lot of fun to weld!
The seat tube was admittedly too slack and contributed to that fabrication shortfall and immediately I realized that I needed to steepen the seat tube angle for a bit better climbing position when seated. But other than that, the bike performed rather well out on the trail. The bike was quick, had pop and was a lot of fun to ride. Front triangle storage space was clearly lacking and I really do not like the aesthetics of a bottle mount under the top tube. But the essence of what I was after was present in how the bike performed overall. The build weight was 30 pounds on the money. Subsequent builds were heavier weighing in at 32lbs which is still reasonable but I really enjoyed that round number of 30 pounds which is a target weight for the production bike. Slowly but surely, after successive rides, tuning and tweaks, a clearer picture started to form in what I wanted out of a bike of this nature. But I wasn’t done here. This was only a starting point to learn, build, ride and repeat. So on to prototype number 2.
Enter prototype number 2: A 160mm front / 140mm rear travel Horst Link big boy bike. Admittedly, I wanted to go the opposite direction of what I started with so I went super long, low and big on this next prototype. A goal of this next phase was to have a complete solid model assembly in Fusion 360 of the entire frame. All of the linkage parts, shock mounts and various bits were to be turned in house and contracted out to a local CNC machinist. Unfortunately, upon delivering part drawings and STEP files to a local machinist, about 1/2 way through machining, they lost interest and went silent. I was on a timeline to have a prototype to ride and test by mid-summmer/early fall and was losing time. Right about this time in 2020, things were starting to evolve in the SLM / 3D printed metal industry. I made the decision to completely scrap everything I had just worked an entire winter designing in Fusion 360 and completely redesign the platforms linkage components, swingarm assemblies and pretty much everything else leveraging 3D printing. So that meant no prototype for 2020. I worked all fall spending long hours again in Fusion 360 and that spring of 2021, I was ready to send files to a 3D printer. Parts arrived within a relatively short turn around and I was now able to start the fabrication process throughout the summer fitting it in between client builds here and there. By the end of July/early August, I had a completed prototype ready to head to powder:
Assembly began in earnest once the frame and swingarm returned from powder:
The geometry cliff notes of prototype number two are below:
· 64° HT Angle
· 77° ST Angle
· 25mm Drop
· 864mm Front Center
· 450mm Rear Center
· 516mm Reach
· 640mm Stack
· 160mm Front Travel
· 140mm Rear Travel
· 1304mm Wheelbase
This bike was indeed big and long and had a stance all of its own:
Physically building the bike was probably the quickest to date out of any bike I’ve built. This has much to do with all of the front end design work I performed as well as the nature of the tube and socket design of the 3D printed parts. I was quite surprised just how quickly it came together and that design time up front paid off. A horizontal shock orientation remained but I added a radius to the down tube to open up more space to be able to run a large bottle in the lower half of the front triangle. I also moved away from aluminum carriers for cable management and used zip tie style (more subtle experimentation). Steepening the seat tube angle solved seated climbing and the bike proved to be quite adept at climbing given its overall length. The bike really loved to go down hill and ate up chunk. The long 450mm chainstays were a bit of a bear to pick up and loft the front end. I did however shorten my cockpit by a good margin to get myself more upright which proved a really good choice moving forward in overall fit/feel of the bike producing a very familiar stance. The testing process was going really smoothly with tuning wrapped up but 10 days in, something I wasn’t expecting brought everything to a halt. The swing arm failed.
After some analysis, speaking with a few fellow builders who are well versed in SLM as well as contacting a few engineers, I believe the issue was two fold: The design created a confluences of forces in a high fatigue prone region of the swing arm yoke and the material (316L) is not up to the task of shelled printed bicycle components that see high stress and fatigue. The SLM process has many variables which include powder alloy type, the number of times the powder is recycled, protocol for handling the powder, part print orientation, resolution of the print, presence of inclusions in the print, and finally strength loss of the parent material due to the printing process. Material choice and post print tensile strength are two very important factors to consider. Generally, post print you’re looking at approximately 70% of tensile strength of the parent material. So for example, if the parent material’s tensile strength is 100,000 psi, post print you’re looking at about 70,000 psi tensile strength post print. So you ideally want a parent material who’s tensile strength is relatively high. So this meant looking for a new printer who provides tensile bar tests for each batch of parts so they have a record of known results and tensile strengths they’re getting post print and starting with a parent material who’s tensile strength is high to compensate for the loss post print. This means moving from 316L (which has a tensile strength of 74,000 psi and a yield strength of 29,700 psi with 60% elongation) to 15-5 which has tensile strength of 200,000 psi / yield of 184,900 psi and elongation of 10% or 17-4 with a tensile strength of 200,000 psi, yield strength of 195,100 psi and an elongation of 8%.
Of note was how quickly the 316L failed at 10 days into testing. With an elongation rate of 60%, it seems odd that the part failed that quickly without actually deformation. I will own any faulty design on my part but I still wonder if there may have also been some inclusions/imperfections in the print.
With all this said, I reached out to New Zealand based RAM3D who is the largest 3D printer of metals in the southern hemisphere. They’ve been a joy to work with, they really know the 3D printing process inside and out and have a thoroughly dialed process along with providing tensile bar test results on request (they do this regardless with each batch of parts).
So with that failure, back to the drawing board to redesign the yoke and make a 3rd prototype.
Parts arrived quickly from RAM3D and I got straight to work. Here’s a shot of all the new parts, the majority of which had internal latticing.
I decided to start fresh and not just replace the section of the swingarm that failed as now it called into question the integrity of all the 316L parts. Plus I was already thinking about revisions, so this gave me a chance to not only start with fresh parts but to revisit aspects of the design to further refine geometry and handling. I realized I went a tad too long in wheelbase with front/rear centers needing to be reeled in a touch. SO the following is what changed between V2 and V3:
· 64° HT Angle
· 77° ST Angle
· 25mm Drop
· 847mm Front Center
· 440mm Rear Center
· 500mm Reach
· 640mm Stack
· 160mm Front Travel
· 140mm Rear Travel
· 1297mm Wheelbase
These changes made a big difference out on the trail between V2 and V3. The bike still lacked some of that pop and snap that I really liked about V1 and by the time I had wrapped up design revisions I already knew I wanted to try a similar set of parameters as V1 in a 140/120mm flex pivot design. Parts count for the Horst platform was an interesting education. The more parts means the more you need to keep up with when maintenance called. We got a lot of rain this past summer and what was great about this experience was I needed to maintain the bike throughout that rainy test period. It became abundantly clear I needed to reduce parts count for maintenance and upkeep, but also for overall costing. So simplicity came to the forefront of design parameters when I began V4’s design process. Here’s a look at the Horst links parts count:
The funny part is there is a pair of bearings missing (I just noted this!) from this photo. That’s a lot for a small brand like my own to manage and for a customer to then maintain. Failure is the best way to learn and improve any design and although this was a tough part of the overall process to get through, I’m really thankful this happened now vs later during ISO testing or even worse to a potential customer. This really forced me to reassess my design process, identify shortcomings, create new goals and make improvements to the broader strokes that help guide the overall process. V3 certainly worked well but one of its short comings along with V2 is it tended to sit into its travel a bit more than I’d like. V1 sat a bit higher in its travel initially so I had more clearance under pedal load of obstacles but I also had more working travel.
While I was waiting parts for V3, I had built up V1 again and by happenstance, this was another great turn of events as it again forced me to reassess what I liked from a previous design, compare that to the current version and then apply those changes and what I had learned to V4. During V3’s extended test period, I was noting under braking load, a lot of flex in the seat stay subassembly of the swing arm. Ironically and once again, as I was assessing whether to update that section of the swing arm from 5/8 OD tubing to 3/4 OD tubing… The seat stay just above the ISO mount failed. More learning process. I thought I was safe without adding any braces under the ISO tab and most of the braking forces would be directed back up into the frame but I was wrong. I’ve actually redesigned this sub-assembly and had parts printed to use .75″ OD tubing for seat stays along with adding a pair of seat stay braces that tie into the link by the dropouts. But again, this is really important to get these failures now rather than later and they appear to be confined to the swing arm. So back on V1 again while I was designing V4. This happened on a friday, and by Sunday, I had actually repaired the swing arm because I wanted to see how the 3D printed parts could be treated in the event of a needed repair where parts were fine, but it was a tube that was damaged. Everything came together quickly and very smoothly. Here’s a look at that C3PO repair:
But that brings us to V4 and what I feel is very close to a final offering.
What I really appreciated about V1 was its pop and snappy feel. Relatively quick handling, light weight and no nonsense feel. Clean, understated simplicity. The V4’s geometry cheat sheet is below:
· 65° HT Angle
· 76° ST Angle
· 25mm Drop
· 813mm Front Center
· 435mm Rear Center
· 479mm Reach (Size Large-ish)
· 620mm Stack (Size Large-ish)
· 140mm Front Travel
· 120mm Rear Travel
· 1248mm Wheelbase
This version is built around a 140mm fork and is designed around a 185x50mm trunnion shock delivering 120mm of rear travel via a flex pivot. The idea here is a customer can choose between a Fox 36 (for example) that uses their 140mm travel kit and a Float X2 for a highly tunable and aggressive trail build or shave weight by going with a Fox 34 and say a Float X for a slightly lighter trail bike.
Additional details are as follows:
· Fits Large Water Bottle
· Tool Mount Under TT
· UDH Compatible
· ISCG05 Equipped
· External Cable Routing
· Internal ST Dropper
All of which is designed around a 30mm stem giving the option for a potential rider to size up on cockpit length with the choice of 30, 35, 40 and 50mm stems. This is pretty close to what I’ll be offering publicly. I missed the pop and snap of V1 and pretty much have nailed how I want the bike to ride and perform which is fast, aggressive trail with pop. Straight forward simplicity with an eye to serviceability and ease of maintenance in a tight package of 140mm F/ 120mm R via a flex pivot. So you like to climb and earn your descents? Jib on features, bunny hops over logs and boost off of stumps? Lean hard and carve turns? Manual and foot plant over rock walls? V4 has it. Here’s a close view of parts count for comparison between V2/V3 and the current version. SO much less to maintain!
One thing to note here is I want to make as much of these parts stock and off the shelf. So whether you’re on the road or at home, you can simply place an order with a supplier like McMaster-Carr or head to your local hardware store to pick up what you need. I worked long and hard on this aspect and I’m pretty stoked that with the exception of the two pivot axles, everything pictured above is readily available online and/or at a local Ace or True Value hardware store.
Bearing count is down from 10 to 8. I moved the bearings from the swing arm yoke to the frame which allowed me to narrow up that area, stiffen the feel of the swing arm laterally and allow room for an ISCG05 mount. Parts count is night and day. I’m considering changing the rockers main pivot width to match the swing arm pivot width of 42mm so I can use the same pivot axle width and cut down of further parts count. I’ve already sent out revisions for a 2 piece rocker design (the current version is a 3 piece design). A 32t chainring JUST fits on the current yoke design. There were some things that I just didn’t “see” in the CAD drawing that became apparent once the parts were a real assembly in hand. So I’ve been working on refining chainring clearance and considering if I max out for 34 or 32t chainrings. 34t chainrings require a bit more thinning of the profile in critical areas of the yoke to gain clearances. 30t and 32t chainrings seem a bit more the average of chainrings for this type of terrain and style of riding I’m designing around but it also allows me to shorten up the yoke a tad, and retain material in a critical area for additional strength and stiffness.
Cable routing is via aluminum clamps again (which seem to retain the cables better on an FS vs the zip tie versions on the down tube. Bottle placement is mid-down tube which actually makes reaching it much easier. The climb switch on the shock is a bit of a stretch to reach, but it has become really obvious that Fox’s line of shocks were designed to be oriented vertically. All of the controls face up towards the rider or are facing the drive side for sit down / in the stand maintenance and tuning. The bike is actually really easy to work on, assembles quickly and upkeep so far has been minimal if anything at all.
The main pivot uses a set of internal snap rings that serve as bearing stops. I’ve turned a bearing spacer so that once torqued, the pre-load on the bearings is centered around the pivot center vs the perimeter of the bearing without one. On a recent ride the M12 pivot bolt loosened up a touch, so I’ve increased torque to 20Nm and went from blue Locktite 242 to red Locktite 271 and all has been holding aok now. And that’s one of the fine details that seems lost in translation: All the small little things that I’ve had to consider in order to deliver a super reliable and easy to maintain mountain bike. What I’m really enjoying now is that I’m seeing light at the end of the tunnel so I can now focus on these smaller details and with a known design direction and suspension platform, I can further refine and hone all the key aspects and details of the bike. And this is the destination I’ve really been looking forward to arriving because I really enjoy the refinement process! Here’s a few shots of the rocker and associated pivots:
As the bike sits, its a tad over 32 lbs. One of the key areas I’m now revising is where and how to apply manufacturing techniques. So CNC vs 3D SLM vs Laser/Water Jet cutting parts that are post machined and welded into assemblies. This has been allowing me to assess each of these manufacturing techniques strengths and weaknesses and then leverage their individual processes and approaches to design uniquely to deliver the most cost effect part as well as the most easily manufactured part with an eye to strength, durability and weight savings. So I’m pretty confident through all of this refinement process I can shave out those 2 pounds to deliver a sub 30 pound complete bike.
The addition of an ISCG05 tab has been a long time coming and it really has made a big difference.
It seems with 11 and 12 speed, I’ve not broken any chains until the FS project got under way. I’ve broken 3 so far and they’ve all been preventable rock strikes and hits. This project has not only challenged my design process and forced me to reassess my own process and methodologies in the shop but its also changed the way I ride. No longer am I seeking the smoothest line but now I’m looking for the fastest and most fun line, which sometimes isn’t necessarily the smoothest line!
One last detail I’d like to touch on is the brake mount. For now, I’ve designed a proprietary brake mount that uses a custom post mount adapter. Basically I’m calling it ISO44 which is 11mm in the Y axis above the axle centerline, 22mm in the X axis to the first mounting bolt and 44mm between mounting center points. Super easy to remember. But this allows for a mount that is directly on the chainstay, is small, relatively low with good clearance and has a small footprint. One big update to this style ISO mount is there is a recessed step that the Post Mount adapter keys into on the backside of the mount on the frame. That means the bolts are no longer in sheer. Whether I formalize this and make it widely available is up for debate. This was something I wanted to try so I could separate the chainstay and seat stay from each other and allow the seat stays to act as the flex portion of the flex pivot without the need for bracing (I could add them if I want of course). I am working on a post mount part that allows me to weld it directly on to the chainstay and not need any proprietary mounting standards (the post mount part would start at 180mm in either case and be integral to the frame if welded vs the machined proprietary mount). Here’s a look
This tucks the caliper between the chainstay and seat stay and getting it out of the way of debris. It also makes the rear end incredibly clean and compact aesthetically (something I really dig). But we’ll see as things take shape and I further refine this phase of the Snakedriver’s journey.
So there you have it. For those interested, here’s a master list of the albums via Flickr of all 4 builds:
Snakedriver V1 Prototype Flickr Album
Snakedriver V2 Prototype Flickr Album
Snakedriver V3 Prototype Flickr Album
Snakedriver V4 Prototype Flickr Album
As I said before, what I’ve been after in a full suspension bike is clean understated simplicity with an eye to easy to access maintenance and upkeep in a light and tight package that is fun, poppy and has snap. So whether you’re hitting local trails or out for a longer ride or on a road trip, this bike will perform. This isn’t a park bike or meant as a burly enduro rig but rather a fun, aggressive and playful trail bike. Build it up with a bit more burly parts for more aggressive riding or strip it down and focus more on weight savings for a fun and playful trail bike.
More soon as I continue to refine this final stage of the Snakedriver project!