Recently I’ve been working with Spencer Thomson, a lead engineer at EOS North America who is a producer to 3D metal printers. Spencer reached out this past summer with a cold email essentially asking if they printed parts for me, would I like to work with them to try and speed up development of SLM (Selective Laser Melting) parts for current projects. I had never met Spencer and he was admittedly new to the cycling industry and community. He had apparently picked up cycling during the Pandemic, and immediately started to see how the 3D metal printing industry could possibly offer new ways of solving problems for cycling companies large and small. I had been recommended during a group ride as a potential collaborator due to the fact that I had been deploying SLM across the range of product I’ve been offering. So after a video meeting to get a better handle on what Spencer and EOS were offering, and to speak to what I was currently developing and how that process was going, I decided to jump at the chance to collaborate with Spencer to advance a few key projects that were just taking some time to develop.
As many of you know I’ve been working diligently designing and developing a stock mountain frame of the Marauder as well as a steel full suspension frame, the Snakedriver. Eventually what I’m working towards is a business model that has stock offerings and I then accept a set number of custom clients (I’m going to make an announcement in the future outlining this in more detail). Eventually I’ll also offer a stock steel version of the Huntsman as well. But my dream of owning a true bicycle company, that builds in small batches with a small line up of USA built frame kits and complete options is starting to take shape. What has been slowing the process down is development and lead times (and a limited budget of course!).For the stock Marauder, the last piece of the puzzle has been the yoke. This eliminates the need to dimple the drive side chainstay, increases lateral stiffness and durability, allows me to use a material that is incredibly strong in a critical area of a bicycle, all of which puts a very thoughtful designed part with parameters designed and built into it using SLM technology.
Additive manufacturing is a completely different way of thinking about the “how” you design parts for complex products such as a bicycle. Traditional CNC / Subtractive Manufacturing has its merits and has a set of parameters and head space design wise you need to work around. But Additive really changes the game: You can now fully realize part geometry and design outside AND inside with relatively zero compromises. With CNC, you’re a bit limited by the process, budget and the machines used that are within your scope of the project. A part as complex as my yoke (albeit its fairly simple in reality) creates a few very challenging hurdles that only a 4 or 5 axis CNC can tackle. And expense: The part is very organic with a lot of surfacing features that would be cost prohibitive to produce traditionally. Its actually a really great candidate for investment casting, but again, budget and now volume are at issue with that sort of part development for that specific manufacturing technique. So this is where Additive Manufacturing can really shine and bridge a gap for small manufacturers like myself.
All the parts I’ve been working with Spencer have been printed from 17-4 stainless steel. 17-4 is a Martensitic stainless steel. In short, this is a class of stainless steels that can be post weld heat treated to achieve incredible strength and hardness (I’ll come back to this). In its machinable state, it is annealed (I’ll revisit this too). Again, this is an incredibly strong base material. With SLM, a general rule of thumb as I understand it from my limited experience deploying the technology is that you get about 70% of ultimate tensile strength from the base material post print. This is why, according to my opinion, 316L is a very poor choice of printed base material for any components on a bicycle other than small non-load bearing parts like braze-ons and such. Think derailleur cable guide or brake line guide for example. Critical areas of a bicycle see a lot of dynamic loads due to pedaling forces applied as well as feedback from contact with the ground and terrain traversed. So starting with a very high strength material is key. All of the parts I’ve received from RAM3D to date have been 15-5 stainless steel and all parts I’ve received from Silca for Titanium printed parts have been 6/4 with a post print heat treatment to normalize the parts. These are also both VERY strong base materials. Think yield strength in the 100’s of thousands (I’d have to look up the true numbers, but you follow me here with 316L being in the 20k to 30k yield strength).
The first part I’d like to talk about the design and development process with everyone reading is the seat stay tip for the swingarm of the Snakedriver FS. Originally, this was designed as a single sided part that was asymmetric left to right of the frame (so side specific). It used a single stock shoulder bolt and washer. I’ll have to go into a separate post about stock hardware! But what was giving me issues out on the trail due to this simple overlapping design was a tad bit of wiggle in very specific terrain circumstances. Think rock gardens at high speed where there’s a lot of space between rocks/hits. This allowed the rear wheel to track and pop or slide a bit between hits. I was feeling a bit of odd sensation through the rears travel. Almost like the bike would take the hit, move on to the next one but between the hits there was a resonance or wiggle feeling. I traced that back to the seat stay tip design and the actual tolerance of the shoulder bolt of the stock bolt. Now that stock shoulder bolt was rather short (its first issue – these are relatively hard to find stock in shorter lengths). Many bicycle companies using this overlapping design have all the hardware custom designed to spec and tolerance. Something I can’t afford to do. So that means stock hardware pretty much throughout the BOM (Bill of Materials). But what I was up against was something I did not have control over and that was the tolerance of the bolt. Its actually a good fit between the Enduro 26900 LLU Max bearing which has a 10mm bore and the shoulder of the bolt (Tolerance of this one in particular is -.04 to -.01mm). That’s not much, but spread that across 435mm to the rear wheel x2 two bolts and connections, and you get a bit of lateral movement that is otherwise imperceptible testing the fit between bearing and shoulder bolt. This meant revising that junction to a clevis style part where both sides of the bolt are captured with a shoulder bolt holding the assembly together. This increases surface contact area and further stiffens the junction laterally. I also can choose to use a longer, more common shoulder bolt that have precision stock offerings (so a TIGHTER tolerance). So back to the drawing board design wise and then collaborate with Spencer to have the lead time shortened so I can try new swing arms on existing prototype front triangles. This is where a rigorous approach to design methodology comes into play: Actual Industrial Design!
After a handful of iterations and 3 physical versions of the new clevis design, I’ve settled on a semi-captured nut design (the image above the the 2nd iteration with a mock up of the semi-captured nut using an existing part – the lip is much more pronounced and sits proud of the nut to capture one face). I had 3 design iterations (and who knows how many intermediate designs in Fusion 360) that really came to fruition. The first had an integrated thread. This posed some tolerance issues in post op machining. I’d bring the bore up to the 10mm spec and machine the pilot hole to spec but if there was any misalignment, the bore and thread would bind (not good for tight tolerance bolts). The second iteration, which had a fully captured nut design, and I will admit this is a clever design on my part, posed equal tolerance issues. Ideally, the part should be held post print in a 4 or 5 axis CNC in custom tooling, the 10mm bore machined to spec and then the captured nut fully machined second so everything is dead on. But because I’m working on a Bridgeport, human error comes into play and again, can create binding between the bore and shoulder bolt. This led me to the final design and a bit of a compromise between the first two ideas. I needed a part that I could register, machine the bore to 10mm and if there was a bit of tolerance between nut, bolt and clevis, I could take a bit off of the clevis to eliminate any stacked tolerances. And that is what the semi-captured nut design does. Any one of the 3 interfaces that need to be toleranced post print can be achieved without interfering with the other 2 interfaces awaiting tolerance machining or if they’ve already been machined. This really solved a lot of headaches in this simple part (recall simple is not easy). And I’m maintaining my design goal of all stock hardware. The final version also eliminates the need for any washers between the clevis and the bearing. This meant updating the rocker a tad, but not by much! A simple change that also did not require new tooling.
Out on the trail with prototype versions of the clevis and new swing arm assemblies confirmed I was on the right track. That lateral “wiggle” I was experiencing was eliminated and the rear end although still retaining that signature pop and snap that steel is known for, now is much more solid and laterally stiff. Here’s some examples of the swing arms which also incorporated a new main pivot yoke that increases both tire and chainring clearance while gaining additional strength and stiffness with ease of post print machining.
The next part I’d like to discuss is the stock Marauder’s yoke. I mentioned up above about 17-4 being a Martensitic stainless steel. This class of stainless steels can be age hardened and solution aged for a stress relief. To date, all 15-5 parts from RAM3D have arrived at the shop “as printed” meaning aside from support structures being removed, surface finished and receiving a media blast, they have not been heat treated (my 6/4 Ti parts have been heat treated for a stress relief – this is a bit different which I won’t get into…). Speaking with Spencer and then Nick at Cumberland Additive, we began to explore options of heat treatment to H900 and a solution age for stress relief. Up to this point, I had not really considered any of this, but it posed an important question that I needed to find answers because what I know of 17-4, and what research confirmed is that it is machined and welded in its annealed state and depending on the parts requirements will THEN see a stress relief solution age with a (for example) H900 heat treatment. This brings up the overall tensile strength from about 160,000psi in its annealed state to about 195,000psi, with a rockwell hardness of about 40C and an elongation rate of 3-5% (that’s not good btw). Thats ridiculously hard and strong. What was essentially being suggested was to reverse a process I was familiar with. Asking some questions of experts confirmed my concern with this method and approach. You do not want to weld 17-4 in its H900 state as that can lead to weld cracks along with the features described above which are not desireable. What I’m after is high strength and a reasonable amount of ductility. If you wanted a heat treatment post weld to have an overall stress relief, you’d choose an H1150 treatment but that may have some negative results in alignment in the entire assembly (along with being a bit unnecessary). So that means and from what I’ve already been receiving to confirm this, parts will come from suppliers in their “as printed” state, no heat treatment or solution age for stress relief. There is quite a bit of heat in the laser so there is some stress in the parts, but the question now is: How much stress? So I’m in the process of having parts printed to make subassemblies to be fatigue tested as well as to understand what the materials base properties from the SLM process in 17-4. Compositionally, its 17-4. But how does that compare with wrought 17-4’s properties? Destructive testing in-house has shown that the 17-4 yokes printed to date are much stronger than their 4130 chainstays that they’re welded to and part of the overall assembly. So the chainstay will fail before the yoke (which is the point). But that’s where I am on the materials end.
On the design end, I have finalized the design of the yoke in 6/4 titanium for 22mm tubes so now the challenge was scaling this part for 19mm 4130 tubing. Again, that’s fairly simple and straight forward but not so simple and straight forward. I had the bulk of the design finished and was able to use that as a base to launch from for the part in steel but again, I have to essentially reverse engineer surfacing from forming techniques in 3D real time INTO Fusion 360 which is a virtual environment. A significant number of the man hours were devoted to getting the transitions between the socket interface and the internal radius of the part to flow and match the formed chainstays that are a bit of a signature to my bicycles. Creating and maximizing tire clearance while allowing room for up to a 34t oval chainring was a challenge (and having it work in both 52 and 55mm chainlines too!). But after significant heavy lifting, I think I achieved a very simple and classy part that captures the overall aesthetic of my forming techniques but in a printed part. What makes this part extra special is whats going on in the inside. The spline has a series of stepped wall thickness transitions which puts material in critical locations while thinning it out in less critical locations to shed weight. The part isn’t much heavier in 17-4 than in 6/4.
Here’s a shot of the internal struts “step”:
And finished results:
Now comparing the physical properties of both 15-5 and 17-4 gets into the nitty gritty of where I’m headed. Annealed 15-5 has a yield of 145,000psi, a Rockwell hardness of C31 and an elongation rate of 15%. Compare that to 17-4 which has a yield of 118,000psi, A rockwell hardness of C28 and an elongation of 10%. If you heat treat 17-4 elongation drops even lower, but rockwell hardness goes into the C40 range I believe and H900 yield jumps to 190,000psi (which we’ve already covered you do not want to do before you weld 17-4). Recall post print, we’re looking at about 70% of all of these properties. Compositionally, the important stuff is 15-5 has 71-78% iron vs 68% iron of 17-4, chromium is about 15% for 15-5 while 17-7 is about 17%, and carbon is sitting equal around .07%. So 17-4 is slightly less hard, but its yield is lower than 15-5 with an even lower elongation rate. What I’m looking for in a yoke material is high strength with a reasonable amount of ductility and good elongation rate. So for this particular part, aka a yoke, 15-5 wins out. But since 17-4 is typically used in high wear situations, like things that rotate, perhaps the clevis is a much better application for 17-4.
SLM has posed a lot of incredible advantages and posed a lot of new challenges and hurdles to overcome. The process has also led me down a rabbit hole that has become a masterclass in materials, their applications and “how” you can leverage those aspects not only in the design of parts but also in picking appropriate manufacturing approaches. I’ve mentioned this above but I’ll reiterate it here in closing: SLM is not a end-all be-all manufacturing choice. SLM is yet another tool in the collective framebuilders’ proverbial toolbox. It sits neatly along side traditional CNC and additional subtractive techniques such as laser cutting, waterjet cutting as well as stamped/folded techniques or combinations of these techniques. And this is what I love about design and fabrication: Design parts but then assessing what is the BEST methodology to realize that same part. Is it CNC? What about waterjet cut with post machine operations? Or Maybe it is SLM! This all considered on the front end of the design process informs your decisions down stream as you progress. And sometimes a shift mid-project from one technique to another can yeild radically different results in the design approach. It truly is a wonderful time to be a framebuilder. It isn’t easy and it does require a multitude of skills on top of those required to market, design and run a bicycle company. Most of us are 1 man/woman shops, myself included. But what has continued to inspire me is the design and fabrication process and the interplay between the two. And then being able to actually ride what I’ve created? I can’t think of a better way to spend my life work. Yes this is about putting more people on bikes, but I have to say for me, its always been about personal skill progression both on and off the bike and in the design and fabrication space. To be as well rounded and knowledgeable as I can possibly be and strive for a higher ground; to seek knowledge and to continually refine my skills.
PS: A very special thank you goes out to Spencer of EOS, Nick from Cumberland Additive, the entire crew at RAM3d, Thomas at Hosford and Co, Sean Handerhan of Handerhan Cycles and Brian from O’Neill Precision Welding. Without your collective knowledge and support, I couldn’t be doing what I’m doing to the level I’m currently designing and building. So from me to all of you: Thank you!
Research and Development
Recently I’ve been working with Spencer Thomson, a lead engineer at EOS North America who is a producer to 3D metal printers. Spencer reached out this past summer with a cold email essentially asking if they printed parts for me, would I like to work with them to try and speed up development of SLM (Selective Laser Melting) parts for current projects. I had never met Spencer and he was admittedly new to the cycling industry and community. He had apparently picked up cycling during the Pandemic, and immediately started to see how the 3D metal printing industry could possibly offer new ways of solving problems for cycling companies large and small. I had been recommended during a group ride as a potential collaborator due to the fact that I had been deploying SLM across the range of product I’ve been offering. So after a video meeting to get a better handle on what Spencer and EOS were offering, and to speak to what I was currently developing and how that process was going, I decided to jump at the chance to collaborate with Spencer to advance a few key projects that were just taking some time to develop.
As many of you know I’ve been working diligently designing and developing a stock mountain frame of the Marauder as well as a steel full suspension frame, the Snakedriver. Eventually what I’m working towards is a business model that has stock offerings and I then accept a set number of custom clients (I’m going to make an announcement in the future outlining this in more detail). Eventually I’ll also offer a stock steel version of the Huntsman as well. But my dream of owning a true bicycle company, that builds in small batches with a small line up of USA built frame kits and complete options is starting to take shape. What has been slowing the process down is development and lead times (and a limited budget of course!).For the stock Marauder, the last piece of the puzzle has been the yoke. This eliminates the need to dimple the drive side chainstay, increases lateral stiffness and durability, allows me to use a material that is incredibly strong in a critical area of a bicycle, all of which puts a very thoughtful designed part with parameters designed and built into it using SLM technology.
Additive manufacturing is a completely different way of thinking about the “how” you design parts for complex products such as a bicycle. Traditional CNC / Subtractive Manufacturing has its merits and has a set of parameters and head space design wise you need to work around. But Additive really changes the game: You can now fully realize part geometry and design outside AND inside with relatively zero compromises. With CNC, you’re a bit limited by the process, budget and the machines used that are within your scope of the project. A part as complex as my yoke (albeit its fairly simple in reality) creates a few very challenging hurdles that only a 4 or 5 axis CNC can tackle. And expense: The part is very organic with a lot of surfacing features that would be cost prohibitive to produce traditionally. Its actually a really great candidate for investment casting, but again, budget and now volume are at issue with that sort of part development for that specific manufacturing technique. So this is where Additive Manufacturing can really shine and bridge a gap for small manufacturers like myself.
All the parts I’ve been working with Spencer have been printed from 17-4 stainless steel. 17-4 is a Martensitic stainless steel. In short, this is a class of stainless steels that can be post weld heat treated to achieve incredible strength and hardness (I’ll come back to this). In its machinable state, it is annealed (I’ll revisit this too). Again, this is an incredibly strong base material. With SLM, a general rule of thumb as I understand it from my limited experience deploying the technology is that you get about 70% of ultimate tensile strength from the base material post print. This is why, according to my opinion, 316L is a very poor choice of printed base material for any components on a bicycle other than small non-load bearing parts like braze-ons and such. Think derailleur cable guide or brake line guide for example. Critical areas of a bicycle see a lot of dynamic loads due to pedaling forces applied as well as feedback from contact with the ground and terrain traversed. So starting with a very high strength material is key. All of the parts I’ve received from RAM3D to date have been 15-5 stainless steel and all parts I’ve received from Silca for Titanium printed parts have been 6/4 with a post print heat treatment to normalize the parts. These are also both VERY strong base materials. Think yield strength in the 100’s of thousands (I’d have to look up the true numbers, but you follow me here with 316L being in the 20k to 30k yield strength).
The first part I’d like to talk about the design and development process with everyone reading is the seat stay tip for the swingarm of the Snakedriver FS. Originally, this was designed as a single sided part that was asymmetric left to right of the frame (so side specific). It used a single stock shoulder bolt and washer. I’ll have to go into a separate post about stock hardware! But what was giving me issues out on the trail due to this simple overlapping design was a tad bit of wiggle in very specific terrain circumstances. Think rock gardens at high speed where there’s a lot of space between rocks/hits. This allowed the rear wheel to track and pop or slide a bit between hits. I was feeling a bit of odd sensation through the rears travel. Almost like the bike would take the hit, move on to the next one but between the hits there was a resonance or wiggle feeling. I traced that back to the seat stay tip design and the actual tolerance of the shoulder bolt of the stock bolt. Now that stock shoulder bolt was rather short (its first issue – these are relatively hard to find stock in shorter lengths). Many bicycle companies using this overlapping design have all the hardware custom designed to spec and tolerance. Something I can’t afford to do. So that means stock hardware pretty much throughout the BOM (Bill of Materials). But what I was up against was something I did not have control over and that was the tolerance of the bolt. Its actually a good fit between the Enduro 26900 LLU Max bearing which has a 10mm bore and the shoulder of the bolt (Tolerance of this one in particular is -.04 to -.01mm). That’s not much, but spread that across 435mm to the rear wheel x2 two bolts and connections, and you get a bit of lateral movement that is otherwise imperceptible testing the fit between bearing and shoulder bolt. This meant revising that junction to a clevis style part where both sides of the bolt are captured with a shoulder bolt holding the assembly together. This increases surface contact area and further stiffens the junction laterally. I also can choose to use a longer, more common shoulder bolt that have precision stock offerings (so a TIGHTER tolerance). So back to the drawing board design wise and then collaborate with Spencer to have the lead time shortened so I can try new swing arms on existing prototype front triangles. This is where a rigorous approach to design methodology comes into play: Actual Industrial Design!
After a handful of iterations and 3 physical versions of the new clevis design, I’ve settled on a semi-captured nut design (the image above the the 2nd iteration with a mock up of the semi-captured nut using an existing part – the lip is much more pronounced and sits proud of the nut to capture one face). I had 3 design iterations (and who knows how many intermediate designs in Fusion 360) that really came to fruition. The first had an integrated thread. This posed some tolerance issues in post op machining. I’d bring the bore up to the 10mm spec and machine the pilot hole to spec but if there was any misalignment, the bore and thread would bind (not good for tight tolerance bolts). The second iteration, which had a fully captured nut design, and I will admit this is a clever design on my part, posed equal tolerance issues. Ideally, the part should be held post print in a 4 or 5 axis CNC in custom tooling, the 10mm bore machined to spec and then the captured nut fully machined second so everything is dead on. But because I’m working on a Bridgeport, human error comes into play and again, can create binding between the bore and shoulder bolt. This led me to the final design and a bit of a compromise between the first two ideas. I needed a part that I could register, machine the bore to 10mm and if there was a bit of tolerance between nut, bolt and clevis, I could take a bit off of the clevis to eliminate any stacked tolerances. And that is what the semi-captured nut design does. Any one of the 3 interfaces that need to be toleranced post print can be achieved without interfering with the other 2 interfaces awaiting tolerance machining or if they’ve already been machined. This really solved a lot of headaches in this simple part (recall simple is not easy). And I’m maintaining my design goal of all stock hardware. The final version also eliminates the need for any washers between the clevis and the bearing. This meant updating the rocker a tad, but not by much! A simple change that also did not require new tooling.
Out on the trail with prototype versions of the clevis and new swing arm assemblies confirmed I was on the right track. That lateral “wiggle” I was experiencing was eliminated and the rear end although still retaining that signature pop and snap that steel is known for, now is much more solid and laterally stiff. Here’s some examples of the swing arms which also incorporated a new main pivot yoke that increases both tire and chainring clearance while gaining additional strength and stiffness with ease of post print machining.
The next part I’d like to discuss is the stock Marauder’s yoke. I mentioned up above about 17-4 being a Martensitic stainless steel. This class of stainless steels can be age hardened and solution aged for a stress relief. To date, all 15-5 parts from RAM3D have arrived at the shop “as printed” meaning aside from support structures being removed, surface finished and receiving a media blast, they have not been heat treated (my 6/4 Ti parts have been heat treated for a stress relief – this is a bit different which I won’t get into…). Speaking with Spencer and then Nick at Cumberland Additive, we began to explore options of heat treatment to H900 and a solution age for stress relief. Up to this point, I had not really considered any of this, but it posed an important question that I needed to find answers because what I know of 17-4, and what research confirmed is that it is machined and welded in its annealed state and depending on the parts requirements will THEN see a stress relief solution age with a (for example) H900 heat treatment. This brings up the overall tensile strength from about 160,000psi in its annealed state to about 195,000psi, with a rockwell hardness of about 40C and an elongation rate of 3-5% (that’s not good btw). Thats ridiculously hard and strong. What was essentially being suggested was to reverse a process I was familiar with. Asking some questions of experts confirmed my concern with this method and approach. You do not want to weld 17-4 in its H900 state as that can lead to weld cracks along with the features described above which are not desireable. What I’m after is high strength and a reasonable amount of ductility. If you wanted a heat treatment post weld to have an overall stress relief, you’d choose an H1150 treatment but that may have some negative results in alignment in the entire assembly (along with being a bit unnecessary). So that means and from what I’ve already been receiving to confirm this, parts will come from suppliers in their “as printed” state, no heat treatment or solution age for stress relief. There is quite a bit of heat in the laser so there is some stress in the parts, but the question now is: How much stress? So I’m in the process of having parts printed to make subassemblies to be fatigue tested as well as to understand what the materials base properties from the SLM process in 17-4. Compositionally, its 17-4. But how does that compare with wrought 17-4’s properties? Destructive testing in-house has shown that the 17-4 yokes printed to date are much stronger than their 4130 chainstays that they’re welded to and part of the overall assembly. So the chainstay will fail before the yoke (which is the point). But that’s where I am on the materials end.
On the design end, I have finalized the design of the yoke in 6/4 titanium for 22mm tubes so now the challenge was scaling this part for 19mm 4130 tubing. Again, that’s fairly simple and straight forward but not so simple and straight forward. I had the bulk of the design finished and was able to use that as a base to launch from for the part in steel but again, I have to essentially reverse engineer surfacing from forming techniques in 3D real time INTO Fusion 360 which is a virtual environment. A significant number of the man hours were devoted to getting the transitions between the socket interface and the internal radius of the part to flow and match the formed chainstays that are a bit of a signature to my bicycles. Creating and maximizing tire clearance while allowing room for up to a 34t oval chainring was a challenge (and having it work in both 52 and 55mm chainlines too!). But after significant heavy lifting, I think I achieved a very simple and classy part that captures the overall aesthetic of my forming techniques but in a printed part. What makes this part extra special is whats going on in the inside. The spline has a series of stepped wall thickness transitions which puts material in critical locations while thinning it out in less critical locations to shed weight. The part isn’t much heavier in 17-4 than in 6/4.
Here’s a shot of the internal struts “step”:
And finished results:
Now comparing the physical properties of both 15-5 and 17-4 gets into the nitty gritty of where I’m headed. Annealed 15-5 has a yield of 145,000psi, a Rockwell hardness of C31 and an elongation rate of 15%. Compare that to 17-4 which has a yield of 118,000psi, A rockwell hardness of C28 and an elongation of 10%. If you heat treat 17-4 elongation drops even lower, but rockwell hardness goes into the C40 range I believe and H900 yield jumps to 190,000psi (which we’ve already covered you do not want to do before you weld 17-4). Recall post print, we’re looking at about 70% of all of these properties. Compositionally, the important stuff is 15-5 has 71-78% iron vs 68% iron of 17-4, chromium is about 15% for 15-5 while 17-7 is about 17%, and carbon is sitting equal around .07%. So 17-4 is slightly less hard, but its yield is lower than 15-5 with an even lower elongation rate. What I’m looking for in a yoke material is high strength with a reasonable amount of ductility and good elongation rate. So for this particular part, aka a yoke, 15-5 wins out. But since 17-4 is typically used in high wear situations, like things that rotate, perhaps the clevis is a much better application for 17-4.
SLM has posed a lot of incredible advantages and posed a lot of new challenges and hurdles to overcome. The process has also led me down a rabbit hole that has become a masterclass in materials, their applications and “how” you can leverage those aspects not only in the design of parts but also in picking appropriate manufacturing approaches. I’ve mentioned this above but I’ll reiterate it here in closing: SLM is not a end-all be-all manufacturing choice. SLM is yet another tool in the collective framebuilders’ proverbial toolbox. It sits neatly along side traditional CNC and additional subtractive techniques such as laser cutting, waterjet cutting as well as stamped/folded techniques or combinations of these techniques. And this is what I love about design and fabrication: Design parts but then assessing what is the BEST methodology to realize that same part. Is it CNC? What about waterjet cut with post machine operations? Or Maybe it is SLM! This all considered on the front end of the design process informs your decisions down stream as you progress. And sometimes a shift mid-project from one technique to another can yeild radically different results in the design approach. It truly is a wonderful time to be a framebuilder. It isn’t easy and it does require a multitude of skills on top of those required to market, design and run a bicycle company. Most of us are 1 man/woman shops, myself included. But what has continued to inspire me is the design and fabrication process and the interplay between the two. And then being able to actually ride what I’ve created? I can’t think of a better way to spend my life work. Yes this is about putting more people on bikes, but I have to say for me, its always been about personal skill progression both on and off the bike and in the design and fabrication space. To be as well rounded and knowledgeable as I can possibly be and strive for a higher ground; to seek knowledge and to continually refine my skills.
PS: A very special thank you goes out to Spencer of EOS, Nick from Cumberland Additive, the entire crew at RAM3d, Thomas at Hosford and Co, Sean Handerhan of Handerhan Cycles and Brian from O’Neill Precision Welding. Without your collective knowledge and support, I couldn’t be doing what I’m doing to the level I’m currently designing and building. So from me to all of you: Thank you!