Page 1 of 2 12 LastLast
Results 1 to 100 of 120
  1. #1
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679

    Show and tell on a... Monday? Milestone: carbon lugs.

    Via sweating out the aftermath of many failures, and with the help of many people, I'm now able to do this kinda bladder inflation molded bizniss, rough cut and unpolished as it may be at the moment:

    Show and tell on a... Monday? Milestone: carbon lugs.-img_20140421_094029.jpg

    Don't know how many times I failed to produce this sort of result. No idea.

    Once I get a proper 3 axis CNC mill, I should be able to prototype these very cost effectively. Now to work out more of the bugs.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  2. #2
    Nemophilist
    Reputation: TrailMaker's Avatar
    Join Date
    Sep 2007
    Posts
    1,873
    Damn dude!

    About time you got off your duff and got some of these made!

    I'm UTTERLY TOTALLY KIDDING of course. You are a testament to dogged determination, if nothing else. I somehow knew you would figure it out eventually, and I am sure you will keep figuring it out for a long time to come as you refine your process. If anyone else has been thinking about doing this, you better be mentally tough enough to keep your head down and your eye on the prize. It's a LOOOONG climb!

    Mucho kudos, Amigo!
    Last edited by TrailMaker; 04-22-2014 at 03:57 PM. Reason: Spellung
    Most people ply the Well Trodden Path. A few seek a different way, and leave a Trail behind.
    - John Hajny, a.k.a. TrailMaker

  3. #3
    mtbr member
    Reputation: Eric Malcolm's Avatar
    Join Date
    Dec 2011
    Posts
    1,014
    Great to get that milestone. Tick off another box. Keep going.

    I know that this is a lug, but my traditional heritage would have me wanting to sort of make the sockets 'pointy'. Its early days to ask this question, but can that be done from a lug like this? Carbon Tubes bonded in, instead of filler as you would use with steel. I just like the art work....

    Eric
    If I don't make an attempt, how will I know if it will work?

  4. #4
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    TrailMaker - I know you kid, but there's some truth to it even if unintentional. Fell on my ass a bunch!

    Eric - I follow what you mean, and I think those frames are timeless. The answer is yes it could be done - diamond edged hole saw and careful manual cleanup. I want to have lots of less ambitious trials done before I take a stab at that style.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  5. #5
    mtbr member
    Reputation:
    Join Date
    Feb 2010
    Posts
    1,774
    That's looking great. Any more pics of the molds and bladders?

  6. #6
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Quote Originally Posted by TigWorld View Post
    That's looking great. Any more pics of the molds and bladders?
    Some attached. In order:

    1) A "2D" or lay-flat bladder.
    2) Wet sanding after polyester sealing the milled area of the MDF.
    3) Put on an epoxy topcoat for a bit of shine, it's not evenly applied as would be had with a proper clearcoat. EDIT - I was SUPER pleased to see the mallard plumage effect after polishing and epoxy coat. I think the iridescent effects of UD fiber is way more interesting to look at in person than twill fiber patterns.

    I'm not actually using this part as I damaged the mold in use, wanted to just get a reasonable idea how a part might turn out once I learn more about the foibles of high pressure mold design.

    For starters, not going to use MDF. I bent some Uni-Strut reinforcement and dimpled the crap out of the exterior of the mold.

    I think one reason that this topic is not better documented the way metal working is pretty well documented is maybe because it is dangerous if disrespected. I think I remember calculating after damaging the mold that there was a near half-ton parting force between the mold halves at a "mere" 25 psi.

    Show and tell on a... Monday? Milestone: carbon lugs.-img_20140410_125259.jpgShow and tell on a... Monday? Milestone: carbon lugs.-img_20140412_185140.jpgShow and tell on a... Monday? Milestone: carbon lugs.-img_20140422_235429.jpg
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  7. #7
    mtbr member
    Reputation: Testmule's Avatar
    Join Date
    Jul 2013
    Posts
    126
    Nicely done!

    I have another CF frame planned for this spring or summer, along with some ideas of my own. Will share when I am under way.

    And by the way, its alway a little sweeter when it works after many failures.

  8. #8
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Hah, definitely. Can't appreciate the good without the bad. Now that I've had some time to stop being pleased with myself, I gave the part a thorough inspection and judged a whole BUNCH of reasons on why it's not ready for prime time. That stuff is really difficult to photograph with the equipment I have.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  9. #9
    mtbr member
    Reputation: ktm520's Avatar
    Join Date
    Apr 2004
    Posts
    1,570
    Room temp cure?

  10. #10
    would rather be ruined
    Reputation:
    Join Date
    Jun 2007
    Posts
    185
    drew

    i was trying to figure out if the radius finished edges were a feature but now i ve seen the tool it makes sense obviously its a front triangle not just a head lug but same system is used

    I have attached a picture to flikr for ya drew this is how we mold a head tube with bearing seats and everything to size

    i will try to crop down another image to show you the overwrap and bonded joint area where the tubes get glued to the lugs


    https://www.flickr.com/photos/compositepro/14034619351/

    one of the reasons i dont really show and tell with bladders is that pressure is dangerous and you get some guys trying to hit 200psi because they heard somewhere thats what company A does it at and then they heat it up to 90 degrees
    well it goes without saying someone will say i saw it on the internet

  11. #11
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Quote Originally Posted by compositepro View Post
    one of the reasons i dont really show and tell with bladders is that pressure is dangerous and you get some guys trying to hit 200psi because they heard somewhere thats what company A does it at and then they heat it up to 90 degrees
    well it goes without saying someone will say i saw it on the internet
    Yeah, that's kinda what I'm driving at. The super high pressure is so whatever small voids you leave during even careful layup, the air bubbles will *dissolve*, right?

    Well - you also need, like, a full size truck to squish those mold pieces together.

    And all I drive are sedans

    Which kinda bums me out - I was initially so bummed out about how little doco there was on the net about this stuff, and I see it might be for a legitimate reason. My hatred for censorship and my appreciation for safety are butting heads here.

    ktm520: the raw material was a prepreg that accepts 180 F to 240 F, I used 200.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  12. #12
    would rather be ruined
    Reputation:
    Join Date
    Jun 2007
    Posts
    185
    Quote Originally Posted by Drew Diller View Post
    Yeah, that's kinda what I'm driving at. The super high pressure is so whatever small voids you leave during even careful layup, the air bubbles will *dissolve*, right?

    Well - you also need, like, a full size truck to squish those mold pieces together.

    And all I drive are sedans

    Which kinda bums me out - I was initially so bummed out about how little doco there was on the net about this stuff, and I see it might be for a legitimate reason. My hatred for censorship and my appreciation for safety are butting heads here.

    ktm520: the raw material was a prepreg that accepts 180 F to 240 F, I used 200.
    Well high psi isn't a prerequisite I know we have solved problems for factories where too high a pressure was assumed to be beneficial frames were light on resin and delaminating even demoulding them

    Using a bladder and incorrect pressure you can trap and lock off air in the laminate it wont escape or boil off any volatiles(older resin systems) as it would in an autoclave

    Do you know about your resin cure cycle through to its tg at some point it will have a lower cp (more fluid) this usually happens before gel though not always hence you can time the bladder pressure to ramp at certain points this means you can run less psi throughout the process

    Don't rule out sticking a vacuum line in their just as a further experiment

  13. #13
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Quote Originally Posted by compositepro View Post
    Well high psi isn't a prerequisite I know we have solved problems for factories where too high a pressure was assumed to be beneficial frames were light on resin and delaminating even demoulding them

    Using a bladder and incorrect pressure you can trap and lock off air in the laminate it wont escape or boil off any volatiles(older resin systems) as it would in an autoclave

    Do you know about your resin cure cycle through to its tg at some point it will have a lower cp (more fluid) this usually happens before gel though not always hence you can time the bladder pressure to ramp at certain points this means you can run less psi throughout the process

    Don't rule out sticking a vacuum line in their just as a further experiment
    I have some rather detailed docs on the prepreg resin I'm using, unfortunately I don't have a heating system (yet) that can go into the sort of detail described.

    Also, what you're saying is consistent with what I saw during my first pressurization test with this mold (which failed), the resin viscosity at time of demold was very thin, sort of like snot. I had to laugh at myself after the fact, I wasn't using a thermocouple or anything, just a few reference plies that I checked on intermittently - yeah well, MDF is wood and glue and therefore has a stupidly low thermal conductivity, I basically interrupted the first molding's attempt when the resin was pre-gel state.

    So yeah. Thermocouples. Duh.

    I've been kicking a few other ideas around in my head to possibly increase the safety factor. Nothing definite yet, and at the moment more focused on getting a halfway decent mill running. With this mold, the machining costs alone... if I were to make say just 20 prototypes in house, I'd break even super fast on a mill heavy enough for high speed non-metals machining.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  14. #14
    mtbr member
    Reputation: ktm520's Avatar
    Join Date
    Apr 2004
    Posts
    1,570
    Interesting that the MDF mold will withstand elevated cures.

  15. #15
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Quote Originally Posted by ktm520 View Post
    Interesting that the MDF mold will withstand elevated cures.
    Forgot that detail: it didn't. The parting faces of the mold were not touching and as such I can't use this cured part (among other reasons).

    There was 2-3mm of curl away from plane on each face at the edges.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  16. #16
    mtbr member
    Reputation:
    Join Date
    Feb 2010
    Posts
    1,774
    Love what you're doing. Keep up the good work.

    If you're not going to use mdf then what next? Big lumps of alu are not cheap. For that head-tube area a mold could be made out of sectioned pipe, mitered, welded and flanged. Miter all the pipes, weld them together, cut in half, weld on flanges so you can bolt back together, bolt back together and then bore in from each open end to your desired dimensions.

  17. #17
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Quote Originally Posted by TigWorld View Post
    Love what you're doing. Keep up the good work.

    If you're not going to use mdf then what next?
    Either a frangible mold design, or something else I haven't even tested yet and people who I mention it to thinks I'm crazy (but my dad doesn't, and he's a nutcase about plastics).

    Jury's out on that subject.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  18. #18
    Nemophilist
    Reputation: TrailMaker's Avatar
    Join Date
    Sep 2007
    Posts
    1,873
    How'bout cement?
    Most people ply the Well Trodden Path. A few seek a different way, and leave a Trail behind.
    - John Hajny, a.k.a. TrailMaker

  19. #19
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Quote Originally Posted by TrailMaker View Post
    How'bout cement?
    Like the Corian variety? Yeah, possibly. I caught a handful of molds composed of that.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  20. #20
    mtbr member
    Reputation:
    Join Date
    Jul 2008
    Posts
    1,615
    Quote Originally Posted by TrailMaker View Post
    How'bout cement?
    Mortar or concrete found in a ready mix bag at your home center ranges from 3000-5000psi compressive strength. Plenty to resist the parting pressure but you'd need to bind the halves with some A36 (or better) steel shapes and high-strength bolting. Unistrut is not the answer as you found.
    You might need some sort of interior metal sleeving as I'm not sure you'd get repeatable results out of the mortar alone. I think the sharp corners and creases would break free because they are thin and brittle.

  21. #21
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Quote Originally Posted by bme107 View Post
    Unistrut is not the answer as you found.
    Indeed so. I've been kicking around ideas on a poor man's hydraulic press. Fight air pressure with air pressure, contained in some ridiculous steel frame.

    Again, one thing at a time - I lack a way to prototype the molds cheaply, working on that.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  22. #22
    would rather be ruined
    Reputation:
    Join Date
    Jun 2007
    Posts
    185
    Quote Originally Posted by Drew Diller View Post
    Indeed so. I've been kicking around ideas on a poor man's hydraulic press. Fight air pressure with air pressure, contained in some ridiculous steel frame.

    Again, one thing at a time - I lack a way to prototype the molds cheaply, working on that.
    Look at cider press 4 posts with a dental screw you will be surprised how many companies use a screw press

  23. #23
    mtbr member
    Reputation:
    Join Date
    Jul 2008
    Posts
    1,615
    Quote Originally Posted by Drew Diller View Post
    Indeed so. I've been kicking around ideas on a poor man's hydraulic press. Fight air pressure with air pressure, contained in some ridiculous steel frame.

    Again, one thing at a time - I lack a way to prototype the molds cheaply, working on that.
    Hydraulic or a "brake press" would work for one axis. Still need to confine the other 2.
    If you found a way to standardize the size of the mold, say 12x12x12, you could invest once in a solid "mold holder box" made out of steel plate and shapes.

  24. #24
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Quote Originally Posted by bme107 View Post
    Hydraulic or a "brake press" would work for one axis. Still need to confine the other 2.
    If you found a way to standardize the size of the mold, say 12x12x12, you could invest once in a solid "mold holder box" made out of steel plate and shapes.
    Agreed on both counts - a square foot buys me a lot of real estate.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  25. #25
    mtbr member
    Reputation:
    Join Date
    Jul 2008
    Posts
    1,615
    You could build a cubic foot holder and always block some of the interior off with solid blanks if the part/mold didn't require that volume.
    Size of the BB/DT/ST/CS lug would govern but drop outs would need far less space.

  26. #26
    Ene
    Ene is offline
    mtbr member
    Reputation:
    Join Date
    May 2014
    Posts
    31
    Quote Originally Posted by bme107 View Post
    Mortar or concrete found in a ready mix bag at your home center ranges from 3000-5000psi compressive strength. Plenty to resist the parting pressure but you'd need to bind the halves with some A36 (or better) steel shapes and high-strength bolting. Unistrut is not the answer as you found.
    You might need some sort of interior metal sleeving as I'm not sure you'd get repeatable results out of the mortar alone. I think the sharp corners and creases would break free because they are thin and brittle.
    Would it not be possible to make the mold in fiberglas and the concrete as support?

    And great work Drew!

  27. #27
    mtbr member
    Reputation:
    Join Date
    Jul 2008
    Posts
    1,615
    Quote Originally Posted by Ene View Post
    Would it not be possible to make the mold in fiberglas and the concrete as support?
    Mortar and concrete are really only good in compression. Tension strength is roughly 10% of compression. They could be the outer supporting material but they'd either need to be reinforced with rebar or their mass (size) increased to the point that their self-weight would hold the 2 halves of the mold together. It'd be a much more efficient use of materials if they were confined by something with better tensile strength to clamp the mold halves together.

  28. #28
    mtbr member
    Reputation:
    Join Date
    Jun 2014
    Posts
    1
    Hi Drew!
    I'm very fascinated by what you are able to do.
    I am trying to find the way to build my own bladders.
    Can you help me?
    I was thinking to a system like spry or liquid latex to be sprayed or brush into the mold.
    I am using latex tubes but I have also to put the silicone intensifiers in the joint areas such as head or seat tubes and bottom bracket.
    What's your opinion?

  29. #29
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Hey man, I'm going to answer your question, I have to draw a picture so it'll be a little bit.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  30. #30
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Quote Originally Posted by Steve Ash View Post
    Hi Drew!
    I'm very fascinated by what you are able to do.
    I am trying to find the way to build my own bladders.
    Can you help me?
    I was thinking to a system like spry or liquid latex to be sprayed or brush into the mold.
    I am using latex tubes but I have also to put the silicone intensifiers in the joint areas such as head or seat tubes and bottom bracket.
    What's your opinion?
    You already have the right idea, but the keyword "the" mold, singular, is where you're wrong in practice. I had the same idea, I was wrong too.

    Molds. You need two patterns. Reason being, you can't line the entire female mold with a gel that will become a semi-solid, and expect it to be a useful bladder. The exterior of the resulting bladder is going to be covered in fiber, which will then be a few millimeters too voluminous to fit in the mold.

    Further, my opinion is that you need a temporarily solid agent inside the bladder, or at least a very thick / stiff bladder with a touch of air pressure in it during fiber layup.

    I tried "lay flat" silicone bladders that mimicked what I saw with welded seam polyethylene film bladders. Three standoffs (or a perfectly level surface to start with), a shallow basin to allow unvulcanized silicone to pool in, wait about an hour for first layer to vulcanize part way, then add a parchment paper pattern that represents your air chamber, followed by a second layer of silicone. Despite the first layer already being partially vulcanized, the two layers will adhere very well to each other.

    You get something like this.

    There are some drawbacks to this simple approach.

    First, you waste some expensive resin depending on how intricate your basin is - if it's a big square like I did, you'll waste quite a few dollars worth. If it closely matches the air chamber pattern, that'd be less wasteful.

    Second, the seam of the two layers means that the air chamber has to double over itself along the seam when inflated against a female mold cavity.

    Like this, the X marks represent cross linking between the two silicone layers, on the right is a cross section of the lay flat bladder with no air pressure, on the left is with air pressure.

    Show and tell on a... Monday? Milestone: carbon lugs.-flat_bladder_problem.jpg

    What happened with the bladder used in the part pictured earlier in this thread is that it withstood pressure for many hours, but the neck eventually burst in the way I just outlined.

    SO...

    So that means "3D" bladders are the correct choice, because they don't introduce little stress risers or tiny areas where a stretchy material has to stretch way too much. My rule of thumb for silicone rubber is no more than doubling the undisturbed length, anywhere. 1.5x would be better.

    Essentially the challenge with making 3D bladders is gravity and viscosity. You want unvulcanized viscosity that isn't so thick that you get a lumpy bladder, but when using a resin that thin, it's just going to spill off a male pattern or make a big pool inside a female pattern.

    The answer is to spin the pattern gently during vulcanization. Got a translucent gallon of milk handy that is mostly gone? Turn it end over end and watch the milk coat the interior surface, right. What if that wasn't milk, and was rather some liquid slowly solidifying? That is "rotocasting" in a nutshell.

    Here's a proof of concept I made with a Gatorade bottle, I chose that shape because of its complex features (which didn't turn out well at the bottom, I just wanted to know) and lack of similarity to a bike part, as in, I wouldn't get attached to it.

    https://www.youtube.com/watch?v=-s2bPytiZ2c
    https://www.youtube.com/watch?v=WdHNn7jhz9A

    Thing is, you don't *need* a female pattern to rotocast. You could use a male pattern. You'll just lose a little bit of resin onto the floor, you'll have to experiment with proper rotation speed and amount of liquid applied to the outside of a male pattern to minimize waste.

    Important caveat: latex won't work well inside a female pattern. It has to evaporate the solvents that are keeping it in liquid form. Silicone, on the other hand, is a thermoset, doesn't need to breathe, so it can usefully vulcanize inside a closed environment.

    A male pattern with a bladder painted on slowly would be handy in the sense that you already have a firm core around which to apply manual pressure when doing the fiber layup over the bladder. You just need to figure out how to get that core out later, in order to remove the bladder at the end. There are a few options for this, I have not settled on one.

    My first rotocaster that I built with lumber, some bike parts, a miter gear, and a $6 freezer ice auger from Poland (lol) ended up dismantled because it was a piece of sh!t. I'm going go design version 2 so that the two rotating planes will have independently adjustable speeds.

    This help at all? I feel like I've given options and no concrete answers, which is useful or totally useless depending...

    EDIT - one other note on latex vs silicone... if you use too much material, latex will be slow to cure because it needs to be thin for each layer of the stuff (you'll need several or many to build up thickness), pretty much like working with latex paint. If you glob it on, you'll get an uncured squishy glob that resembles a zit that is ready to pop. If you use too much silicone inside a female mold, remember that it is a thermoset, and too much mass means exothermic acceleration, which means you'll get a thick blob of silicone somewhere, and the heft of it will mess up the rest of the casting. Learned both of those the hard way.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  31. #31
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Cripes, man.

    Almost two years since I updated this thread.

    What the @#$* is up with that?!

    Answer: struggle and repetition like you suggested. In the last year, I'd been having big big troubles with my sustainable molding concept. I finally have worked out the majority of the bugs. Unseen behind this shiny (unpolished, btw) lug are two Cub Foods bags of crappy, half infused, hole filled, bladder collapsed, seam showing bunch of rejects.

    Each failure was a learned thing. I learned SO much, most of it written down, some of it in my new muscle memories.

    The biggest thing I ran into with the use of prototyping wax as a mold material is the drastic deformation shown when subjected to heat. ...Even a little bit of heat created a lot of warping.

    I had to switch to Resin Infusion of some kind. Room temp, yo.

    It just happens to be much, MUCH more difficult to achieve a good result than working with prepreg carbon. The fiber is much more prickly when dry and unbound by anything.

    I was becoming paranoid that, after failing to make the prototype bicycle frame I'd hoped for by December, I was feeling that maybe I wouldn't so much as make ANY good infused parts in all of 2014 and 2015 combined.

    With just a shred of time to spare, I proved myself wrong.

    In other words, I did a LOT of the "do it again" practice. I now feel strong. Part failures are so much easier to react to objectively now.

    Show and tell on a... Monday? Milestone: carbon lugs.-20151230_115139.jpg
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  32. #32
    mtbr member
    Reputation:
    Join Date
    Dec 2014
    Posts
    198
    There are "casting resins" that are used in the auto and aviation industry that OUT PERFORM even cnc's AL molds.
    AL has a nasty habit of thermal expansion. Can be quite a bit.

    I can't believe more people in the bike and guitar industry don't use the stuff???

    For making lugs, I ALWAYS use mandrels (solid) with the female molds.
    As you are finding; wet lay-up and bladder molding don't get along too well.
    Sure it CAN be done, but why?

    Waxed and PVA coated hardwood mandrels are used by mold makers for dozens if not hundreds of parts.

    Round stock AL also works well for mandrel material. (can't do complex radii)

    You can still remove consistent radius "tubes" from the mold.
    That gives you accurate ID along with OD. You end up with (some) flange from pressing the female halves together.

    But, carry on any way that works for you.

    Resin Services/Reklin Plastics is a good source for commercial grade casting resins.
    You can get room temp cure, or high heat stuff.
    They have other material added for strength and desired use.

    It is used for complex plastic castings as well cause some parts can't be machined.

    Nice looking lugs!

    JM

  33. #33
    mtbr member
    Reputation:
    Join Date
    Dec 2014
    Posts
    198

  34. #34
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Quote Originally Posted by JohnnyMagic View Post
    As you are finding; wet lay-up and bladder molding don't get along too well.
    Sure it CAN be done, but why?
    Well, I'm doing dry layup, but it's still very unfriendly compared to prepreg.

    I'll attempt to answer why, the big thing is I started with bamboo bikes and realized pretty quickly that jigging up some rare, expensive bamboo only to mess up a wrap-over type reinforcement because of some vacuum problem... is a bunch of crap.

    I needed Separation Of Concerns in more ways than one. I didn't want to subject the bamboo to extreme heat (outside of the bamboo hardening process itself), I instead wanted to do secondary bonding.

    I have an odd background with injection molding of plastic, so I decided to see how much the skills transferred. A fair amount. I will probably end up with a hybrid approach of part internal mandrel, part bladder inside a female mold pair, because as you said, known ID and OD.

    I also want to be all What You See Is What You Get regarding customer previews, and female molds with their defined outer surface and everything, they really push a bunch of my hot buttons.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  35. #35
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    I've never before achieved two-in-a-row with some of my conditions validity checks removed - for instance, no dry pressure two hour run to try and exhaust the bladders before wasting any epoxy. Just infused "blind".

    They're within a few grams of each other. Using flash on my camera revealed some places where I had flow media too close to the surface (the red fibers). You can also see white tracer material where tiny surface air bubbles remain. The idea is to not know the tracer is even there, but this is pretty close to good enough. (But it is not good enough, just... feelin' good)

    Show and tell on a... Monday? Milestone: carbon lugs.-20160101_105901.jpg

    It's an odd bit of tech. It doesn't work unless conditions are all Goldilocks just right. And when that balance is achieved, it looks as though it will work every time.

    I'm pretty happy that I can now get a part to cure overnight at room temp. Get all the air machines shut off and unplugged before sleep, wake up the next morning with a snazzy lug in hand. I can live with that, yeah.

    Happy new year! I am friggin' PUMPED!
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  36. #36
    mtbr member
    Reputation:
    Join Date
    Aug 2015
    Posts
    52
    I'm amazed that you managed to make this complex shape as a single part!

  37. #37
    mtbr member
    Reputation: Eric Malcolm's Avatar
    Join Date
    Dec 2011
    Posts
    1,014
    Drew

    You mention having started with Bamboo bikes. The purpose of the lug making is to create a Bamboo tubed/carbon lugged frame? Where are you intending to go with this project?

    I'm curious, but impressed with the tenacity of this project. It looks very good and a credit so far.

    Eric
    If I don't make an attempt, how will I know if it will work?

  38. #38
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Quote Originally Posted by Eric Malcolm View Post
    Drew

    You mention having started with Bamboo bikes. The purpose of the lug making is to create a Bamboo tubed/carbon lugged frame? Where are you intending to go with this project?

    I'm curious, but impressed with the tenacity of this project. It looks very good and a credit so far.

    Eric
    Thank you for the encouragement! I was hoping to impress, this niche has been kicking my ass for too long.

    Improving my bamboo frame building craft was the original goal. I wanted the bamboo to undergo particular heating phases that did not jive at all with prepreg carbon requirements. Maybe if I was Boo and had access to the highest quality bamboo on the planet, then it'd be a different story. But I can only get what I can get in the lower 48.

    Once I started getting closer to high quality carbon, the thought crept in "I could use this for a lot more than bamboo bikes", so the motivations morphed.

    The bamboo stuff isn't going away. It's just that I've figured out potential bamboo bike customers don't need to be sold on the idea, they either love it or are repulsed by it... so, it will no longer be the sole game in town. I'm very interested in suspension and gearbox stuff, and bringing the cost of the latter down to Mere Mortal levels.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  39. #39
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Ever watch a baby deer slip around while learning to walk? Like holding themselves up reliably is more of a parlor trick than anything else.

    That's how developing my understanding of composite processing has felt. Accomplish something, think you're cool, try to repeat the process and it gut punches you and takes your lunch and steals your girlfriend.

    I often remind myself I no longer have a commute and to STFU and get back into the kitchen, so to speak. Hard knocks in the shop, yeah so what, SHUT UP and do it again. Shut up and repeat. Repeat. Think you're cool? SHUT UP, DO IT AGAIN, AGAIN, AGAIN. Ari Vatanen didn't learn to shift in a day.

    I've hit seven infusion processes in a row, with repeatability that is finally on the cusp of journeyman / semi pro area. I'm making molds four times faster than when I started learning how to do milling. What has felt like massive turbo lag for a long time now feels like it is starting to spool. Habits and muscle memory are coming along, while the daily bumps and hiccups are becoming fewer and fewer.

    All this hard work is starting to bear fruit. Feels good man. Feels good.

    Resin content is somewhere in the 41% to 43% range (still honing the math stuff). If I really want to compete with the big boys weight wise, I need to hit 38%, which... isn't going to happen soon. I'm prioritizing bullish toughness over light weight. I tend to break stuff, which is really dumb because I barely push a buck seventy even while stuffing my face with pizza constantly.

    So, it's time to take this tech and make my seventh frame / first one like this approach. It's time to implement. Gonna make a thing that's like a Zerode G1, even though it's going to be a pedal watt sucking monster. My mission is to liberate the rear axle of any barnacles. No more derailleurs for this guy. The cheapest most available way to go forward with that is to suck up the efficiency penalty and cobble and Alfine into the beer belly of a weird looking bike frame.

    Show and tell on a... Monday? Milestone: carbon lugs.-screenshot-2016-03-12-02-28-27.jpgShow and tell on a... Monday? Milestone: carbon lugs.-screenshot-2016-03-12-02-27-24.jpgShow and tell on a... Monday? Milestone: carbon lugs.-screenshot-2016-03-12-02-25-52.jpg
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  40. #40
    mtbr member
    Reputation: Eric Malcolm's Avatar
    Join Date
    Dec 2011
    Posts
    1,014
    So....You're finally going to make THAT bike!!!!


    Cool.

    Eric
    If I don't make an attempt, how will I know if it will work?

  41. #41
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    HELL YES

    I've been wanting this since I first trashed a rear derailleur on the first mountain bike I bought with my own money. I worked at a bike shop when I was 15 and learned that I was a terrible mechanic, they fired me in a month. Was just enough to afford a Schwinn Moab 3, which I used as an escape mechanism. And not very far into its life, I had a suddenly lame bike out in the woods.

    And now it's 2016. The bicycle market is NOT flooded with gearbox bikes, and that really pisses me off so I'm gonna do anything I can to change it.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  42. #42
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    View from a "side B" mold that has a featureless parting face. The carbon surface has some flaws evidently created by the CNC's chassis wobble, which isn't helped by the primitive NC code that I generated.

    Worse than that is the flaw at right where the edges are sharp. That sort of part feature will reliably kill inflation bladders *after* the epoxy curing is done. An unceremonious death. I need to improve my technique here.
    Show and tell on a... Monday? Milestone: carbon lugs.-20160324-dscf5336.jpg

    Here is a shot of the interior when things go well. I can see your underwear! About halfway up, that's where the mold parting face would be. Note that you cannot see a hint of it. I used to have such trouble with the result of this is supposed to be a closed shell piece, and all I've done is conjured an unnecessarily complex way to make two disjointed half shells. Not anymore!

    Note from the white particles that I need to bolster my shop cleanliness. Those are just particles I missed when cleaning the bladder prior to layup.
    Show and tell on a... Monday? Milestone: carbon lugs.-20160324-img_5730.jpg
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  43. #43
    mtbr member
    Reputation: bluechair84's Avatar
    Join Date
    Sep 2008
    Posts
    180
    Mega impressed here!
    I'm currently working on a carbon garage project as well. I'm not at the stage of experimenting with carbon yet - this week I've gotten most of a main triangle mandrel made. I'm just polishing it up this weekend to start making a female mould from it.
    It's been interesting reading your posts on the mould. My plan is two female halves in fibreglass and I'm expecting a bladder pressure of about 14psi. Have you got any advice on mould building or some reading I can do before I begin this venture?
    Cheers

  44. #44
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Quote Originally Posted by bluechair84 View Post
    Mega impressed here!
    I'm currently working on a carbon garage project as well. I'm not at the stage of experimenting with carbon yet - this week I've gotten most of a main triangle mandrel made. I'm just polishing it up this weekend to start making a female mould from it.
    It's been interesting reading your posts on the mould. My plan is two female halves in fibreglass and I'm expecting a bladder pressure of about 14psi. Have you got any advice on mould building or some reading I can do before I begin this venture?
    Cheers
    Regarding mold pressure: I'm using a bit more than 14 psi (but still a lot less pressure than in a road bike tire), but I've been told that it is possible to do closed molding with just 14 psi. After all, if it is possible to do open shell vacuum bagging with a natural limit of a touch over 14 psi, then it follows that you could feasibly do it with a bladder.

    It's just, I had trouble getting it to work. I tend to use a little too much temporary tack adhesive. One or two little puffs covers a surprising surface area. If the plies are stuck together too firmly, they won't want to slide along each other during compaction. If the plies are not stuck together at all, it's absurdly difficult to get the preform into the mold... like bathing a cat, or trying to put an angry baby squid into a jar.

    Also on mold pressure: check your math regarding clamping forces. Get an idea of the surface area of the empty parting face of one mold, get it in square inches and multiply that by what your psi is. You'll probably be like what the heck, that much force?!from that little pressure, yikes.

    These are good clamps. Note the compression rating. I use a backing plate to avoid killing my molds, and I fasten each clamp in an iterative way... like when you put a wheel onto a car, you don't just tighten one lug nut fully and immediately.

    Regarding making the molds, I am coming at this from a machinist perspective, which is possible when you have a CNC handy. Whereas you are taking an inverted copy of a male plug. I don't have any useful advice on what specific tooling epoxies to use, just some general advice about surface tension. Eddy currents happen the world's rivers and they happen here too. Walls, hard transitions, pockets - expect some possible crappy mold results in those zones first. I've read that some tooling epoxies are "no shrink", but everything shrinks a little bit.

    Locating pins for the A and B sides should be placed as far apart as you can without damaging the trueness of the dowels. It's weird but if you crack open one side of the mold only, and try to lift fully from there, you'll be bending the @#$* out of either the dowel or the dowel hole at the other end. You have to scoot the mold up... a little up over here, get something to hold it in place, a little up on the other side, get something to hold it, repeat.

    I dunno if that was helpful, if you have any specific mold making questions do go ahead and ask.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  45. #45
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679

    C is for Compromise

    I did a dry test fitting for the first time. I made the straight sections recently, using a caul plate design. Basically the tubes are just half tube profiles, they're going to be secondary bonded, and later overwrapped after... well, lots of secondary (third, fourth, etc) bonding is going to happen.

    It's not ideal, but it moves the ball forward for me. I lack a BUNCH of composite-specific equipment for machining, and I want to move at high speed, so there's a little rock and a hard place thing going on there. So, for now, right, use a caul plate so that the interior tube is known, the exterior tube is known, the parting plane is known (friendly to bond flat), I can use the frame components themselves to aid in alignment.

    The slip / "shunk" noise that the last socket made was viscerally pleasing, like letting go of a bowling ball and your arm KNOWS that something cool is going to happen, even though you can't really touch the ball anymore you still WATCH the damn thing go.

    Note the vertical seat tube socket (short stay design) and pronounced belly (gotta make room for an IGH to fit where a normal down tube would be).

    I find the caul plate thing a little cumbersome in terms of effort per yield, but WOO the interior and exterior results are convincing, the resin ratio is convincing. Tedious but works firmly.

    Show and tell on a... Monday? Milestone: carbon lugs.-geo_compare.jpg
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  46. #46
    mtbr member
    Reputation: bluechair84's Avatar
    Join Date
    Sep 2008
    Posts
    180
    Quote Originally Posted by Drew Diller View Post
    After all, if it is possible to do open shell vacuum bagging with a natural limit of a touch over 14 psi, then it follows that you could feasibly do it with a bladder.

    It's just, I had trouble getting it to work. I tend to use a little too much temporary tack adhesive. One or two little puffs covers a surprising surface area. If the plies are stuck together too firmly, they won't want to slide along each other during compaction. If the plies are not stuck together at all, it's absurdly difficult to get the preform into the mold... like bathing a cat, or trying to put an angry baby squid into a jar.

    Also on mold pressure: check your math regarding clamping forces. Get an idea of the surface area of the empty parting face of one mold, get it in square inches and multiply that by what your psi is. You'll probably be like what the heck, that much force?!from that little pressure, yikes.

    Locating pins for the A and B sides should be placed as far apart as you can without damaging the trueness of the dowels. It's weird but if you crack open one side of the mold only, and try to lift fully from there, you'll be bending the @#$* out of either the dowel or the dowel hole at the other end. You have to scoot the mold up... a little up over here, get something to hold it in place, a little up on the other side, get something to hold it, repeat.
    We're definitely on the same page regarding a few things - I'd also taken 14psi as the (atmospheric) pressure used in compressing open moulds. The only problem with this is they are often cured in autoclaves which raise the pressure further I think. I've seen other no-clave cures and vac-bagging so I'm pretty confident it'll be reet.
    I hadn't considered the need for the sheets to slide along each other during compaction - I can see why this might be relevant for the changing circumference of curvatures for each layer. It's something I'll look at in the first trials!
    Because I'm not currently planning on a preform (I will be wrapping some Alu mandrels for the bearing faces) I shouldn't have your 'baby squid' problem. 95% of my fibre is going straight into the female mould and the bladder will be the only internal shaping force.
    As for the dowels, I might have a solution. I'm planning on running a 'bead' around the face of one mould and have a concurrent groove in the other mould. This will mean that there is very little height to clear when separating the moulds, but there will be a full seating guide all the way around.

    When you are combining your halves, are you using an internal flange on one side that is pressed against the other half via the bladder? Or are you wrapping the two halves externally?
    You've given me some confidence to share what I'm doing, so I'm going to revist an old thread and put up some details of jig and blank. Keep your eye out

  47. #47
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Quote Originally Posted by bluechair84 View Post
    We're definitely on the same page regarding a few things - I'd also taken 14psi as the (atmospheric) pressure used in compressing open moulds. The only problem with this is they are often cured in autoclaves which raise the pressure further I think. I've seen other no-clave cures and vac-bagging so I'm pretty confident it'll be reet.
    I hadn't considered the need for the sheets to slide along each other during compaction - I can see why this might be relevant for the changing circumference of curvatures for each layer. It's something I'll look at in the first trials!
    Because I'm not currently planning on a preform (I will be wrapping some Alu mandrels for the bearing faces) I shouldn't have your 'baby squid' problem. 95% of my fibre is going straight into the female mould and the bladder will be the only internal shaping force.
    As for the dowels, I might have a solution. I'm planning on running a 'bead' around the face of one mould and have a concurrent groove in the other mould. This will mean that there is very little height to clear when separating the moulds, but there will be a full seating guide all the way around.

    When you are combining your halves, are you using an internal flange on one side that is pressed against the other half via the bladder? Or are you wrapping the two halves externally?
    You've given me some confidence to share what I'm doing, so I'm going to revist an old thread and put up some details of jig and blank. Keep your eye out
    Gotcha, I misunderstood a few things. I think you have a realistic course ahead of you. I was getting confused where you were using "bladder" and I think you meant "vac film". I'm splitting hairs of course, both of them act as flexible-membrane-soft-clamps. Just one type is a balloon, the other type relies on a sealing flange.

    My halves could be called half/full (different than half-full). Part is laid in the mold in halves. Part is laid on the bladder as a complete unit, but "complete" is still a relative term here since we are talking about plies of overlapping fabric.

    That said, a mating flange is a legit way to do it as long as there is some overwrap. 3M secondary adhesives are really something, but even then they are not really designed for peel strength. The caul plate tubes I made just the other day are mated in this same flange-y kinda way, and will be over-wrapped at the seam later.

    Regarding autoclaves, yeah it is basically for the ease of use with vacuum bagging, but goes well beyond the atmospheric pressure limitation that otherwise goes with vac bagging. For your purposes, I don't think you need it. Does your first part need to be perfect 38% resin? Nahhhh. If you're trying to make some world record road bike --- it's been done. Make now with whatcha got, autoclave later.

    I like your groove idea for the registration.

    Have you purchased a vacuum pump yet?
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  48. #48
    mtbr member
    Reputation: bluechair84's Avatar
    Join Date
    Sep 2008
    Posts
    180
    I won't need a vacuum pump as I'm using a pressurised bladder inside the tubes. The bladder will likely be left crusoe'd inside the frame but, I'm not bothered about saving a few grams to get it back. The process I'm planning on is to lay up mould A to the edges, lay up mould B to have a few inches of overlap, place the inflation bladder into mould B beneath the flanges and fold them down onto the bladder. As the bladder expands it'll press the flanges onto mould A.
    It does leave me with a problem though - how to get the atmospheric air out of the frame as the bladder pressurises. If I don't remove it I run the risk of dissolving it into the epoxy. It's a big advantage your vac bagging has!

  49. #49
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Ahhh, ok, I misunderstood twice! Gotcha. You're worried about that trapped air, give one mold face some cardio lines. I don't know how well this applies to close mold wet layup, but once piece of advice I read was "if the epoxy doesn't move, the air bubbles won't move". Some feed lines to the outside world will give a path of least resistance when you inflate the bladder. They don't have to be deep channels, because air pressure finds a way.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  50. #50
    mtbr member
    Reputation: bluechair84's Avatar
    Join Date
    Sep 2008
    Posts
    180
    Hey that's not a bad idea at all... You know, I was going to run some full length cable guides inside the frame. They can be multi-freakin-purpose! If they are porous, the air can escape through the internal cable inlets. That's bloody genius.

  51. #51
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    That thing I was saying, about caul plates being accurate...

    Caul plates are accurate.

    Show and tell on a... Monday? Milestone: carbon lugs.-20160410_015825.jpgShow and tell on a... Monday? Milestone: carbon lugs.-20160410_020004.jpgShow and tell on a... Monday? Milestone: carbon lugs.-20160410_015711.jpg
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  52. #52
    mtbr member
    Reputation: bluechair84's Avatar
    Join Date
    Sep 2008
    Posts
    180
    They're fantastic, nicely done. I'm looking forward to see how you're going to bond them!

  53. #53
    mtbr member
    Reputation: Velobike's Avatar
    Join Date
    Jun 2007
    Posts
    6,551
    Quote Originally Posted by bluechair84 View Post
    They're fantastic, nicely done. I'm looking forward to see how you're going to bond them!
    I'm watching with interest too.
    As little bike as possible, as silent as possible.
    Latitude: 57º36' Highlands, Scotland

  54. #54
    mtbr member
    Reputation:
    Join Date
    Apr 2016
    Posts
    4
    Quote Originally Posted by bluechair84 View Post
    Hey that's not a bad idea at all... You know, I was going to run some full length cable guides inside the frame. They can be multi-freakin-purpose! If they are porous, the air can escape through the internal cable inlets. That's bloody genius.
    bluechair - If you are only planning on 14PSI (same as 28 inches of Vacuum) I think you have to seriously consider using a vacuum bag. If you are making a mold that is able to take a bladder then you will be able to use that mold with a vacuum bag. The added advantage will be to use a breather cloth and padding backing in order to suck out all bubbles.

    To do this you would make the bag part that goes inside the closed two piece mold similar to the shape of the bladder you are expecting to use. Then there is the outside of the mold to worry about. Make another part of the bag that fits the whole mold. The only issue is that you have to connect these two pieces with a length (through your pressure opening) that will be able to close it all to take the single line vacuum. So the vacuum will be pulling on the inside of the mold as it pulls on the outside of the mold. Imagine it appears as a bag inside a bag. Without the mold halves in there it would look like an empty balloon if vacuumed down. As mentioned before make sure you have no sharp edges.

    If you are not interested in buying a vacuum pump consider getting a Venturi Vacuum Pump from Harbor Freight for about $20 (Vacuum Pump - AC Vacuum Pump w/ R134A & R12 Connectors).

    Later,
    Dan

  55. #55
    mtbr member
    Reputation: bluechair84's Avatar
    Join Date
    Sep 2008
    Posts
    180
    Thanks, I understand the process. thinking about it the benefit is that my clamshell mould isn't having to hold in 14psi. I do need to leave access to the in and out of the finished frame to remove the the vac-bag when actually I want to create a complete, almost fully closed frame. The theory is that without large holes needed to remove the vac-bag, or a 'monocoque' design, the frame will be stronger.
    To do the vac-bag technique I'd need to leave a hole... possibly inside the headtube... I'll consider it for sure.

  56. #56
    mtbr member
    Reputation:
    Join Date
    Apr 2016
    Posts
    4
    Quote Originally Posted by Drew Diller View Post
    Thank you for the encouragement! I was hoping to impress, this niche has been kicking my ass for too long.

    Improving my bamboo frame building craft was the original goal.

    Once I started getting closer to high quality carbon, the thought crept in "I could use this for a lot more than bamboo bikes", so the motivations morphed.
    Drew Diller: I have been on a similar path to what you are doing for quite a while. I started a long time ago with latex bladders that did not last past a one time thing that took a while to make each bladder. I then started making plastic film bladders using a soldering iron to seam them...works great. I have been successful making parts with pressures up to 60 PSI. Also failed with high area parts using 10 PSI.

    I would encourage you to take a look at Adam Pequette's video on Inflation Bladder Molding ()

    Later,
    Dan

  57. #57
    mtbr member
    Reputation:
    Join Date
    Apr 2016
    Posts
    4
    Quote Originally Posted by bluechair84 View Post
    To do the vac-bag technique I'd need to leave a hole... possibly inside the headtube... I'll consider it for sure.
    Even if you used only the inflation bladder you would want to get it out of the inside somehow. The inflation connection width may be big enough to dig out the bladder.

  58. #58
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    dandelave: you're spot on regarding the positive-air-source vac pumps, I love them. Solid state, zero maintenance, no bad smells or bad compounds in the air other than what is produced by your compressor.

    The near-full-vac one really gulps down air, though.

    Show and tell on a... Monday? Milestone: carbon lugs.-20160412_145235.jpg

    Regarding the seam film bladders, I believe they work and I'm glad you go them to do so, but I always had one tiny spot that would hinder an otherwise good bladder. I had previously followed exactly with Adam's techniques. If you can get the seam strong enough, I have no doubts that it works.

    The reason I went with 3D bladders is that I knew I wanted deep V and U type shapes. The down tube I built the other day is 56mm viewed from the top, 38mm from the side. I knew I wanted something that would take a near vertical angle straight away from the parting line, which is why I turned away from the seam bladder approach.

    EDIT: I would even say that the film bladders are better in some respects, for instance they have lower friction under pressure. Whereas my silicone bladders tend to be high friction under pressure, a ply can get dragged elsewhere under pressure if I do the layup poorly, which can lead to cuts and eventual rupture.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  59. #59
    mtbr member
    Reputation:
    Join Date
    Apr 2016
    Posts
    4
    Quote Originally Posted by Drew Diller View Post
    dandelave: you're spot on regarding the positive-air-source vac pumps, I love them. Solid state, zero maintenance, no bad smells or bad compounds in the air other than what is produced by your compressor.

    The near-full-vac one really gulps down air, though.

    Regarding the seam film bladders, I believe they work and I'm glad you go them to do so, but I always had one tiny spot that would hinder an otherwise good bladder. I had previously followed exactly with Adam's techniques. If you can get the seam strong enough, I have no doubts that it works.

    The reason I went with 3D bladders is that I knew I wanted deep V and U type shapes. The down tube I built the other day is 56mm viewed from the top, 38mm from the side. I knew I wanted something that would take a near vertical angle straight away from the parting line, which is why I turned away from the seam bladder approach.

    EDIT: I would even say that the film bladders are better in some respects, for instance they have lower friction under pressure. Whereas my silicone bladders tend to be high friction under pressure, a ply can get dragged elsewhere under pressure if I do the layup poorly, which can lead to cuts and eventual rupture.
    Drew:

    They do take a fair amount of air compressor time but they are cheap.

    Sorry to hear your luck with the film badder technique. I would only occasionally have issues with them, and that was always spelled out when I tested them, before putting them in the mold.

    Lately I have really liked oogoo for making bladders, easy cheap, strong, holds a pretty good shape when not inflated, and easily repairable.

    Later,
    Dan

  60. #60
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Quote Originally Posted by dandelave View Post
    Lately I have really liked oogoo for making bladders, easy cheap, strong, holds a pretty good shape when not inflated, and easily repairable.
    Clever, I'll have to try it.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  61. #61
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    I'm no longer embarrassed by sharing failed parts. This time, I used too much temporary tack adhesive during dry layup. Doing so creates a condition where the plies have too much shear friction for the bladder to break apart and flatten. The fiber bundle doesn't quite make it to the mold cavity surface all the way - sometimes by 2mm or more - and the epoxy is given no incentive whatsoever to continue infusing. It just fills the empty space, path of least resistance.

    On the upside, I did achieve the possibility of daisy-chained molds. I can now make parts that are twice as long as the cutting envelope of my CNC.

    I'm really kicking myself over using too much tack, I've done this before, absolutely a repeat failure. It's a skill, usually I nail it, sometimes not. Much to learn.

    Show and tell on a... Monday? Milestone: carbon lugs.-%5Bmp%5D-things-got-soft.jpg
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  62. #62
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    I made some superior pressure bladders, hooray, and got some decent handlebars as a result.

    Note I say decent and not good or great or even commercially viable. It's a decent, but ultimately scrap, part.

    So I cut it into many pieces and looked at the insides. It was mostly really good, with two symmetrical bad parts. The bad parts were, happily at least, free of any air, but they were bogged in excess epoxy that is just sort of... there, not doing anything.

    The aspect of this part that I'm really happy about is the minor Z flashing at the middle. This was made with four molds, like A1-A2 meets B1-B2, all hydraulically linked. The theory works, which means I can now legit make parts that are larger than the size constraints of my CNC mill. While this may be a scrap part overall, what is represents is huge: this paves the way to true monocoque front triangles for full suspension frames, because I could feasibly daisy chain more than two molds in this same fashion. It could be N number of molds, it just comes down to how much fancy BS am I willing to do / how motivating is the part in question.

    Show and tell on a... Monday? Milestone: carbon lugs.-2016-05-31-14-57-32.jpgShow and tell on a... Monday? Milestone: carbon lugs.-2016-05-31-14-59-12.jpg
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  63. #63
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Forgot to note, the thicker flashing along the length of the bar, currently it is there as a crutch. Initially, I thought it might be some competitive advantage because it decreases overall infusion times, but... it's actually not a good thing. The reason they persist for me at the moment is lingering shortcomings in my layup technique - the channel gives an emergency exit to little wild hair fibers. If you DON'T account for these loose fibers, they mess with the hydraulic seal of the mold. They are very hard to keep track of, and until I am able to track them all, I need the training wheels mode with the seam relief channels.

    Also, the flashing left behind is extra work to clean up. It'll be better once improved technique nearly eliminates one task.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  64. #64
    mtbr member
    Reputation: Velobike's Avatar
    Join Date
    Jun 2007
    Posts
    6,551
    I can see some nice tapered carbon curved fork legs that fit into a lugged crown in the future...
    As little bike as possible, as silent as possible.
    Latitude: 57º36' Highlands, Scotland

  65. #65
    mtbr member
    Reputation:
    Join Date
    Aug 2015
    Posts
    52
    Congratulations on the progress! Did you determine the fiber volume content that you achieved?

  66. #66
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Quote Originally Posted by Daniel Thomas View Post
    Congratulations on the progress! Did you determine the fiber volume content that you achieved?
    On this one, wasn't too concerned since scrap, but the ugly answer before chopping it up was greater than 50%.

    Good infusion of lugs usually gives somewhere around 41%. It may be a reality where I *never* hit 38%, and based on the stuff I've seen so far, I'm not too concerned about that 3%. I'm MUCH more concerned about the bladder mortality rate.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  67. #67
    mtbr member
    Reputation:
    Join Date
    Aug 2015
    Posts
    52
    I'd be fine with consistently getting 41% (resin weight percentage I assume)! It is more important to a have reliable results than saving an ounce of weight at the risk of having dry spots.

  68. #68
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    So Daniel Thomas, I thought about what you said and tried a literal application: I used a very low cost, recyclable, and mildly inaccurate method of internal compression. To elaborate on that last notion, the ply compression is well-enforced in the Z direction of the mold, and poorly compressed in the X and Y directions. The X and Y are expected to take up some "fat" or excess resin pockets.

    I made a bar the other day that turned out heavy, relatively speaking - something north of 47% resin.

    That said, it doesn't feel like a boat anchor. It is light enough for me, sure. Like Walt has correctly said ad nauseum, the weight weenie game is stupid. The way I see it, thanks to the wind turbine industry, there is all this standard modulus carbon fiber available for low cost now. Why not make some affordable and light enough frames? If you want to throw in oodles of the super-high-tech carbon fabrics to chase ultra low weight, then you have to pay for it.

    I want to take the notion of forced elitism out of carbon. I'm sick of the superiority complexes, I just wanna make some weird shapes, all right?

    So I was thinking a bit - for the oddball customer who *doesn't* want a super shiny and well painted expensive result... rather, they just want to test an idea and they are willing to exercise their bicep muscles... is there any way I could provide a sort of "C grade" prototype.

    It would need to check the boxes:

    Light enough?
    Strong enough?
    Reasonably not-ugly enough?

    Show and tell on a... Monday? Milestone: carbon lugs.-20160614_223402.jpg

    This handlebar checks the light enough (well, according to me) and attractive boxes. The real question is, does the mediocre X and Y compression make for a weak part? This remains to be seen, I expect the answer is yes but I hope it is no.

    In this way, I hope I can appeal to more than one kind of customer income bracket, while still making a livable income myself.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  69. #69
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    I tried attaching a metallic logo / clamp anti crush zone thing today.

    Ehhh it went okay for a first attempt. Not bad. There was an air inclusion which messes with the flatness of the logo surface. I went very modest with the secondary bonding adhesive, too conservative. With no abrasion, most of the clamp surface was acceptably close to spec, but one section at the strip meeting point, that was 0.1mm out of round.

    I tried straight filament winding over top twill fabric - I don't really care for it. I'm going to take a tighter pattern with the straight filament instead, make it one smooth continuous thing. I don't much care for the twill fabric finish, it makes me a little dizzy / vertigo sensations. I'd rather chase after a "grayscale mallard plumage" tone.

    330 grams in all at 760mm wide... time to put it on an epoxy diet to some degree, and more importantly fix the flashing problem.

    Show and tell on a... Monday? Milestone: carbon lugs.-20160619_151833.jpg
    Show and tell on a... Monday? Milestone: carbon lugs.-20160619_190208.jpg
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  70. #70
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    July was cruel to me. I thought there were things I was "over", apparently not! Vac and bladder leaks of new types, one of which was cause by the instability of my main work table. Some moments were mean enough that I was contemplating those dark "Am I good enough to hack this? I need to pay some friggin bills" moments.

    August is being slightly more kind so far. I've taken some heavy handed steps at fixing the new leaks, and they're working well.

    That said, I did something stupid this morning. I had a legit handlebar laid up and infused, and I repeated an old mistake by demolding too early. The epoxy was still considered an extremely high viscosity liquid and I didn't realize it. The carbon decided it liked the hydro lock of the mold on either side a lot better than the epoxy wanted to stay in one piece, and I was rewarded with a bad stretching goopy sound instead of the desired cracking popping sounds. I could get the halves out as they were stiff enough once carefully handled, but it was clear that they would still self-stick in a very slightly gooey way.

    It ended up being a good mistake, oddly enough. In this manner, I could see some interior fiber warp, indicating bladder stress. This was location matched to white-opaque (micro tearing) zones on the silicone bladder. So I gotta figure out how to artificially increase wall thickness in certain zones without changing much about part strength or weight, ONLY for the sake of the bladder being asked to do something sane and achievable.

    Thankfully, when it comes to adding artificial thickness for the sake of interior flatness, boat hull builders are already on that like flies on sh!t. My naive past self used to look at those filler materials all "Why the heck would you want to do that?" and here I am again with another "Oh. This is why."
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  71. #71
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Like I said, August has been more kind. Feel like I am touching rubber to the road again. The epoxy I used in this one is harder to work with, but it has a usefully high service temperature. 297 grams pictured with flashing, with just barely enough room to stuff a smidgen of extra carbon in for the next bar.

    First carbon part I've made where I've struggled to make it heavy enough.

    Zero mold damage. After a solid month of fails, feels damn good to have a win.

    Show and tell on a... Monday? Milestone: carbon lugs.-20160809_024638.jpg

    EDIT - forgot a major thing. This is a 39% resin content bar. I did it!! I think I'm going to plump it back up to low 40% because I really really like the look of carbon veil surface compared to the twill look, and that stuff is really epoxy-hungry.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  72. #72
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Had a failed bar, oh well. Repeated an infusion mistake based on some new advice. Turns out it was only half-good advice, so time to head back to the notion of "nothing is true, everything is permitted".

    I used to think one vac gauge was good enough, by the time I'm done here there will be five gauges.

    So I decided to take this most recent bad bar and chop it up. What did complete looks pretty good. It certainly is thick and that's to keep me from having to honor warranty claims. I want it to be where I don't process any warranties because hardly anything comes back to me.

    The light colored ring around the donut hole is a single layer of aramid fiber. Wow, now THAT stuff is a pain in the ass to work with.

    Show and tell on a... Monday? Milestone: carbon lugs.-botched-bar-wall-thickness-near-clamp.jpg
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  73. #73
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Well since no one posted to this subforum over the weekend, here's a scene from the shop:

    Show and tell on a... Monday? Milestone: carbon lugs.-%5Bmp%5D-staring-into-abyss.jpg

    I like to supervise the filling of a box mold (with machinable wax) by witch wanding with a heat gun on full blast. It makes the top surface cool in a relatively smooth way.

    In the future I'll have some snazzy auto-heated boxes to offload this process away from me, but I cannot afford that yet.

    A few more steps after this image, and the resulting blank can be CNC'd.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  74. #74
    would rather be ruined
    Reputation:
    Join Date
    Jun 2007
    Posts
    185
    Quote Originally Posted by Drew Diller View Post
    Had a failed bar, oh well. Repeated an infusion mistake based on some new advice. Turns out it was only half-good advice, so time to head back to the notion of "nothing is true, everything is permitted".

    I used to think one vac gauge was good enough, by the time I'm done here there will be five gauges.

    So I decided to take this most recent bad bar and chop it up. What did complete looks pretty good. It certainly is thick and that's to keep me from having to honor warranty claims. I want it to be where I don't process any warranties because hardly anything comes back to me.

    The light colored ring around the donut hole is a single layer of aramid fiber. Wow, now THAT stuff is a pain in the ass to work with.

    Click image for larger version. 

Name:	botched bar - wall thickness near clamp.jpg 
Views:	131 
Size:	66.0 KB 
ID:	1090677
    dude that thing will take a plane landing,

  75. #75
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Quote Originally Posted by compositepro View Post
    dude that thing will take a plane landing,
    See that's exactly what I'm after. I've been told "you shouldn't sell carbon bits to people who crash a lot".

    I crash. I'm clumsy. I'm inattentive. Sometimes, I become enraged and resemble someone high on PCP. I want carbon bits that will survive the worst of my human faults.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  76. #76
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    I had an idea after looking at my epoxy catch pot too many times. I noticed that the wasted epoxy resin was "neat" or free of gas bubbles. And they call this waste epoxy?!?! How about perfect epoxy that is stuck in a bad place?

    What if I could pause / open up the infusion pathway for a short time, collect wasted resin that has been immaculately degassed after being forced through the fine mesh of carbon fiber, and then put it back in at the beginning of the pathway? Feasibly, I could recirculate resin through the system with very little waste, as long as the pot time of the resin was still alive.

    Seems to work. I'm using an extra slow epoxy on this one, so I can't view the result til Saturday. I'm guessing I recirculated 30 grams or more out of a 250 gram resin payload.

    Basically this custom catch pot that I made is an unexciting shallow tub with five holes in the back for various tubing lines and a gauge. The slanted line with seemingly nothing underneath it is a small pool of epoxy with no visible gas in it.

    Show and tell on a... Monday? Milestone: carbon lugs.-shallow_resin_catch_pot.jpg
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  77. #77
    would rather be ruined
    Reputation:
    Join Date
    Jun 2007
    Posts
    185
    Quote Originally Posted by Drew Diller View Post
    I had an idea after looking at my epoxy catch pot too many times. I noticed that the wasted epoxy resin was "neat" or free of gas bubbles. And they call this waste epoxy?!?! How about perfect epoxy that is stuck in a bad place?

    What if I could pause / open up the infusion pathway for a short time, collect wasted resin that has been immaculately degassed after being forced through the fine mesh of carbon fiber, and then put it back in at the beginning of the pathway? Feasibly, I could recirculate resin through the system with very little waste, as long as the pot time of the resin was still alive.

    Seems to work. I'm using an extra slow epoxy on this one, so I can't view the result til Saturday. I'm guessing I recirculated 30 grams or more out of a 250 gram resin payload.

    Basically this custom catch pot that I made is an unexciting shallow tub with five holes in the back for various tubing lines and a gauge. The slanted line with seemingly nothing underneath it is a small pool of epoxy with no visible gas in it.

    Click image for larger version. 

Name:	shallow_resin_catch_pot.jpg 
Views:	70 
Size:	61.3 KB 
ID:	1094763
    IM NOT SURE WHAT YOU MEAN BY OPEN UP eeek Pretty much once its sealed you want it to stay sealed 30grams of resin is a good sacrifice for a part you know is free of air?

  78. #78
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Quote Originally Posted by compositepro View Post
    IM NOT SURE WHAT YOU MEAN BY OPEN UP eeek Pretty much once its sealed you want it to stay sealed 30grams of resin is a good sacrifice for a part you know is free of air?
    I cheated for this test, see what I get kinda thing. In the future one of the extra tubing lines is only for clamping, so I could isolate the mold from the catch pot for a time, maintaining the vac integrity for the part while relieving the pot in order to drain it.

    Either way, the feed line stays clamped so there's no start point entry of air.

    Not sure what I'm going to get!

  79. #79
    would rather be ruined
    Reputation:
    Join Date
    Jun 2007
    Posts
    185
    Quote Originally Posted by Drew Diller View Post
    I cheated for this test, see what I get kinda thing. In the future one of the extra tubing lines is only for clamping, so I could isolate the mold from the catch pot for a time, maintaining the vac integrity for the part while relieving the pot in order to drain it.

    Either way, the feed line stays clamped so there's no start point entry of air.

    Not sure what I'm going to get!
    roger roger , i read it wrong :-)

  80. #80
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Pretty okay result. I made some human errors during layup (again). The inlet side showed better infusion than the outlet, which suggests I might want to just recirculate longer.

    The sharpie points to said human error, basically there is a limit to how far the bladder is allowed to expand, on purpose. I didn't fill it out properly, capillary action was missing. Consequently, the fiber volume fraction is not what I wanted, a rather unexciting 50/50.

    Gettin' closer. Time to try (yet) again and concentrate on my weak zones.

    Show and tell on a... Monday? Milestone: carbon lugs.-20160917_195559.jpgShow and tell on a... Monday? Milestone: carbon lugs.-20160917_195602.jpgShow and tell on a... Monday? Milestone: carbon lugs.-20160917_202424.jpgShow and tell on a... Monday? Milestone: carbon lugs.-20160917_150257.jpg
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  81. #81
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Found myself arguing with a friend the other day about me being ready for production. I was claiming to be not ready yet, based on my acceptable parts vs crap parts ratio being way too low.

    So I snapped this pic, and realized that I had saved more failure / learning pieces than I thought. At least two others weren't pictured, so I guess that makes fifteen failures of some kind or another between June and now.

    I'm unamused, but at least I gleaned important info from each. Some over compression here, under compression there, leaks here, a different kind of leak there, a resin batch that was inappropriately fast curing, a third kind of leak...

    Pareto principle all up in here.

    Show and tell on a... Monday? Milestone: carbon lugs.-20160920_220804.jpg
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  82. #82
    would rather be ruined
    Reputation:
    Join Date
    Jun 2007
    Posts
    185
    "I'm unamused, but at least I gleaned important info from each"

    And with that you can say each was a success ,your learning very well

  83. #83
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    After damaging a mold component, I had to mill a replacement, delaying a second test of the epoxy-recirculation idea until recently.

    Looks pretty good to me. There are a few lack-of-capillary problems again, but this time I am pretty sure it is because the bladder ruptured some time during the night. This was the bladder's seventh pressure cycle. Next goal: twenty cycles.

    Each bladder costs me about $20 in materials, so if I can have them cost $1 per use I'd be pretty okay with it.

    This is also the first bar that I've made where both the left grip and the right grip are properly registered. I can clean this bar up and actually put things on it. Neat. That much closer to gettin' paid.

    Show and tell on a... Monday? Milestone: carbon lugs.-20160925_155359.jpg
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  84. #84
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Forgot to note: the resin recirculation has been a huge success in my eyes. I got some bits of data, the most fun aspect is that the ENTIRE batch of resin was recirculated in 30 grams increments.

    This is one of those fun times where "on paper this should work" and in real life IT DOES!! Some months ago I was reading about how air bubble "spontaneous nucleation" is a problem for resin infusion, and one way to deal with it is to include a fine mesh screen in the degassing pot. I was like "Fine mesh screen you say?" .... what if the part itself was the fine mesh screen? POW!

    I will allow myself this momentary victory. Now, the next thing...
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  85. #85
    tiny rider
    Reputation: cartographer's Avatar
    Join Date
    Jun 2006
    Posts
    623
    I understand virtually none of this, but really appreciate your sharing your work. It's always exciting to see those moments of inspiration turn in to something functional!

  86. #86
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    The last three bars in a row have turned out fine. Phew.

    It's time to move on: clean 'em up, badge the stem clamps, and start bend testing them. I've already done the thing where I hang them on an overhead beam and I do rough-shod pull ups with them. Buuuuuut that's pretty unscientific.

    Show and tell on a... Monday? Milestone: carbon lugs.-2016.10.06-bar-stack.jpg
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  87. #87
    mtbr member
    Reputation: Velobike's Avatar
    Join Date
    Jun 2007
    Posts
    6,551
    Looking good Drew.
    As little bike as possible, as silent as possible.
    Latitude: 57º36' Highlands, Scotland

  88. #88
    Gigantic Hawk
    Reputation: dubthang's Avatar
    Join Date
    Apr 2009
    Posts
    1,405
    Bars look great. Is there a possibility that the metal clamps could cause stress to the bars? The metal looks to ride in a slightly narrower section for a flush fit. Would bar flexing cause the carbon to rub the metal resulting in a premature failure?

  89. #89
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Quote Originally Posted by dubthang View Post
    Bars look great. Is there a possibility that the metal clamps could cause stress to the bars? The metal looks to ride in a slightly narrower section for a flush fit. Would bar flexing cause the carbon to rub the metal resulting in a premature failure?
    The bars pictured are unfinished, as they sit, the stem clamp zone doesn't reach 31.8mm. Once I bond a 26 ga sheet metal strip (new ones have self alignment wobbles), the 31.8mm will be reached, and there will indeed be a flush zone across, there's a 75mm length of straight tube sharing that same 31.8mm OD (I like my headlights to point forward).

    I'm not too worried about it, due to the ball nose end mill I'm using for the mold making, there's a little pocket between the edge of the metal sleeve and the nearby carbon surface, which I'm going to fill using chopped fiber and thickened epoxy.

    Further, the wall thickness in the area is excessive.

    Even so - that's what testing is for, I'm just going to have to see what develops there.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  90. #90
    Gigantic Hawk
    Reputation: dubthang's Avatar
    Join Date
    Apr 2009
    Posts
    1,405
    Cool. Best of luck with the testing.

  91. #91
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Yarrr, testing be a harsh mistress...

    I didn't do very much science here. Just took a standard aluminum bar that came off an unnamed mountain bike, and gave it a full bore test, which it failed. My first visually passable bar also failed, in that pleasantly disgusting way that carbon likes to do.

    Time to build a better bar. If I can pass the Too Powerful For The Standard Bars test, I'll sleep easy at night.

    Show and tell on a... Monday? Milestone: carbon lugs.-20161207_203948.jpgShow and tell on a... Monday? Milestone: carbon lugs.-20161207_201743.jpgShow and tell on a... Monday? Milestone: carbon lugs.-20161207_202218.jpg
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  92. #92
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Forgot to note, the snapping failure mode shown by my first tested bar is totally unacceptable to me. I know some very, very big men who are interested in being customers. Regardless of what point it breaks, I want a carbon bar that is going to bend.

    I've seen a composite that does this: bamboo. I'm convinced there is a way to dope carbon + epoxy with an additional ingredient for the sake of a safety oriented failure mode.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  93. #93
    mtbr member
    Reputation: Eric Malcolm's Avatar
    Join Date
    Dec 2011
    Posts
    1,014
    How are you doing the test Drew? By lifting the bar to a certain height and dropping it so the clamp hits the block of wood? The weighted ends then having a flexual moment beyond its capablities? SNAP.

    Eric
    If I don't make an attempt, how will I know if it will work?

  94. #94
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Quote Originally Posted by Eric Malcolm View Post
    How are you doing the test Drew? By lifting the bar to a certain height and dropping it so the clamp hits the block of wood? The weighted ends then having a flexual moment beyond its capablities? SNAP.

    Eric
    Correct. In the future I want something like a very strong spring pad that would compress over a very short distance, enough to still cause the same kind of failure, but usable for maths.

    I'd be able to say the bar is surviving ______ newtons. This is currently a knowledge hole I need to fill.

    I am working on testing samples of aramid / fiberglass / Innegra for more likeable carbon failure modes. I'm also cutting new molds that have gentler features and will produce less fiber warp.

    Of the aluminum reference bars I've tested so far, a Race Face bar was very impressive. I could keep on bending it repeatedly, and it never snapped. That is the kind of performance I wish to imitate.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  95. #95
    mtbr member
    Reputation:
    Join Date
    Mar 2013
    Posts
    63
    Looking forward to what you come up with.

    From my experience in aerospace with Carbon, I think it'll be tough to replicate a failure mode similar to aluminum... mainly by the nature of material. As you're well aware I'm sure, the stress/strain curve of AL has a much longer "yield" area before failure, whereas carbon, while fantastic in fatigue - tends to creep up to that limit and fail catastrophically... Part of why on new airplanes they don't do the full wing static test to failure like the metal wings of past. Partially due to the mode of failure being gnarly.... I think that is why you tend to see just a huge amount of safety margin built into the carbon stuff in recreational applications...(that's just IMO though...)

    regardless of my jibberish, definitely enjoy seeing what you're doing, and maybe there are some cool layups you can play with to get it closer...


    Quote Originally Posted by Drew Diller View Post
    Correct. In the future I want something like a very strong spring pad that would compress over a very short distance, enough to still cause the same kind of failure, but usable for maths.

    I'd be able to say the bar is surviving ______ newtons. This is currently a knowledge hole I need to fill.

    I am working on testing samples of aramid / fiberglass / Innegra for more likeable carbon failure modes. I'm also cutting new molds that have gentler features and will produce less fiber warp.

    Of the aluminum reference bars I've tested so far, a Race Face bar was very impressive. I could keep on bending it repeatedly, and it never snapped. That is the kind of performance I wish to imitate.

  96. #96
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Quote Originally Posted by MannaDesigns View Post
    Looking forward to what you come up with.

    From my experience in aerospace with Carbon, I think it'll be tough to replicate a failure mode similar to aluminum... mainly by the nature of material. As you're well aware I'm sure, the stress/strain curve of AL has a much longer "yield" area before failure, whereas carbon, while fantastic in fatigue - tends to creep up to that limit and fail catastrophically... Part of why on new airplanes they don't do the full wing static test to failure like the metal wings of past. Partially due to the mode of failure being gnarly.... I think that is why you tend to see just a huge amount of safety margin built into the carbon stuff in recreational applications...(that's just IMO though...)

    regardless of my jibberish, definitely enjoy seeing what you're doing, and maybe there are some cool layups you can play with to get it closer...
    My brain agrees with you. I chose the word imitate carefully, as I think you're correct that I will not be able to replicate aluminum in the ductility department.

    But I do want to imitate it. At least to the extent of sending the end user a visual cue of "this is still mostly in one piece but you should really stop using it immediately".

    The snapping completely off behavior is @#$&ing scary.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  97. #97
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679

    Continue testing.

    Show and tell on a... Monday? Milestone: carbon lugs.-drop_test.jpg

    Recent bars are not yet quite strong enough, though they are certainly much stronger. I'm fairly certain this is because my winding angle is not correct, due to some dumb ass calculations on my part when ordering carbon sleeving most recently. So the next two weeks are going to be spent cutting new molds while I wait for stuff in the mail.

    Interestingly, the carbon sleeve approach has a "rescue" failure mode, I'm not sure what to call it exactly. It's not yielding, it's not ductility - technically it is fracture, but only some of the fibers failed while others did not. Once it goes limp from an initial overload, the subsequent repeat (and increased) loads didn't rip off the flimsy limb.

    I think I need to take some material out of the wall thickness as it approaches the stem clamp. My hunch is that area is being too stiff, forcing the secondary bend on either side to shoulder an excess load.

    One thing I'm still missing from my test jig (aside from a proper twin-rail setup that doesn't rotate) is a means of measuring contact time. The falling-impact formulas use time as one of the components. I know the impact speed and the mass, but not the duration applied. Changing the value even slightly in a theoretical environment drastically changes the peak load, so I'd love to know what it is.

    Or I could throw (increasingly scarce) money at an impact sensing instrument.

    Up next, construct different internal pressure tools (this part is getting easy, hooray!), use more appropriate winding angle.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  98. #98
    Most Delicious
    Reputation: dr.welby's Avatar
    Join Date
    Jan 2004
    Posts
    1,113
    Why not just use an air cylinder? I broke a lot of handlebars with that, an adjustable regulator, and a ruler to measure displacement.

  99. #99
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,679
    Quote Originally Posted by dr.welby View Post
    Why not just use an air cylinder? I broke a lot of handlebars with that, an adjustable regulator, and a ruler to measure displacement.
    I was advised by a vastly-more-experienced-than-me composites guy to use a method that involves sudden shock rather than gradually reaching ultimate strength. Going beyond strength limits with a large displacement will tell a more visceral story about its post-failure behavior.

    As a bonus, it makes for an easy to understand visual that will be useful to me later for marketing purposes.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  100. #100
    mtbr member
    Reputation:
    Join Date
    Dec 2011
    Posts
    55
    Off chance have you considered integrating some Innegra fabric in your bars? It is supposed to dampen vibrations but does a pretty good job for impact resistance. Hai Velo is using it with some good results it seems.

    innegra

Page 1 of 2 12 LastLast

Similar Threads

  1. Monday Show and Tell
    By adarn in forum Frame Building
    Replies: 11
    Last Post: 02-09-2014, 05:34 PM
  2. Replies: 16
    Last Post: 11-21-2013, 12:45 PM
  3. Where can you buy carbon lugs??
    By mtblucas in forum Frame Building
    Replies: 18
    Last Post: 08-15-2013, 04:42 AM
  4. Friday - Monday show and tell.... tooling
    By RCP FAB in forum Frame Building
    Replies: 4
    Last Post: 03-07-2012, 07:52 AM
  5. Show and tell monday
    By j-ro in forum Frame Building
    Replies: 17
    Last Post: 10-15-2011, 08:47 PM

Members who have read this thread: 133

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •