filament winding progress- Mtbr.com
Results 1 to 29 of 29
  1. #1
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,884

    filament winding progress

    Making tubes has finally become something of a convenience, given the right tools: a winder, a large-enough heating appliance, some "towpreg" from TCR Composites.

    I don't even know how many years my X-Winder has been collecting dust for a litany of reasons. I ordered it during the crowd funding phase at version 1, so I got it for a song. They're rather expensive now, which makes sense considering what they can do.

    You can see some flaws. The Dunstone shrinkage tape is tricky to use. It wrinkles too easily, which are the surface deviations imprinted in the finish where there is intense light reflection mostly toward the left where it is catching my ceiling lights. There are compression and expansion alternatives I have yet to try.

    Another thing I need to sort is rewinding the carbon spool during certain kinds of direction reversal. Sometimes it matters. In the first two images you can see some slight lack of tension pre-cure. That's no good. The compression is meant to be the final step, it's bad to have the fibers move very much. There is evidence of fiber "crimp" in certain places despite the overall good looking finish.

    Good progress - but I can do better. I wanna see some steel front triangles with carbon rear swingarms that I'm gonna build. Can't wait!

    filament winding progress-filament1.jpg
    filament winding progress-filament2.jpg
    filament winding progress-filament3.jpg
    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
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,884
    Also, the middle wheeled support is there as a "zero day" fix for the notion that an 18mm OD polypropylene tube is SUPER flexible. I managed to yank it out of the chucks sans support.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  3. #3
    mtbr member
    Reputation:
    Join Date
    Dec 2006
    Posts
    225
    I know nothing about this, but find it really interesting. Please keep updating.

    I wanted a filament wound seat tube for a frame Iím making, and in the process of nibbling info I found a German guy on Instagram who made them. He wanted Ä400 for a single tube, I instantly redressed my ďneedĒ for filament wound.

  4. #4
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,884
    Quote Originally Posted by Cord View Post
    I know nothing about this, but find it really interesting. Please keep updating.

    I wanted a filament wound seat tube for a frame Iím making, and in the process of nibbling info I found a German guy on Instagram who made them. He wanted Ä400 for a single tube, I instantly redressed my ďneedĒ for filament wound.
    Yeah, that is a bit steep. Towpreg is approximately $30 / lb. Assuming you wanted a 1 lb seat tube (somehow I doubt it), the labor is not massive. I'm guessing his process involved high precision chromed mandrels and he needed a new size, which would be an expensive up front thing, but any consecutive tubes thereafter would be much cheaper. I'm wanting to avoid that material choice for avoiding that kind of catch-22 economics.
    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
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,884
    I almost forgot: Cord, in the mean time, why not use Rockwest Composites bike tubes? If you only need a small quantity.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  6. #6
    mtbr member
    Reputation:
    Join Date
    Dec 2006
    Posts
    225
    Rock west had the size I wanted, or the finish I wanted, or the length I wanted. Unfortunately not all 3 at the same time. I ended up discovering a local company that specialise in fishing rods (tri-cast) to make almost exactly what I wanted for £110.

    The Ä400 was with a mandrel I was supplying, the metal part of the process was the only bit I could do. I politely declined, obviously.

  7. #7
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,884
    Ok so what do we have here. Experimented with a different mandrel.

    Previous mandrels have been with a polypropylene rod as I mentioned briefly. The trais of using PP (plus + for good, minus - for bad) are as follows:

    + Large thermal expansion makes for a grow during heat / shrink during cool effect that can be used as a cheap "inflation bladder", albeit on a smaller ratio of expansion compared to an actual air membrane. Kind of a big deal, because the fibers retain good tension from being pushed outward during cure.
    + Does not bond to epoxy.
    + Relatively easy to machine, but the swarf sure is weird.
    - Bendy.
    - Will undergo plastic deformation if not stored perfectly flat or hung vertically from something.
    - There is a practical limit to the thermal expansion effect - since CTE is a ratio, if your overall base unit (diameter) is small, then the amount of diameter lost during cooldown is not great enough to pull out easily on its own.

    So, to some extent, screw using PP for covering great distances with a small diameter. It is good for large tubes where the rod diameter is such that the flexibility is not very important.

    What to do then for small tubes. This time, tried a steel rod surrounded by Teflon. The idea is to wind, cure with heat, cool, pull the steel core out, then stretch the Teflon by pulling on it.

    And that's pretty much what happened.

    Only problem? There's no significant expansion. All the fibers get pushed inward from the outer pressure membrane (in this case, Dunstone shrink again). My fear here is that the fibers will lose tension slightly and make little wavy motions.

    And that's pretty much what happened.

    So what needs to change with my winder? I think the low-friction eyelet surfaces need to change to rollers. The towpreg is crazy tacky compared to wet winding. Also, I think the amount of tension is too low - friction alone does not seem to be working right. You know how when you peel packing tape or duct tape from a fresh roll, and it makes that peel-pause-pop-pop-pop-pop-peel sort of pattern? Like a ... seismograph - in terms of tension? The feed looks like that.

    You don't want to bounce the line, just like when towing a vehicle with a tow strap in an emergency situation. You want to ride the brakes. Except, in this case, the mass involved is so freaking small that riding the brakes isn't doing the job. I think I need a rewinder device on the gantry. A bunch of stuff I've been reading about filament winding makes similar claims: tension must remain even.

    Apparently there are yet other reasons aside from just keeping the carbon tight for good loading: if the tension is too low, the angle of separation of the towpreg coming off the spool can micro-crack the carbon fibers. An ideal separation angle is tangent. What I'm getting is more like 45 degrees. I'd be happy with a small number of degrees because tangent ain't about to happen.

    In other words, a stock X-Winder can make good parts but not ideal parts.

    On the plus side, I'm starting to learn sequencing. Note the Charlie Brown wavy pattern of the fiber crimps. Fewer of them than last time, woot woot.

    I'm still not too keen on the Dunstone stuff. The epoxy spirals on the surface don't look cool and retain a bit of mass. Maybe I'm applying it wrongly. Maybe it isn't meant for tapering wall thicknesses.

    I think dr welby said something once about hating your equipment for years. I'm grateful for my current equipment set --- but it needs improvement if I'm going to reach predictable strength numbers.

    Images show: Charlie Brown patterning, under-tensioned carbon tows, surface air pockets doing a poor job of escaping (tiny white scratch mark lookin things), epoxy pool spirals. Second image shows ridiculously bent first attempt with a long polypro mandrel compared to the absolute straightness of using a thick as possible steel mandrel core.

    Sorry not sorry for novel. Apparently filament winding is more than just putting a string around a tube. It's putting lots of weird strings around a tube in a hyper specific fashion.

    filament winding progress-20191122_230201.jpgfilament winding progress-20191122_230606.jpg
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  8. #8
    mtbr member
    Reputation:
    Join Date
    Dec 2006
    Posts
    225
    Charlie Brown pattern? Eh? Who? What???

    Donít apologies for the novel, it kept me entertained. Any chance of a video of the machine so we can see the unwinding seismograph in action.

  9. #9
    Most Delicious
    Reputation: dr.welby's Avatar
    Join Date
    Jan 2004
    Posts
    1,208
    Ha yeah, "You will hate yourself and all your tools for your first 10 frames."

  10. #10
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,884
    Quote Originally Posted by Cord View Post
    Charlie Brown pattern? Eh? Who? What???

    Donít apologies for the novel, it kept me entertained. Any chance of a video of the machine so we can see the unwinding seismograph in action.
    Like Charlie's shirt. Pattern highlighted by pink line. Red splotches indicate air flaws at surface.

    filament winding progress-charlie-brown-featured.jpgfilament winding progress-filament4.jpg

    A high class example can be found at carbonfibertubeshop.com as follows:

    Name:  fw1small.jpg
Views: 91
Size:  16.7 KB

    It means the pattern is extremely uniform and well covered. When the tows tuck under each other and make that pattern, it is a crimp in the fiber. Fewer crimps = better. An even better situation would be to eliminate the crimp completely, which would involve running the mandrel backwards as the carriage changes direction at either end. This is hard to do, but not impossible. The pattern I'm chasing can be thought of as a very-very-very-good compromise done at medium effort.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  11. #11
    mtbr member
    Reputation:
    Join Date
    Dec 2006
    Posts
    225
    Charlie Brown, i like it!

    So your pink extended "W" shows 4 tows passing from right to left over/under woven with 4 tows coming from bottom to top. Presumable the more even and uniform your W, the flatter and more consistently your tows have been laid/woven. And i guess a tow is a group of fibres together, like a strip of tape?

    Sorry to pester you with basics, it's all new to me, but very interesting!

  12. #12
    would rather be ruined
    Reputation:
    Join Date
    Jun 2007
    Posts
    202
    Dunno if you have access to ground rod we used to use or place UHMWPE around a ground steel mandrel won't whip and deflect at high rpm there's a ratio of diameter to length and rpm where things get lairy when tension winding

  13. #13
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,884
    Quote Originally Posted by compositepro View Post
    Dunno if you have access to ground rod we used to use or place UHMWPE around a ground steel mandrel won't whip and deflect at high rpm there's a ratio of diameter to length and rpm where things get lairy when tension winding
    The one test I've done with a steel mandrel indeed showed incredibly high stiffness. I tried a PTFE tube over that, and that worked pretty well, except that it was slightly oval cross section. The stiffness of the steel mandrel revealed the weak points in my winder, specifically the gantry, wherein the tension is not maintained at every single moment. I had a thought the other night, you know how with derailleurs on bikes, the tension arm is built really light weight with a strong spring to quickly take up the slack from moving chain mass. And now clutch derailleurs are more likable because their tension abilities are even stronger than the previous generation. Maybe an analogy to the light weight tension arm also applies here.

    The way I've seen some winders use REALLY long gaps between idler wheels to naturally baffle the tension variations, but I don't have the room for that. I think my shop is 13 feet across given the rest of the tools/storage already present. So I need some compact way to maintain tension so that I can do a good job of winding on a non-expanding mandrel.

    I'll try the UHMW, thanks for the tip.
    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
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,884
    Despite my relative discomfort with recent things with Walt leaving these online shores, I guess I'm gonna keep posting here.

    I another test with a centerless-ground steel mandrel wrapped in a PTFE tubing layer. One thing I noticed with the previous molding in this manner was that the OD of the PTFE had changed after one processing cycle. It's a relatively thick bit of tubing.

    Well - after a second cycle it got screwed up even worse. Observe the interior surface of the tube. Twirling the PTFE tubing under the light reveals such warbles in its reflection as well.

    filament winding progress-20191126_051935.jpg

    It is becoming deformed well below its melting temperature.

    Unfortunately, after looking closely at UHMWPE as a low friction liner option, it doesn't have the temperature rating needed by my current client. Gotta stick with PTFE, but maybe a thinner layer of it. But it is giving me a rough time, I wonder why? After talking with a guy, apparently PTFE has some funny behavior -- a very low softening temperature but a high melting temperature, and poor stability in a thick solid form, which contributes to it generally being used as a surface treatment. So I'm going to try silicone backed PTFE thin gauge tape. Applying it in a spiral with zero gaps is going to be... pppppfffffttt... well, that's why I tried the tubing, basically. Denial, bargaining...

    I also figured how to apply more even tension on the tow spool with my current supplies, which... promptly resulted in applying so much torque that the carriage belt stepper motor started losing position, resulting in wrongly sorted filament wrap and exposed zones. I inquired for tech support from X-Winder, and Turner responded (PARAPHRASING) that the default config of the machine is for wet winding, and that adapting the machine to towpreg is possible, but don't expect it to do that in stock form. Which, hey, I'm fine with that now that I know for absolutely certain what angle to attack.

    The centerless steel core approach really is great for linearity across a distance. 1.5m pictured:

    filament winding progress-20191126_051957.jpg
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  15. #15
    mtbr member
    Reputation: Feldybikes's Avatar
    Join Date
    Feb 2004
    Posts
    897
    Have you tried an Al mandrel? I remember hearing from someone who knows what they're talking about that steel has too small a CTE to release well but Al works well.

    Also Iím happy to collab on a FS frame when you get to that point.
    I make bikes. www.feldybikes.com

  16. #16
    would rather be ruined
    Reputation:
    Join Date
    Jun 2007
    Posts
    202
    Never caught you were curing past 90 deg c

  17. #17
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,884
    Quote Originally Posted by Feldybikes View Post
    Have you tried an Al mandrel? I remember hearing from someone who knows what they're talking about that steel has too small a CTE to release well but Al works well.

    Also Iím happy to collab on a FS frame when you get to that point.
    Makes sense to me, I'll try an aluminum mandrel in the near future. Heard of one guy that uses dry ice to shrink his AL mandrels out. Client wanted to try with steel first for sake of high stiffness over significant length at small diameter. Kind of a moot point now that I've figured it's relatively easy to support the suspended mass along the middle with rollers.

    @compositepro: well, I probably didn't mention it -- mea culpa. In this case it's not a bicycle tube, but, whatever I learn from this implementation can benefit future bicycle specific tubes.
    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
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,884
    I received some very-thin PTFE taped back with silicone adhesive, so it will be tolerant of a lot of heat. As I feared... even at this thin, it is very unlike packing tape for example. Packing tape is useful for making very straight lines with little effort because it is stiff along its width. This PTFE stuff is more like a gum in tape form. While I do not exactly need to handle as if it is extremely fragile, it is also pretty easy to pinch fingers, pull, and watch it distort readily while the silicone whitens. This is going to be incredibly hard to apply to the mandrel without any gaps or overlaps, and I already have doubts about it being a successful experiment. Will know soon enough.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  19. #19
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,884
    OKAY, so, time to break down the good, the bad, and the changes of the recent test.

    Changes:

    New software installed. Carriage speeds reduced. Teflon tape applied to all contact points along the towpreg path.

    Not changed:

    Towpreg tension problems and Dunstone related spiral compression problems yet remain.

    Good:

    The ordering of fibers was much improved. The most recent software allowed for some carriage motor tuning, which eliminated the over torque / missed steps problems (which probably had everything to with the fiber ordering). Lots of chevrons / Charlie Browns. Woo woooooo

    Bad:

    As I expected, hand application of wide (2") teflon tape is a nightmare, the finished tube had visible tiny grips to hold onto. I could see them, I could feel them with my fingernail through a nitrile glove. Maybe I should try again with a narrow tape that's easier to handle, and spiral it into a sort of makeshift screw thread... but... I have reservations about trying it.

    This got me to thinking, anyway. My bladder molding stuff has been pretty successful. While that process may have involved resin transfer, it obviously translates to prepreg stuff very well, that's how almost everyone else does it after all. So, maybe I need to take my bladder related knowledge and apply it to making straight tubes. I didn't want to involve that level of complexity... but maybe it's what I need to do. @#$*

    There's still a few things to try, such as the aluminum mandrel idea. I could spice up that notion with a few details, such as a VERY slight taper, and/or some lengthy polishing until I can see my face in the surface. The idea is to make it more slippery than an oligarch.

    But... when it comes to making slightly curved shapes (such as chainstays), the slippery mandrel thing just isn't going to work. I need to make mandrels that either mold in place permanently like I've done with my handlebars, or I need to make some dissolvable tooling ($$$$$$$$$), or I need to make trapped rubber tooling.

    Attachment 1296781
    Attachment 1296783

    At any rate, I think this thread has largely run its course, and it was to answer this notion: can you, for a budget of less than $10K, get into high quality filament winding at home? Yes you can. Should you? Ehhhhhhhhhhhhhh... just because it's going to fit my future situation very well doesn't mean it's worth it for everyone. I know one guy selling his winder so YMMV.
    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
    Most Delicious
    Reputation: dr.welby's Avatar
    Join Date
    Jan 2004
    Posts
    1,208
    Quote Originally Posted by Drew Diller View Post
    or I need to make some dissolvable tooling ($$$$$$$$$)
    Could you 3d print PVA tooling?

  21. #21
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,884
    Quote Originally Posted by dr.welby View Post
    Could you 3d print PVA tooling?
    That is a fine idea, though I lack a functioning printer. I'd need to own one for that approach to make sense. Buying anything at all is suddenly questionable after my wife was just delivered news a few hours ago that her job (along with hundreds of her peers) is gonna disappear in February.

    If we were to pretend cost is no object, print speed is still a concern. So if I could mill a negative of a mandrel and cast something in it that is stiff-during-layup and soluble-after-cure, that would be ideal for curved shapes. There is a substance for doing exactly this but I think it got bought out by the US military, though I cannot seem to get a straight answer on the subject. At least I have not gotten killed for asking, hooray for speech.
    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
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,884
    Pleasant surprise: after some cold treatment, the dimply steel mandrel covered in wrinkly teflon is moving after some hard strikes on the ground. Moved it all the way to where pounding simply wouldn't push any further because the carbon is contacting the ground. Well, then, Teflon. You win this round. This tube should not be coming off, but it is. That's just weird. Gonna make myself an improvised mandrel pusher. I've used a bottle jack and a large syringe to do resin transfer -- I want to find out what a bottle jack and a bit of sturdy hardwood can do.

    I guess this reinvigorates my willingness to experiment with Teflon tape -- albeit of a different size, next time.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  23. #23
    mtbr member
    Reputation: dRjOn's Avatar
    Join Date
    Feb 2004
    Posts
    2,593
    Quote Originally Posted by Drew Diller View Post
    There is a substance for doing exactly this but I think it got bought out by the US military, though I cannot seem to get a straight answer on the subject. At least I have not gotten killed for asking, hooray for speech.
    its always fascinating following your progress but when it gets spy vs spy, it's getting serious! cheers and sorry to hear about your wife's job - that level of uncertainty is never good....
    For a rock steady Gas Tank bag > the DeWidget

    bit.ly/BuyDeWidget

    https://www.instagram.com/drj0n_bagworks/

  24. #24
    mtbr member
    Reputation:
    Join Date
    Dec 2006
    Posts
    225
    .

    Attachment 1296781
    Attachment 1296783

    .[/QUOTE]
    Attachments donít work, is it just me?

  25. #25
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,884
    @Cord -- yyyyeeeeah I don't know what's going on with that. I attached images in the usual way. Maybe next post I'll use my own website host to store the images

    @dRjOn -- to be fair I wouldn't trust my own memory on this. I think that's what I read in some press clipping, but... *fuzzy*, and at the end of the day I still don't have the material in my hands. There's a few other materials out there that I am already qualified to buy, but the minimum order size and expense is CRAZY. Like if you're doing something other than building race cars or aircraft, don't bother asking. Which is a real bummer because some of those advanced materials can handle compression forces and bike frames have all manner of zones that obviously deform in compression.

    Regarding my wife's job, I think her circumstances will allow her to land on her feet, but the situation pains her emotionally because it has been her one and only professional job, she was there more than a third of her life so far. She doesn't call them coworkers, she calls them friends. She feels sad for other people more than she feels bad for herself.

    Just means I need to get product out the door all the more swiftly.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  26. #26
    would rather be ruined
    Reputation:
    Join Date
    Jun 2007
    Posts
    202
    You need a mandrel puller

    Honestly there's also a cheap ass way where you have a threaded collar on the mandrel , two flats to get a spanner on and pop it off that way

  27. #27
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,884
    Mandrel puller, eh? I was like hmmmmm... I have a half-ton rated little hoist that I use to get my winder table off of my main table and up along the wall (the winder itself is relatively flat).

    I took the hoist off the wall and found some heavy lumber, fastened some blocks to constrain the end of the carbon tube. Things started moving right away, hot damn! Not much effort. It was harder to manage the chains, as I quickly found that the hoist wanted to choke on its own guts due to the lack of gravity that would otherwise be keeping the chains relatively oriented toward their inlet.

    Let's see if pic attachment works this time.

    filament winding progress-20191207_212403.jpg
    filament winding progress-20191207_213841.jpg
    filament winding progress-20191207_213845.jpg
    filament winding progress-20191207_214107.jpg
    filament winding progress-result_tube_small.gif
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  28. #28
    will rant for food
    Reputation:
    Join Date
    Jan 2010
    Posts
    3,884
    BTW, getting the animated GIF to show up in the previous post was a chore. Upload limit for GIFs on this forum is 879kb, I had to make a custom video output in Kdenlive and then run the resulting GIF image through EZgif. Total chore.
    Disclaimer: I run Regular Cycles (as of 2016). As a profiteer of the bicycle industry, I am not to be taken very seriously.

  29. #29
    mtbr member
    Reputation:
    Join Date
    Dec 2006
    Posts
    225
    Quote Originally Posted by Drew Diller View Post
    BTW, getting the animated GIF to show up. Total chore.
    Worth it, look at all that Charlie Brown!

Similar Threads

  1. Progress is.... well progress
    By kinsler in forum Women's Lounge
    Replies: 4
    Last Post: 09-25-2012, 10:42 AM
  2. help - Rockshox fork winding down while riding
    By ringo in forum Shocks and Suspension
    Replies: 4
    Last Post: 02-03-2012, 08:07 PM
  3. Wonderful winding windy Wednesday
    By Funrover in forum Colorado - Front Range
    Replies: 2
    Last Post: 02-02-2012, 10:35 AM
  4. Replies: 21
    Last Post: 11-06-2011, 05:26 PM
  5. Winding Trails of Farmington
    By Tony777 in forum Connecticut, Rhode Island
    Replies: 8
    Last Post: 04-13-2011, 04:53 PM

Members who have read this thread: 68

Posting Permissions

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

THE SITE

ABOUT MTBR

VISIT US AT

© Copyright 2019 VerticalScope Inc. All rights reserved.