Fiber-reinforced resin prints - how much STRONGER are they? 

I liked the print head bouncing in sync with the beat of the music 😁

Concerned about the methodology here... There is a huge difference between Tensile Strength and Flexural Strength, and this test (which is purely flexural) is kind of a worse-case scenario for the orientation of the layers and the direction of force, mechanically speaking. I'll ramble a bit but don't want one wall of text so... If I could recommend one improvement it would be to print two eyelets in your test pieces and do a Pull Test in a similar orientation to your scale. This would be testing Tensile strength instead of Flexural, and I suspect you'd find the overall tensile strength greatly improved over the default resin. When working with material composites, orientation is everything. To put into other terms: imagine a wet rope gets frozen all the way through. The newly rigid rope might, under the right circumstances, break if set parallel to the ground and force is applied to one end, as it's being attacked in it's thinnest/weakest dimension. If, however, the frozen rope was used to lift a load (as a rope realistically might be asked to do), failures in the ice may be more numerous, but they should be overall smaller, more evenly dispersed, and overall less impactful on the rope itself or it's ability to lift the attached load.

Please wear a respirator when working with glass fibre, especially when you have frayed edges and you are making cuts.

CNC Kitchen should repeat this. We need multi site validation of results. Also the standardized test machine with replicates and standard deviation appeal to my inner scientist

Maybe it would also be good to test the orientation of the fiber weave as well? Alternate the layup 45° each layer perhaps? Great work as always!

We add 10% milled 1/64" glass fibers to the resin by weight, it works great.

Great video. I actually use reinforced fibers a lot, we build bicycle frame, velomobiles (sometimes called "banana bike") and special bicycle components. We use glass, carbon and kevlar for this. Strength depends on the type,thickness and weave of the material but even more important are the resin/fiber ratio and how compact the end result is. Carbon is way stiffer than glass and kevlar is almost like a rubber band. Kevlar is mostly used in combination with carbon to prevent it from shearing apart on impact. When using fibers, we always apply pressure to the end product either mechanically (by pressing the resin/fiber product in a 2-part mold), using vacuum (by placing the resin/fiber product in a bag and vacuuming all the air out or by applying pneumatic pressure (the product is made in a mold and the core is a balloon that is blown up to a few Bars of pressure). In this way the weave is pressed together and the layers are pressed on top of, partly into, each other. For us a 70% fiber ratio works best, so the product contains about 30% resin, but sometimes we make carbon sheets of 8-10mm thick that contain even more fiber. If you tap on that material with a hammer it sounds like a ceramic tile and it is very hard (does not flex at all) whereas for most things (like a bicycle frame, a steering bar or a front fork) you do want some flex. The type of resin is also very important. Some resins are very brittle whereas other resins have more flex. A brittle resin with too few fiber content will result in the resin cracking up and the fibers shifting with tears and cracks in the fibers as a result. A more tenacious resin with a carbon/kevlar mix will end up like spring steel; hard to break, very flexible and very durable - this is even used as spring leaves by some companies. I like to see what happens when you use short cut up fibers as suggested by another user too. We use 0.1 - 0.3mm glass and carbon fibers as fillers and those are mostly used in combination with glass bubbles and cotton fiber. The carbon or glass will add strength, the glass bubbles make it flow more easy and the cotton fiber make it thixotropic.Cotton is not suitable for 3D resin printing but mixing small fibers or glass bubbles is something that could be worth a try

I think 3D printing a mold and just making an epoxy/glass part would be easier at this point. Very cool tests!

Looove the editing! @3:15 printer disco? Yes please!

Truly, a great idea! And definitely something to keep an eye on. Below I gave you some suggestions, based on my composites experience: I think there were two main possible problems: 1.) Compatibility between the UV-resin and the glass fibres might be very poor. You would need to check the compatibility between the applied glass fibre silane-sizing and your matrix material. Sizing compositions are usually proprietary, so this is really hard to find out. I suspect that most UV-resins cure by radical mechanism, so a vinylsilane or methacrylsilane (or even Volan cloth) might work best. If, for example the fabric had an aminosilane applied, there is a chance that it would not react with the UV-resin. 2.) Impregnation quality could be poor. We do not know if the resin has fully wetted the reinforcement. There might be large voids within the fibre bundles, since no pressure was applied to compact the reinforcement layer and push air out. Looks like it at 7:32. Usually the individual fibre-bundles should appear almost invisible when properly impregnated. Maybe your cleaning agent at 6:27 has soaked into the fibre-reinforced parts prior to curing, which could cause a lot of problems. 10:50 and 14:06 The main failure mode looked like debonding and delamination, rather than fibre fracture - which would strongly agree with my theory. Therefore, the potential of the fibre reinforcement was not even closely exhausted. A continuous fibre-reinforced part must be both stiffer and a lot stronger than pure resin (thats common knowledge). A well-engineered fibre-reinforced composite usually fails in catastrophic manner. Here, the roving-reinforced composite failed by gradual delamination/debonding

I really enjoy these videos Tom. 👍

This was a great idea! I've not seen anyone else on RU-vid do this. Great video and I can see this technique gaining popularity. Thanks Tom.

Can't be the only one from the community asking for a collab with Stefan from CNCkitchen ✌️

Thank You for performing these experiments. Reinforcing resin prints is what is on my mind.

Interesting application towards engineered composites. Generally fiber composites have a resin volume fraction of 30% or less, and you'll only see improved strength and stiffness if the raw fibers are themselves stiffer and stronger than the matrix (maybe why the random orientation drywall fibers failed)

Nice video! Another reason why the strenght of the reinforced samples is not very high could be the fact that the printing resins do not adhere to the glass fibers like epoxi or polyester resins do. To transfer the external load that is applied on the resin to the fiber, a chemical bond between resin and fiber is needed. A specific resin designed for fiber reinforcement could get better results.

I think the reason the more coarse weave works is that those layers have holes/gaps where the resin can connect between layers. With the glass fiber paper, you create a total break inbetween each resin layer. So you loose some of the effect of it being a composite material? So trying the paper again with holes or lines cut in would be very interesting.

Better than any netflix documentary. Nice editing. Nice shots. Nice music sync. Deep and experementary infos. Really gread vid.

Hm... on plastic (and filament) shredded fibers are often used. You could try those, but should be a similar result i guess. Great video 👍

Bruh this definitly deserves a thumbs up, I can see all the work you put into this

Incredible video!

Wow, this kind of videos about 3d printing is expected from CNC Kitchen and not from you, Tom. You actually did great !!!

I have thought of doing this for some time but I was thinking of mixing the separated fibers into the resin so the print would just run with no pauses. Would be interesting to see this if you can do it (may be messy?).

Drop the beat! I LOVE how you selected and synced the beat to the thumping of the printer. Epic !!! :-)

Could you try unidirectional fiberglass strands knowing your bending load. Also could you try small chopped fiber just mixed into the resin?

Very nice, you clearly worked really hard

Great choice of bed-banging music, Tom!

Time to replicate CNCKitchen's test rig? Good concept, sketchy test results, I want to see a load curve. And I bet the fiber ones will go up sharper than pure resin ones.

Great job Tom. I would recommend getting ahold of a Instron test machine to run a proper tensile and compression test.

I really like the blending background color. It gives such a nice and smooth contrast.

It almost seems like the eventual realization of this idea would be some sort of mixed-mode/media printing where the printer could insert sheets or strands of fiberglass from the side automatically during a print. Either way, super fascinating results and well played for showing your results even though they didn't match your hypothesis.

I was FRP inspector for years use the RTP1 book as a guidance and astm for pipeping will give you the right amount layers of material to have the desirable strength. Roll out the material first make sure that air is out for max strength with the same resin then 3d print on top of that. Do a tensile test, barcole test and a burn out to see how strong is your compositive. Like your show I do 3d print as a hobby. Thanks for your dedication.

Fascinating. I would like to see a follow-up video where you test resin-fiber slurries. FormLabs' Rigid is a glass-fiber slurry resin from what I remember.

Seeing the printer paper reinforced part looking perfectly clear made me wonder; what would it look like if you used a 2D printer first? You could incorporate some really neat designs.

I'd be interested in seeing how loose fibers would work.

Great stuff! I have been using composites for years, combining with resin printers opens some very new directions!

lol loving the editing.... bouncing along with it :D

A bench grinder would work wonders for that fiber glass work. Great test!

How about chopped fibers mixed into the resin? Should remove the hastle of trying to insert fiber layers into the print and lead to more isotropic performance.

You really synced the resin stamp to the music.. OMG details.

I would have loved to see a stretch-breaking test, rather then a "bending" stress test with the reinfoce parts! Shouldn`t the reinforced parts be able to endure much more of a stretch-stres so to say?

Tom, that's a genius idea! Imagine the ease of fabrication for home grade quads! Man, that's great! Thanks a lot!

Hardness, strength, toughness, brittleness, and ductility. All of these to consider for the material property needs of whatever you're making.

Great video and a very interesting experiment!

6:12 This is some good printing music xD

Would be interesting to see how these parts performed in tension etc, great video and fantastic concept, thanks!

CNC kitchen has interred the chat

That why I love Thomas. That was awesome. One thing to try if you feel like it. In boat building they stack the glass fibers directly on top of each other with minimum resin between them so it ends up like a single fat glass fiber piece. Another thing they do is that they press the layers to push the air out of the glass fiber but I can’t think of a way to do it in the printer.

Glass fibre doesn’t have the exact or same strength as carbon fibre.

I used fiberglass cloth in my model airplane builds, especially over foam air frames. Something that we learned to do is to turn the fiber direction by 45 degrees to the layer before it, to change the directions that the stresses act upon it. This helps in the strength of the part being made or covered.

Netflix: Are you still watching? Someone's son: 3:26 Edit: And the synced music is very high effort content creation, just noticed it!

A suggestion... When you are doing the bend test, the tensile stress in the upper region is taken by fibers only and compressive loads at bottom part is taken by the resign. Here, the resign is more flexible and can take more strain in compression. This shifts your neutral stress plane above the center plane and less fibers are taking majority of the load (that is the reason you are seeing top side of samples failing much often). The maximum load carrying capacity during bending then depends on the total tensile load carrying capacity of those contributing fibers. Further, the strains on upper fibers are more(due to bending curvature) compered to fibers close to neutral plane. This results in upper fibers being loaded earlier than the one below it which results into progressive failure from top to bottom which you can see in the sample with 6 layers of glass fibers. During tensile testing the maximum tensile load the sample can handle is the combined force of all of the fiber layers. the maximum bending loads in similar way has relation with the fibers contributing to loads. A way of improving flexural rigidity is by adding reinforcement layers close to each other in a way that distance form the center plane remains nearly same for each of them. Using more flexible sheets at layers further away from center plane also helps, as those can sustain more strains before failure which are present at surface layers, and thus distributing the load evenly among remaining fibers. Resigns with more compressive strength would prove to be more effective in improving flexural rigidity. You can directly relate it to the RCC beams where rigidity of concrete provides strength against the compressive loads in the lower part of the neutral stress plane while tensile loads are handled by the steel rods.

So I though you were going to add individual fibers to the resin (like a lung cancer powder). This is interesting too though!

Another suggestion for chopped fibers within the resin! Awesome vid though!

How were you able to plug the USB from the first try ? that is impossible!

I've worked extensively in the composites field. Fiberglass and other composites have excellent tensile strength but very low strength for compression or bending. If you repeated these parts and used a testing method that pulled from either end, you would likely see a very significant increase in the strength of the part, especially with the weaves as the fibers are aligned. You may also see somewhat better results in a bend test like this by only reinforcing the top side which is in tension during a bend. Bending tests do favor less rigid parts, as the flexibility allows more deformation of the part before actual failure. Think of a comparison between metals such as steel and titanium. Titanium is considered to be as strong or stronger, but because of its increased rigidity it is more prone to failure when subjected to bending.

Honestly, at this point I'd just invest in or make my own little cnc milling machine and just buy CF stock. This looks waaay to messy.

Pretty cool idea! One reason the wallpaper stuff was underperforming might be the coatings that are (not) applied to it. Fibers meant for strength in laminates usually are coated in one way or another to provide good adhesion with the intended resin system. You could quite easily imagine a setup that automates a process like this as well. Just keep the fabric on a roll, and cut and submerge a new layer every time.

I just look at resin printing. The gloves, the mess, the chemicals, and think no!

Carbon fiber added to resin is the same as putting metal rebar to concrete. It's more durable and much stronger if applied correctly. Nicely done!

I really appreciate this mix of traditional fiber composite and resin printing. It's something I never thought about but is a really interesting idea. For next time I would read more into the process for more traditional methods of making/designing fiber resin composite structures. It should give some clarity to the results, like the effectiveness of the loose fiber(tape) VS the woven variety, and give some ideas to explore. Also instead of paper try old clothes for a cheap accessible way to reinforce your prints. And as many people have commented safety first.

okay... so this gave me ideas. try adding glass and carbon fiber to the resins. shake it up & print with it. might get some similar properties to fiber filaments.

Very clever! I don't see an immediate application, but a nice option I'm sure will come in handy.

Wow, the amount of time and work that this testing must have taken is amazing, thank you for the video! I'm not surprised by the results having studied composite materials quite a bit and I'd like to impart a bit of material science knowledge here. As you astutely point out, the issue here is the huge disparity in flexibility between the resin and the fibers and generally the way to fix this is to change the volume ratio of fiber to resin. For glass reinforced composite, this is typically never lower than 20% fiber and often much higher like 70% fiber, and getting 70% fiber with this method as your video shows would unfortunately be more time consuming than it's worth. The second factor is the resin used and since photopolymer resins (even Sirayatech's Blu resin) are far less tough than the two part curing resins used for fiberglass production, I don't know if you could ever really make a useful composite out of them. I think the fastest way to test this would be to just lay down many layers of fiber then pour the resin over it and compress it all in a clear container, then expose the whole thing to UV light all at once. Lastly, there is the concept of the void ratio (the volume of voids over the volume of bulk material minus the void volume). In this case, the overall void ratio would be pretty low but unfortunately, all of the voids would have been concentrated right at the boundary between the fibers and the rest of the resin which also reduces the strength significantly. Anyway I hope this is helpful and thanks again for putting so much work into improving the knowledge base of this hobby!

Given the... crudeness, no shade intended, of the way the part was created it's extremely exciting to see a nearly 50% improvement.

It might make sense to place the reinforcement towards the middle rather than the outsides. Brittle materials (high compession strenght, low tensile strenght) normally get reinforced by putting something in the middle that either has a higher tensile strenght, or pretensions the outer surfaces by contracting after curing. The goal is to use the inner phase as a lever to convert the bending stress into compression on the inside curve without reaching the required tensile stress to crack the part on the outer curve. The more brittle the material, the more it benefits from this kind of reinforcement. Theoretically, one could pretension the print by variing the layer thicknes and exposure time throughout the print. (As the resin contracts differently with different exposure.) Reinforcing the part without adding reinforcement. (See gorilla glass, or tempered glass to a lesser degree).

I like how the fiber reinforcements make any eventual failure much less catastrophic.. instead of splitting completely (and sometimes with pieces carrying potentially dangerous amounts of momentum), the heavy glass fiber arrangement makes the failure take a lot longer to occur from its start to finish, indicating that it may useful in applications where catastrophic failure of the printed part might result in destruction of something much more valuable (like mounting cameras, for example).

Why hasn’t anyone already done this? Seem so logical. Well done! I wonder where else it’ll go; Nomex fibers, nylon, polyester, Kevlar, etc.

I think that under a straight tension load (pulling from each end, not bending) this technique would likely do better. Another idea is to mix the resin with a lot of individual glass fibers, less work than manually tending to the print and adding layers, and it might result in a more amorphous mix of reinforced directions, which could be useful for some applications.

Tom this was a GREAT set of experiments! I know for this you used just a *generic* brittle resin, but if you take it further, I would love to see how much of a difference it makes with *tough* resins, such as Siraya Blu, or even a Blu/Tenacious mix.

I've never worked with any kind of 3D printing but happen to know just a bit about using glass fiber and resin. When doing this the traditional way your goal is to use as much glass fiber and as little resin as possible while still making sure there are no air pockets. When making really high spec items such as masts for sailboats you use high pressure, often several tons of pressure to press as much resin as possible out of the laminate. These can end up containing more than 90% glass fiber per weight. It's the same with carbon fiber or Kevlar weave which can also be used to reinforce resin. I'm not trying to teach anyone how to fiber reinforce their 3D print, but I'm pointing out that using 2 or 6 layers isn't very much like how it's usually used.

Props on the "shear" ;) amount of work you inputted into this project video.

Literaly got the add for exactly this

Probably the difference between this method and the classic reinforcement one where we use vacuum chamber is that the stack of layers using the vacuum we can assure that it will work like a whole part. However, with the method that you use here, we can not be sure the stack will work together due to the interface between layers and, consequently, its will have a worse performance. Great video by the way and an interesting workflow Thanks!

Flexural vs. Tensile debate aside, really high quality video production, great job!

very cool video.... in this application(among most) using "normal" fabric the crossover transitions in the weave greatly reduces stiffness. 2x2 "twill" weave means having a minimum of 50% properly orientated fiber not crimped at the over/under weave transitions. Going to a 5 harness weave (over 4 under 1) means 80% continuous fiber and a "unidirectional" weave means all the fibers are properly oriented for stiffness (in this case 200% as there is no cross weave). It really does make a huge difference which almost has to be seen to be appreciated. Unfortunately these more exotic weaves are not so easily sourced especially in glass. Carbon is much stronger with less elongation and easier to find more exotic weaves but UV cure is a problem. Maybe in this case with the low fiber content CF could be cured out the outside layers anyway. There are precisely cut "milled" fibers available and would be an interesting experiment but I think parts would turn out rough & fuzzy and require too much post processing to end up with accurate parts. As resin''s elastic properties are so different than reinforcements, resin's primary function is to maintain the reinforcement's position/orientation. quality FRP parts are generally about 50% reinforcement by mass. I'd be very interested to see how high of ratio of milled fiber could be achieved as at anything near 50% would be much to viscous to flow.

Huh, thats a really interesting and counterintuitive outcome! Also, love the printer bouncing to the beat

Really interesting introductory video about composite materials. I had more hope in the 23x layer thin drywall fibre glass but i think the main issue was the alinement's of the fibres. In the other clothes the fibres where crossed like in a fabric but on the drywall fibres where all parallel to each other and were put in the wrong orientation where the part would break between fibres. A part made with 23 layer of drywall fibre glass may hold way better with the fibre orientation crossed between each layer or simply keeping all the fibre parallel but with a 90° rotation where the part would have to break along the fibre and not between, same for paper which have only one fibre direction.

Okay I have to admit I didn't think it would even print I thought you were going to crack your LCD or it wouldn't adhere to the build plate very impressive video

I wonder what would happen if you used fiber threads non woven? This would allow them to potentially cross between layers instead of all being horizontal. Fantastic video!

One thing I've considered is finely chopped fibers added continually throughout the print. Theoretically as long as you keep the proper ratio it shouldn't make the resin too cloudy, but the random orientation of the fibers should help bond between multiple layers throughout the print

Thank you for making a mess on our behalf! Keep doing what you do!

Interesting. If I could recommend: 1) bias the fiber cloth rather than align with the direction of stress; 2) basalt fiber is supposed to have coarser surface features, which may improve the resin-fiber interface yield strength.

I always feel great about my brain when I see people work fiber-glass or carbon without gloves and/or respirators. Have you ever thought about not ingesting extremely toxic and carcinogenic materials? It rocks, dude!

Great video! I hope 3D printing embraces multiple material builds more in the future. Too many designs try to accomplish everything with the print alone. Even Hermes uses filter paper as a stiffening element in their watch bands. A different kind of 'composite' entirely, but it's fair to say paper has its place in builds.

You should take the high and median strength methods for the strongest two layups and repeat those four prints using the two different alternative resins you mentioned at the beginning. Thanks.

You read my mind!😂. I've been thinking about this and it looks every bit as difficult and messy as imagined.😂. It would be nice to try different resins too. I bet the tough flexible resins would do better with the added rigidity of the fiber because the heavy fiber seemed to fail when the resin shattered. Maybe add some tenacious to the resin so it's more flexible and the fiber will add the stiffness back.👍

Very cool video :) If you want to experiment further it would be fun to see some results for UD (unidirectional) fibers. Having all fibers along the 0-axis should give you a much stronger part. I might have some laying around at my workshop if you need a small sample :)

Very interesting experiment! Some advices from a composite engineer: The sizing (silan primer) of the glass fibers need to be epoxy compatible. Yours probably are, but it's not obvious. And when you test such small items, there is a strong "edge effect" of the material. Imperfections, such as unsaturated fibers at the edges can cause an early failure. I have seen that in other composite testing. A larger item would be more useful. And maybe bending test until first shear load failure isn't the best indicator of material improvements. As you say, higher material stiffness might be the property you are looking for. A very resin rich composite material will be weak in general. When testing, you see the resin parameters instead. That's because when areas of resin cracks from ultimate elongation or ultimate shear load, the relatively few fibers cannot take over the high concentrated load in the crack area . You should aim at an active fiber volume share of at least 20 - 30%. I don't know the exact figure. It varies from many different factors. But resin rich areas in composites are definitely seen as weakening failures for the same reasons. Maybe you can even use short fibers mixed in the resin. In thermoplastics you see an effect of added fibers, which are shorter than 1mm. Are no resin suppliers offering such a mixture like in some filaments for 3D printing? Maybe there is a negative effect on the curing process, if no resin suppliers do!? At least maybe in the effective volume share. A low volume share of fibers does not offer a significant improvement of material properties. Sometimes even the opposite.

Using a fiber layup that is entirely in the orientation of the load would improve your results tremendously. The glass mat weave you're using has half of the fibers perpendicular to the load, which means that for those segments it is entirely resin trying to do the job and it is no longer even continuous.

So cool to see someone try this! i always thought about this as i work with composites!

You could also try Chopped or milled Carbon or Fiberglas and mix it directly with the resin. R&G Faserverbundswerkstoffe has something like this directly from germany

Editing was on point

This would be great for making reinforced connections between parts on large segmented FDM prints - You could model slots into the connecting area between two parts and add fibre reinforced resin pegs, hooks or even rings to hold it together depending on the shapes you're working with.

The drywall cloth sucking makes sense. It's an inclusion that's probably weaker than the standard resin, and no long fibers to provide extra rigidity. The glass fiber cloth adds strength and rigidity by adding strong tendrils throughout the length of the part, and adding more layers reduced the localized stress around the inclusions. Overall, excellent dive into some more complex material science topics. I'd love to see the crazy numbers you'd get with a pure tension or energy absorption test.

Its worth noting that fibers add strength when under tension, but don't do much (if anything) under compression. So if you want to get the best sense of how much the fiberglass improves the strength of a part like this you would want to load the object parallel to the fibers. The bending test being applied here is shearing the layers of resin against the fibers and only leveraging a small fraction of the strength of the fiberglass. Its testing the shear strength of the resin matrix as much as it is testing the effect of glass fibers. Thomas - Thanks for all of the great content! Perhaps this is a collaboration opportunity with Stefan over at CNC Kitchen? His rig would be ideal for testing these parts purely in tension. :-)

just today i used the SIRAYA Tech simple clear resin, what i like about it is that it's super thin so it flows easily and is easy to cleanup. also very fond of Phrozen water washable rapid resin. it kinda sucks to wash it with straight water but you can cut your alcohol with water big time so its more economical, and it prints pretty quick i was getting away with 5 or 6 second thick layers on my creality ld-002r

Hi. This was a great experiment. The fibre is meant to make resins more ridged and not to strength in they way you test it. Eg cast iron is way more ridged than steal. If you had to test the two different metals in the same way you carried out this test then you would get the same result. If you need less flex in parts then use this method. I think this would be great method for building quad frames.

Yeah I've already played with this to get the most strength underprint the part to make it match the amount of layers you want to coat it with. Then paint your resin over the outside of it and lay your layer in cure it I was curing for about 30 seconds to where it was still just a tad tacky adding the next layer paint it on and so on and so forth. Then do a full cure and now you got your part wrapped in fiberglass which covers all sides and gives you a hell of a lot more strength. It's also a lot quicker.

Very good video! I was asking my self the difference weight between each piece