Introduction: Strengthening a 3D Printed Part

Picture of Strengthening a 3D Printed Part

As we all know, 3D printed parts produced by our ordinary home type printers are not so strong and when used in place of actual parts, they don't last long. I thought it would be great If it is possible to somehow strengthen the 3d printed parts and worked on the subject a bit.

Strengthening a 3d part may give us the ability to use those parts in place of actual working, load bearing parts. So instead of trying to find an original replacement part for a broken one of our gadgets, robots, vehicles, mechanical toys or in any mechanical tech project we develop, we can simply go with the 3d printed parts.

The need for me to develop this instructable arose in fact when i bought some bicycle accesories from ebay. Recently I've bought two cycling front pannier bags . One for my gf's bike and one for mine. The bag's support was suitable for 1" handlebar thickness and barely fitted to my gf's bike. And didn't fit to mine as my bike had a 31.5mm thick handlebar.

I planned to manufacture the parts with matching dimensions to my handlebar by 3d printing them.

Step 1: The Part Required

Picture of The Part Required

This is the disassembled part which I planned to produce by 3d printing.

To produce a 3d printed part, you need to have the 3d model files for it. So you need to either find and download them from the internet or get to work and model it yourself. Way 2 was the only choice I have, as 3d model for a specific part like this is impossible to find on the internet and I needed to modify it to fit my handlebar anyway.

But as very well known, 3d printed parts are not so famous with their strength. I've thought that an ordinary printed part willl not be strong enough to work properly at my bike, with heavy stuff in the bag, going on bouncy roads. So I thought that it will be necessary to strengthen the part somehow.

In my previous boatbuilding experience several years ago, I saw how much some layers of fiberglass and epoxy give extra strength to a thin plywood part, enabling it to withstand strong forces easily.

So I decided to use fiberglass and epoxy to do the job.

First plan was to model the part, do the necessary modifications on the dimensions and cover all around it with fiberglass fabric and epoxy. So I started modeling my first design for the part.

Step 2: The First Design

Picture of The First Design

I ended up with a part like this. I've arranged necessary details with required dimensions, and overall enlarged other dimensions to make it a bit bulkier. I've enlarged the surfaces for them to be able to carry reinforcing fiberglass fabric.

I also made grooves on all sides for the epoxy resin to fill and give the part an extra strength. I've modeled holes that go through the part from one side to the other, to connect the sides together internally, provide a link with fiberglass and epoxy to support from the inside.

I've printed the part using an Up printer and tested whether it fits to other original parts.

I choose a lattice type fill while printing leaving quite emty space in the printed part. This normally should reduce the strength of the part considerably but it was done for a purpose. I've drilled several more holes on the printed part, between the 3d modeled ones for the epoxy resin to flow inside part, giving the part much greater strength.

Step 3: Applying Fiberglass and Epoxy to the First Design

Picture of Applying Fiberglass and Epoxy to the First Design

This is the dirtiest part of the job. Epoxy and fiberglass is hard to work with. Sticks to everything and very difficult to clean.

First I took several fiberglass threads from the fabric and sew the parts from one side to the other to provide a glassfiber reinforced connection between the sides.

Then I laid the parts on a piece of nylon as epoxy fortunately doesn't stick to nylon well. Then applied epoxy resin to the surface, put the fiberglass fabric cut to required shape and size over the wet surface and applied more epoxy on top. Until the fiberglass fabric becomes transparent.

Repeated this for each side on all parts and left them to cure around a day after each application. Took a while for the parts to be glassed completely.

Step 4: More Epoxy for Added Strength at the Sides

Picture of More Epoxy for Added Strength at the Sides

I've poured some more epoxy to the side grooves of the part for sideways strength.

Step 5: The Second Design

Picture of The Second Design

While making the first design, the idea was to partially fill the inside of the part with epoxy thru the extra holes drilled strengthening it from the outside by covering it with epoxy soaked fiberglass. While printing the first part, this second thought came to my mind to 3d model the part as a bare shell only and build the inside structure with epoxy and fiberglass later on.

So I started modeling this while also still waiting for the layers of epoxy-fiber to cure over the first one. I wanted to manufacture both designs and compare them in strength and ease of production. I also used different approaches towards filling the part 1 and 2 of the second design( two parts were necessary for mounting the bag to the bike).

As seen, the second design is an empy shell with 1mm wall thickness. The round part is separated from the big part to make the filling process easier. It is possible to screw fit the round part back to the main part during the curing of the epoxy fill inside.

Step 6: Filling the Second Design With Epoxy and Fiberglass

Picture of Filling the Second Design With Epoxy and Fiberglass

First I shredded some amount of fiberglass and stuffed them into both of the parts until they are more than half full. Then started pouring small drops onto them and applying a bit of pressure to soak them.

Step 7: More Filling Into the Design#2 Part#1

Picture of More Filling Into the Design#2 Part#1

Continued this until all fabric inside is wet and then added more fabric and epoxy step by step making the inside of the part is full. Then Screwed the round part on top of it. I continued filling the part, rising the level of epoxy-fiber mixture to the edge.

I did the same to the smaller part.

Step 8: Filling the Second Design's Part 2

Picture of Filling the Second Design's Part 2

For this part, I followed a little bit different approach, I didn't fill it to half full from the start. I've placed only one layer to the bottom of it, than a second layer covering both the bottom and also the sides. Wetted these two layers with epoxy until they are completely transparent and continued adding layers one by one and appying pressure to get them wet with epoxy. Added a bit more epoxy for every 2-3 layers.

With the previous mehod, the part was filled with more epoxy and a little bit less fiberglass fabric. With this method, the part is filled with a lot of fiberglass and less epoxy in between the layers.

Step 9: Left All the Parts to Cure.

Picture of Left All the Parts to Cure.

I checked all the parts from time to time and filled the spaces which were formed due to epoxy flowing inside the gaps. Added more fiberglass fabric where necessary. Time for the parts to completely cure and harden is around a day. There are faster epoxies on the market or you can add the hardener more than necessary to speed up the process but I don't recommend as it gets even more difficult to work and disables the epoxy to flow inside and fill the empty spaces.

Step 10: Completed Parts and the Assembly

Picture of Completed Parts and the Assembly

Well the completed parts turned out to be rather nice, especially with the second design. I assembled them along with the existing parts of the pannier. They matched perfectly. Ready to be mounted on the bicycle.

Step 11: Removing the Plastic Parts

Picture of Removing the Plastic Parts

This step was not intended to be done in fact. I was planning to use the 3d printed parts with the reinforcing inside them. But while I sanded one of them, the printed skin at the outside of one corner just broke off and the rest peeled off very easily. As it provided no extra strength to the part, I removed nearly all of it, converting the part into an epoxy-fiberglass composite part.

Step 12: Results

Picture of Results

I've used the parts during a 20 days trip around Sweden, Poland, Germany and Denmark without problems. The front bag was absolutely full with food during the whole trip and there were a lot of bumpy roads in the path too. The parts worked great and they are still rock solid. I've listed some results from this study below.

1-Interestingly, epoxy doesn't stick to 3d printed ABS plastic well. So extra holes and negative details that provide holding points for the cured epoxy is mandatory.

2- Covering a 3d printed part is much difficult than simply filling an empty part. It takes a lot of time. In the end this produces a lighter part. And there is certainly a great strength improvement over a 3d printed part without reinforcement. It is easy to see this, comparing a reinforced and non reinforced parts by simply trying to twist them by hand. The reinforced part is absolutely rock solid while the other is not.

3- The second method of reinforcing is much easier and produces a much neater result. There is nearly no need for cleaning after curing. And the produced part is like granite. :)

4- I've tried two approaches with different amounts of fiberglass, when handled, they both give the same feeling for strength and solidity. But the one with more fiberglass should be able to bear more weight before breaking I strongly guess.

5- The second design produces a heavier part, as it is completely full with epoxy and fiberglass. Removing the abs plastic parts surrounding the epoxy reduced some weight. But it also changes dimensions of the completed part. So dimensions must be adjusted during design, taking into the change the lack of abs plastic will bring.

6-Grooves and empty spaces can also be put into the design of the second part to prevent epoxy to fill into them, thus creating a lighter part.


jpfalt (author)2015-11-29

I have been using 3D printed ABS parts since about 1994 and made functional parts for landscaping equipment. I found three methods that vastly improve the performance of the material.

1. The first is to do a 10 second dip in acetone. The ABS equals nylon in strength in the x and y directions, but the z direction is about half strength, like plywood. The acetone dip fuses the individual abs strands into a solid and increases the z direction to full strength. This also has the effect of making the part water and air tight by sealing together the pores in the deposite strands of abs.

2. On rotary ad hoop parts, I leave a groove and drill small holes if required and wrap the part with two wraps of .041 horrible fright stainless steel lock wire with a twisted end using lockwire pliers and clipping to leave about two full twists. I made a series of backpack blower impeller fans up to 8" diameter that successfully ran for extended testing at 8500 rpm.

3. Fiberglass or epoxy reinforcement can be applied directly to the ABS by painting the surface with acetone and laying on the fibers. I have also wrapped the fibers and then painted with acetone to melt the ABS underneath to capture the fibers. Finally, I keep a bottle of ABS dissolved in acetone to paint over fibers ir metal inserts. The acetone fully evaporates overnight and leaves full strength ABS over the top of the .reinforcement.

A final method that I don't count is to make a metal part and 3D print a wrapper. I did this for a bakelite reversing handle on a large disc sander. The original got smashed during a machine move and the replacement used an original metal core and printed two parts that were assembled around the core and fused with acetone.

I found three different things that really

DaoD1 (author)jpfalt2017-02-23

I am considering doing your method 1 then coating it with method 3 with my 3d printed part. Is it possible to do so?

jpfalt (author)DaoD12017-02-23

If I'm doing reinforced parts I use both methods 1 and 3 on the same part. Method 1 makes the ABS uniform strength in all directions and method 3 adds additional strength in whatever dimension the fiber strands run.

Something I will do in addition if I want a smoother surface after fiberglass or kevlar reinforcement is to pain the part with ABS dissolved in acetone to build an additional layer of plastic for sanding and finishing. I use a small glass jar about 2/3 full of acetone and add the ABS printing filament until the solution starts to thicken. Then I just paint it over the surface and it fuses to the part.

DaoD1 (author)jpfalt2017-02-23

Thank you very much! I will try to see how much stronger my part improves.

I am doing a project at school that is using post treatment to make my 3d printed tensile bar the strongest. I looked everywhere on the internet and came across your comment. Thanks a lot for sharing!

puneethrj (author)jpfalt2015-12-10

This is indeed useful. This deserve to be an instructable rather than a comment.

hobbyman (author)jpfalt2015-11-30

This is very valuable information, thanks for sharing it with us.

smrat alleck (author)2015-12-10

Thanks for sharing; this is a clever idea that I could see being useful in many situations. Glad it held up for your bike tour!

JON-A-TRON (author)2015-12-10

Great idea! I bet this part is rock-solid.

pkrueger (author)2015-12-10

I have thought of just using them to make sand castings....

drinkypoo (author)2015-12-10

All I want for christmas is a 3d printer that can print me molds for lost wax casting. I haven't been able to find any inexpensive printers/extruders which claim to be able to do the job.

spaceman874 (author)drinkypoo2015-12-10

i have used PLA to do that. it works. the temps are just a bit higher and it takes a bit longer for the PLA to have evaporated. works a treat :)

GergelyD (author)2015-12-10

Nice job, I'd like to add some ideas you may find useful.

working as a propmaker for the film industry, therefore usually all I
make are 'prototypes' within a very tight timeframe. Some of them must be strong some of them just

I rarely use 3D printing since the process is still unacceptably slow.

sometimes I make layered slosh casts (in a silicone mold) for the nice
surface which is reinforced from the inside (regarding a one sided
mold). For this I use the smoothest possible glass matting (like silk)
with epoxy resin and fill the rest with micro-fibre filled Polyurethane

In your case I'd use Polyesther resin since it bonds to ABS
(since Polyesther's and ABS' main agent is Styrene), and or the
Polyurethane casting resin mentioned above. Sensitizing the ABS surface
with Acetone is well tested method too like an other commenter wrote.

I'd cast a block from the same material and mill the piece out on my
CNC. Since you have to make the 3D model first on your computer all you
could gain with this method is time and surface finish quality. After
having the first piece made, I'd mould it in silicone and cast as many
pieces later I'd need...

I buy my stuff from PS-Composites if anyone care.

moein (author)2015-11-29

why you dont try cnc milling instead of all these hard work ? !

there are tons of parts and designs that can simply be made with machining and milling.

i am a fan of 3d printing but i think earlier methods still work nice in many cases. you have spent hours and and used various materials in a time and energy consuming way. as i see the part has a rather simple design that can be done by any cnc machining. you may say cnc milling machines are way more expensive than domestic 3d printers and a few people may have access to them. but as i can tell you a domestic low power cnc milling machine can to this nicely. plus working with fiberglass and resin is not that simple and needs some facilities. a benefit of 3d printing vs milling and machining is its elegance and ability to print complicated shapes and volumes. if i want to add fiber glass to my designs after i printed it, i am limited to simple and crude shapes and all 3d printing benefits will be eliminated.

i think you are just trying too emphasis on including 3d printer in your projects because it is cool maybe. otherwise there are far better solutions for this issue that do not need a 3d printer.

hobbyman (author)moein2015-12-09

you are absolutely right. We have two 3d printers and and two CNC milling machines of different sizes at school. Hired a mechanical engineer to be able to run the CNC machines last year. He worked for a full month just to mill a part with CNC, there were big problems with the machines. He managed to fix the problems of the small one(size of a 4 person table) and gave up working on the big one (triple the size) saying that it had an inherent design flaw. Later the firm we bought the machine from also said the same thing and proposed a method to fix the problem that costs several thousand dollars. Well, it is a bit difficult to arrange funds so it is still broken. But 3D printers are easy. we just click print, and they print. Would have loved to use the CNCs as well but as i've said. they are much more difficult to operate.

jjsorensen (author)2015-11-29

Another option would be to make an epoxy cast. These kits go for about $50 at Hobby Lobby. You could have skipped all the modeling work entirely.

Rule 34 applies to a lot of small parts like this. Your local bike shop probably has a box of these. Mine usually just gives me stuff like this.

That being said, this is an awesome way to create a epoxy cast for something you don't have other resources for. A thin-walled mold would be pretty cheap and quick to print. Thanks for the info.

hobbyman (author)jjsorensen2015-12-09

thanks for the info, didn't hear that before could you give a link for a sample epoxy cast?

ronjohnmartin (author)2015-11-29

the Cool stuff. I was wondering if you could just modifly the infill to make a pattern so the epoxy could just flow in between the gaps and the infill would act as the reinforcement.

hobbyman (author)ronjohnmartin2015-12-09

Thanks Ron, I've tried to do just that in my first version. I've printed with a loose filling and drilled holes at the outer skin and tried to let the epoxy flow into the print, hopefully making it stronger. And yes it is absolutely stronger than it without the epoxy fill. In fact I plan to use and waste (break) my first version parts in an experiment of weight bearing, to see how strong they are.

clazman (author)2015-11-29

I wish we there were more inventors with mechanical engineering backgrounds.

You are not strengthening a 3D printed part, you're just using the fancy 3d printed part for a simple form.

In bending and portion it is the outer fibers handle the majority of the load. with the stress going to center at the center of mass. The moment of inertia used to define the stress. The section modulus used to define the deflection. I leave it to the reader to the actual relationships. Please see:

In shear and tension the entire cross section carries the load equally. Stress is simply defined by load/area.

Please study the original part. The designer utilized excellent engineering design using the concepts I have mentioned ever so briefly.

Why do you think that the handlebar is a tube?

jimrittenour (author)clazman2015-11-30

Dude, go to a mechanical engineering site.

clazman (author)jimrittenour2015-12-08

Not sure what you are implying. Maybe you didn't understand what I was trying to say. I must admit that my submission left a lot to be desired hitting only very basic engineering concepts. For me to discuss more coherently would fill a text book.

jimrittenour (author)clazman2015-12-09

What I meant was nobody here is trying outsmart the rules of physics and engineering.

When you wrote "'re just using the fancy 3d printed part for a simple form," folk might infer that you judged the spirit and fun of this guy's project as wasted; as if here were expected to create something revolutionary.

Your certainly correct about principles and I have no doubt you could fill a textbook on engineering concepts; that's why I said to go to an engineering site - if that's the level of expertise you expect to find in a site like this.

There's plenty of heady stuff on instructables, but nobody ever wrote "you just used a fancy super-computer to make a Raspberry, loser."

If you wanted to make a critique of the project, you might have offered some tips to make it even better using simple improvements along with a principle or two that wouldn't need a degree in structural engineering.

Lastly, I bet it works just fine as it is and somebody else might have fun making or improving it.

For that, it's a nifty instructable.

neurostinger (author)2015-11-30

Thank you for sharing your results! I think we all learn a lot from experiments like this.

I think another interesting addition to this could be using a foam core in the product.

I mean: 3d print the outer shell, put a layer of epoxy with fiberglass on the inside, cut a foam core to fill the inside and then finally close the product with another layer of epoxy on the foam core.

That way you save epoxy, weight and probably have almost the same strength because the outer shell usually bears most of the load.

hobbyman (author)neurostinger2015-11-30

this is a very good idea to place some foam inside. I'll try it next time.Thanks for sharing.

brian32768 (author)2015-11-29

When working with epoxy, it's recommended to wear gloves and a respirator mask. I have a friend who can't even walk into a shop
where epoxy is in use anymore because just the smell makes him break out
in a rash.


"You may become sensitized to epoxy after many exposures or just one. It
could take ten days of exposure, a month, or even years. It is best to
avoid all exposure because you cannot know ahead of time how much you
can tolerate before you become allergic."

If you machine or sand parts with plastics and glass you will be releasing lots of nasty dust, you need to be aware of that too.

hobbyman (author)brian327682015-11-30

This is valuable information Brian, Thanks for sharing. and sure yes, dipping small parts into epoxy will strengthen them for sure. there are some very good info above at jpfalt's comment about using acetone instead of epoxy.

Mark 42 (author)2015-11-29

At my work, they 3D print mineral & metal filled plastics. They can 3D print metal parts if they are not too large.

lucpet (author)2015-11-29

You should be using petg for strong parts

eprincipe01 (author)2015-11-29

I wonder if you could actually pause the build and place layers of fiberglass into the build to create a composit matrix. It may require adhesive and curing time and a cheat on adjusting the build platform. Getting complicated now. Might be better to make the object from "N" slices and then insert the fiber reinforcement and "glue" it up using a plastic adhesive. An oven cure might be an alternative way to mix and bond composit

brian32768 (author)2015-11-29

Lots of stuff to think about here! Lots of interesting comments too. I've made parts for bicycles so far just by printing them, knowing they will break eventually. It had surprised me how durable plain old PLA can be for things like head and tail light brackets for example, but I always plan on it breaking.

I bet some small parts could be strengthened sufficiently with a simple epoxy fill (leave out the glass). What do you think?

gunguru (author)2015-11-25

I wonder if the hollow form printed part could be filled with layers of ultra thin sheet metal sandwiched between steel bedding epoxy (like J B Weld or the Harbor Freight copy)? Then the part could be machined, drilled, or tapped just like a metal part.

gunguru (author)gunguru2015-11-25

If the melting point of the ABS is low enough the part could be printed and packed into a sand form and then molten aluminum could be poured into the form making a perfect cast aluminum part. (Kindda like the lost wax method of jewelry making) the King Of Random has a great youtube video of the technique.

Sigmaz_1 (author)gunguru2015-11-29

The best material to use for a lost casting process would be PLA. Since you can't directly pour into the plastic pattern like with lost foam the pattern must be invested in a block of plaster/sand and burned out in regulated kiln. Then youd be able to pour aluminum into the block. Why PLA? ABS gives off toxic smoke when burned, the burn out process takes up to 12 hrs and all that time youd be fogging the neighborhood. Its just a better idea. I cast an intake manifold with this process. YouTube: Sigmazgfx

rundmcarlson (author)2015-11-29

You should mount both of your designs and hang weights from them, progressively increasing the weight until they break. Mark down the weights at which the parts start to bend and when they break. Then you will have definitive numbers for your outer covering vs solid epoxy part. Just "feeling stronger" doesnt always mean something is stronger, and if the coated part is close in strength to the filled part, you could save a lot of epoxy.

arpruss (author)2015-11-29

I wonder if tossing some scrap metal - screws, nails, etc. - into the epoxy might increase strength.

proileri (author)arpruss2015-11-29

Epoxy resin seems to be very forgiving about materials, so you could use anything from sand to wood dust to fill it up. Seems that fibre-like materials like chopped fibreglass give best result, though, as they form sort of a interwoven grid structure. Metal could be used, but might be a bit expensive per volume.

MattGyver92 (author)2015-11-25

I wonder if a soluble material like PVA or HIPS might work for dissolving away the 3D printed parts after being filled with resin/fiberglass. If it held up long enough for the resin to cure, it would be easier than manually removing the material.

alames (author)2015-11-24

Great idea, by the way. Makes a 3d printer much more useful if the parts can be rugged. Printing a shell is faster, too.

alames (author)2015-11-24

You complain of epoxy being messy and difficult to clean up. Try methyl (denatured) alcohol. It can be used as a thinner, so epoxy pours easily and can be used like paint, and easily cleans up uncured epoxy.

NanoRobotGeek (author)2015-11-21

This was a really interesting instructable!! Thanks for doing this and especially for sharing. Ive heard filling a hollow print with hotglue also helps strengthen a part but this method seems a lot sturdier.

hobbyman (author)NanoRobotGeek2015-11-24

thanks for the feedback. Yes hot glue also works to a degree but of course, glass-carbon-fiber + epoxy is in a totally different league. When applied properly, it produces results lighter and stronger than steel. Well at least the supercar makers say that.

Stryderx (author)2015-11-23

I think most of 3d printer parts are just good for making molds you could make sand mold and cast aluminium for instance.

hobbyman (author)Stryderx2015-11-24

aluminum parts look really great, I wish I could have done that but that kind of metalworking is out of my repertouire for now :)

maybe in the future.

rickdod3 (author)2015-11-24

Nice job! Once I seen that you intended on using fiberglass to strengthen the plastic I knew that it would not stick...however...using the print as sort of a "mold" is a fantastic idea! Something else you can try in the future to strengthen the resign would be to use play sand and or bits of rope. It will make the fiberglass/resin SUPER strong. Just a tip for the future :-)

hobbyman (author)rickdod32015-11-24

Thanks for the valuable feedback. I've tried to overcome that abs-epoxy not sticking problem by modeling multiple holes on many places of the first model for the epoxy to reach and bond to the other side. That way the epoxy surrounded the part from all sides hopefully making it much stronger. I was planng to also insert some glass fibers into the holes but gave up after inserting 2-3, that was a bit too much labor intensive. I'll try to perform some force and weight test with the first part to see how strong it is. I also think that there might be some kind of epoxy or polyester product with small fibers already put into it during production. Then reinforcing the 3d printed part will be as easy as squirting the resin into it. hopefully :) .

Flecks (author)2015-11-24

Very clever idea! Maybe to make the filling of the part with fiberglass easier, you can try using a CSM fiberglass mat. It's already chopped, and may be easier to fill the part with

proileri (author)2015-11-24

This was something I was thinking about as well.

Additional thoughts: You could always make a casting mould out of water-soluble support filament, cast it and throw it into an ultrasonic cleaner. Also, if relatively thin parts are needed, you could just brush on a small layer of epoxy.

Having a 3D printed filling in solid parts might help you find an optimal balance of 3D printing and brush-on epoxy. You could even inject closed sections with epoxy and slosh it around to spread it, using a process similar to rotary casting.

bluenoiseMF (author)2015-11-24

I really like the way you've shown the various ideas for adding strength. I'd not considered using the printed part as a mold, but it's certainly something I will keep in mind going forward. Thank you!

Yonatan24 (author)2015-11-23

Or just 3D print your own epoxy! Nope :(

jfarsdale (author)2015-11-23

Thank you for this information. I am looking at using 3d printed molds for polyurethane parts. my first print was very large. I see now I only need to use some shells to make it work. I think a lost pla mold would work great for this method.

About This Instructable




More by hobbyman:Multi-Mode Table Clock With Graphics LCD ScreenSteampunk Solar Powered Air Conditioner and Identity Card Holder Ghost Busters PKE Meter
Add instructable to: