Bamboo Bike Frame Strength Test

Above is a video of my homemade frame testing apparatus. It’s far from scientific, but good enough to satisfy my curiosity as to whether or not the frame passes the AS/NZS 1927:1998 Frame & Fork Assembly test. The method described for the test is as follows:

Method

(a) Anchor the rear wheel axle attachment points.
(b) Apply a force of 890 N to the front axle attachment point towards and in line with the rear wheel axle and from the deflection reading compute the energy absorbed in joules.
(c) Should the energy absorption reading at 890N be less than 40 J increase the force until this figure is attained.
(d) Release the loading.
(e) Examine the test specimen for any signs of fracture or permanent deformation. The examination shall include the fork steering tube. The examination for fractures shall be done at ×5 magnification.
NOTE: This test applies equally to rigid and suspended frames. When testing a suspended frame the energy absorbed in reaching an applied force of 890 N can be expected to be substantially more than 40 J.

Theory

Based on my understanding of physics Joules are calculated by the formula: Work(Joules) = Force(N) x Distance(m). So to get the frame to absorb 40J of energy I have to make it move 45mm (40J = 890×0.045). So I set up my rig to shift the frame 45mm up from its resting position, then applied a force to the rear axle that pushed the frame forward by 45mm. The frame was blocked by the end of the rig thou so it couldn’t move, thus the force acts through the frame, causing the front forks to bend a lot!

Result

It seems to have passed the test. There was no cracking noises or breakages. I’ll pull apart the fork tonight and make sure the carbon steerer has survived and inspect the frame for cracks or deformations.

Finishing touches

Today I’ve put the first of the final touches to the frame – a carved bamboo head-tube Cognitive logo. It’s a bit simplified, but that’s ok as it will probably change down the track anyway.

The only things left to do now are two more layers of carbon on the seat tube and then to unwrap all the tape, sand it back properly, and add some final clear coat as a protective varnish layer.

Carbon fibre wet layup – stage 2

Unwrapping the first layer this morning and it’s all gone well. The cling wrap’s left lots of wrinkles in the resin, which isn’t ideal, but for these base layers it’s ok as it’ll all get sanded back. Hopefully by the time I’m on my final layer’s I’ll be shrink-wrapping with better precision.

I’m now onto the 2nd and 3rd layers. Each layer I’m making from a slightly smaller template so the edges of the carbon fibre will gradually taper off rather than being a sharp drop off to where the bamboo is.

Carbonfiberizing

So tonight I’ve started down the road to the final wrap.  I didn’t document last night’s fibreglass layer, but it’s very similar to this process, just with white fabric instead of black. The layer of fibreglass is used between any steel parts to prevent the carbon fibre from corroding the steel (that’s a very long PDF on the topic, probably best you don’t click that link).

So, with an array of carbon fibre, some dish cloth and a penchant for late-night carbonfiberizing I got stuck into it.

First I used the dish cloth as a template around the joint I was wrapping (in this case the head tube). The idea is to use as long a piece of carbon as possible without cutting it. I managed to make something that looks like a space invader. It’s done the job splendidly thou, with very little overlap and good coverage around the awkward tight areas. The comments within each picture explain the process.

Seat post in

I’ve cut, drilled and notched the seat-post sleeve and epoxied it into the seat-tube. Hopefully the carbon fibre will arrive tomorrow and I’ll start the mega-sisal-carbon-bamboo composite process!

Tune in, drop out

I’ve completed the wrapping of the drop-outs with sisal. I’ve also added stay support bars near the bottom bracket and seat-post to give the frame more stiffness. Here’s some pics:

The clearance between the last cog and the chainstay wasn’t something I thought about that much. Luckily for me there’s about a 3mm gap to spare. This will be a single speed bike, so there will be plenty of room anyway.

The chain tension screws were also a very close fit. I actually had to cut away at the joins a bit to make them screw in properly.

That’s a wrap

Started wrapping the sisal twine around the joints tonight. Here’s the process for the seat-tube. I’ve done it in two layers. The first using lengths that run lengthwise down each pole, following the curves of each filleted corner. The second layer then holds these strands in place by wrapping around and around them. I finish it off with a few diagonal turns across the main joint and give it a haircut to neaten it up.

There’s no epoxy on any of this. I’m mainly practicing the wrapping process but am also considering making some sort of mould that I could then inject the epoxy into the sisal and joint as it is now. I’ve tried brushing and drizzling epoxy into twine like this before and then wrapping it in PVC tape to set and it sort of works, but due to the tightness of the wrap and the double layer it might need more pressure than what PVC can provide in order to make its way to the bamboo and not have any air bubbles in it.

Sanding back

Today I’m sanding back the tack epoxy to a smooth finish. I figured I probably shouldn’t have used as much epoxy for this stage as it’s not very strong, due to the micro-fibre filler that’s added to make it more paste-like. The final wraps will use undiluted epoxy along with sisal twine and should be the main force-bearing layers.

It’s messy work sanding this stuff back. You need to wear a mask as the un-cured epoxy is still toxic. Only after it’s properly cured, which takes 4-5 days is it harmless.

Tacky

So it’s taken approximately 3 months but I’ve finally tacked together my first frame!

Here’s the process from start to finish:

Surface preparation

I roughen up the bamboo by using the edge of a Dremel sanding wheel. For the steel I sandpaper first with 80 grit, then follow the same sort of deal as with the bamboo. This will provide more surface area and texture for the epoxy to key into.

I then coat the steel parts in a thin layer of pure epoxy and set them aside to get tack-dry.

Tack glue

Starting with the chainstays, I fill the rear end with epoxy and then squish onto the threaded rod I’ve attached to the drop outs. I then position the bottom bracket and stick the other ends of them to that (forgetting to sand back the epoxy on the BB… oops!?). Next comes the seat-post and the down-tube. The glue’s pliant for about 30minutes so it’s pretty easy to move stuff around if you make a mistake.

I then go to the pub for dinner and a few beers with mates while these parts take hold.

Upon returning and checking that all is well I then attach the head-tube to the down-tube, along with the top-tube to the seat-tube. The angles of the head-tube won’t be exactly 72°, but it’ll be close enough I’d say. I really need a jig that holds the head-tube in place!

Finally it’s the attaching the seat-stays to the seat-tube. Proving harder than it should be I find out that they’re not exactly even in length. I make a few adjustments and compromises on the seat-post angle and set them anyway. If it’s wonky in the morning I can always cut thru the epoxy and try again. Fingers crossed I won’t have to!

Let’s set something straight.

Sharks with frickin’ lasers. That’s what I’m talking about. Ok, well minus the sharks. It’s more like jigs with frickin’ lasers. Close enough. Check out the hi-tech laser-level action below.

The mitres are all done, so now it’s a matter of setting my somewhat cumbersome frankenstein of a jig (can someone buy me a gift voucher for Maytec please?) into place for each pole. I’ve made a series of height adjustable stand-offs to hold up each piece as well as come conical shaped rubbers from some chair legs for the head tube placement. It’s all a bit too dependent on the table being level thou, making it quite painstaking to set.

With any luck I’ll have it all set tonight. Then it’s on to roughing everything up so epoxy will stick to it. Good times.

Mitre’s well show you

Here’s the results of today’s mitreing. Pretty happy with the results so far. Just have the top tube to go and then into the more finicky work of the rear triangle. The main thing I learned today is that getting both sides of the mitres straight and level (ie. so the BB sits at exactly perpendicular) is much more important than getting the curves right. Sure getting the curves right is a good thing too, but it’s ok if there’s small gaps as it will allow the epoxy to fill into the inside of the tubing and strengthen the join.

Choosing the bamboo for your frame

This might seem obvious, but choosing the exact bits of bamboo you’ll use for your bike frame is one of the most important things to do in the frame building process. You need to know a lot of stuff before you can choose the right bits thou! Read on for a few pointers I’ve picked up along my bamboo selection journey.

So you’ve got your geometry right and you know what lengths of each bit of bamboo you need. So off to the shop/woods right? Wrong! You’ll need to know what diameters suit which bits of the frame and then work within a set of tolerances that you know won’t cause problems (or be willing to fix those problems in creative ways!).

The rear chain-stays are by far the hardest bits to work with, followed closely by the seat-stays, then seat-post. The top and down tubes are fairly open to how much trust you have in the strength of bamboo. You could go for skinny poles if you want a retro steel looking frame, or fat chunky one that might weigh a bit more, but are probably stronger, like on an aluminium or carbon frame.

The main problem areas (and questions to find answers to) are:

  1. Tyre clearance – what’s the widest tyre you want to put on your frame?
  2. Chainring clearance – how many chainrings? How many teeth?
  3. Crank-arm clearance – what’s the Q-Factor (width) of your cranks?
  4. Seat-post – what diameter and length will you need? Are you going to use a metal sleeve inserted into the bamboo for the seat-post to go in?
  5. Disc rotor clearance (if you’re using them) – how big are the rotors?

Only once you’ve got all those questions solved are you then ready to go in search of bamboo with the right diameter!

A rough diameter guide

The figures below are a rough guide to what you might need for a road or cyclocross frame. The figures in brackets are what I’ve used on my first frame.

Top tube: 26-36mm (35mm)

Down tube: 35-45mm (41-43mm tapered)

Seat tube: 40mm (the inner diameter needs to be more than the metal sleeve, which if you’re using a 27.2mm seat-post is around 30mm)

Seat stays: 20-25mm* (22mm)

Chain stays: 20-25mm* (25mm)

*Tyre and chainring clearances end up being very precise things, so try to find bits that will exactly match your spec.

What to look for in choosing your bamboo.

So you know what diameters you need. What else is important in your bamboo selection process?

Nodes

I still can’t find out decisively if nodes are strong or weak points in bamboo. Books tell me one thing, the internet tells me others. It’s confusing. If anybody has any definitive information on node strengths & weakness please let me know. What I think is right is that they add strength from crushing forces, but cause weakness from bending forces.

Wall thickness

Bamboo varies wildly in wall thickness from species to species, and due to its tapering nature will often be thick at one end and thin the other. Use the thick ends for areas of your frame you think will be under lots of force I guess. Where are those areas? All over the place!

Straightness

They don’t have to be dead straight, but my feelings tell me that any bends and kinks in bamboo will only make it weaker and more likely to fail when put into the triangle formations of a frame. A triangle with a bent side can easily be crushed!

Roundness

Bamboo is often not round! Lots of the pieces I’ve dealt with are quite oval shaped. This can be a good thing. Oval shapes provide more strength in certain directions. Use them on chain-stays and down tubes to your advantage.

Defects

Wood borers seem to love bamboo and often you’ll find pieces with trails left by these little critters. Most of the time they just eat the surface “skin” of the bamboo and don’t do much structural damage, but this surface is the strongest part or the culm, so if they’ve eaten away large chunks of it, or ring-barked it, be warned!

Cracks, scratches and splinters are also things to look out for. Remember this thing is going to be on a bike for a long time. Find the best bits you can!

Mitre 10

Ever wonder where Mitre 10 got its name from? Neither have I. But mitres are what I’m working on this week.  I’ve printed off my mitre templates from the ever awesome BikeCad and wrapped them around the poles I intend to use for the frame. Here’s a pic:

Tomorrow I’ll use my trusty Dremel to cut, gouge, grind and sand those curves into things of beauty. Joy!

What to do with cracked bamboo?

If you’ve read my previous post you’ll see I’ve cracked a few bits of bamboo.  Not wanting to throw them away I’ve made use of them by splitting them into two and then cutting them down further to form two less-than-semi-circle pieces that can then be glued together. I got the idea from Brendyn, another Melbourne bamboo bike maker, who’s done something like this in his latest build. Sweet stuff!

Below are some pics of what could be an oval shaped chain stay or seat stay. You can see in the first picture where the PVC tape I used to wrap it all up as the epoxy dried has left small amounts of epoxy behind. This would be sanded off later.

Next time I try this I’ll spend more time making sure the two pieces have bigger flat edges facing each other. Bigger area means more epoxy bonding area!

Enter the jig…

Started work today on a flat wooden jig that will hold the frame in place while I tack the joints and generally put it together.
I found an old cupboard door, which to my luck had 33mm holes drilled in it from where the hinges were. 33mm is the size of a bottom bracket ID, making it a perfect fit for the PVC tube which I’ll extend out to hold the BB in place.
I’ve drawn out the centrelines for the frame, to within 1-2% accuracy I’d say. I wonder how accurate I need to be?