Welcome, Guest
Username: Password: Remember me
  • Page:
  • 1

TOPIC:

Composite Rib Construction & Load Testing 10 Dec 2013 20:56 #468

exelent...is your dragon going to be hevier or lighter?
russ.


On Sat, Mar 3, 2012 at 2:52 AM, Phil <This email address is being protected from spambots. You need JavaScript enabled to view it.> wrote:


Hi Folks,

I've been a bit quiet recently but that's because I've been busy trying out different rib fabrication techniques and testing them. It's been an interesting learning experience and I think I've arrived at a formula that I'm happy with for constructing production parts for my Carbon Dragon. Please forgive the long post, but I want to give as much information as possible to give a clear picture of what I did and how I did it. Comment and constructive criticism are welcome!

You will find all the photographs referenced in this post in my photo album within this group - It might be worth having these photos visible in another browser tab while you read on...

groups.yahoo.com/group/Carbondragonbuild.../1647894528/pic/list

My initial attempt to build a carbon wrapped foam core rib was an unhappy success - it worked (surviving 8g load testing) but suffered badly from bridging along all the edges, which I had to grind off, exposing the foam core. Although the rib survived 8g load testing it exhibited too much bowing and distortion for my liking. This rib was was made without any moulds by laying a 5mm thick closed-cell PVC foam cut-out of the rib (with caps and internal braces 25mm wide) directly onto a hand-wetted single layer of CF Twill cloth, with another single layer laid over this. A layer of perforated release film, followed by a layer of breather fabric and finally the vacuum bag film was placed over the lay-up, and a vacuum drawn down. Despite my best efforts, and working against the clock (before the epoxy started to gel) I found it impossible to get all the layers to conform perfectly to the foam core and there was an unacceptable amount of bridging on all the internal angle edges.

Given the extraordinary length of time it took me to lay-up this one rib using the wet lay-up method I decided to switch to using the Vacuum Assisted Resin Infusion method instead as it allows you as much time as you need to get the lay-up perfect before you even start mixing the epoxy. The Resin Infusion method, in my opinion, is infinitely superior and safer than the wet lay-up method. I will be using the Resin Infusion method from now on for all my parts.

It was suggested, after my first test report a few weeks ago, that I try increasing the thickness of my foam core (and reducing the depth of the caps and struts) to improve the stiffness of the rib, so I constructed my next root rib test piece out of 10mm thick closed-cell PVC foam board and also reduced the cap and internal strut depth to just 10mm. The weight of the new rib remained about the same as the 5mm thick rib. Again I was plagued by bridging along all the edges where the upper layer of cloth came down to meet the lower layer of cloth on the flat work surface. The foam core was exposed in places after the part was trimmed, but ended up weighing more than the first 5mm thick rib due to the excess resin concealed within the untrimmed bridges - not a pretty sight. See the following images:

Root Rib 10x10mm foam core - no flanges - 1
Root Rib 10x10mm foam core - no flanges - 2
Root Rib 10x10mm foam core - no flanges - 3
Root Rib 10x10mm foam core - no flanges - 4

Surprisingly, when this 10mm x 10mm rib was mounted on the test stand and loaded to 8g it exhibited an unacceptable amount of twisting and bowing - far more than the first 5mm thick rib - although it did survive to 8g loading. See the following images:

Root Rib 10x10mm foam core - no flanges - 5
Root Rib 10x10mm foam core - no flanges - 6

I was not happy with this rib at all... but mostly I wasn't happy with the method of construction, which resulted in unavoidable bridging and excessive weight. A new method of construction was plainly needed if I was to overcome the bridging problem.

For my third root rib test piece I went back to using the 5mm thick foam core, but reduced the cap and internal strut depth from 25mm to 24mm wide and constructed it using a male mould. The male mould was made out of 12mm thick MDF sealed with around 5 or 6 layers of thinned out polyurethane floor varnish - fine sanded between coats to provide a high gloss finish to the mould, which was then wax polished and treated to a coat of PVA release agent. See the following images:

Main Wing Root Rib Mould - Male - 1
Main Wing Root Rib Mould - Male - 2
Main Wing Root Rib Mould - Male - 3

My thinking behind using a male mould was three-fold: firstly I wanted a means to fully encapsulate the foam core; provide the rib with a 10mm wide flange all around its outer perimeter (and a 5mm flange along all the internal strut edges to reduce bowing under load), and finally to remove the problem of bridging... or at least move the bridging to an area of the part that would get cut away during the finishing process. This worked very nicely, as can be seen in the following image:

Main Wing Root Rib Mould - Male - 4

This rib was load tested to 8g (32.55kg) and performed very well, with only a little twisting and bowing of the part under full load. Unfortunately the plywood batten I had glued the rib to (for mounting onto my test stand) delaminated and sent the whole lot crashing to the floor - water everywhere! However... the rib survived!

As this new rib had performed so well I decided to grind off the 5mm flanges from the internal edges of the caps and struts; the edges of the foam core remain fully encapsulated in CF cloth. With that done, and the rib a few grams lighter, I re-glued the rib to a fresh piece of plywood and put it back on the test stand.

The second test of this rib also went very well, again surviving 8g loading with only slightly increased twisting and bowing now that the internal flanges had been removed. My plan now was to increase the load incrementally until the part failed. Unfortunately, again, the plywood batten the rib was glued to delaminated at around 9.5g leaving me with another small ocean to mop up off the kitchen floor! However, I think I have proven that the new rib configuration is more than successful - I will re-test this batten yet again to discover exactly at what load it fails. Watch this space.

On the strength of these tests I now plan to use a 5mm thick core for all my wing, tail and H-stab ribs. The root rib of the main wing has caps and internal struts 24mm wide and I plan to reduce this dimension by 2mm for every rib as you move out from the root - so, rib #2 will have caps and struts 22mm wide, rib #3 - 20mm wide, rib #4 - 18mm wide, and so on down to rib #8 which will be reduced to just 10mm wide. At this point the 5mm thick foam becomes too delicate to to cut and sand without damaging it before lamination. So, at the same time that I made the second 5mm thick root rib I also fabricated a #8 rib for testing - again, using a male mould.

The test loads for rib #8 are 2.81kg at 1g, 14.04kg at 5g and 22.47kg at 8g, but holding the rib in my hands and hardly able to feel its weight I seriously doubted that it would survive to 5g let alone 8g - having run the numbers, I was conscious that I was reducing the depth of the rib caps and internal struts faster than the load was being reduced out along the wing. However, not only did it survive to 5g with only minor twisting, it survived 8g loading without complaint! See the following images:

Wing Rib 8 Load Test - 1
Wing Rib 8 Load Test - 2

I left the rib loaded to 8g (22.5kg) while I went off for a cup of tea and came back to find it still smiling happily, so I continued to increase the load until it finally failed at just over 28kg, a load of 9.97g!! See the following images:

Wing Rib 8 Load Test - 3
Wing Rib 8 Load Test - 4

Satisfied with these results (though not 100% happy with the male moulds, which are reluctant to release their grip on the part being fabricated) I turned my attention to the wing tip rib (rib #13) which requires its outboard surface to be fully covered and the flange around its outer perimeter to point back inboard instead of outboard like the rest of the ribs (the root rib particularly.) I decided to try fabricating a female mould for this part and use the foam core, carefully sanded to an exact fit, to gently press the CF cloth into place.

The results were astounding. Where I had spent nearly an hour trying to get the various layers of CF cloth, peel-ply, flow-media/mesh and vacuum bag film to conform to the complicated internal geometry of the root rib and rib #8, it took me literally two minutes to assemble the laminate stack for the tip rib. Sealing the vacuum bag with gum tape and pleats took another ten minutes, but when I applied the vacuum, the whole part conformed to the female mould perfectly - and I do mean *perfectly*! The resulting part, once it was trimmed and finished was flawless... and weighed in at only 21grams! See the following images:

Wing Tip Rib - 1
Wing Tip Rib - 2
Wing Tip Rib - 3
Wing Tip Rib - 4
Wing Tip Rib - 5
Wing Tip Rib - 6
Wing Tip Rib - 7

Although this was only supposed to be a test part, it is more than good enough to keep and use as a production part.

The female mould has proven so easy to use (though more care is needed to sand the foam core to an *exact* fit) that my next job is to fabricate yet another root rib, but this time using a female mould.

I also want to experiment with using a white epoxy pigment so that I end up with a more heat-reflective (less heat absorbent) part than the standard black carbon fibre. When flying hang gliders in the French Alps during summer, the carbon fibre cross-booms become almost too hot to touch even though they are protected inside the glider's sail - one can only hope that they have been post cured to a high enough temperature! The post curing regime that I have for the infusion resin that I am using takes it up to 80degC. I haven't enquired if this can be increased to >100degC.

The one draw-back of the female mould over the male mould is that you need to fabricate separate moulds for both the left and right wing components, and so doubling your work - no big deal if you're handy and careful with a router! I'll take a few photos of the process of making a female rib mould and report back.

So far, so good - Phil is a happy boy!

Please Log in to join the conversation.

  • Page:
  • 1