# CriCri Composite

Discussion in 'Composites' started by ULF, Jun 16, 2019.

### Help Support HomeBuiltAirplanes Forum by donating:

Tags:
1. Jun 24, 2019

### ULF

#### Well-Known Member

Joined:
Oct 24, 2015
Messages:
58
19
Location:
Vienna , Austria
Hi Bob,
Searching for carbon fibers I came across bidirectional non- crimp fabric https://shop1.r-g.de/en/art/192150127-EBA, with unidirectional layers +45°/-45, there I saw, some +45°/-45°, and others -45°/+45°.
Ihave never used such bidirectional stuff, I used plain weave, cut 45°.
Do you know about non-crimp fabric? Which one is better,+45°/-45°, or the other way? Do I have to cover it with thin fabric?

John

BoKu likes this.
2. Jun 24, 2019

### BoKu

#### Well-Known MemberHBA Supporter

Joined:
Aug 16, 2013
Messages:
2,176
2,702
Location:
Douglas Flat, CA
First off, a reminder of my usual warning: I am not an engineer. I do a bit of my own engineering at a very basic level, and I have figured out a regime of composite construction that works well in my remote workshop. But for real engineering, ask Billski or someone like him.

The fabric at your link is what I have heard called a "knitted double-bias." We use fabrics like this in fiberglass, but not in carbon. I think the "non-crimp" feature is because it is not woven, so the fibers stay straighter and have better stiffness. However, I suspect the overall strength and stiffness is probably only modestly greater than conventional woven fabrics. But the weight and price look reasonable, so it is probably worth a try.

Unless you are stressing the material well into its capacity, I think that the order of orientation is probably immaterial; I don't think it matters at all whether it is +/-45 or -/+45.

The 6oz cloth we use a lot of is the 48" 3K Carbon Biaxial 2x2 Twill on this page. The neat thing about it is that it has tiny polyester interstitial threads along the 0 axis so that it does not distort as easily as when you cut +/-45 swaths out of conventional 0/90 cloth. We pre-cut entire 7.5m wing skin swaths, set them aside, and unroll them into the mold when needed. Using this material has reduced the wing skin layup time by more than half compared with cutting individual bias swaths from conventional fabric. It has also made it practical to make each new wing skin set with a new, untrained crew as a training exercise.

Whether you cover the carbon with some other cloth depends a lot on your methodology. Some shops use a thin veil cloth between the carbon and the mold surface. Some shops use a randon-strand veil mat. What we do is use a vinylester in-mold primer (like gelcoat, except filled with microballoons instead of talc so it is lighter and softer), roll on a tie coat of microballoon-thickened epoxy, then put down the first ply of carbon. The in-mold primer gives a good exterior surface on the outside and a good epoxy wetting surface on the inside to eliminate pinholes. The tie coat fills the weave of the carbon to reduce print-through. It's not perfect, but perfectly good for parts with a long in-mold dwell time that get polyurethane paint after assembly.

Thanks, Bob K.

dino, wsimpso1, pictsidhe and 2 others like this.
3. Jun 24, 2019

### Vigilant1

#### Well-Known Member

Joined:
Jan 24, 2011
Messages:
3,551
1,528
Location:
US
A slight diversion: We've discussed the minimum exterior skin strength/toughness needed to withstand the real world. If I've got this right, it sounded like Bob had found that 6 oz CF worked well if Divinycell or other substantial core was behind it. Billski has found that a more substantial skin (17+ oz FG, per Rutan el al?) is typical and needed when using massive 2.2 pcf XPS cores.
What would be the likely problem/failure mode in the real world if 6oz CF is laid onto a solid core wing of typical 2.2 pcf XPS? Divinycell has greater compressive strength than XPS, but the skin (thick or thin) is so much stiffer than either it would seem to be taking the point load of the ball-point pen or errant child regardless of which foam is backing it.
I'm not suggesting 6oz CF will work in this application, I'm just trying to understand why it wouldn't.

4. Jun 24, 2019

### Marc Zeitlin

#### Well-Known Member

Joined:
Dec 11, 2015
Messages:
424
571
Location:
Tehachapi, CA
I've used this material, or at least something extremely similar (knitted +/-45 carbon) on a large aircraft project last year. There is no "other way" - +45/-45 is just -45/+45 turned upside down. If you care which way the first layer goes, just turn it over. In practice, you don't give a crap.

You put a layer of Fiberglass over the carbon if you care about what's going to touch the carbon from a conductivity/corrosion standpoint, or if you want to be able to have a sanding surface that doesn't ever impinge on the carbon.

My \$0.02.

BoKu likes this.
5. Jun 24, 2019

### BoKu

#### Well-Known MemberHBA Supporter

Joined:
Aug 16, 2013
Messages:
2,176
2,702
Location:
Douglas Flat, CA
I think you'll find that if you flip over a piece of +/-45 over, what you see is still +/-45. If for whatever reason you actually need -/+45, you have to turn the cloth 90 degrees. It's like how if you take a tractor prop and put it on the crankshaft backwards, it's still a tractor prop--just not a very good one.

wsimpso1 and sotaro like this.
6. Jun 25, 2019

### Marc Zeitlin

#### Well-Known Member

Joined:
Dec 11, 2015
Messages:
424
571
Location:
Tehachapi, CA
Sheesh. You're right - thanks for the correction . But in any case, whether the +45 or -45 is on the top, for normal wing skin type layups, it doesn't much matter.

It MAY matter if you're trying to make a balanced layup with a lot of plies and you can't turn the cloth 90 degrees...

BoKu likes this.
7. Jun 25, 2019

### wsimpso1

#### Well-Known Member

Joined:
Oct 19, 2003
Messages:
5,623
2,866
Location:
Saline Michigan
You are aware that PU foam when heated releases isocynate and related gases, polyol, and other nasty chemicals. This is a high risk process for you and anybody downwind of your operation. Please do some research on the materials and effects in exposure to temperatures reached in hot wire cutting.

For everyone else, we do not hear about people hotwiring polyurethane foam for very good reasons - it will poison you. Just don't.

Billski

8. Jun 25, 2019

### wsimpso1

#### Well-Known Member

Joined:
Oct 19, 2003
Messages:
5,623
2,866
Location:
Saline Michigan
Listen to Bob. He has way more experience than I in graphite, plus several of his designs are flying. That counts for a lot...

Technically, if you took all of your BIAX cloth off one roll and used it for both sides of both skins on the wings or on both facings on a massive core, you will have non-symmetric laminate that will have a tendency to twist when temperature changes or is curved under wing bending moments. Thing is that the asymmetry is tiny and so is the twist. Here is why. A BIAX cloth is one layer at +45 and the next at -45, with these two plies 0.0045" apart. Then a 1/4" away you do it again. The twist effects go with the square of the separation, so the bending stiffness compared to the twisting effect is (0.250/0.0045)^2 = 3086. You will have to work really hard to even measure the twist effect while you will see the tips bending under flight loads.

Billski

Battler Britton and BoKu like this.
9. Jun 25, 2019

### wsimpso1

#### Well-Known Member

Joined:
Oct 19, 2003
Messages:
5,623
2,866
Location:
Saline Michigan
I am not saying I have found 3 plies of 7 oz UNI to be adequate, the entire world of Rutan canard ships and their many derivatives have found it to be adequate.

The issue with blue foam at 2 pcf is soft and low strength compared to 6 pcf or 8 pcf Divinycel. Impact damage appears to be the big issue, and the deflections that produce the forces that break the foam from the fiber-resin layer are related to EI of the skins. If 3 UNI in glass is adequate on blue foam, then anything else with that much EI should also be OK.

I was using thickness per ply of 6 oz graphite fiber cloth at 0.0075" and E on the bias of 8.4E6 psi (vacuum bagged). One ply has much lower bending stiffness, two plies is about the same as the glass. That was my basis for same tolerance for impact.

Put one ply of anything on Divinycel foam, and it is way sturdier than that same ply on blue foam. Use one ply 6 oz CF biax cloth on blue foam, and I would expect it to be much more sensitive to impact than 3 UNI (glass).

That being said, 2 plies of GF 6 oz cloth is still much lighter than 3 UNI 7oz glass.

Billski

Vigilant1 likes this.
10. Jun 25, 2019

### BBerson

#### Well-Known MemberHBA Supporter

Joined:
Dec 16, 2007
Messages:
11,405
2,092
Location:
Port Townsend WA
I would try a 30" bow saw (for logs) instead of a hot wire. Turn the blade 90° in the bow.
I haven't tried it so I don't know if it would work.

11. Jun 25, 2019

### Marc Zeitlin

#### Well-Known Member

Joined:
Dec 11, 2015
Messages:
424
571
Location:
Tehachapi, CA
Just to clarify, while there are areas of Rutan derivative canards (top of wings, fuselage forward of front seatback, etc.) where there are 3 plies of UNI glass, the bottom of the wings and the majority of the fuselage are 2 crossing plies. So from a damage tolerance standpoint, 2 plies is reasonably acceptable. The only place where it isn't (from a damage tolerance standpoint) is on top of the strake where folks kneel to get into the rear seat - there, damage DOES occur on occasion. 3 plies (one spanwise) is use on the top of the wing for buckling resistance. So depending upon the usage, likelihood of finger/pointy object poking, buckling resistance, etc., some places need 2 plies, and some need 3.

And yeah - DON'T hotwire Polyurethane foam. Some folks did that once at Scaled and I blew a gasket - didn't happen again.

12. Jun 25, 2019

### ULF

#### Well-Known Member

Joined:
Oct 24, 2015
Messages:
58
19
Location:
Vienna , Austria
Hi there,
Short pieces of 2ft, or under give excellent results hotwired. Very accurate cuts.
John

13. Jun 25, 2019

### wsimpso1

#### Well-Known Member

Joined:
Oct 19, 2003
Messages:
5,623
2,866
Location:
Saline Michigan
The Master's Degreed Material Scientist and career long Polymers Engineer in this house (she is my wife) says you are exposing yourself to isocyanates, polyol, and other toxic stuff when you hot wire cut PU or otherwise heat to melting. Please stop...

Billski

14. Jun 25, 2019

### PiperCruisin

#### Well-Known Member

Joined:
Jan 18, 2017
Messages:
91
28
Location:
Idaho
Burning/melting a polyurethane foam will yield hydrogen cyanide, benzene, carbon monoxide, and other fun stuff. So, yes, I concur, hotwire cutting the stuff is simply out.

fyi. Polyurethane comes about when you mix a polyol (organic longish-chain molecule with some OH groups) and an isocyanate. The isocyanate (there are many types) is what you can become sensitized to...so use proper PPE. Of course polyurethanes are very versatile. They can be the foam you are sitting on to high performance paint.

15. Jun 25, 2019

### ULF

#### Well-Known Member

Joined:
Oct 24, 2015
Messages:
58
19
Location:
Vienna , Austria
haha, well once every 5 years, on a windy day, Ill probably survive without prtmanent damage! I always wondered what that smell is.
Many thanks to the master!

16. Jun 25, 2019

### BJC

#### Well-Known Member

Joined:
Oct 7, 2013
Messages:
8,875
5,728
Location:
97FL, Florida, USA
Unbridled optimism and dismissiveness are among the first symptoms.

BJC

17. Jun 25, 2019

### ULF

#### Well-Known Member

Joined:
Oct 24, 2015
Messages:
58
19
Location:
Vienna , Austria
Hi Billski, Bob mentioned you are an engineer, and I wanted to ask for your opinion.
On aircraft, it is not uncommon, to oversize parts sometimes, for practical use, groundhandling and so on.
I did some preliminary calculation about the size of the upper carbon spar capstripe, for 9g.
The aircraft weighs 150kg without wings, that comes up to about 17KN.
With 600MPa published compressive strength of CFK, 17000/600 ~ 28mm² (.043 sq in) of carbon, at the root. Pretty small, but its only a 7ft wing.
Now if I double it, so its an area of 56mm² (.087 sq in). Still not a weight factor.
Now I looked at the original size spar, made of aluminum 2024 T4, with a cap stripe dimension near the root, of 190 mm² (.29 sq in), that is almost seven times that of my carbon cap stripe calculation.
Taken, that carbons compressive strenght is some what double to 2024 T4 (it isn`t really), looks like the original spar could be good for almost 30g, not 9. ?

What would you say, that one should aim for on the capstripe dimension at the root of a cricri composite wing ?

John

18. Jun 25, 2019

### wsimpso1

#### Well-Known Member

Joined:
Oct 19, 2003
Messages:
5,623
2,866
Location:
Saline Michigan
The same stuff is in solvent borne polyurethane paints, and you must use supplied air to be resonably safe with that - no filter masks.

PiperCruisin likes this.
19. Jun 25, 2019

### wsimpso1

#### Well-Known Member

Joined:
Oct 19, 2003
Messages:
5,623
2,866
Location:
Saline Michigan
Some folks will use analytical methods to fully size main structures, then add in another ply for grandma (Scaled is reputed to do this). For the most part, you will not find spars to be grossly oversized - the next smaller increment either failed analytical checks of some sort or failed in test and was bumped to the next available thickness. Most oversize I have seen in parts is the result of either shop expedient or min gage restrictions. I would not assume that the original was randomly oversized - Columban did good work in the original.

Basic sizing of composite spars (like you have done for your caps) is just the starting point on composite flanged beam designs. I wrote on it here and elsewhere:

https://www.homebuiltairplanes.com/...ite-plates-beams-and-bigger-structures.29030/

First problem with basic sizing is usually that folks forget that the centroid of the cap is where that force has to be calculated and they just use wing depth. Start with wing thickness, subtract two skin thicknesses, subtract two glue line thicknesses, subtract two spar outer wrap thicknesses, subtract one cap thickness, divide all that by two - now you have about the right arm for sizing symmetric spar caps. Then the stress at the top and bottom fibers of the caps is still higher than that sizing's assumption. And if you go with asymmetric cap sizes, more cap beefing is needed.

Then the spar designed with basic sizing for caps (compression load / compressive stength) and shear web (shear load / shear strength) will almost always indicate failure at lower than intended loads in the web next to the flanges. You gotta beef up something to get to 2.0 times in the analysis, and usually the lightest way to get to strength is beefing both the caps and the web. I advocate surveying the design space for the optimal mix at each of several stations - Left pin, every few inches in the overlap, right pin inboard, right pin outboard, then every few inches as you go towards the tip until you get to MIN Gage on caps and web. Doing this with FOS of 2.0 might find you still with very favorable weights on the main spar.

PM me if you want to share data...

Billski

Battler Britton likes this.
20. Jun 25, 2019

Joined:
Dec 16, 2007
Messages:
11,405