Composites vs aluminum

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,944
Location
NJ
But the cowling, wing tips, root fairings, and other compound surfaces may still be best translated into composites because they are compound curved surfaces.

Billski
Soft aluminum and English wheel or sand bag and mallets, even welded/soldered/filed, might still compete with the composites. Also innovative design might make all of them flat wraps, and then hard aluminum and rivets would work. These items are not considered structure, although certainly do not want your cowl to fail and hit the prop.

To me it is builder preference being most important. What do you want to work with?
 

Little Scrapper

Well-Known Member
HBA Supporter
Log Member
Joined
Jan 3, 2014
Messages
5,894
Location
Wisconsin
Yesterday my son and I went to the bike store to get a new tire for his unicycle. While waiting, I picked up a couple CF frames. They are incredibly light. I'm baffled by how light they are.

Again, plumber talking here, but I suspect it certainly can be lighter and stronger than aluminum but it would take some pretty serious engineering.

Those carbon fiber bike frames had steel bungs molded in the frame quite seamlessly. Amazing product.
 

autoreply

Well-Known Member
Joined
Jul 7, 2009
Messages
10,763
Location
Rotterdam, Netherlands
Yesterday my son and I went to the bike store to get a new tire for his unicycle. While waiting, I picked up a couple CF frames. They are incredibly light. I'm baffled by how light they are.

Again, plumber talking here, but I suspect it certainly can be lighter and stronger than aluminum but it would take some pretty serious engineering.

Those carbon fiber bike frames had steel bungs molded in the frame quite seamlessly. Amazing product.
For bikes; absolutely. About the worst shape to "carbonize" and I speak from experience.

Wings are a lot simpler. Nice smooth monocoque potential shape. Just throw out the rest*; no need for it, except controls and maybe the wing attachment.

By rest I mean ribs, spars, doublers etc.

Obviously that still requires some significant engineering effort. But it can be a lot simpler than an alu wing.
 

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,944
Location
NJ
Obviously that still requires some significant engineering effort. But it can be a lot simpler than an alu wing.
The engineering for aluminum has numerous texts with worked examples (actual calculation of sub-assemblies). Maybe I need to see a similar texts for composites. I believe that much of the data is for autoclaved parts, which is usually more than a homebuilder does. The composites are always being improved and there are more resins every day with older ones being retired. Aircraft Aluminum and steel has not changed much since the 40's and much of the data on it is actually from that period. I also marvel at the CF stuff but having worked the "top of the line" with the AV8-B (Harrier) I also know those results would be much more difficult to achieve without the "pre-pregs" and 250 PSI high temperature autoclave.
 

Laromin

Member
Joined
Mar 16, 2015
Messages
6
Location
California
Some of the most lightweight applications of hand layup carbon can be seen with RC handlaunch gliders.
https://www.rcgroups.com/forums/showthread.php?2832791-Introducing-the-Vortex-3
These gliders weigh under 8 ounces RTF with a wingspan of 59 inches.The wings/tails are solid core Rohacell 31,
and the wingskins use 1.5 to 2.5 oz/sqyd spreadtow or special biax carbon fabric using IM or HM carbon.
Let me repeat:Every single elliptical wing core is machined from a solid block of foam.All RC gliders are now painted
in the mold.
 

cheapracer

Well-Known Member
Log Member
Joined
Sep 8, 2013
Messages
6,294
Location
Australian
Again, plumber talking here, but I suspect it certainly can be lighter and stronger than aluminum but it would take some pretty serious engineering.
For the ultimate complete structures that people like Autoreply and Boku do for sure, but there's many simpler applications also available to homebuilders, even as simple as a dash panel.


Aircraft Aluminum and steel has not changed much since the 40's and much of the data on it is actually from that period. .
The consistency and purity of aluminium and steel has increased quite a bit which is great because you have an extra safety factor following the same data.


fig4.gif
 

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,944
Location
NJ
:Every single elliptical wing core is machined from a solid block of foam.
The math of scale vs full size ....volume is a cubed relation to linear dimensions, in a full size aircraft that would be a lot of extra weight of foam.
 

autoreply

Well-Known Member
Joined
Jul 7, 2009
Messages
10,763
Location
Rotterdam, Netherlands
The engineering for aluminum has numerous texts with worked examples (actual calculation of sub-assemblies). Maybe I need to see a similar texts for composites. I believe that much of the data is for autoclaved parts, which is usually more than a homebuilder does. The composites are always being improved and there are more resins every day with older ones being retired. Aircraft Aluminum and steel has not changed much since the 40's and much of the data on it is actually from that period. I also marvel at the CF stuff but having worked the "top of the line" with the AV8-B (Harrier) I also know those results would be much more difficult to achieve without the "pre-pregs" and 250 PSI high temperature autoclave.
The math is the same. The solutions are different.

Take a spar. Calculate I and stresses in the caps. Calculate local shear. Check web and caps for buckling.
A CFRP spar is exactly the same math, except that buckling for the web (and often) the caps is avoided by simply making the core thick enough.

Pre-pregs are certainly stronger, but not much stiffer, so their overall weight reduction isn't nearly as much as one would expect at first sight. The LBA allowables are achievable by any decent "amateur" and have 35 years of proven history, so I wouldn't sweat those too much if you're building technique is solid.

If we get away from "black aluminium", it gets simpler. Take a monocoque, sparless wing. Simply calculate local bending and shear in all 3 directions, check stresses and do a single buckling check on the part of the skin with the highest stresses. Thicken the core there enough to not buckle and you're done with the calculations.

Here's an example that covers about a third of the math for a full wing. You obviously have to run it through multiple load cases (flaps up/down, ailerons up/down, gust, combined loads etc, which means that local lift will vary non-linearly), but that's simple iteration.
wing-example.jpg
 

lathropdad

Well-Known Member
Joined
Sep 12, 2007
Messages
47
Location
Indiana
I am absolutely not an expert on composites but I have decades of experience both in working with composite materials and actually doing the layups.

Based on my experience, I would not recommend using carbon any where in the cockpit area that is not necessary for structural purposes. In a crash, carbon shatters, like glass, into chards .010" or less thick with very jagged edges. Any cut will also involve carbon bits in the cut itself. Doubly Ugly.

A couple ways to think about composites: each layer, be it carbon, Kevlar or glass is something between .008 and .010 thick when properly wetted and then vacuum bagged. Strength wise it is about the same as aluminum of the same thickness. For penetration, single laminates (multiple layers) are way stronger than cored panels. So for something close to .032 aluminum, you need 3 to 4 layers, bagged into a single panel.

For years, I have included a single layer of Kevlar between layers of glass for any place I intend to have fasteners. Example is the nose bowl on my Bearhawk 4p. The mounting flange is Kevlar and glass that has been added to the edges of the nose bowl where the cowl fastens. For protection, we use 8 layers in the body work surrounding the driver.

Where weight in an issue, we now use single layers over a core. This is race car stuff where carbon is restructured to crush structures only.
 

autoreply

Well-Known Member
Joined
Jul 7, 2009
Messages
10,763
Location
Rotterdam, Netherlands
  • Like
Reactions: BJC

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,944
Location
NJ
Check web and caps for buckling.
A CFRP spar is exactly the same math, except that buckling for the web (and often) the caps is avoided by simply making the core thick enough.

The LBA allowables are achievable by any decent "amateur" and have 35 years of proven history, so I wouldn't sweat those too much if you're building technique is solid.
Thicken the core there enough to not buckle and you're done with the calculations.
Are the LBA allowables the minimum loads or do they relate to CF strengths, I was not able to find any specific information on Composite Design in LBA.

When checking aluminum for buckling there is much data and charts and funny formula....do you use these same data,? or do you rely on Fcy and or Euler as your buckling allowable.

I also have the impression there is no Fcy, or it is only very close to Fcu....the stuff does not yield but just breaks, in fact I have seen specification for CF with no Fcy listed. Maybe you rely on deflection calculations for your allowables

Basic stuff but it is hard to pick out of the texts sometimes.

Fcy= Compression Yield in pounds per square inch or metric units
Fcu= Compression ultimate failure in pounds per square inch or metric units
CF= Carbon Fiber
LBA=http://www.lba.de/EN/Home/home_node.html;jsessionid=EC4C59286C862641100D8694BB1F786F.live11291
 
Last edited:

autoreply

Well-Known Member
Joined
Jul 7, 2009
Messages
10,763
Location
Rotterdam, Netherlands
AMC VLA 572 (b)
https://www.easa.europa.eu/system/files/dfu/CS-VLA Amdt 1 combined.pdf

Multiply by 1.5 to get ultimate allowables. Unlike in the US, in Europe you use an additional KD factor with composites instead of a higher safety factor. Two times a half vs 3 times a third.
KD typically is 1.15 for established manufacturers, so 1.5*1.15=1.725

So that'd be 400 MPa*1.5/1.725=348 MPa as treshold yield stress
400*1.5/1.15=522 MPa as treshold ultimate.

The American way is way simpler and less bookkeeping; 600 ult, 300 yield for these kind of numbers.
 

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,944
Location
NJ
Thanks a quick look came up with this ....numbers look different. But I do not speak metric very well.

csvla allowables.jpg

edit: more internet (google) research 40 dkn/mm^2=400 mpa=58k psi
348 mpa= 52k psi. yield

Is that tensile or compression yield, in aluminum they are different by a significant amount.

At 55% the weight of aluminum certainly weight savings is possible. 7075-t6 has a compression yield of up to 68k psi

However looking at the bucking and crippling chapters of Bruhn, often the crippling allowable is a fraction of the Fcy based on test data, and developed formula.

How is that handled in the Composite world? "just make the core thicker" sounds like TLAR
 
Last edited:

autoreply

Well-Known Member
Joined
Jul 7, 2009
Messages
10,763
Location
Rotterdam, Netherlands
Thanks a quick look came up with this ....numbers look different. But I do not speak metric very well.

View attachment 64615

edit: more internet (google) research 40 dkn/mm^2=400 mpa=58k psi
348 mpa= 52k psi. yield

Is that tensile or compression yield, in aluminum they are different by a significant amount.
In composites too. The lowest of the two. Rarely a disadvantage because most structures and loads are symmetric. If not, you can always beef up the tensile side with stiffer/more fibers.
However looking at the bucking and crippling chapters of Bruhn, often the crippling allowable is a fraction of the Fcy based on test data, and developed formula.

How is that handled in the Composite world? "just make the core thicker" sounds like TLAR
You calculate the stress levels at which buckling for that structure would occur. Then check with the actual stress levels.

Here's an actual example for a tail boom under bending in two directions plus torsion:
http://www.nieuwenhuize.com/wp-content/uploads/2017/08/Tail-boom-calculations.png

You start with the stations (mid, left, calculate for a given input all the EI's). Second bunch of lines calculates the stresses in the caps and skin.

Exactly in the center (Buckling stress Mp) are the stresses at which the skin would actually buckle. Compare actual stress levels with buckling stress levels. If they're low enough you're good. Mind you, this is a CFRP tail boom without cores and a wall thickness on the order of 0.02" With buckling roughly scaling with t^3, you can imagine that even the thinnest cores (0.06") would obliterate any concern about buckling.

Actual failure and deformation was <5% from calculated ;)

A full wing calculation would only have a few extra columns but would otherwise be similar.
 
Last edited:
Top