Foam vs. honeycomb for carbon fiber sandwich

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
6,722
Location
Saline Michigan
Okay. So, the epoxy runs into the honeycomb cells, so you need pre-preg, so you need big ovens, so that's out. That's the kind of info I was after.

It's going to be dark blue, so I'm probably going to need to do a cure at elevated temp (Aeropoxy at ≈250, I think). Will need to figure out how to do that.

I'm hoping Vne will be pretty zippy. Wouldn't mind 300 mph. Too early to know. Based on what you were saying, Bill, maybe I'll need to look at some thicker foam in the fuselage. 1/2"? Parts of the fuselage—e.g., turtledeck—will have much tighter curves, so I hope it's okay to switch to thinner foam in those areas. I'll get to that point sometime. Sounds like 1/4" in wing panels might be adequate. As you said, the carbon fiber fabric I use will influence things. Jim Marske talks some about the different recommended fabrics in his book.

Speaking of Jim Marske, I'm going to use his book in helping design the spar. I've gone through a few different ideas about how to make the spar(s). At this point, I'm thinking of making it one piece (if possible). The question is if the Graphlite rods are flexible enough to do the bends in the gull wing. Maybe I'll be able to achieve that by using a greater number of smaller rods vs. fewer larger ones.

Thanks,
Jay
This is serious thread drift, but it is your thread. Several comments:

Make sure that the foam will stand the cure and operational temperatures;

Vinylester resins have higher Tg than room temp cure epoxies without the hot cure;

Marske is a good place to start on design. Please read my stuff on beams and composites too. http://www.homebuiltairplanes.com/forums/showthread.php?t=28953 and http://www.homebuiltairplanes.com/forums/showthread.php?t=29030

Stress in an elastically bent piece is kappa*E*y, where kappa is 1/bend radius, E is Young's Modulus, and y is the distance from the neutral axis of the rod you are bending. So each rod in your bent spar will have built in stress going from compressive at the inside of the bend to tensile on the outside. Flight loads will be superimposed on top of those stresses. Your design will require a beefier spar than would otherwise be the case. In the extreme, you can have no load carrying capability at all. Using Jim Marske's 21Mspi for E, 200 kpsi compressive strength, and 0.06" thick rods, at 3.15" bend radius, you get to compressive strength of his rods just by bending. His small round rods will give more margin. If the scale bends result in rods being bent to these kinds of radii, you could "cheat" and make the radii larger. This may still be a much more acceptable solution than attempting wet layups of spar caps from carbon tapes...

Billski
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
6,722
Location
Saline Michigan
Foam can be scored on the inside for tight radius curves.
Understand that the scored lines turn into wedges of epoxy, and the weight add can be significant. Same is true of holes punched in foam, joints between pieces of foam, etc.

Billski
 

birdus

Well-Known Member
Joined
Apr 3, 2012
Messages
185
Location
Tacoma, WA
Birdus
If you feel you you can get a good bond by rolling out a film, what would stop you from flipping it over for the other skin.

Ben
Air Creation Clipped 582
When the Green Light pops the BS stops
I'm not sure what you mean. My plan is to lay a sheet of carbon fiber into a mold, then foam, then another sheet of carbon fiber, then apply vacuum.

Jay
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
6,722
Location
Saline Michigan
Based on what you were saying, Bill, maybe I'll need to look at some thicker foam in the fuselage. 1/2"? Parts of the fuselage—e.g., turtledeck—will have much tighter curves, so I hope it's okay to switch to thinner foam in those areas.
No, what I said was that your speed and panel size and facings combine to determine what core thickness you will need. Please do not oversize things. The panels I refer to are the sizes of the skin between other supports, like longerons, spars, and ribs. You have homework to do on how big the panels will be. You will have a rib at each end of the wing, defining fuel bays and equipment bays, maybe one or two as fuel tank baffles, and probably have one at each end of each bend. Main and drag spars will establish other edges. If you arbitrarily use 1" foam, and 3/8" would have done, you will have excess weight and you will have to extend your fuel tanks further out the the wings too.

Billski
 

birdus

Well-Known Member
Joined
Apr 3, 2012
Messages
185
Location
Tacoma, WA
This is serious thread drift, but it is your thread.

Billski
Well, I think my initial question was answered adequately. I'll plan on going with foam. Everything beyond that I'll be asking eventually anyway, so I don't mind the extra info, and since some folks asked for more project details, I don't mind sharing.

Regarding bending the Graphlite rods, I can get a very gradual bend in the center section. It's hidden and I have the entire interior of the fuselage to straighten out that curve, so I'm not too worried about that (similar to the original, really). Regarding the bends at the lower points in the wings, I'm not sure how much I can straighten those curves. I'll probably just beef up those curves using more rods, or more layers of carbon fiber cloth, etc. That's where the landing gear mount, anyway, so more meat may be just what the doctor ordered. I'll get into those details when I'm farther along. I had thought before about making a multi-piece spar and maybe I'll just have to abandon the one-piece spar. That would make for easier construction, anyway. I could build the center section, the an outboard section for each wing. I could join the outer spars to the center section using steel or something like that. That would handle that curve and help with landing gear loads. Anyway, it's a lot of fun to think about and appreciate the thoughts I'm getting here.

Thanks,
Jay
 

BBerson

Well-Known Member
HBA Supporter
Joined
Dec 16, 2007
Messages
13,107
Location
Port Townsend WA
Understand that the scored lines turn into wedges of epoxy, and the weight add can be significant. Same is true of holes punched in foam, joints between pieces of foam, etc.

Billski
My Grob has random small holes punched in the foam near the wing root. Apparently on purpose for sandwich disbond resistance.
 

birdus

Well-Known Member
Joined
Apr 3, 2012
Messages
185
Location
Tacoma, WA
No, what I said was that your speed and panel size and facings combine to determine what core thickness you will need. Please do not oversize things. The panels I refer to are the sizes of the skin between other supports, like longerons, spars, and ribs. You have homework to do on how big the panels will be. You will have a rib at each end of the wing, defining fuel bays and equipment bays, maybe one or two as fuel tank baffles, and probably have one at each end of each bend. Main and drag spars will establish other edges. If you arbitrarily use 1" foam, and 3/8" would have done, you will have excess weight and you will have to extend your fuel tanks further out the the wings too.

Billski
As you said, Bill, we're drifting into other topics here, so I moved the conversation to a new thread on some of my more general design questions for this project.

Thanks,
Jay
 

BoKu

Pundit
HBA Supporter
Joined
Aug 15, 2013
Messages
2,598
Location
Western US
My Grob has random small holes punched in the foam near the wing root...
That's a standard Divynicell option that I usually get. When I don't get it, I just punch similar holes through the foam with an icepick. They make it easier to get the air out from under the foam on large layups. Yes, they eventually turn into little pylons of solid epoxy that you have to deal with when doing repairs or installing post-mold inserts. But the overall added weight is inconsequential.
 

birdus

Well-Known Member
Joined
Apr 3, 2012
Messages
185
Location
Tacoma, WA
Marske is a good place to start on design. Please read my stuff on beams and composites too. http://www.homebuiltairplanes.com/forums/showthread.php?t=28953 and http://www.homebuiltairplanes.com/forums/showthread.php?t=29030

Billski
By the way, Bill, I've been going through those threads over the past week or so. Some of it I get, some of it is beyond me. When I get farther along, I'll either get some more concrete help here, or at my local EAA chapter, or I'll pay a professional to help make sure that if I die, it's not because the wings fell off my plane.

Thanks,
Jay
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
6,722
Location
Saline Michigan
Regarding the bends at the lower points in the wings, I'm not sure how much I can straighten those curves.
Birdus,

You do not straighten them out, you just make sure you have a generous bend radius... The originals looked to have a fairly large bend radius, and that might be OK. If not, increasing the radius a little more might get you there..

Billski
 

DeepStall

Well-Known Member
Joined
Dec 26, 2010
Messages
136
Location
Colorado
The question is if the Graphlite rods are flexible enough to do the bends in the gull wing. Maybe I'll be able to achieve that by using a greater number of smaller rods vs. fewer larger ones.
When the time comes to detail out your spars, recognize that loads that act to open a curved lamination also induce inter-laminar tension forces that can be significant and lead to delaminations. So for your Corsair, negative g flight loads for instance could cause problems. One of a few curved beam effects that doesn't always get the attention it needs...
 

birdus

Well-Known Member
Joined
Apr 3, 2012
Messages
185
Location
Tacoma, WA
Birdus,

You do not straighten them out, you just make sure you have a generous bend radius.

Billski
That's what I meant by straighten them out. Perhaps I shouldn't speak so informally when discussing engineering topics.

Thanks,
Jay
 

birdus

Well-Known Member
Joined
Apr 3, 2012
Messages
185
Location
Tacoma, WA
When the time comes to detail out your spars, recognize that loads that act to open a curved lamination also induce inter-laminar tension forces that can be significant and lead to delaminations. So for your Corsair, negative g flight loads for instance could cause problems. One of a few curved beam effects that doesn't always get the attention it needs...
I'm actually concerned about both positive and negative Gs, as there are bends in both directions! I'm thinking about just building 4 spars (or maybe 3 with a single center section) and joining them with steel or aluminum brackets at the bends. Anyway, it's fun to think about and I'll definitely consider the effects of those bends with respect to using the Graphlite rods. Thanks for your input.

Jay
 

birdus

Well-Known Member
Joined
Apr 3, 2012
Messages
185
Location
Tacoma, WA
...so each rod in your bent spar will have built in stress going from compressive at the inside of the bend to tensile on the outside. Flight loads will be superimposed on top of those stresses. Your design will require a beefier spar than would otherwise be the case. In the extreme, you can have no load carrying capability at all. Using Jim Marske's 21Mspi for E, 200 kpsi compressive strength, and 0.06" thick rods, at 3.15" bend radius, you get to compressive strength of his rods just by bending. His small round rods will give more margin. If the scale bends result in rods being bent to these kinds of radii, you could "cheat" and make the radii larger. This may still be a much more acceptable solution than attempting wet layups of spar caps from carbon tapes...

Billski
Jim just sent me these numbers:

---

A .060" thickness rod can bend down to a 3" radius before a fracture occurrs in the outer surface. The tension side always fails first. Now based on that:
A 6" r would be stressed to 50%
A12" r would be stressed to 25%
A 24" r is stressed to 12.5%

---

He definitely seems to think a one-piece spar is the way to go. I just did some measuring on a line drawing of the Corsair and I believe I'll have a radius of at least 40" on the bends, much more than I had been thinking. The shear web height will be reduced where the landing gear folds up into the wing, as it crosses the bottom of the spar, but I feel encouraged about using Graphlite rods and making this a one-piece spar.

Jay
 

BoKu

Pundit
HBA Supporter
Joined
Aug 15, 2013
Messages
2,598
Location
Western US
As a data point, in the HP-24 spar stubs we bend the 0.092" Graphlite strips of the lower cap upward with a 30" radius.
 

PiperCruisin

Well-Known Member
Joined
Jan 17, 2017
Messages
148
Location
Idaho
A few issues with a bent composite spar, doable, but something to think about.
1. Interlaminar loads will be higher...wants to delaminate. Take a stack of Post-it notes, hold the edges and bend it. Notice the "plys" want to separate in the corner.
2. The bend creates a kick load as the beam wants to straighten when the load turns the corner.
 

PiperCruisin

Well-Known Member
Joined
Jan 17, 2017
Messages
148
Location
Idaho
It is just trying to straighten the beam potentially causing a lot of extra stress. See the image of the F4U wing box with all of the web reinforcements. Potentially gets complicated. Don't remember where I got the image.F4U_WingBox.jpg
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
6,722
Location
Saline Michigan
Check this out:

http://courses.washington.edu/me354a/Curved Beams.pdf

Shigley also has some stuff on curved beams... Roark's has a whole section on the mechanics of curved beams. I still do not know what this "kick" load is. Aero guys rename things that were already identified and analyzed by mechanical engineers before the Wrights ever flew, so sometimes we need translation help. I will look at Shigley and Roark's and see if it will help.

Billski
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
6,722
Location
Saline Michigan
OK, got into Timoshenko (Mechanics of Materials), and nothing on curved beams. Then Shigley, and the standard method of calculating neutral axis shift and stresses. Then Juvinall and nothing on curved beams, but a nice review on constant stress cantilever beams (applicable to Wittman style landing gear legs).

Then Roark's, Chapter 9 is huge and useful. It goes through the whole topic. If the curve radius is 8 or more times the depth of the beam, the errors from using straight beam theory is 5% and less. If the curve radius is smaller than 8 times the depth of the beam, then you need to wrap curved beam theory into your analysis. There are formula for neutral axis shift of wide flange beams and deflections. The expansion of the calculations to composite beams is straightforward. Well, once you get past the composite beam theory, which may warp your brain more than science fiction novels.

Oh, and I have looked and looked for "kick" loads for tapered and/or curved beams. The other engineer in the house did too. The only place our internet searches have turned up anything is - wait for it - a couple threads on Homebuiltairplanes.com. Sure, I found how to calculate recoil forces in guns - simple Newtonian physics. Somebody want to talk about "kick" loads in curved and/or tapered beams, their origin, cite textbooks, learned articles, professors making up expressions that did not get widely used, etc? I am all ears.

Billski
 
2
Top