Load Test Data: Balsa-Sheeted Foam Cores with Carbon Reinforcement

[green66]

This may be of interest to those who design or build with balsa-sheeted foam cores, specifically concerning weight reduction by using lighter foam for the core.

I load-tested 6 sheeted foam-core wing panels (3 different foam types, with and without reinforcement) in order to see whether a lightweight foam incorporating two small strip reinforcements to support the sheeting against buckling could be used in place of much stronger (and heavier) non-reinforced foams.

The foams tested were: 1#EPS, Dow High Load 40, and Spyder foam.

A sheeted foam wing typically fails via bending-induced compressive buckling of the upper sheeting. Since buckling is influenced by stability of the sheeting and local surface condition, the critical buckling load can be appreciably lower than the load based on the material's compressive strength. With that, and considering an airfoil's geometry, buckling loads would be difficult to reliably predict with purely analytical methods, hence the testing.

Points to note:

1. The data obtained shows that the addition of two thin carbon strips spanwise in a balsa-sheeted 1# EPS foam core made the EPS panel 50% stronger than a non-reinforced Dow High Load 40 or Spyder panel of the same size. :)

2. 1# EPS weighs less than half of Dow 40 or Spyder. :)

3. The addition of the spanwise strips increased the strength of the EPS panel by 80% ! (moment at failure: 275 in-lb non-reinforced, 495 in-lb reinforced).

4. The reinforced EPS panel is only 10% weaker than the similarly-reinforced High Load 40 and Spyder panels. This suggests that the difference between the various foams' ability to support the sheeting is not nearly as great as the difference between their strengths and stiffnesses ie compressive strength and modulus along the panels' thickness direction.

Edit added: Foam's stiffness does affect stability of the sheeting, however the foam's modulus is so low relative to the sheeting or the carbon that the contribution of the foam to the sectional modulus (EI) of the combined foam / sheeting / reinforcement system is small (even when considering the small sectional area of the carbon), so the sheeting and carbon react the major share of the bending-induced compressive load. Ref: modulus of carbon is approx. 20 million psi vs. 600,000 for balsa (10 lb density, end grain) vs. 7,000 for Spyder along panel thickness (1,200 spanwise).

5. Where the EPS panel suffers is its stiffness, but if normal flight conditions are at relatively low loads, this may be tolerable. The decreasing stiffness with increasing load affecting only the reinforced EPS is interesting - Not sure, but I'm guessing that this may be due to the softness of EPS allowing the carbon reinforcement to move inward as bending increases, pulling the upper sheeting inward with it (but nonetheless still supporting it against buckling), thus effectively reducing the panel's thickness.

6. Camber increases the buckling resistance of sheeting, therefore the reinforcing strips should provide even greater benefit on wings with less camber.

[green66]

The reinforcement consisted of two 1/8" square carbon strips embedded spanwise and flush to the top of the foam (see pic). One strip was positioned 1/2" in front of the high point of the panel, and the other strip 1/2" behind the high point (2.50" and 3.50" behind LE). These positions were based on the assumed location where buckling would initiate ie high point of top surface.

Geometry of all panels: 48" span, 8" chord, no taper, 3% camber, 10% thickness.

The 1# EPS panels were sheeted using 3M #924 transfer tape, and vacuum bagged at 5" Hg. The Dow 40 and Spyder panels were sheeted using Pro-Bond glue, and bagged at 22" Hg, except for the non-reinforced Spyder panel which was mechanically pressed since a bagger was not available at the time.

Some blurbs on transfer tape:
1. Considered among the lightest, maybe the lightest, method to sheet EPS cores, it was popularized by now-defunct Dodgson Designs' with its line of gliders in the mid-80s.
2. It isn't literally tape, just the medium by which the adhesive is applied, and it can only be used on EPS, not on hard foam.
3. Although its bond strength is considered lower than glue or epoxy, the transfer-taped reinforced EPS still demonstrated superior strength. A bond stronger than the foam should be all that matters.

Each panel was loaded in 3-point bending per the setup shown. Deflections at mid-span were measured to the nearest 1/16" after each addition of weight, in 5 lb increments until the panel failed. The weight was supported with cloth padding in order to avoid locally stressing the sheeting which would sensitize it to buckling.

It is important that the reinforcements be bonded to the sheeting, with no gap between. Epoxy the strips sticking slightly above the foam just before applying the sheeting, then lightly coat the top of the strips with epoxy. The force of bagging at even 5" Hg should be enough to press the carbon strips flush with the foam. Trial fit the strips before bonding to ensure that the grooves are deep enough and don't require too much force to push the strips in. Remember to bag 1# EPS at no more than 6" Hg to avoid permanent deformation.

[green66]

Video clip (http://www.rcgroups.com/forums/showthread.php?t=753299#post8310258) shows reinforced Spyder panel failing at 50 lb applied load.
Caution: Do not duplicate this test procedure unless you're wearing open-toed sandals :)

[vintage1]

Very excellent work.

Actually you PROBABLY would get more benefit from a full depth vertical grain shear web that 1/8" deep spars.

Basically you are, as you say, trying to remove the Euler buckling mode of failure.

The spar method relies on the resistance to bending of a narrow spar..its far better to uitilise the full wing depth and tie top and bottom skins together IMHO.

I would be very interested in comparisons with a wing cut in two along the span, with a plate of vertical grained balsa inserted before sheeting.

I suspect the results might surprise you.

[green66]

The spar method relies on the resistance to bending of a narrow spar..its far better to uitilise the full wing depth and tie top and bottom skins together IMHO..... I would be very interested in comparisons with a wing cut in two along the span, with a plate of vertical grained balsa inserted before sheeting. Agree, that may be stronger, but in a different context as the failure mode would likely be shifted from buckling of the sheeting to bending of the spar.

The benefit of using small carbon strips, which are themselves rather flimsy, is not to gain strength as much as to eliminate a substantial amount of weight by replacing a blue foam or Spyder core with EPS weighing not even half as much ...... while preserving (actually increasing) the original non-reinforced Spyder-based strength margins, with very little additional cost or effort. I've always been leery about cutting apart a wing spanwise, or even making a deep cut, to accommodate a spar.... in fear of not getting everything back together in perfect alignment.

I mean we're talking about cutting 50% of the weight of a wing here, and the wing is what, maybe 1/3 of the total aircraft weight? Needless to say, some major benefits to be had in terms of performance, control response, sink rate, run time, etc.

I suspect that there's a sizeable group of r/c'ers (including myself, before doing this testing) who are fond of sheeted hard foams based on nothing more than knowing that the compressive modulus and strength of these foams is several times that of EPS, who then wrongly assume that their wing will also be that much stronger in bending than an EPS-cored one..... when it's buckling of the sheeting that's the weak link all along.

With that, and recognizing that buckling initiates over a small portion of chord, why use hard (heavy) foam as the core of a sheeted wing? Just use light foam and support the sheeting where it's needed.

One might also ask "OK, how about reinforcing a Spyder wing to obtain super strength, then thin it down to the original strength margins to save weight?" => Nope. The test results show reinforced Spyder only 10% stronger than reinforced EPS because, once again, buckling of the top sheet is the failure mode. Although hard foams have much higher comp. strength and modulus than EPS, they're still relatively weak / compliant materials, and apparently aren't much better in supporting the sheeting against buckling...

Edit added: ...likely because the harder foam doesn't appreciably increase the section modulus (EI) of the combined foam / sheeting or foam / sheeting / carbon system.

[nuevo]

However... you get the best strength-to-weight in the spar by putting all the carbon as close to the surface as possible. That is, thin and wide carbon strips, not square carbon strips. If you build a spar like that with the foams you used, I suspect the 1# EPS will buckle much earlier than the higher compression foams.

[green66]

However... you get the best strength-to-weight in the spar by putting all the carbon as close to the surface as possible. That is, thin and wide carbon strips, not square carbon strips. A spar is designed to take bending loads. The square strips are not spars, they're just restraints to stabilize the sheeting to increase the critical buckling load. Buckling of the top sheeting is the primary failure mode, not bending of the core.

The thickness of a square strip provides additional stiffness to stabilize the sheeting, whereas a thin strip would tend to locally bend and give way at the point of buckling.

If you build a spar like that with the foams you used, I suspect the 1# EPS will buckle much earlier than the higher compression foams. I did do that :) Look at the charts => The 1# EPS wing with the small square strips is 50% stronger than the non-reinforced High Load 40 and Spyder wings (45 vs. 30 lb load at failure), and the EPS wing weighs less than half as much! Also, the reinforced 1# EPS panel is only 10% weaker than the reinforced High Load 40 and Spyder wings (45 vs. 50 lb at failure) despite 1# EPS having compressive strength and stiffness not even close to the other two - I think Spyder has something like 10 times higher modulus than 1# EPS.

Putting the above in another perspective, the strength to weight ratio of the reinforced 1# EPS wing is more than triple that of the non-reinforced High Load 40 and Spyder wings.

The thing to remember is that the sheeting is what fails, not the foam.... and because of that, the higher strength or stiffness of the hard foams, although providing some additional support of the sheeting, doesn't translate into proportionately higher bending strength of a finished (sheeted) wing. Based on the test results, it doesn't even come close to proportional.

I certainly didn't expect to see benefits of the reinforcing strips to the extent shown, so I keep looking for conceptual or procedural errors that I might have made that could bias the results.... but I'm not seeing any.

Edit added: Assuming buckling based on the sectional modulus (EI) of the combined foam / sheeting or foam / sheeting / carbon system, it does make sense that the type of foam had little influence on the strength of the tested panels, and that the carbon strips, given their much higher modulus than the sheeting or foam, contributed greatly to the buckling resistance and overall strength of the panels.

[vintage1]

Er..if the sheeting fails due to Euler buckling, its the foam that has failed to allow the buckling.

Hence my suggestion for full depth vertical grain shear webs.

In fact, do away with the foam altogether and make a rib/shear web sheeted wing. Just as strong and lighter ;)

[green66]

Er..if the sheeting fails due to Euler buckling, its the foam that has failed to allow the buckling. All of the reinforced panels demonstrated similar strengths regardless of which core material was used; ditto for the non-reinforced panels, therefore the foam's strength or stiffness had little influence on panel strength. Further, the sheeting only needs to experience a small deflection to initiate buckling, and that deflection I believe is well below what the foam can handle before failing, thus not likely that the foam failed prior to buckling.

Hence my suggestion for full depth vertical grain shear webs.

In fact, do away with the foam altogether and make a rib/shear web sheeted wing. Just as strong and lighter ;) Yes, a wing with a legitimate spar / web would outmuscle anything I tested, as it would be like extending the carbon strips full depth. However, the carbon strips demonstrate how a simple reinforcement enables a 1# EPS wing to outperform a similar High Load 40 or Spyder panel ...... AND where that same reinforcement applied to the hard foam would NOT yield similar benefit ie only 10% stronger than EPS, but more than twice as heavy :(

A built-up structure has its merits but that's another context - This info is for those who prefer sheeted foam cores for various reasons, a couple of mine being durability and an inability to build open structures with the profile consistency of a sheeted core.

[markschaffin]

I think everyone here is in violent agreement with each other. I've caught hints in the hand launch forum that the higher density foams aren't really needed for strength as much as for maintaining a nice surface. These wings generally have no balsa sheeting, but are bagged with a combination of glass/kevlar/carbon and some kind of carbon spar cap. Supporting the spar caps is critical to the strength of the wings.

That said, I haven't studied structures in about 20 years.

Mark