Testing XPS foam for epoxy adhesion/bonding

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

stanislavz

Well-Known Member
HBA Supporter
Joined
Aug 21, 2016
Messages
926
Location
Lt
Anyway, the best surface prep would be a mini rake that cuts square edged channels into the foam . Putting the epoxy joint a little more into shear , rather than tension.
Have you calculated how extra epoxy you have due to this ?
 

Protech Racing

Well-Known Member
Joined
Jul 10, 2020
Messages
404
No. I have not built the tool. I use Glue around the edges and epoxy in the center. The Glue and epoxy seem to mix with no issues.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
5,861
Location
US
Anyway, the best surface prep would be a mini rake that cuts square edged channels into the foam . Putting the epoxy joint a little more into shear , rather than tension.
I was thinking something like ganged micro-gouges. Or, a simple powered version: a palm-sized tool with a bunch of very thin circular blades (Dremel-size) on a common shaft cutting kerfs in the foam at the chosen spacing and depth.

Am I the only one that finds that polystyrene foam is hell on blades? Seems that after I cut about 15-20 linear feet, my utility blade starts to rip rather than cut the foam. Maybe I need to spend more for better blades.
Have you calculated how extra epoxy you have due to this ?
That's an issue, and the weight is why the folks who are most concerned about weight first fill the voids in the large-cell XPS with an epoxy/micro mix rather than straight epoxy. But, that either adds another bagging session or requires a well-timed application of the layup (after the micro has cured enough so the microballoons won't migrate into the layup, but early enough that the new layup will still bond chemically to the micro layer underneath).
With the desired, specified large-cell foam, the voids are whatever they are, and designers/builders have found the absorbed epoxy or epoxy/micro weight to be acceptable. If we, instead, are making our own deliberate voids in too-smooth foam, at least in theory they could be tailored (depth, frequency/pitch, width) to give the desired "tooth" without more absorption than required. In theory...
 
Last edited:

stanislavz

Well-Known Member
HBA Supporter
Joined
Aug 21, 2016
Messages
926
Location
Lt
If we, instead, are making our own deliberate voids in too-smooth foam, at least in theory they could be tailored (depth, frequency/pitch, width) to give the desired "tooth" without more absorption than required. In theory...
+ Stability of results..

And you need 1/5 of you foam shear strenght for our sub 200 mph applications.
 

peter hudson

Well-Known Member
Joined
May 24, 2020
Messages
50
+ Stability of results..

And you need 1/5 of you foam shear strenght for our sub 200 mph applications.
I think that's a good point in that the skin to foam bond's shear strength, and flat-wise tension is what should be tested rather than peel. If it is consistantly better than the foam shear and tensile than it's as good as it will ever perform.
 

stanislavz

Well-Known Member
HBA Supporter
Joined
Aug 21, 2016
Messages
926
Location
Lt
I think that's a good point in that the skin to foam bond's shear strength, and flat-wise tension is what should be tested rather than peel. If it is consistantly better than the foam shear and tensile than it's as good as it will ever perform.
But that 4/5 is dead weight, and high penalty in sub 200mph region. Where your benchmark is fabric wing..
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
5,861
Location
US
+ Stability of results..
" Stability' over time? That's important, and I'm not sure how to test/measure that. It would be simple enough to apply cyclic shear loading of some type at an accelerated rate to simulate years of use, but we'd still have other issues (thermal cycling, chemical compatibility, etc) as an unknown. There's no doubt that the 40+ years of field experience with Dow billet foam provides a lot of confidence. For consideration:
1) When the akaflieg folks and later Rutan started using XPS as a core, was it always the large cell stuff? Did they have the ability to simulate adhesion in subsequent decades?
2) Has Dow (now DuPont) changed the formula for this product over the years? It seems certain they have, we know positively that at least the blowing agents have changed. Did the moldless composite community learn of the proprietary changes in the material and perform prospective tests to assure the new foam would hold up for decades? I'd bet, instead, designers and builders are accepting some risk in this area.

And you need 1/5 of you foam shear strenght for our sub 200 mph applications.
Sorry, could you explain this a bit more? Do we have a rough estimate for the required shear strength of the epoxy/foam bond?

For what it is worth, Dow XPS with compressive values and densities similar to buoyancy billet have listed shear values of approx 45psi (310kPa) and tensile strength of 85psi (590 kPa).

From a practical standpoint, I don't think it would be much harder to test the epoxy/foam bond in shear than to test it in peel. Bigger weights needed. It wouldn't be lab-quality testing, but could be okay in a comparative sense and shed light on some issues.
 
Last edited:

stanislavz

Well-Known Member
HBA Supporter
Joined
Aug 21, 2016
Messages
926
Location
Lt
Sorry, could you explain this a bit more? Do we have a rough estimate for the required shear strength of the epoxy/foam bond?

For our 1.02 psi load one 0.2" rib spaced each 6" is ok, and it have 8 reserve factor. For load from ribs to spar.

1/5 was from cri cri example - 1/4" rib each 1.8" . 2/14 of original 100 kg/m3 pvc.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
5,861
Location
US
For our 1.02 psi load one 0.2" rib spaced each 6" is ok, and it have 8 reserve factor. For load from ribs to spar.
Sorry, slowest kid in the class here, holding up the class. "Load from ribs to spar" ? In our stressed skin wing aren't all the loads that go to the spar being passed from the skin (not the ribs)? The ribs are needed to reduce skin deflection and keep them aligned to carry those loads. Or, I misunderstand.
1/5 was from cri cri example - 1/4" rib each 1.8" . 2/14 of original 100 kg/m3 pvc.
This is the "inflation load' pulling the skins out from the aerodynamic chordline of the airfoil? If so, and if your skins are no foam FG/epoxy only and the ribs are pure foam, then it would seem (to my naive eye) we have primarily a tension load on the rib (if negative external air pressure) with a peeling load to the degree the skin is bulging out and pulling on the first line of contact with the foam rib. If our rib has FG capstrips that fold down over the rib sides, or if our skin/rib joint would be loading the foam rib sides in shear.
 

stanislavz

Well-Known Member
HBA Supporter
Joined
Aug 21, 2016
Messages
926
Location
Lt
Sorry, slowest kid in the class here, holding up the class. "Load from ribs to spar" ? In our stressed skin wing aren't all the loads that go to the spar being passed from the skin (not the ribs)? The ribs are needed to reduce skin deflection and keep them aligned to carry those loads. Or, I misunderstand.
Ok. Not a problem. Solid foam can take much more load, than it need. We have that 1.02 psi on surface. And in my example - pvc foam was able to hold concentrated load - 30 times more - 0.2" thick ribs spaced each 6". But it was pvc foam. Xps should be 3-4 weaker..
 

peter hudson

Well-Known Member
Joined
May 24, 2020
Messages
50
Ok. Not a problem. Solid foam can take much more load, than it need. We have that 1.02 psi on surface. And in my example - pvc foam was able to hold concentrated load - 30 times more - 0.2" thick ribs spaced each 6". But it was pvc foam. Xps should be 3-4 weaker..
As we strive to reduce the amount of foam with thin ribs (which is a good goal for weight savings) it starts to make the skin to foam bond more important.
AND as Vigilant1 points out it adds a little peel to the bond at the corner of the ribs. In keeping with this thread's title; what test coupon would combine the flatwise tension and peel in the corners from the pressure load, and the shear from the rib bending? I still think the goal is to assure the foam fails rather than the bond. If that is consistently achieved, I think the joint can be considered good and the XPS sheets could be made into effective ribs.
 

stanislavz

Well-Known Member
HBA Supporter
Joined
Aug 21, 2016
Messages
926
Location
Lt
As we strive to reduce the amount of foam with thin ribs (which is a good goal for weight savings) it starts to make the skin to foam bond more important.
AND as Vigilant1 points out it adds a little peel to the bond at the corner of the ribs. In keeping with this thread's title; what test coupon would combine the flatwise tension and peel in the corners from the pressure load, and the shear from the rib bending? I still think the goal is to assure the foam fails rather than the bond. If that is consistently achieved, I think the joint can be considered good and the XPS sheets could be made into effective ribs.
But full xps core leaves us with one problem - it weights around 30 kg for two seater ul. Where you wing budget is about 60kg per both wing... Spar is in 10-20kg range... And we have no reserve for controls skins etc..
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
5,861
Location
US
In keeping with this thread's title; what test coupon would combine the flatwise tension and peel in the corners from the pressure load, and the shear from the rib bending?
The degree of peel at the rib edges may be influenced significantly by the inter-rib wing skin deflection (more deflection = higher peel angle) and also by the wing skin stiffness as an independent variable (stiffer skin increases the bond area taking the peel load). So, simple, general tests may only be able to provide a trends/indications of the quality of the epoxy/laminate-toam bond. From that, candidates could be selected and detailed tests with actual proposed wing skin/rib section could be tested.
The build method is important. If the wing is built up and the interior structure is accessible, then it is practical to affix corner tapes, flanges, or even gussets to make peeling into shear loads.
For a monolithic foam core (with or without lightening cutouts), the epoxy/laminate-to-foam bonds are almost exclusively in shear, right? And, because of the area involved, the per-area magnitude of the loads would be comparatively small.
Thanks for the note.
 

Protech Racing

Well-Known Member
Joined
Jul 10, 2020
Messages
404
I tested ribs with glass on one side and 1in glass tape on the covering surface . When I ran the weight of failure, I was 10x over 3G flying weights.

The ribs tested to about 90# on the tail section of the foil naked . Glass tape the edge went to 140#. Glass on the rib face went to 150+#. The fail was the rib collapsing just about 160 or so . 1 in XPS foam .
I made the plane with ribs , glassed on one side , the 1 in glass tape, and cloth retainers to the ladder rungs .
Another improvement was to glass tape the rib edges and then wrap the rib with a 6 in section of either cotton or glass at the max lift area of max depth plus 3% My case was 28% .
All of my glass was adhered with Gorollia glue on the edges and epoxy in the center. This was after my engineer Dad watched the glue test of epoxy solo. I consulted the Long EZ guy near me and he said the peel test will always be a fail with Epoxy.
My next wing will have just the 1 in glass tape. For covering retention and the additional load. It about doubles the fail test .
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
5,861
Location
US
But full xps core leaves us with one problem - it weights around 30 kg for two seater ul. Where you wing budget is about 60kg per both wing... Spar is in 10-20kg range... And we have no reserve for controls skins etc..
All true. But:
- Many other alternatives end up just as heavy. As Billski has sketched out several times, if we have a sandwich skin, ribs, joining tapes, standalone spar, etc all the other stuff can equal the weight of the solid foam block. Building the spar right on the foam block saves a little spar weight compared to a standalone spar.
- As discussed in another thread, for lighter weight, it may be practical to remove about 45% of the monolithic core and still not exceed the limits of the remaining foam. So, if the solid foam wing was a lighter alternative before (see above), it will be ”lighterest' after some of the foam is removed. :) If that works...
- Weight is a consideration, but not the only one. For one-off builds in a garage, the time saved by avoiding a lot of ribs, jigs, separate wing molds, separate spar molds, no "will it fit right?" blind structural closures, etc can be worth a few pounds.

A lot depends on the aero loads and desire for laminar flow. Solid foam (or rigid skin for that matter) probably isn't the most efficient choice for a low-wing loading, slow plane with a non-laminar airfoil.
 
Last edited:

stanislavz

Well-Known Member
HBA Supporter
Joined
Aug 21, 2016
Messages
926
Location
Lt
- Many other alternatives end up just as heavy. As Billski has sketched out several times, if we have a sandwich skin, ribs, joining tapes, standalone spar, etc all the other stuff can equal the weight of the solid foam block. Building the spar right on the foam block saves a little spar weight compared to a standalone spar.
It is. And as best solution / benchmark - Rigid D cell + fabric covered aft. portion of wing. For any sub 150 mph airplane. And in a wing - foam or build-up, spar have similar mass if done from same material. Or spar + strut combos.. So we need just light way to transfer loads from Cl to spar.. And stay in desirable skin deflection region.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
5,861
Location
US
And as best solution / benchmark - Rigid D cell + fabric covered aft. portion of wing. For any sub 150 mph airplane.
This way has a lot to be said for it. We still get low drag laminar flow over the "easy" portion of the wing (approx front 30%) and probably don't even need a laminar airfoil to do it. Rigid skin D-cell takes drag loads and covers the place where much hangar rash and bird strikes occur (the LE). Fabric + ribs aft where loads are lower. Very light. Too bad the ailerons, flaps, and the need for something to attach them all to re-complicates the picture.
From a building and owning perspective, I still prefer a rigid skin airplane. Maybe I need to work on that.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
5,861
Location
US
Protect,
A few questions:
1) What is your wing skin/covering that is attaching to these ribs?

I use fibreglass drywall tape on the rib edge , glued on with gorrila glue and stippled with little tiny holes to increase surface area.
Is this the wide-open grid, relatively stiff fiberglass drywall tape, or the silky tape often used for reinforcing basement crack waterproofing? I can see GG working with the former, it would seem too thick for the later.
The coverings get glued on. to the tape.
What coverings are you using, and what glue?

I tested ribs with glass on one side and 1in glass tape on the covering surface . When I ran the weight of failure, I was 10x over 3G flying weights.
? "Glass tape on the covering surface"? Is this the glass tape you bonded to the edge surface of the rib, or did you attach more tape directly to the wing covering where it touches the top of the rib?
When you glassed the large "faces" of the ribs with epoxy/FG, did you remove the factory facings/skin? How did you prep the foam before glassing? When the ribs failed in your tests, did the FG come off at the bondline, or did it rib off a lot of foam with it? Do you remember if it was easy to pull the FG off the foam?
The ribs tested to about 90# on the tail section of the foil naked . Glass tape the edge went to 140#. Glass on the rib face went to 150+#. The fail was the rib collapsing just about 160 or so . 1 in XPS foam .
In any of your tests, did you take a wider tape (say 3") centered on the rib edge and then folded over to wrap down the rib faces about 1" on each side? That would seem to take all the peeling load off the foam due to the uplift of the wing covering and instead give a more moderate shear loads to a larger portion of the rib face. As a bonus, if a stiff adhesive/matrix (e.g epoxy) is used, you'd also get a stiffening "C" channel around the rib's perimeter.
When the rib collapsed in your tests, could you tell how it failed?

What brand XPS did you use for the ribs (pink, blue, or green, and compressive psi claimed)?

Thanks much.
 
Last edited:

stanislavz

Well-Known Member
HBA Supporter
Joined
Aug 21, 2016
Messages
926
Location
Lt
Very light. Too bad the ailerons, flaps, and the need for something to attach them all to re-complicates the picture.
It is not..


As an example have one rib weighting 150 grams. Ribs are spaced each 300mm. Some are twin one for ailerons mount etc.. but : 8.2 / 0.3 = 27 + 4 for twins = 31 * 0.15 = 4.65kg.

Covered with fiberglass single skin 300gsm. Flying, certified in Russia.
 

Hephaestus

Well-Known Member
Joined
Jun 25, 2014
Messages
2,127
Location
YMM
Very light. Too bad the ailerons, flaps, and the need for something to attach them all to re-complicates the picture.
From a building and owning perspective, I still prefer a rigid skin airplane. Maybe I need to work on that.
Remember in the monolithic foam thread where I mentioned varying skin thickness? If you layup root/tip/Dtube in a heavier, then finish with a single ply of <100gsm over top pf all. 80gsm/10mm XPS backing is enough to require not-accidental force to push through a surface.
 
Top