Pulling a precision part from a male mold...

Homebuilt Aircraft & Kit Plane Forum

Help Support Homebuilt Aircraft & Kit Plane Forum:

Grimace

Well-Known Member
Joined
Jan 2, 2009
Messages
319
Location
Chicago, IL
This is what I'm planning to do.... unless someone wiser convinces me that it won't work....

As per my post in the design/structures forum, I am planning on building my wing without twist and without taper. This means that I can use the same mold for both wing halves.

My plan is to build a wood male mold... basically an exact replica of the wing, minus the thickness of the fiberglass and foam. The skin I pull from the mold will be the complete top of the wing and the bottom, all the way to the rear/false spar. I will wrap the wood mold in mold release film... Then lay up a layer of glass or two... then bond the foam sheet to the glass using a vacuum bag. Then micro the foam for a good bond. Then vacuum bag a couple plies of glass on the outside of the foam sheet. Once it cures, just slide out the plug and make the next skin.

Sure, it's inefficient for mass production, but I'm thinking it should work well for a one off design. But of course, I've always used female molds because, obviously, they give you more control over the outer shape. However, I'm figuring this technique has to be at least as precise as regular wet lay-ups.

Anybody think differently?
 

PTAirco

Well-Known Member
Joined
Sep 20, 2003
Messages
3,669
Location
Corona CA
I think you will have great difficulty "sliding" the mold (plug,in this case) out of the finished wing - removing glass structures from any mold can involve a fair degree of cursing, hitting with rubber mallets, digging into it with flat, sharp objects, cursing, trying to blow pressurized air into it, cursing etc etc... .Remember , you have no draft angle, a huge friction area and not much to grab unless you mold in handholds on the plug and the wing skins. A tapered wing would make it a zillion times easier since you have that built in draft angle. (Or "Draught angle" in Olde Englishe...)
 

addaon

Well-Known Member
Joined
Feb 24, 2008
Messages
1,697
Location
San Jose, CA
What are you gaining here by using such simple wing geometry? When you mentioned that you were doing a candy-bar wing, I assumed aluminum; you gain a huge amount by having only one rib shape to make a form for, etc. But for composites?
 

Grimace

Well-Known Member
Joined
Jan 2, 2009
Messages
319
Location
Chicago, IL
The shape won't be a complete loop.... It will terminate at the rear/false spar. So I will be able to stretch the skin upwards and downwards to release it from the mold.

And the advantage for composites is that I can use the same plug for both halves of the wing....
 

PTAirco

Well-Known Member
Joined
Sep 20, 2003
Messages
3,669
Location
Corona CA
The shape won't be a complete loop.... It will terminate at the rear/false spar. So I will be able to stretch the skin upwards and downwards to release it from the mold.
Ah - ok, that will eliminate that particular problem entirely.
 

orion

R.I.P.
Joined
Mar 2, 2003
Messages
5,800
Location
Western Washington
There's one or two issues you should consider in making your wing skin in this manner. First, the most critical issue I see is that you don't indicate what your plan is for the spar location. Based on your paragraph above it looks like you plan on maintaining the foam core over the spar - this is not good practice and has already caused several fatalities in the industry, the first, if I recall right, coming from some prefab skins being produced for KRs some years back.

Maintaining the core over the spar can be problematic in that it creates sort of a discontinuous beam. Regardless of your spar configuration, the behavior of beams is such that a significant portion of the load is taken up by the outer-most fiber of the structure, which is the skin. Most designers are often quite surprised when they do a detailed analysis of a typical composite design since it generally shows that the skin can be loaded much higher than they suspect. This loading requires a predictable shear distribution in the buildup but if the buildup suddenly ends, as you'd have with a relatively flexible core, the said core may fail since it is not capable of carrying the shear flow generated between the substructure and the skin.

The failure of the core at that point might not be noticeable for some time but eventually it could lead to outer skin buckling or tearing, which will of course destroy the wing's ability to create lift. But even relatively light shear loads could cause minor wrinkling in the skin, which will over time propagate and affect the wing shape, even if the skin does not totally fail. Imagine having a normal flight but having the skin suddenly wrinkle as you flare for landing. Not a pleasant surprise.

The other issue in this type of construction is simply the amount of work you'll get into in eventually doing the final surface finish. And this is especially the case if you're looking at cored fabrication. It is highly probable that your finished airfoil will look nothing like the shape you analyzed the plane for. Also, if the shape is too different from side to side, you might run into unfavorable trim issues.

If you do male plug construction, it is best to do so for just the outer skin sheet (no core). Furthermore, working towards the most fair surface is paramount so a slow setting epoxy aided with a vacuum bag and possibly a surface sheet (thin aluminum sheet) might be a worthwhile consideration.

But considering the amount of work ahead of you, your best approach might be to do the plug as you envision and then simply take a splash, creating a nice female mold. then you outer surface will be nearly perfect and on the inner surface you can use a layer of peel-ply, which when removed will provide you with a nearly ready to bond surface.
 

Grimace

Well-Known Member
Joined
Jan 2, 2009
Messages
319
Location
Chicago, IL
Thanks for the excellent replies. I'm really hoping to avoid making a female mold if at all possible. If I can't avoid it, I'm hoping to make it as simply and disposably as possible. One of the goals is to make it nearly all-composite while still making it feasible as a "from plans" project.

I know Rutan has used sheet metal, bonded to wood stringers, to create a female mold for a wing. It sounds like an easy enough concept, but have you priced aluminum lately? Not exactly a cheap way to go... and delivery for such a large sheet wouldn't exactly be a bargain either...
 
Last edited:

orion

R.I.P.
Joined
Mar 2, 2003
Messages
5,800
Location
Western Washington
For a well reinforced and supported framework assembly, a sheet of formica works pretty well too. There though the leading edge curvature might be a bit of a challenge.
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
8,593
Location
Saline Michigan
As someone who has used male molds for my tub, turtledeck, and forward deck, I can tell you to expect significant deviations from intended shape on the outside, particularly around the edges and where you allow for bond lines.

I was going to talk about the core at the spar too. Another way to think about this is to think of the wing unloaded and straight. Then when you load it up with airloads, it will take on a curve with the tips lifted. The whole wing has to do that together, with the top skins being shortened and the bottom skins being stretched. Now if there is no glass to glass connection between the inner and outer skin layers, the deflected shape of the spar and inner skin has to be carried to the outer skin through the foam, which is known to fail most foams over time...

Better to have no core around where the spar goes, at the leading edge, and back at the drag spar. This also allows the main spar to be deeper, which for a certain strength is lighter too. Once this is done in your plug, well, some roughness becomes more roughness around the discontinuities, which means you will end up developing your wing surface in the micro paste... And then, you will have to add some access panels, etc. If you have any ribs besides the ones near the ends of these wing panels, well, it will be plenty sporting to get them in place and bonded in.

For female molds, you can make a splash from a (perfect) plug, or you can do it my way. Hotwired foam and roof flashing. Yeah. I made up templates for three segments that ran several inches beyond my panels in all directions. Careful jigging, seal them to each other with micro slurry, set a glass tape around the periphery, and vacuum bag in the coated roof flashing. Some folks have used plastic laminate. Mine took about 10 hours each for 11 foot long molds, a lot less for the sub skin molds...

Either way, you end up with a female mold. You can make your access panels right on the mold, trim them, tape them down, and build your sandwich part over all of that with vacuum bagging. The outside is right to within some pretty fine precision, your access panels fit perfect right away, and the presence or absence of foam cores only shows up on the inside of the cores.

One other thing to think about... How are you going to make this accurate wooden male mold? This is non-trivial. It almost has to be solid, so it will be glued up, and then carved. I am imagining a big table with a stiff track for a router to run along, and method of moving the block of lumber for successive passes, then many hours with big sanding sticks, filler, etc.

Billski
 

Grimace

Well-Known Member
Joined
Jan 2, 2009
Messages
319
Location
Chicago, IL
As someone who has used male molds for my tub, turtledeck, and forward deck, I can tell you to expect significant deviations from intended shape on the outside, particularly around the edges and where you allow for bond lines.
For non critical dimensions, such as a fuselage where a quarter inch here or there is not a big deal (assuming the errors are not cumulative), would you say this is a reasonable approach? Would you do it this way again?

The reason I thought it would be feasible for this wing is that there are no compound curves. No taper, no twist. I figured the foam should readily take to such a slight curvature and could be bonded with an acceptable level of precision since it's only being curved in one axis.

I was going to talk about the core at the spar too. Another way to think about this is to think of the wing unloaded and straight. Then when you load it up with airloads, it will take on a curve with the tips lifted. The whole wing has to do that together, with the top skins being shortened and the bottom skins being stretched. Now if there is no glass to glass connection between the inner and outer skin layers, the deflected shape of the spar and inner skin has to be carried to the outer skin through the foam, which is known to fail most foams over time...

Better to have no core around where the spar goes, at the leading edge, and back at the drag spar. This also allows the main spar to be deeper, which for a certain strength is lighter too. Once this is done in your plug, well, some roughness becomes more roughness around the discontinuities, which means you will end up developing your wing surface in the micro paste... And then, you will have to add some access panels, etc. If you have any ribs besides the ones near the ends of these wing panels, well, it will be plenty sporting to get them in place and bonded in.
I think you all have got me pretty well sold on this. I was peripherally aware of the problem, but figured that such a light aircraft wouldn't have an issue... I'd seen this type of full core across the skin construction referenced a number of times and while it never really seemed like the "right" way of doing it, I'd seen it referenced enough that I didn't think it would be an issue. Always better safe than sorry though, I suppose.

For female molds, you can make a splash from a (perfect) plug, or you can do it my way. Hotwired foam and roof flashing. Yeah. I made up templates for three segments that ran several inches beyond my panels in all directions. Careful jigging, seal them to each other with micro slurry, set a glass tape around the periphery, and vacuum bag in the coated roof flashing. Some folks have used plastic laminate. Mine took about 10 hours each for 11 foot long molds, a lot less for the sub skin molds...

Either way, you end up with a female mold. You can make your access panels right on the mold, trim them, tape them down, and build your sandwich part over all of that with vacuum bagging. The outside is right to within some pretty fine precision, your access panels fit perfect right away, and the presence or absence of foam cores only shows up on the inside of the cores.
Interesting.... let me ruminate on that for a bit...

One other thing to think about... How are you going to make this accurate wooden male mold? This is non-trivial. It almost has to be solid, so it will be glued up, and then carved. I am imagining a big table with a stiff track for a router to run along, and method of moving the block of lumber for successive passes, then many hours with big sanding sticks, filler, etc.
Good question... and one that I think I have pretty well worked out. The plan is to make it out of solid wood. Multiple strips of varying height glued together into the rough shape and then bondo or a slurry to smooth the curves. 8 foot sections of wood are easy to come by. The wing is only 4.32" tall at the deepest point.
 

Topaz

Super Moderator
Staff member
Log Member
Joined
Jul 29, 2005
Messages
14,768
Location
Orange County, California
...Good question... and one that I think I have pretty well worked out. The plan is to make it out of solid wood. Multiple strips of varying height glued together into the rough shape and then bondo or a slurry to smooth the curves. 8 foot sections of wood are easy to come by. The wing is only 4.32" tall at the deepest point.
If you want to do a solid core plug, it seems like it would be easier to literally do it 90° around - cut out airfoil shapes in wood and stack them spanwise. You can cut several rough shapes at once, stacked on a band saw. There's a lot less shaping to do that way. For a laminar-flow wing surface, you have to maintain the correct airfoil shape to 0.01"-0.02" in the laminar region on the final wing to actually get the low drag values you're expecting to see from the airfoil data.

I'm also curious - even if you intend to sell plans eventually, why not simply go to a Rutan-style hotwired full-depth foam and glass wing? This tiny wing seems to be a natural for that. Even with the thick airfoil, the wing is so small that any weight penalty for full-depth foam would be minor.
 

Grimace

Well-Known Member
Joined
Jan 2, 2009
Messages
319
Location
Chicago, IL
I'm also curious - even if you intend to sell plans eventually, why not simply go to a Rutan-style hotwired full-depth foam and glass wing? This tiny wing seems to be a natural for that. Even with the thick airfoil, the wing is so small that any weight penalty for full-depth foam would be minor.
I think full-foam makes sense for the middle third of the wing... assuming I have truly abandoned the retractable gear as I swore I would.

However, it doesn't seem particularly useful for the rear third which will have flaps or ailerons across the full span. It's a small wing and cutting a solid core so thinly would be a lot more difficult than using a thinner sheet. This would also allow the shape of the slot to be changed... which means the low speed aerodynamics could be easily tweaked if necessary without ripping into the entire wing.

As for the front third of the wing, I would really like to make it a solid core. It would solve a lot of issues. And it actually would be perfectly feasible for so long as I was using a fixed leading edge.

Maybe you've convinced me to go with a solid core over the first 2/3rds with a thin sheet for the top half of the rear-most third. It really would be easier to build, wouldn't it?

I guess the method I proposed here is just residual from my ultimate plan of using a drooped LE. I guess the solid core does make a lot of sense. Cross this bridge when I come to it...

So uhhh... How does the rutan-style wing account for wing twist? Obviously, you run the fabric at 45 degrees, but is there another structure for this purpose commonly used in such designs?
 

Topaz

Super Moderator
Staff member
Log Member
Joined
Jul 29, 2005
Messages
14,768
Location
Orange County, California
...So uhhh... How does the rutan-style wing account for wing twist? Obviously, you run the fabric at 45 degrees, but is there another structure for this purpose commonly used in such designs?
In most cases, as I understand it, the wing skin itself forms the entire torsion box in a full-depth foam structure. I believe that, in most of Rutan's earlier designs, the entire wing panel from the leading edge to the aft spar is the torsion box. It bonds to the spar caps top and bottom, and the spar begins to concentrate the torque loads as it approaches the root into a more conventional torque box between the front and aft spars.

Structures aren't my strong suit, so take that with a bit of a grain of salt. I'm still learning.
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
8,593
Location
Saline Michigan
Would I use it again? For a one-off, you betcha! If I was considering production, I would build a plug and female mold. My alternatives for a one-off airplane were:

Plug, Female Mold, Vacuum Bag Part;

Male Mold, Vacuum Bag Part;

Foam Box, Wet Layups.

When I got into it, I decided that I wanted complex compound shapes and light weight, so a mold became necessary. Since it was a one-off, the male mold made for a lot less glass/ epoxy/ fairing... The parts look good, their shape is not highly critical, I used the same templates to cut the plan view for top and bottom halves, so the parts fit together.

I used strips of foam hot glued to the mold at part edges to offset edges for bonding, and even glued the three pieces (sides and belly) of the tub together on the mold. It will require as much fairing and finishing as a Rutan style fuselage. With vacuum bagging I have 0.6 pounds of epoxy for every pound of glass in my parts. Light!

For wings? The foam supported flashing skin mold makes a quick, accurate mold. I tried the rumored wooden frame and two layers of sheet metal in smaller parts, and did not like it. Went to foam core support, and it worked. I knew that somebody had to have already come up with this method, and I was right. At Sun-n-Fun the next spring, and attended a talk by Alan Shaw, and he discussed a similar method that he was using for making large parts for the Velocity. He uses formica to skin his molds.

On other topics, well, If I was not filling my wing with landing gear mounts and fuel, I would have used a solid foam wing. If you vacuum bag the glass to the cores, it looks like the single lightest way to build a flying surface until it gets really big. And there is no reason not to do it, even with a slot in the aft part, although I suspect there are better ways to keep the aileron's working down into stall. I actually initiated a thread on that topic too. https://www.homebuiltairplanes.com/forums/design-structures-cutting-edge-technology/3168-aileron-design-what-works-why.html

Good luck!

Billski
 
Last edited:

Grimace

Well-Known Member
Joined
Jan 2, 2009
Messages
319
Location
Chicago, IL
Sorry guys.. after looking at the prices for solid sheets of pvc, I'm back to the male mold idea... :ermm:

$623 for a sheet of 3 inch foam the length of the wing??? And I'd have to add some extra material close to the spars because that's still not thick enough. That's for a 48"x96" sheet. Given that I'm now talking about a 30" chord, granted, there would be some cut off and I could maybe use thinner sheets bonded together to optimize it... but as a ballpark estimate of what I'd need, that's pretty high for a plane meant to have a 9 foot halfspan... It's actually more than what I figured the complete wing could be built for... I could build ribs out of fiberglass or even spruce and be way ahead. So I'm back to the male mold idea...

I'm not yet convinced that a quarter inch PVC core, between two plies of glass on each side, is going to signficantly challenge my ability to come up with an accurate final shape... I mean, if the EZ builders can make an airfoil shape that accurately with solid foam, I should be able to do the exact same thing with the male mold... it's basically the same thing... just hollow in the middle...

Of course though, the leading edge would be a challenge if you were going to bend it out of foam... and that was giving my brain a tickle... obviously, the thickness of the foam could vary dramatically as you squish it around a tight leading edge... So that got me to thinking... what if you score the core every eighth or quarter inch all along the entire leading edge, getting say 2/3rds of the way through... use a heat gun to get the last bit of bend out of it... then slurry in the score lines.

The cost would work out to something like 30 yards of unspecified fabric (64 feet, plus waste, depending on the bias used and whatnot); $200 in 1/4" PVC to skin both wings; and a gallon of epoxy.

That's a heck of a lot cheaper than the solid foam core. The price difference would pay for a vacuum pump and a compressor...
 
Last edited:

Topaz

Super Moderator
Staff member
Log Member
Joined
Jul 29, 2005
Messages
14,768
Location
Orange County, California
Sorry guys.. after looking at the prices for solid sheets of pvc, I'm back to the male mold idea... :ermm:

$623 for a sheet of 3 inch foam the length of the wing???...
Try these instead:

Polystyrene

Styrofoam (What the EZ guys use)

Both are considerably less expensive than PVC. If they can be used on the EZ's, you can use them, too.
 

Grimace

Well-Known Member
Joined
Jan 2, 2009
Messages
319
Location
Chicago, IL
Try these instead:

Polystyrene

Styrofoam (What the EZ guys use)

Both are considerably less expensive than PVC. If they can be used on the EZ's, you can use them, too.
But the density for both of those cores doesn't really go up. So why not build a wing that can take a little more abuse if the costs are comparable?

A traditional higher-density cored structure would be far more optimal than a low-density solid core, I would think... and the weight would be about the same....

Frankly, delamination scares me...
 

Topaz

Super Moderator
Staff member
Log Member
Joined
Jul 29, 2005
Messages
14,768
Location
Orange County, California
But the density for both of those cores doesn't really go up. So why not build a wing that can take a little more abuse if the costs are comparable?

A traditional higher-density cored structure would be far more optimal than a low-density solid core, I would think... and the weight would be about the same....

Frankly, delamination scares me...
Works for the Vari-Eze, Long-Eze, and just about any other full-depth foam airplane out there. They don't seem to have any problems with it.

And this stuff is about as "traditional" for a composite structure as it gets. Rutan has been using styrofoam like this since the '70's, and the Darmstadt guys that invented the full-depth foam technique did so back in the '60's.

Use your gut for a reality check, and a calculator for a design tool. ;)
 
Top