Long before the advanced composite world developed into a very complex design for emerging aircraft there was laminated wood where one could take a product that came from a tree and shape it into a very strong structural member. Wood is classified as either “hardwood” or “softwood.” This determination is based on the shape of the leaf of the tree only. Hardwood comes from broadleaf trees while softwood comes from conifer bearing trees with needle-like leaves. Wood, in its simplest state has maximum strength parallel to its annual growth rings and for this reason wood only had good strength in its longitudinal direction. When plywood was developed it combined good strength in two directions, longitudinal and transverse.
FIG.01 If one refers to the WARP CLOCK as developed for advanced composites to show fiber direction in a layup, we can apply this new technology to old technology. The warp clock is used to show fiber placement direction in a laminate, the zero line being in the direction of the major load, in the case of a fuselage it would be nose to tail and in a wing it would be from tip to tip. Thus the zero degree would resist bending moments of lift and wind gusts. The 90-degee line would be chordwise in a wing or at right angles to the zero line in a fuselage. The + and – 45-degree fibers would resist twisting moments in the structure.
Now, with all that data satisfied, let’s look at a common piece of aircraft grade plywood. Longitudinal grain direction (zero) is determined by the outer or face grain. That would put the core grain direction at 90-degrees. When viewed from the edge, each outer or face grain veneer is generally half the thickness of the core and thus if the thickness of both face veneers are added together they will equal the thickness of the core. That is done to make plywood strength the same in both the zero and 90-degree direction. If all three veneers had the same thickness, then the plywood would have most strength in the zero degree direction and somewhat less strength in the 90-degree direction.
For standard 1/16” plywood, the core would be 1/32” thick (usually poplar) and the outer face plies would be 1/64” thick for a total of 1/16”. This gives the plywood the same strength in the zero and 90-degree direction. For a 3-ply veneer the grain direction would be 0-90-0, for a 5-ply veneer the grain direction would be 0-90-0-90-0 and for a 7-ply veneer the grain direction would be 0-90-0-90-0-90-0. The dark numbers indicate the center plane of the veneer stack called the core and the layup is symmetrical about the center line.
Aircraft grade plywood is manufactured in 4’x8’ sheets and the longitudinal (zero-degree) grain direction is along the 8’ length. There is 45-degree grain plywood available and in that fabrication the face grain direction is 45-degrees to the length of the sheet but the angular placement of core and face is still 90-degrees. FIG.02 A beautiful partial sheet of 3/32” 45-degree grain mahogany plywood I have had since around 1960 when restoring my first airplane, a Fairchild PT-19. I recall paying an outrageous price of $35.00 for a full 4’x8’ sheet. Today the cost is around $250.00, therefore you don’t see it much anymore. Note the different colors in each of the veneers. The face and back- sides were mahogany veneer that was 7-7/8” wide while the core is poplar.
FIG.03 During early manufacturing there was specialty plywood made specifically for the OEM. I am thinking of the Lockheed Vega airplane of the late 1920’s and plywood used to skin the wings and empennage. Lockheed used a two-ply veneer spruce plywood for wing skins making the wing lighter in weight than if mahogany or birch was used. Sitka spruce is a member of the softwood family of woods.
FIG.04 The left wing panel for the Hughes HK-1 flying boat being assembled in the Culver City, California factory. Note the very large 45-degree grain birch plywood panels glued to the structure. This plywood is fabricated using the Clark/Fairchild Duramold process previously discussed. In the lower center of this photo is a man standing next to one panel that gives a good reference as the size of these flat plywood sheets and the wing structure. Pressure on the glue joints are provided by nailing strips that can be seen crossing each panel in the spanwise direction. Grain direction of the outer ply would appear to be +45-degrees.
A wood planer stands in the lower right corner of the photograph that is vented through the wing and outside of the building. Note the cut-outs in each main wing rib that would allow for a crawl-space so a man could access the backside of each engine firewall and even crawl out to the wing tip. When the aircraft made its one-and-only flight, there was a man stationed behind each engine firewall and at each aileron. Look very close and you will see part of the four engine mounts attached to the front spar. Aircraft plywood is a stack of thin veneer made up mostly from the hardwood family – birch, mahogany, basswood and poplar being the most common. However, members of the softwood family can be molded into shapes that produce very strong structural components. Laminated wood is described as several plies of thin softwood with a parallel structure. Many wood wings have a laminated wing bows constructed of several layers of spruce, each laminations being about 1/8” thick. With some wing bows it is necessary to soak or steam the wood in order to make the correct curvature design.
FIG.05 A sketch showing laminated bulkheads, sometimes called formers. These formers were often used in monocoque fuselage designs, such as the Lockheed Vega. Laminated formers were generally spruce strips steamed to take the shape of the mold. Aircraft grade spruce has a moisture content of 8-12 percent and the general rule for laminating is that there should be not more than a 2-percent spread in moisture content when laying-up the part. Therefore it is best to cut all strips (when possible) from one piece of spruce so that moisture content is all the same.
Assembly of the formers was in a complex fixture that could be disassembled for removal once the molded plywood skin was glued in place.
FIG.06 A sketch of an internal fuselage fixture designed to hold each laminated spruce former in place so pre-molded plywood skins can be glued using a cold-setting adhesive and suitable bands to hold pressure on joint until it cures.
FIG.07 A Lockheed factory
photograph showing a monocoque fuselage having been removed from the fixture. Note bands around formers to hold pressure until glue cured. This was the process of constructing molded wood fuselages back in the late 1920’s. From this photo it would appear that the outer ply of veneer has grain direction of zero degrees or parallel to the longitudinal axis of the aircraft. Inner veneers were no doubt + and – 45 degrees because veneer more easily conforms to the mold in 45-degree direction.
For large wood structures such as wings and center section, box spars were widely used because they afforded great strength but were light compared to a solid spar of the same dimensions. Box spars consisted of laminated spruce flanges separated by filler blocks. The interior was coated with varnish and spruce or birch plywood veneer called webs were glued to the flanges to form a structural box assembly. This spar was quite strong but light in weight. The Hughes HK-1 used massive box spars in the wings, horizontal and vertical stabilizers.
FIG.08 Center section box spars from an old friend, the Fairchild PT-19 series aircraft – my first restoration project back when I was 18-years old. These photos show front and rear box spars with the back-side webs glued in place. There are upper and lower heavy spruce laminated flanges separated by spruce filler blocks where ribs and major fittings attach, such as wings and landing gear attached. These spars were originally fabricated using Casein glue that, at the time of manufacture back in the early 1940’s, was the very best technology had to offer.
I have repair experience on solid, laminated, externally routed, box and “I” spars and can report that the ease of the repair begins with the solid and laminated spars, with the most difficult being the box spar. I’ve never seen an internally routed wing spar. As I research and write this column the question that keeps appearing in my mind is, “How do you make repairs to these components once they are damaged or deteriorated, especially molded plywood veneer with special grain direction imbedded in the wood. I must say that the easiest spars to repair by splicing are the solid and laminated variety. Splicing an externally routed spar is more difficult, as is splicing a built-up “I” spar. I have made repairs to box spars but not splicing the entire spar. In most cases of design, little thought is paid to the poor mechanic who has to maintain and repair spars. I suspect that splicing of wing spars is a dead art with most mechanics simply replacing the entire spar.
FIG.09 From an old AAC manual on wood repairs, this sketch of various methods for manufacturing wood wing spars. Most spars I’ve come in contact with were either spruce or douglas fir. The easiest to hand work and bond is spruce with Douglas fir coming in second. Douglas fir is slightly denser and has a tendency to splinter when hand planed or sanded.
FIG.10 Methods employed by designers to firmly attach wing ribs to wing spars. Detail A shows a molded plywood angle that is glued to both the rib and the spar. These veneer angles were formed on a heat brake and consisted of two or more veneers bonded together. The angle is very light and offers excellent strength to the glue joint. Detail B shows a typical triangular spruce fillet that is more commonly seen in wood structures. It too adds strength to the rib-to-spar glue joint.
FIG.11 A two-veneer mahogany 90-degree angle from my collection. It is the same as sketched in detail A. I was told this sample came from the Hughes factory when they were building the HK-1 flying boat, but have no certification of that fact. In any case it’s just a neat souvenir. The bend radius is approximately 5/32” on this sample and the grain within the bend lines is very smooth. It is obvious that this mahogany angle has been steam bent on a special bending brake designed just for wood. Here is how it was done.
FIG.12 Reference is from ANC Bulletin 19 showing a sketch of a steam bending machine, circa 1940’s. This machine was automatically controlled by cycle timers, so that it is only necessary to insert flat sheets of plywood veneer that had been steamed or soaked and then remove the bent pieces after an interval of 2 to 5 minutes. For severe curvatures the plywood was soaked for 1 to 2 hours at 150-degrees F and then bent. The mandrel was usually maintained at a temperature of 300-degrees F or higher. The Duramold process of forming and bonding ultra thin wood veneer was adapted to making many different products during the 1940’s. In particular wood air ducting segments and propeller spinners were fabricated.
FIG.13 From ANC Bulletin 19, a Duramold air duct fabricated from birch veneer. Note 45-degree grain orientation to the major axis. In order to fabricate a part like this, bag molding was used. Most bag molding processes used positive air pressure up to 100 psi to tightly compact and veneer to the mold and hold it there until the adhesive cured. In the case of Duramold, the temperature to cure phenolic resins was 250-280 degrees F.