Tips for wood frame restoring
Tip no.1 - Gluing wood
It is strongly recommended that animal glue is not used, even for the sake of originality. When the car rots away again, it will end up less original! Most PVA adhesives are also not suitable (although cross-linked PVA is used by coach-builders). The ideal type of adhesive is phenol-resorcinol formaldehyde, a two part formula of resin and hardener. Often recommended (and used for boats and gliders) is the brand still commonly known as Aerodux 500 (as produced by Cyber Geigy in the past). It is now known as Prefere™ 4050 (by Dynea). It has very high water resistance to WBP (Weather and Boil Proof) grade. It is also good at filling gaps if a joint is not a perfect fit, without much loss of strength. It is less suited for repairing trim, because of its dark colour.
"Aerodux" adhesive needs to penetrate the surface of the timber, and it is worth spreading it onto any surface firmly. This is especially true for joints that cannot be clamped very tightly and have no screws, such as small repairs to edges. The surfaces of the wood should be freshly cut or abraded, but smooth, for the best adhesive performance. Although listed by its makers as "cold setting", some sources of data suggest that warm ambient conditions (about 21 deg.C/70 deg.F upwards) are advisable during curing, although it seems to perform well down to slightly lower curing temperatures. I usually keep joints clamped for 48 hours to ensure that the adhesive has cured sufficiently. It takes a few days to reach full strength.
Some restorers prefer not to glue at all, and rely upon the screws. When jacking up a corner of an unglued frame, you will hear lots of creaks as the joints move. Not only will the car creak and 'crack' a lot, but the panels will be more prone to cracking. When the car twists, the rectangular boot lid opening will distort to a rhomboid shape, because the joints accomodate the overal twist. This over-stresses the panels near the corners of the boot lid opening and they crack at the welds. With glued wood joints, twisting of the car is taken up more by bending of the wooden members, and thus moving the stresses away from the corners. Unglued joints have considerably less strength too. Unglued joints also let in moisture and start the rotting process over again. In fact, unglued joints are even worse than the original animal glued joints.
For more information on adhesives, scroll down this page.
Tip no.2 - Choice of wood screws
Screws need to be driven into wood frame joints very tightly. Therefore slotted heads are preferred, and the material should be either mild steel (possibly zinc-plated for rust resistance) or stainless steel. Aluminium and brass screws are not strong enough.
The sizes of the screws need to be chosen carefully. You can get a rough guide from AC's choice of sizes, and their practice matches present day structural calculations on screw joint strength. Most major joints used no.10 and no.12 screws - no.12 being used for fixing pillars to the sills. No.6 and no.8 screws are only used for smaller components. There are limits to how long a screw needs to be, since an overly long screw would break under tensile load long before it would tear out of the wood. For dense hardwood, no.6 and 8 screws don't exceed 1.5" (38mm) in length. No.10 gauge screws go up to 2" (50mm) and no.12 screws up to 3" (75mm).
If the depth of thread penetration of a screw is short (due to the wood part being fairly thin) then a large gauge should be selected. There are empirical calculations for the withdrawal force to tear a screw out of wood, assuming that it is driven into the side grain and that the timber is seasoned. For seasoned ash, and a screwthread penetration of about 0.5" (13mm), the withdrawal force would be approximately 540lbs (2.4kN) for no.6 gauge, 650lbs (2.9kN) for no.8, and 760lbs (3.4kN) for no.10. The loads are proportional to penetration length and screw diameter. The withdrawal forces for smooth nails are much lower. AC used nails on a few joints and it is advisable to replace them (or reinforce the joints) with screws.
Tip no.3 - Inclination of wood screws
Wood frame joints are fairly crude but effective. Joints are pulled up tightly by the angle of the screws. For halving joints, the screws are not quite square to the plain of the joint, which allows them to slide the two pieces of wood - during tightening - until any gaps are closed up. Annoyingly, gaps may still appear when tightening screws and this could be due to the clearance hole being drilled off-centre to the pilot-hole. It is better to drill out the clearance hole by removing the component and drilling from the other side, carefully aligning the drill with the directon of the pilot-hole.
Tip no.4 - Removing rusty wood screws
If you are lucky, many screws will undo with a screwdriver! An impact screwdriver might help on large screws that are only slightly rusted, but care needs to be taken not to either break the screw, or destroy the wooden thread. More badly rusted screws will need to have their heads drilled off. This is awkward when a component is held by screws that are at various angles (maybe at right-angles to each other). In such a case, some screws may have to be fully drilled out. If you are lucky, you may be able to get pliers onto the screw head to undo it.
The worst screws will have rusted to almost nothing just under the head, while the thread has gained a large ball of rust debris, locking the screw in place. They will probably break, leaving the threaded section of screw buried. A small hole drilled alongside, will allow you to dig it out, and then the hole can be plugged up with new wood glued in.
The rear wings are secured by large wood screws. One option is to force the aluminium washers off the heads (they can be straightened out again), and then pliers can be used to remove the screws.
Tip no.5 - Timber
It is important to use only the best quality timber, although I confess to be no expert on this subject. Air seasoned English ash is rated as the most suitable, along with American white ash. AC (like many manufacturers) also used beech and even some mahogany. You will probably find a mixture of the three timbers on your car. However, ash is the first choice. There is little to choose between a number hardwoods for strength, but ash has additional properties of value, such as always springing back to its original shape after being heavily loaded. It also resists splitting when hammering in panel-pins (beech is inferior in this respect) and is quite a hard wood. Hardness matters if you have to fold and hammer panel edges over the wood frame. Ash also has a very high shock-load resistance (although woods often have a higher strength for transient loads than for continuous ones). Oak is less favoured because it is more acidic and likely to corrode any metals in contact with it. Don't panic if you spot woodworm holes under the bark of new timber as this is something that affects newly felled trees. Mahogany is slightly less strong, although more resistant to rot than ash and beech.
Be careful that any new timber is fully seasoned, and ideally air-dried (rather than kiln dried). It is a good idea to purchase timber months in advance to be more certain. Any thick timbers that are not fully seasoned, may warp after cutting into thinner sections. The curvature of the warping is in the opposite direction to the curvature of the growth rings.
For plywood, there are two considerations: The timber, and the adhesive. The latter is where the AC was seriously lacking. You need to look out for Weather and Boil Proof (WBP) grade adhesive (or "exterior grade") in plywood. Be careful that you really do receive exterior grade when ordering. Inspect the plywood and check that the glue lines between the plies are a dark reddish brown. If you can't see any dark glue, then it is probably interior grade. The timber should be birch throughout for strength. Note that some birch plywoods only have the outer plies (or alternate plies) of birch and rest of softwood.
It is generally recommended that the edges of plywood are "sealed in". For our purposes, I would suggest using the "Aerodux" adhesive to seal the exposed edges.
Tip no.6 - Joining boards end-grain to end-grain
End grain to end grain joints will be stronger if they run diagonally across the boards. Door steps and sill-boards tend to rot away from one corner, so this is convenient anyway.
Tip no.7 - Joining rails to the edge of plywood
Most of the plywood panels have a number of hardwood rails attached to their edges, by driving a screw into the board parallel with its plane. This tends to split the plies when the screws are driven in. A more upmarket solution used in coach-building, was to embed a short length of dowel in the plywood, at right-angles to both the board itself, and the screw to be inserted. A tapping hole would then be drilled through the dowel, and a clearance hole through the plywood. A less elaborate alternative, is to firmly clamp the plywood while drilling the tapping hole, and also while driving the screw in, which also maintains originality for the AC.
Tip no.8 - To paint, or not to paint
Wood frames of coachbuilt cars were generally not painted for protection. Welding up of panelling on the frame might have had some influence on the lack of paint used. Maybe the top coats or primers available were not considered ideal? I think Morgan painted their frames at one time, as do some restorers, and I consider it advisable to do so for longevity of a restored car (especially to keep woodworm away!). As well as protecting the wood, the paint will also protect any steel or aluminium that is in direct contact with the wood frame. It is important to use a paint that allows the wood to "breath", as humidity changes. Marine paints that seal the wood below the water-line should not be used for a car. A one-part polyurethane will probably be best for the top coats, as this is very hard-wearing. These are often used on boats above the water-line and so marine suppliers can come in useful. The ideal primer is the aluminium based type, which can cope with various wood painting challenges, including old wood that was creosoted (such as the AC's floor panels).
Usual practise has been to paint the frame black, but I prefer something lighter. This is partly to make it easier to see and clean from under the car, and also because the woodwork in the boot will be flock-sprayed a light brown on my own AC. A pale colour also shows up any cracks in joints that need to be attended to, such as old joints that have lost glue adhesion. Panel squeak is a potential problem, with the bodyshell resting upon a glossy paint finish. I intend to lubricate the appropriate areas with wax polish. The shell touches the wood frame around the window openings, and at the top of the boot area. The roof framework already has strips of padding to prevent rubbing. Silicone rubber sealant may also help to reduce squeaks and keep rain water out, if applied under the panel edges.
The above photo is a close-up of where the right-hand cantrail meets the rear shelf. The gusset block is nailed on, but you can see that the original glue has lost adhesion by the crack in the paint. The paint has also cracked on the lower nail-head. This joint needs to be reinforced with screws and re-glued. In fact, it may be a safety issue to reinforce this particular joint, since it affects the strength of the roof.
Tip no.9 - Protecting screws and bolts
The wood screws and the bolts that pass through the wood frame will rust very quickly unless they are protected or use is made of stainless steel. Mineral oil/grease should not be used, because this will speed up rotting of the wood. Some restorers seal in the fasteners with epoxy adhesive. Another solution is to coat the screws, coachbolts, etc. with Waxoyl. Credit for this tip goes to Alan Alderwick, author of "How to Restore Wooden Body Framing" (Osprey - 1984).
Tip no.10 - Nuts and bolts
Never mix nuts and bolts that have different threadforms, even if they appear to fit (and even if some merchants recommend it - sheesh!). So, DO NOT mix UNC and BSW, or BSF and BSCy, even though some sizes appear to fit (similar general dimensions).
Also, NEVER cut a thread with a die - at least not on anything remotely important, which covers most of a car. A thread cut this way will lack any kind of run-out or stress relief groove, etc, at the end of the thread, and so increases the risk of fracture. The thread itself will also be poor quality. The best external threads are rolled ones, because the 'grain' of the metal flows along the roots and crests of the thread.
Don't replace high-tensile bolts with stainless steel, even if the ultimate strength of the stainless steel appears to be high enough. The effective yield strength of stainless is relatively low. If replacing mild steel fasteners with stainless steel, avoid contact with aluminium components, otherwise corrosion will be accelerated.
Tip no.11 - Screw head recess
It is important that screw heads cannot touch the inside of body panels, so the holes should be countersunk deeply (possibly using a small diameter countersink).
Tip no.12 - Panels creasing over the wood frame
The body panelling should not touch the edges of the wood frame components, except where the panel edges are fastened on. Otherwise the panelling may crease slightly, which will show up on shiny paintwork. This is why some faces of wooden parts are at a slight angle to the panelling that covers them. There are a few exceptions to this rule, where the bodyshell does rest upon the top surface of some wood frame parts, and these parts should have their edges rounded off (if they are not rounded by design already). The roof frame also has strips of waxed card for the roof panelling to rest upon.
Tip no.13 - Panel-pins
The body panels are secured with hundreds of panels-pins. These are often near the edges of wood frame components, and so the pins should be inserted at an angle to reduce the likelihood of the wood splitting. Drilling pilot-holes may help, especially if the wood is beech which is prone to splitting.
Galvanised pins might be a good idea, and/or dipping them in waxoil. Stainless steel pins are theoretically worse for causing aluminium alloy corrosion by electrolytic action. However, with a large ratio of panel to pin surface area, this is only going to be a problem if water gets trapped around the pins (electrolytic corrosion depends on the ratio of wetted areas of the 2 metals). So, mild steel or stainless steel pins should be satisfactory as long as panel edges are sealed against water ingress.
Tip no.14 - Steaming and preserving
Steam bending is often discussed in the context of car frames, but this practice reduces the strength of the wood. There were no steam-bent parts on the AC's frame. Morgan don't steam parts that they laminate. Reduction of strength also arises if you apply water-based preservative, so be sure to use oil-based preservative on a structural frame.
Adhesives: More information
One's choice of adhesive will come down to various factors, including moisture resistance (when cured), moisture tolerance (when curing), strength, curing temperature (and your local climate), ease of assembly, choice of timber, surface finish of timber and ability to penetrate dense hardwoods.
This 2-part adhesive has the highest rating for structural strength and is used for buildings, boats and aircraft. It has high heat resistance, and the very highest moisture resistance. Some types provide gap filling properties so that joints perform well in real life rather than just laboratory tests. It is tolerant of timber with high moisture content during curing. It requires moderately high clamping pressure, but should not be overly tight.
Dynea manufacture PRF adhesives including Aerodux 185 (with powdered hardener) and Prefere 4050 (formerly Aerodux 500) with liquid hardener. The latter is my personal choice. Both have gap-filling properties. Prefere 4050 is listed by the makers as "cold setting", but most data on PRF adhesives suggest at least 21 deg.C (70 deg.F) for curing. Practical experience suggests that slightly lower temperatures are satisfactory. I've glued external house repairs at around 10 to 15 deg.C (50 to 59 deg.F) which are still sound after 6 years of the British climate. If you purchase the fast setting version, Prefere 4050F, it can be used at slightly lower ambient temperatures.
Other formaldehyde-based adhesives
Phenol-formaldehyde (PF), melamine-formaldehyde and urea-formaldehyde cure at elevated temperatures and are therefore unsuitable for our purposes. Resorcinol-formaldehyde (RF) cures at room temperature. PF and RF resins have a similar high resistance to moisture plus a high structural strength to PRF resins.
Epoxy is another 2-part adhesive with high structural strength and cures at room temperature. Although rated well for moisture resistance, it is not rated as highly as PRF resin (at least less resistant to wetting and drying cycles). It is also less tolerant of high moisture in the timber when curing. However, it is still worth considering as many boat builders use it successfully (cold moulding hulls), as do car builders, and it is considerably better than the original animal glue. It requires lower clamping pressures than PRF resin.
This has a similar rating to epoxy, with high structural strength and a moisture resistance only slightly limited by repeated wetting and drying cycles.
Most PVA adhesives have only moderate moisture resistance, weaken at high temperatures and are not rated well for strength in prolongued loading. Cross-linked PVA performs better in these respects than regular PVA, and is used by coachbuilders, but is not as highly rated as PRF, epoxy or polyurethane.
The original wood glue for our cars, but very poor water resistance, hard and brittle, and not officially recommended for structural work. Definately not recommended for coachbuilding, unless the car is going to be kept indoors and rarely used.