Tuesday, October 27, 2009

In the Frame ~ Part One.



Shaping a Set of Molding Knives.

So let’s get down to business here. Making moldings. What I show may not be directly practicable to many, but it may inspire you to think about finding your own way to do framing, and I think it worth documenting, as a personal adaptation of woodworking techniques first developed in the mid nineteenth century, which gradually replaced those old wooden molding planes that you still can find in the antique shops. Every cabinet maker used to have his own collection of these. I’ve owned and used them myself, but long ago gave that up in favour of a Sears Craftsman three-bladed molding head, which will mount onto either a table saw or a radial-arm, which is what I have. This naturally does a far more even and accurate job in a fraction of the time. This is still available, and cheaply. Here’s what the set looks like. You can still get it from Sears for the same price I paid about thirty years ago - $99; (DSCN1714). - Please refer to my 'Flickr' site to see all the illustrations: http://www.flickr.com/photos/a_henwood/sets/72157614565091611/.
Disappointingly, Sears does not seem to sell the cutter bit sets separately any more as accessories, and neither could their parts department help. Fear not, however; all is not lost. There is a company called Corob Cutters - http://www.corobcutters.com/ - which offers an up-to-date molding head of the same type, and, they offer an extensive range of molding knives, which, (bless their hearts), also fit the Sears head. All this at most reasonable prices, so if you want to try this approach, and have a table or radial-arm saw, you can get into it without breaking the bank. If, on the other hand, you decide this particular approach is not for you, you can instead create a very respectable selection of mouldings using a router and some of the huge variety of bits on offer these days.
I have put my own personal stamp on framing by designing my own mouldings and then filing (by hand) sets of three cutter blades to reproduce the chosen profile. The first, and easy, step is drawing a profile, sometimes copied or adapted from one I have seen and like; (illus - ’Frame Designs 4’ - on 'Flickr'). For most watercolours I usually use a simple, slim frame moulding, for larger, more important pieces or oil paintings, I choose a wider, more ornate design.
Here’s how I made a matching set of three cutter blades to create the picture frame molding shown here : (DSCN5472). These are in High Speed Steel. In this example I start with the plain, square-ended ‘planer’ blade (DSCN1707 - 'Flickr'), though often I can save effort by modifying another ready made shape. The trick, of course, is to achieve the identical shape on each of three blades. This is not as hard as you might think. The task is made quite simple by using a stepped holder to carry the three blades at once, stacked together; (DSCN1705). The wooden holder is designed to support the blades with the bevels flush with each other, and is held in a vise so that these surfaces are level and ready to be worked on; (DSCN1708). Observe from this same illustration that quality control of the delivered cutters is pretty spotty: these three are supposed to be level and even! Remembering that the blades are going to cut at an angle of around 40º, draw an approximation of the profile on the face of one of the blades; no need to do all three. We are not going to use computers or machine tools here, just files and maybe a grinding wheel or point in a Foredom or Dremel flexshaft. I prefer filing, as with grinding it is easy to overheat the metal and lose the temper of the steel; also it is harder to attain an even shape. There will be no precise gauging or measurements here: in the end the eye will tell all. A good selection of small files will do all the work; (DSCN1710). With the blades held as described, just start filing, taking care to keep the tool as straight and level as possible; attempting to cut, in the same pass, an equal amount from each blade. You won’t achieve this without a most important next step: after removing a reasonable amount of material it is time to shuffle the deck! By this I mean – unclamp the blades from the holder and change their order: e.g. if we say they were stacked as 1,2,3, then change that to 3,1,2 for instance. When you do this small discrepancies will show up, (DSCN1711) and you take these differences out with the next filing. Repeat this step from time to time as you progress. Because of the angle of the blade when it contacts the wood, the cut moulding will be shallower and less pronounced than the shape appears on the knife, that is, you must grind the knives to a slightly deeper, more exaggerated extent to achieve the desired cut.
When the knives approximate the shape you want, remove them from the holder and dress them individually with waterstones or oilstones; (DSCN1716).
Now it is time for a trial cut. Mount the cutter blades in the molding head and mill a cut on a short length of wood. You don’t need to cut more than a few inches. Unless you are extremely lucky, you will not be pleased with the result; (DSCN2022X). Note the double ‘tracks’ shown by the middle arrows. The upper and lower arrows point out that the sides of the molding, which should be evenly rounded, are anaemic or missing. The cutters are not registering together. The first thing to do is swap the blades around in the cutter head and run a couple more trial cuts to achieve the best result. Having done this, now mark the blades across the body or edge with a file, 1,2 and 3, according to the particular slot they fill in the molding head; (DSCN2053). (I have numbered the three slots on my molding head with a small punch). From here on, always mount the knives in their own same slot. Next, turn the head by hand over the trial cut and closely examine where each blade falls. Sometimes millscale is left on the base of the blade which interferes with properly square seating, or the blade may not in fact be cut properly square. This in turn may cause the entire blade to track slightly left or right of the others. This can be corrected by a little judicious filing of the base edge of the blade; (DSCN2054). Most of the discrepancies will by now have been taken out. What remains is small differences between the cutting edges of each blade. Turning the head by hand over the latest trial cut will reveal the reasons. Dismount each blade individually and file out the problem area. A final trial should now yield a perfect profile, but one yet marked by striations caused by the filing process; (DSCN2055). Polish these file marks out now using small slipstones; (DSCN5463). Your set of cutters is now finished, and should produce a nice clean stick of molding needing minimal sanding prior to applying a finish; (DSCN2486). The rabbet to hold the picture is easily run with dado blades, the stick ripped from the board, and the back edge of the framing smoothed on a jointer.
This may all sound like a terribly long, slow, involved process, but actually it is not. The methodical approach I have just set out will be just an interesting morning’s work, and the result will be a set of knives that will be useful for a lifetime.
There are ways to further decorate plain moldings. But I’ll get into that later on.

No comments: