These two questions came in this week’s online mail. I understand where they are coming from. As a vegetarian, I often eat something made and usually intended to be a counterfeit for the real thing. Food technology can add and subtract tastes as well as they can textures and looks. Experts can invent food to demand according to type not as an alternative but just another food. I have eaten bacons made from plants that would stop even the most discerning bacon butty eater dead in their tracks. But I have tasted some that are absolutely terrible too. I’ve learned to accept no substitutes but to see any alternative as just another type. Camp coffee, a Brand, can never replace a Columbian coffee. It is just another drink type in the same way tea is to coffee. With so many choices for plant-based anything you care to name, I am aware that we in woodworking have the same considerations coming from those who engineer flat boards to look like any wood we want to name too.
Now how did I get there? Ah! Yes, I know. It takes me back to my often-said thing and that is that often we have to see what something is not to see what it is. Vice versa, you have to what something is to see what it is not. In my world, as a vegetarian, I rarely if ever look for a food as a substitute but treat all things as merely other types of food or drink. In similar vein, MDF will never be wood even if a large percentage of its constituents might well be from wood and neither can any other engineered material for that matter. Below are a couple of questions needing answers. Comparing the outcomes of machining materials with the outcome of hand tools is not one and the same either. There are remarkable machine processes that hand toolists might feel they need to parallel in their woodworking at the workbench; they can get the impression that the outcome of hand tools should be one and the same. Mostly that is not realistic but it also needs a bit of explaining for clarity as well as determining our ultimate goals. This might lead to a longer blog (my usual bad).
Happy New Year,
Do you have any suggestions on how to successfully plane boards to the same thickness? This has been a challenge for me in my woodworking. I always check boards for true flat surfaces and use this surface to square other surfaces. I use a #5 or #6 plane to do this and a shooting board for the edges. The problem is trying to keep the flat surface dimensions of various pieces the same.
I guess that this comes with experience and taking smaller thickness corrections as you zero in. Any assistance you can offer would be helpful.
John
Hi Paul,
I have begun watching your 1st episode in the new bed series and noted the techniques you are using with the bandsaw and planes to efficiently dimension your material. I have often struggled to efficiently dimension longer material in my attempts to achieve flatness/squareness using your technique of planing a section at a time starting at the nose. I often find myself dealing with humps I develop along the face or tapers on the end. Admittedly, I am pretty focused on flatness/straightness in the quality of my boards, only to find out that the wood will move a bit a day later. Can you help me to better understand the precision goal, in light of where it may or may not be as critical? That part caught my attention as you were planing the legs and rails.
I appreciate the help.
God’s blessings to you, sir,
Andy
Answering John.
Thank you for the greeting. Same to you, John. In the machine world, especially at the more commercial end, machines can true and deliver sticks and stems of any size perfectly flat parallel and fully squared by simply allowing the conveyor belt to deliver the wood to the power feed and in so many feet run per second the sizing of all dimensions is within minute fractions of thousandths of a centimetre. Less sophisticated systems requiring an operative to hand feed the wood in and out of the machines as a system will also give similar quality of cut but without the speed. Home workshops often have a series of four or five machines that rip, crosscut, surface plane and then thickness plane in only minutes. the feel test will show no noticeable difference in parallelity and when machined properly adjacent faces will indeed be dead square.
We hand toolists can have a dozen reasons for non-machine woodworking in our lives. High-rise apartments, neighbour objection, space alone and so on, but then there is the reality that we just don’t want the machine work in our lives. There can be another dozen reasons for this too: Health, safety, noise exercise, etc. The challenge alone is enough. We don’t catch the bus when we go for a hike and neither do we take the ski lift if we want to climb a mountain. In my present world I just don’t want to be a machinist and so keep them to a minimum. Over a 20-year period, I have reduced their presence to one 16″ bandsaw in my single-car garage-sized workshop. It works great.
Looking back in the history of vintage furniture when the bulk of making was making using hand tools and you quickly see the economy of making using body energy only. The unseen parts beneath drawers and the insides of some cabinets own the tool marks left in the wood by the maker. By this, we soon learn from what was made by hand by the masters of their craft, realistic woodworkers who developed specific methods and standards for truing their wood within the restriction of a time frame. You tell me that you straighten and flatten one face and then work the edge using a shooting board. No matter the method or the tools used, these two surfaces are prepared purely to register the layout tools to. The marking gauges reference one or both faces as will the stock of the square. It is also against these two trued faces that we establish parallelity to saw and plane to and it is here that we have to make decisions. How parallel is parallel and what are the parameters we need to work within?
Watching my videos might well give the impression that everything is thickness planed to perfection so that we have equality throughout. In general, I don’t. In some circumstances, I will. Making doors I am pretty high in self-demand. For my recent bed, the bar was much lower. What was the difference? Size, distance between components, strengths in terms of flexing under pressures, tolerance, wasted energies, the interplay between facets and how they relate in all of their key parts; my decision-making regarding parallelity in thicknesses, etc are all made minute by minute. With a bed or table, the components and distances are far apart, long, distanced from one another. I let things slide more because no one can draw a comparison on such parts as the long side rails as to parallelity and thicknesses. It makes no difference if one end of a six-foot rail is an eighth of an inch narrower in thickness or half an inch in width. Nobody can see to visualise or gauge the difference visually.
The face-mark, face-edge marking spans centuries of woodworking. There are other systems too, but in western woodworking, these seem to have stood the test of time. Planing that larger first face true and giving it its marking attests to our work and approval. This will be the face we will use for registering all of the tools too as much as possible. The adjacent face will be trued string and true too, with the addition of squaring it to the first proven face. This too gets its face-edge mark as shown. Using the marking gauge to establish the line of the final dimension for thicknesses from these lines is our task now. Taking a look inside vintage cabinets, the backsides of them often show many discrepancies cabinet makers and joiners lived with. there was no wasted energy resulting in wasted time. A rail to a door might be a millimetre thicker but the prominent face will be flush with the stile. taking a straightedge to the inside will often show that it was planed to a very slight slope to feather in the discrepancy and that was acceptable. Modern machining will work only with identical thicknessing otherwise it just will not work to mass-make in the hundreds and thousands. But we are not working in or even near those realms. Imagine belt sanders three feet wide where all of the doors pass through and pass on to a spray system that delivers the finished doors at the other end.
What am I saying? We hand toolists can be and are for the main part far, far more tolerant than others might often think but at the same time we have to follow certain rules as closely as possible. The two faces face- and edge-marked ensure that we can make a perfectly square and untwisted door. The frame or box will flush or smooth or fully aligned with and by these two reference surfaces. The opposite sides may well have minor discrepancies, small differences in thicknesses and such, but these are taken care of with the plane before or after glue up. Our system using well-proven methods to establish registration faces to synchronise all of our layout and surface planing, etc means that the important faces fully align whereas the opposite faces on the insides will ultimately need some remedial planing to level the parts and this varies according to our level of tolerance when we planned for parallelity.
I have seen many a piece of fine furniture from some historic past where deep kerf from circular saws, adze and axe chops and then pit saw sawing too have been an integrated part of back panels, inside unseen faces and so on. This economy of time and effort was a factor of handwork. The finer the demands of their customers, the less these might be.
Answering Andy:
Following on from my answer to John, it’s a judgement call on many things straightness. Long bends in wider boards for things like the side rails of the bed and bottom rails of the head and footboard really do not need to be dead straight. I might even say reconsider twists in such pieces too. The sheer weight of large frames combined with adjacent boards can tame bends and bows by sheer weight alone.
The two rails and posts to the headboard felt quite light once framed and fitted but adding the slats seemed to about double the weight. I had not expected that, but that mass of weight from all areas would indeed force compliance easily when assembled and standing in place with the mattress installed. Whereas we aim for twist-free stock, it doesn’t take long to lose the thickness for the sake of removing twist. Let’s take the bed. I made certain that the posts for the head and footboard were twist-free, straight and true and the edges were all square. Any small degree of discrepancy here will telegraph into the rails multiple times and those multiple times will throw the whole frame into a gigantic twist. The stoutness of the posts will stand rigid and immovable. The thinner rails on the other hand will flex far more readily. Adding the slats, all 19 of them, totals a width distance of almost 30″.
Combine the pulling power of the screws through untwisted slats over such a combined width will serve with the posts and the untwisted and straight other rail to pull the whole into submission. A hinged door on the other hand must be made untwisted because it only has one fixed and immovable stile. The free or floating stile must be made in direct alignment to the fixed (hinged) stile otherwise the door will appear, as it is, twisted. Tabletops follow a similar consideration. Bows placed in opposite directions while exaggerating the problem can be used to counter one another in glue up. I tend to let some bows go and simply level them to flat across-the-board width only. In 95% of cases, this strategy works just fine. On the bed, I turned the long, curved bows inwards so that putting mattress support boards in pushed the sides out to equalise parallelity. There are many tricks of the trade we use to pull and push the parts to compliance. If a tabletop ends up with a dipped hollow then adding a cross member underneath, between the apron rails will push the hollow up to level the same way if it is cambered the rail serves as an anchor to pull it to with turn buttons, cleats or screws.
And you are right in finding wood that is straight one day, after truing, often takes on some unpredictable curve the next.
