In an earlier life in Silicon Valley, part of my role was what one might call “design for manufacturing” - collaborating with the engineering & deployment teams far enough upstream in the process to provide useful and actionable feedback about the products so that it can be built and deployed in a cost effective manner.
Little did I expect that type of thinking and guidance would be used in my woodworking business as well, but over the last several years, I have had several inquiries to mass produce wood products. While only two of these inquires have resulted in prototype production, and only one of those has turned into an ongoing collaboration, I’ve been happy to provide initial guidance and comments to every inquiry that comes in.
With this post, I’ll touch on a few of the common questions that have come up with these projects:
Can it be made with wood?
What would it cost?
Even though these are just two of many considerations, pages could be written on each of these. And for the sake of brevity in this post, I’ll be leaving out approaches like bent laminations and veneering an engineered surface, though both are valid and common (and can be prototyped at my workshop).
Ikea Poäng Chair, make with bent laminations for the curved parts.
Ikea birch countertop, with a thick veneered top and thinner veneers on the sides and ends, but an engineered wood core.
But I hope to give at least a brief intro to the topic, and always welcome questions over email or via comments on the blog.
Solid Wood vs. Plywood
Wood that would be used for manufacturing comes in two main “flavors”: solid wood boards, and engineered panels. For a visual, you can think of pine or oak boards you might pick up from a Lowes or Home Depot, vs. a sheet of plywood.
Hardwood stack at a lumber supplier; note the uniform thickness, but the variations in widths. The boards typically also have a variation in length. This lumber is sold by the “boardfoot” - one square foot of one inch thickness.
Stack of plywood at a lumber supplier. All the sheets are of a uniform thickenss and length and width, but most plywoods have unattractive edges that typically need to be hidden or covered in some way.
Generally, boards come in random widths (within a range), and random lengths of 8-12’ long, with thicknesses of 1” to 4” depending on species. Engineered panels (i.e.: plywood) come in 4’ x 8’ sheets (or for some products, 5’x5’ sheets) of different thicknesses, typically, 1/8” up to 1-1/2” thick, though most commonly 1/4” - 3/4” thick.
You can purchase plywood with a variety of veneers pre-applied to the surfaces, and potentially also pre-finished, saving a lot of time and effort in production. However, most plywoods have unattractive edges that often need to be covered for reasons of aesthetics, longevity, and to avoid splinters, typically with a process called edge banding. Plywood also has the advantage of being “dimensionally stable” with moisture changes, is typically of uniform strength in both width and length, and is typically cheaper for a similar surface area of wood.
Hardwood-veneered particle board options; veneered panels also come with plywood layers or with an MDF core, depending on thickness required and finishing options or requirements.
Two plywood options that have attractive inner cores that are often left exposed - mitigating one of the major plywood disadvantages - are bamboo-based plywood (trade name of Plyboo) and baltic birch plywood.
Plyboo plywood with different colorations and an attractive core
Baltic birch plywood with attractive and consistent core layers
Plywood is thus the obvious choice for larger flat objects, if the limitations around the edge quality can be addressed - either by choice of material (such as with plywoods that have more attractive inner cores) or by a post-processing step such as edge banding. However, when manufacturing something thicker than the available plywood options, solid wood is more often the way to go, so long as you can address concerns around wood movement.
Of course, a solid wood block has a unique grain pattern throughout its body, which can be both an advantage (i.e.: features cut into the surface at different depths do not need to be treated to hide potentially unattractive edges), and a disadvantage (i.e.: designs that have a lot of “short grain” can be very brittle and compromise strength).
Wood fibers follow the grain, and are themselves very strong. However, the wood is much weaker between the fibers. Where the cuts, for whatever reason, mean that force will be applied in such a way to split across the grain, breakage is much more a risk than if the grain flows across the direction of force. This chair leg broke because of the “short grain”.
This is also why plywood is advantageous, where it can be used, in that each ply is rotated 90 degrees from the layers above and below, making for a much more stable product.
The board is strong in flexing along the grain.
The board easily splits when flexed across the grain.
Additionally, because of the variability in sizes that are typically delivered when purchasing lumber by the board foot, you need to either be able to accommodate that variability (i.e.: by gluing narrower boards together to make wider boards or repurposing for other projects boards that are not wide enough), or to work closely with the supplier to hand pick the lumber from larger stacks, introducing a labor intensive process that also may not scale well depending on how closely your size requirements meet the range of supplier inventory.
Picking up & transporting four large specifically-sized walnut boards - rather than leveraging the free delivery service when purchasing by the board foot - adds time and cost to fabrication.
Cost
At quantity, the cost is comprised of two key elements: the raw materials, and the processes that are applied to those materials.
The raw materials, of course, have a wide range of costs: a 4’x8’ sheet of plywood can range from $30 to $250 depending on the veneer type and quality and the thickness; hardwoods cost between $3 to $40 per board foot (one square foot of material 1” thick), similarly depending on species and thickness. But just to provide some anchors, as of August 2021, an unfinished sheet of 3/4” maple plywood costs about $60, and one board foot of walnut costs about $10.
The cost of the processes depend on the machines involved, the cost of any consumables, and of course, the underlying cost of labor in the market. For instance, sanding and applying a finish to the product might cost $30/hr in the San Francisco Bay Area, whereas machine time on a large CNC might cost $100/hr or more, to reflect the cost of using a machine that can easily cost tens or hundreds of thousands of dollars to install.
Thus, designing in such a way to minimize machine or processing time is also important. One obvious savings is if a particular surface will be hidden post assembly, then it needn’t be finished or even sanded. Less obviously, if CNC time is a big component of the manufacturing process, minimizing tool changes can speed up production, especially if the component can be made by larger end mills (which can remove more material at a time, and be run at faster speeds).
And of course, the few initial prototypes will cost more than the at-volume manufacturing, to reflect multiple underlying costs the shop incurs for the first few iterations, such as building any necessary jigs; advising the client on limitations; and tool changeover time inherent in making a small quantity. This may be reflected as an “NRE” (non-recurring engineering) charge, prototyping fee, or an hourly rate on necessary design time.
A Few Examples
Serving Tray
I was asked to consider making a higher-end wood version of a 1/2” thick tray that was already for sold in a molded plastic form. Given the cost goal considerations of the client, as well as the width of the tray and the fact that some features were milled to partial depth of the tray, the only material option was a plywood that would not require edge treatment like plyboo. Additionally, some of the features of the tray that had sufficient strength in plastic would be too narrow in wood (i.e.: short grain), so the design needed to be reconfigured to have thicker “handles”.
Tea Service Tray
Similarly, another client wanted modest quantities of a tray with raised sides as a tea service tray. The bottom of the tray would definitely need to be a 1/4” plywood given the cost and weight goals; given that the bottom’s edges would be hidden, a lower cost option like maple plywood could work. While plywood would be nice to use for the sides as well, the best option for joining the sides together was an a box joint, so an inexpensive wood like pine or perhaps poplar was the better choice.
Dumbbell
A client approached me to consider manufacturing an attractive dumbbell (the interior of which would be filled with iron filings for additional mass). The only option for something this thick was solid wood, and even with solid wood, thick enough stock was not available. So a glue-up of 2-3 thick boards to achieve the necessary diameter “blank” would be necessary, in addition to all the machining time to remove all the extra material; the raw material costs alone, unfortunately, tanked this project, before even factoring in the cost of the machining.
Picture Frame
I built a few hardwood and MDF frame prototypes for a client a few years ago. At the time, I had a lower cost - and lower capability - CNC. As it did not have a tool changer, any at-volume quantities would need to be milled with a single bit for both roughing and finishing, which limited some of the sharpness that could be incorporated in the corners. However, running on a lower cost machine allowed me to offer a more competitive price that fit within the client’s budget. I have since upgraded my machine at considerable cost to include a tool changer and allow for much faster cutting speeds; though I now need to charge more per hour of machine time, that is made up for the fact that cutting the same frame will be faster, or I could programmatically swap tools (adding some time back in) to get even greater detail in the frame corners.
Wrapping It Up
What questions do you have in the “design for manufacturing” realm? I’m always excited to help with new projects, even if only as an initial advisory engagement or a few initial prototypes before handing off to another manufacturing facility.
In fact, though there are a handful of large manufacturing facilities that can churn out 1000s of an item per week, where I excel is in the initial design to prototype stage, up to production of a few dozen per month. If that’s where you are at in your project, please contact me via email or leave a comment below.