Door and Window Building: CNC Versus Traditional Machinery

Here's a thoughtful and sophisticated conversation about the proper place of CNC equipment and traditional shapers, mortisers, etc., in custom door and window production. October 3, 2011

Alright, I want unbiased opinions from door builders on CNC versus conventional shaper/moulders/tenoner. I don't want you swearing CNC is the best because you plunked down 200K on the latest greatest machine, just honest remarks if you feel it has improved production, quality, or maybe even both. Or do you regret it?

Hereís where I'm at; we have a flat table router we use for cutting panels, templates and we do jambs on it sometimes. We also bevel our doors on it. We don't bore the door or do the hinges on the router. All our cope and stick we do on conventional shapers along with shaping arched top rails against patterns. We raise all our panels on a shaper. I have watched all the videos and read all the literature and still can't decide if there is an advantage to CNC.

Forum Responses
(Architectural Woodworking Forum)
From contributor Y:
I have a flat table also, but since I do custom doors programming becomes a pain in the butt. I know there is software that can handle it but the part libraries can become quite large to set up. To add a new custom door to it might take more time than itís worth. To get an exact order for that same door is and has not happened. I do all the template work on router the rest by conventional means.

From the original questioner:
Thatís where I am at too. There are some attractive deals out there on pod type machines, but with the increased cost of tooling and other things that go into setting them up I'm not sure the value is there. We do custom exteriors and whole house packages of interior doors. A lot of our interior styles will repeat but not necessarily order to order.

From contributor D:
We have two machines that run all day. I guess I'm biased though, so maybe not much help. At times one machine has been down, and we always had to work overtime in each case to keep up. Imagine setting up a shaper with different profiles versus dragging and dropping a tool on a toolpath and typing in depth and speeds. Can't remember the last time I made a template.

Making the actual part takes about the same amount of time. I program in Masterwork and in Alphacam. Takes about 15 minutes to program most custom doors, but I'm not ashamed to say I'm really, really good at it too. I've been to software demos at Biesse and at Stiles, and left laughing at how difficult it was to program compared to what I'm using. Then again, that is a biased perspective. Maybe if I'd learned Biesseworks first I'd think differently - maybe not. If you can't draw an arc and an offset of it in fewer than five minutes and you are the teacher then something is very wrong.

From the original questioner:
Wow two machines, what type of volume are you doing? I was also wondering if you are a convert, or did you guys start out strictly CNC? Do you dowel, bore, hinge and everything on the router?

From contributor D:
Anywhere from about 4-10 mortise and tenon slabs, including frames, casing, etc per day. The CNCís also do other odd tasks throughout the day. We bought our first one around 2002, and another one in 2008. Anywhere we can see a time savings, the CNC is used. I dislike full pre-hung units with complex multipoint locks installed, and sweep on the bottom of the slabs, as we don't finish our units. Half of our clients I know for a fact aren't even unhanging before finishing, which voids the warranty, but still causes problems.

CNC saves huge time for complicated lock mortises, as some companies provide dxf files. We also build dozens of windows and sashes a day, which the routers play a small role in. The most doors I've ever produced in a day is around 30-35, a rush order. Pretty long day though, and they were all pretty much the same simple two panel design going all in the same house. Before getting a CNC router I'd get a RF press if you don't already have one. And there is the learning curve. Don't expect to go guns a blazing the first day, or even the first year. I program probably ten times better and faster than when I started. Each year I look back on things and can't believe I was so stupid doing it that "old way."

From contributor Y:
One more thing I wanted to add. Don't forget time it takes to glue up panels or any parts that need pre-machining before it can go to the router. When I make the stile and rails I use s4s on moulder with outside head for profile. One pass through the moulder and I'm done. Rails just get s4s then to cutoff machine, one shaper for tennon, second shaper for profile. Mullions get cut off, tennon cut, and then two passes through the Hussey for profiling each side.

From the original questioner:
The fact you are chopping and breaking down your lineal stock rough cutting it to length to then have the CNC shape it to length is hard to get my mind around. We precision chop to length as we are cutting out a job, end cope, and through the shaper. As far as equipment, I pretty much have it all including an RF gluer. Iím just trying to figure out if I am missing the boat. I personally know of one shop who went to CNC from conventional (out of business), and two that got into the market by focusing their production around a CNC. The caveat of one of the guys is he spent a lot of years working for a major door manufacturer.

From contributor D:
Rectangles are faster using a moulder and tenoners/shapers. Most of our work is segment double doors, roundtops, ellipses, serpentine french, and other odd geometric shaped stuff, including carvings in-panels, rope or grape leaf moulding, scallop and spiderweb grills, etc. CNC is king for this stuff.

From contributor S:
Normal doors (square raised panel and arch top) are a lot faster than using shapers. If you want a faster way to make doors look at a multi-head profile shaper with plunge coping head. There are also dedicated CNC door machines that are relatively affordable, take up a small space, and can cut pretty much anything for a door very fast. I still think a shape and sand machine is by far the fastest.

From contributor O:
I am an admitted old-school door maker that has made doors with shapers and tenonerís for 40 years. Our volume is around 1m a year, and about 60% of that is door and related work. All custom doors, interior and exterior, so it is rare to get many repeats, except when we do a house full of doors.

I have been collared by plaid pants CNC salesmen that tell me to buy their machine or go out of business. Mass customization will take only minutes, and will put me out of business unless I get onto their particular product. What software to use, actual real world experience, and real performance numbers is where their pitch falls apart.

Our shop is skills based and we can build a curve head door and frame as fast as anyone, and with the quality we don't see in other products. Swinging arcs on the bandsaw, curved moldings, full cope and stick mortise and tenon, all equal superior millwork. We also have proprietary design elements that add greatly to the longevity and weatherabiltiy of our work, as well as design details that help our products stand out and build value.

I have tried to work with CNC shops over the years, but can almost never get a price from them, whether it is for carving, curved molds or whatever. Always a sea of variables to affect pricing, and I cannot move forward if I don't know costs. If I'm going to experiment, I'll do it in our shop and gain the craft/skill equity that such open-ended experimentation can provide. Our shop is flexible and fast and even fun. We build doors and work the way we want to - a way that is sustainable and productive and rewarding. We aren't getting rich, but we knew that long before we started.

We can think on our feet and truly enjoy our product and the efforts that go into it. Mutual respect, cooperation and pride are an integral part of our day. Somehow, I think a CNC might corrupt this nice arrangement. While there is nothing inherently wrong about such machines, they just are not needed in our shop - plenty of other neat equipment to lust over.

This is not true across the board, and as we see, the CNC can be a valuable addition to a shop - the core of its machining efforts, as described. I'm sure it is a fine product - that result still goes back to the maker and his or her intentions and integrity. However, I increasingly see people with a CNC that decide to jump on the door bandwagon that know nothing about doors (or even woodworking), their intricacies and details, yet deign to forge ahead with that big digital hammer. Today, the landscape is littered with their efforts in both closed shops and poorly performing products all over the US, though there are success stories, I'm sure.

From the original questioner:
As mentioned earlier I own a CNC router which is a central part of our door production and would never be without it, but we didn't learn on the router. Oddly enough we use it for our own in house production and do a very small amount of CNC work for anyone else, it always seems to be a loser. I also have the full complement of conventional machines and probably more shapers than one man should have. Nostalgia aside, I still can't determine if all CNCís are a passing fad or the way of the future. Perhaps the conclusion is itís for the new generation who never has learned to swing an arc on the bandsaw or setup a shaper and never will, but will be able to point and click their way into an industry once reserved for the hands on craftsman.

From contributor D:
While shopping for our second machine, I had the opportunity to tour many door shops across the country. It was a two year journey, and a very enlightening one. I can say for a fact I was not really impressed with most shops, but always came away with a few ideas. Most shops geared their product towards a catalog of available products, with the purely custom stuff done for a premium. More custom for us, I guess.

Pinecrest is one shop I was inspired to be like. They use both manual and CNC machines, and have the skilled staff to operate them. We value skills as well- one of our guys built a good deal of millwork in the king of Thailand's royal palace. Many of those kinds of shops have gone under in the past couple of years as well as those who couldn't afford their new CNC machine payments. Seems only the strong are left standing. But I will say one thing many will agree on- CNC machines are great fun as well.

From contributor O:
Yes, I think you are right, the future will definitely include CNC, and it can be the core of production or an ancillary part. The key is finding one's own path and making that work and not force a method on the product. And there will be craftsmen doing excellent work with such equipment. I do think it a danger that some know little or nothing about wood in general and doors in particular, yet they jump into door production in a big way, with tons of problems as a result. Doors don't lend themselves to that approach as cabinets or moldings will.

I will take issue with the word nostalgia when talking about the equipment or even our attitude here. Yeah, I'm a romantic, but somehow I don't care for that term, though I know you meant no disrespect. I am doing today just about exactly the kind of work and doing it the way I like after many years of work. I really love my job, my shop, my coworkers, my customers (well, most of them), and my products. Most of our customers are amazed beyond words. I look forward to spokeshaving a curve, the hum of the shaper with that particular set of knives, the very satisfying thunk of a mortise and tenon pulling tight, and wouldn't trade it for anything.

This is why I do the work we do, and I can't imagine doing anything else. It is passion pure and simple, for better or worse. Passion is what makes it happen for me, the products and entire circle. While that may sound nostalgic, it is not. Passion can also infect that young guy programming a CNC to do things not thought of yet. At least I hope so.

From the original questioner:
Never any offense meant, but I posed the original question because I don't personally like to be complacent either. I love every part of this business also, but it is still a business and as I get older it also does become about how we continue to keep our position in the market secure and make money.

I know my abilities and what I can do, it's the other 25 people working here that are always at question. Although I doubt I will hear it, if someone said a new quarter million dollar Busellato is going to improve your production and quality exponentially, I would feel foolish to not to consider it.

Although I live and work in a resort destination for the rich and famous, whether or not I get the job has never hinged on whether I cut that arch on a bandsaw or a CNC router but that the quality expectations are met and the price is right. Thatís just my world, your situation may be different.

From contributor O:
I donít own a CNC but looked real hard for three years at putting one in our two-three man door and window shop. My business plan has always been to produce as much quality work as possible with no more than four or five people working. It seems like a router could fit this situation.

In the end I figured to make it pay I would need more orders and more people upstream and downstream of the router to keep parts flowing. In addition I would need to duplicate the 60K of insert tooling I already have. Itís still a possibility for me but a lot of things would have to change with the economy and my shop.

In a 25 man shop your parameters for justifying the cost will be totally different than mine. You need to figure more than the cost of the router. In my case it was about 180K minimum for a door and window machine - about half that for tooling and about half again for software and learning curve, with six months to a year to get it working right.

To make this work you also need the right person to operate and program the machine. You already have a flat table machine so you should be ahead of the curve on this. I did the Homag router tour in Europe, and itís pretty eye opening to see what these can do when totally optimized with tooling and software. A router is just a tool. No different than a hand plane or a CNC tenoner. Used correctly any of these tools can give pleasure and profit.

From contributor D:
To the original questioner: you may have a sweet spot right now without using a CNC router. Honestly, if we hadn't gone down this road, we'd still be in good shape. Myself personally wouldn't have received so many job offers to jump ship, in places like Hawaii to tempt me further. Woodworking is all about making the customer believe you are the best, and serving it up on a silver platter. A few of those in the past couple of years tried to show off this talent with a shiny new machine, which got the sale but maybe not the product. I encourage you to keep doing what you know how to do. We do have a few new feathers in our cap now, nothing we wouldn't have eventually managed.

I just believe after using this new technology my thinking has changed, and I apply it all over the shop. Super accurate jigs are made on the CNC machine whenever someone needs one. Forty pushsticks can be made in a couple minutes. Playing locksmith with that new multipoint lock to be installed is nearly automatic. Your assemblers are so trusting they never dry-fit anything, and just go full blast. For multiple runs, you can kick back and look at actual machining times, to factor in with other labor.

When you get to the point where you know you can machine perfect parts day after day, you can start cutting corners and look at time savings.
Really machining the parts for doors is not the most labor intensive part, regardless of your path. But after you acquire an arsenal of different approaches, you can pick the best for each task more effectively.

New Busellatto? Maybe start looking at what software is intuitive for your type of machining, and tinker with a demo for a year. If you are still hungry, then dive in and embrace it. The first year I kept jumping off to manual machining out of fear. Gradually I have gone to nearly all CNC cutting. The mitered half laps that Steghers do for 90 degree joints? I now do that for radius TDL intersections. You have to believe in it to make it work.

From contributor I:
I agree with most of what has been said here. We decided to go CNC for our door production for a number of reasons, the main reasons being quality, shortage of skilled craftsmen, and flexibility. The CNC is the most flexible machine in the shop. It can do almost anything you just need to have an imagination. We had our 5 axis installed six months ago and are now starting to see huge time savings in a lot of processes. We could be machining parts for a mahogany entrance way one minute and making parts for a custom closet organizer or wall unit in the next. All parts made to an accuracy that is as perfect as you can get.

From contributor N:

We have a 5 axis CNC. We don't build doors in a high production environment like a lot of you guys do. We have, however over the years built a lot of doors. They are always one of a kind and the quantities are low. With the ability to make moldings and details in the panels without profile cutters the CNC does have a advantage. It does however take up a lot of time to machine that way. For the standard shaped panels using a shaper is by far the most efficient way to do the job. I guess a shop that has a setup with both the traditional methods and a CNC could go a long way to satisfy the clients needs.