Rocking Chair Design and Construction

Creating a good rocker is more art than science. April 30, 2006

I'm interested in hearing some theories on getting chair rockers to rock nicely and look good, with different styles/designs.

Forum Responses
(Furniture Making Forum)
From contributor D:
Good rockers are like wagon wheels. What works and what doesn't has been derived from years and years of people building rockers, keeping the good ones, and burning the bad ones. My rockers are based on a pattern I got by measuring up an old chair that looked good and was comfortable. I use a story stick to build them, so I don't even know the dimensions off the top of my head. The rules of thumb I go by are basically that the chair looks better if it is slightly narrower at the top than the bottom, the rockers should be one inch closer to the seat in the rear, and don't skimp on backslats - one too many looks better than one too few. Old chairs are the best teacher.

From contributor B:
I've got a question regarding the rockers on a rocking chair. I have two new rocking chairs that don't rock too well. Some folks say the radius is not that important. Others do. I'm thinking about making some temporary add-on curved parts to find an acceptable rocking motion.

From the original questioner:
I was just thinking today about something like you're saying, except I was thinking "shoes." Removable rubber (or similar) covers to help with walking and possible floor damage. Maybe these are available - I don't know. Anyway, if the shoe were firm enough or thick enough, I believe it might also work as a corrector? Because as you probably know, just a little bit can change so much. But who knows - something like this might also be disastrous. My conclusion is that there are far too many variables to try and come up with any sort of standard for rockers. Anything goes.

From contributor K:
Seems I read something that the apex of the arc should be positioned somewhere like the nose or chin of the average person. Have you found anything like this?

From contributor T:
Back to seat angle, back height, seat depth, arc of runner... all are variables and while it would be great if there were some magic geometric rules to encompass those, there would still be some consideration of the persons who will occupy the rocker. I have had to alter the runners on several rockers, either to correct a really bad factory job, or to better suit the stature of the occupant. The idea to clamp on test curves is what works for me. By the way, "radius" would not describe the variable curve need for good rocking.

From the original questioner:
I've got this Neanderthal method of intersecting two or three different size arcs to get a nice rock. Just a section of a circle seems to usually rock too fast to suit me.

From contributor W:
I use a 42" radius for my rockers. Only about 8" of the rocker is actually used for rocking. Once you go beyond that, it becomes more style than function. One fellow did say he used a catenary curve, but when I looked at where the rocker contacted the floor, it was really close to a 42" radius. A shorter radius is found to rock too fast and a flatter curve (longer radius) too slow. With a shorter radius the individual's head moves through a greater distance and longer radius gives a sense of less motion.

I use about 5 degrees as the front to back angle on each side. However, the most critical thing is to make all of the dimensions exact side to side. An error of 1/16th inch can cause the chair to "walk" sideways.

As for making the seat lower in the back, I have seen everything from flat to 1+ inches. I use a flat chair (front and back legs the same length) but the center of gravity of the chair tilts it back about 10 degrees and a person sitting in it rocks it back another 10 degrees. This is with a 5 degree back angle relative to the seat. The customer can change the back angle to 10 degrees if they wish, by a simple adjustment.

From contributor L:
I'm interested in making an armless rocker. I like to use flat reed for both the seat and back. Is there any reason why the back post should be bent? Can one just have a straight back?

From contributor W:
I make an armless rocker for which I weave the back and seat in cotton cord. I use a dinning chair for the rocker. I use a 5 degree angle and some people like a 10 degree angle. The Shaker rocker designs all used a straight (90 degrees to the seat) back chair. They needed to tilt the chair about 3/4 inch back on the skids to get the center of gravity in a comfortable position.

From contributor L:
Thank you for your quick response to my question. You indicated you make a dinning chair. How do you rack a chair? How do you align all four legs to touch the floor? One always seems to be a little higher than the others, yet the measurements are right on. I have done a few of the Number 5 Tape Back Chairs and seem to always encounter this problem.

From contributor W:
Ah, yes. For chairs (versus benches), I glue up the back after carefully dry fitting to assure that the back is not racked. Once it is dry I glue up the front legs to the back and while still wet, I rack to make the legs all level. Since I use all round dowels for seat rails and stretchers, they can slip slightly until the glue sets. You have about 5 minutes to complete this task. I also clamp for about an hour or so.

The biggest hurdle is making all the rails and stretchers precisely the right lengths, all the mortises drilled the same depths, and the mortises/tenons on the back rails all in the same plane. Dry fitting the back is absolutely a necessary assembly step. Since it has four cross pieces, it becomes the basis for the rest of the chair being level.

Every Windsor chair maker I know comments that a chair after glue up never sits level. They all level the seat, finding the three shortest legs, hanging the longest off the flat assembly table, marking and cutting it.

Except for rockers, my designs are intended to flex a bit to conform to the typically non-flat floors of homes. The cotton cord seats 1" dowels that typically are longer than 18" can flex slightly as one sits on it, making it stable. Rockers need to be reasonably stiff so as not to squirm under the person as they rock.

From contributor L:
Thanks for the response. Right now, I'm doing a beech European (wood) chair. It's light colored and I think it looks good. You can see the grain in the wood real easily. What kind of finish would you put on it? Keep in mind the seat and back have a flat reed that has been installed. The last chair I finished I used Waterlox and I just don't like the tacky feeling it gets after it has been applied. Do you like linseed oil and paint thinner? If so, do you mix it one part linseed to two parts paint thinner?

From contributor W:
In the past I used hot linseed oil and turpentine (I have heard all the jokes). Starting with 2/3 turp and 1/3 oil, then 50/50, then 1/3 turp and 2/3 oil. Tedious. I went to 50/50 tung oil and turp. Better, but still... Both of these did not work well in very humid conditions. Tung was better and did handle day to day abuse. Now I use Danish Oil 90% of the time. Watco Natural for all but walnut, when I use Dark Walnut for the first two coats and natural for the third. The walnut I have been getting lately has been a little streaky/blotchy, so the Dark helps even out a bit, as that is what customers expect for walnut.

From contributor L:
After using linseed oil and turpentine, did you follow up with a paste wax? What is the difference between paint thinner, turpentine and mineral spirits? They are all used as a dryer with linseed and tung oil. Which of the three do you like best as a finish on chairs?

From contributor W:
I do not have the chemical knowledge for a detailed discussion of paint thinner, turpentine and mineral spirits. However, paint thinner and mineral spirits appear to evaporate faster than turpentine.

I only use boiled linseed oil (contains dryers) and turpentine to thin the oil for better penetration. My parents, in refinishing antiques, used to heat (on a hot plate) the oil/turp blends to get further penetration. I would think that is not something you would want to do with PT or MS. Also, I can tolerate turpentine more and longer than either PT or MS.

After all my oil finishes, I use a hard paste wax, specifically Lundmark's Clear Paste Wax, which is 100% carnuba wax that uses turpentine as the softener. It seems to last longer on a piece, especially where hands rub like arms of chairs. Also on the rails of my sliding seat benches, the wax helps the UHMW plastic abrade less.

From contributor L:
Have you ever found yourself ready to stain or finish a chair when all at once you see you have a dried glue spot? When glue dries and is fairly hard, what do you do to remove it? Start sanding again or what?

From contributor W:
Luckily, I try very hard to keep that glue blob from forming, but... On walnut, I have found that a combination of black and brown Sharpies can be used to darken the spot. For some reason, red oak and white oak usually do not create too much contrast, maple does leave a light blotch. Cherry responds well to using a stain pen from MinWax. I haven't tried their pens for walnut. As for removing the blob, I do not sand, as the usual location is in the corner between the stretchers and leg and you will be cross grain or you cannot really get into the corner. I use a thin pointed detail carving knife to carefully cut the glue out, then use it to scrape the area as best as possible.