I am building a Queen Anne chair from the Lynch drawigs of one from Williamsburg. I am having a terrible time fitting the crest rail to the tops of the sides. You can’t make the shoulders of the tenons perpendicular to the front or rear surfaces since the pieces are curved, and had to be cut that way, nested in each other to consrve expensive stock. I find the joints have a slight gap since the shoulders are slightly angled too much either from front to back or side to side. I tried Clarkston’s method form his book on Chippendale chairs where he uses spring clamps to hold a steel rule across both side rails and the splat at the same time, then scribes a line, but this didn’t seem to get me any more accurate results. I have spent about 5 hours planing a little here or there and feel like I’m making a mess of some beautiful curly air dried walnut. PLEASE ADVISE ME!! Thanks a lot.
Jay
Replies
Jay,
Sorry to hear about your luck. Now you know why it is nice to have a flat surface when dealing with chair joinery. Any way, you need to establish a reference line to pare to. Clamping a straight edge is one method. If you could plane a small flat on the back of the post this would help you to use a square to scribe a line around the post.
Or maybe lay the posts on a flat surface like your bench and put a square on the side to scribe a line on the side of one post and then square across both posts.
I find it very interesting that the plans you speak of have angled tenons for this joint. All of the period chairs I have seen, both English and American, have square shoulders and a short flat section at the top of the post for the ease of fitting the joint. Even if you get the crest rail fit to the posts you still need to fit the splat. Hopefully this will go a little better.
One last thing, use a sharp marking knife to layout the shoulders. Pare to this line with the last cut registering in the scribe mark.
Let me know how it goes and if you need more clarification on any of my descriptions.
J.P.
I thought the sraightedge line would be my reference line, which I scribed with a marking knife. I then tried to scribe the line all the way around the top of the posts on both sides, but unfortunately there was no way to know whether the angle of the shoulders was exactly the same on the sides. When the joint closes in front it gaps in the back and vice versa. Even with one side flat, that might not be exactly perpendicular to the axis of the tenon. I then tried to use the cheeks of the tenon as a reference, and with a small machinist's square, pared the shoulders to be square to the cheek, registering the chisel in the scribe cut. There were still errors. I'm still hoping someone will have a more standard method to avoid this problem, and tell me how to start with square surfaces, without wasting a huge amount of wood, for next time. (I didn't even have a big enough piece to separate the parts without getting into sapwood - which as you know is lighter in color in walnut.)
For this time, how exactly woudl you fit the splat? Do you glue the splat into the crest rail and shoe and put that on as one assembly?
Jay
Jay,You could make a block of wood to clamp to the post that would be square with the tenon and allow you to pare accurately. I attached a small sketch of what I am talking about. Also, if you make the top of the block square it will allow you to use your square to layout around the back and front of the post and connect the lines from front to back on the inside. When doing this clamp both posts together and work as a single piece.As for the crest rail. Having a shoe will help simplify the process. Don't glue anything until the whole chair back is dry fit. Put the back seat rail in and clamp it between the posts. Then put the splat in the shoe and the crest rail on top. Now put the crest rail on the posts and see if there is any mis alignment with the shoe. You may have gaps at the shoulders of the crest rail. Plane the shoe to close these gaps. Since you have been paring the post shoulders this is the most likely scenario that you will encounter.However, if there is a gap where the shoe meets the seat rail then you will need to pare the tenon shoulders on the post tops. Once you think you are satisfied with the fit, then clamp the shoe to the back rail and put the whole seat back together to see if it will work. If it's all good then go ahead and glue it up.J.P.http://www.jpkfinefurniture.comEdited 10/6/2005 11:50 am ET by JP
Edited 10/7/2005 10:46 am ET by JP
Me too needs help on this.
Jay, last time I had to fit odd shapes in timber I got it as close as I could, placed some pencil-carbon paper in the shoulders as I assembled, pulled the papaer out, opened the joint, and the high spots were marked.
In much rougher work, I have deliberately cut the shoulders just on the waste side of my line, assembed the work, and used a fine saw to create an identical width kerf in each of the joints. A kerf has (or should have) parallell sides, so the joint drops together. Never tried this on furniture.
So.. look forward to the responses to this question
David
The following jig can be used to scribe a line around with a knife and then serve as a platform for using a chisel to par the tenon shoulder
Looks like a great idea. My question is this: How can you know if your platform (which will create the plane of the shoulders), is square to the tenon cheeks? If it isn't then there will still be a gap in the joint, either on the front side or back. The joint will only close all around if the tenon is perpendicular to the shoulders.
Also, what is your technique and sequence for fitting the top of the splat to the crest rail?
Jay
When you created the side rails you should have created a few reference faces as well. One of these reference faces should be a 90 degree flat created at the top of each side rail to allow the jig to square. I don't know where you got the plans or procedure you are using but if I remember the book correctly and you are following anything from the Ron Clarkson book you have problems. Many times his procedures seem to leave you hanging with no real reference area to work from.
I believe I have a few more pictures and I will look for them and post them later. This will probably show you the basic steps better then any description alone can. There are many steps which need to be to preformed before getting to the point that we are discussing and if these aren't done in a logical sequence with good reference areas you will have problems.
<!----><!----> <!---->
Where did you develop the plans from?
I purchased the plans from a company that sells drawings done by Carlon (?) Lynch, done from an original at Williamsburg. The plan included a measured drawing and I made a large size xerox copy at Kinko's, cut out the profile of the posts, legs, splatt etc and rubber cemented them onto 1/8 inch ply, which I bandsawed to create patterns, for use in laying out and tracing the patterns onto the stock. (and for use if I ever want to make the piece again) The plans don't explicitly say anything about sequence of work or anything about flat reference surfaces, although I know from an article in FWW a few years ago on Chippendale chairs, that there must be a flat on the front of the posts to join to the seat rails. It also wasn't clear as to whether there is any "wind" in the seat rails. (ie, if the back posts are slightly inclined, compared to the vertical face where the side rails join the front legs.)
Jay
I have a book with some of Carlyle Lynch drawings and they are pretty good. <!----><!----><!---->
Cutting out the patterns the way you did it is a good way to go. The Queen Ann legs should not be canted inward which is one of the marks of a Chippendale chair but obviously back then there were exceptions as new styles were developed on top of old ones. Not having the legs canted makes the tenon a little easier to fit to the back rail as I'm sure you have discovered. The back rails on a Queen Ann have an s curve while the typical Chippendale had a simpler curve but the major issue here is getting the back splat to follow the curve. A few more little jigs can help here. <!----><!---->
I won't be able to get to the picture until sometime tonight and maybe I won't be able to post them until tomorrow but regardless you are already trying to fit what you have and it's probably too late to make any real changes now. The general rule is to try to plans your procedures so that you have a reference surface that you can index from with either a power or hand tool and use it before that reference surface disappears in some later operation. Cutting the back rails first and establishing flats where you can use them later on will save a lot of effort. I think the pictures will be able to help but it should be noted that the way I learned to construct a Queen Ann chair is one way and there likely more and even better ways to do it. <!----><!---->
I did have a woodworking instructor who used to say if you see a lot of a particular item designed in an antique then the makers must have found an easy way to do it because they had to be quick in order to make a living. Today we tend to concentrate on jigs for our power tools but jigs can also be a big help in the hand tool world and the old masters were probably very good at it.
What part of the country are you located in?<!----><!---->
I am in Atlanta. When you look at the Lynch drawing, it shows the distance between the back legs (at the turned spindle rail near the bottom) as 12 inches. BUT, the distance between the back legs at the rear seat rail is 12 3/8 inches. That would imply a slight outward splay from bottom to top. Hence my question about whether the legs should be parallel or canted. Visually they seem parallel but the dimensions create a slight angle.
The way I lay out the rear tenon is, as I understand it, the way the old timers used to, and is simple. That is, you do the front tenons first, put the side seat rails into the front legs, and hold the whole assembly up to the back. Mark the "wind" on the end grain of the side rails by tracing along side the rear seat post, then plane the side rail to the line. Use this fresh face to register a mortise guage to mark the tenon (which will be parallel to the outside edge of the rear leg and not plumb). Lay out the shoulder and the tenon on the top surface of the rail and cut with a back saw.
Jay
Such a small difference between the two its probably a mistake by the original maker carried forward in the drawings. <!----><!----><!---->
Because my legs are not canted the mortises are cut and only the side rail tenon is angled as per drawing Pic1. In the attached pictures you may notice a plug as been inserted in the through mortise to protect it while working on the rest of the back rail. Pic2 shows the marking of the rail top in preparation for butting the tenon. Pic3 shows the crest rail black sitting on the back rails. Part of the tenon is angled to insure clearance for the shaping of the crest rail to follow. Between Pic4 and Pic5 I should have a picture showing the rough shaping of the crest rail prior to glue up ears are left on to help the clamping and gluing . Pic6 is the chair roughed out ready for shaping, transition block etc.<!----><!---->
I don't know if this can be of any help as there are many ways to build a chair but this procedure that I learned seems to make sense to me. <!----><!---->
Ron,
Very cool, I just dont think well enough in jigs.
Still like to know how the old fellas did it though
David
Still like to know how the old fellas did it thoughProbably used some kind of a jig if they did it all the time..
This forum post is now archived. Commenting has been disabled