Home Articles Forums Gallery Marketplace Videos

WoodCentral Forums

Knowledge You Can Build On — Est. 1998

Abingdon, VA – Tree Cutting & Portable Sawmill

Tracing a moulding geometrically (Roubo)

Posts

Tracing a moulding geometrically (Roubo)

#1

Tracing a moulding geometrically (Roubo)

Bjenk Ellefsen

>So I dusted off my geometry skills and got to work on a Cyma molding for the cupboard I am building. The molding will be done with my Mosley and Sons hollow and rounds.

In the first sections of Volume 1, Roubo gives a crash course in Geometry after saying that woodworkers don't know much about it and that it is important to know the basics. This is my first attempt following his instructions on a molding which has as much height as width (square). The next step is figuring out his instructions on tracing the Cyma (doucine) when the molding has more height.

Roubo says that we could do it by eye but it is best to know how to do them geometrically. His "mentor" and greatest influence was a geometrician. Roubo says that even though we could trace profiles by eye, "it is important to learn to do them geometrically so that we can get accustomed to bring order and truth in the composition of profiles (moldings are like the letters we use in writing: their combination forms an infinity of words and diversity of languages) so that we do not expose ourselves to the faults of Gothic architects who are moving away from the perfect and regular forms the Greeks and Romans gave to their profiles and moldings. They (goths architects) have created new profiles of bad taste that have reigned in France for too long". His words not mine!

Ironically, I believe Chippendale included many Gothic influences in his Director and you can find in the presentation page the line: "In the most fashionable taste".

So on with it (Cyma recta and reversa):

1) Trace a diagonal ie from point i to e

2) Find middle of ie and place point f

3) trace a perpendicular gh passing through point f

4) use points g and h as a swing point to trace the curves with a compas.

The reverse concave and convex curves:

extend a line from point e to a new point down to create a square with point f and use the new point as a swing base. Vice versa for point i.


img

Re: Tracing a moulding geometrically (Roubo)

#2

Re: Tracing a moulding geometrically (Roubo)

Chuck Nickerson

>Do you have a translated Roubo, or are you doing the translation?

Re: Tracing a moulding geometrically (Roubo)

#3

Re: Tracing a moulding geometrically (Roubo)

Bjenk Ellefsen

>I'm doing the translation myself.

Also, while we are on the subject, you might want to go check Chris' blog at Woodworking magazine. There is some bit of News about Jack Metcalfe, a marqueteur and Beatrix, a retired french teacher who will be doing the translation of the volumes.

Re: Tracing a moulding geometrically (Roubo)

#4

Re: Tracing a moulding geometrically (Roubo)

George Makowski

>Bjenk, I used this method in making the details of the feet on some single board type chests. I had it from a Roy Underhill presentation. It worked so well that I still admire the lines on those simple chests.

Good proportioning!

George in AL

Re: Tracing a moulding geometrically (Roubo)

#5

Thank you

Bill Tindall, E.TN

>It is the willingness of participants at WoodCentral to share their considerable efforts with others that makes this Forum valuable and enjoyable. I can appreciate the work to prepare this post and share it. I have made a copy to add to my files when I need to make one of these moldings. I don't draw well, so geometry is a comforting crutch to help me design.

Re: Tracing a moulding geometrically (Roubo)

#6

Re: Thank you

Bjenk Ellefsen

>My pleasure Bill! This is just too much wicked fun not to share it!

I'll update the post with the making of the actual molding!

Re: Tracing a moulding geometrically (Roubo)

#7

Re: Tracing a moulding geometrically (Roubo)

Dean J in MN

>That is great stuff!

Moxon pages 268-270 describes how to lay out ogees, cavettos, etc in much the same manner for those who have Moxon. Until we get the exciting work you (and others) are working on Moxon is a good reference. Chippendale's plates on moldings are of use as well, the have module info, and armed with Moxon (or Roubo) one should be able to lay out any molding they desire.

Thanks for posting!

-Dean

Re: Tracing a moulding geometrically (Roubo)

#8

Re: Tracing a moulding geometrically (Roubo)

Bjenk Ellefsen

>Hey Dean,

Good thinking about Moxon. Let's get some moldings done!

Re: Tracing a moulding geometrically (Roubo)

#9

It's my understanding that Greek and Roman shapes

Bill Houghton, Sebastopol, CA

>differ, with Roman molding tending toward the circular and Greek toward the elliptical. Does Roubo discuss how to do elliptical layout?

Thanks for this, Bjenk; your discussion of how to make shapely shapes is great stuff.

Re: Tracing a moulding geometrically (Roubo)

#10

Palladio does...

Stephen Shepherd in the State of Deseret

>in his first book on pages 31 to 33 with a leaf in between with illustrations on both sides. This is a bit early, 1570.

You can do the same thing with major and minor axis on an ellipse. Or you can use an oval engine.

Hogarth also mentions this I believe.

Stephen

Re: Tracing a moulding geometrically (Roubo)

#11

Re: It's my understanding that Greek and Roman sha

Bjenk Ellefsen

>I will have to check it out in more details. I am no expert in antiquity architecture but I remember in reading Kimball's book History of architecture that there is no sharp distinction between Greek and Roman architecture. Italy did fall under the very strong influence of Greece and that from the beginning of Greek Civilization. I was always told that Romans who came to dominate were not too big on new aesthetics and derived their culture first from Spartan ascetic principles and then from Greece and the Orient.

Supposeddly, from what I remember, in conquering Greece they came to accept and admire the Greeks' advancement in architecture. So their relation to Greek forms was very strong.

Sorry, long post. I'm no expert in this topic so does it make any sense?

Re: Tracing a moulding geometrically (Roubo)

#12

Re: It's my understanding that Greek and Roman sha

Peter Hyde

>Here's a link with info about mouldings and layouts.

http://www.traditional-building.com/article/moldings.htm

Re: Tracing a moulding geometrically (Roubo)

#13

Re: Palladio's the man

Robin Corell in Atascadero CA

>In one of his books he even has an illustration similar to the one refered to by the origianl post in this thread. The difference was that the two radii are not exactly the same. to me this gives a much more elligant and less machine made looking moulding. With all due respect to Roubou Palladio is defianately the expert in the area of layout geometry.

Re: Tracing a moulding geometrically (Roubo)

#14

Pallaidio

Dean J in MN

>We can go back even further to Vitruvius for more interesting layout tips. For example, layout of flutes:

Each column should have twenty-four flutes, channelled out in such a was that if a carpenter's square be placed in the hollow of each flute and turned, the arm will touch the corners of the fillets on the right and left, and the tip of the square may keep touching some point in the concave surface as it moves through it.



...the couture of the channels (flutes) may be determined thus: draw a square with sides equal in length to the breadth of the of the fluting, and centre a pair of compasses in the middle of the square. The describe a circle with a circumference touching the angles of the square, and let the channellings have the conture of the segment formed by the circumference and the side of the square.

He discusses the moldings on the base, volutes, &c. Palladio drew on Vitruvius -- both are worth a read. Vitruvius is an fun read really. It is amazing what they knew -- surround sound in theaters using vases and niches! Great stuff.

-Dean

Re: Tracing a moulding geometrically (Roubo)

#15

Update: Tracing a moulding (Chippendale)

Bjenk Ellefsen

>My earlier post about tracing geometrically a Cyma is only one example amongst many more. Roubo shows how to trace ellipses and such, how to trace a shorter convex curves and longer concave curves inside a geometrically correct form.

However, Chippendale is also going over moldings and their tracing minus the explanations and details Roubo is known for. So I traced the actual molding I am going to use with Chippendale as a reference. My molding is not square but I still used the basics taught by Roubo.

Of note: Don McConnell did a fantastic DVD using Chippendale as a reference on how to trace and push the molding with Hollow and rounds. There is nothing like seeing someone actually do it to get started.

So here goes:

1) Trace a diagonal ie

2) find the middle point of ie and place point f there.

3) use point e as a base for the compas and place the marker side on point f. Now swing a line down from point e.

4) Use point f as a base for the compas and swing a line down from e

5) Use point i as a swing point (without changing the angle we have just given to the compas which is the middle distance of ie) and trace a line up from point i

6) Use point f as a base and swing a line up from i.

This will have done two Xs that we use a swing point for the curves of the cyma. The diagonal from the two swing points are equal to the diagonal ie.

This is what you will find in Chippendale's director without the details I have given which were taught to me by Roubo. (credits to Don for showing how to do it in a video of course!)

The two lines on the upper and lower sides are a 1/8 inch limit to define the molding and step. (Thanks Don!)


img

Re: Tracing a moulding geometrically (Roubo)

#16

Re: 1curve more pronounced than the other? Simple.

Bjenk Ellefsen

>Now, if you would want to trace a more pronounced concave curve and a shorter convex curve on this profile you would have to do this first:

Divide ie into nine equal parts. Give 5 parts to the concave and four to the convex curve and place point f accordingly.

Now, just do the same as I explained previously.

Re: Tracing a moulding geometrically (Roubo)

#17

Re: 1curve more pronounced than the other? Simple.

Pam Niedermayer, Austin, TX

>Bjenk, I don't understand what makes this such a difficult process that requires such elaborate solutions. I think I'm wrong, but I still don't understand. If you want a particular curve, you draw it and plane/carve it.

Pam

Re: Tracing a moulding geometrically (Roubo)

#18

Re: 1curve more pronounced than the other? Simple.

Bjenk Ellefsen

>Drawing by eye? Such barbarism! ;)

I am actually trying to show how simple it is to draw them in respect of the forms. From reading Chippendale and Roubo, moldings were important and the architectural forms were not arbitrary. Hence why one would learn how to trace them properly.

Roubo did say that you could do them by eye but that would be subjecting forms to the arbitrary. In 1768, the movement of the enlightenment was already strong in France and its no surprise to see Roubo's preoccupation with geometry and form. You can trace them that way if it suits you. I chose to learn how do it and I am showing it to anyone who is interested.

Food for the brain and for the shop if you care enough.

Re: Tracing a moulding geometrically (Roubo)

#19

The argument could be made for either approach

Bill Houghton, Sebastopol, CA

>of course. It depends on what design philosophy/effect one is striving for.

Re: Tracing a moulding geometrically (Roubo)

#20

Incomplete thought there...

Bill Houghton, Sebastopol, CA

>The value of these discussions, for me, is the whole issue of design. There's a great deal of attention paid in the magazines (and here sometimes) to technique, but getting the right shape in the right proportions is the deeper challenge. Most all of us can learn to cut, plane, and shape wood. Having it come out right's the hard part. Bjenk has offered his thoughts on one way to refine a molding shape; Pam has offered another. Geometric development of a shape will tend toward regularity; freehanding will tend toward naturalistic irregularity. Each has its place. Ruhlmann, Federal furniture, and any number of other furniture styles would look "off" if not symmetrical and geometrically clean. Art Nouveau would look pretty silly if dead regular.

Re: Tracing a moulding geometrically (Roubo)

#21

Re: Incomplete thought there...

Pam Niedermayer, Austin, TX

>I really didn't intend to "offer another way," certainly didn't intend to trivialize any process, more to gain an understanding as to why such methodology was useful and/or required. I think I do tend to ascribe more design/artistic capabilities to people than they often feel comfortable with. It's that lack of comfort, or stiffening up, that I think interferes with our (as in all people) natural abilities, very similar to contour drawing (as in "Drawing on the Right Side of the Brain" fame). Or, if you look at the shape you want to draw and draw it on the wood without looking at it, you stand a pretty good chance of making an accurate drawing.

I ordered Don's dvd a few minutes ago, will let you know how...

Pam

Re: Tracing a moulding geometrically (Roubo)

#22

Re: Incomplete thought there...

Dean J in MN

>Pam --

I try to use the right side of my brain when I can. Dovetails for example, I don't mess around with dividers, just draw them were they look good. I agree with you, this is something we all can strive to do. Free yourself.

With moldings however I'm not good enough, nor have the years of experience to just draw the profile. The benefit of the geometric method is one can look to the column orders to guide you to the size/proportions. If I need a base molding I can choose a column order, look at Chippendale and know how tall, deep, &c to make it. From there you can use the methods to create a pattern or draw directly on the stock.

That said, I'd love to build the design vocabulary to be able to just draw the profile on the stock and get to work. But I'm still in grammar school on that front.

-Dean

Re: Tracing a moulding geometrically (Roubo)

#23

Re: Incomplete thought there...

Pam Niedermayer, Austin, TX

>Yes, no disagreement with what you said. I opine that it's not necessary to make one of everything that's previously existed before going on with one's own stuff. :) I do try to approximate the golden ratio and approximate numbers of twists and turns that appear appealing to most (the magic number is 20).

Pam

Re: Tracing a moulding geometrically (Roubo)

#24

Update 2: done with the design

Bjenk Ellefsen

>I am not good at designing so in following the methodology and looking at Mike Dunbar's molding for the cupboard I am making, this is what I arrived at.

It's not a reproduction but it's close enough for this piece, I guess.

I will push it in wood this week and will report back.


img

Re: Tracing a moulding geometrically (Roubo)

#25

Nice looking pattern

Bill Houghton, Sebastopol, CA

>The little I've done with handmade molding has really been fun, and a lot more peaceful than using the whirling screamy thing (though I will confess to having used that plenty of times too).