Computers transform the cutting edge of sculpture

[Merlion]

This is an introductory write-up based on an actual job. You can skip the details if you are already familiar.

Computers transform the cutting edge of sculpture (http://technology.guardian.co.uk/weekly/story/0,,1965387,00.html)

Techniques hundreds of years old are being updated and made faster and more accurate by laser scans

Sculpture is probably the oldest of human arts. It seems a human instinct to make shapes out of whatever material is handy. Yet computers are changing even this ancient art that produced Michelangelo's David, Trafalgar Square's Lord Nelson, and the Statue of Liberty.

Those sculptures all have something in common: they are big. .... How do sculptors do it? This is the question Manhattan sculptor Meredith Bergmann faced in creating an eight-foot bronze portrait of the first African-American opera singer, Marian Anderson.

The first step is to make a maquette, a clay model a few feet high - large enough for detail work, but small enough to see around. That took Bergmann two years of research. Once the maquette is finished, the technical fun starts. ...

..... The computerised method is done by several companies. Bergmann sent the Anderson maquette to Kreysler and Associates, a small company north of San Francisco. Kreysler painted it grey and put it in front of a laser scanner made by Cyberware, the company that laser-scanned actor Robert Patrick's movements for the liquid metal cyborg in the 1992 movie Terminator 2: Judgment Day.

The scanner traces a line from top to bottom, turns the sculpture a fraction of a millimetre, and does it again until it's scanned the entire thing. If the maquette is too big, or if some of the block detail is overhung from a particular angle, the sculpture is turned on its side and scanned again, or sawn into pieces to scan separately. Eventually, the computer holds a complete three- dimensional representation.

The software then draws a milling path, imaged as a network of fine lines. The cutting tool itself is a long, smooth metal lance-like object, much like a drill bit without the scrolling. A chunk of very dense urethane foam is placed on a turntable or flat bed, and a computer-driven machine uses the spinning tool to cut those lines back into the foam block, or blocks.

It was the latter with the Anderson sculpture, which came back to Bergmann as a jumble of puzzle pieces to put together. The artist and her budget determine how close the cutting distance will be: the closer the lines to each other, the more detail - and the more time, which costs more. .....

[John Evans]

Frankly this seems to me to be a very expensive and very dumb way to make an enlargement. I work with sculptor, Luben Boykov. He gives me a maquette that is marked with about 200 fine ink dots. My first job is to measure the exact X Y Z position of each dot. These are measured from a rigid frame built around the maquette. The points are measured with rulers and a depth gauge. Then he gives me a magnification number. Each X Y Z location has to be multipied by the magnification number.

The magnification starts out as a strong central steel core skeleton. This skeleton is placed on a ruled X Y grid. Each square equals 1 centemeter times the magnification number. A vertical laser is placed on the grid in the position of one of the points taken from the maquette. The Z location is found by using a ruler (each centemeter enlarged by the magnification number.) When the enlarged X Y Z position in space has been found a 1/4" steel rod is welded to the core skeleton. (Actually the end of point is placed about 2 cm in towards the center of the figure. This space is filled when the clay is applied.)

When all 200 points are placed the enlargement looks like a porcupine. Then using 1/8th inch steel rod the tips are connected to each other in a series of triangles.

Chicken wire is wired to the triangulated surface and on this chicken wire surface 2 cm thick slabs of clay are placed. The sculptor then sculpts the enlarged figure.

The whole enlarging process take 2 workers about 3 days to complete.

[BMBourgoyne]

Using digital technology in the enlargement process need not be expensive or complicated, and infact offers greater accuracy and detail with less time and labor than purely manual methods. In fact, when used in conjunction with manual techniques, digital scanning and modeling can make the process cheaper as well as faster.

A digital scan will measure over a million points on a model, not just 200-- in minutes, to an accuracy that can only be achieved manually with extreme care and skill and a lot more time. So when the computer triangulates those points to create its wireframe, it does so with a lot more points, and a lot more detail. That means, less detail for the artist to re-create in the enlargement.

That digital model can then be used to model the armature, undersized to accommodate the clay surfacing, revised and modified digitally however desired, and then fabricated a number of ways, from high detail (expensive) CNC routering in foam to lower detail (low cost) "sliced bread" methods where each cross-section is cut out manually from printed plans and glued up around a steel armature. But even this method, though utilizing low cost manual labor, is faster (because the computer does everything but the cutting and assembling) and more detailed (you get a high definition contour every cross-section) than a purely manual method, where you have to measure and re-measure every point. And its cheaper too, because the fabrication is faster and requires only a skilled craftsman for the scanning/modeling and during the final assembly. The rest is low skill cutting on an ordinary scroll saw, and gluing layers together on a jig.

When I enlarge a six foot figure from a 2 foot model, it takes me (one worker) about 5 days for the rough enlargement (plus a day or two for the final hand surfacing for more detail), from scanning the maquette, modeling the armature and cross-sections, printing the plans, welding the armature, cutting the cross-sections in styrofoam, gluing it up, and final measurement re-checks and adjustments. And that gives an enlargement made up of 3/4" layers (that is, a contour of continuous detail every 3/4" inches, which is a suprising amount of information). You can see more about my process at
http://www.bourgoynestudio.com/bourgoyneSTUDIO/BMBSculpt_enlargements.html

And for those who want to do the fabricating themselves, the model can be scanned and meshed, and the enlargement plans emailed to the customer for a "do-it-yourself" service.

If you're open to the possiblilites, and recognize where the advantages of these new technologies are (as well as their limitations), they can help make the process a heck of a lot easier. Its all about finding the sweet spot-- where there is more benefit than cost. If you'd like to know more, please let me know.

Brad M. Bourgoyne
www.BourgoyneSTUDIO.com

[John Evans]

OK lets compare the computerized method of enlarging and the method I describe below. Luben Boykov, the sculptor that I work with, recently recieved a commission to create a life sized figure of Amelia Earhart, the first woman to fly accross the Atlantic Ocean. Luben created a 16" maquette that was approved by the people who commisioned the work. When we got the go ahead Luben gave us the maquette with 200 points marked where he wanted measurements to be taken. Two workers, under my supervision, spent 20 hours (at $10/hour), the materials cost about $45. The total material and labour costs were $245. In 3 days a strong steel enlargement, covered with clay, was delivered to Luben's studio. He then proceeded to finish the sculpture.

Now lets look at the laser scan enlarging. From what I can gather from the articles below they can measure over a 1,000,000 points on the model. But when using the low cost "sliced bread" method these points are not scattered evenly over the model, but in lines 3/4 of an inch apart. According to Mr. Bourgoyne this takes between 6 and 7 man days to complete. I would be very interested to know what the total cost of these services. Also I would like to know the transportation costs of getting the maquette to the scanner and sending the enlargement back to the sculptor.

A maquette is a work of art in itself. Luben's maquette of Amelia Earhart is cast in bronze. He would not be happy to have it sliced and diced in order to get a scanned enlargement.

How about capital costs. All the equipment used to create our enlargement, The digital depth gauge, the laser level and a TIG welder set us back about $2,000. What is the capital cost of all the equipment required to create a low cost "sliced bread" enlargement,

According to the "Computers transform the cutting edge of sculpture" article digital scanning did not exactly provide Ms. Bergmann with an easily finished sculpture. I quote from a paragraph in the artice --"The further you go [in enlargement], the more undescribed areas and surfaces you find," says Bergmann. "When it's big, you want more details to the drapery, and you have to fill in the expression." Bergmann spent four months correcting and adding detail to the Anderson piece, 10 hours of it on the fingertips alone."

[BMBourgoyne]

You have to also consider the amount of work the chosen method saves the artist. If the artist is happy with 200 pts of data, and plans having to essentially resculpt the entire figure from that limited information (which is certainly an important start), then that is certainly the better method. You have saved him the trouble of making an accurate armature, and not wasted time and resources on work he plans on re-doing anyway. The sweet spot for his process.

But if the artist would like to save a week (or really more) both building the armature and roughing the forms out accurrately to a much higher level of detail than 200 points, but still plan on significant refinement of details smaller than 3/4", then the digital slice bread method makes sense. It accurately and reliably replicates the artist's own work in their stead, just as a highly trained studio assistant. If they have a good maquette, there is reason to be as faithfull to it as possible. The artist's time is the cost you are overlooking in your comparison.

[If they have a very good maquette, and really need to save their own time, CNC routering will give even better detail (at a higher cost, of course)]

Capital investment in the equipment is of course much higher for digital technology than just a welder, but it is used for a great deal more than as just a fancy pointing-up machine. It is multi-tasked, and used at every part of my studio work, just as all the equipment in a manual shop is multi-purposed. And most manual fabrication shops charge a great deal more than $10/hr for their services.

I have used the method you champion in my own work, as well as the digital sliced-bread, as well as CNC routering in foam. For myself, the time that is saved in the roughing out stage is well worth the costs of equipment. And the detail of 3/4" layers gives enormous information about the details (including drapery folds, facial structure, expression, and the like). I am able to spend more time working on the maquette (experimenting, refining, detailing), and know that that work will be easily reproduced in the enlargement. But my time still doesn't cost enough to make CNC routering viable for me. The digital sliced-bread method is the sweet spot for me, and I believe for many other artists.

Now John, you are clearly someone who values doing everything by hand, and are not as concerned about the cost in time. That is not an unreasonable position. A pair of smart hands and a few simple tools will always be the most versatile machine at anyone's disposal. But be fair about it, and take a real look at the cost and the benefit for the artist, especially in time because that is the bottom line.

Most artists who hire out enlargement work do so because they would rather spend their time on the creative end of the process-- the original development and the final execution-- leaving the less creative, laborious work to someone else. I find that the digital sliced bread-method does more of that more efficiently than the alternatives. Just my opinon.

Happy Holidays,
Brad

ps
3d scanning is entirely non-invasive/non-destructive. The maquette is not cut up in the process, it is the the digital model that is sectioned (or rather, sectional contour lines are taken from it). 3d scanning is exactly like taking a photograph of the maquette. You don't even have to put little dots on it. And not only is the maquette left untouched, but reproductions at other scales (smaller and larger) can be made with a 3d printer for about what it would cost to make a rubber mold and cast it.

Also, the sliced-bread method is only one way to make a manual enlargement from a digital model. Any kind of custom framework or cross-sectional matrix can be designed and printed as 1 to 1 plans.

pps
I typically charge about $2500 to $3000 for a full-service 6' figure at 3/4" layers, which is a little more than 10x the cost of your method. But I figure it gives at least 10x--no make that 100Xs-- more detail (about 200 cross-sections with at least 100 points each contour= about 20,000 points of data used). That includes the scanning, the digital model (which can be re-purposed at other scales), the steel armature, and the hand-finished and sealed surface (ready for 3/16" of clay). The scan with DIY plans are about $600, the armature about $400, the foam rough cut and laminated about $1000, and the hand finishing about $500. Shipping will generally add about $150-$200. These prices will vary depending on the nature of the pose, of course.

If you added the artist time to get the enlargement you provide to the same level of detail as this, I have a feeling it would come out to more. It does for me when I do it that way in my own work.

[John Evans]

I am sure that both methods have their merit. Both Brad and I deliver a service to sculptors and naturally prefer our own methods. I think it would be very interesting to hear from sculptors who have experience with these different methods.

I got the idea that digital scanning could not handle over hangs from the first article in this thread. It is copied below.

"If some of the block detail is overhung from a particular angle, the sculpture is turned on its side and scanned again, or sawn into pieces to scan separately".

I have visited several large foundries in the USA and was told that scanners were unable to scan under cut regions of a sculpture.

[BMBourgoyne]

If a scanner does not have a clear line of sight to a surface, then it cannot scan it (much like trying to mold an undercut). The simplest solution is to cut the model if that is a problem, and if that is an option. However, probe digitizers can typically reach in and "feel" the surface to get data as a secondary approach. And of course, good old calipers can do the job manually, in the clean-up stage of the foam carving.

No tool will do everything perfectly. Its a matter of finding the ones that get you closest to what you want most efficiently. Until a few years ago, digital tools were indeed way too expensive for most artists to even consider. But that is quickly changing, as other industries have developed enough of them that they are beginning to filter down to other applications (such as sculpture). Ceramic shell casting, and CNC cutting of steel are good examples of how once expensive technologies can become a valuable artist resource once they become common enough in other industries.

Brad

[John Evans]

If you really want to go digital check out <www.bathsheba.com> and <georgehart.com/rp/rp.html>

Bathsheba is a mathametician/sculptor who uses digital printing and rapid prototyping to create her magical sculpture. Her web site is the best and deepest I have ever seen.

Best of the season to all

[dilida]

I find this to be a very insightful conversation, thanks guys. I have carved and clayed-up some enlargements because of lack of time on the artists' part mainly. The place that does the scanning uses wax maquettes that they cut into pieces, that way they can scan those undercuts, and then the foam pieces are glued together. Since I'm the moldmaker also, many times, if the artist is agreeable, I won't glue together two sections that I would separate later in the molding process, thus once again saving time which translates into saving money at the foundry. When we are bidding the cost of a mold, We take into consideration the difficulty of handling a large piece, cutting supporting, ect. My first question to the artist is "What kind of armeture do you have inside?" The foam enlargements are usually the easiest to handle, I love them from that point of view. The smaller detail does have to be totally re-sculpted, but the time saved on the larger expanses that are reproduced perfectly in the enlargements is a trade-off I think. My main worry is making the big one have a life of it's own, and not just look like a stiff blow-up of the little one. So far the artists have been happy with my attempts, or at least say so and pay the bill, but that is the biggest struggle with enlarging as I see it. In fact, I'm going in to work today, a holiday, to try and get two enlargements done for their DEADLINES, a nasty, evil word. Ya'll have a Happy New Year! May it be full of time and ideas! Lisa

[Meredith]

Hi- This is an interesting discussion and I appreciate the introduction to Bathsheba’s and George Hart’s amazing mathematical work. John, thanks for contacting me.

I make figurative sculpture and have tried different enlarging methods including the time-honored grid and plumbline, the pointing machine, and paying someone else to rough out a clay for me. The enlargement from a 32” model to an 8 ft. statue of Marian Anderson was my first time using digital enlarging and I found it entirely worthwhile under the circumstances of time and money and method for that project. First of all, despite the essentially columnar shape of my statue I had taken some artistic risks with the twisting rhythm with which the figure is posed, and because I assembled the foam statue and painted it to match my clay, I was able to study it, crude as it was, and get an idea of the look and effect of the completed statue to make adjustments before I began working on it with clay. I was very happy to be able to work without any external armature to block my view or the need for any internal armature that would complicate disassembly for molding. I was also happy that the (detachable) head was so lightweight, as I had to clamber down and up a ladder with it in order to work on it and to set it back in place to look at it. The rigidity of the foam was helpful in the sculpted hands as well. I still had to alternate periods of work with periods of refrigeration but had much more working time without fear of warming and distorting the entire form, as I would have had if the hands were mainly clay. The overall lightness of the piece meant I could easily and often turn the dolly on which it stood, and in fact I felt much more at ease with the statue, despite its size, than I’ve felt with other large figures and I believe that made it a better, more confident sculpture. I was also able to work alone, and that was priceless for my concentration.

As for time, there is no comparing one artist’s process with another’s. I haven’t seen the Amelia Earhardt sculpture you mention. My figures are finished to a delicately textured smoothness, with intricate rhythms and long, spiraling curves in the drapery that require many hours of meticulous work. I worked on the enlargement with great intensity for four months and thought that was fast, for me. I made my deadline and am very proud of the finished product. Please take a look at the statue.

The article:

http://technology.guardian.co.uk/weekly/story/0,,1965387,00.html

contains some errors. The whole project (not the creation of the maquette!) lasted two years, including periods of waiting for feedback and approval, during which time I could do other work. I continued to research Miss Anderson’s appearance, life, music and impact on history throughout (she was an astonishing artist and led a inspiring life), and was changing details of her hairstyle and gown in the final weeks. The enlargement arrived in about seven sections- hardly a “jumble of pieces”. Anyway, I’m glad the article led to this discussion.

[Meredith]

So that you can, if you're curious, see the level of finish, here's a link to a page of photos of my statue of Marian Anderson:

http://web.mac.com/meredithbergmann/iWeb/Site/Marian%20Album.html

[BMBourgoyne]

Hi Meredith,
I'd like to compliment you on a remarkable sculpture.

For others interested in the process, if you have any photos of the foam enlargement in process, or before the clay was added, it might be nice to see them.

Also, as a side note, did you have only foam with no internal support?

I include a steel armature inside the foam, as for most figures, the foam alone would be too weak for the weight of the clay and rubber and rigid mother mold. However, I design the the steel to come apart into sections (for shipping, for re-adjustment, and for making sectional molds). One of the advantages of the digital part of the process is the ability to design the armature within the sculpture, and have the holes cut for it in the foam at the same time that the cross-sections are cut out. The foam then goes around the armature with very accurate registration (but still enough play for re-alignment).

Also, because the forms are built from cross-sections cut out individually, there is very little wasted material, and the styrofoam used is low cost, mass produced construction styrofoam sheets rather than expensive foam blocks. When you carve from a block, you have to cut away and waste all the material around the desired form. In contrast, if you lay out all cross-sections on a sheet arranged as closely together as possible, very little foam is wasted (small cross-sections fitting in the spaces between the large). So not only are you using a less expensive foam, you use it more efficiently.

These advantages hold true because the enlargement is only looking to reproduce proportions and major detail (3/4" typically). If you go for more detail (thinner cross-sections) the increase in work makes it more expensive. However, you can have selective areas (faces and hands for example) 3d printed, or CNC routered to higher detail and attached to the lower detail figure. The process can be easily tailored to the artists needs for each particular process.

[Baraka]

I would like to thank each of you for taking the time to explain these methods, and for showing us your amazing sculptures. This discussion has been very ensightful for me.

Delida (Lisa), I suspect that you work for the same foundry that will be casting my horse sculpture soon. This will be enlarged at Crucible a little later in the spring, and be my first experience with this method. The maquette will be brought in to be cast in it's original size, in a week or two. I'll look for you and introduce myself.

For artist like myself, this is the only option. I am not so young anymore, and a little restricted in what I can handle. Frankly, the scanning sounded like a dream. I would also assume that most, if not all, large pieces are commissioned, so the difference in cost is covered by the client.

I live 5 hours away from the foundry, so I will use the studio they provide their artist to do the details on the large maquette. I'm told it should only take a few days, at most. My time is extremely valuable, and I cannot imagine not taking advantage of this service, for myself that is.

Again, thanks to all of you. It is very helpful to hear these kinds of details, and gain more understanding of the process, both from the artists and foundries point of view.