Appliques

[paul_z]

I made some appliqués for Extreme Makeover: Home Edition. I thought telling the story of how they were made may help others to use the same technique, or hopefully improve the technique.

Nancy Hadley nancyhadley.com (http://www.nancyhadley.com) is a designer for the show and she was aware that I was doing some other work for the show. Nancy hand drew the following saying this was a simplified version of a theme she wanted to use in the nursery. It was sent it to me as a jpg directly from a digital camera. I don't know if she drew this free hand or if she traced it. In either case, Nancy is an expert in design and seemed to know what level of detail would work.


2533

I took the image into Photo shop and used brightness and contrast to blacken the lines and white out the background. I then used a Gaussian blur filter to get rid of some of the "jaggies" and to make the lines of consistent width, followed by another round of brightness and contrast redefine the boundaries of the lines one more time. I used Photoshop for this but one could use almost any picture editing software for this step.

Inkscape (free on the web) was used to trace the drawing and create a .eps file for V Carve. Not only does it do an excellent job of tracing, it also seems to further smooth the drawing.

I used V Carve to create series of previews of the appliqués from 0.125 to 1.0 inches thick. The following is 1.0 thick.


2534


2535

The previews were sent to Nancy with a message that the "frames" wouldn't be in the delivered item. She liked the 1.0 inch thick preview and the artwork was revised to include a wider margin between the "frame" (which is scrap), and the appliqué.

I glued up two 3/4 sheets to form a 1 1/2 thick block of MDF 20" x 31.5". (Use lots and lots of glue. The MDF just sucks it up.)

I told V Carve that I would be using a V bit and that I would be using an end mill to clear the pockets which were to be 1.050 deep. V Carve puts the V Bit “stuff” and the end mill “stuff” into to two different files to accommodate the bit change. I lied to V Carve; I had no intention of using the end mill but it prevented V Carve from trying to cut the flat portions with the V Bit. Then I started cutting … well … OK, I didn't do the cutting, the shopbot did all the cutting and I dropped by occasionally to check on the progress. Swinging a 1 and 1/4 inch V bit, the design horsepower of my router was soon exceeded and the feed rate had to be slowed from 3"/sec to 1"/sec.

With the V Carve complete, the frame and some pretty ugly waste chunks still remained. The work was turned upside down and the back was milled until the piece was 1.1 thick. This left just enough strength to handle the piece to get it over to my drum sander. The sander was set to 1.0 inch and a slow feed rate. The finished piece emerged from the sander with the scrap falling away. The sander may not have been necessary, but I was afraid to use the shopbot to cut the appliqué loose.

Since I had left and right pairs (two sets of two) I taped the pairs back to back to protect the edges. After I left, the production crew attached the appliqués to the doors and this is the final product.


2536

Two things of note:

1. Nancy said she didn't really expect the V Carve preview to be representative of what was going to be delivered. She was surprised when “the delivered parts were exactly what the preview had predicted”. I have seen similar comments on the forums. This is a tremendous benefit where the product has to be "sold" before it exists.

2. At no time did the artwork get "reworked" to get the final product. No hand manipulation of vectors or any such work. The final piece was the direct result of the original sketch. The entire process could be completely automated with no human in the loop. I'm not sure that I like that.

Overall, it was an interesting experience; I met a lot of good people and learned a great deal. I hope the process will be useful to others.

Paul Z

[gary_n]

Paul,

That looks great. What angle v-bit did you use?

[den73160]

I saw that episode, where the host was working with a trim router and then held up a preview of the finished work. It didn't look to me like it was done with a handheld router.

Looks great and glad to see someone helping out.

[paul_z]

It was a 90 degree bit. Afterward, I wondered if a 60 degree would have been better.

I think Ty Pennington intended the router scene as a joke. It would be nearly impossible to cut the appliques by hand. Ty is a highly skilled carpenter by the way.

[robtown]

The stuff turned out nice Paul. What about the other side of the door? You had some adventures there also did you not?... ;)

[paul_z]

As Rob stated, the other side of the doors are another story. Rob had a pattern that they wanted to cut into two doors.


2537

The pattern is a 3D cut and I was going to use Mill Wizzard to make the cuts. If I remember correctly, Rob had estimated that each door would take about 16 hours to cut. When the doors arrived on the building site, they were solid core, not solid wood doors, and it’s 5:30 on a Friday. The places that one could get replacement solid wood doors won’t be open until Monday morning and the doors have to be cut and back to the site Tuesday at noon. Several alternative plans fall through one by one.

By Sunday morning I realized that we are going to have to cut the doors we have on hand. If I use the design Rob gave me, most of what is going to show from the door will be the core of the door. At best it will be MDF or particle board, and at worst, it will be scrap wood with gaps and voids; however, there are no alternatives left.

After a few hours trying some patterns in V Carve, I sent some previews to Rob and the other designers.


2538

2539

The new design retains some of the 3D look but with minimal depth cuts. Rob and the designers liked the previews, but wanted to check with Ty. More time passes and finally Rob says go for it, there’s no time left.

I made a trial cut on a piece of MDF:


2540

And here are the doors:


2541

We were very fortunate; the core was particle board with no gaps or holes. Big Dave of the production crew did his magic and filled, sanded, stained, and varnished the doors.

Because we switched to a V Carve design, each door took about 2 hours to cut. I think the original design would have been gorgeous in solid wood, but this was a very nice alternative in solid core.

I will readily admit that I was gravely concerned about the reliability of the ShopBot as a system. So many things could go wrong. There were hundreds of thousands of vectors. Since the design was a simple geometric pattern, any flaw would be readily apparent and there wasn’t any time left to fix errors. In the four hours it took to cut the doors, there wasn’t a single error that I could detect; the cuts were perfect. I sent my thanks to Ted and I think there is still some sawdust on my lips from kissing the gantry. What an incredible machine!

[jhicks]

Great stories and excellent idea photos. Thanks