View Full Version : 3d printer as semester project
01-26-2006, 09:27 AM
This next semester at my high school I was planning on making something with my BS2. I did not want to make a generic robot so much as something usefull... And that is how I came about wanting to make a 3d printer. It would not use lasers or any extremely temperature oriented parts, instead, I planned on using wax and dental cement (the stuff they use to glue braces to your teeth) The dental cement can be dried fast with a UV light. and the wax can be melted away later, leaving just the hardened cement behind.
I thought the best idea for the mobility of the "head" of the printer would be to make a claw-like assembly (you know, the claws in stores like target, which take all your money?) except that instead of a string or cable letting the head move up and down, there would be another rod whose end would be attached to the head, and the other part would be moved up and down via a motor at the junction of the x/y rods. This would let the printer head move anywhere in a 3d coordinate plane within a specified area.
I was planning on giving the coordinates for each peice of dental cement via a serial cable from a PC. The "image" to be created would be in raw form, not any other 3d modeling file form like .3ds or .lwo. This format would go something like: (x,y,z)(x,y,z)(x,y,z). this would make 9 numbers per triangle.
->One of the problems here is that i need to tell the printer to print all the area inside this triangle, not just on the individual points.
->Another hard spot for me would be the programming of the PC program that hands out the points and the BS2 program that interfaces with the pc
all the rest of the project, building, materials, motors, and programming the less important parts of the program. I can do. I can also make a concept-art image, later- because I have homework to do.
So I guess I am asking for any help and constructive criticism you guys can give me, and help on the bulleted items above ;)
01-26-2006, 11:27 AM
A couple of thoughts for you:
First, in terms of movement, check out the following thread on a "Community Supported PCB milling machine". This project does xyz positioning. The link is: http://forums.parallax.com/showthread.php?p=539099
Another track to take would be to move the "print head" only in X and Y, and have a plate move in the Z axis. This way, you would build up the model layer by layer.
One other thing to think about would be using the Parallax/HP ink jet print head. You could fill the cartridge with something other than ink, and this might work. (don't ask me what to fill it with, I only have the concept) The product page is: www.parallax.com/detail.asp?product_id=27949 (http://www.parallax.com/detail.asp?product_id=27949)
This is quite a project for a high school, but if you have the funding for the hardware, should be doable. There is information in the above thread on the milling machine, as well as other sources on the X, Y and Z movement and control. If you want to google for things, you might have some good luck with DIY CNC or something like that. On one hand, you're making a printer. On the other hand, the movements you're after (both in terms of X, Y, Z control and getting all the points in a triangle) are very close (i.e. identical) to CNC milling.
The biggest challange might be figuring out what to build the model out of, and how to deposit it. You mention wax and dental cement, and this might work well. As another possibility, some of the original "rapid prototype" machines just used wax. The parts were fragile, and were not "functional" but it allowed folks to see the "finished" 3-d part.
Good luck and let us know how you do.
8 + 8 = 10
01-26-2006, 01:14 PM
I was going to do this for a high school TSA project, but then I figured it wasn't worth the trouble and the costs of maintaining the system.· I really didn't need to build one anyways since my school has both a CNC mill and a 3d printer ( http://www.zcorp.com/products/printersdetail.asp?ID=1·).
The 3d printer at school uses a special hp print head with a special external 'ink / glue' tank that feeds the print head durring the print cycle.· The ink is squirted onto a very thin powder layer, layer by layer.· The glue and the powder is very high quality and extremely expensive.· It costs hundreds to thousands of dollars a year to run the machine, not to mention the cost of the machine itself was about $15,000 on an educational discount.
01-26-2006, 01:45 PM
http://forums.parallax.com/images/smilies/smile.gif the printer head might be a good idea, but i have a very limited budget of about $$0 if I can help it... besides, it would print letters, and i would have to print periods in order to get some sort of line.
I think a syrenge idea would be good... I made an image in blender (www.blender3d.org) a very cool program, i might add...
The red boxes are the motors, the blue and yellow cone-things are the wax and dental-cement dispensers (perhaps some sort of syrenge or pump system, probably a pump system) The wiring and tubing and BS2 arent included in this render.
01-26-2006, 08:36 PM
I understand your budget, but just as a point of interest, the ink jet head is not limited to letters, it can also do "graphics" and/or lines of various widths. It is fully programable.
8 + 8 = 10
AutoCad will save the (X,Y,Z) values in a text file with a .DXF extention. This would be easiest to parse and send to your Stamp. You would have to 'rasterize' each layer some how. You could just draw it as a raster for each layer.
You might be able to find 'AutoCad Lite' heavily discounted for school projects and I'm sure there are other 3D programs that export to the .DXF standard as well. Blender may even have .DXF export available! I haven't used it in years but yes its a very cool (still free?) 3D program. Good luck on your project.
01-28-2006, 04:03 AM
indeed, blender is still free and is now in version 2.41, hopefully it will remain free. And yes, it does export to DXF
To solve the problem of having it print all the points in a triangle, not just the vertices or edges, I will have the computer do most of the work. It will take the current Z level of the mesh, and create a 2d array of all the points on there that need to be printed. it will then send the coordinates to the stamp, which will go out to that point and print a "pixel" or "dot" at that current location, it will then tell the computer that it is ready for the next point, and the program repeats. The stamp knows at what position the head is at through the use of some sort of rotation sensor on the wheel that moves it on an axis. I am thinking it will be an optical sensor, it will shoot infared light at a circle of paper that has black and white stripes radiating out of the center. every time a black stripe goes across, it has moved one position. Depending on how the printing head prints, I will set a resolution for the printer. the higher the resolution, the longer it will take to print.
I am going to try printing just wax first, then use a combination of wax and cement to get the "good" stuff ;)