No announcement yet.

spin casting

  • Filter
  • Time
  • Show
Clear All
new posts

  • spin casting

    I'm about to set up a spin casting operation to make a deep parabolic shape. By deep I mean about 1/2 inch diameter at the exit point (the top of the container, which is basically a plastic test tube) and about 2 inches deep. In the bottom will be an led. I'm expecting that much of the light output will be focused by reflection from the surface, since the angle of incidence will be small. I'm going to use ordinary fiberglass resin to begin with so I can get the rpm right and to get a feel for the process.

    Once I have the shape correct I can experiment with reflective coatings, and determine if any reflective coating is even required. The goal is to turn the light output from the led into a spot of reasonable quality. Optical quality is not required, but a nice looking spot is the goal. It's not going to be a laser beam, I'm well aware of that. If I can get much of the light into a spread of about 10 to 15 degrees, without the typical optical aberations that most flashlights exhibit, that will be great.

    If I can get a decent result, the next step would be to cast a resin inside of an aluminum tube. The led will mount on an aluminum plug which will be inserted into the 'bottom' end of the tube. Because leds require heat sinking, this structure will offer that, with little else being required to make it a finished product. The plug to which the led is mounted will probably have two pins molded in to make external connections to.

    I may have the option to include a reflective material within the resin, with the hope that it would enhance the reflective qualities of the resin surface- I have no idea how this would work, if at all.

    This experiment is only going to work with leds which have a small light emitting area. This is usually enhanced by having a lens shape molded to the led. The particular ones I'm using are 3 watt with the emitting area appearing to be about a mm in diameter.

    Bear in mind that this is an experiment at this point, though it does have a useful application for me. I would like to be able to machine the parabolic shape into the aluminum itself (starting with solid rod) but I think that's going to be quite daunting to achieve acceptable results. If I could get the required shape, with a fine enough surface finish, then it would work well. Another option would be to do the spin casting, call that a mold, then cast a plug into that. The plug would be removed and coated on its outside with a reflective coating, and the led would be inserted into a hole in the end of it using a clear epoxy. Heat sinking would have to be provided secondarily, and the cast medium would have to be very clear and void-free- but it could also then have a slight lens created on the exit end of it. Lots of hassles with this method.

    I've done some spin casting years ago and got a very encouraging result. In this case the diameter was about one foot and the focal distance about 8 inches or so. The optical quality was more than good enough for focusing light, but at the time I stopped short of having it metalized. That thing had a diameter to depth ratio of about three- the one I'm going to try now will have a depth to diameter ratio of about 5 or so. Reminds me more of a cosmic ray gathering device than a lens or mirror.

    All thoughts welcome.
    I seldom do anything within the scope of logical reason and calculated cost/benefit, etc- I'm following my passion-

  • #2
    I've seen videos of astronomical telescope mirrors being spun cast, very slowly and with controlled cooling of the molten glass to produce parabolic blanks. This reduces the time consuming grinding and polishing stages.


    • #3
      I find the whole process fascinating. When I did my first spun-cast experiment I was struck by how optically smooth the result came out. I could put a tiny bulb near the focal point and get it to focus on the wall where I could see the coiled shape of the filament. That's without doing any reflective coating on it. It was proof of a concept for me. I rigged up an old video tape deck motor that had a toothed disc on it that was used for speed control previously, so I made up a phase lock circuit to run the motor and lock the speed very precisely at something like 25 rpm. I think that part of the good result I got was because the bearings in such motors are by necessity very smooth. I did my casting in two parts- the first was to define the shape of the mold to begin with, and the second was to add the final layer at a more or less even thickness. I used coffee table epoxy because I thought it was probably formulated to flow out nicely, which it did. I used a torch on the surface to remove bubbles, just like is recommended when pouring resin on a slab to make the top for a coffee table. I kept it spinning for two days just to allow a full hardening. This latest experiment will use higher speeds, and I'm hoping to get the focal point somewhat behind the bottom end of the casting so there's a proper place for the led. I don't expect to get anywhere near the kind of focus I got with my original casting, and I know that much of the light will be reflected off-angle, probably creating a huge coma. Still might be workable, that's what I'm hoping for.
      I seldom do anything within the scope of logical reason and calculated cost/benefit, etc- I'm following my passion-