How Are Size-based Costs Determined When Printing?

Discussion in 'General Discussion' started by mikeoui, Oct 20, 2016.

  1. mikeoui
    mikeoui Member

    When figuring out a price to print a design, there are usually three elements: part, material cubic cm, and machine cubic cm. How are the cubic cm numbers determined? For the material, is it based on the bounding box or the total amount of actual material used? For the machine I assume it is bounding box. Trying to estimate costs for future designs (I won't waste the time designing them if the cost will be too high.)

  2. CopperBezel
    CopperBezel Well-Known Member
    If you go into the 3D tools and select "Machine Space", you'll see how that figure is calculated - it's a skin projected around the parts with a particular clearance depth. For instance, my latest print was a vehicle miniature kit that showed up as 103.3 cc, where the bounding box would have been 378 cc. Other models can be printed into the space not occupied by these blobs, and I believe these individual shells are rearranged individually for a best fit.


    Material volume is the actual total interior volume of the mesh. In this case, my model only had a material volume of 28.9 cc, under a third of the machine space calculation; that would be the volume I'd read by doing a displacement measurement on the final printed model, the actual amount of plastic used.

    In the case of WSF, that material volume is only a proxy for material cost, since the unused powder from that whole "machine space" bubble can't be reused anyway, while the calculated machine space is only a proxy for actual machine space, since a pile of dozens of different shells from different people are not going to arrange into a perfect fit and consume all of the available space. Part count is of course used as a proxy for labor (and I could have saved myself $8 above with a couple more runners.)
    shawn_halayka likes this.
  3. shawn_halayka
    shawn_halayka Well-Known Member
    That's awesome information. :)

    Edit: I wonder what kind of algorithm they used to generate those blue hulls. Is it an isosurface extracted from some 3D scalar field?
    Last edited: Oct 21, 2016
  4. MrNibbles
    MrNibbles Well-Known Member
    And of course prices are calculated differently for different materials. Review the material information sheets to get the specific details. For example porcelain price is calculated by surface area so having a solid piece would be cheaper than a hollow one even though it uses more material in the final product.
  5. Youknowwho4eva
    Youknowwho4eva Shapeways Employee Community Team
    Key point "in the final product" Seeing we print a mold, and cast the porcelain in that mold, the mold itself would have more printed material in a hollow porcelain part :)

    I don't know how much @AlanHudson can say about the algorithm.
  6. MrNibbles
    MrNibbles Well-Known Member
    Also some materials like frosted detail are priced using the "classic" method of material volume. But strong and flexible and metallic plastic are priced by material volume plus machine space, where machine space is always greater than material volume and is highly dependent on the geometry of the object.

    The bottom line is that the original question of knowing whether or not to waste time on a design (because of final cost) depends on various factors that depend more on experience than being able to provide a simple list that is guaranteed to work for all materials. Sometimes making a hollow object larger can actually reduce price!
  7. NoahLI
    NoahLI Well-Known Member
    easiest just to upload your rough design and get an estimate. Block it out using basic shapes, upload it and get a cost estimate. You can be confident the estimate will be the upper bound, assuming your final design scale doesn't change drastically. if estimate is acceptable, proceed with final design.

    most 3d packages should have a measure tool to give you basics like bounding box dimensions, volume, and surface area. leverage that as you are designing.

    Porcelain is a little weird. Spec sheet says maximum wall thickness is 50mm. but in reality it's any section of volume must be less than 50mm in the longest dimension. e.g. a slab that is 100mm x 100mm x 50mm would likely get rejected as the longer dimensions exceed 50mm. Only reliable way to check it myself is having a 50mm cube that I move around inside the design and make sure there're aren't any chunks larger than the cube.
    Last edited: Oct 21, 2016
    CopperBezel likes this.
  8. CopperBezel
    CopperBezel Well-Known Member
    I just want to +1 that making a mock-up of the model and testing that is a very good first step. It certainly makes things a little less mysterious and guessworky, and I've been doing it as a first step myself with everything I'm modeling.
  9. siquod
    siquod Member
    To me, it looks as if they do something similar to a closing, except that they perform the dilation with a 40 mm diameter ball and the subsequent erosion with a 38 mm ball, in order to get the 1mm clearance. In other words, inside of the blue shell are all points that a 38 mm sphere cannot reach if its center keeps a distance of at least 20 mm to the model. This will keep all holes open that are larger than 40 mm in diameter, as the description says.
    One of my (currently private) models has a circular hole with a diameter of 20 mm in a plane of thickness 1 mm.
    As would be expected if my guess is correct, the hole is kept open even though it is too small, but the blue shell extends way more than 1 mm into the hole:
    Also note how the blue shell hangs down into the six cut-out semicircles (they have a radius of 9mm). The curvature radius of the blue shell over these gaps seems to be about 20 mm.