Model to big-need to split it

[chappie99]

Looking to print an object the shape of a frisbee, however, it is too large for the printer therefore I have to split it into 2 pieces. I'm new to 3D printing so I'm not sure how/where to split it so that it doesnt lose its structural integrity when I put it back together.

Any feedback is appreciated.

[fred_dot_u]

Without an image to reference, I'm going with a wild guess here. Shape of a frisbee implies a saucer/circular shape, but with contours approaching that of an aircraft wing in cross-section. Too large for the printer requires some guesses as well, in that you may mean you can't lay the "frisbee" flat on the surface of the printer. Such a layout would result in layer lines being placed in parallel with the "line of flight" and may be the strongest method to print.

By the same token, one may not require that level of strength, allowing for layer lines in other directions.

If you split the disk radially, in two pieces, across the diameter of the disk, will the model have sufficient thickness to adhere to the print bed and sufficient thickness to be glued together? As an extension of this concept, if you make the cut across the diameter of the disk but slant the cutting plane to 45°, there will be greater surface area on the cut, more surface area to stick to the bed and to stick to the glue, plus the layer lines will be diagonal rather than perpendicular to the disk plane.

SEdisk.jpg

The image above has the cutting plane across the diameter, but has been tilted 45° to the right. The resulting cut is of greater area than if it were sliced vertically. Of course, this is simply a flattened sphere and a frisbee would have much less material, but it serves as a suitable demonstrator.

Another option, if you have the height in your printer and at least one of the horizontal dimensions of your bed is large enough is to tilt the disk. It will obviously require support, but you'll have a good compromise between layer line strength and ease of build.

How about a drawing, sketch or photo of the model?

[chappie99]

Hi-

I've attached the .stl file. Please keep in mind that this is a rough sketch, however, its pretty close in terms of dimensions. Thanks for the feedback

Hubbie.stl

[chappie99]

fyi: when i tried to get a pricing quote from online 3d printer, i got an error message saying "exceeds maximum dimensions". plan to print in plastic. since i'm doing a prototype i figure 3D printing is the way to go...?

[fred_dot_u]

Those thin walls are a killer! Meshmixer thinks the maximum dimension of the model is just under 43 mm, although the alternative indication is that it's slightly smaller than 43 inches. What is the true diameter? What program did you use to create the model? How thick is the dome? What is the airspeed velocity of an unladen swallow?

Certainly, 3D printing is a good method to use for prototyping, although for thin enough material, one could create a buck, a forming device and hammer aluminum or steel sheet to create the dome, adding a 3D printed clip, but that's outside the scope of this forum.

If it is necessary to chop the disk up into quarters rather than halves, the thin portion of the dome will create a weakness. If cleats, also known in the radio control world as doublers are used, the dome gets thicker where they are attached. It can be unsightly, but strong. It appears that an internal cleat/doubler would be invisible in your design and provide the necessary strength for the prototype.

I sure hope that thing isn't 43 inches diameter! I can print up to about 8" with a good bit of luck, but would still segment the disk and tilt it to get a better fit on the print bed. The re-assembly would require a bit of sanding on a true surface but the results are likely to be acceptable for a prototype. Printing in ABS would make for a strength improvement, and allow for acetone smoothing. Acetone smoothing would also hide some of the glued joints if carefully assembled.

A challenging project, to be sure.

[fred_dot_u]

I suspect that it's not 43 mm, as the commercial resources mentioned earlier would not have rejected it as too big. If it's 43 inches, there might be a scaling problem with the originating software.

According to Siri, it's about 25 mph for a European swallow, so you're close.

[chappie99]

Its supposed to be 18" in diameter with 1/4" thickness. Created the file in Inventor and then converted to a .stl file for upload to get a pricing estimate. New to Inventor as well so the thin walls is a design error.

Long story short. Looking to print dome shaped object 18" in diameter with 1/4" thickness, and 3 clips that extend out as fasteners. Drawing I attached only has 1 of the 3 clips. Looking to get a ballpark cost est but then ran into the max dimensions error which I need to resolve so I can get a cost est.

Google says airspeed velocity of an unladen swallow is 31-40mph

[fred_dot_u]

457.2 mm in diameter with 6.3 mm thickness is indeed a large print. With three clips, you could split the model radially along 120° arcs and get parts that can be printed on many of the larger consumer grade printers.

In the STL file, the clip is inboard of the raised rim. Is that part of the design or could the outside of the clip be flush with the rim edge? If it can be flush, the dome third could be printed with the clip in vertical orientation reducing the amount of support needed.

Further with that thought, the "pie slices" could then incorporate a tongue and groove joining or a few tabs and slots. I think tongue and groove would be a better choice, as it would provide easier post-processing. The groove or recessed slots are not going to be a perfect fit and the groove would be easier to sand out than tiny slots. On the other hand, a series of longer slots with shorter tabs would also work and perhaps provide a compromise level of design. With 6.3 mm thickness, the slots and tabs can be 2 mm with slightly more than 2 mm on top and bottom for glue bonding.

If printed in ABS, an acetone solvent can created a welded join, probably the strongest for such a model.

Postscript: I was wrong about the size. I used Meshmixer, resized it to match the dimensions noted, then chopped it up at the 120° points. It's an inch too tall to be printed in my printer. That would mean 60° segments, one triple set with clips, one triple set without. Double the workload in post processing, but still something that can be done with determination and patience. After running it through the slicer with the above considerations in mind, I would expect that you'd pay more than one hundred dollars and perhaps triple that. I'd give it a shot at a hundred bucks, but only with the understanding that the resulting model may not be perfect appearance and perfect fit with the six segments being the biggest factor.

You'd want to re-design the model to incorporate the individual segments for accuracy's sake. My butchering was done by eyeball, as I didn't have the source material and don't know how to use that program. The tongue and groove or slot and tab portion might be better off supplanted by flat face gluing, just to save labor.

[chappie99]

Thanks for the info...

1. If I am able to reduce the diameter by say 1 to 3 inches would that make a significant difference in printing costs? I may be able to go as small as 15" diameter. I would do scaled down version but I want to be able to test it out so it needs to be as close to full scale as possible.

2. Yes, the clips can be incorporated with the rim (I think). Since I'm in the initial design phase, anything is possible at this point.

3. Down the road if I were to produce say 20-30 of them, is there a way to print/produce them as one piece w/ the clips so I dont have to put them together. I.E. injection mold or even a larger scaler 3D printer?

I'm brand new to this. Hair-brained idea has been stewing for some time and I'm finally taking action so lots to learn.

I appreciate the feedback.
M.

[fred_dot_u]

If you can go to 15" diameter, I'd be able to print it, but it would still have to be in six pieces, as the amount of time involved for such a large item is prohibitive in a single piece print. I didn't pay too close attention to my slicer results, but at one point it was in the neighborhood of 75 hours for the entire model! Reduction to 15" or so means a reduction in support material and printing time by a substantial amount.

Good that the clips can integrate at the rim, as it prevents the requirement for support material under the clip when the segment is printed in a vertical orientation. Additional printing time and material reduction as a result.

If you can find a printing service that will do 15" in diameter, the entire model should be printable in a single piece, with supports, as the model will have to have a tilt and supports for the clip portion. If the clips could be added as an assembly, many things improve with respect to quantity production.

I don't know injection molding practices per se, but recognize that integrated clips would make for a very complex mold. Read "very complex mold" as "very expensive!"

I'm working on a project in my alleged mind that involves producing 160 pieces of a much smaller model. Less complex design means less difficulty, but still almost two hours per unit on the 3D printer. I am hoping instead to create a silicone mold for either four or eight pieces and create resin casting parts rather than 3D printed ones.

Your clip design would also complicate such a production plan. The disk would be easily resin cast in silicone but for the clips. A big advantage to casting is that you could create the part in the aforementioned six pieces, glue them together, sand and acetone smooth the surface to a beautiful finish and make the mold from that. The result would be no layer lines and no glue lines in the resin part created from the mold.

I've had dozens of hair-brained ideas in my life. Some of them are now far more possible than before, simply because of 3D printing, resin/silicone casting, home hobby mills and lathes and so many maker-related developments. It would be great to be in my teens or twenties now!