Design rules for FDM objects?

Discussion in 'General' started by UlrichKliegis, Feb 11, 2016.

  1. UlrichKliegis

    UlrichKliegis Well-Known Member

    Joined:
    Sep 5, 2015
    Messages:
    191
    Likes Received:
    48
    Hi all,
    is there a place or generally recommendable source of information for the proper design of mechanically loaded RP objects, specially such made in FDM / FF / Fx technology? The background of my question is the failure of some parts that looked nice and sturdy but were rather not.

    What has to be observed to ensure that lateral, torsional, dynamic, sheer etc. forces don't lead to a collection of fragments? How can one introduce mechanical strength in certain directions, how should shape and surfaces (and infills) be designed to achieve the desired result? Methods, tricks, secrets?

    I understand that one would generally need an engineering training for that, but is there some lower level available and reachable somewhere?

    Thanks, cheers,

    Ulli
     
  2. mike01hu

    mike01hu Well-Known Member

    Joined:
    Apr 22, 2014
    Messages:
    957
    Likes Received:
    167
    Hi Ulli, I have done the searches for this problem too, as I have access to FEA through SolidWorks but there is no model for 3D printing, only the ability to define solids. The dynamics of a 3D printed model are many and varied, as you know, and I cannot see a way forward at the moment. It is a subject for a university engineering graduate to research for his/her PhD.

    Mike
     
  3. UlrichKliegis

    UlrichKliegis Well-Known Member

    Joined:
    Sep 5, 2015
    Messages:
    191
    Likes Received:
    48
    Mike, thanks for the desillusion :) ! But I guess and hope, nevertheless, that there are some collections of basic tricks of what to do and more what to avoid to achieve some reliable results. What does the best printer to be delivered realsoonnow help, what does all that fancy 3D sculpting software help if the wannabe designer creates objects that break, fall apart, fail.

    In my case a compilation of some boxes and plates and boreholes that did not pass the first power-on-test - they broke at an edge where a block arises from a plane, which again is the 'deck level' of another flat block. There, where the largest forces appear, was the weakest point: The wall of the upper block just sat on the three-layer-lid beneath it, with nothing but some infill underneath. Lesson learned in this special case. We discussed here already the anisotropic force resistance properties in x,y, and z.

    But how can you overcome that? My idea is now to create an internal anchor structure by multiple tubes running parallel to Z through the block. Maybe even criss-crossed in space to achieve more rigidity. Does that make sense? Build time is no issue, aesthetics neither - form follows function.

    Cheers,
    Ulli
     
  4. mike01hu

    mike01hu Well-Known Member

    Joined:
    Apr 22, 2014
    Messages:
    957
    Likes Received:
    167
    Ulli, I have beaten the same trail as you but with little success in the z-plane. To combat this I have resorted to using steel wire reinforcement by designing in a hole/s through the structure then pressing and gluing a wire, typically 3mm welding wire, into the hole. Yes. it increases weight slightly but the z-plane strength is very good or even excellent if glued. I also use this to give stiffness in long structures embedding stainless steel rods (non-magnetic) in them; tricky but it works well. So there you have it, steel reinforced plastic. Now if I could buy cheap carbon fibre rods this would address the weight problem.
    Mike
     
  5. UlrichKliegis

    UlrichKliegis Well-Known Member

    Joined:
    Sep 5, 2015
    Messages:
    191
    Likes Received:
    48
    Mike, thanks, that's exactly one of the tricks I was having in mind without knowing them. A threaded rod and some nuts might also work.

    How do you glue them to what kind of plastic? Or threading some glassfibres through a hole, get them tensed and fill that hole / canal with epoxy. Far away from RP then. Thermal binding comes also to mind. Did you try aluminium tubes?

    And did you see this and this ?

    Cheers,

    Ulli
     
  6. mike01hu

    mike01hu Well-Known Member

    Joined:
    Apr 22, 2014
    Messages:
    957
    Likes Received:
    167
    Ulli, I only use PLA and I glue with epoxy or superglue but I am sure it would work with most filaments. As for the rods, aluminium is acceptable too and I have used it on occasions, also brass. I did try using heat to push the rods through tight holes but PLA is so sticky it proved difficult, so I made test pieces with slightly varying holes sizes to allow an easy fit the accept the glue. Depending on the size and strength required threaded stud and nuts is OK.

    Mike
     
  7. UlrichKliegis

    UlrichKliegis Well-Known Member

    Joined:
    Sep 5, 2015
    Messages:
    191
    Likes Received:
    48
    Mike, this leads me to the idea to insert metal tubes, no matter if steel, brass, alumi- or titanium in pretty tight, yet penetrable canals and then use a heat gun with an appropriate nozzle (and suitable temperature) to heat up the tube from inside in situ. It should result in a perfect hold. Might even work with epoxy-carbon tubes.
     
  8. Miasmictruth

    Miasmictruth Well-Known Member

    Joined:
    Sep 4, 2015
    Messages:
    804
    Likes Received:
    118
    If you write in some gcode to pause the print at the right layer and move the head away you can fully embed objects in the plastic aswell (or change filiment). After you have inserted the rod, resume the print and allow the printer to close off the holes.

    This may not directly impact you but if I have unsupported overhangs inside the part, I will make a single layer membrain bridge that you can break away in post real easy. That way the structure on has something to build off of. I have used this on some of the more complex models I have desinged to great effect allowing me to print geometry that would normally require supports.

    Lastly I am sure you are probably aware but the thicker the trace the better the interlayer bonding (think volcano nozzel) also certian types of material has much better interlayer bonding then others.
     
  9. mike01hu

    mike01hu Well-Known Member

    Joined:
    Apr 22, 2014
    Messages:
    957
    Likes Received:
    167
    @UlrichKliegis
    Ulli, I had thought of this too but tried putting electrically heated nichrome wire down the centre of the tube but electrical insulation was the problem as I did not have a high temperature insulator and it was difficult to control temperature, so I stuck to glue ;-) So much depends on the size of support tube/rod, in my case the largest used was 6mm but most were 3mm to 5mm, the size was determined by tensile or shear protection needed. Have a look at the two STL's here and if you slice then view the result by layer you will see the channels for the rods. In the idler model the two cheeks are strengthened by the vertical holes that run from top to bottom. The focuser plate is for a 355mm astronomical telescope I built that has about 200 printed parts and the telescope pictured is the subject for remodelling to Mk II with the BigBox.
    Mike
     

    Attached Files:

    #9 mike01hu, Feb 13, 2016
    Last edited: Feb 14, 2016

Share This Page