3
$\begingroup$

Note: Before any negative flagging, we asked permission on Meta prior to posting & got a YES: Could this Printing Material Recommendation Question be or shaped to be valid on 3D SE?


Which common 3D printing materials/methods will suitably replicate (with durability) this injection molded Polypropylene item?

Item to replicate via 3D Printing:
Club Handle - 5/6 image album (a few images are attached below as well)

Physics, Mechanics & Forces in play on object:
Usage Intent is not to create an Impact or Hit but to Flow in Circular motion like this: YouTube - 10 Best Indian Club Exercises

  • Item with PCO 28 PET Threading
  • 20 cm Length - Avg. 28 mm Diameter

Instead of me trying to figure out materials, I seek advice from experienced experts here.

Please advise and suggest on materials:

  • Ideal top 3 materials recommended/ most suitable for this item?
  • Other top 3 materials that are easier to get through "Entities" on 3dHubs?
  • Add Thought: Is "negative" printing is the way to go (Sculpting, lathe, threading, CNC types)?

Note: The immediate answer people gave was to use PP printing, however I am looking for alternatives because PP printing is not common or easy to find nearby and expensive.


Self Homework:

So I looked through these material guides which show different attributes and ratings for various parameters:

  • Print-ability
  • Strength
  • Stiffness
  • Durability
  • Price

For example, Polypropylene - Simplify3D - Polypropylene

Polypropylene is great for high-cycle, low strength applications due to its fatigue resistance, semi-flexible, and lightweight characteristics.

Entire Material Lists looked at:


Injection molded Polypropylene item - image 1 Injection molded Polypropylene item - image 2

| improve this question | | | | |
$\endgroup$
  • 2
    $\begingroup$ Besides being expensive and uncommonly found, PP is also pretty difficult to print! $\endgroup$ – 0scar Jun 26 '18 at 21:47
  • $\begingroup$ @0scar - Exactly my experience, but someone threw that answer without any alternatives when I asked the same question on reddit. $\endgroup$ – Alex S Jun 27 '18 at 3:09
  • 2
    $\begingroup$ Hmm CNC seems logical but those threads cant be done on a CNC . This part has to be moulded or 3D printed .I don't see much of a challenge in terms of manufacturing its just your material that is a limitation . $\endgroup$ – Axel Fernandes Jun 27 '18 at 4:08
  • 2
    $\begingroup$ @AxelFernandes There are special thread cutting tools, but they need a fully free outside for one half. but if you CNC, one could weld the grip... or include the cap in the design. Just the interior thread remaining, and that could be CNCed on a 5-axis machine. $\endgroup$ – Trish Jun 27 '18 at 7:49
  • $\begingroup$ Ah ok, I sincerely apologise for my mis-understanding, and completely getting the wrong end of the stick. I have rolled back the deletion. $\endgroup$ – Greenonline Jun 28 '18 at 8:24
3
$\begingroup$

No FDM print at all.

The problem of your design will not be the materials, but a basic property of FDM printing: FDM Printers do create a structure by placing a long string of filament next to itself and ontop of itself, creating tons of boudaries.

These boundaries between the layers are the weak points for this application: Even if the material like ABS could withstand the blow handled with such a club, the print will break at its weakest point - which in this case is any layer boundary. This is amplyfied by the basic design we have here: The elongated shape will serve as a lever on each of the weak boundries, until one gives way and results in catastrophic failure and a flying clubhead.

Non-FDM for the rescue.

To counteract this, you need to use a different method than FDM printing to get a more homogenous material than the bound deposited filament. Such methods could be for example SLA (Stereolithography) or SLS (Selective Laser Sintering). Both could easily offer even tiny details.

SLS uses Nylon or metal powders, sometimes even ceramics - Tungstencarbide for example.

SLA/Resin

Using a Resin printer using the SLA methods results in an object almost as homogenous as an injecion molded object. Proper aftercare and curing is required to get the best results. Also, Resin prints usually age under UV light, which can negatively impact lifetime. SLA printers are expensive (for home printers), print shops that offer them relatively rare and costly (in comparison to FDM) but usually offer superb resolution and almost perfect smoothness. A lot of the exact material properties is resin and aftercare dependant.

A way around the aging could be that the results of an SLA print could be used to create green-sand molds countless times, which can be used for casting metal or even some thermoplastics. Remember though, that cooling metal shrinks.

SLS Nylon

Nylon would be a medium rigid, light solution, but it ages and has a quite rough surface. It does offer some flex, almost perfect for this application. While most SLS machines for nylon are commercial to industrial, print technology of this kind is widespread enough to make them somewhat affordable (for an industrial printer) and printshops for these relatively common, prints are not cheap but well priced.

DMLS / SLM

Direct Metal Laser Sintering and Selective Laser Melting - an evolution of SLS - allows to create structures from various metals by sintering/melting powders of metal at the right spot to gain shape. The benefit would be, that you get a part that could withstand much more destructive testing than your bottle used as a clubhead - you get a workpiece of solid metal (Steel, aluminium, Titanium and a lot others are available) after all that has the same properties as a cast item. The big downside is, that only few companies currently delve into DMLS, among them the former patent holder of many of the FDM printing patents, Stratasys. This means, that a machine for this is industrial rated and priced, and that print suppliers charge accordingly.

| improve this answer | | | | |
$\endgroup$
  • $\begingroup$ Thanks for such a detailed response. Before I delve into reading further I'd like to update with how its meant to be used (Intent is not to create Impact or hit something or someone but to Flow in Circular motion). Like this - youtube.com/watch?v=fIEOWh87ahY - Hope that helps give more insight and maybe changes your answer a bit. $\endgroup$ – Alex S Jun 27 '18 at 3:07
  • 2
    $\begingroup$ The thing is, that thdre is a heavy weight (club head) mounted to the shaft. Any change of motion introduced via the shaft will induce a force that is similar to an impact but a little more gentle. The fact that you want a maximum homogenity and bonds along the length of the structure stays - FDM does not provide those. SLA, SLS and DMLS do provide these according to my knowledge. $\endgroup$ – Trish Jun 27 '18 at 7:28
  • $\begingroup$ Fair enough - FDM is out. PS: I just searched around via 3DHubs.com for SLS & Nylon & Resin and there's some possibilities. Will post back with the Resin & Nylon options that I get from the nearby vendors. $\endgroup$ – Alex S Jun 27 '18 at 7:41
  • 1
    $\begingroup$ @AlexSIf When you trimmed your material choice down, you might better do a new question, asking for a dirrect comparison of your small material choice. Mind you, we are all biased, and a lot of the actual print is design related: one could do an FDM print that won't fail under load, if cleverly designed and introduced to the proper aftercare and maybe a strengthening insert or two along the length, but that is an entirely different thing from "what is the best durability material combo". I added to SLA a little too. $\endgroup$ – Trish Jun 27 '18 at 7:47
3
$\begingroup$

If you want a 3D printing answer then

I see that the part has a few major points of concern to me:

  1. Threads (Gosh I hate threads).

  2. Tolerance - +/- 0.1 mm-0.2 mm should work fine (in terms of those threads fitting on the top of the bottle)

  3. Strength - This needs to be strong, so as to be able to effectively not break when spinning around.

In terms of tech your best option is - SLS

Layer resolution - 0.120 mm

For material options (In descending order of choice)

  1. PA 11 - Impact resistance is the major property for these parts. This is the material for your application . Here is the material data sheet - EOS PA 1102 Black.

  2. Glass filled PA-12 (I'll be taking this as second option) it has the strength but is extremely brittle . Material Data sheet - EOS PA 3200 GF.

  3. DuraForm ProX HST Composite (The expensive option) Yeah this will work. Check out the data sheet 3D Systems DuraForm ProX HST Composite

Now the Non 3D printing solution

So, to manufacture this design any other way is actually very simple but the threads are a serious area of concern.

  1. The body section can be cut on a standard 3 axis CNC.

For material options please choose a no brainer like polycarbonate or delrin. Find information here: Plastic CNC machining.

  1. Threads would have to be manually lathe machined onto it. (You need some beautiful hand work here). This might not be feasible by suppliers on 3dhubs.

Now my argument is to go for Additive manufacturing over CNC as you get the part straight out of the machine with the threads and material strength innate.

You can even make the part hollow using some design skills. This will make it lighter.

It would be very nice if you could get your 3D file ready so as to start this whole process and actually go about to manufacture it.

I can add more details if you want to. Just comment and ask.

| improve this answer | | | | |
$\endgroup$
  • $\begingroup$ F - I have some PET thread files from online sources but instead of putting all my energy into finalizing a 3D file and then finding out that I cant realistically or economically get it done nearby I figured lets get a reality check. PS: I just searched around via 3DHubs.com for SLS & Nylon & Resin and there's possibilities. I'll look up your material suggestions also to see availability. $\endgroup$ – Alex S Jun 27 '18 at 5:28
  • 2
    $\begingroup$ My friend professional advice . Dont waste your time on anything without a 3D file . You have no idea where things go once a file is ready . Spend some Time on getting the tough CAD work done everything else is a piece of cake . $\endgroup$ – Axel Fernandes Jun 27 '18 at 6:48
  • 1
    $\begingroup$ Slight side note: all the materials are nylons, while some SLA machines can also handle different materials. $\endgroup$ – Trish Jun 27 '18 at 7:32
  • $\begingroup$ @AxelFernandes - Now its more realistic & doable, because I have local vendors that do these "material options" suggested by both of you, so I can take few of the existing CAD files and merge them. Lots of thanks. Will post back with detailed material (lots of variations) info $\endgroup$ – Alex S Jun 27 '18 at 7:44
  • $\begingroup$ @Trish yup your right PA can also be called a type of nylon . Though I don't like to call it that because Dupont has a patent for the name and the specific type of nylon . So stick to call it PA ( polyamide ) . $\endgroup$ – Axel Fernandes Jun 27 '18 at 8:43
2
$\begingroup$

I do think that FDM is possible for this item.

Polyamide (nylon) would be my first choice for this, as it has great interlayer strength and it is very impact resistant and durable. Its failure mode will be permanent bending and stretching rather than cracks or a clean break like common FDM plastics like PLA and ABS.

I also needed a filament to simulate PP for FDM use recently, and found that Polymaker's nylon co-polymer CoPa in its conditioned state closely matches PP in its properties. Conditioned means that it has absorbed moisture which softens polyamide.

Regular nylon for 3d printing should have similar properties to CoPa, if a bit softer than PP in my experience.

However material is only secondary to geometry

I believe that in general the actual 3d model and its geometry is more important than the material itself. Indeed this item could be made using regular brittle PLA if the design is good, but it might be heavy or bulky. Thus choosing a better suited material is often a trade between weight, dimensions, and cost. Printing with PLA using very high infill might work, but would take too long and be too heavy compared to nylon, as an example.

This also means that the design of the part would inform the choice of the printing method to be used. Using FDM printing the Z-axis is a lot weaker than X-Y. But other methods are not completely isotropic either - they still have layers that can introduce stress risers and weaker bonding, but they are much less prominent than in FDM. Changing the design can enable you to choose a cheaper process.

Two important principles for designing for 3d printing to be considered before doing a final selection of material are:

  • Increasing the dimensions/wall thickness in weak or highly loaded areas.
  • Design with printing orientation in mind. Choose a proper orientation for the load the item will experience.

To exemplify for your particular case: Increase the thickness of the handle around the connection to the bottle. Lay it down with the long axis in the X-Y plane which will increase the strength tremendously, perhaps enabling it to be printed in PLA, ABS or PETG. Either split it in half and glue/screw it together, or use supports. You may need to print the threads upright as a separate part that is mated to the handle in the assembly process.

Finally, I would like to encourage you to embrace the main benefit of "rapid prototyping" - its speed and low cost to try things out. Make a first prototype and test it to destruction before settling on the design and material selection.

| improve this answer | | | | |
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.