Notifications
Clear all

How to engineer for part strength  

  RSS
JMcK
 JMcK
(@jmck)
Reputable Member
How to engineer for part strength

Short of "Just print a test piece and see what it takes to break it," is there an approach for determining how strong a part will be in the design phase?

So I'm trying to design a carrying handle for a part. (a beer bottle carrier, but that's not relevant) I figure the load is a bit under 3 kilos, the load would be not entirely static, but probably close enough.

If I were to print a rod to hold it; where the rod is printed vertically, i.e. layer adhesion is the "weak link", then is there a way of computing the load bearing ability of the layers? so like, "x diameter with n perimeters can hold up to y kg." I'm sure the tensile strength would be based on the cross-sectional area, but I never really took any statics or material design classes...

When someone asks you if you're a god, you say, "YES!"

Posted : 06/01/2019 4:16 pm
Peter L
(@peter-l)
Honorable Member
Re: How to engineer for part strength


Short of "Just print a test piece and see what it takes to break it," is there an approach for determining how strong a part will be in the design phase?

Not really, at least for hobby-level 3D printing. There might be some professional machines where the material properties of printed parts are sufficiently well studied to do this analysis.

There's just too much variability in how things are printed: even small changes in print settings can make a big difference in the part strength, so even if you had the data to do the analysis, it would probably be useless in the real world because of small variations in the printer and environment.

You're right that the bonds between layers will be the weakest part of any FDM print. So engineering for strength in a 3D printed part usually means trying to arrange things so the layers are parallel to whatever tension or shear loads the part has to take. (Compression loads can be either directly aligned to the layers or normal to the layers, but a compression force applied to the layers at an angle can become a shear force along the layers and fail pretty easily).

For example, I would not print a load-bearing rod vertically, I would print it flat. I would also suggest using a hexagonal cross-section rather than a cylinder, so it prints flat without supports (if you need the rod to rotate in a hole, I've found that hexagonal rods rotate in round holes just as well as round rods).

Posted : 06/01/2019 5:30 pm
JMcK
 JMcK
(@jmck)
Reputable Member
Topic starter answered:
Re: How to engineer for part strength


For example, I would not print a load-bearing rod vertically, I would print it flat. I would also suggest using a hexagonal cross-section rather than a cylinder, so it prints flat without supports (if you need the rod to rotate in a hole, I've found that hexagonal rods rotate in round holes just as well as round rods).

Completely agree, and that's how I'd normally do it, but for this particular project, I'd like it to print in a single pass, no assembly. The orientation is dictated by the rest of the part.

If I shed the "fully assembled" design, and go for multiple pieces, I could also go with making it not 100% 3d printed. Perhaps make the rod hollow and thin, so that a threaded rod could run up the center. Then the plastic is just aesthetic, and not actually load-bearing.

When someone asks you if you're a god, you say, "YES!"

Posted : 06/01/2019 5:48 pm
bobstro
(@bobstro)
Illustrious Member
Re: How to engineer for part strength

You can do some basic stress analysis using Fusion 360 - even with the startup/student license. I used to to check out the stresses on a cable clip I was working on for my deck holiday lighting. I only mastered the very basics, but was able to get some useful insights:

There are some good tutorial on YouTube. Check out Lars' tutorials.

My notes and disclaimers on 3D printing

and miscellaneous other tech projects
He is intelligent, but not experienced. His pattern indicates two dimensional thinking. -- Spock in Star Trek: The Wrath of Khan

Posted : 06/01/2019 6:09 pm
John
 John
(@john-6)
Reputable Member
Re: How to engineer for part strength

Maybe post an image and you might get some feedback on other options.

Other factors to consider might be degradation and creep over time or if exposed to UV light.
That said I am surprised how strong just PLA Is.

Also it is possible to add metal inside the print with a suitable void so you can’t see the metal.

Printing in two halves and gluing or male/female press fit is also possible if that helps to print in the best orientation.

i3 Mk3 [aug 2018] upgrade>>> i3MK3/S+[Dec 2023]

Posted : 06/01/2019 10:24 pm
Chocki
(@chocki)
Prominent Member
Re: How to engineer for part strength

PETG sticks very well to itself and is not as brittle as PLA. If you print it slow and with reduced cooling, it will stick to itself really well, to the point that any failure would actually be stretching rather than de-laminating. Obviously you would need to play with settings to find the balance of perfect bonding vs slumping.

In an ideal world you would want to deposit a layer onto a still soft previous layer, but not so soft that it just oozes outwards.
Temperature tower calibrations and stringing tower calibrations are well worth doing for your particular filament to find the balance between asthetics and strength and print-ability, then destroy the towers to find the strongest point.

I've done this and sometimes the nicest looking part is amongst the weakest with some material, yet 20 degrees difference and I could not break it without some mechanical assistance.

With PETG I've found that when it starts to look glossy, it tends to be stronger, but stringing becomes a problem and overhangs can sag. I haven't learnt slicer yet as I believe it is possible to alter speeds / temps for certain sections so cooling could be increased for tricky sections.

Recently I have been making parts using SUNLU PETG, with 30% cooling fan and 240 deg hot end, 90 Deg bed and I cannot de-laminate the sections, they have actually been breaking more like a solid hard plastic rather than PET when I have introduced them to the water pump pliers. the parts were a 36mm diameter tube shape with up to 2mm wall thickness and I wanted to see if I could pull the layers apart, the answer was no, I ended up snapping out chunks from the sides, but I had to give it quite some abuse.

Normal people believe that if it ain’t broke, don’t fix it. Engineers believe that if it ain’t broke, it doesn’t have enough features yet.

Posted : 07/01/2019 8:06 am
lawrence.g2
(@lawrence-g2)
Active Member
Re: How to engineer for part strength

The best text I have found by far is Functional Design for 3D Printing: Designing 3d printed things for everyday use - 3rd edition by Clifford T Smyth (ISBN: 978-0692883211). I bought it about six months before I bought my first 3D printer, and it has remained a useful reference.

Posted : 20/01/2019 7:04 pm
bobstro
(@bobstro)
Illustrious Member
Re: How to engineer for part strength


The best text I have found by far is Functional Design for 3D Printing: Designing 3d printed things for everyday use - 3rd edition by Clifford T Smyth (ISBN: 978-0692883211). I bought it about six months before I bought my first 3D printer, and it has remained a useful reference.

Great reference, I've ordered it. Thanks!

My notes and disclaimers on 3D printing

and miscellaneous other tech projects
He is intelligent, but not experienced. His pattern indicates two dimensional thinking. -- Spock in Star Trek: The Wrath of Khan

Posted : 20/01/2019 8:13 pm
Share: