Wood vs 3D parts: New video from CNC kitchen

I just viewed this new video and it clearly explains why PLA is a great material for most of the parts for our machines.

Wood vs 3D parts



Yeah. Very informative. The mdf and particle board strength is pretty surprising to me.

1 Like

Real Numbers!

PLA 25% more rigid than PETG. Let’s just bookmark that one.


I really like Stefan’s approach - but this time it raises more questions than answers I think. Mostly due to the practicality of testing at a different scale - and valid for his purposes of building now quite a large database of testing of identical elements.

Like most things (and Stefan says as much) there are a couple of caveats - had the mdf and particle board been turned through 90° the results would have been a little different - however they are not designed for any particular strength across the grain- they are best used as flat sheets in shear and constrained - so as a component of a box beam (or torsion) table with a wider cross section they are quite efficient as pieces of structure. Much later this year (or early next) I will try a few different versions using maybe a standard 50mm x 12mm section - that’s a promise! :wink:

In our context though - the plates on my LR2 in MDF can be easily deflected with thumb and forefinger - I am sure that they would be a lot stiffer printed in PLA - and perhaps “stiff enough”.

Maybe a permanent link is needed!


In the last couple of years, I’ve found significantly more datasheets published for FDM filament. Looking at just the Young’s Modulus on these sheets, I find the self-reported numbers to vary widely between different manufacturers. For example, for PLA I’ve found values as low as 2000 MPa and as high as 3400 MPa. As a counter example for PETG, I found SD3D PLA filament to have a reported modulus of 2300 MPa, whereas IEMA PETG has a reported modulus of 3000.

Looking at just the Young’s modulus, the most interesting thing I found was IEMA CF-PLA that has a Young’s Modulus of 5000 and is the same price as the Hatchbox PLA that is my primary printing material ($25 per KG).

1 Like

Oh if I swap out nozzles I will try that out!

1 Like

Do you think the reporting is accurate? I wonder sometimes if it’s just made up to keep us happy?
I am learning that there’s a huge difference between filaments - I use Prusament when I can and when I do comparison tests with other brands (I haven’t compared it to other premium filaments) some of the others have to be very finely tuned to get anywhere near the same end result - Prusament just seems to have a much bigger window to get things right.

That observation goes hand in hand with - if the secret ingredients make that much difference are the numbers being reported correctly?

I also have a spool of some “silk” filament that is just awful to print (PLA) has to printed below 200 and at that temp is useless for anything that needs strength - layer adhesion is very dubious.

I don’t think the numbers are made up, but I also don’t think they are a way to do an apples-to-apples comparisons. For example, some datasheets included more extensive information like a range of Young’s Modulus values, and include tests both across and with the layer lines. When seeing just a single Young’s Modulus number on a datasheet, it is hard to know what that number represents.

I’ve printed PLA from around ten different companies over the years. As you indicate, they vary in printability, but the properties of the finished pieces also vary. Given this variability based on “secret” ingredients or formulation, I have no trouble believing there is a wide range of Young’s Modulus values for PLA filaments from different companies.


Hi guys just a little input from an injection molding veteran. The specs that are on plastics data sheets are more of a guide line than actual hard number. The specs are very dependent on molding\processing conditions used for the test bars that are used. I no for a fact that process and testing methodology can make the numbers swing wildly. If you see great # from a spec sheet take them with a grain of salt. Those #s are real but only for the optimum processing conditions. Most of the spec sheet that you find for filament are pulled directly from the resin mfg. testing which is injection molded test subjects.


Very rarely do see true lab testing on 3d printed test subject.

I’m just curious, if PLA, especially carbon fiber and other voodoo blends, are at least as rigid as plywood or mdf, and most likely much more rigid than fiberboard and other (to me) over glorified cardboards, why isn’t it used as the brace material in a LR3?

Is it the contiguous nature of the brace that gives it the strength vs individual brace(s)?

Could perhaps overlaying sections of PLA braces be as good, or better? Or am I just lazy in preferring to print out braces vs machining them?

As I write this, I am thinking of all the reasons that this is a dumb idea. But I am curious as to the answers from this much smarter than I group :slight_smile:

If you got the seam very tight, or glued, and very straight I see no issue. But if there is any play in that connection, I believe you would lose significant rigidity. The Strut plates do 95% of the work on the beam.

A belt printer on the other hand, well now here is a great excuse to go get one.


I think he’d have had different results if he tested the plywood and osb at a different orientation. They are much stronger in shear/compression loads than bending loads. The way it’s used on the struts it will be in shear, and compression, just like a tube.

Even if you look at his plywood test rigs, he has the plywood oriented so the forces act along plies.

1 Like

Also for a material that’s not homogeneous, it won’t be invariant to scale, as stress/strain usually assumes. In other words half the cross sectional area might have less than half the tensile strength. Larger pieces will average out the inconsistencies for a higher average, and smaller pieces will have more of a ‘weakest link’ problem. Bending stresses can translate into local tensile/compressive loads over very small areas.

So something like this in “theory” might work and then again might not.

I get it, no worries. I might try it - if nothing else they might make nice temp. struts…

1 Like

At least with the printed parts you can crank down the screws to embed them into the plastic. I keep seeing these slots on surfaces that are made to be shear stiffeners and think “that’s going to slip”.

Biggest weakness for any polymer used in a practical application is fatigue.
And that was not tested.
And so it is useless unless you want a single use part.
Also PLA looses like 20% of it strengths when you get it too 40 degrees celsius.
So any usage close to any heat source such as a heater can ruin your parts.

1 Like

Hey, that’s my biggest problem after lunch!. All kidding aside, you raise some good points. I don’t think plywood cares how hot it is.

And realizing that I would have to print 16/24 strut plates (still haven’t figured out if all 3 sides get them or only 2) , and glue / pray that there is enough structural rigidity in the whole mess to make it worthwhile has convinced me that no, it is not worthwhile. I don’t think this is something to experiment with, I am by no means an engineer, I only play one on a train…

But, having printed two out already, they do seem stiffer than 1/4" ply. To be more scientific, a lot less “bendy”. Who knows, tomorrow I may wake up and decide to print a bunch more “just to see”. At least the holes line up with the braces, how bad could it be :face_with_raised_eyebrow:


The next strut plate takes care of the previous’s bendy (along the thickness, length) planes, that is why there are three. The real strength is in the width plane, as long as the plate is supported it will not bend that way. so 1/16" carbon fiber might be able to be rolled up into a tube but when bolted on all three plates with will not move.
The only weakness we have is twist/torsion, that is where plate rigidity comes in and hardboard/thick CF/plastic is king. Compared to the LR2 We should be leaps and bound better in bending along a plane, and better in twsit as well, how much is debatable since I was not concerned I did not test it even though I should have.

1 Like