Stiffness versus Torsion

[RobsonBH 1]

I have a part (model airplane aileron) that is thin, narrow and long. I need this part to be light, rigid enough to avoid twisting because it is long (161.6 x 22.6 mm). I tried several configurations of layer thickness, infill, temperature and unfortunately I was not successful. Could anyone help me.

[Slashee_the_Cow 438]

What material are you with printing with? Also remember that print speed and cooling can affect the strength of a print.

 

I find that Gyroid infill is pretty good at avoid twisting. Octet is good for models that aren't very thick but on a piece this small you'd need to turn the density up so not good for this particular case. Tri-hexagon is another one which would work if your piece was bigger.

[RobsonBH 1]

Hello, good morning. Thank you very much for your interest in helping me. I am using Esun's ePLA LW. I made some prints also varying the temperature because I noticed that the higher the temperature (I have worked with up to 245 degrees Celsius - and the print already turns yellow) helps in stiffening the part. I have also made variations in the thickness of the layers and also in the infill (lines, triangles and Gyroid), but without good results. The entire model is printed and ready to be assembled, however, the ailerons are not of sufficient quality to avoid twisting. I believe that the big problem is that the aileron is very thin and this prevents all types of infill from producing good results. In wood construction, what matters is the internal structure (stringers and ribs), as it is this that keeps the piece rigid. I thought about asking the project developer for help, but in another question I asked, he did not respond to my email. So, I am looking for help from someone who is really interested in helping and cares about other people's problems. Once again, thank you very much.

[Slashee_the_Cow 438]

I wouldn't know the first thing about woodworking (literally quit high school because of a woodworking class) but we can get some nice internal structure which for all I know might be like stringers and ribs.

 

So here's my 160x20x20 bit:

Note that the direction it's going on the build plate is important. For this method you need it to be straight along the X or Y axis.

Set Infill > Infill Type to Grid.

If this is what stringers and ribs are like, I think I was justified in my school ragequit.

We can straighten this out though: set Infill > Infill Line Directions to [0,90]:

It's still just thin lines. So let's turn up Infill > Infill Line Multiplier (in this case I'm setting it to 3):

Okay, now each one should be stronger. But it just added a ton of filament. Thanks to the extra lines we can drop Infill > Infill Density. I started at 20%. Here it is at 10%:

Are some thick straight lines like that anything like what you're thinking of?

[gr5 2,235]

First of all gyroid is the strongest infill for compressive strength only - it gives you the best strength to weight ratio but only if you plan to stand on it.  Not twist or bend it. But that doesn't help you much.

 

1) I understand the mechanical properties that relate to bending strength very well and with that the thickness of the part matters a ton.  The strength goes up by the thickness to the fourth power I think so 10% thicker is 46% stronger. I'm not sure about twisting.  With bending the part can be hollow and you lose no strength.  That's why bones are hollow (well there's marrow in there but it has no strength).  So if you could make it just a little bigger.  I know: not your design. 

 

2) for twisting it really does seem to me that ribs are indeed important - when you wring a towel it has to compress the inner portions and ribs would help a LOT.  So I love the advice from @Slashee_the_Cow

 

 

3) You could also insert carbon fiber rods strategically in the design: leave a void where you can insert the rods and glue them in there. Again, I don't even know which way they should run.  For bending you run them the long way.  For twisting?  I think the same.  Those ribs will help with compression associated with twisting but I think you also will have tension the long way and carbon rods are stunningly good at that.

 

4) Stiffness in 3d printing materials is known as young's modulus (or tensile modulus or modulus of elasticity).  There are not-as-good setups to measure the modulus so sometimes it has other names that translate somewhat badly because they are consistently higher or lower values (I forget which).   I'd definitely look at that value for different materials.  Pretty much every filament you can buy has a material property data sheet but I usually google "modulus" and the name of the material and can usually find a pdf or a web page on the manufacturers website for your filament.  Higher numbers are stiffer.  CF (carbon fill) and GF (fiberglass fill) materials tend not to be a lot stronger in tension but are usually significantly stiffer.  Also it can be measured in mpa, gpa, or kpsi.  google can help you translate among the units (just a constant). These filled materials should only be printed on a printer with hardened steel feeders and nozzles or you will wreck your nozzle and feeder both.  There are other materials not as impressive as GF such as polycarbonate which are also quite difficult to print.

 

I have an interactive web page that helps you look at materials and shows their young's modulus on the X axis.  You can zoom in with mouse drag.  You can hover over any material for more info.  I created this quite a few years ago and haven't added to it in much too long.  But again, google the material you are interested in and "modulus".  Maybe add the manufacturers site in the search e.g. "site:esun3d.com  modulus" google search gets you info on lots of esun products.

[gr5 2,235]

Oh - and with modulus, higher numbers are stiffer.  Note that if you want a "tough" material you want the modulus lower (more flexible).  Not what you want there but good for say a nose cone that might hit the ground at high speed to absorb shock. Nylon parts are incredibly tough and you can even drive over most nylon printed parts as they just flex and then bounce back after.  But this is the opposite of what you want for wings or for ailerons - you want stiff.

 

So definitely avoid any "tough" materials for your wing or ailerons.

[Slashee_the_Cow 438]

Not having looked up moduli this is just an "off the top of my head" thing but I don't know what they do to PLA to turn it into lightweight PLA but it's possible that might lower the rigidity. You could try using regular PLA (slightly heavier part that works > part that doesn't work, right?), but not "PLA+" - usually that has parts of some materials (often TPU, which is pretty damn flexible) to make it a bit harder to break.

 

I should know what's in eSun PLA+. I'm literally printing with it as I'm figuratively speaking to you.

[RobsonBH 1]

I would like to thank you all once again. Following your instructions and some of my knowledge about aviation structures, I am testing the information you have provided me. I will be posting updates soon. Thank you all.

Edited October 2 by RobsonBH

[RobsonBH 1]

Dear all, I was successful. Unfortunately, I didn't save the file and lost it. I tried in several ways to find how I found this result (from the attached photo) but I can't. This piece is printed according to the attached print. It is tilted approximately 30 degrees. Another observation is that when turning the piece on the table, the infil is not vertical but tilted. The result was excellent and the rigidity improved even more when I spread a little cyanoacrylate on all surfaces of the piece. Bingo. Could anyone give me an idea of how I got the result in the photo? Thanks again.

[Slashee_the_Cow 438]

Do you still have the gcode file you used to print it? Cura can load quality profiles from those, but when it comes to positioning, your best bet is to open two instances of Cura side by side, load the gcode in one and the model in the other and try and match the angle manually.

[RobsonBH 1]

Windows screwed up my SD card right after printing this part. There is no other solution than to make thousands of more combinations in an attempt to find the solution again. Since the aileron is large, it is printed in two parts. One of the parts fits entirely on the build plate and that is why I am trying to redo the file with it. However, the other part only fits on the build plate if it is at 45 degrees and besides (because it has no support) it rests on the build plate at about 30 degrees. It is complicated.

[Slashee_the_Cow 438]

Unfortunately instructions on using forensic-level data recovery tools to try and rescue your SD card is probably a bit beyond the single-text-message-at-a-time forum medium but I hope you figure it out without tearing out too much of your hair 🙂

 

55 minutes ago, RobsonBH said:

It is complicated.

Don't worry, I never thought printing a plane was simple 😆 although there's a guy up the street doing one who sometimes need some help so I know it can be a lot un-simple.

[gr5 2,235]

There are a few ways to use the rotation tool in cura.  I forget how to switch among them.  If you click the arrows I think it only does 90 degrees.  There's another mode where it does only increments of 15 degrees.  And another does increments of 1 degrees (maybe hold down some key like shift, ctrl, or alt).  It depends if you click versus drag versus where you click and so on.

 

Anyway there's a strong chance that your rotations where in multiples of 15 degrees.  If this is true, it shouldn't be very hard to duplicate the rotations you performed.

[RobsonBH 1]

I believe I found the solution. The problem is in the position of the part on the table (X axis and Z axis), see the print. I used the Triangles as Infill Pattern. I just can't understand the number of print lines. The models have the same thickness and width characteristics, but the length is different between them. What I can't understand is that the number of print lines is different between them. I believe they should have the same number of print lines, but one file has 75 print lines while the other has 79 lines. I went through the Print Setting item by item and couldn't find the reason for this. I checked the scale and its correct.  Another detail is that I can't increase the length of the infill lines. In the part that I printed (photo I posted earlier), the infill line was closer to the end of the part, which is ideal. Now I can't do that.  

[RobsonBH 1]

58 minutes ago, gr5 said:

There are a few ways to use the rotation tool in cura.  I forget how to switch among them.  If you click the arrows I think it only does 90 degrees.  There's another mode where it does only increments of 15 degrees.  And another does increments of 1 degrees (maybe hold down some key like shift, ctrl, or alt).  It depends if you click versus drag versus where you click and so on.

 

Anyway there's a strong chance that your rotations where in multiples of 15 degrees.  If this is true, it shouldn't be very hard to duplicate the rotations you performed.

In the Rotation function, simply uncheck Snap Rotation and you can rotate the part by up to tenths of a degree. It's difficult to do this using the computer's touch pad because I don't know what the shortcut keys are. But this is how I managed it.

[RobsonBH 1]

2 hours ago, Slashee_the_Cow said:

Unfortunately instructions on using forensic-level data recovery tools to try and rescue your SD card is probably a bit beyond the single-text-message-at-a-time forum medium but I hope you figure it out without tearing out too much of your hair 🙂

 

Don't worry, I never thought printing a plane was simple 😆 although there's a guy up the street doing one who sometimes need some help so I know it can be a lot un-simple.

I believe I found the solution. The problem is in the position of the part on the table (X axis and Z axis), see the print. I used the Triangles as Infill Pattern. I just can't understand the number of print lines. The models have the same thickness and width characteristics, but the length is different between them. What I can't understand is that the number of print lines is different between them. I believe they should have the same number of print lines, but one file has 75 print lines while the other has 79 lines. I went through the Print Setting item by item and couldn't find the reason for this. I checked the scale and its correct.  Another detail is that I can't increase the length of the infill lines. In the part that I printed (photo I posted earlier), the infill line was closer to the end of the part, which is ideal. Now I can't do that.  

[Slashee_the_Cow 438]

The number of layers corresponds to the total height of the print. When placing an object like that, a couple of degrees of rotation could make up that difference (since it won't be sticking up in the air quite as much).

[RobsonBH 1]

Yes, you are right and today I was able to better understand this difference between the two files. I am here printing another aileron with a new configuration. I believe it will work although the infil will not be identical to the one in the photo. After printing I will see if it is rigid enough for it to fulfill its function. Thank you very much for the information.

[gr5 2,235]

A reminder: as slashee mentioned earlier you can change the rotation of the infill as well.

[gr5 2,235]

10 hours ago, RobsonBH said:

but one file has 75 print lines while the other has 79 lines.

I suspect you can tweak these by adjusting the infill percent.

[RobsonBH 1]

On 10/3/2024 at 11:36 AM, gr5 said:

I suspect you can tweak these by adjusting the infill percent.

I don't remember what I did, but this problem was solved. It was something related to rotating the part. Well, do you know how it is possible to rotate the infill so that it follows the same rotation of the part?

[RobsonBH 1]

On 10/3/2024 at 11:34 AM, gr5 said:

A reminder: as slashee mentioned earlier you can change the rotation of the infill as well.

Hi! Do you know how to do that?

[gr5 2,235]

Yes.  Read her post.  Above.  From Wednesday.  Her second post in this thread.

[RobsonBH 1]

On 10/4/2024 at 2:22 PM, gr5 said:

Yes.  Read her post.  Above.  From Wednesday.  Her second post in this thread.

Hi! I read it again and did some tests here. Little by little it's happening. Thank you very much for your help.