Notifications
Clear all

Weird print failure - advice sought  

  RSS
kenryan
(@kenryan)
Eminent Member
Weird print failure - advice sought

So I had a weird print failure.

The print is a grip for the Switch 2.  I printed one in all-TPU (red), came out fine bu is too floppy.  So I made a version where the wide flat part is painted with PETG (clear).  The print included PLA supports which worked great, and aren't relevant to my question.

When I tried the tpu-petg combo it started out good but then developed a strange lean.  One side started the issue with a slight layer shift, but the rest has this smooth curve almost as though it was designed that way.  The tpu and petg curved together.  Note I had the additional problem that the spool tangled & broke, which is why the tpu stops before it got to the top.

In the photos below I'm holding both the original tpu-only print and the tpu-petg combo with the problem, just to show what the structure is supposed to look like.

I thought at first maybe it was the tpu getting extra resistance as it started to tangle, but that should not affect the petg.  Another thought was maybe the print started to lean, but when I took it off the bed it was still attached well, including the supports. I didn't sense any significant play in it.

Anyone have any thoughts as to what I should look for?  The print takes almost a full day so I would prefer to have something to try tweaking before I restart it.  But if I do retry it with no changes, what should I be looking for?

Other info: pla in extruder 1 (supports), tpu in extruder 2 (grip), petg in extruder 3 (back).  Latest prusaslicer.  Wipe tower set for extruder 1.  Printed standing right-side up with supports both for the back and grips.  My XL has the prusa enclosure, the lid was closed but bellows down.  I've attached the slicer project.

Thanks in advance!

ken

 

Posted : 13/06/2025 2:37 am
kenryan
(@kenryan)
Eminent Member
Topic starter answered:
RE: Weird print failure - advice sought

Not sure why the project file is refusing to attach...

Posted : 13/06/2025 2:43 am
Neophyl
(@neophyl)
Illustrious Member
RE: Weird print failure - advice sought

You have to zip most file types files to attach them. 

Posted : 13/06/2025 4:54 am
Diem
 Diem
(@diem)
Illustrious Member

It does look as if a small layer shift was followed by a slow warp - please show us the underside of the failed print.

Zip your part file before attaching.

Cheerio,

Posted : 13/06/2025 8:46 am
kenryan
(@kenryan)
Eminent Member
Topic starter answered:
RE: Weird print failure - advice sought

Here are a couple underside views, and an attempt to attach the project as a zip (thank you!)

Posted : 13/06/2025 1:29 pm
Diem
 Diem
(@diem)
Illustrious Member
RE: Weird print failure - advice sought

Thank you.  It seems the part partially detached and slowly aligned to a new position under the stress from the cooling contraction of the longer extrusion runs.

Although you have set a brim it's not connected to the part - Set: Print Settings > Advanced > Elephant's foot compensation:    to zero.

If it's possible, also adjust the design so the grips have a larger footprint - maybe substitute some of the under-curve for a shallow chamfer to increase the diameter at the print sheet.

And protect the printer from drafts.

Cheerio,

Posted : 13/06/2025 3:30 pm
kenryan
(@kenryan)
Eminent Member
Topic starter answered:
RE: Weird print failure - advice sought

Hmm, OK thank you.  I thought it seemed like it was still firmly upright but now I'm doubting myself.  The brim was attached to the supports at the handgrips, I'll try changing that setting and check it in the slicer.

(The design itself is from Bruce Lucas on printables, I don't have the skills to make something with ergonomic curves myself 🙂 )

Posted : 13/06/2025 5:38 pm
Diem
 Diem
(@diem)
Illustrious Member

I thought it seemed like it was still firmly upright but now I'm doubting myself.

The lift line is visible in your pictures.

I don't have the skills to make something with ergonomic curves myself

It looks fairly simple, quite a good subject for a first project, suitable for an OpenSCAD beginner.  I'd be tempted to enlarge the top surface enough that the whole thing could be printed upside down and so avoid support.

Cheerio,

Posted : 13/06/2025 6:25 pm
kenryan
(@kenryan)
Eminent Member
Topic starter answered:
RE: Weird print failure - advice sought

It is?  I'm not understanding what you're seeing, even though I have the thing in my hand.  Can you elaborate on what you mean by "lift line"?

I greatly appreciate your help!

ken

Posted : 13/06/2025 6:31 pm
Diem
 Diem
(@diem)
Illustrious Member

Look for a change of texture or drop a straightedge across, one side of the line has detached from the bed:

Cheerio,

Posted : 13/06/2025 10:40 pm
1 people liked
Diem
 Diem
(@diem)
Illustrious Member

I don't have the skills to make something with ergonomic curves myself

Here is a quick and dirty OpenSCAD script that should, once you have copied and pasted it and edited in your correct dimensions, make a useable part.  OpenSCAD is free to download and open source so well worth learning the basics for something like this, tinker with this as you learn.

 

// Drop-in controller holder

//  All of these figures are user editable in millimetres
//  measure the parts and enter your desired values.

//  Edit to suit your hands
gripdiamax=40;
gripdiamin=30;
gripheight=85;
gripmaxheight=70; // How far from bottom is widest part

// Measure your controller...
controllerwidth=280;
controllerdepth=15;
// controllerheight not set, it is assumed support height is enough

socketdepth=10;

supportheight=60; // height of support bar
supportthick=3;   // and thickness

// curve smoothing
$fn=50;

// Designed upside down for easier, support-free printing.

difference(){
  union(){
    grip();
    translate([0,controllerdepth/2,0]){
      cube([controllerwidth+socketdepth*2,supportthick,supportheight]);
    }
    translate([controllerwidth+socketdepth*2,0,0]){
      grip();
    }
  }
  // Subtract the shape of the controller from the holder to make a socket
  translate([socketdepth,-controllerdepth/2,0]){
    controllershape();
  } 
}


module grip(){
  // Shape the grip with two cone sections
  cylinder(d1=gripdiamin,d2=gripdiamax,h=gripheight-gripmaxheight);
  translate([0,0,gripheight-gripmaxheight]){
    cylinder(d1=gripdiamax,d2=gripdiamin,h=gripmaxheight-(gripdiamin/2));
  }
  // and dome the bottom end
  translate([0,0,gripheight-(gripdiamin/2)]){
    sphere(d=gripdiamin);
  }
}

module controllershape(){
  // very crude, should be edited to actual shape with about 1mm of tolerance
  cube([controllerwidth,controllerdepth,supportheight]);
}

Cheerio,

Posted : 14/06/2025 12:38 am
1 people liked
kenryan
(@kenryan)
Eminent Member
Topic starter answered:
RE: Weird print failure - advice sought

You have a heck of an eye ... if I put the light on it just right I can kinda see the line you're talking about.

Thanks again for the help!

Posted : 14/06/2025 3:39 am
kenryan
(@kenryan)
Eminent Member
Topic starter answered:
RE: Weird print failure - advice sought

 

Posted by: @diem

Here is a quick and dirty OpenSCAD script that should, once you have copied and pasted it and edited in your correct dimensions, make a useable part.  OpenSCAD is free to download and open source so well worth learning the basics for something like this, tinker with this as you learn.

 

Great!  I'll give that a try - I would like fatter grips, I'll see how far this gets me!

ken

Posted : 14/06/2025 3:41 am
Share: