Suggestions on Orientation or Supports on Chassis Rear Case?

Hi All.

I’m struggling with printing a specific part and could use suggestions on using the right combo of orientation and support structures for a clean print.

I print primarily in PETG Basic and would like to continue learning to improve my prints with that filament.

I have been using ASA as a support interface structure. It breaks away very easily, but as you’ll see below, that comes with layer adhesion issues on XZ / YZ planes.

Stainless 0.4mm Nozzle (currently installed)
Hardened Steel 0.4mm Nozzle
Hardened Steel 0.6mm Nozzle
Glue Stick on Textured PEI plate
I haven’t ventured far from the default PETG Basic settings.

Here’s the part:

When I orient like this: (using ASA interface with tree/auto supports)

… I get bad bed adhesion on the ‘upper deck’ of the shelf support structure and I end up with this:

When I orient like this: (with PETG interface)

I get poor quality on the 45d overhang:

Any suggestions?

Orientation of the 4th photo, use organic or hybrid tree supports on the plate only and increase the brim size to 10 or 15, check the plate when you slice it post a interface photo of what it slices please

Here’s the image you requested:

Note that my original attempts required me to paint in some manual supports for the wider holes highlighted here:

Here are the settings I used:

Note that my printer only offers the following support types:


Here’s the photos you asked for but in the vertical position (which I think is what you actually requested?)

You actually asked for an “interface photo.” Does that mean you only want to see the interface layers? Do you also want the main body visible? (I assume you don’t want the supports themselves?)

Studio 1.8.0 Public beta and OrcaSlicer also have Type Styles, which give you more options. I’ve had good success with the Tree Organic style - effective support, easy removal, minimal surface damage.

I think the poor 45° surface (and small defects elsewhere) with the vertical orientation could have several causes. Be certain that the PETG is dry, and try using a gyroid infill instead of grid. You might find further improvement using OrcaSlicer filament calibrations - temperature, flow rate and pressure advance in particular.

Use this one. Try it in a less expensive filament type that you have plenty of to test, but this should work pretty well with minimal errant strands on the bottom of the top most surface, then you can change the material to the part specification material of choice

I would likely cut the model along this line and flip both halves onto the bed. You’ll get the nicest finish on the top and can glue the halves together. If it’s too thin, then add in a wider seam along the line to give some points you can add connectors or glue. You shouldn’t need supports for the front holes if you print at a fine layer height 0.16 or lower and may just need a small support for the top half slot and square hole.
Or redesign with consideration for the slicing.

I can’t see which one you are referring to. Can you reply again and quote me on the post you’re referring to, or embed the corresponding image in your reply?

The Vertical one, clean the plate so it holds well

I had considered this. Unfortunately overall size and available space inside are both limiting factors. I don’t think I’d be able to add enough of a seam to get a proper hold, and there would be no space for mechanical fasteners in there.

Still, your comment brings up another question for me: Why do you expect an improvement in print quality for the 45d overhang by printing it closer to the bed? Do you suspect a lack of stability as the cause?

Thanks for clarifying!

My current best print of this model was indeed in the vertical position.
In my case, I used the “normal” support style and not the ‘tree’ style you’re recommending, and if I used a brim it was smaller.

Can I ask how the choice of brim size and tree support style is likely to improve the print quality of the 45d overhang? Is it because the stability of the loftiest sections is better using ‘tree’?

I don’t doubt that you are correct, I’m just trying to understand the correlation.

For me I go off the height of the model orientation for the brim, my “default” is on the plate only so 20-40 depending on the size and weight of the entire model, it just makes 30+ hr prints less stressful, if you look at the branches on the top at the interface level it does it sparsely over a wider area, so when you remove the tree’s most of it stays on the tree and the layers above are supported for the build up of future layers, which is why the K values are extremely important for each filament so it’s the right amount to fuse them fully and using the correct temps dialed in helps achieve that too.

More that if the model is split into two, you can lower the layer height which will make printing 45° easier to get a nice finish. Also, there’s more time for the layers to cool if you can print both halves on the bed at the same time. Because it prints so fast, the layers don’t get enough time to cool properly on sloping walls. I believe this is a part cooling fan issue myself. The only way I can get it to reliably look nice is to slow down those layers a lot and at a lower layer height.
It’s unfortunate but there’s a few ways around it but I believe you’re going to have to redesign. Maybe just increasing the wall thickness will give you enough to glue? Personally, I’d probably print in ABS and use acetone to join the halves together, or try to reduce the angle of your sloping wall so it prints nicer.

1 Like

Thank you! Much of this is really relevant to how I use my printer, so your comments are very appreciated.

I have no problem setting the speed to ‘snail-crawl’ mode for those layers if that is likely to help the build quality. I had feared that the opposite might be the case (that layer adhesion required the previous layer to still be warm.) Very helpful.

I’ve had some success bonding PETG with methylene chloride. I’ve also got ASA or even PAHT-CF if PETG is the problem.

The wall is angled to keep the bulky D-SUB62 molded connector from interfering with the mounting mechanism and racking tube. If I lessen the angle, my mounting is obstructed. If I increase the angle, I can’t get a soldering iron in there.

Follow-up question: Why does splitting in two allow me to lower the layer height? Can I not lower the layer height and print monolithically?

Sorry, very early here I’m not 100% woken up and just realised what I’ve written lol. Splitting in two won’t make it any different to being able to lower the layer height, of course you can do for the whole model. It may not even increase the time a lot. You can also try variable layer height to achieve the same result. What you want is to slow down printing that slope, or give each layer enough support underneath so it doesn’t sag when the next one is put on top (extra weight, heat and half the filament in mid-air). Actually thinking more about it, printing it with a 0.6 nozzle at lowest layer height may work well for strength but again will depend on your models wall thickness if it will end up printing 1 or 2 walls and also another factor is if you use classic or arachne. When designing I now try to make the wall thickness a factor of the nozzle printing width I’m going to use, which results in strong, good looking prints.

1 Like

Thanks again. And good morning!

So, to be clear, is there still a benefit to cutting the model in two if layer height can be used without the split? And if so, what is the benefit splitting would grant me?

It should print fine if you don’t split it and can slow it down on the slope or by using lower layer heights. It’s just my preference to orient the pieces so I get the nicest look at the end which usually when the top is against the plate. IMO, It would also look nice splitting like I said but then printing on a textured plate and adding fuzzy skin to match the plate texture. Someone on here had a great write-up on the best settings to use which results in no visible layer lines and an overall nice looking shell.
EDIT: found the post

1 Like

I did end up with a more cohesive 45deg overhang with fewer imperfections when I set the layer height and speed to much slower speeds during the overhang part of the print. Thank you for the suggestion!