##### How do I calculate optimal wall thickness when modeling in CAD?

I just installed a 0.8mm nozzle and am trying to print parts as fast as possible so I'm reworking my CAD to optimize the number of paths. I'm using 0.5mm layers as that's the best I was able to dial my printer settings in, but I am super confused as to how to calculate for this.

Originally I made a square tube with 2mm thick walls, printed vertically. I had set my default width to 1.0mm but left everything else on auto, and it created a tube with walls that are exactly 2 layers thick. The tower printed fast and was very strong, I couldn't even break it by hand.

Since 2 walls worked so well, I want to know how to calculate it so that it never adds the center wall. At first I thought it was subtracting the 25% overlap, so I made a 1.75mm tube and set the perimeter and external perimeter both to 1, expecting 2 walls exactly, but that didn't happen.

So I'm really confused as to how this calculation is suppose to work, and what is optimal. Should I want that center wall even though it massively increases print time? Is it possible to only get full walls and never center walls beyond 2?

##### RE: How do I calculate optimal wall thickness when modeling in CAD?

I’d be interested to know as well

##### RE: How do I calculate optimal wall thickness when modeling in CAD?

Isn't it as simple as 1 perimeter at 1.0 mm equals a 2 mm wall (assumes the wall has two external faces)?

Then for multiple perimeters, set both perimeters to 1 mm, and a 3 mm wall will have two external perimeters and one internal perimeter. A 5 mm wall would have 2 external, 2 internal, and will need 1 mm gap fill.

And, as far as I know and I could be wrong, the 25% overlap applies to infill, not perimeters.

##### RE: How do I calculate optimal wall thickness when modeling in CAD?

When I first read this question, I thought the answer was simple. It relates to gap fill. Turns out there's a ... bug-like feature in PrusaSlicer that goes way back to Slic3r. More below...

The over-simplified answer to @david-3's question is: "reduce the perimeters printed" in order to speed up print times. There are some nuances though.

When it comes to walls (which I think is what david-3 is after), you can specify a minimum number of perimeters under **Print Settings->Layers and perimeters->Vertical shells**. (Why they refer to the same thing as "shells" and "perimeters" I have no idea.) This controls the number of perimeters printed in thin walls where they will fit. It's important to realize that PrusaSlicer will use infill to fill narrow gaps between walls where full perimeters won't fit. Take a look at this sample:

A few things to note:

- With a wall thickness of 1.35mm using a default perimeter extrusion width of 0.45mm, there's not enough room for 4 perimeters (2 inside and 2 outside), so PS prints 1 each and uses gap fill between them.
- Similarly at 2.00mm, there is room for 4 perimeters, but a gap between them remains. PS prints gap fill between them.
- At 1.67mm, 4 perimeters are printed. PS twiddles the width (not sure where) to make things fit.
- At 3.00 mm, there's room for our 4 perimeters, but the gap between them is large enough that infill is used.

Meanwhile, the large inner area is printed exactly as specified. 2 perimeters + 15% infill in this example:

So when designing our part, you want to be aware of the extrusion width you'll be using if possible, but be aware that gap fill will be used for small spaces, so you needn't worry about an exact fit. I'm unaware of any strength penalty for using gap fill, so it's a "real" perimeter for all intents and purposes. Design your part to have walls as strong as you need but "mind the gap" if going too thick lest sparse infill be used.

Tim, I think you were referring to the overlap controlled by **Print Settings->Advanced->Overlap->Infill/perimeters overlap**. This is how much sparse infill overlaps perimeter walls, but doesn't impact the width of those perimeters.

So knowing this, I thought I'd do some easy screen shots to show all this and then... couldn't find "gap fill" anywhere. How the hell do you turn this on or off? Turns out it goes way back to Slic3r. You can disable the feature under **Print Settings->Speed->Gap fill** by setting it to 0. Obvious, no?

I could have sworn I saw a setting to control how wide a gap it uses, but I may be thinking of another slicer. I can only assume gap fill will be used for any gaps narrower than the specified perimeter extrusion width.

I've opened up a couple of github issues based on these issues.

*He is intelligent, but not experienced. His pattern indicates two dimensional thinking.*-- Spock in

*Star Trek: The Wrath of Khan*...

##### RE: How do I calculate optimal wall thickness when modeling in CAD?

Just designed a simple test - and of course - learned something in the process. But here's the print:

Top is 2.5, 1, 2,3,4,5 mm walls; bottom is 2.5, 0.9, 1.8, 2.7, 3.6, 4.5 walls. 0.42 first layer width (I was trying for 0.45, but oops).

* - while the forum devs seem to have fixed the default link to media issue, they broke the capability to adjust post format ... so now images are clipped rather than stretched, and a user has no way to correct it.*

##### RE: How do I calculate optimal wall thickness when modeling in CAD?

@bobstro

Wow, thanks for going into detail.

It is strange that 1.67 resulted in no need for gap fill since its not divisible by 0.45, the software clearly is fiddling with some parameters behind the scenes.

##### RE: How do I calculate optimal wall thickness when modeling in CAD?

@bobstro

I appreciate the detailed explanation. I was trying to understand the process and now it's very clear.

##### RE: How do I calculate optimal wall thickness when modeling in CAD?

@bobstro

I appreciate the detailed explanation. I was trying to understand the process and now it's very clear.

Glad it's helpful. I'm just passing on knowledge I've accumulated over time myself.

*He is intelligent, but not experienced. His pattern indicates two dimensional thinking.*-- Spock in

*Star Trek: The Wrath of Khan*...