Is there a reason to choose 10-20% grid over 30% adoptive grid or not use adoptive layer height  

Yes

Are you software engineer or mathematician 😉

RE: Is there a reason to choose 10-20% grid over 30% adoptive grid or not use adoptive layer height

He's right.

What prevents you printing for 12 hours or more?

I have another 23 hours to go on my current job but apart from checking in when I get home tonight and maybe first thing tomorrow morning I shall just let it run ... I'm using a 2kg spool so no running out worries.

Cheerio,

RE: Is there a reason to choose 10-20% grid over 30% adoptive grid or not use adoptive layer height

• Fire safety concerns means I don't want to run overnight or unattended - at least not until I have an automatic suppression system in enclosure on (which requires enclosure).
• That should be solved today once I get a precision wrench to tighten the nozzle but my nozzle is leaking so it needs cleaning every 4 hours or so to prevent clamping.
• I'm still beginner so there is a bit of things that can go wrong. For example yesterday I tangled filament which ruined job mid print. I lost the job but it less of an loss if you do it with 4 hour job than
• I have cats which means that without enclosure I need to keep them separately. Which means that either I need to let them out for midnight zoomies across living room or let them in for morning snuggles into bedroom.
• I don't know how much filament is left and I'm still waiting for ressuply.

So the question was more about settings than reslicing.

I have no clue why Prusa uses 20% grid infill for those models, in particular in light of their own knowledgebase article about infills: https://help.prusa3d.com/en/article/infill-patterns_177130/

To quote their article:

The grid infill is more solid (and has better layer adhesion) than the rectilinear infill, however, it sometimes can cause annoying noise or even a print failure due to the nozzle going over the crossings where material accumulates.

I for one have certainly experienced exactly those crashes from the grid infill crossings while printing several enclosures. Now I'm using 15% gyroid and have not seen one crash. More recently. I've shifted increasingly to adaptive cubic infill, with good success, but I have not (yet) printed an enclosure with it. I don't see why it would be an issue.

Infil.

I try to print my print with the lowest infill that the product will take and remain viable after the standard use.  Heck I print rocket parts with 5% infill to make them as light as possible and they will take 500 mph flights.  I also print with low infill to print faster.  Printer time is money.  

Ok. I printed a lot of enclosure with 30% infil. Will try 10%.

In general I found the defaults (at least set in .gcode) to be very strange and set the support offset to 0.2 mm etc. (with their defaults I had hard time removing them, after tweaking a bit for next one it started coming off in one movement). Since printing enclosure is first thing people do probably they should be resliced...

10% infill caused warping so 20% or 30% is needed.

Infill decision is not cosmetic

Lowering infill percentage compromises final part strength.  If you are going for high strength, go with infills that are designed for strength - such as the octagon infill, hilbert curve, 3D octagon when you need strength.  IMO, print time is not something you trade off when you want the outcome to be strong (resist flexing in multiple directions is my present definition of strong).

Weak infills - easy to tear a print in parts with some twisting - include gyroid, adaptive, concentric - all of these create a light, fast print, but have very little internal support needed to withstand some abuse of the final print.

Yes, to shorten print time, mess with infill - the fast print settings in slicer help even more (combine infill every 2 or 3 layers, for instance).  This makes a spongy and rather weak infill - but you get fast prints.

Your reasoning for 12 hours are not uncommon for a first month owner.  It won't catch fire over night if you built it well, tightened all of the thick wire screws on the power supply, einsy board and on the heat bed.  Printer in your bedroom?  Yup, is your first month surely.

Re: Infill decision is not cosmetic

@dan-rogers - I think you misunderstood me:

• I'm moving from 10-20% infill grid (default) to 30% adaptive cubic as time is the same. My understending is that latter should be stronger and yet it doesn't seem to increase print time.
• Regarding 12 hours - that's why I'm splitting the build not decreasing the infill.

There are several techniques you can use to optimize print times without compromising strength

A few thoughts:

• If you want stronger parts, add perimeters and layers. Stefan did a great test of infill pattern strength over on his CNC Kitchen YouTube channel and concluded that adding a perimeter mattered most.
• A perimeter printed with a wider extrusion width will increase wall thickness and strength without adding a lot more time. Depending on your specific print and nozzle choices, a lower number of wider perimeters may provide stronger walls with fewer perimeters, thus saving significant print time.
• Variable layer height can make a huge difference in print time if the print consists largely of vertical surfaces or walls that can be printed at higher layer heights. Stick to a max of 80% of your nozzle size for any extrusion and results will be solid.
• Calibrate your actual maximum volumetric speed for your chosen nozzle & filament. Most of the filament presets have rates that are chosen for a good balance of print quality and time. If you can increase your hotend throughput (e.g., using higher temps, less viscous filament, or larger nozzles), print times can drop dramatically. 
• If your infill is for strength, Adaptive Cubic can be a good choice. If your infill is just to support the top layers, don't forget only support where needed and Support Cubic infill.
• Don't forget you can set separate extrusion widths for infill (resulting in fewer, more robust lines) and combine infill layers (assuming the combined total stays below 80% of your nozzle size) for further optimization.

I've compiled a bunch of notes on optimizing prints here that might be of interest. The biggest cube print time can be reduced from an estimated 5d+ to roughly 13.5h.