PLA shrinking and maybe a important rule !?

[Elysio]

Some important questions and observations / impressions in terms of PLA shrinking.

Yesterday I reactivated my Max, after getting a new glas bed, because I broke it for 2 weeks ago.
This glas plate is now 6 mm tight. I did the calibration again and it work very well (better than a 3.5mm glas plate ! )

So I send yesterday night a new print job to the printer and it starts , this morning I saw that the edges of the object were bended and I killed the job immediately. Okay lessons learned I never will start such a big object without bed heating, with smaller objects this is possible. Anyway, this seems to be logical.

But there is a disturbing fact !

It seems to be that if you print without heating the bed, big objects are bigger than they should be. **

I really dont have an explanation for that, but I saw this morning that this object does not fit to its counterpart of a previous printed object, which were printed with a heated bed and with a glas plate with 3.5mm.
There were over the whole object constant differences of a few milimeters 3-4mm , that has consequences in details and diameter.

So this part is so or so crap.

But I dont have logic assumptions for that ?
Do you have any experiences how much PLA can shrink ? Is it that much if you print with and without heated bed ?

An overview of questions :

- What is the max. coeficient for PLA shrinking ?

- When i print on cold glas the shrinking happens immediately, So I think this will help to get a object which is more accurate in dimensions on the base than a object that is printed on a hot bed ?

- Can I create a rule, lead out of this observation -> "print cold to cold and hot to hot" ! ? Wich means if you have to split an object in many parts, it is important to take care which side belongs to each other, and they should be printed as the rule tells it above !?

- Tiny objects can be printed cold? -> Big obects has to be printed hot?

- Does the slicer calculating with the seated temperature of the bed ? So is it possible that the software modifies the G Code in terms of a heated bed and If i print then without a heated bed this will fail ?! Seems to be alocigal thing, but maybe this is a wrong idea ?

I hope this are the right questions, and my english is clear enough to understand them ?

Best regards and thank you in advance for your answers.

** ( Attention, this seems to be the answer for the question, i guess the dimensions are more correct printing on a cold bed than printing on a heated bed )

[McSlappy]

Firstly, I have tried PLA without bed heating, but it's just too much hassle in my book. 60º doesn't require much power and saves a lot of work so I do it by default.

- Not sure on the shrinkage coefficient, but it's kind of hard to say "PLA shrinks by X" since there's a large variety of manufacturers producing PLA filamant and each use different ingredients. Cheap filaments contain more dodgy fillers and who knows how much they affect shrinkage. In the end decent PLA should shink unnoticeably.

- The generally accepted idea is that you want to slow the cooling of the body of printed filament so that this sort of sudden shrinkage does not occur. Moreso with ABS, but it does apply to PLA. While we do try to cool the newly laid filament with a fan, we're just aiming to get it closer to 60º which is the glass transition zone - this allows new filament to be laid on top without it drooping or lifting.

- I would agree with printing all parts of a single object using as close to the exact settings as possible. On all fronts not just the bed temps.

- Yes if you're printing small objects, I'd say they're safer to be done cold than large ones. Also the infill amount matters. Objects with sparse infill are more suited to cold prints than solid or heavy infilled ones. Personally I say print them all at 60º and never worry about it

- I don't think the slicer makes any extra modifications to gcode when you use a heated bed or not. I could be wrong, but I don't think it does.

If you're finding your dimensions are wrong when printing hot, perhaps try a different brand of PLA - there shouldn't be that much variation.