Test Part Possibilities
1719 forum posts
There has been a number of comments over the months that ideal printing temperatures for PLA (and other materials probably) vary with supplier, batch, colour part configuration .... and possibly day of the week and whether there's an "R" in the month.
It has me wondering whether it's possible to come up with a small part that could be printed at different extruder temperatures when starting a new reel. Possibly multiple copies of the same part with the different temperatures printed on, to avoid confusion when comparing the resultant parts.
I wonder if that's useful and what such a test part should look like. Ideally, it should be something that could be printed in 5 or 10 minutes.
|Dick H||06/05/2020 22:52:22|
|95 forum posts|
There are things (tower shapes with bridges) for this in Thingiverse, search calibration and pla. The problem is to get the printer to print different parts at different temperatures, so as far as I can see you have to import their g-code and run it, i.e. no stl and slicing. I run at 210°C and 65 °C bed.
|Former Member||06/05/2020 23:57:24|
[This posting has been removed]
1719 forum posts
210/60 °C is nominal (according to Cura) yet I have some black that will string badly at 200°C and up. 190 seems good for it. I don't think I've ever printed over 210°C. (Of course, one machine's 210°C isn't necessarily another's).
I've seen and printed some calibration models on Thingiverse but many of them take several hours to print. I was looking to come up with something small that could be printed quickly one after the other at different temperatures to give some basic temperature guidance for specific filament batches. I don't expect it to be totally definitive (even if that's possible) but some starting guidance would be helpful.
I'd also prefer something modelled locally so I have CAD files with a feature-history rather than someone else's STL that I can't modify.
Edited By Bandersnatch on 07/05/2020 02:01:14
|jimmy b||07/05/2020 04:23:03|
678 forum posts
I run 205°C for the first layer/raft (if used) with no fan, then drop to 195°C, heated bed at 45°C.
The other thing to consider is the accuracy of the displayed temp, people do tend to believe digital as being correct!
|not done it yet||07/05/2020 07:11:55|
|5130 forum posts|
I’m no expert (only just started with mine).
STL files are not printer files - they need to be processed by the slicer program to be converted to code for the printer, to organise exactly the order of operations understandable by the machine. The slicer program will contain the temperatures, not the STL, which is basically an engineering drawing from a CAD program. Not all slicer programs are alike.
Small items will likely give different results to very large items. I cannot see that printing directly on, or close, to the bed is going to be quite the same as printing as much as 300mm (or more on some larger machines) from the bed. Layer-to-layer adhesion may be different, by virtue of this.
Adhesion to the bed seems a particular issue (it is for me, so far). As I see it, once the first layers are placed in position, heat transfer between the bed and print will change. Contraction can affect the adhesion properties. I like the idea of a cold plate, but they don’t mostly have heated plates for no good reason? As I see it, once printed on a heated bed, they need it to stay the same temperature to avoid the plate and workpiece parting company.
Enclosed printers will undoubtedly operated more consistently than open frame machines - especially if situated in less than ideal situations.
Comparing machines by net ‘influencers’ and often by good ‘neutral’ reviews may well be flawed by choice of parameters due to lack of truly exhaustive testing under different conditions.
Comparing ‘stringing’ between machines is one area that I have found to be questionable. My results, compared to those reported by others are different. I suspect my filament suits my machine and the particular slicer/machine settings.
The standard ‘benchy’ can even give variable results, and each machine likely needs tweaks (to settings) to optimise that little task on that machine. Mine did.
My conclusion is that whatever item, for a test, that may be chosen might work for one model and one PLA supply, but not aross multiple machines and variable filaments. But I wouldn’t bank on it.
|Cabinet Enforcer||07/05/2020 10:07:53|
|93 forum posts|
My experience has been that PLA is fairly tolerant stuff, and that temperature variation is something to finesse the last few percent of print quality, most of the time.
There are way too many other factors which affect issues such as stringing for temperature to solve alone; retraction settings in particular, since many printer maufacturers turn these down, or some of them off to speed prints up. Cura in particular seems to default to turning off the lift Z for travel and associated retractions, and alters the timing and routing of these moves all in favour of speed.
At least half the different reels of PLA I have bought have had temperature recommendations on them, which have been satisfactory, though to be honest I find 190degC is good enough most of the time that I only vary it if I have a specific problem.
Basically, the temperature question is like a "speeds and feeds?" question where you are still able to vary the diameter of the part, and what type of cutting tool you're using. The ballpark figure should be good enough 99% of the time, and when it isn't there will be part and other specific issues to consider.
|Neil Wyatt||07/05/2020 12:54:30|
18316 forum posts
It take a wee bit longer to print but look at this:
It has bridges, holes, bosses, overhangs, compound curves etc. and there's a downloadable sheet with all the dimensions on so you can compare your results to what they should be.
The compound overhang at the bow is a good test for curling up layers. The text underneath shows how good your first layer is and if your z-height is too low. The text at the stern is a good test for small nozzles and fine layer heights but will barely appear at 0.2mm layers with a 0.4mm nozzle.
Another advantage is that there are online guides to different print issues that use Benchy as the example making comparison easy.
|Neil Wyatt||07/05/2020 12:59:25|
18316 forum posts
Interesting point on temperature accuracy.
The thermistors used on most printers seem to be surprisingly consistent. I just took a reading off my bed with an IR thermometer and it gave 32C. The readout says 30C.
This is with the bed unheated as I'm printing TPU, ambient temperature reading off objects in the room is 26C.
Unfortunately the print head is too small to take a reliable reading off, especially a sit keeps moving!
|Andrew Entwistle||07/05/2020 13:40:11|
83 forum posts
I also found agreement within a couple of degrees between the bed and hotend thermistors and a temporarily attached type-K thermocouple. However on two different machines I have had the hotend thermistor spontaneously change calibration by more than 20C. Enough that on one the filament would not extrude and just ground up the filament with the extruder gear(s), on the other PETG was overheated enough that it produced prints with a burning smell that were very brittle.
The sensors were both the type that are I think still glass encapsulated Pt100, but in metal cartridges Neither had been damaged by grub screws and just had to be replaced.
Edited By Andrew Entwistle on 07/05/2020 13:42:20
|I.M. OUTAHERE||07/05/2020 13:51:19|
|1468 forum posts|
Ok for PLA on my printer i start with a nozzle temp around 210c and drop to 200 after the second layer but if i find the filament is a bit stringy i might drop to 180 - depends on how it prints .
Bed temp is 80c and drops to 60 after second layer
Using buildtak a build surface which is on a piece of borosilicate glass on my hot bed .
Even batches of the same material from the same manufacturer can like different temps and you can sometimes pick this up on the first layer as one will put down a nice flat layer but the next filament will look all knobly and rough along the edge as it lays it down - up the nozzle temp by 10 -15 deg and see if it gets better.
If you buy cheap rubbish filament you will suffer - trust me i know !
Speeds - start off slow work your way up - you have to crawl before you walk!
Don't buy into the exact temp crap as every printer is different so use what works for your printer , you may need to run a nozzle temp like 220 and a bed temp like 90 - it all depends on what your machine likes but the lower the better .
When you have maximised your constants you will have invariably minimised your variables
Edited By XD 351 on 07/05/2020 14:32:14
Please login to post a reply.
Want the latest issue of Model Engineer or Model Engineers' Workshop? Use our magazine locator links to find your nearest stockist!
You can contact us by phone, mail or email about the magazines including becoming a contributor, submitting reader's letters or making queries about articles. You can also get in touch about this website, advertising or other general issues.
Click THIS LINK for full contact details.
For subscription issues please see THIS LINK.