I have a filament that is shiny, when I do the auto-calibration, it sets the flow rate kind of high. When I do a manual calibration in general, I find I have a hard time judging the best flow rate, it seems too subjective. I will see a tile that looks the best, but the print results afterward don’t seem right.
I was recently printing a part with tight tolerances and I noticed that some of my filaments printed perfectly while one in particular I tried was too tight to fit. I noticed that the flow rate was pretty high and when I ran an auto-calibration again, it came with almost the same value(1.0226). I reduced the flow rate to the standard rate for Generic PLA (0.98), and now the part fits but is slightly too loose. So I tried again with a flow rate. of (0.99) and it was too tight. Then I tried 0.985 and it fits but is slightly tighter than it should be. I am going to try to fine tune it later.
One thing to note about my fit test that may be helpful in returning a more accurate value is that the male part is printed at a 90 degree angle vertically to the female part (done to increase the strength for the male part), so the layers of the male part are perpendicular to the female part, which I think will make the fit test better at judging flow rate because the layers don’t match.
Are the mating parts round? If so, I think that flow rate is not the right parameter to dial in the fit. Typically in FDM, concave walls (e.g. walls of a hole) become slightly outset (e.g. holes too small). I usually compensate that by designing holes 0,2mm bigger diameter in CAD. If that isn’t possible, you could try X-Y hole compensation ( i believe the concept of that feature should be done differently, but it works in many cases nonetheless).
This is one thing that shows up in bad ratings often, fitted parts work fine on the test bench but the moment it goes public all bets are off. I manually tune most my filaments as the auto is not entirely faithful in my experience as you’ve encountered here. I tend to give a little extra but it works against me in some models. It would be nice to have faith in a calibration reading and for it to be consistent but I guess I’m dreaming
Is there a quick calibration print that covers PA and flow?
I think one thing that can be done is to have a small calibration part that is two pieces that fit together with specific tolerances. I think printing each part in different layer orientations is a better way to reveal flow rate issues than printing both parts in the same orientation. Then before the person makes the print they can print the small test part. If the test part fits too loosely or is too tight, then the person’s filament calibration will have to be adjusted to work with the model.
It might be good for people to test with a part like that before publishing a print.That would be a good way to standardize all models if everybody used the same standardized part.
Here are the tolerances I use for my models, depending on how tightly I need a part to fit. The tolerance… is applied to both parts and tolerance…Clearance is when the tolerance is only applied to one part, so it is the full length of the space between the parts. The part that I am talking about in this thread uses my tight tolerance.
That would be the ideal solution, having the test print set up and then select print-profile 1/2/3 dependent on the test fit. Of course getting the user dialled into the mindset/patience is often harder particularly when the problems are filament calibration originally.
I think it needs to be made into a trend that gets widespread adoption among makers and that will hopefully translate into downloaders getting on board. I think people will come to appreciate it because it will save them spending a lot of time and money on failed prints.
Maybe @BambuLab@MakerWorld can make some efforts to popularize the idea on MakerWorld. It will fit in well with their concept of easy one touch printing.
If the fitting test part is one standard part for all MakerWorld, then there shouldn’t be a need for a 1/2/3 size version for models. Once people have their filaments dialed in to properly print the standardized test fit model, then at least in theory, they should be able to just print any future model without reprinting the standardized model.
Its wild how crucial fine tuning can be with some filaments/models, I avoid uploading a few of my models because I hate setting people up to fail. My side spool uses the X1 metal bracket and I’m having some issue with a couple of users who are having fit issues which I can only put down to this.
I didn’t realize before how much an impact a small flow rate difference has, in my example, 0.980 was too low, 0.985 too high, and, 0.982 good. I could probably fine tune it even more, but as it currently stands, it is a pretty small difference between a good value and too high or too low.
Most parts I make have looser tolerances, but this one needs to be tight. I usually like using screws, you can use looser tolerances that will be good in spite of printer and filament differences, and the hold is more reliable than glue or snap fit. And it is easy to take apart and put together to replace parts.
I am going to use this part of mine in the future to help fine tune my flow rate. I will just adjust it so it is easier to test the fit instead of just being a small cut out of my larger model.
It looks like my tight tolerance is too tight, at least for the filament that I’ve been testing.
When I get the flow rate low enough to fit the part, the top surface of prints are obviously under extruded.
I am going to experiment some more with the filament I am testing to see what tolerance will give me a tight fit while having a flow rate high enough so the surface doesn’t look under extruded. Then I will calibrate some other filaments with that fit test and see if the surface looks properly extruded. I wonder if I will find that different filaments need different tolerances to get a good fit while still having a flow rate that also makes a good surface.
I still think that flow rate is not the right tool to adjust fitting of parts. Flow rate should be calibrated to extrude the right amount. Good first and top layers are one indicator. It also has an influence on e.g. layer adhesion and others.
I compensate in the drawing stage, .010 total over/under provides a very nice snug slip fit. I usually check fit with a small test part to ensure the filament type does not mess me up.
I don’t adjust flow rate to correct for dimensional inaccuracies. When I need acuracy, I measure a sample print (or use one of the many dimensional calibration models) then calculate a shrinkage factor for my OrcaSlicer filament profile. Future models printed with the same filament are usually accurately sized within at most ±0.1 mm. but often within ±0.02mm.
Do filament settings get stripped from uploaded models? Additionally if they use their own filament profile. With holes/threads I tend to use xy compensation to get a happy medium, loose is usually better accepted than tight in my experience.
I have been using this test print to calibrate for fitted parts:
There is a spreadsheet to use with the .stl file for calculating the values to use. I always do a confirmation print to make sure it is correct. It is a quick print and does not use much filament. Make sure to use the recommended settings (i.e. no infill, etc.) when printed the test coupons.
I was looking at the comments for that Thingiverse model and a few people said that it made the parts fit, but they saw under-extrusion on the surface. That was my main concern after seeing the results of my own fit test.
One reply to the comment was that the Ideamaker slicer allows you to have a different extrusion rate for the top and bottom surfaces.
In that case, I would assume that none of these methods will help. At least nothing that is to your liking. As a rule, if you use a test that guarantees that the parts fit together, they will be slightly under-extruded. I was going to suggest my test, but it is not suitable for your requirements (1) only provides a relatively rough guide and 2) may be under-extruded).
Their requirements: the surface should be perfectly smooth (not under-extruded) and the accuracy of fit of the parts should be 100%.
In this case I would look for a preferred filament, there are really very good filaments, like PLA from Creality (please disagree with me, but I am always amazed at the prints from a new roll of CR-PLA). Once you have found your favourite of the filaments as a reference, dry the filament so you have a starting point for evaluation. Then calibrate this filament in the best possible way. Finally, test how well two parts fit together. make a model for this yourself and find out by trial and error how big the gap needs to be when designing the 3D models so that the parts fit together as you want them to. I suspect that you will end up with around 0.2 mm, maybe even 0.3 mm. However, always remember that materials can shrink after printing has been completed for a long time. What once fitted may be a little too loose 4 weeks later.
Unfortunately, FDM (but resin also has its pitfalls) is not one hundred per cent, even if we would like it to be and it would make many things easier.
