Issues around screw holes.

Hi guys, I would like to see if anyone has any suggestions for me to sort my issue out regarding pretty poor quality printing around screw holes.

My initial thoughts were that I have my wall thickness/line width in correct, they are set at 0.4mm line width with a 0.4mm nozzle and the wall thickness is 1.2 on this model. I have also tried 0.8 but that still isn't working very well either.

The images are here: [imgur.com]

Do you think i'm correct in thinking that this is the area that my issues is at? Or do you think it could be something else? The reason i thought that it would be an issue with the wall thickness, line width or flow rate is because it only happens around thin sections such as around the outside of screw holes, basically anywhere that is too thin to have infill.

My top and bottom layers are 1mm thick (5 layers, as i'm printing at a layer height of 0.2mm for this model). I have the top and bottom layers printing at 15mm/s and the walls are at 20mm/s. The infill density is at 40% grid style.

I have calibrated my flow rate for this setup and its currently sitting at 86% with a filament gauge of 1.72mm. I calibrated my flow rate by printing the outside/shell of a cube at 0.4mm line width and 0.4mm nozzle, and kept printing until the walls measured 0.4mm. So, although this flow rate seems low, its correct for this test. Is there anything else in terms of settings that I could change to increase the flow rate back up?

I'm using Cura for the GCode generation.

Is not advisable to use the same width of the nozzle on the extrusion, usually for a 0.4 nozzle your extrusion nerds to be at 0.48 for best results, I have gone down to 0.45 without issues but once you go lower than that you might face perimeter bonding issues.

Now 86 percent in a 1.72 mm filament is telling me that your steps per mm are way off, it should be over 100. You need to calibrate your extruder so you don't need to lower your flow rate to compensate for over extrusion.

Right okay, i should bring my line width back up to around 0.48. So you seem to think that my extruder steps per mm is incorrect? I set it up so when i press extrude 10mm, it does pretty much gave me 10mm movement on the filament before it enters the extruder. I could check that measurement for the steps per mm on the extruder in the morning. I need to get some figures for the MK6 (i think its mk6) direct drive bit/gear.

is it me or do those holes look not quite round? I'd check the belt tension as well as the grub screws for the pulleys.

I have a problem with pressure dropping in the nozzel with cura for screw holes. I have my retraction settings set to where it should retract, but instead just preforms a travel move and oozes out extra filament, even at low temperatures. This causes it to under extrude around interior features. But as ggherbaz said, your multiplier should be over 100% with your measured filament diameter, meaning either your steps are off some where or you are consistently under extruding.

Thanks guys, at the moment i'm battling with major under extrusion. I have my steps set to what they should be, using a mk7 drive gear directly so my steps should be 96.45 but when i extrude 50mm, only 30mm is going through.

Check your micro stepping, sounds like it's half of what it needs to be, don't know what board you have but you might be missing a jumper under the driver or isn't set correctly.

I'm using a Duet Wifi, its set to 1/16 micro stepping in the config file, maybe something else is activated, i may need to find something else out regarding that. Its not really moving the amount being feed through so i would imagine you seem right. But it doesn't seem to be effecting the other steppers. Maybe it is and i'm not noticing though. The duet wifi is running a fork of the Reprap Firmware.

I was having an issue where the drive was slipping on the filament and this was the reason that when I extruded 50mm, i was only getting 35. With an MK7 drive the steps should be set at 96, but i now have mine set to around 113, and when i extrude 50mm, i get around 49mm. So although the steps are not the figure that i would expect, i am getting the right amount flowing through when asked.

So i went for a little test of some cylinders, inside hole of 3.5, and a wall of 1.8mm and a total diameter of 7.1mm. The actual print came out with 2.9mm center hole, 2.4mm wall and an overall size of 7.43.

The flow rate is set to 100%, the correct filament thickness is used of 1.72mm. Do you think that i could now be over extruding as the overall size if .3/.4mm too large, i would normally expect a 0.1/0.2mm tolerance but not a 0.4. Do you advise that i continue calibrating the steps per mm for the extruder to get it close as i can for the correct sizes in my little test or do i now use the flow setting to dial it in?

Before changing anything, do a single line extrusion command, do it several times with different multipliers from .9 to 1.0 as you mentioning having a 0.4 nozzle do one with 0.45 and the multipliers from .9 to 1.0 in .02 increments, then do another one with 0.48 find which multiplier gives you the right measurement and if for both settings is the same. Once you find the correct numbers print again the test cube, if your measurements comes out wrong your X and Y steps might be at fault.

Regarding of the extruder steps per mm difference don't worry too much, machining have tolerances and your extruder gear might be a little bigger in diameter and the 20 or 30 steps more just account for that, the important part is that it moves the exact amount you are asking. My kids printer (Prusa I3 generic) Z steps are 4040 instead of the theoretical 4000.

I may have found the biggest issue. I think that the filament is not being fed through properly, sometimes slipping on the filament and what not. I have also found that if i crank the temp up to 220, the correct amount is extruded when i use a multiplier of 1.05 (105% flow). Now, i'm going to leave it set at this steps per mm as when i request 50mm, it extrudes 50mm. Its just not enough when printing an actual part and is coming out under extruded, but the amount fed when asked is correct.

At some point i may decide to change the steps per mm instead of the flow, but i will do that with a quality filament as the stuff i'm using at the moment is varying from 1.75 to 1.7mm. Pretty shocking really. I have ordered some MK8 drives, these have a smaller effective diameter and therefor allow more torque to be produced on the filament. Essentially anyway. However, by using a mk8 over a 7, my current extruder setup won't work, so i'm having to reprint a extruder setup for the new drive size. This could prove to be a challenge with the current issues i have.

I'm now trying to perform your test method. 0.9 to 1.0 multiplier in steps of 0.02 increments, that is a total of 50 lines per test, am i understanding this correctly?

I don't know what type of extruder you have,but my "wade's type" with indirect gearing gave me all sorts of feeding problems.
I found that the grub screw holding the small gear on the stepper shaft was not sitting on the flat of the shaft.
This caused it to slip at odd times,leading to the inexplicable feeding troubles.
I put washers on the main gear till i was able to move both gears out far enough for the grub screw to grip the flat on the shaft.
I'm not saying that this is your problem,just giving you something else to check.

I appreciate all the feedback possible, so thank you!

I have checked all of that, i now have a new drive gear, a MK8 drive. I have changed the steps per mm to the correct setting, when i extrude 100mm i get around 99.5mm now. Super happy at that. However, i print a test cube (just perimeters) and get 1.08mm walls instead of the 0.96 expected at 100% flow rate. If i bring the flow rate down to 90%, i then get 0.99/0.98mm walls. The width of the cube is perfect within 0.05mm when at 90% feed rate. So at 90% flow rate, i'm getting 0.03-0.06 too thick walls (over 2 perimeters) and the width of the cube to be within 0.05mm.

At the moment, my x, y and z figures are those given by the prusa calculator, belts and lead screws. I'm thinking that they may be slightly off like suggested by ggherbaz above. I'm a little confused which i should calibrate first though, because surely they are effecting each other and therefore if one is out, it will knock the other out in terms of the x and y effecting the placement of the walls in the perimeter test. Because they both effect each other directly, i'm confused to which i should sort out first, or if they need to be tweaked together....

Not as any as 50, with an extrusion set at 0.45 you do 0.9, 0.92, 0.94, 0.96, 0.98 and 1.0 then you repeat it with a 0.48 extrusion. Now if you have a good vernier caliper you can do line extrusion in both X and Y direction and measure them to find which one is giving you an error, so you can for example command a line extrusion from X20 to X80 and measure then repeat from Y20 to Y80 and measure, if both give you same measurements then just adjust both X and Y to that steps per mm, if both are different then adjust both to the shortest adjustment needed and use the slicer settings to compensate for the other one (you can use different X and Y steps per mm, but is not advisable).

Printing a test cube without having your extrusion lines correct is a waist of material and a nightmare for calibration because you are assuming that your feedrate, your extrusion width and multipliers are perfect and 99% of the times are not.

Also doing single line extrusion allows you to adjust the jerk and acceleration for X and Y, too much speed and your stepper would not stop where it needs to, too much jerk and the extrusion will not be even.

I have it printing a lot better now. The only thing that i have a little issue with is it wasn't quiet perfect on the Z axis, so i changed the steps from 1600, to 1614 and now a 50mm height print design is 49.97mm. I/m happy with that.

Would you advise any other way to change that setting rather than the steps per mm or?

Well Z is probably the easiest to calibrate, the main thing is to pick a layer height that matches the size of the part to print and that your first layer is set to 100% height. For me a +- 0.05 margin is acceptable.

When printing with a 0.2mm layer height, my first layer is 0.25mm to get it to stick to the bed.

Don't understand, it is always recommended to go lower for better adhesion, a 0.2 layer will have better adhesion than 0.25, you usually go lower or with higher extrusion rate but not higher layer specially not with a 0.4 nozzle.

Everywhere I look, it advises more plastic is better for first layer. I think our points are the same and the wording is throwing us out. Basically, more plastic is better for more heat and a slower cooldown for the plastic on the first layer. Lower layer height and more extrusion would also work i suppose. Just an extract from Slic3r: [manual.slic3r.org]