Difference In Printed Cubes

[desman]

The cube on the left is my calibration test for pet+. The only thing I did different for the cube on the right is add three layers to the top, Three layers to the bottom with two perimeter shells. The infill is at 80%. All four corners on the right cube have that extra plastic. These were sliced using Simplify3D. The calibration cube looks fantastic. Any thoughts on what is causing the anomaly.

[desman]

Attached is the cubes factory file.

[Tim]

I think your extrusion multiplier is set too high. For reasons I have never been able to track down, the nominal/average value for the extruder multiplier tends to sit somewhere around 0.9. What is probably happening is that you're extruding too much filament, but when there is only a thin wall to print, the excess can expand as it needs to. But when you have 80% infill, there's only one way to expand, and that's outward.

One purpose of thin-wall calibration objects is to measure the wall thickness and compare to the model. If you printed with zero infill, and your extrusion width is manually set to 0.4mm, did you get 0.4mm (per perimeter) wall width? If it's more than that, that would be a clear sign that you're over-extruding. But you'll also want to check your filament diameter, which if not tightly controlled in manufacturing, can be another source of over- or under-extrusion.

[desman]

I started with the file provided by Madesolid they did not have it set to .40mm manual they had it set to auto. I changed it to .40mm manual and I thought I did a calibration after that. Maybe I did not and just ran the solid cube. I will run another calibration print and see what the wall thickness is using these settings.

[Bratag]

I started with the file provided by Madesolid they did not have it set to .40mm manual they had it set to auto. I changed it to .40mm manual and I thought I did a calibration after that. Maybe I did not and just ran the solid cube. I will run another calibration print and see what the wall thickness is using these settings.

Tim is referring to the extrusion multiplier not the width. Its the number directly below the width in S3D

[Tim]

Tim is referring to the extrusion multiplier not the width. Its the number directly below the width in S3D

Well, I did also mention the extrusion width, in reference to the calibration object.

Theoretically, if you print a single wall, and S3D is set to 0.4mm manual extrusion width, then the wall should be about 0.4mm thick; and if you have it set to auto, it should be reading 0.42mm, and should give you walls that are about 0.42mm thick. And if that's actually a statistically significant difference for you, then you're getting better prints than I am, and/or have a better set of calipers. I see more variation than that from point to point on any calibration print I do.

[desman]

My understanding was you use the extrusion multiplier to calibrate the actual width of the print of a single wall to the .40mm manual width in S3D. You run the print and divide the S3D width by the actual width to get the proper multiplier.

[Jules]

You run the print and divide the S3D width by the actual width to get the proper multiplier.

Close....you are creating a multiplier for the multiplier that is already showing in S3D.

The default multiplier for S3d is 0.90, so you would multiply 0.90 times "0.40/actual width" for a specified 0.40 mm thick thread.

Say your actual width was 0.43 mm. So the calculated adjustment is 0.40/0.43 = 0.93

But 0.93 is not your extrusion multiplier. The correct extrusion multiplier is 0.93 times 0.90, (which was what you used to extrude), so 0.93*0.90= 0.84

Next time you adjust the multiplier, you would adjust the 0.84.

[jsc]

Also, depending on your cooling settings, low infill layers may print at significantly slower speeds than filled layers. So when you turn up the infill, the time per layer goes up and the cooling slowdown goes away. Printing at higher speeds makes sharp corners more bumpy.

[Tim]

You run the print and divide the S3D width by the actual width to get the proper multiplier.

What Jules said.

But to get the 1.1 value you have as your extruder multiplier, even if you started with a default of 1.0 from the MadeSolid example file instead of the usual 0.9, you would have had to have measured a wall thickness of 0.36 (i.e., 0.4 / 0.36 = 1.1, approximately). Given that hot filament is squishy and tends to flatten out into an oval as it is laid down, it would be very difficult to obtain a single layer thickness of 0.36mm from a 0.35mm nozzle opening. Are you sure you didn't invert that equation?

(edit: I really shouldn't say that it's so very difficult to get single layer thickness less than the diameter of the nozzle, but that usually happens due to blockage in the nozzle or badly controlled filament diameter, and it's usually easy to tell because the process no longer has any real control over the amount being extruded, and the result looks very messy.)