TPU95A vase mode with wider-than-normal extrusion settings - for a 0.7 mm pipe sleeve

I needed a 0.7 mm thick sleeve between a pipe and a bracket to prevent metal-on-metal damage on paint coverings.  TPU seemed like good choice, being strong and with good layer adhesion.

My research indicated that the 0.4 nozzle should be able to readily  print up to 0.8 mm extrusion width, perhaps even wider (see the video link below).

In PrusaSlicer spiral vase mode, achieving a wider, reliable wall requires balancing extrusion width against volumetric flow limit and speed. This ensures the printer doesn't outrun the melting capacity of the nozzle, which may cause TPU jams.   

The model

A thin-walled cylindrical pipe with diameter 34 mm and length 90 mm. Wall thickness had to be 0.7 mm. To be printed in vase mode with zero top and bottom layers.

Setting a wide line width

To get a wider wall from a 0.4 mm nozzle, requires overriding the default math. Go to Print Settings > Advanced. Find Extrusion width. Then change External perimeters to 0.7 (depending on how thick you want the vase). PrusaSlicer might warn you that it is too wide, but for vase mode, somewhat wider still works fine. 

The slicer also suggested using a brim for my pipe design (zero bottom layers), which I ignored, because TPU adhesion was found to be sufficient, and also because the vase mode printing is smooth and always staying on top of the previous layer - there is no risk of the nozzle bumping sideways into the part.

Controlling Speed via Volumetric Flow

TPU is limited by the volume of filament that can be pushed through the nozzle per second. If you set the linear speed to 60 mm/s, but the wall is 0.8 mm wide, it will likely exceed the flow limit and the filament will buckle.  The formula for volumetric flow speed is

Volumetric flow speed (mm^3/s) = Layer height (mm) X Line width (mm) X Linear speed (mm/s)

It seems that Maximum Volumetric Speed, (MVS) in the slicer settings sets an upper limit, limiting the linear speed to fit the above equation. Which means that the product of layer height, line width and linear speed must be less than the MVS limit. If the product exceeds the MVS, the linear speed will be reduced to stay within the MVS.

Set the MVS value in Filament Settings > Advanced, then Maximum Volumetric Speed, to a value of 2.5 to 3.5 mm3/s for TPU95A.

Next set the linear speed in Print Settings > Speed,   External perimeters to a fixed value (e.g., 25 mm/s). If the linear speed is too high, it will be limited to maintain MVS within the specified limit.

Vase mode settings

Print Settings > Layers and perimeters: Check Spiral vase.

Print Settings > Layers and perimeters: Set Bottom solid layers to 4, 5 of whatever.  For my use case for the sleeve, this was set to zero to print a pipe without a bottom. It worked well for the TPU because of the good bed adhesion (0.7 mm wide ring on a 34 mm diameter).

Filament Settings > Cooling: Normally, keep the fan low (0%-20%) to ensure the thick line fuses perfectly.  Fan setting for the first three layers was 0%. For this pipe print with the short layer print time (around 5 seconds), the upper fan speed was set to 50% to cool down sufficiently quickly.

Inspect the sliced model

For interest, look at the Feature Type color map in the G-code Preview after slicing. Change the "View" dropdown to Volumetric flow speed. If the whole vase is one solid color, the speed is consistent. If it changes colors, the volumetric limit is actively slowing you down in certain spots. 

For this particular model with thin layer height, very short layer time, and low linear speed, the MVS was very low, well below the set value.

Material

Nozzle temperature was 230 C for the first layer, increased to 240 C to support the faster flow rate. Bed temperature was 50 C. Printed with the door open.

The model printed repeatedly with no issue on the Core One Nextruder with (1) filament sensor disabled at the feed-in point,  and (2) idler screws screwed out by two turns. To load the material I removed the PTFE tube from the Nexruder and pushed the filament by hand to ensure it being gripped by the extruder (then pushed the PTFE tube back in).

The print is quite flexible but not really stretchable at 0.7 mm thickness. I could not tear the print by hard pulling with a pair of pliers.

The eSun TPU came on a cardboard roll.  I heard irregular soft 'clicking' sounds, as if the extruder was slipping. It was not, the sound came from the stiction and release between the cardboard roll and the spool holder.  The feed rate was so low that the filament spool was not continuously rolling. A case for a lower friction spool mount can be made here.

Print and application

Printed in eSun TPU95A. The filament was dried, straight from the sealed box, for six hours at 55 C. After opening it was stored in a vacuum bag with about 100 cm^3 silica desiccant. During printing it was open in a room kept at 40% RH at around 23 C. There was negligible stringing in my other TPU prints under these conditions.

The pipe sleeve was cut open with a pair of scissors, slid over the pipe, glued with a small amount silicon sealant to prevent it moving and the two ends tied down with  zip ties to keep it in place.

The limit?

You may find this video on printing vase mode with extremely wide extrusion widths, by @LostInTech3D, entertaining.