Umpteenth Filament Shrinkage Calibration Models

Umpteenth Filament Shrinkage Calibration Models'

Why These Exist

This is probably the umpteenth set of filament shrinkage calibration models on the internet. I made my own anyway because I make most of my own stuff and wanted dedicated, repeatable models with longer measurement spans without wasting filament on solid calibration cubes. I also wanted them to fit my typical design language.

These models are designed to measure proportional shrinkage across the X, Y, and Z axes after an FDM/FFF print has cooled. This can be important when tuning filaments intended for dimensionally accurate parts and structural designs with critical dimensions.

Model Features:

• 50 / 100 mm nominal measurement spans across X, Y, & Z

• 8 x 8 mm beam cross-section on the 50 mm model

• 12 x 12 mm beam cross-section on the 100 mm model

• X, Y, & Z labels modeled into the corresponding end faces for clarity

• Flat opposing measurement surfaces

• Chamfered outside edges to prevent corner artifacts from interfering with caliper placement

• Material efficient design, uses less filament & print time than a similarly sized cube

• Hash marks on the Z axis for more in-depth Z axis tuning

What This Test Is For:

These models are intended to determine proportional material shrinkage after the filament and printer profile have already been dialed in.

For accurate results, your filament should be dry and your temperature, extrusion multiplier, and pressure advance should already be tuned!

Check out my full filament tuning guide: 

Shrinkage compensation is especially useful for ASA, ABS, Nylon, PC, and other materials that can produce dimensionally undersized parts after cooling.

NOTE: This test is not intended to calibrate printer steps-per-millimeter (rotation distance), correct poor printer geometry or skew, or replace hole and contour compensation.

Included Files:

• 50 mm Umpteenth Shrinkage STL: Faster to print and uses less material.

• 100 mm Umpteenth Shrinkage STL (Recommended): Provides a longer measurement span for greater resolution and makes small dimensional changes easier to identify.

• CAD Files: Cuz I dislike when they aren't included.

Recommended Print Settings:

• .2mm layer height

• 3 perimeters

• 15% Adaptive Cubic infill

• No supports

• Print the model in its supplied orientation

Adaptive Cubic is recommended for low material usage. Infill can slightly affect the final dimensions, so using the same infill you intend to use for your actual parts may provide a more representative result.

For the most representative results, use the same nozzle, layer height, walls, infill, temperatures, cooling, enclosure conditions, and other settings you normally use with that filament. These models were designed around a 0.20 mm layer height. If you use another layer height, check the sliced layer count and final Z height in the preview before interpreting your measurements. Changing any of these settings can affect the final result.

How To Use

1. Set filament shrinkage compensation to neutral before printing.
   - OrcaSlicer: 100.00%

   - PrusaSlicer: 0.0%

2. Print either the 50 mm or 100 mm model at its original scale.

3. Allow the print to cool completely. Do not actively cool the model. Let it cool naturally.

4. Measure the complete outside span of the X, Y, & Z axes with calipers.

5. Calculate the required shrinkage compensation using the formulas in the supplied PDF.

6. Enter the results into your filament profile.

7. Print and measure the same model again to validate the updated profile.

8. Print it over & over while questioning your sanity because you never hit the exact numbers.

9. Skip step 8. At some point, accurate enough is accurate enough...

Use the complete 50 or 100 mm outside spans for these calculations. The beam cross-sections are secondary reference features and should not be used for the main shrinkage calculation.

The 100 mm model is the better choice for a more accurate result, but it requires more print time and material.

A Note About Z Compensation

Printed Z height is constrained by layer-height increments. A small calculated correction may not change the resulting print until it is large enough to add another layer.

Because of this, Z compensation should be treated as a practical approximation. Do not chase endless validation prints trying to correct a difference smaller than your selected layer height can reliably represent.

Support My Work

Please consider checking out my other social media accounts or supporting my work directly on Patreon!

http://www.MandicReally.com/links

https://mrly.link/patreon

Need Design Support? Check Out My Discord:

https://mrly.link/discord