PySLM Version 0.3

PySLM version 0.3 was released last week to coincide with a large number requests from users of the library. The release is consists of many updates, fixes and examples accumulated across the last 6 months since last summer. Additional work was done to refine the release of the sister library libSLM and resolve some bugs that couldn’t be determined until exporting the machine files and testing on the machine, with acknowledgement of support from researchers who have got in touch. Many thanks for their assistance on this development journey.

The release notes can be found on github.

The original release was scheduled to include support generation, but this has been postponed for v0.4 to ensure that there was a underlying stable release of PySLM as reference to ensure users can utilise the library without waiting for the support structure element to stabilise.

A summary of notable features amongst fixes are as follows:

  • Added class geometry.utils.ModelValidator with functions to validates the inputs (models, layers) are consistent and coherent prior to exporting to machine build files using libSLM.
  • Added an alternative method BaseHatcher.clipContourLines for clipping a list of contour scan vectors. See the previous post about generating custom sinusoidal scan strategy using this method.
  • Added method hatching.simplifyBoundaries to simplify boundaries using Sci-kit Image method based on Douglas-Peucker algorithm.
  • Added a methodvisualise.visualiseOverhang to visualise overhangs – in preparation for support structure analysis
  • Added function argument index to visualise.plot in order to visualise the scan vector parameters (e.g. length, laser parameters, build style id)

Any requests for additional features or other improvements feel free to get in touch.

3 Replies to “PySLM Version 0.3”

  1. Hi Dr Parry, great script you have here! Is there anyway to extract the coordinates of the scan paths ? I’m trying to create a gcode script based on PySLM. Any assistance from you is greatly appreciated.

    1. Hi Mike,

      Yes, the scan paths generated are just a sequence of connected coordinates, so it is possible to adapt this to GCode. The code is fairly flexible to accommodate this aspect.

      I’ll send an email over to discuss further,

      Luke

Leave a Reply

Your email address will not be published. Required fields are marked *