Jump to content
AutoDesSys Forums

Armin Stroß-Radschinski

  • Content count

  • Joined

  • Last visited

  • Days Won


Armin Stroß-Radschinski last won the day on December 11 2021

Armin Stroß-Radschinski had the most liked content!

About Armin Stroß-Radschinski

  • Rank

Profile Information

  • Gender
    Not Telling
  • Location
    Troisdorf, Germany
  • Interests
    Industrial Design, Programming, Generative Algorithms, Python, Architecture, Manufacturing

Recent Profile Visitors

178 profile views
  1. Armin Stroß-Radschinski

    Curvature continuous rounding in FormZ

    @bbuxton💪 Excellent wrapup, needs to be collected elsewhere (my archive is not enough) It was worth to kick this topic off.
  2. Armin Stroß-Radschinski

    Curvature continuous rounding in FormZ

    Good point your remark on usability of the T1 and T2 sliders. If I find the time I would like to address this issue from python.
  3. Armin Stroß-Radschinski

    Curvature continuous rounding in FormZ

    I indeed took the way to extend the curvature into the flat surface and used the G4 level. As you said it is a bit cumbersome but worth the effort if you aim at real world production with CAD/CAM. I had to adjust the control points by hand. My usecase is a tabletop 2D contour shape. The usual way in a prototyping workshop for furniture and even when eroding steel tools with copper master in tooling to handsand out the smooth blend between a "technical" radius and the surface. In traditional contour shape ruler copy milling and havig people with these skills around, this was an absolute acceptable solution to invest this time. When you want to produce a production ready and reproduceable shape directly out of machining or create better realistic renderings (parts looking as if these real world skills were applied) it is worth to play around with these adjustments. Every time when your guts say: This is an artificial image, one reason can be the missing of cool looking smooth blends. One issue with 4th Grade in FormZ is the missing of adjustment of the reference radius. In FormZ you have to create a real radius helpline first, adjust the actual rounding shape and then scale it including extension into the planes to meet the expected radius by approximation. This is acceptable for square 2D base shapes when the same corner can be repeated 4 times after creation and you just need to optimize one part. I agree that smoothing out full 3D shapes and lots of them is a different job. In this case I expect a complete different approach to tackle the challenge. The red curve shows the smooth blend, the black one the mathematical radius (this is an inbetween step and not the final shape)
  4. Armin Stroß-Radschinski

    Curvature continuous rounding in FormZ

    I finally managed to find the rounding tool opions in FormZ that I had in mind. I is the N-rounding tool in the nurbz tool palette. There you can select the grades of rounding I expected. I now start to play around an come back here later. One issue is that the origin object was no Nurbz object and therefore some conversion may be needed.
  5. Armin Stroß-Radschinski

    Curvature continuous rounding in FormZ

    OK seems that this was saved as an repeated action or kind of macro. The link is still missing ;-). Also the steps in Sketch are a kind of "dull" . Saving a rectangle as SVG will not be enough.
  6. Armin Stroß-Radschinski

    Curvature continuous rounding in FormZ

    @vva Can you provide a link to the Affinity forum? This option seems not to be in the current Affinity Designer Desktop UI, but Figma is mentioned. Can it be possible that the Screenshot is from another Software Application? In this post Sketch is cited here https://forum.affinity.serif.com/index.php?/topic/144496-ability-to-add-smooth-corners-to-rectangles/ they use the term squircle .
  7. Armin Stroß-Radschinski

    Curvature continuous rounding in FormZ

    @vva Thank you for this hint. For this time it would be sufficient.
  8. Armin Stroß-Radschinski

    Curvature continuous rounding in FormZ

    My use case is a prominent smoothed out edge rounding with a large 75mm radius at the edges of a table. I can manage to meet my needs wit a sequence of manual operations, but that is not the point here! As far as remember there was maybe a way to create Class-A curvature continuous rounding in older versions of FormZ (I use it really for a long time starting with 3.x or so). I am not sure if this was integrated in stitch rounding (makes more sense for Nurbs) or simple and controlled rounding. The purpose for me is actually to smoothen out the join of the radius shape and planar surfaces to avoid the usual visible ugly "frontier" when the surface goes from planar to a constant bending. I the real world this is mostly smoothed out by professional modelmakers. In a complex CADCAM workflow it is not the best idea to do this manually both CAD and after milling. I searched the FormZ docs, forums (old too) and the web but did not find any hints related to FormZ, maybe looking for the wrong term. Class A Rounding Curvature continuous rounding G2 Rounding I found the facetting schemas in the FormZ docs and rounding tools, but this seems it only affect visualisation. When playing with the options and exporting the rounded shapes as DXF and check in other software, I get just a plain stupid radius. Maybe I need to stop to "play". Any hints? You find some content on "Curvature continuous rounding" in the context of SolidWorks, Catia and Alias, but the findings are rare.
  9. Armin Stroß-Radschinski

    Exporting from Affinity Designer to Form Z

    AFdesigner SVG as splines to Postscript Level 2 EPS conversion via Inkscape renamed as AI. Beneath the QCAD approach (nice hint) you can use the free Inkscape vector Illustration program as missing link I am not a lover of the GUI for creative work, but it allows to be run for file conversion via commandline as well (like QCAD). Install Inkscape from https://inkscape.org for your platform (Mac, Win) Export from afdesigner das SVG Open SVG in Inkscape (currently 1.1.1) Export (save) as Postscript Level 2 EPS and rename the resulting file suffix from .eps to .ai FormZ 9.2 will import it as splines! IMPORTANT: The import has a different orientation than expected from AI. Not tested that much with all kinds of primitives and other intermediate formats (DXF is not working well). Note: The Postscript Level 2 EPS Export of AFDesigner does not work with the FormZ AI 8 import filter. Benefits of an Inkscape install: Commandline automation by using the inkscape command Conversion can be triggered from FormZ Python (not testet yet) Conversion of the vector export of FormZ as .emf files. Suggested to import FormZ 2D vector data in Microsoft Office Apps (Powerpoint works fine to present CAD Details for others without pixels this way) Path modification features Rich Plugin ecosystem, because you can write your own extensions in Python or modify others yourself look for the Plugins installed and online : https://inkscape.org/gallery/=extension/ Examples (useful in Export Automation) Adding Knots for older CAM Systems: Menu: Extensions -> Modify Path -> Add Knots (default Extension), could be done in FormZ, but nicer as postproduction. Tools for creating gearwheel shapes, boxes etc (to go directly to 3D try OpenSCAD https://openscad.org/ Maybe people at AutoDesSys can reuse their OpenSource Code (SVG is the native file format of Inkscape) to add SVG Import to FormZ. Missing in AFDesigner EPS or SVG Export: No Layers from AFDesigner can be preserved (Generally supported via AI Import) because Layers are lost during AFDesigner SVG Export. Use a top Level Group inside of every AFDesigner Layer to seperate in FormZ later (Inkscape preserves gouping) IMPORTANT: 2D illustration programms think of layers different than CAD. In AI and AF layers are actually top level groups with special flags. This is due to the importance of 2D object stacking to keep proper hierarchical display of objects. Groups cannot span multiple layers like in most CAD applications You can use Advanced SVG Export in Inkscape to cluster elements by attributes! and start from there.
  10. Armin Stroß-Radschinski

    formZ 9

    I am very happy to see Python upcoming in formZ. To leverage all these concerns: I am an Industrial Designer and 20 years formZ user. I am also a "semiprofessional" workflow maniac that is lazy and tries to be smart. I dived into Python and the whole Open Source ecosystem in 2002 during the use of Plone, a very secure object oriented Web CMS with over 15 years of success. Hope the following lines are not to boring and open the mind of even "no nerd" users to empbrace python. I have done a lot of communication work around Python and Plone. One is the official Python Brochure by the Python Software Foundation that can be downloaded here. http://brochure.getpython.info/media/releases/psf-python-brochure-vol.-i-final-download.pdf We researched a lot of stories around the usage of Python in the Industry (Including Industrial Light and Magic, Houdini, Blender). Some stories did not make it into the brochure (including the one from Disney) due to some legal permissions missing. We know that companies like Electronic Arts, McNeel (Rhino), Maxon (Cinema4D), GOM (3D-Scanning) and much more use Python as a glue language to connect everything to productive environments. You find nearly unlimited solutions to solve almost every problem you can imagine. Microsoft offers more than 10h of training videos on Python and you find excellent books and trainings everywhere. You can bridge from Python to almost every database, spreadsheeds and even scrape webservers. You can process xml and JSON. You can easily document your work using clear and modular programming patterns if you know that, but also save your life with simple few liners working with the OS clipboard. Python is very succesful especially with non full time programmers like scientists and technicians. The easy readable code and the interpreted execution as script boosts your productivity even if you just read and modify existing code. Plone is still using Python 2.7.x but on the way to Python 3.x. Why they are so late? Lazyness? Dumbness? Are there similarities to formZ? The Python Standard Library has a very big and proven ecosystem around it. You can do almost everything in either Web Programming, Embedded Controlling, Scientific Computing (including AI and Big Data) in a more efficient way than using Java or C and with lesser code. There are reasons to use C if the code needs to be executed fast very often. But if you just need get things done several times, short development time beats minor execution time improvements. This is one story behind the success of Python. So why Python 2.7 (e.g. in MacOSX)? (Background why 2.7 is still in use and is supported) There are some rare critical Libraries that were only ported to Python 3 recently and they are not interpreted code (which runs almost fine in Python 3 with some easy tweaks) but compiled into C to run fast) So if you do not have the power to port the code to Python 3 compatibility yourself (what you could do theoretically with Open Source) you have to wait for the maintainer to do it. I fact Python 2.7 is supported until 2020 and everybody knows this. It works as reliable (or better) than Python 3. You can already write code that runs fine under both major releases Python 2 and 3 and you can "Test" that! There are converters to do that automatically as well. Unicode is already supported in Python 2.7 but is treated different. Since formZ scripting is not exposed to the web (even it could do so!) there are no security concerns / reasons against to extend that deadline for some time if you know what you do. Are PyQT etc relevant? If formZ offers nice access to manipulate the native GUI the productivity will be much better because you are in the same object namespace to query the state of everything in formZ directly. If necessary you can use PyQT or WXPython if you have strong needs. There will be no limit as far as I can guess after what I read here. Get a grip now? The only bad story is, that every 3 mails I sended to the ADS crew in the last year to grant access to the SDK were not answered. I suggested the use of Python or/and to implement a bridge from the old SDK myself. Inspiration? – I think they catched it themselves before... but still a bit poor... I will start to work with Python in formZ it as soon I get a grip on a beta. This was the last reason to switch to Rhino, now gone.