backpack design challenge

Show off your FreeCAD projects here!
Forum rules
Be nice to others! Respect the FreeCAD code of conduct!
User avatar
looo
Veteran
Posts: 3941
Joined: Mon Nov 11, 2013 5:29 pm

Re: backpack design challenge

Post by looo »

hammax wrote:… some days ago I saw some longdistance-paragliders in TV with something like an inflatable trailingedge
in/on their bagpacks, I suppose to reduce leeward drag.

Hi hammax, thanks for your input. I like the "eggy"-design. This will give some extra-space on top for putting the helmet in. Still I would like to integrate the straps into the design.

The harness is another topic. But you are right, adding some inflatable tail will decrease the drag [1]. In our case drag is not very important. It's more about lightness [2]. But the backpack is only for carrying the glider...

[1] https://flybubble.com/media/catalog/pro ... ir_004.jpg
[2] http://i.ebayimg.com/images/i/152032123 ... -l1000.jpg
triplus
Veteran
Posts: 9471
Joined: Mon Dec 12, 2011 4:45 pm

Re: backpack design challenge

Post by triplus »

looo wrote: Thu Aug 16, 2018 11:41 am To make tools usable efforts have to be merged and compromises have to be found. But the step from getting a workflow useable for my own tasks to getting something ready to be as generic as possible and acceptable from the average user is quite big...
I am sure that when there will be some FreeCAD surface related tools you will find easy to use and the experience to be streamlined. For others to accept it just like you will for your tasks. But there should be some common goal discussions going on. On what we would actually like to achieve. And once that discussions will mature it could turn out the solutions won't end up being all that hard to code. Now everybody is trying to tackle this area alone and on some i guess higher level. Like for example trying to provide tools to build the whole products in one go. But that is just too specialized and limited approach. Basic spline/surface manipulation. That is what it is needed and that is how (big) things end up getting done.
Another problem I see with the interaction nodes them self: If we add them to FreeCAD master, we will somehow depend on pivy... As we have seen support for coin and pivy won't last for ever. So I don't know how much sense it makes to add additional stuff which will be very hard to port to something like vulkan-osg. I guess porting the sketcher interaction-nodes will be already difficult enough, so nobody want to do this... In my mind we should abandon the idea of ever switching from opengl to vulkan!!!
What will likely happen. In the post Coin3D viewport era. That FreeCAD likely won't be compatible with todays FreeCAD anyway. But waiting for that to happen first doesn't make much sense.
freecad-heini-1
Veteran
Posts: 7788
Joined: Tue Jan 07, 2014 11:10 am
Contact:

Re: backpack design challenge

Post by freecad-heini-1 »

looo wrote: Sat Aug 18, 2018 4:57 am
freecad-heini-1 wrote: Like the Moi3D helmet.
I am not sure if curve-design alone will be enough. Currently I am manipulating the poles of the surfaces directly. But I guess with curves it's possible to create a more parametric workflow

This is how the current design is made:
The slow fps is coming from the capturing tool. But the computation is quite fast. The nurbs representation is computed in python, but the heavy part is done as a sparse-matrix (scipy) product, which is quite fast.
interaction.gif
Whow looo, very nice. 3D curves and a tool to manipulate the poles and the curve and a tool like you showed in your animgif to manipulate the surface would be very nice to have in Freecad.
Is it possible to integrate such features in an existing workbench like Surface or Curves?
triplus
Veteran
Posts: 9471
Joined: Mon Dec 12, 2011 4:45 pm

Re: backpack design challenge

Post by triplus »

I remembered a discussion about the spline editor:

https://forum.freecadweb.org/viewtopic. ... 30#p152518

Once such low level control for splines was made available it was easy to get such results:

https://forum.freecadweb.org/viewtopic.php?f=24&t=19772

P.S. If instead developer would focus on how to provide a tool to create sunglasses or backpacks or cars ... Likely until today i still wouldn't have modeled sunglasses in FreeCAD. ;)

Therefore back to basics i guess. Creating splines, surfaces and manipulating them in some straightforward and seamless ways. As for the rest. I guess the end users will figure that part out!
Jee-Bee
Veteran
Posts: 2566
Joined: Tue Jun 16, 2015 10:32 am
Location: Netherlands

Re: backpack design challenge

Post by Jee-Bee »

https://www.youtube.com/watch?v=J0Fy-V0GfmY

Here a video of rhino they have some nice surface properties... Don't know if FC tools are that strong... but nice as idea (for developers and ).
Last edited by Jee-Bee on Sun Aug 19, 2018 11:06 am, edited 1 time in total.
User avatar
Chris_G
Veteran
Posts: 2579
Joined: Tue Dec 31, 2013 4:10 pm
Location: France
Contact:

Re: backpack design challenge

Post by Chris_G »

Nice project, looo !

triplus wrote: Sat Aug 18, 2018 10:46 am
looo wrote: Thu Aug 16, 2018 11:41 am To make tools usable efforts have to be merged and compromises have to be found. But the step from getting a workflow useable for my own tasks to getting something ready to be as generic as possible and acceptable from the average user is quite big...
I am sure that when there will be some FreeCAD surface related tools you will find easy to use and the experience to be streamlined. For others to accept it just like you will for your tasks. But there should be some common goal discussions going on. On what we would actually like to achieve. And once that discussions will mature it could turn out the solutions won't end up being all that hard to code. Now everybody is trying to tackle this area alone and on some i guess higher level. Like for example trying to provide tools to build the whole products in one go. But that is just too specialized and limited approach. Basic spline/surface manipulation. That is what it is needed and that is how (big) things end up getting done.
...
I would be happy to help, if I can.
But, as far as I am concerned, I am not able to lead such a project.
I am not good enough to start from a blank sheet, and build the foundations of a workbench.
I can only contribute to a project that is already existing.
My CurvesWB is just a sandbox of experiments. And I am just waiting for a "serious" WB project to contribute with what I experimented.
Of course, there is the Surface WB, but right now, it is pure C++ :(
triplus
Veteran
Posts: 9471
Joined: Mon Dec 12, 2011 4:45 pm

Re: backpack design challenge

Post by triplus »

Chris_G wrote: Sun Aug 19, 2018 10:00 am I would be happy to help, if I can.
But, as far as I am concerned, I am not able to lead such a project.
I am not good enough to start from a blank sheet, and build the foundations of a workbench.
I can only contribute to a project that is already existing.
My CurvesWB is just a sandbox of experiments. And I am just waiting for a "serious" WB project to contribute with what I experimented.
Of course, there is the Surface WB, but right now, it is pure C++ :(
I guess as all of this was done in Python therefore ATM Python likely is the preferred choice. At least among the developers pushing the FreeCAD Surface capabilities forward. In the past lets say 2 development cycles.

Image

phpBB [video]


phpBB [video]


Image

That is if all this would be packed into a polished and streamlined workflow an average FreeCAD end user could use. I am sure backpacks and things like that would come after. ;)
User avatar
looo
Veteran
Posts: 3941
Joined: Mon Nov 11, 2013 5:29 pm

Re: backpack design challenge

Post by looo »

triplus wrote:I guess as all of this was done in Python therefore ATM Python likely is the preferred choice. At least among the developers pushing the FreeCAD Surface capabilities forward. In the past lets say 2 development cycles.
yes, python is for sure the preferred way to explore and learn new stuff. With things like numba, cython, numpy it's also possible to implement algorithms in a fast way. So I guess everything needed can be done with python.
triplus wrote:That is if all this would be packed into a polished and streamlined workflow an average FreeCAD end user could use. I am sure backpacks and things like that would come after.
We will see how things evolve. Different tasks need different tools. I don't think there is the fully parametric, generic nurbs editor mechanism... In my mind it's better to throw away the idea to make the nurbs manipulation parametric and concentrate on the tools to get the job done.... (tools like: adding additional poles at specific uv values, setting poles to a plane, setting poles to a line, snapping,... there are quite a few difficult tasks.

Also blender with subdivision surfaces could be a nice alternative...
User avatar
looo
Veteran
Posts: 3941
Joined: Mon Nov 11, 2013 5:29 pm

Re: backpack design challenge

Post by looo »

I would like to create a pocket on the side ofthe backpack. I guess this should be possible by drawing a uv-curve in the uv-space. Are there any experiences with mapping a bspline from uv-space to the surface?
User avatar
microelly2
Veteran
Posts: 4688
Joined: Tue Nov 12, 2013 4:06 pm
Contact:

Re: backpack design challenge

Post by microelly2 »

looo wrote: Tue Aug 14, 2018 1:17 pm Today I made some progress with the surface-creation. I guess what I really need is an interactive tool to manipulate the poles of a bspline surface...

So the workflow will be like this:
- create a 2d-face from sketcher-bsplines
- convert the face to nurbs
- get poles, weights, knots and compute the influence matrix (for given uv-points)
- drag poles and compute a quadmesh (interactive design)
- once result is satisfying get back a shape

no idea if this can be done fully parametric.
Interesting task.
I use Bezier curves based on sketches
the face is calculated as a product of the border beziers (something like a gordon surface)
everything is parametric - sketches and datum planes
phpBB [video]
Post Reply