yes but low-res meshes are not the point I think? Obviously you can get in trouble if you have a low-poly mesh and try to replicate a smooth surface. Also, since a tool can only move from one point to the next in a linear way you always have a (theoretical) resolution problem.
Do you know a CAM software which works with surfaces internally and not convert it to mesh to generate the toolpaths?
I have gotten "smooth(always depending on your constrains obviously) pieces out of mesh-based CAM software and would like to compare.
This is not the point. Of course a Nurbsmodel gets converted to a polygon for most purposes. But the ability for dynamic and curvature-based tessellation is one reason for Nurbs. However at least there are some implementations which skips mesh conversion and computing toolpaths directly. However I have limited knowledge here.
As my second sentence said and which is also part of my previous reply. Its about the dynamic and curvature based tessellation which is possible through representing a shape with parametric equations.
If you model a mesh you usually don’t model optimised for fabrication. For fabrication you want a very dense tessellation on strongly curved areas. If you get a dead model, you have no way in changing the density. You can increase or decrease the points, but you change the shape by doing so. You can only guess.
In reality you want to model the shape as light as possible, but if you model dense there is only limited space for change. Sometimes a manufacturer needs to change the shape because something is not working well. How can he do if he gets dumb data?
You may also want a different tessellation for FEM purposes, Analysis and Visualisation etc. So it absolutely makes sense to work with Nurbs.
Of course you can maintain a certain “parametric” when doing things in Grasshopper only. So this weakens the older arguments. But the real point is flexibility and contracting to a certain industrial standard.
I know of certain inhouse implementations of certain companies, but I cannot name these tools, nor I know if how far these are and if you can buy them on the market.
Edit: However I admit of having only limited knowledge, since most of this was learned with chats of people doing this kind of work. So I’m walking on “thin ice” here. Still all of them always required very clean Nurbsdata for further processing…
That’s a good question and I could be wrong here, but i thought SolidWorks CAM uses the surface data. Maybe it converts to meshes to create the toolpaths behind the scenes but I would be surprised if it does this. I would have thought the toolpaths for 3d surfaces would be generated from contours.
Everything else it does in CAM is feature based - holes and pockets, fillets etc. but of course it can derive the 2d geometry from each feature from the underlying 2d sketch to then create toolpaths.
No, I agree with you that meshes might not be ideal for some production processes. As mentioned above, I’ve only dealt with small, local manufactures. I only think harping on that point, when in fact production with inexact meshes is possible, is pointless.
What I don’t agree with though, is your point about a design being futile, if it can’t be produced, or translated into the physical world.
For me, “parametricism” is more of a tool, not a philosophy. Each tool brings new challenges that drive innovation, but that’s not news either. As far as I know, many people are working on this, in academia and the industry.
What is and what isn’t a “waste of time” is very subjective!
Sure, I absolutely agree. I just wanted to point out that production, starting from meshes is possible.
I’ve been to Vietnam, but on vacation. I haven’t visited any factories that produce for Nike, however I do voluntary work for the local branch of Amnesty International, and over the last years the brand name has popped up more than once in relation to human rights issues.
Sure, but that’s a common practice among multinationals that publicly and legally want to appear clean, but in fact simply turn a blind eye to issues like forced or child labour, or environmental issues linked to production, because their only care is profit.
To everybody that wants to sweat-talk these issues, I’d simply suggest to imagine a world where the roles were inverted, where you and even your children would be brutally exploited, just to satisfy the insatiable hunger of some people, and you’d have to bow to this system simply because you’d have no other choice.
I realise that my comment about child labour was probably inappropriate for this discussion or even forum, and I want to state that it wasn’t directed towards you, Michael, being a Nike employee.
What you have to understand about university programs, especially architecture and design, is that they need to pump out interesting projects that they can publish and thus attract new students with.
However, practically all of them surely have courses that are steered towards teaching technical execution and production. Most even research new fabrication techniques and processes, which is exactly what a university should be, a driver of innovation.