escalator

Also curious how to go about these sort of standard pieces in general.

Where can I find good ones and how to best manipulate them?

>>553699
is it gonna be animated?

>>553702
not necessarily, but would be cool to have the option.

make the first step then array to make the others.

the other parts you can model regularly. the panels on the side can be made with curves

>>553699
I would handle this by breaking apart the stairs into components. For example: each step can be made individual, and the rails/walls can be copies of each other since those parts are equal/ symmetrical. I would use a texture to make the details for the panels and steps.

So in the end I'm making two pieces: the rail and a single step. Then copy paste.

There are better ways of going about it, but this is the route of least resistance.

Argument to your needs. Break down the geometry, and piece it together.

>>553704
>>553723
Modeling one individual escalator isn't a problem. It becomes a problem when you have to do it 8 times because you're modeling a building from the 80s that is being remodeled and every floor has a different height.
Ever since I got into Grasshopper I became very lazy so I'm looking for a more parametric solution.

>>553739

Model it out of modifiers, then you only need to model small pieces that follow a Curve and are repeated via an Array, then you can adjust the features on the fly.

Or if you can't do that then like the other guy said, keep it all floating geometry so you can adjust lengths of things to match the real world specifications

Honestly it sounds like you're a bit of a novice to 3d modeling and somehow conned people into hiring you for work you weren't qualified to do?

If you can't even do those basic things, then I'd hire someone off fiverr or some site like that to do it for you and then try to leave that job while you work on your core skills

Yeah, Houdini would be great for this. Search for few procedural Houdini modeling videos on Vimeo / Youtube, you'll see it's perfect for this kind of task

>>553765
I'm obviously still learning. I'm doing nurbs modeling with Rhino. I didn't con anybody tho.

Like I said, modeling and adjusting the thing itself is not a problem. It's just not as efficient compared to a parametric solution. I'm a novice to that.

I am currently using grasshopper to parametricize (is that a word?) large parts of geometry that is appearing in almost every project we do.
And when you can create entire facades in a matter of minutes by adjusting a few sliders, 'analog' modeling starts to feel very inefficient.

>>553768
Thanks. Any tutorial you could recommend in particular?

>>553739
Huh. As in it is all one piece? I'm not sure I get what you want to accomplish. If you have a reference or some picture you should post it.
I don't see why you can't make 7 copies of the original. Path of least resistance.

>>553789

This is just preview of one GameTutor tutorial, but you can quickly see in what way you can use Houdini:

https://www.youtube.com/watch?v=D4ECynoYuaU

Here's a simple one, just to see basic workflow:

https://www.youtube.com/watch?v=2VIutrsYK5M

>>553794
Not in one piece.
What I want is to pretty much specify start and endpoint of the escalator and have whatever program generate the geometry.

I built a script in grasshopper to generate a paneled facade. I just put in the point where it should start, the position of doors and windows. Then I have sliders to adjust panel size and amount, door and window sizes etc. Then I bake the geometry in the respective layers.

I also got a script for stairs and handrails, guess I can rework that into escalators.

I could model one and make 7 copies and slightly adjust for different height, but if I can make a script I won't have to model one ever again.

>>553818
Thanks a lot, gonna watch now

>>553819

okay quick rundown on one way to do it in grasshopper

>point on lower ffl, point on upper ffl of connecting floors
>line between
>measure distance and then divide the line into points based on your local requirements for escaltor risers and tread distance.
> use those points to build risers and tread curves off
>extrude out.
>place start and end geometry of escalator

Basically break up your modelling into a series of steps is the key to parametricising stuff. At the start do it in separate definitions and then link them up into one. will help in resolve any bugs your definition has.

Also im pretty sure escalators come in set sizes / at set angle increments. Go to a manufacturers website and see what documentation they have, could be useful.

>>553867
Thanks for the rundown. I will set out to create such a definition today.

Escalators do come in set angles. They don't seem to fit in this particular project tho. This might be because the building is a few decades old and standards may have changed, or because of inaccuracies in drafting of the plans.
As per my current workflow, I obviously don't model a building around the escalators. I model the basic geometry based on the sections and floor plans and add in details such as escalators at the end. So that's why standard escalators don't fit perfectly most of the time. For still shots this isn't an issue, you won't notice if an escalator doesn't perfectly meet the top floor, or you can adjust it for a particular rendering. But for real time visualisation it is a problem when you go up an escalator just to be greeted by a gap or a ledge.
This problem would be eliminated by having a parametric escalator that could fit itself to any length, angle and height while maintaining realistic handrail and stair proportions. As a whole, it would not conform to real world standards anymore, but it would be great for visualisation purposes.

>>554181
Now, I could simply 'cheat' and use a standard escalator and remove any gaps by adding geometry to the building, or manually adjust the escalator. And that would probably be the easiest solution for one project.
But looking ahead, a parametric solution would speed up subsequent projects and remove a lot of unnecessary finetuning.
I might also be a bit autistic about this.

>>553699
RailClone could probably do it.

>>554203
Didn't know about that, thanks

File: escalators.png (20KB, 731x498px) Image search: [Google]

20KB, 731x498px

>>554183
yeah makes sense. The only parts of the model which need to be parametric then are the escalator steps and rails. You can model up premade end pieces and just move / mirror them according to the width of your steps. Use internalise data to save the geometry into the grasshopper definition, that way you don't have to reference a rhino file.

This shitty drawing explains it. Red stuff is parametric ( steps, rails) black parts you can just make once and save into the defintion.

>>554303
Thanks for breaking it down

I will also need the bottom part, the 'chassis' if you can call it that. It will also need to be parametric so it can align with the bottom of a ceiling package. It's easy tho, pretty much just 3 boxes.
Great point with the end pieces not having to be parametric. Makes the whole definition a lot easier

File: ANON 3.png (204KB, 2656x715px)

204KB, 2656x715px

>>554334
Heres a definition which makes a set of stairs between two points. Adjust the line division until you get the riser height you want. The expressions on the list length components are ' i-1 '. Otherwise the list item wraps around to the first item instead of choosing the last one.

>>554441
Thanks, this will be very useful. I already have a stair script from grasshopper3d but it's all baked into a cluster and not as clean.

Unfortunately I will have to postpone the escalator script a bit, got some changes in the project that need to be dealt with first.