That is the same basic thing, just the math is different. That would be no more complicated to make in grasshopper than the original problem. Start with the basic input parameters in the diagram. The exact math can be built into your shape. After that, you can shatter the various curves if needed. I am not sure what your question is?
The issue is that Iâm new to Grasshopper, so I donât know how to create the .gh file, could you please help me
Sure, I will look at it later and post a script. It is something different. Even my Tekla CC creator found the problem unique, and he has been doing this for a long time. The new diagram will largely use point inputs and expressions that are your forumlaâs. Once you have the curves, it is no more different than revolving in Solidworks. You can do this in Solidworks as well using variables, but if you are planning to use this with Tekla, grasshopper is better.
If you see a benefit to grasshopper in your work I suggest taking a course. If you search some of my posts from a year ago you will find the struggles I was having. Thankfully, I found a number of really helpful people here that were kind to help. This program is a bit of a beast to learn, but it is worth it. I still fight with data maniupulation at times, but my scripts now are far more efficient.
Talk later.
1 Like
I created your new dome geometry.
Some notes:
- There are sliders to adjust the dimensions of the shapes.
- You can see the basic curves geometry used to find the dome geometry in the front view.
- The script creates the inner shape of the dome by finding the intersection points and creates half of the inner dome profile.
- The dome outer profile is a simple offset, and the dome is a revolve.
- Both the 2.0 domes and 1.9 domes are included. One is green, and the other is blue.
- The formula values are not to sufficient significant figures to find exact intersection points. There will be a mathematical solution, but I did not have the time. I assumed the SF values, and DH values would be correct and adjusted the KR formula until there was one intersection point between the curves.
DOME-V2.3dm (530.5 KB)
DOME-V2_FORUM.gh (28.3 KB)
Front View:
Dome
To preview domes click on the custom preview.
3 Likes
Thank you for your helpâit worked perfectly in Rhino!
However, Iâm not sure how to import the model into Tekla Structures.
Could you please assist me with this?
Thanks in advance!
There is no simple solution to import directly into Tekla from this point as mentioned earlier. If the goal is to build this I would do most of the plate work in Rhino, and import the flattened platework into Tekla to track all the parts. To do that you will need to split the various arcs into the dimensions needed before you do the revlove. If the goal at this point is a simple illustration in Tekla, then you can split the dome geometry into mutliple levels (10 - 20) and loft between the circles. The plate geometry from lofts will not be very useful, but you wil get a decent picture.
You may want to post on the Tekla forum to see if you find any that specialize in this type of work. There may be another solution. Given @sebastian.lindholm comments, I donât think there is a simple solution.
I only need it for geometric modeling, not for flattened platework or detailing.
The best method for illustration purposes is to import a baked rhino version into Tekla. I have done that many times with site data. If you want native tekla parts you will need to split up the arcs and create reference curves to use in Tekla as lofts. I am not sure how well the lofts will section at intersection points. I expect there will be gaps where the plates intersect and you will need to edit the drawings. The rhino data will be a watertight shape that should section perfectly. I do not have time today to illustrate this.
1 Like
HI Sebastian,
is there a way to build C.Cs for this using C# or there a limitation in tekla?
Hi, whatever you can model by hand in Tekla (or indeed by using Grasshopper) should be possible to recreate with the API.
So if itâs only for geometry, you could create a C# tekla plugin that creates the geometry as a ShapeItem and then inserts a Brep whose profile is that shape. The ShapeItem needs to be defined using vertices and edges though, so for that youâd need to calculate a mesh that represents the geometry (however you chose to define it in the first place). Probably not the easiest thing to do without Rhino or some other external library.
Some example code here:
Cheers,
Sebastian
Tekla is about how you actually make parts. Rhino and solidworks are what is referred to as sketch based modelling. Any sketch is mathematically defined. There are mathematical relationships when you use revolve, extrude or any other operation. This means every point is exactly defined in space. The math behind the scenes is not simple. Tekla uses shapes defined by a small number (small in comparision to sketch based modelling) of points to define most shapes. Scroll thru the profile catalog and you can see this when you look at any profile. These shapes are extruded along relatively simple paths. Sketch based is important for mechancial people since they use 5 axis CNC and other equipment to build small parts. That is far less common for structural work. We use 5 axis equipment to prep part ends, but the size it can handle is limited.
90% of what I see posted here is not constructable at any scale.
Have you discussed this with the fabricator? The equipment to roll 1200-1600mm wide piece of sheet steel to the radii you are talking about, with the precision needed for welding is not common and the cost to roll it would be notable. Maybe there is a different method you use. The people we work with make parts like this using sheet steel. I believe fabricators we work with would make this using flat sheets. The custom component expert we asked has made CCâs for companies building structures like this and he said they were using all flat steel.
1 Like
This is what it looks like in reality.
That is thick plate. I do not see a lot of high tech here. I think that plate is bent in a single plane and if you were to laser scan a finished roof it would not be exactly as your model. It will be reasonably close given the size of the parts that are used, but unless each plate is bent in two planes you will find gaps between the model and the finished shape. You seem to have access to the fabrication floor. Did you ask them the steps used before they set the plate into their form?
The image below highlights the problem with this method. The gap here is quite large and will require more welding than one wants. Without more high tech equipment I doubt that can ever be corrected. Parts would need to be machined to tolerances less than 0.5mm to avoid that.
If the plates are only bent in a single plane this is easy to model in grasshopper and Tekla. You can use grasshopper to drive your plate shapes. If you track the gaps with a few projects you can improve the weld gap issues by varying the parameters in your script.
If you have never looked at a english wheel, I suggest you google that. This is how they roll thin sheet steel manually for automotive applications. This does not work for this thickness, but the basic principle needed to shape this plate in two planes would be the same. The difference for this is your radii are huge.
The large format plate rolling makes me think of âThe Beanâ in Chicago. See this article for a general summary:
1 Like
taz, that would be so much more difficult to roll. All sorts of problems.