Section VPs: improve clarity of "structural objects", merging and "profile line"

[line-weight]

At the request of @Matt Panzer a wishlist item, spun off from a discussion thread.

 

Basically: I think confusion arises around what "structural objects" are, as defined for the purpose of section viewports and the various options in the "Advanced Section Properties" dialogue. But this wishlist item goes a little further I think. I am going to try and keep this concise to start with, partly to stop it being too confusing, partly because I'm running out of time to write this just now.

 

Here's that dialogue box:

 

 

I have attached a VWX file with this post. It'll probably be necessary to open it and look at it in order for this post to make any sense. Screenshots are taken from it. The array of possible section viewports looks like this:

 

 

Each of those section viewports is a section through this hypothetical wall/roof buildup:

 

 

That's a standard cavity wall here in the UK, brick outer skin (not loadbearing), insulation in cavity, block inner skin (loadbearing). I've modelled a kind of decorative brickwork corbel at the top of the brick. The roof is structural timber joists, plywood deck on top, insulation & membrane above. Plasterboard ceiling fixed to bottom of joists. Decorative coving at junction of wall and roof, internally. What I've drawn is just illustrative for the purposes of this wishlist post.

 

 

Most of my section drawings (in practice) are either

- something like section type B where I want to show some level of internal buildup detail, and where the scale is 1:50 or larger. In these examples I've shown material fills as solid colour, but in reality I would usually be using hatches instead.

- something like section type E where I don't want to show buildup detail. Anything smaller than 1:50 will usually be like this. Sometimes I might use it up to 1:50 for early stage drawings.

 

Points to make:

 

Note the difference between section types A and B:

 

On section type B the brick corbel detail is "merged" with the brick component of the wall. I'm more likely to want this, than the lines shown in section type A. I've achieved this by telling VW that those extrudes I used to model the corbel courses are "structural objects" because this allows it to merge them with the "structural object" that is the wall. But of course, those corbels aren't in reality "structural objects" at all.

 

What are/aren't defined as "structural objects" for this purpose doesn't make sense to me. Either VW has got some wacky ideas about what's structural (ie holding the building up), or that's just a bad name for that type of object, or there needs to be some additional method(s) of choosing what to merge in sections.

 

Below, section type F on the left shows what VW thinks is and isn't structural. Section type G on the right shows what I think is and isn't structural.

 

 

Then we come to what the "profile line" is. I actually haven't really used this in real life. I'd assumed it was the outer profile of the whole section...but it seems it's not. A complication is that it doesn't seem to work in VW2023 SP2. So... here's a screenshot from VW2021 to show where it is drawn (as the red line) in section type C :

 

 

But in section type D I indicate (by using stacked viewports) where I think a "profile line" should be and would be useful, and, I think, fits with a fairly standard drawing convention where the overall section cut line is drawn with a thick line (I've just shown it in red for emphasis) and then the internal detail (also in section) is drawn with a thinner line:

 

 

Ok, I'm going to leave it at that for now, no doubt multiple subsidiary questions will come up in discussion.

 

I do think the way that sections are drawn (and the amount of fine control given to the user) is really crucial in allowing us to create good, clear documentation from the model with as little touching-up in annotation space as possible.

 

 

 

 

VWsections.vwx

Edited December 19, 2022 by line-weight

[Amorphous - Julian]

5 hours ago, line-weight said:

 

 

 

 

 

Thank you @line-weight

In my office, we manually trace over viewports to achieve this thick section cut line you showed in the above image. It is time-consuming and drives me nuts!

And everytime the model changes, we have to modify this manual think cut line. I hope something more intuitive and relevant to real-world use case can would soon be available.  

 

 

VWsections.vwx

 

[Tom W.]

6 hours ago, line-weight said:

I've achieved this by telling VW that those extrudes I used to model the corbel courses are "structural objects" because this allows it to merge them with the "structural object" that is the wall. But of course, those corbels aren't in reality "structural objects" at all.

 

What are/aren't defined as "structural objects" for this purpose doesn't make sense to me. Either VW has got some wacky ideas about what's structural (ie holding the building up), or that's just a bad name for that type of object, or there needs to be some additional method(s) of choosing what to merge in sections.

 

'Structural Objects' are Walls, Slabs, Roofs, etc i.e. parametric objects used for representing building structures: 'part of the building' rather than the elements which are load-bearing.

 

'Nonstructural' objects are 3D solids.

 

By enabling 'Merge With Structural Objects in Sections' for a 3D solid you're just telling VW to regard the 3D-modelled corbal brickwork as part of the Wall which is precisely what you'd want it to do in this case. Otherwise you'd have to model the corbal as separate Wall objects or as Wall components with bottom offsets to get the same effect which would be far less convenient.

 

Having said this I do find that the merging behaviour in section VPs is not always consistent + often get objects which should merge but don't. Conversely, I sometimes want some objects not to merge but have no control over this so it does often feel like a bit of a blunt instrument + I agree very much with this:

 

6 hours ago, line-weight said:

I do think the way that sections are drawn (and the amount of fine control given to the user) is really crucial in allowing us to create good, clear documentation from the model with as little touching-up in annotation space as possible.

 

[line-weight]

3 hours ago, Tom W. said:

 

'Structural Objects' are Walls, Slabs, Roofs, etc i.e. parametric objects used for representing building structures: 'part of the building' rather than the elements which are load-bearing.

 

'Nonstructural' objects are 3D solids.

 

Yes...which is why it's confusing to call them "structural" objects.

 

But even if they were renamed, say as "parametric building parts", I still don't really see that it's a useful or relevant distinction to make, when drawing a section. When I draw the section, I don't want to tell the reader of the drawing about how objects are generated in my CAD programme. And I don't feel that it matches any particular drawing convention that I'm aware.

 

There are however things that are useful to have control over:

 

- Whether or not certain objects get merged with adjacent objects that have the same fill

- A distinction between an outer main profile line thickness, and the thickness of other lines within the section cut.

 

 

 

 

[Matt Panzer]

Thank you @line-weight, et al, for your thoughts on this!

 

I created VE-103844 for this and linked it to this thread so we can follow the discussion.

[setdesigner]

Thanks @line-weight. I confirm that profile line is not showing in 2023 SP2. Here is a 2022 file with profile line, when opening in 2023 and update the viewport, the profile line disappear. I understand now why I was so confused, the feature just does not work and did not tried in 2022. Is there a way to check if a bug has been filed ?

2022-12-21-HORIZONTAL-SECTION_v2022.vwx

[rDesign]

3 hours ago, setdesigner said:

Is there a way to check if a bug has been filed ?

According to this other thread I found, a bug has been filed.

 

[setdesigner]

Thank you @rDesign

[line-weight]

On 12/20/2022 at 2:47 PM, Matt Panzer said:

Thank you @line-weight, et al, for your thoughts on this!

 

I created VE-103844 for this and linked it to this thread so we can follow the discussion.

 

Having given it a bit of thought, here is how I think "merging" can most usefully happen. This is assuming sections where internal build-up of things like walls, roofs and floors are shown.

 

 

- There are materials (say, poured concrete, or plaster) that generally exist as a kind of contiguous mass. They are not made up of individual units.

- In most cases, with these materials, if I draw two objects made of them, and the objects are immediately next to each other without a gap, I am drawing them as two objects because that's what is convenient to draw, rather than because I want them to appear as separate objects in the drawings. In this case, I'll almost always want them to be "merged" in section (and elevation).

- Then there are materials (say, timber framing) that generally exist in the form of individual objects/members that are fixed together.

- In most cases, with these materials, if I draw two objects made of them, and the objects are immediately next to other without a gap, I am drawing them as two objects because they really are two objects. Therefore I don't want them to be "merged" in section (or elevation). For example; two timber beams immediately next to each other. I'd nearly always want them to be drawn unmerged because I'm drawing a doubled-up timber beam, not a single timber beam that's the dimensions of the two objects added together. And that distinction is important constructionally.

- There are certain materials which aren't exactly "contiguous mass" materials, but are generally drawn as such. For example, brickwork - it's actually made up of discrete units with mortar in between them, but it's not usual practical to model it that way, and in construction sections it's usually shown as contiguous rather than with each individual brick drawn. Sometimes individual brick units might be drawn in large scale sections, where it's important to show it, and in this case the bricks could be modelled individually, or such detail could be added manually in annotation space.

- Therefore, in principle, whether or not adjacent objects should "merge" should mainly be determined by what material they are made of.

 

- The existing approach of merging or not merging objects based on whether they have the same fill works up to a certain point for me, because I use "material classes" already, and every component in my model has a material class, and that's what determines the fill it has in a section. So when I want objects of "contiguous" type material to merge with each other in section, I can mostly get this to happen, by classing them the same, and telling VW they are "structural" for the purposes of sectioning if necessary.

- However, I'm not given the control I'd actually like, which is to define merging behaviour per "material class" rather than per object via the "merge with structural objects in sections" button. Logical to me is to tell VW what kind of material something's made of, then merging behaviour to happen as a result of this. I shouldn't have to tell it what to do with each object individually.

- I don't (yet) use the newly added "materials" concept in VW. That's partly because it doesn't seem yet to be applied consistently to all objects, partly because I haven't invested the time in understanding how to start using it, and what advantages it offers over my current strategy of defining material via class. If some kind of sectioning behaviour can be implemented based around the materials concept, I'm open to adjusting my workflows in order to take advantage of it.

- The existing approach of certain objects being called "structural" doesn't make much sense to me, because it seems to mean that non-structural components of things like walls get included, and I don't see in what way it's useful to anyone. Of course, it may be useful to some people and maybe there's a reason to retain it, but I think it should be better named.

- In all of this I'd want to stress that it's crucially important that these kinds of sectioning behaviours are solidly reliable, predictable and consistent. It's not just for aesthetics or clarity - the presence or absence of a line between components will often convey important information from a constructional point of view. I don't want to be on a site visit and find something built incorrectly because VW has drawn a section differently from how I told it to. We've already identified that the "profile line" has been messed up in the transition between 2022 and 2023 and there are also ongoing issues with lines appearing in elevations where they shouldn't. This is fundamental stuff and it should be one of the first things to check and double check before releasing any new version of the software into general use.

 

 

I would be interested to hear where others do/don't agree with me on any of this.

 

[Tom W.]

Most of the time for me I am sectioning Walls/Slabs/Roofs + so what would work for me would be a 'Merge in Section' column in the component settings where I could place a check next to the components I want to display merged + those I don't I can leave unchecked. Like you, certain components I want to merge + others I don't + at the moment there's no fine control over this: it's either all or nothing.

[line-weight]

5 minutes ago, Tom W. said:

Most of the time for me I am sectioning Walls/Slabs/Roofs + so what would work for me would be a 'Merge in Section' column in the component settings where I could place a check next to the components I want to display merged + those I don't I can leave unchecked. Like you, certain components I want to merge + others I don't + at the moment there's no fine control over this: it's either all or nothing.

My suggested approach I think could do what you want without you needing to set things up for each component of each wall style - it would be decided by what material each component was.

[Tom W.]

I'd personally prefer to do it on a style-by-style + component-by-component basis as I want to see different things depending on the circumstances. For example I might have two insulation components side by side (insulation between timber studs + an insulation-filled cavity) + not in this case want to have them merged to maintain the distinction between the two components. Whereas in other circumstances - where wall insulation runs into roof insulation for example - I would want them merged. Also concrete: a concrete wall sitting on a concrete foundation I prefer to see a line between the two objects whereas most other times I'd rather see adjoining concrete objects merged. 

 

Not even sure if my component check-box would even let me achieve these examples. Perhaps a Component Join Tool which worked in section VPs is the answer... 🙂

 

The beauty of Materials is that you can use them across the board in a variety of different contexts + so to have the merge-or-not-to-merge control set per Material would be too blunt an instrument for me. 

 

[rDesign]

I would not want to have to manually set which components in a SLVP will merge or not merge. This may be a very big request, but I’d like Vw to be smart enough to know when components should be merged.
 

I think my preferred approach is mostly addressed in the long description from @line-weight above advocating Materials as the deciding factor.

[line-weight]

41 minutes ago, Tom W. said:

I'd personally prefer to do it on a style-by-style + component-by-component basis as I want to see different things depending on the circumstances. For example I might have two insulation components side by side (insulation between timber studs + an insulation-filled cavity) + not in this case want to have them merged to maintain the distinction between the two components. Whereas in other circumstances - where wall insulation runs into roof insulation for example - I would want them merged. Also concrete: a concrete wall sitting on a concrete foundation I prefer to see a line between the two objects whereas most other times I'd rather see adjoining concrete objects merged. 

 

Not even sure if my component check-box would even let me achieve these examples. Perhaps a Component Join Tool which worked in section VPs is the answer... 🙂

 

The beauty of Materials is that you can use them across the board in a variety of different contexts + so to have the merge-or-not-to-merge control set per Material would be too blunt an instrument for me. 

 

 

In these scenarios, you maybe could use duplicate "materials", for example "between-stud insulation" / "cavity insulation", and "wall concrete" / "footing concrete".

 

I already do something like this under my current approach.

 

Of course this becomes one of those things where you trade off the admin of having multiple materials vs multiple styles. And the balance will be different for different people according to use case.

[Tom W.]

Just now, line-weight said:

 

In these scenarios, you maybe could use duplicate "materials", for example "between-stud insulation" / "cavity insulation", and "wall concrete" / "footing concrete".

 

That would be an option but not a route I'd personally want to go down because of the way I use + report on Materials where I want to count a single material or product regardless of how or where it's used in the model.

 

Presumably if I wanted I could actually do this right now by having two Hatches/Tiles that were identical graphically but had different names as then VW would treat them as two different fills + wouldn't merge them?

 

At the end of the day, thinking about it further, it's very easy to add a line in annotations to split two merged components so perhaps the current arrangement is actually quite pragmatic. It's a lot easier to unmerge two components by adding a line than it is to merge ones which are displaying separately when you don't want them to be.

[zoomer]

I currently try to model an obviously simple looking, small, existing residence house,

planned in the late 60ies. Although it was done for legacy building standards, it

has some pretty architectural quality, functionally and visually and there are some

nice physical building details you would not expect.

(Like a thicker outer Brick Wall for the bath room !? Because of humidity and

cutting in Water Installation)

Anomalies which make it hard to understand and model, from measurements only

possible from room too room. But when you got it it all makes so much sense.

 

I want to model it at a high LOD in 3D that I can understand and export all kind of

Components Areas or Volumes, Door/Window raw sizes and such things to

estimate costs and realistic options what you can reasonably improve, or not.

 

 

I do it in Bricscad as it seem more appropriated, but may test in VW too afterwards.

As I have pure Solids Direct Modeling and can easily switch between overall Solids

and by Ply Direct Modeling.

(Like VW Wall Connection by Components - on steroids and in all 3D)

I even have flexibel Component dimensions inside a Wall/Slab Style, which saves me

from some redundant Style Duplicates. But ....

 

Again, my general BIM conclusion is - Wall/Slab Styles (overall) are useless anyway.

 

What I am interested in is how it will be build.

First come the Structural Elements like Slabs and Walls, then Windows (and Doors)

and then comes Finishing/Covering like Insulation (Compounds) or Plaster,

Finish Floors, .....

 

Even for such a small and pretty modernistic and reduced Building ....

I personally suffer when I have to interrupt a running Wall just for inner outer outer

Finishes. Even ignoring that the Stair Case has a special interior Acoustic Plaster

vs standard mineral Plasters. I will still get too many Wall styles.

Or Floor Styles. The same Travertine Floor Finish, directly on structural concrete in,

Stair Case, as panels under Doors and usually on floating screed ....

 

 

The only problem hindering BIM modelers from working that preferred way is that

BIM/CAD usually doing hard with Insertion Elements - cutting through multiple objects.

I learned, a bit too late, but Bricscad even offers inserted Elements in Walls and such

(Doors/Windows/..) to even integrate more elements to cut.

It is quite hidden but already possible.

Unfortunately did not yet switch to that opportunity for now, as I am still fighting with

understanding and determining the physical dimensions of the Building Components

and Parts.

 

So I am still propagating, BIM Modeling has to separate Structural vs Finishing Elements.

Interior Finishes may be controlled by the Room Definitions from a table of Wall/Ceiling

and a Floor Finish(es), Exterior Finishing or Insulation Systems mostly surrounding larger

areas may need some extra rules for vertical and horizontal appearances.

Edited December 28, 2022 by zoomer

[line-weight]

17 minutes ago, zoomer said:

I currently try to model an obviously simple looking, small, existing residence house,

planned in the late 60ies. Although it was done for legacy building standards, it

has some pretty architectural quality, functionally and visually and there are some

nice physical building details you would not expect.

(Like a thicker outer Brick Wall for the bath room !? Because of humidity and

cutting in Water Installation)

Anomalies which make it hard to understand and model, from measurements only

possible from room too room. But when you got it it all makes so much sense.

 

I want to model it at a high LOD in 3D that I can understand and export all kind of

Components Areas or Volumes, Door/Window raw sizes and such things to

estimate costs and realistic options what you can reasonably improve, or not.

 

All this sounds very similar to something I've been doing this year. And run into many of the same questions. I don't know Bricscad so I'm stuck with Vectorworks.

 

 

 

Here's an example of the quite complex construction, the sectional view is only for my own purposes (so doesn't have to look pretty or even be very understandable by other people) but there are lots of things I'd have to fix in order to make this useful as documentation. For example it's a flat roof which started out as 3-ply roofing felt but then had a fibre-glass roof installed over the top of that. So there are large areas of flat roof, with multi-layer buildup, but with various drip or upstand details along the perimeters. I can model the main part of this roof with a multi component slab or roof face object... but then the perimeter details have to be modelled directly with extrudes. And this is exactly the kind of location where I run into the problem of merging those perimeter details with the top layer of the roof buildup, because fibre-glass is a contiguous material, and I don't want a line at the transition.

 

At the moment each time I reacj this kind of detail, and I want to make it into a "proper" sectin drawing I have various options...

- model most of the roof as an eg. slab object (with convenience of multilayer buildup, etc) then make the perimeter as extrudes, and then fiddle about with settings and viewport options and maybe annotations to force the parts to merge where I want them to

- give up on using a slab object, model the whole roof buildup directly as solids & use solid additions to "merge" the relevant bits together

- maybe use a slab object for the bottom-most layers, and then direct modelling for the more complex outer layers.

 

The best option always depends on the project and what I want to produce at the end and how much I expect to have to edit things once I've modelled them....but sometimes things change and the better option turns out not to be the better one, but I'm too far down the line to change it.

 

More direct control over the merging of objects would save me having to make those painful decisions early in the process (and then discover later that I made the wrong gamble).

 

 

 

P.S. producing the clip-cube image above reminds me ... please can we have the clip cube let us see full section detail on its cut plane, ideally just like a non merged section viewport would???

[line-weight]

38 minutes ago, zoomer said:

 

Again, my general BIM conclusion is - Wall/Slab Styles (overall) are useless anyway.

 

What I am interested in is how it will be build.

First come the Structural Elements like Slabs and Walls, then Windows (and Doors)

and then comes Finishing/Covering like Insulation (Compounds) or Plaster,

Finish Floors, .....

 

Even for such a small and pretty modernistic and reduced Building ....

I personally suffer when I have to interrupt a running Wall just for inner outer outer

Finishes. Even ignoring that the Stair Case has a special interior Acoustic Plaster

vs standard mineral Plasters. I will still get too many Wall styles.

Or Floor Styles. The same Travertine Floor Finish, directly on structural concrete in,

Stair Case, as panels under Doors and usually on floating screed ....

 

I basically agree with this too.

[zoomer]

I also agree.

 

I have basically two Slab levels.

One is cast in place for the "residential/private area, while another is a

more thick and lowered concrete Beams Slab filled with gas concrete

modules for larger distances. The Walls sit on top of each specific Slab,

but the Plaster my go down to the other Slab level.

In Door opening Direction, each Floor Package will climb down 0.5

to 1.5 cm, to also give a stop for the Doors in all 3 Dimensions ...

 

You are not able to model such details with just a few Slab and

Wall Stylesin VW.

 

 

Yes, it depends on what LOD is desired.

In VW you will have the Story Level heights for fast height adaption and

moving Walls will auto heal the by Component Connections.

And we have Wall/Window connections and Plasters running in.

But in reality I still see so many extra situations asking for new Styles.

[line-weight]

Exactly as @zoomer says, the way the BIM model is built up should reflect the way the building is built up.

 

The more closely they are connected, the easier it is for the logic of connections to automatically follow.

 

A building might have 2 or 3 types of wall core/structure. Maybe it will also only have 2 or 3 types of external wall finish, and it might have 5 or 10 types of internal wall finish. But these all overlap with each other, so the number of possible combinations of external finish + core + internal finish becomes larger than any of those numbers, and most of us I'm sure end up with 10s of wall styles with elaborate names, for each project, by the time you get to a certain level of detail (or you just give up on wall styles altogether).

 

This is going a little bit off topic but the same reasoning is relevant to the question of whether things should be merged in section or elevation.

 

For example a common frustration is joint lines appearing on elevations where they shouldn't. Firstly VW should be able to understand that if the materials are the same, and the elements are coplanar, there's no joint line. And it is kind of programmed like that already. But also, all these potential joint lines in the external finish of a model, most of them are only there because of the logic behind building VW walls which says that junctions in the external finish generally correspond to junctions in the internal finish. Walls are concieved as things that span from storey to storey.

 

But constructionally, walls aren't usually like that. The core may span storey to storey - or it might be continuous. It depends how the floor structure is attached, and the sequence of construction. Internal finishes do generally span storey to storey (with offsets for the floor depths). And external finishes very often don't care much about internal floor levels at all. They might span from some point relative to external ground level, to some point relative to the edge of the roof.

 

If the logic of how walls were modelled wasn't entirely based on the starting point of a wall being something from one storey level to the next, then perhaps we could have more efficient methods of applying external wall finishes. And ones that didn't introduce so many opportunities for portions to be 1mm off co-planar. Because by definition the whole thing would be coplanar.

 

This is often where I end up when I finish up with a model where a wall object is used for the core but directly modelled solids are used for internal & external finishes or cladding. It's a pain to have to manually cut wall holes ... but I can draw it as it actually is, and often that's with multistorey portions of walls drawn as one object or group. If I want to look at the building storey-by-storey (isolating each by layer), then of course these large wall cladding objects can get in the way. But... if I am looking at the building storey-by-storey, I'm looking at internal stuff so they can simply be turned off. And for that reason they get classed as external wall finish objects, or maybe they even get their own layer.

 

In any case the point is that in my experience, whenever I start deviating from the way VW expects me to build a BIM model, and I start using more of my own custom classes or layers to control visibility or material or whatever, then I always seem to end up at the same place: things are separated off from each other based essentially on construction sequence. Standard VW objects like walls and floors don't currently allow this. And the "merge with structural objects" tick-box is perhaps illustrative of how concepts have got muddled - or how there's a failure to recognise what "structure" means, in the context of a BIM model and the logic that underpins it.

Edited December 28, 2022 by line-weight

[Tom W.]

1 hour ago, line-weight said:

At the moment each time I reacj this kind of detail, and I want to make it into a "proper" sectin drawing I have various options...

- model most of the roof as an eg. slab object (with convenience of multilayer buildup, etc) then make the perimeter as extrudes, and then fiddle about with settings and viewport options and maybe annotations to force the parts to merge where I want them to

- give up on using a slab object, model the whole roof buildup directly as solids & use solid additions to "merge" the relevant bits together

- maybe use a slab object for the bottom-most layers, and then direct modelling for the more complex outer layers.

 

What about using the 'Add 3D Object to Slab' command? Draw Extrudes etc where you want them + add them to specific Slab components one by one. Looks great in section. That being said, making changes down the line - even minor ones - can be a massive pain.

[line-weight]

1 minute ago, Tom W. said:

 

What about using the 'Add 3D Object to Slab' command? Draw Extrudes etc where you want them + add them to specific Slab components one by one. Looks great in section. That being said, making changes down the line - even minor ones - can be a massive pain.

I should probably experiment a bit more with that but yes the main concern is about how easy it is to make changes further down the line - compared to, say, a solid addition that I maintain such that I can edit and move the various elements within it.

[zoomer]

11 hours ago, line-weight said:

And external finishes very often don't care much about internal floor levels at all. They might span from some point relative to external ground level, to some point relative to the edge of the roof.

 

If it is a larger or industrial multi story building,

I really model like that.

Facade is on an extra Layer and over the whole building height.

 

You can switch that Layer off to keep your Layer Visibility comfort

to switch between Stories, without the need of Clip Cube.

 

But a bit tedious and error prone if you need to adapt your Facade

Layer to potential Story height or Opening changes.

[line-weight]

Coming back to this.

 

I think most of the comments I made in this thread still stand, but I have now got a little bit more clarity on how this "structural object merging" works currently.

 

I realise that used carefully, it does (I think) solve a recurring problem I have, which is that there are always bits of walls, roofs and slabs that I just can't model using the parametric tools (or at least, it becomes incredibly convoluted to do so). In these cases I have to resort to modelling directly from solids. But I would always end up with unwanted lines in section and elevation, where these directly modelled portions joined with the wall/roof/slab objects.

 

Having realised that I can tell those solids (in their individual settings) to "merge with structural objects", most of these problems disappear. I can get rid of those unwanted join lines in sections - and - a little to my surprise, in elevational views too. This means that I can make my custom wall parts behave nearly as if they were actual wall objects, for presentation purposes.

 

Unfortunately though, this behaviour is not carried over to shaded view as far as I can work out. And because I do sometimes use shaded view for presentation drawings, the join lines become an issue. So I'd add another wish:

 

Allow "structural objects" to merge (ie lose lines at junctions between them) in shaded view as well.

 

This would free up my workflow very significantly because it would remove a whole load of stuff I have to strategise about when deciding how to draw something.

[line-weight]

By the way, I don't think "structural objects" are actually defined anywhere in VW help, or am I missing something?

 

This means that I still don't know exactly which objects are and aren't "structural objects". I don't think it's only walls, roofs and slabs. But have no way of finding out what the others are.

 

 

[Tom W.]

2 minutes ago, line-weight said:

By the way, I don't think "structural objects" are actually defined anywhere in VW help, or am I missing something?

 

This means that I still don't know exactly which objects are and aren't "structural objects". I don't think it's only walls, roofs and slabs. But have no way of finding out what the others are.

 

 

 

'Structural Objects' are Wall, Slab + Roof objects. 'Nonstructural Objects' are 3D Solids. It's definitely specified somewhere as that's how I know, but how deep you have to dig to find it I don't know...