Compound Cell

Hi Pedro,

Yes, it is possible to generate new compound cells. Once you have created and identified the cell library to which you want to generate the compound cells, if all inputs are provided, for each update of the CompoundCell node a compound cell is generated into the library. Of course, only compound cells with unique names are creating new entries. The attached video shows how it is possible to create a set up for controlled creation of compound cells. The compound cell is added as the last transaction. Cell names are created by an expression, reflecting the changed inputs. Cells with the same inputs create the same name, overwriting any previous versions of that cell with the same name (in case there is some other change in what the compound cells look like).  

There are two ways how this set up can be utilized: do not play the last transaction, find the settings for the cell that you want to create, play the last transaction and that cell will be written to the library. Unplay the compound cell transaction (the new parameters will create a new transaction above the compound cell transaction and can form a record of the compound cell versions that have been generated in case updates are required later), find a new parameter set, play the compound cell transaction, etc. In order to see the generated compound cells toggle to another cell library and back to refresh the list.  

The generated parameter set transactions can be suppressed and their headers amended to provide additional information.  

The second way is to activate the last transaction and then change parameters. This will generate compound cells for any unique parameter sets that are activated. The same parameter set generates the same cell name, effectively overwriting the cell that was previously generated with that same name already. When unplaying/deactivating the compound cell transaction, only the last parameter set will be saved in a new transaction below the compound cell transaction. This means there won't be any record of all but the last version (aside the fact that the compound cells have been generated).

Attached are the video, the ABD-GC V8i SELECTseries 6 Update 1 (version 08.11.09.436) DGN and GCT files of what is shown in the video.  Note that it may be necessary to replace the compound cell node in order to make the correct connection to whichever compound cell library you have created or are using.

Regards,

     Volker

ABD-GC 436 CompoundCellFactory.dgn 

https://communities.bentley.com/cfs-file/__key/communityserver-discussions-components-files/360/ABD_2D00_GC-436-CompoundCellFactory.gct

Volker,

Nice one !! 

I can see the Compound Cell Node being a very useful productivity multiplier for ABD users.

Compound Cells or View dependent Cells are a key part of BIM authoring apps like ABD. On one hand, they provide the means for scaling and speeding up drawing views of the the model. On the other, as cells, they allow BIM attribute/business info to be attached to the object that the Cell represents.

Some questions:

1. Hopefully, the GC C-Cell node can be enhanced to include an input to allow the ABD user to define 2d plan, section, RCP etc graphics. ABD's Compound Cell tool already allows for this, so I am hoping that this would be fairly straight forward. Plan view symbols are especially important.

2. I think that there is also a masking element in the Compound Cells that will be used to hide unwanted graphics below the Cell. For example, a window that is set into a wall may have a 'cill' masking element to hide any cavities in the wall buildup. It would be good if GC could access that as well. I think that ABD MEP drawing symbols also do this.

3. A lot of ABD users are very dependent on C-Cells and avoid PCS whenever possible. It would be good to push GC to see how far it can automate the C-Cell generation and updating process.

• Sort Plan, Section, RCP symbol graphics by Element Template. The user would just select an ET for Plan graphics and the C-Cell node would automatically collate and assign the graphics accordingly.
• Sort Section, RCP symbol graphics by model name. The plan etc graphic may be complex/ large and would be best generated by using a hidden line or CVE generated in separate model (probably in the same file as the Cell.gct dgn). The C-Cell node would play thorough its transactions, generating the 3d model, export to the Cell dgn as shown, then trigger a remote CVE update (apparently Mstn has key-ins for this) and select and assign those graphic elements to the Cell library.
• BIM objects will need to be imbued with business data. See the long labourious process that PCS forces the user through. See page 57-71. It would be good to script up a separate node to generate the DataGroup definition for the object. This would include extracting any relevant parameters in GC.... sparing the user having to manually re-type them in the DGS xsd etc.
• Family and Parts? How would the user assign F+P or rendering materials in GC?

4. Updating: How would the user modify and update the C-Cells? As it is often the case these days, it is not easy. It would be great if GC could step in and help the unwashed masses. Please note that you would need to go through the DGS tools to update/maintain the business info / instance data attached to the cell.

Looking at the power already provided by GC, I would have thought that this would be a relatively quick C# project for anyone on the GC dev or ABD dev team. Maybe, a lot of tools can be prototyped using GC plus a little C#...and converted into ABD / MstnAPI C++ at a later point, once the workflow stuff has been sorted out.

Hi Dominic,
Great suggestions. I'll see how we can convert your suggestions to enhancements that we can queue up in our development queue. Some may be possible to implement just on the GC side, some may require enhancements to ABD's API which may have ripple effects in ABD, i.e. more work than one would assume necessary to achieve that capability, like wanting to add a floor to a building and finding that first the foundations require reinforcement.
Regards,
Volker

Hi Volker,

Thanks very much. I hope that this kind of 'cross border' collaboration will benefit both parties, and the Bentley family of apps as a whole.

ABD seems to be enjoying some attention in China, where at least one large 'design institute' is developing / cutomising ABD. So, hopefully a more comprehensive ABD SDK would not only help GC

Hi Volker,

Thanks very much. I hope that this kind of 'cross border' collaboration will benefit both parties, and the Bentley family of apps as a whole.

ABD seems to be enjoying some attention in China, where at least one large 'design institute' is developing / cutomising ABD. So, hopefully a more comprehensive ABD SDK would not only help GC