Until now, we managed the revision index in the "Revision" property of the 3D file (part or assembly). This clue was then taken up on the drawing via the use of a note containing $PRPSHEET "Revision" which fetches the property of the 3D file.
At the same time, a "Revision" property exists in the drawing and is linked to the SW revision table of the BOM. At each revision, we first modify the 3D file and then we increment the SW revision table of the drawing to put the description of the modification, date. Incrementing the table updates the "Revision" property of the drawing.
The main advantage of this system is to have the revision index on the 3D and therefore without having to open the drawing.
The move to EPDM means that version and revision management will be directly accessible via the EPDM interface.
So my question is: is there still any point in managing the revision via the 3D file (it forces you to edit a revision on the 3D each time you revise the 2D)?
Should we change our methodology and only manage the " Revision " property of the 2D file?
This would allow for example to mount a revision on the 2D without having to touch the 3D in the case of an addition/modification of note.
How do you manage revisions with EPDM for your drawings ? For your 3D files?
In my old company, I had set up with @flegendre the following system for the management of 2D/3D indices with EPDM.
The revision information (index, date, projector, commentary...) was stored in the 3D. This information was entered in the 3D data map. The 2D retrieved this same data by mapping via EPDM (copy of properties) and its data map had the same fields as the 3D, but none of them were editable, so that the elements only came from the 3D.
So when a state change via EPDM to mount an index, the system incremented the 3D index, and left the field free for the description of the modification.
So, if the modification only concerns a modification of the 2D, we could modify the description by the data card of the 3D, without having to open the 3D with SW.
Other points, as the parts could change regularly, we kept a history of the modifications on 4 lines:
Has---
B A - -
C B A -
D C B A
E D C A
F E D A
...
Hoping to have been clear and helpful :)
Edit: for us, the 3D and the 2D necessarily had the same index, even if the modification only concerned one of the 2.
At the moment we manage indices with EPDM, but the revision table is stored in the properties of the plan. Reading Benoit.LF's solution I realize that our solution is really not the right one. Indeed, as it is possible to modify the part without modifying the plan or vice versa, I end up with plans that do not have the same index as the room, for me it is not logical.
It is good to have a "Quickchange" to make a change without taking a hint to be reserved for minor changes (spelling, updating...)
The description of the revision retrieves the comment (mandatory on each transition) that must be put during the validate->modification passage, you can still modify it after the fact.
I just regret that we use the drawing review table more because we could link the revision bubbles.
(it's possible to link this table with EPDM properties but it's not great)
@yoan: the SolidWorks revision table is rotten even without the EPDM, it is better to manage the index table with custom properties in the part and not in the drawing: it is the 3D part that is "master" and the drawing is "slave".
In the method described by @Benoit.LF, there is the possibility of making minor modifications without changing the index of the part. In this case, in the WF, we had the choice between a modification with a hint or a minor modification.
Agree with you flegendre you have to be able to make minor changes. I worked for several years without it and it was very annoying: the minor corrections were not made out of "fear" of taking a clue with the work that it implies for the rest of the company, updating the range, manufacturing program...
What's good about the Solidworks revision table (and this is the only positive point I see) is the link with the revision bubbles: you erase the line, the bubble(s) are deleted, practical when you start from an existing plan with many revision bubbles.
I'll have to change our workflow to handle the revisions in the room.
Thanks to Yoann and Flegendre for their participation in the debate ;-)
We have planned for modification without revision (we manage parts with a single plan for several length configurations, we didn't want to have to mount revisions each time a length configuration is created).
I think we will also provide for the revision decrement (our plans are approved by clients who are sometimes difficult on the indices). With something like this: http://blog.capinc.com/2013/06/how-to-roll-back-a-revision-in-enterprise-pdm/
In your system, all revision information (date, names, comments...) is therefore in properties of the 3D file.
On the other hand, how do you guarantee the sunchronization of EPDM revisions between 2D and 3D? Since 2D and 3D are different files, it is possible to edit a revision on the 2D without editing one on the 3D and conversely it is possible to mount a revision on the 3D without editing one on the 2D, right? Have you set up throttling in the WF preventing the rise of a revision independently between the 2 files? (and in this case a little explanation on the throttling methodology would be welcome) Is it up to the designers to be careful and to tick the right boxes when going through the revision? (and now how do you make up for the inevitable mistakes?)
Yes, everything is available from 3D. The 2D only receives the copy of the information from the 3Ds.
On EPDM, when you request a change of state on a SW file, a window appears with the "complementary" file: if I change the state of a 3D, the 2D comes with it and is automatically checked to follow the same evolution (question of configuration), and vice versa 2D/3D.
So if we were to desynchronize the EPDM indices, it would be due to a voluntary action of the users, in particular by having unchecked one of the two files in the dialog box.
Because of this feature, no throttling is necessary at the WF.
If this were to happen, for example the 2D to a higher index, you would have to pass a index to the 3D alone, by unchecking the 2D in the state change dialog.
You can also create loops in the WF such as "Index -1" and "Index +1", reserved for administrators. But you can always add them later if the need arises.
If I understood the elements of @Benoit.LF and @flegendre correctly: 2D and 3D files go through the same workflow transitions simultaneously in order to keep the same index, right?
Is change history management workflow-driven?
During our development, 2Ds are created after several clues about the 3D. Is it still possible to synchronize the 2D and 3D indices?
Yes, we transition 2D and 3D at the same time: so their revision indices go up at the same temps.et each file has its own revision index.
It should be noted that having a clue on the 2d and on the 3d makes it easier to detect problems with the automatic generation of pdf where one of the 2 files is not in the right version: just compare the revision table (which retrieves the index of the drawing), and a property of the cartouche that fetches the revision index of the 3D. When the two are not in sync, it means that the pdf was generated with one of the 2 files in an earlier version. (unfortunately too frequent sosu epdm 2015. (I hope this point will be improved with the 2020 version of the script since we have just migrated).
NB: we have also created a bypass in the workflow to update the 2d or 3d without editing any indexes. This allows you to make very minor changes to the files without having to rewrite (adding dimensions for information...)
Is it a variable/PP and workflow-driven review table or PDM/SW review table function ?
Is it possible to synchronize (without going through manual +1 loops), the index of the 2D with the 3D if the 2D is created after several indices of the 3D (for example the 3D is at index 3 and the 2D is created at this time?)?
I take up the speech again because I can't use the SW revision table in PDM: I can increment a row, but no more.
Until now, we managed the revision tables on 4 lines, with 4 rotating variables. We would like to change the way we don't lose the clues and be able to add more space (I'm exaggerating..).
I created the variables, updated the workflows for increments, the 3D map for inserting texts, modified the "Revision Table" settings in SOLIDWORKS so that it points to the right variables and modified the variables recalled in the table that we insert in the MEP.
I manage to insert the first line (Emission of the shot, with a "hyphen"), then on the board side it's silence. On the cartridge no problem because I recall the variable of the index.
I have read the various posts on the subject and it seems to me that using SW's "Revision Table" is a bad idea and a worse experience and therefore variables should be used. So my question is: how do you use variables if you don't know the final number of revisions? Can PDM create a matrix of the data that is inserted for revision indices? If so, how do you retrieve the data and transcribe it into a table?
For me, I don't necessarily need to do anything if it's just to access the information I'm looking for in a very punctual way.
Thanks to the EPDM history you can find the values previously entered (you just have to open an edrawing of the version considered in the history of the file to have access to the different properties of this file at this version)
.
If you want to have the values on the plan, it's a different story: you have to add rows (and the associated properties) to be able to fill in the revision table.
Don't forget to also modify the workflow to offset the revision properties (when you edit a revision, we say that rev1 becomes rev2 and so on all the variables filling the revision table: it's very painful to do but mandatory for it to work).
That's a few lines to add (below is my transition with 'only' 6 levels on the revision table....):
